The easiest way is to make a run. Then obtain the mass spectrum (scanning or SIM) at the beginning of the peak. Note the scan number. Do the same at the end of the peak (i.e. determine scan number). For best results want at least 10 to 15 scans across a peak. If the number is less, decrease the dwell time. If the number is more increase the dwell time. It will take a few runs

By scan number do you mean the peak position and retention time???

I am using an Agilent GCMS.... I am in agreement that my sample profile should be known prior to adjusting for max sensitivity.... I ran a standard with all my compounds, after being satisfied with the chromatographic separation, I identified the target ion and qualifier ions... I included these in the method... dwell time is on default settings at 100ms.... the start time, I figure should be related to when I want the MS to pick up signals for the ions at a particular time... Is that not based on when the retention time, as in, when the peak starts and when it ends? In that case, how many minutes leeway should there be, and in instances where co-elution takes place how does one determine the start or end time or does one create the method in such a way that one groups the time? (I did that)... I also retention locked my method... where do i see the effect of dwell time????

Thanks

In general you want to use the lowest dwell time possible that still provides you with adequate sensitivity. I'd start by setting every ion to 10 ms. Try to group your ions in groups of 10 or less, so that you get a higher number of "scans" across your peak. You can use increased dwell time to adjust your sensitivity if need be. Remember that the stated dwell time is NOT the actual dwell time; you have dead time in the detector when you jump from ion to ion. Also make sure that you start your ion window at least 6-10 seconds prior to the start of your peak elution, as there is processing dead time when you switch ion groups.

SIM is not at all like scan, despite the fact that on Agilent systems "scan" mode is really a rapid SIM mode. You will get what appears to be increased tailing with your peaks (it's a function of the detection mode); you may also lose some apparent separation. Be prepared for ion overlap, especially if you have isomers.

We see anywhere between a 20 to 100 fold increase in sensitivity using SIM. The more fragmentation there is and the more high mass ions that are present the higher the sensitivity increase. Also, depending on your compounds you can selectively tune the MS to give you better ion response in your specific range, rather than using a general wide-range tune like autotune or BFBtune.

Thank you!!! :)

Windows and dwells depend on the number of ions you want to look for. If you only have a few ions to look for then you can have just one time window to look at all of them. But realize that if you have a 'large number' of ions in a window you may lose sensitivity. Dwell times are also a function of how many ions you have. In most systems you will have a lag or lock on time as well as the dwell time - take these two times and add them for each ion you want to look at - this will give you the total time for a scan. The total scan time will then affect how many scans you get across an analyte peak in the GC run - ideally you will get at least 5 scans across a peak - that is at least 5 points for each peak - to give good integration and quantitation. Within this you can have shorter dwell times for more abundant ions and longer dwell times to give better sensitivity for specific ions.

Hi Paul, I am used to operate on 7 points per peak. but that might be an issue of convention.

On the dwell times I agree widely with Mark, I very rarly use 100 ms dwelltime, however I experience 10 is often on the low side, one basically needs to ballance between how many compounds do I need to do and how munch sensitivity do I need.

For one series it is usually fine to have 2-3 peakwidth front and back of the peak maximum. (i.e. about 10 seconds). However if you wnat to run the method soemwhat longer (i.e. every once in a while 100 samples) you will be happy if the method does not need massive adjustment each time you change a precolumn or an injecore liner. So if there is the need (number of series) and the opportunity (depending on the number of peaks to be analysed) I try to get some 30 s or even one minute between groups of peaks that shall be analysed together in one function/window orwhatever the respective suftware calls it.