SIM/MRM - The rule of diminishing returns

[Travisdog2]

Hi hi,

I have recently been having a bit of play with some MRM methods, and after doing a comparison between 2 of my methods, noticed that one produced a greater response than the other despite running the same sample.

This wasn't particularly surprising (they were after all slightly different methods), but what did surprise me was that the method which was set to detect a greater number of daughter ions seemed to provide greater response than an otherwise identical method which was designed to detect a more limited set of daughter ions.

It was always drummed into me that the fewer ions selected in a SIM/MRM type method, the greater the proportion of the cycle time available for their detection and thus the greater the response; so the results appeared counterintuitive. I now suspect that the reason for the result is that the methods were set to display response as a function of both quant and qual ions and hence the method detecting more ions was simply displaying a response encompassing signal from more ions and hence was bigger.

However this got me thinking about whether or not there is an optimum dwell time for MRM and SIM beyond which the law of diminishing return rears its head and where this is? The SIM work I do requires only a quant and a single qual ion (2 ions + RT) which means that the dwell times for each ion are in the region of about 120ms. I was curious as to whether these dwell times were excessive and whether I would lose anything by dropping the dwell times to say 80 ms and adding another ion which would increase response? I will obviously try this out, but would be curious to get an idea of what to expect beforehand if anyone would be willing to let me know their thoughts on this?

Kind Regards

TD2

[per_oxid]

I would recommend the article by agilent
"Instrument detection limit at ultrashort dwell times"
By Charles Nichols

In essense there is clearly a point of diminishing returns. At what point there is a rapid decline depends on the instrument. At low concentrations low dwell times will give high rsd%. For modern QqQ 80 ms is a very high dwell time i try to stay above 5ms lower than thid and ild is quickly increased. Point is that you wont notice much difference with dwell time intill you have low signal intensity. But as shown in the article longer dwell time result in lower instrument detection limit but it is not a linear relationship.

[Travisdog2]

Many thanks for the reply P_O,

I've had a look through the contents of the linked document and it's certainly eye opening - I've never considered dwell times that short before but I guess IDLs don't lie!

Another interesting set of content which I found today comes from this forum back in 2016 https://www.chromforum.org/viewtopic.php?t=27935 in which about half way through various posters (James_Ball included) discuss how dwell time works with Agilent GC-MS/MS systems and the reasons behind the limited improvement in response when dwell time is increased...

All this stuff happening under the hood is fascinating!

Kind Regards

TD2

[James_Ball]

I think I was looking into that back then because I had the same observations you are having now. I once thought long dwell time just keeps adding to the counts but it actually just gives more samples to be averaged together. The more samples will give a smoother peak to some extent, since you begin averaging out the noise. Too long dwell time and your peak doesn't have enough "scans" to form a smooth peak. I use "scans" because that is what we would look at if doing full scan and wanting short enough cycles to have enough scans across the peak.

When doing full scan, the instrument is usually looking at acquiring a count at each 0.1amu increment, so if you are scanning from 35-350amu then you are taking 3150 measurements in each scan, and if scan time is 2 scans per second then you are taking 6300 measurements per second which would be a dwell time at each 0.1 amu increment of 0.000159 seconds (0.159msec). If you look at scan and just SIM with a large number of consecutive masses, then the dwell is very very low.

[Travisdog2]

Hi James,

Thanks for the confirmation. Glad to see that I'm not going completely mad!

Kind Regards

TD2