Curve flattening 8270E

[MrPottsWith2Ts]

Hello,

I'm running 8270E analysis on a 6890A GC and 5973 MS. Splitless injection. I just installed a brand new column and did source maintenance, like I have done a million times before, and I'm getting some really odd lines of fit for my curves on analytes that I don't normally have trouble with. I had a couple of ideas that I wanted to run by anyone here:
-My response values have been in the process of skyrocketing over the last year. A year ago when I changed a column, the response of the last analyte on my list had a response value of just over 3,500,000 on the high point of my curve. On the curve I ran yesterday, this value was 7,500,000. The earlier eluting analytes of my curve are raising, too, but not nearly as high. I think it's worth noting that that later eluting compound curves look better (still not great though), despite me using that as an example. The naphthalene photo I provided is just the worst example of my normally good looking PAHs.
-My repeller value is kind of high, at 34.81 on the DFTPP tune that I did, so I am going to switch that out.
-It's also been a second that I've changed the split-vent trap, so I'm changing that, too.
-I read up on posts on various website about curve flattening and they mentioned that I could be oversaturating my source/column. I find that hard to believe since I have been making the same curves and running on this same instrument for almost 5 years now. Don't know why/how I could suddenly be doing anything different, but I could obviously be wrong.
-I admittedly have an air leak somewhere. I've been trying to nail it down for a while now, but everytime I think I get it, I don't. This could obviously be the root of all issues, but in the time I've had this leak, I've not had my curves be this bad.


Thanks

[cjm]

I would first try and fix the air leak, using canned air searching for the right ions is usually a quick way to find exactly where it is; likely somewhere on the sideboard, but could be at the other usualy leak points.

Splitless injection for 8270 may lead to quadratic-like calibration curves. Our lab runs 8270E using a pulsed 5:1 split, injection volume 1.0 uL, cal range 2.0-120 ug/mL. Even with the split, some compounds at the 120 level are clearly starting to show column overload. This also assumes we keep DFTPP tuned to about 400k counts for m/z 69 prior to calibration.

What is your injection volume?
What is your calibration range?
How does the general chromatography look like, i.e. does benzo(b)(k) resolution check look decent, does DDT break down at all? is pcp or benzidine tailing abnormally?

[MrPottsWith2Ts]

I would first try and fix the air leak, using canned air searching for the right ions is usually a quick way to find exactly where it is; likely somewhere on the sideboard, but could be at the other usualy leak points.

Splitless injection for 8270 may lead to quadratic-like calibration curves. Our lab runs 8270E using a pulsed 5:1 split, injection volume 1.0 uL, cal range 2.0-120 ug/mL. Even with the split, some compounds at the 120 level are clearly starting to show column overload. This also assumes we keep DFTPP tuned to about 400k counts for m/z 69 prior to calibration.

What is your injection volume?
What is your calibration range?
How does the general chromatography look like, i.e. does benzo(b)(k) resolution check look decent, does DDT break down at all? is pcp or benzidine tailing abnormally?

Yeah, the air leak is an ongoing thing, I've narrowed it down to somewhere before the MS, but my leak detector can't find anything, and spraying compressed air at many connection points isn't yielding the compressed air peaks. My guess is the EPC has a leak, but I'm not sure. I must have used enough compressed air to fill a new Hindenburg at this point.

I inject 1.0 uL with a cal range of 0.5-60 ug/mL. General chromotography in general looks good. I would say that my higher concentration points do look like they overload, with some analytes right after my midpoint (20 ug/mL) on my calibration showing the first indication, at as low as 35ug/mL showing some overloading. I took over these GC-MSes back in 2022 as my first GC-MS job, and the GC settings and cal points have stayed the same as I was taught. Tailing factors of PCP and Benzidine of around 1 at my low point on a new column, and benzo(b)(k) resolution looks great, with the valley between them being very low compared to the peaks.

Might be worth noting my columns are 30m Agilent columns. 0.25 mm diameter, Film Thickness 0.5 um. Part number 122-9736. Restek just gave us on of their new RMX GC columns with the same length, diameter, thickness, but we have yet to try it out.

-Also just put in a brand new repeller the other day, and the repeller value is 32, which is odd, makes me think other parts need to be changed instead.
-Also replaced the split-vent trap, but I didn't really anticipate this doing anything.

[MichaelVW]

Can you lower EMVolts so that your responses are in the range they were when everything was working?

[MrPottsWith2Ts]

Can you lower EMVolts so that your responses are in the range they were when everything was working?

Well, I just replaced the EMP with a different one and the EMVolts are 200+ less than with the old EMP, at just above 1200. Unfortunately, based on one curve I've done so far, the problem is persisting. In fact, despite lower EMVolts, the response values were actually higher, which I did not expect. I'm going to lower them even further and then report back.

[MrPottsWith2Ts]

Can you lower EMVolts so that your responses are in the range they were when everything was working?

Well, I just replaced the EMP with a different one and the EMVolts are 200+ less than with the old EMP, at just above 1200. Unfortunately, based on one curve I've done so far, the problem is persisting. In fact, despite lower EMVolts, the response values were actually higher, which I did not expect. I'm going to lower them even further and then report back.

I would say that there was improvement after lowering the EMVolts by about 150. Response of my highest concentration points has dropped back to levels that I have seen in the past. There is still flattening that is making me use quadratic lines of fit on my normally well-performing analytes, but I was able to move some of the analytes that were quadratic lines of fit back to Average of RF lines of fit.

There was one thing I noticed that I could use an opinion on. I tuned yesterday morning and my base peak of 69 gets tuned to an abundance of about 500,000. After I ran my first curve, and before i lowered the EMVolts, I entered into the tune control and scanned to check for any irregularities. The base beak was at about 1,000,000 now! It dropped to about 800,000 over the next 15 minutes or so. Something is wrong here but my boss didn't seem to concerned by that revelation. Looking for other opinions. Also, here is my PFTBA level, in case that was a thought