Sinc(x) wave at tail edge of TCD peaks

[drexel13]

Hi all,

Running an Agilent 7890 w/ TCD. Things were fine for some time, then the instrument wasn't used for a while. Got it back up and running and everything goes OK except that the trailing edge of peaks have a sinc(x) looking wave shape to them. This is more obvious on larger peaks. There's 2 screen caps that illustrate this attached.

We've checked the column installation and that seems fine. Any other thoughts as to what could cause this? My FID troubleshooting experience is a lot more than TCD troubleshooting...

Thanks!

[GasMan]

What are your GC conditions? What column are you using and what are the peaks? It looks at first glance that you are overloading the column.

Gasman

[drexel13]

Thank you for the reply. The method was reproduced from parameters obtained from an outside analytics company that helped determine how best to separate the peaks of interest. It has not changed and it didn't do the sinc(x) tailing before, but here are all the details:

All gasses are He

Inlet: 340 C, 10psi, 105 mL/min total flow, 3mL/min septum purge (standard mode), plot flow ratio 47.149:1 (split flow 100 mL/min)

Column: DB-5 phase, 30m x 320 um x 0.25 um, Flow = 2.12 mL/min, pressure = 10 psi, ave velocity = 34.9 cm/sec. holdup time = 1.4324 min, constant flow

Oven: 50 C for 4 min, 10 C/min to 100 C hold 0 min, 25 C/min to 290 C hold 5 min.

Detector (TCD): 300 C, 30 mL/min reference flow, 2 mL/min makeup flow (4.12 mL/min const col + makeup), 11.5 (25 uV) filament, unchecked negative polarity, no TCD signal subtraction

The peaks in the first spectrum are (from L->R): acetaldehyde impurity in acetic acid solvent, water, very small unidentified peak, huge acetic acid solvent peak, toluene (~2wt%), furfural (~5wt%)
The other spectrum is just zoomed in on the acetic acid peak to better illustrate the sinc(x) wave

I know overloading the column gives you non-ideal peak shape (like for the solvent and furfural), but I've never seen it cause the sinc(x) squiggles like that before.

Let me know if I forgot anything. Thanks again!