Help with ELSD

[wormwoody]

Hi all,

In early 2010 I developed and validated an analysis for degradation products of gabapentin in an oral liquid pharmaceutical product using LC-ELSD.
The method was developed and validated in a very short timeframe as part of a licence submission.

The method was capable of detecting down to a 0.05% level of the stated dose and as such was at the limits of sensitivity forthe deetector.

As such it was never pretty but largely fit for purpose. After a year or two of running it the ELSD started to lose sensitivity, numerous service engineers visits later and the situation never improved (at the time it was the only ELSD within the company so comparison to other detectors was not possible).

Finally this year the company the company bought another detector.
Having got it set up I've immediately tried the method on it..................

At the moment even after optimisation of gas flows evaporator temp at best it gives half the response of the old detector ( originally we'd get a peak height of 45-50mV for the calibration Std on the new detector its 22mV at best).

So does anyone have any ideas as to whats happened?

Conidtions are as below

Agilent 1200 with Varian (now Agilent but was Polymer Labs 385 ELSD).

mobile phase 99:1 10mM Ammonium formate @ pH 3.2 with formic acid:Acetonitrile

Flow of 1ml/min with column at 50°C

Column is a 150x2.0x4 Synergi Polar RP

Thanks for any replies

[DR]

1) I suspect that your flow rate is a little higher than optimal for a 2mm column packed w/ 4µ particles but I also suspect that this is minor and not a potential solution to the problem.

2) Have you tried another example of this column?

3) Is there any difference between the current chromatography and what it looked like before (aside from heights)? Specifically, look for changes in peak shape, retention times, resolutions... stuff like that that would indicate a column issue.

4) Did the detector meet manufacturer specifications when the technicians left? Does the new one meet specifications now? If the new one meets specs now and you can rule out column problems, it may be time to look for a better ELSD and send the new one back... (anyone know of better options? I am not well versed in the current ELSD "pecking order").

[wormwoody]

Thanks for the reply, the flow is possibly a bit high but it runs conform ably enough at 200bar. I was more concerned about flooding out the detector, but really 1ml/min shouldn't be a problem. The method has been in use for about 2 years so a number of columns wil have been used.

The manufactures specs were all met on install, though the result I get for it currently is on the edge of the spec for noise. In all the various visits of the service engineers they've been worse than useless, I suspect this stems from the fact that they're ELSD engineers and as such have no idea about using one in conjunction with an LC.

By and large the chromatography is comparable there is no obvious tailing or detioration of peak shape, it's just half the height (roughly) of what I'd expect to see.

[Uros]

Did you alter the PMT gain. In Varian ELSD, you have two knob on the mainboard, to turn them around and set the size of the peak (set voltage to the PMT).

[Carvone]

I am not sure about the design of the Varian ELSD as I have a Sedex ELSD, but for this instrument I found over time the corrugated black plastic exhaust tube was building up water condensation inside the tube ribbing and this seemed to be the problem causing lower sensitivity and greater noise.

When I took the exhaust tube off, instantly the sensitivity increased and just by placing my hand in front of that exhaust close to the outlet I could affect the signal. I ended up replacing the factory tube with a much wider diameter (1") clear non-corrugated flexible plastic tube that was just long enough to go out the back of the instrument when the cover was placed back, and I have not had a problem since. I only use water as my mobile phase so it is not a big issue venting into the lab, but if other solvents are used you will need to vent this into a fumehood for proper exhaust.

[wormwoody]

Hi thanks for your replies, well I've got the new detector working not sur ehow or why I suspect it was a combination of issues.

The only problem now is the old ELSD now gives no peak responses for anything, I've followed all the various wash and low out procedures for the detector. I've tried performing the instrument OQ procedure with glucose probes and essentially I get no peaks. I have a baseline i.e its not totally flat but whatever analyte I inject with or without a column give no response.

Carvone I'll have a look at the outlet and see how it impacts the detector, thanks for this suggestion.

[wormwoody]

There appears to be no issues with the outlet, PMT won't and doesn't have any impact on signal response. I guess its probably of most use using a chart recorder as data capture, our data is captured by a CDS so it will scale accordingly.

I've been reviewing our data and depending on the method used I can trend upto a 30x loss of signal depending on the mthod used with no comparable loss of chromatography. Yet everytime we have a service engineer visit (at considerable cost) we are told there's no problem with the detector and it passes it's OQ fine.

At a loss really.