Dionex UVD340 Diode Array Detector Problem

[chrsdrake]

Does anyone on this forum have a Dionex UVD340 Diode array detector, or experience of this type of detector? If so I wonder if you could help us out with a lamp intensity problem.

I have recently started in the lab here (Reading University, UK) and we have one of these detectors on an old HPLC system, connected to Chromeleon software. I have recently changed the lamp as the old one was giving an error message "low lamp intensity" before the start of each sample batch. However the new lamp has not really cured this problem - the lamp intensity warning threshold is 250,000 counts/sec and the brand new lamp gave only 280,000 when first installed. It has rapidly dropped back down below the threshold.

(I am sure it is not just a poor lamp as I was told that this is typical of what happens when a new lamp is installed). Apparently the system was used for indigo many years ago and it was thought that the problem was due to a "stained or dirty flowcell". I cleaned and then replaced the flowcell with a brand new one - neither gave any improvement.

I have just replaced the "lamp connection tube" and cleaned all the mirrors on the optical bench - apparently the older type lamp connection tubes were made of Pertinax some of which caused precipitation on the mirrors. Anyway this still has not resulted in any substantial improvement, and I have no idea what to try next?

Here are some figures from our old logbook and my recent investigations.

October 2008 - lamp intensity (flowcell removed) 1,491,000
November 2008 - lamp intensity (flowcell cleaned and in place) 445,000 - Reading between the lines it looks as though this was considered low as there are several readings like this during a period of flowcell cleaning with a variety of different solvents/acids.

September 2013 Old lamp (Ref intensity 1,045,890) Lamp intensity 176,910
Old lamp flowcell removed (Ref intensity 1,966,050) Lamp intensity 781,080
New lamp flowcell in place (Ref intensity 1,217,940) Lamp intensity 281,100

I would be very interested to see the equivalent figures for other peoples detectors and to hear anyone's suggestions of what I might try next.

(Note this detector is no longer supported by Thermo/Dionex and my efforts to find out what the lamp intensity and reference intensity specifications were have got nowhere)

Many thanks

Chris

[chrsdrake]

Just been looking through some more old documents and found the original (factory) test on the flowcell gave a lamp intensity of 775,000. The limit value was 700,000.

Our detector is miles away from this.

[Csaba]

We had that problem many years ago but at the time, Dionex REPLACED the mirrors (not washed them) for a resonable cost and changed some other plastic details to Teflon. It is UVD170/340 with serial number lower than a certain limit that got that problem (lost information about the serial number limit). Sorry to hear that Thermo do not support the UVD340 any more so you will have to buy a new (used) detector. Don't you have any other unused detector (Agilent, Waters etc) in the lab? One advantage with Chromeleon is that you can freely mix components.

[LCbob]

Sounds like you have tried everything possible. What solvent do you have in your flowcell, this may help. Check if possible you did not move a mirror out of alignment.
In general with older systems it is usually the detectors that struggle because of the 'blazing' on the mirrors and we cannot get replacements... Search eBay and auction sites for another for parts.
Also if possible ignore the warning, as long as your signal to noise is adequate it should be okay ( check wavelength accuracy if poss ).

[chrsdrake]

Thanks Csaba and LCbob.

It is UVD170/340 with serial number lower than a certain limit that got that problem

- Yes this is the "Lamp connector tube" problem that I mentioned in the original post. Unfortunately we dont really have the money for even a used DAD at present. It is now something of a challenge to see if I can get this one working well again, but as I said I am out of ideas.

Do you still have your UVD340?

What solvent do you have in your flowcell

These readings were taken with 1% acetic acid in water in the flowcell - but the intensity is read at 254 nm so I figured the acetic acid wouldn't make any difference

Check if possible you did not move a mirror out of alignment.

I am fairly sure I didnt move a mirror out of alignment but would love to know how to check the alignment.

In general with older systems it is usually the detectors that struggle because of the 'blazing' on the mirrors and we cannot get replacements

I think it is something like this that is happening, a gradual decline in energy over the years rather than a sudden drop in intensity as you might expect if a mirror had got knocked out of alignment.

However by eye all the mirrors look in excellent condition.

Also if possible ignore the warning, as long as your signal to noise is adequate it should be okay

Yes we are ignoring the warning but it would be very nice to get the detector back to something that we had confidence in. The lab is looking at flavanols in plant samples, the chromatograms are a mass of (mostly small) peaks and the DAD is key to peak assignment. The better the signal the better/easier this peak assignment will be.

Thanks again for your suggestions.

Chris

[Csaba]

Yes, UVD340 are our workhorses. But old Agilent 11000 and Waters 996 has been very robust as well during the years.

[chrsdrake]

Hi Csaba, could I ask a big favour please. If you have any of your UVD340's connected to data systems that have the functionality could you measure the reference energy and lamp intensity (ideally both with and without the flowcell in place). I am not entirely sure what that will tell us but it might suggest some other things to try.

Many thanks

Chris

[ACrawshaw]

Hi Chris,

Can I suggest you contact our service department (Unity Labs)?

0870 2411034
service.cmd.hemel@thermofisher.com

Parts are still available for the UVD340 and these detectors are still serviced, however I would strongly advise making contact with service as they would be able to avise you correctly on the best course of action.

I see a lot of these detectors about on my travels....there may still be life left in the old horse yet!

Thanks

[Csaba]

Hi,
We follow the old Dionex limit that the intensity shall be > 500 000 counts with a cell (lower then that casue investigation if cell i dirty etc).
Usually a new lamp has nowadays has around 800 000 counts.
(Some 10 years ago there were a batch of bad lamps around that had light intensity problems. I do hope you have new lamps.)
Regards
Csaba

[chrsdrake]

Anthony, Csaba

Many thanks for your replies. Antony - I am in contact with the Thermo service people and have been pestering them with this problem for a while. They seem to have run out of ideas too, but I will try and pester them some more till this thing gets sorted.

Csaba - Wow 800,000 for the lamp intensity. With a brand new lamp mine was 280,000 and it is now down below 200,000. Do you know what the reference intensity was/is and what these values are if the flowcell is removed?

Thanks again

Chris