I am experiencing an interesting detector related chromatography problem I have never seen before. I am developing a method for a drug (an acid) and am trying to decide whether to use UV or fluorescence (FLR) detection. In literature I have seen both methods used. So, figuring I would kill two birds with one stone, I split the flow (0.5 mL/min) post column (C18, 3u, 2.0 x 50 mm) so I could detect my peaks by UV and fluorescence simultaneously. The UV results looked really good, adequate retention, good peak shape (PW = 0.2 min), and good resolution of my analyte and IS (0.5 min baseline separation). However, the FLR results were terrible. The peaks were 0.5 min wide and no longer baseline separated because the tailing was so severe.

I thought the problem was flow rate related. Maybe the split wasn’t even and the flow to the FLR was too low, causing broad peaks. I measured the flow to the FLR and found it to be ~ 0.1mL/min. So I then ran the tubing directly from the column to the FLR (0.5 mL/min), no split. The result: great peaks, shaped just like my UV results. Still wanting to develop my method using both detectors simultaneously and thinking it was just a flow rate issue, I added in the split, increased the flow to 0.7 mL/min and checked to make sure the split was pretty even, 0.4 mL/min to the FLR, 0.3 mL/min to the UV. The result: great peaks by UV, terrible tailing by FLR. It looks like the FLR is going to give me far better sensitivity and specificity than the UV so I would really like to figure this out.

Method: Column - C18, 3u, 2.0 x 50 mm
MP - 45:55 ACN:0.1M Acetic Acid
UV flow cell: 10 uL
FLR flow cell: 16 uL

Any ideas? I have some example chromatography in pdf form that I can e-mail to you, if you like.

Angela Wehr
awehr@shirelabs.com

Check that you have the correct tubing connected to the fluorescence detector. Sounds like you have a larger internal diameter tubing on the FL side. Check it out by swapping the tubing going into each one. Just switch the lines going into each one and check the results. Another thing could be a poorly made fitting which causes a dead volume and mixing of eluent and analyte.

I've been using red peek tubing (i.d. 0.005") so I don't think that the tubing diameter is a problem. I'm not sure how possible the dead volume option is either. The connection at the FL remained the same whether splitting or direct connecting. And if there were a gap at the splitter, wouldn't the UV be affected, too?

When did you measure the flow rate split? Before you connected the detectors, i.e. just with the tubing, or after connecting the detectors? Especially did you check the flow split during or directly after the bad peak shape showed up at the FL detector?

Something else: It is also possible that the volume of the detector cell or the bandspreading in the FL detector is in a critical range. When you ran it at 0.5 mL/min, the peaks had a certain volume. If you split the flow by two, the peak width will be only half the volume. If the detector cell volume of the FL detector is in the critical range, this could cause the effect also.

Instead of splitting the flow (parallel detection), why don't you connect the FL to the UV outlet (sequential detection)? If proper plumbing is applied, the delay between the UV and the FL signal will be just a few hundredths of a minute. Band spreading should not occur any longer (UV cell volume is normally lower than FL cell volume, as Anonymous 9/17/03 03:18 pm pointed out) and back pressure to the UV cell should not be a problem

Agree with 10:12pm; I'd run in series instead of splitting; easier, too.

I always measure my split flow rate as it elutes from my detector, so as to account for the back pressure exerted by the detector.

I never did fix my tailing problem while splitting, but the serial connection was a great idea. That's what I am using now. I still see slightly more tailing by FL than by UV, but I can at least get by with this chromatography. Thanks for the help!