Dirty mobile phase?

[kobie]

Following is the chromatogram of the blank I run. The mobile phase I used is A(1% formic acid) and B(ACN). Gradient program is as follows: 0-1min (2%B), 10min (25% B), 12min (25%B), 15 min (2%B).
I run my sample using this gradient program but found that the elevated baseline interfered with the peaks eluting from 10-15min period. So I run the blank and similar results came out as below. The baseline elevation trend looks exactly like the gradient program with the dwell time delay considered. I wonder how I can bring the baseline back to around zero. Is this caused by dirty mobile phase? I tried to clean my glassware thoroughly and filter the water with Millipore before adding acid but that did not help.
I hope to hear advice from the experienced.

[sepscientologist]

This is a lot of background and quite unusual. The fact that your 254 is larger than your 200 nm background is really odd!

You may have dirty water. While pumping mostly water a contaminant in your water may collect on the stationary phase
and then when you ramp up your ACN it elutes off. I would suggest run the blank gradient twice in rapid succession and see
if it mostly goes away the second time. That would be an indication of dirty water or formic. 1% seems a little high for formic
and I think most people run .1 or .25% or so. You might also add it to the ACN to keep the 200 nm baseline flat. I think
it needs to be a bit lower in the ACN like .1% in aqueous then 0.075% in ACN.

[lynzjm]

The units on your graph suggest you're using an ELSD as a detector?

If so it could be a change in volatility between the volatile formic acid and the MeCN. The baseline seems to start to drift upwards when you start introducing the increased MeCN and come back down again when you begin to decrease it. Line A which has your formic and water is the portion of the run when its smooth.

Other possibilities include (as the change in baseline is pretty small) not 100% clean mobile phase containers etc. I'd keep the mobile phases on and run some test injections varying the switch times in your gradient, perhaps even trying a straight switch to tie down if the changes in baseline follow the changes in your gradient.

x

[tom jupille]

I have always attributed the "mid gradient hump" to RI effects, and my experience has been that they *are* typically more pronounced at longer wavelength. Flow cell alignment can make a big difference in the amount of RI sensitivity.

[8jw8]

I think that is 280nm not 200nm

This is a lot of background and quite unusual. The fact that your 254 is larger than your 200 nm background is really odd!

You may have dirty water. While pumping mostly water a contaminant in your water may collect on the stationary phase
and then when you ramp up your ACN it elutes off. I would suggest run the blank gradient twice in rapid succession and see
if it mostly goes away the second time. That would be an indication of dirty water or formic. 1% seems a little high for formic
and I think most people run .1 or .25% or so. You might also add it to the ACN to keep the 200 nm baseline flat. I think
it needs to be a bit lower in the ACN like .1% in aqueous then 0.075% in ACN.

[John Lim]

Good suggestions by all.

1. The background trace overlays do look like 254 and 280 nm; the 280 nm (purple trace) has a lower profile than the 254 nm (black)---just hard to read the fuzzy numbers

2. I agree that the first thing to check is the water quality. I was not able to see any confirmation that HPLC grade (spectrophotometric) was used...we just need to confirm this. In any case, pop open a fresh bottle of HPLC grade water; or, if using a DI RO water system, have that checked, especially the TOC level for organic breakthrough.

3. Another consideration is the column set. If the column guard (or if no guard present, the separator column inlet) is saturated then running the gradient program on the blank will elute the retained material as the eluent strength increases from the gradient program. The column (and guard) can be cleaned according to the column vendor's instructions...this may take some time; replacing with a new column and after conditioning, then running the gradient will give you some helpful information right away (after verifying the water quality in #2 above).

4. If #2 and 3 check out, then back to square one. Hardware, perhaps. Running the system under hardware IQ conditions is the recommended approach; when presented with this type of situation, our approach is to verify the hardware functionality first, then add the chemistry next...

Hope this helps and good luck!