I have an interesting problem. I am running a reverse phase related compounds analysis with UV detection at 220 nm. I am using a Luna C8 column with a mobile phase at pH about 2. The diluent is mobile phase. After injection of the 1% standard the baseline wanders downward until the first analyte peak elutes. After the peak is eluted the baseline is now the point where the baseline began the run and appears to stay there. I have seen this in coductivity detection but I don't have a clue how this can happen with UV detection. I can't even theorize a corrective action. Also this is an assay that we have successfully performed and validated for over a year. Finally we have remade mobile phase, standards, and switched columns. I hope someone who understands this issue can help me out.

Hi Dan,

some questions to your problem:
- does the shift appear at every or only at the first injection?
- at which temperature is the method running? (guess ambient temperature)
- does the baseline shift disappear at higher temperature or is it increasing?
- could you give some details to the compounds and mobile phase?
- do you have the possibility to check an "ordinary" C8 (with lower %C and without endcapping)?

Right now I've got a plain speculation but I need some more information.

Regards, Andreas

What is the time of the first analyte peak? Are you running isocratic? How much is the baseline "wanders downward" mean?

Andreas
The shift happens on every standard injection.

The method is at ambient temperature.

Don't know what happens at other temperatures since this is the first time we have seen this in over 12 months of weekly running this assay.


The functional groups on this molecule are:

secondary amine
ester
Benzene ring

The isocratic mobile phase is prepared as follows:Dissolve about 6.8 g of monobasic potassium phosphate and 10.1 g of 1-heptane-sulfonic acid, sodium salt in 1 liter of water. Adjust the pH to 2.0 +/- 0.05 with phosphoric acid, and mix. Prepare a suitable degassed and filtered mixture of buffer and acetonitrile (55:45).

The alternate column is a symmetry.

The problem could be due to either overloading (ion pair systems often exhibit rather unusual peak shapes under overload conditions) or due to the fact that your sample solvent isn't focussing (i.e. your sample and solvent have the same aceonitrile concentration). I can't explain why you haven't seen this before but the former can be checked by using a diluted standard while the latter can be checked by diluting the sample so that the sample acetonitrile concentration is <35%.

In ion-pair chromatography, you often get negative peaks from the lack of ion-pairing reagent. I would think that at your wavelength - 210 nm - you see the heptane sulfonic acid as a negative peak. It is nearly unavoidable, since there is a co-migration of the ion-pair reagent with the oppositely charged analyte, which results in a hole somewhere else in the elution of the heptane sulfonic acid.
Solutions:
1. Ignore, if there is no interference with the analytes!
2. Change wavelength to a higher wavelength!
3. Do not use ion-pair reagents!
Since this is a validated method, option 1 may be your only choice. You then may need to be extra careful in the makeup of mobile phase and solutions.

Dan;

The problem you are describing is something I have seen many times. In IP chromatography the column is always in a state of delicate equilibrium with the mobile phase components. Any time you inject something, particularly at high sample concentrations, the equilibrium is interrupted and it takes sometime for the column to re-equilibrate. The observed drift or negative system peak can be due to a change in RI. Do not forget that UV detectors are sensitive to RI changes specially at low wavelengths.

This problem will be more evident when the sample concentration is high, or when the sample solvent is much too different from the mobile phase. I have also seen that the presence of salts in the sample also results in the same problem. For these and other reason it is always a good practice to dissolve the samples in mobile phase, and even better in mobile phase that has been collected from the column efluent.

Good Luck;

Benjamin

After I wrote the conditions the other day I answered my own question. There is some methanol in the standard prep. It appears that the PIC was being washed off. The bottle of PIC was not fresh. We opened a new bottle of PIC and it ran like a champ. Sorry for the delay in my finishing the story. Uwe, why in general don't you like PICS. We have had overall better luck when have both ionic and nonionic analytes than when we use a gradient on an Alliance. Analysis times much faste once we dial in.

Today, with the variety of columns available, I haven't resorted to IPC in a long time. Personally, I would rather change selectivity by changing the column or pH before using ion pairing.

In some ways the newer polar embedded columns can be thought of as reverse phase columns with covalently bonded ion pair reagent. Which prevents many of the equilibrium related problems.

We don't have any problems running gradients with the Alliance systems but I can understand the preference of isocratic IPC over gradient RP. Given that preference, I think the percentage of methods that end up being IPC will reflect the underlying development methodology.

If I take my preferred column and start development with scouting gradients with various organics I could quickly determine that the k' range is to large for an isocratic sep. Now, not wanting to continue with a gradient method I could try IPC.

However, if I started with a column screening routine, I would have a much higher probability of being able to choose a column/pH with an acceptable k' range with which to continue isocratic development.

In the old times, ion-pair reagents were good for a lot of things, from fixing peak shape to create resolution to getting all the peaks into the right retention window. In the old times, you also got lousy peaks when you moved away from an acidic mobile phase.
Today's packings do not have this problem any more, and I am using pH freely to get the peaks to where I want them to be, and - as Tom said - we can use gradients today.
I have nothing against ion-pair reagents, but I consider them to be a rather limited tool, with problems.

My applications will eventually be performed routinely in a pharmaceutical quality control laboratory. The requirements are that the method run by many different chemists on different equipment for a long time. Also the methods have to be run in one of the district laboratories for method validation. With these constraints I focus on developing methods that are relatively independent of column and instrument brand. That is a C18 is a C18 and HP (oops Agilent) instrument runs like a Waters instrument. These goals can usually be obtained, but when there are both ionic and neutral compounds the challenge is great. The problem that we have had with gradients is that they seem to be very dependent on instrument brand and model and column brand. There has also been a great deal of sensitivity to solvent variability batch to batch. In general, the gain has not been worth the trouble. Also when I take into account reequilibration time the methods always take longer to run. The problem that I have is that my young chemists like gradients because they are easier to develop and require less thinking about chemistry. Most of the courses I send people to talk about scout methods and a method development algorithm to follow. I would like to know how other people deal with these issues.

i have USP method the column is 4.6 mm X25 cm 5um packing L1. i dont have this type of column i use instead 4.6 mm X 15 cm 5 um