Chromatography Forum

Mobile phase: 78/22 Methanol/Water
Flow 1.0ml/min
Column 100x4.6 3micron Thermofisher ODS-2

If any other parameters may be necessary please let me know.

In a period of just a couple weeks, one peak in the samples we analyze has drastically changed shape. Initially it eluted as two peaks, isomers of each other. Now I get one big glob.

Fortunately, it seems to quantify the same. But even so that doesn't answer the question as to why it looks like this. I have attached a picture of a chromatogram of the standard solution.





One possibility is the standard solution has degraded - but the peak shape is the same in fresh formulated samples from our formulation lab as in my standard solution. Could be the raw material? But did it really go bad in just a matter of weeks to where it impacts the peak shape in such a manner?

I have also tried multiple colums, another identical column as well as a thermofisher hypersil gold C18 column - same problem.

I've tried fresh water and methanol in the mobile phase as well - same problem.

So at this point, while it hasn't impacted my results yet, I think it's a problem that should be solved, and I'm not sure where else I can check. I haven't made any other changes to the system between the two analysis where the peaks look good then poor.

Open to suggestions or ideas.

The second (bad) chromatogram looks as if those two peaks are still there as they should be, but each one of them is split into two peaks - so that "blob" in fact are four unresolved peaks. Looking closely at the early eluting peak at ~3-3.5 minutes, there seems to be the same problem, visible as a shoulder at the right tail of the peak.
By trying different columns and getting the same bad result you ruled out a column defect as the culprit. Did you also try different HPLCs? If all runs were performed on the same machine, I'd suspect an injector problem. Rotor seal in the injector might be worn out leading to channeling.

The second (bad) chromatogram looks as if those two peaks are still there as they should be, but each one of them is split into two peaks - so that "blob" in fact are four unresolved peaks. Looking closely at the early eluting peak at ~3-3.5 minutes, there seems to be the same problem, visible as a shoulder at the right tail of the peak.
By trying different columns and getting the same bad result you ruled out a column defect as the culprit. Did you also try different HPLCs? If all runs were performed on the same machine, I'd suspect an injector problem. Rotor seal in the injector might be worn out leading to channeling.

I actually have not tried running on our other instrument (I try to avoid using it until we replace it because it constantly gives communication errors between the system and the detector) but I will give that a shot.

I went back and did look through our recent runs and in the beginning of April the peaks looked great, then about a week later they started to have a more round apex, then a week later they looked a bit worse, and now they look like I posted. So it looks like it wasn't as sudden as I thought but did gradually get worse.

It could be as you mentioned, an injector problem and the seal has gotten worse and worn down over the past month. What's interesting is I don't see any kind of odd peaks in any of our other samples we run that all have different active ingredients. It's mostly this one compound that is looking very poor. Would a poor rotor seal and injector issue not effect all compounds that get injected?

I'll check out the injector and see what I can find.

Just to update, I ran the same injections on our other instrument and the peak still looks as if it's 3 - 4 peaks co-eluting. So it's most likely not a rotor seal issue unless I also have a rotor seal issue on the other instrument.

I don't think it's hardware. Any pH control in the mobile phase?

I don't think it's hardware. Any pH control in the mobile phase?

No pH control. We've always been running this method with Methanol:Water, 78:22 respectively, isocratic.

I'm going to ask some more parameters.

- Is there temperature control of the column?
- Injection volume and solvent composition?
- Solvent composition of needle wash?

As a diagnostic test, i think it might be worthwile to inject ~10x less of the same solution (or dilute 10x with mobile phase and inject the same volume).

I'm going to ask some more parameters.

- Is there temperature control of the column?
- Injection volume and solvent composition?
- Solvent composition of needle wash?

As a diagnostic test, i think it might be worthwile to inject ~10x less of the same solution (or dilute 10x with mobile phase and inject the same volume).

Sample concentration of the molecule in the solution being injected is 0.8mg/mL with a 5uL injection volume.

Temperature is controlled at 30ºC and does not fluctuate more than 0.3 - 0.5ºC during an entire sequence, let alone one injection.

Injection solvent is Acetonitrile and the needle wash is also Acetonitrile.

Pure Acetonitrile as injection solvent? Then I'd try definitely what Rndirk suggested: Inject much less or dilute your sample solution with mobile phase to see if this is due to the strong solvent effect.
Is the needle wash only to wash the outside of the needle? If yes you could probably keep neat ACN here, but if the inside of the needle or the sample loop is also swept with it, I'd change it to something weaker, too.

Pure Acetonitrile as injection solvent? Then I'd try definitely what Rndirk suggested: Inject much less or dilute your sample solution with mobile phase to see if this is due to the strong solvent effect.
Is the needle wash only to wash the outside of the needle? If yes you could probably keep neat ACN here, but if the inside of the needle or the sample loop is also swept with it, I'd change it to something weaker, too.

Yes the sample is a liquid formulation, so I just dilute the formula simply with acetonitrile. I realize that the solvent in this case is pretty strong compared to the mobile phase. I just find it interesting that always having done this, we've never seen this issue until just recently.

As for the needle wash, I'll be honest I don't know. I don't know enough about the engineering of the PerkinElmer Altus (the instrument I mainly use) to know whether the needle wash cleans the outside of the needle or gets swept into the loop. I will ask one of the engineers on that.

I will go ahead and both dilute the sample with mobile phase by 10x and I will also run the analysis by mixing the formulation with mobile phase and injecting that at the same concentration to see how thins look.

If I do run the samples with Acetonitrile/water though the peak doesn't completely split but it's two very sharp co-eluting peaks opposed to one big mess.

We rarely used ACN as a solvent for standards and samples, but we very often used methanol or DMF for that, with water-methanol and water-ACN mobile phases. Oftentimes we added some acetic acid to the aqueous phase to stabilize pH.

And when we injected 5 µl and had flow rates even lower than the poster, we never had splitting, but like I said our experience using ACN for standards and samples was limited. But injecting less and seeing if peak shape improves is never a bad idea.

I just noticed the scale in the OP chromatograms, the upper one gives peaks ~3x larger than the lower one. Is this the same sample? Could it give a hint to what's going on?

I just noticed the scale in the OP chromatograms, the upper one gives peaks ~3x larger than the lower one. Is this the same sample? Could it give a hint to what's going on?

The first image, with good peaks, was actually a 10uL injection. My co-worker had forgotten to change the injection volume. The second chromatogram is a 5uL injection volume.

I just noticed that as well, thanks for pointing that out.

I will try a few things today/Monday and update with some news with what changing solvents or lowering injection volume or diluting with mobile phase.

I'll take a look at using a weaker solvent as well. In most of the methods I've developed since I've gotten here I try to stick with isopropanol or methanol as my solvents of choice for standard and sample prep.

We need to full chromatography details. Please provide the EXACT Detection parameters (Wavelength & bandwidth,use of any optional reference wavelengths) for this isocratic run. Please share with us the exact make and model of detector plus the flow cell size. This is critical data to have and may or may not explain the results. W/O it, you will never know.

A few problems in technique are noted.

(1) Using an injection solvent that is stronger than the mobile phase does not follow proper chromatography guidelines. Use the mobile phase.
(2) No mention of how the column is washed down in between analysis runs. It should be washed with a solvent system that is stronger than the one used for the method to remove any accumulation of material which will invalidate the method.
(3) The sample size (conc) seems very high for such a low volume column. Was a proper loading study run to confirm it is within the linearity and detection window of the detector?

We need to full chromatography details. Please provide the EXACT Detection parameters (Wavelength & bandwidth,use of any optional reference wavelengths) for this isocratic run. Please share with us the exact make and model of detector plus the flow cell size. This is critical data to have and may or may not explain the results. W/O it, you will never know.

A few problems in technique are noted.

(1) Using an injection solvent that is stronger than the mobile phase does not follow proper chromatography guidelines. Use the mobile phase.
(2) No mention of how the column is washed down in between analysis runs. It should be washed with a solvent system that is stronger than the one used for the method to remove any accumulation of material which will invalidate the method.
(3) The sample size (conc) seems very high for such a low volume column. Was a proper loading study run to confirm it is within the linearity and detection window of the detector?

Analysis is performed at 215nm, no reference wavelength, bandwidth of 1.2nm.

The detector is a PerkinElmer Altus A-10 PDA. To be honest, I'm not sure the flow cell size, I will have to look into that and get back to you.

To address the rest:
(1) Yes I am aware that using a stronger injection solvent than the mobile phase can cause problems. Unfortunately the method was validated under GLP in this manner and I haven't had time to revalidate it. So when we are running GLP studies this is what we are using. I did not create this method.
(2) My column wash is a gradient that goes from 50 - 100% Acetonitrile and back down to 40/60 Water/Acetonitrile for storage.
(3) Linearity was performed on all active ingredients of interest that we analyze in this analysis and this method has been GLP validated. I have been running the method as is for the past 5 to 6 months and it is only until just recently that this one peak seems to become distorted on both HPLC instruments we have, when analyzed on different columns of the same make/manufacturer (including different lot # for columns).