I get quite a few customer questions concerning peak tailing during LC analysis, and how to best troubleshoot this issue. Peak tailing may be attributed to a variety of different causes including secondary interactions, contamination, column loading, etc. This list goes on and on. I usually ask a few key questions, and generally can give some good advice based on the responses.
- Is this a new problem?
- What are your analytical conditions?
- Mobile phases?
- Which analytes are troublesome?
During some applications work I was doing in the lab, I came across a very familiar issue: All of my peaks are tailing...
I asked myself the above questions:
- Is this a new problem? Yes, I just started method development.
- What are your analytical conditions? Flow rate – 0.5 mL/min, Temp – 30 °C
- Mobile phases? MPA: 0.1% formic acid in H2O, MPB: 0.1% formic acid in MeOH
- Column? Raptor Biphenyl, 2.7 µm 50 mm x 3.0 mm I.D.
- Gradient? 10-75%B in 5 minutes, re-equilibrate to 10%B for 2 minutes
- Which analytes are troublesome? All of them. A mixture of 6 beta-blockers.
Nothing struck me as particularly odd with the analytical conditions, all analytes were well retained and eluting during the gradient, and there was sufficient re-equilibration time. As I looked at the analyte structures, I noticed that these were all very active compounds.
Based on this information, I believed the peak tailing was attributed to secondary interactions caused by silanol activity. The cartoon below depicts how the secondary amine in the pindolol molecule could be potentially interacting with residual silanols on the surface of the silica.
With the addition of a buffer to my mobile phase, I should be able to mitigate this issue. Since I’m already using formic acid, choosing a complimentary salt, Ammonium formate, will give the best buffering capacity at the desired pH.
By changing the composition of my mobile phase, I was able to reduce the secondary interactions, resulting in much more symmetrical peaks, and even got the bonus of some increased resolution between labetalol and oxeprenolol. With a flow rate of 0.5 mL/min, the total volume of mobile phase used per injection is 1.4 mL of MPA and 1.1 mL of MPB – for all intents and purposes these are relatively equal. The addition of buffer to both aqueous and organic mobile phases ensures that the secondary interactions will be mitigated throughout the entire gradient – solving the peak tailing issues for both early and late eluting compounds.
See http://www.restek.com/chromatogram/view/LC_GN0550 for the final optimized method using 0.1% Formic acid and 5mM Ammonium formate modified mobile phases.