Gradient and acetonitrile with TFA

Part 1. Introduction – Gradient and acetonitrile with TFA

Gradient elution using mobile phase(s) composed of water / acetonitrile containing trifluoroacetic acid (TFA) is often employed for the analysis within most of the laboratories across the country. In gradient chromatography, baseline drift and ghost peaks are most likely to appear, and there are cases where it is difficult (nearly impossible) to decide on objective data processing approach. In this June edition of our chromatography blog, we will take a closer look at baseline drift, wavelength selection and impact of organic mobile phase content on your chromatography.

Part 2. Baseline profiles – wavelength, bandwidth – let’s take a closer look

Figure 1 below presents a 3D chromatogram obtained in gradient elution mode with mobile phase A being 0.01 M TFA in water and mobile phase B 0.01 M TFA acetonitrile. Column used was a C18 column with gradient programme from 0% Mobile Phase B to 100% of Mobile Phase B over 21min. Looking closer at a wavelength of 200 nm in the below figure 1 we can observe a large peak (a) at around the time at which Mobile Phase B first reaches the detector, further a rise in baseline (b) and a baseline drop (c) occur. Moving our point of view a bit further to around 225nm the whole profile changes and a baseline rise (d) becomes evident with further increase of Mobile Phase B (more acetonitrile with TFA).

Figure 2 above was generated just by removing the C18 column with the same chromatographic conditions as in figure 1. One major difference is that peak (a), shown in Fig. 1, disappeared. We can conclude that this originated from the elution of TFA and impurities present in the mobile phase(s) retained inside the column, which absorb in the shorter wavelength range of 200nm – 240nm. On the other hand, it is evident that (b), (c), and (d) occurred at a more fundamental level, meaning they are related to the background created by presence of acetonitrile, water and TFA. We must remember that solvents including their inherent impurities will also absorb the UV light creating certain profiles. Therefore, it is expected that the baseline in gradient chromatography will never be a flat line for entirety of the run time. Additionally, from the data presented in Figure 2, number of chromatograms at different wavelengths were extracted and presented in Figure 3 below.

Figure 3. Chromatograms for various wavelengths (200nm – 235nm) for conditions in Figure 2

Chromatograms from 200nm up to 235nm with increment of 5nm were compared in figure 3. At around 215 nm, the baseline fluctuation is the least erratic. With short wavelengths, 200nm – 210nm, the baseline rises before dropping down to negative values. With longer wavelengths, 220nm up to 235nm, the baseline starts moving upwards after approx. 10min. It seems that the best choice for the analysis will be detector set to 215nm wavelength, assuming the baseline stability is the requirement. Based on the Figure 3 and 215nm chromatogram, the analysis set to run using the 215nm wavelength even with the gradient conditions specified above should be fairly problem free, but will it?

Let’s have a closer look at the selected 215nm chromatogram and check the Figure 4 below. We are still using the same chromatographic conditions as above. All chromatograms produced in the Figure 4 below are of the same wavelength (215nm) however each was collected with different bandwidth settings, from 4nm to 32nm. It’s obvious that different settings of bandwidth will have an impact on the baseline profiles. This gives us a different perspective to acquisition at 215 nm. The wavelength bandwidth for a standard single wavelength UV-VIS detector usually is around 8 nm, therefore you must additionally consider the bandwidth setting impact of your UV detector when performing analysis at 215nm.

Part 3. Organic content with TFA and UV profiles

To further support the impact of the higher organic content on the UV spectrum and selection of 215nm wavelength for the analysis refer to Figure 5, where the change in absorbance at various mobile phase B concentrations is compared.

The content of organic portion was varied from 20% up to 100% and UV spectra were recorded. You can see that this has a vast impact on the UV spectra profile with 80% and 100% organic content having the biggest impact on the profiles after 220nm. However, the “crossing point” of all spectra (215nm) is the common lowest absorption point which in fact displays the least disturbed baseline, as we discussed above. 

Part 4. What about different Mobile Phase A?

Let us have a look at the individual UV spectra of mobile phase A and mobile phase B, as presented in below figure 6. Both mobile phases contain TFA at the same concentration (0.01M) yet the UV spectra differ greatly. Therefore, it can be concluded that the changes in the gradient baseline, mentioned above, occur mainly because of changes in the UV spectra for the mobile phases themselves, rather than because of the way in which they were mixed.

Why do the absorption spectra change when the proportion of acetonitrile increases? Unfortunately, there is no one explanation and various possible causes. One is the suppression of TFA dissociation reaction. Another is mutual interaction between TFA and acetonitrile.

An experiment was performed to maintain the dissociated state of TFA in Mobile Phase A as much as possible. Mobile phase A, instead of water with TFA, was prepared as 0.1 M sodium phosphate buffer solution (pH 2.1) with 0.01 M of TFA. It must be noted that pKa value of TFA is approx. 0.23. This means that with mobile phase A containing only water and 0.01M TFA yielding approx. pH = 2 without any buffering control the dissociated state of the TFA is not controlled. Introducing sodium phosphate buffer at known pH = 2.1 with buffering power of 100mM phosphate salt, means that TFA dissociation will be properly controlled. Additionally, the UV spectrum of the new Mobile Phase A was also recorded in figure 6 for comparison and it is obvious that water and sodium phosphate mobile phases exhibit similar UV profiles. In this experiment the gradient elution was performed going from 0% to 70% concentration of mobile phase B. Only 70% of mobile phase B was explored to prevent phosphate buffer “crushing out” within the system or on the column. Refer to Figure 7 for a 3D chromatogram generated using gradient elution conditions specified above (0% – 70% of mobile phase B).

It is clear, that irregularities and baseline fluctuations observed in Figure 2 are not found in Figure 7. Therefore, it can be concluded that the changes in the spectra are mainly due to changes in the dissociated state of TFA, if the acetonitrile percentage is high, the equilibrium shown below shifts to the left. See the comparison of figure 2 and figure 7 side by side below.

Now it’s a good time to consider substitutes of TFA in mobile phases, also. The same experimental approach was taken using acetic acid, instead of TFA, and a similar form of baseline fluctuations occurred. Before assuming that there is a problem, first consider whether or not this fluctuation is acceptable and whether or not it presents a problem regarding qualitative and quantitative analysis.

Part 5. Summary

I hope that our June edition helps you to understand a little bit more about gradient baseline and the source(s) of the baseline fluctuation. Not always you have to use the best wavelength for your compound of interest but maybe there is a compromise between the full chromatographic system (compound of interest and mobile phase(s) composition). Additionally, the above can help you to troubleshoot baseline fluctuation and maybe reoptimize the method(s) looking at different wavelengths thus improving the potential detectability of your compound(s).

Additional Resources and Further Reading

Product: i-Series Plus | High Performance Liquid Chromatograph

Liquid Chromatography – Master the Basics

This article is part of our “Liquid Chromatography – Master the Basics” series, your go-to resource for comprehensive and insightful updates on the world of liquid chromatography. Each month in 2024 we will dive into a Liquid Chromatography topic, offering content that is both accessible to beginners and beneficial for experienced scientists.

Explore the Series

For information on Shimadzu instrumentation: 

Sebastian Jurek is an application consultant with Mason Technology with specialist knowledge in the Shimadzu range of instrumentation. He holds more than 22 years experience in chromatography techniques and analytical method development, optimisation and troubleshooting.

Get in touch with Sebastian today if you would like further information on our range of Shimadzu products.

Sebastian Jurek
Application Consultant for Shimadzu Chromatography
E:      sjurek@masontec.ie
M:     +353 87 436 4185
T:   +353 1 4154422

Connect with Sebastian on LinkedIn