Effect of Organic Solvent on Selectivity in LC Separations
Different organic mobile phase modifiers produce different chromatographic results, and the foundation of good separations is leveraging choices in mobile phase and stationary phase chemistries to create different degrees of interactions between them and the analytes of interest. In this work, we will explore how the differences in organic modifiers result in the chromatographic performance you observe when using acetonitrile or methanol with reversed-phase LC columns. Specifically, we will be looking at a classic C18-type stationary phase as well as another popular reversed-phase stationary phase, biphenyl. Let’s start with the C18.
One of the first differences you often learn about organic modifiers is “elution strength,” which is sometimes represented as an “eluotropic series.” The eluotropic series is just a list of solvents in order of elution strength where a solvent with a strong elution strength is one that solutes have a high affinity for relative to the stationary phase. The consequence of this high affinity is that the analytes stay in the mobile phase more, reducing interactions with the stationary phase, which results in less retention. Analytes typically have lower affinity for organic modifiers with weaker elution strengths, and so, in this case, the same compounds would have more interactions with the stationary phase and exhibit greater retention, all other things being equal.
When it comes to an eluotropic series for reversed-phase chromatography, acetonitrile has a higher elution strength than methanol. Therefore, we would expect, as we see in Figure 1, that at the same percentage of organic modifier in the mobile phase, we would see less retention for the analytes with acetonitrile than we do with methanol.
But, there are times when someone may be interested in switching from one organic modifier to another while still trying to achieve the same separation in approximately the same time. The question is, can that be easily done?
To account for differences in elution strength between methanol and acetonitrile, you need to change the amount of the organic modifier in order to match retention times. When switching from acetonitrile to methanol, you will need to increase the overall elution strength of the mobile phase by increasing the percentage of methanol. There are tables (e.g., see Figure 4 in reference [1]) that give estimates of what percent of methanol, for instance, matches the elution strength of a given percent of acetonitrile. Using these tables, we were able to figure out a methanol mobile phase composition that matched the elution strength and chromatographic performance of the original acetonitrile conditions fairly well (Figure 1).
Figure 1: Comparing the retention times of various reversed-phase analytical probes (a) using acetonitrile as an organic modifier, (b) using methanol at the same mobile phase composition, and (c) using a percent methanol empirically determined to match the elution strength of the acetonitrile analysis. Column: Raptor ARC-18 (2.7 µm, 50 x 2.1 mm); Peaks: 1. uracil, 2. toluene, 3. naphthalene, 4. biphenyl.
So, in this example, if you want to switch from acetonitrile to methanol (or vice versa), you can figure out the organic modifier composition that will match the elution strength of the method you are trying to convert. However, there are times when that approach does not work out so easily. The following cannabinoids analyses show that the effect of organic solvent on selectivity in LC is not always simple to predict.
Let’s say we have a method in acetonitrile that we want to switch over to methanol. If we do the same experiment that we did in the example above, where we just switch modifiers and keep the percentage of organic modifier the same, we see our first clue that matching elution strengths alone might not work (Figure 2).
The initial thing you notice is that, as expected, the run is longer (a lot longer, in this case) with methanol. Methanol is the weaker elution strength solvent, so it is expected that compounds elute later. But, if you look carefully, you will see that a number of compounds in this example move significantly in relation to nearby compounds, and, in some cases, they even switch elution order all together! So, we don’t just have a change in the retentive character of the chromatography (strong elution strength, less retention; weaker elution strength, more retention), we have a change in selectivity (how compounds elute in relation to each other).
The prominent examples of that, in this case, are instances where the acidic compounds (e.g., CBDVA, CBDA, THCVA, and CBNA) show greater retention than their neutral counterparts (e.g., CBDV, CBD, TCHV, CBN) when methanol is the organic modifier compared to when acetonitrile is used. Again, that everything would show greater retention is expected, because methanol is the weaker organic modifier, but some compounds are showing an even greater degree of retention than their neighbors when methanol is used when compared to acetonitrile at the same percent composition.
Figure 2: Comparison of the effect of organic solvent on selectivity in LC, where the percent composition is kept the same for both, on commonly monitored cannabinoids.
| Peaks | Conc. (µg/mL) | Acetonitrile tR (min) | Methanol tR (min) | |
|---|---|---|---|---|
| 1. | Cannabidivarinic acid (CBDVA) | 50 | 1.877 | 3.117 |
| 2. | Cannabidivarin (CBDV) | 50 | 2.086 | 2.563 |
| 3. | Cannabidiolic acid (CBDA) | 50 | 2.592 | 5.003 |
| 4. | Cannabigerolic acid (CBGA) | 50 | 2.750 | 6.678 |
| 5. | Cannabigerol (CBG) | 50 | 2.912 | 4.373 |
| 6. | Cannabidiol (CBD) | 50 | 3.084 | 4.233 |
| 7. | Tetrahydrocannabivarin (THCV) | 50 | 3.391 | 4.904 |
| 8. | Tetrahydrocannabivarinic acid (THCVA) | 50 | 4.279 | 11.237 |
| Peaks | Conc. (µg/mL) | Acetonitrile tR (min) | Methanol tR (min) | |
|---|---|---|---|---|
| 9. | Cannabinol (CBN) | 50 | 4.609 | 7.643 |
| 10. | Cannabinolic acid (CBNA) | 50 | 5.437 | 17.535 |
| 11. | Δ9-Tetrahydrocannabinol (Δ9-THC) | 50 | 5.815 | 9.866 |
| 12. | Δ8-Tetrahydrocannabinol (Δ8-THC) | 50 | 6.002 | 10.747 |
| 13. | Cannabicyclol (CBL) | 50 | 6.916 | 10.865 |
| 14. | Cannabichromene (CBC) | 50 | 7.263 | 14.387 |
| 15. | δ-9-Tetrahydrocannabinolic acid-A (THCA-A) | 50 | 7.612 | 23.975 |
| 16. | Cannabichromenic acid (CBCA) | 50 | 8.510 | 28.943 |
| Column | Raptor ARC-18 (cat.# 9314A65) |
|---|---|
| Dimensions: | 150 mm x 4.6 mm ID |
| Particle Size: | 2.7 µm |
| Pore Size: | 90 Å |
| Temp.: | 30 °C |
| Standard/Sample | |
| Inj. Vol.: | 5 µL |
| Mobile Phase | |
| Flow: | 1.5 mL/min |
| Detector | UV/Vis @ 228 nm |
|---|---|
| Instrument | Waters ACQUITY UPLC H-Class |
| Notes | Mobile Phase Details Acetonitrile (top) A: Water, 5 mM ammonium formate, 0.1% formic acid B: Acetonitrile, 0.1% formic acid 9 min isocratic run (75%B) Methanol (bottom) A: Water, 5 mM ammonium formate, 0.1% formic acid B: Methanol, 0.1% formic acid 30 min isocratic run (75%B) |
Next, we will try to match elution strengths and see what happens. Figure 3 shows the results of that experiment (note that the concentration of the ammonium formate buffer is different in two conditions, but it was changed to keep the overall amount delivered on-column the same, like it is in Figure 2). We can see that in a few cases, CBN being a good example, we do get a pretty good match in analyte retention when we match elution strength, but the change in selectivity that we observed in Figure 2 is still there. The nature of the analyte itself has a lot to do with how effective an approach like matching elution strengths is going to be when it comes to trying to get similar results with different organic modifiers.
Figure 3: Comparison of the effect of using a different organic modifier, but at the same elution strength (via higher concentration), on cannabinoid selectivity.
| Peaks | Conc. (µg/mL) | Acetonitrile tR (min) | Methanol tR (min) | |
|---|---|---|---|---|
| 1. | Cannabidivarinic acid (CBDVA) | 50 | 1.877 | 1.998 |
| 2. | Cannabidivarin (CBDV) | 50 | 2.086 | 1.700 |
| 3. | Cannabidiolic acid (CBDA) | 50 | 2.592 | 2.803 |
| 4. | Cannabigerolic acid (CBGA) | 50 | 2.750 | 3.479 |
| 5. | Cannabigerol (CBG) | 50 | 2.912 | 2.522 |
| 6. | Cannabidiol (CBD) | 50 | 3.084 | 2.522 |
| 7. | Tetrahydrocannabivarin (THCV) | 50 | 3.391 | 3.030 |
| 8. | Tetrahydrocannabivarinic acid (THCVA) | 50 | 4.279 | 5.876 |
| Peaks | Conc. (µg/mL) | Acetonitrile tR (min) | Methanol tR (min) | |
|---|---|---|---|---|
| 9. | Cannabinol (CBN) | 50 | 4.609 | 4.137 |
| 10. | Cannabinolic acid (CBNA) | 50 | 5.437 | 8.158 |
| 11. | Δ9-Tetrahydrocannabinol (Δ9-THC) | 50 | 5.815 | 5.170 |
| 12. | Δ8-Tetrahydrocannabinol (Δ8-THC) | 50 | 6.002 | 5.605 |
| 13. | Cannabicyclol (CBL) | 50 | 6.916 | 5.605 |
| 14. | Cannabichromene (CBC) | 50 | 7.263 | 6.828 |
| 15. | δ-9-Tetrahydrocannabinolic acid-A (THCA-A) | 50 | 7.612 | 10.969 |
| 16. | Cannabichromenic acid (CBCA) | 50 | 8.510 | 12.399 |
| Column | Raptor ARC-18 (cat.# 9314A65) |
|---|---|
| Dimensions: | 150 mm x 4.6 mm ID |
| Particle Size: | 2.7 µm |
| Pore Size: | 90 Å |
| Temp.: | 30 °C |
| Standard/Sample | |
| Diluent: | Methanol |
| Inj. Vol.: | 5 µL |
| Mobile Phase | |
| Flow: | 1.5 mL/min |
| Detector | UV/Vis @ 228 nm |
|---|---|
| Instrument | Waters ACQUITY UPLC H-Class |
| Notes | Mobile Phase Details Acetonitrile (top) A: Water, 5 mM ammonium formate, 0.1% formic acid B: Acetonitrile, 0.1% formic acid 9 min isocratic run (75%B) Methanol (bottom) A: Water, 7.14 mM ammonium formate, 0.1% formic acid B: Methanol, 0.1% formic acid 13 min isocratic run (82.5%B) |
In our cannabinoid case, many of the compounds seeing the greatest change in relative retention are acidic in nature. And, that is what you might observe if you try this yourself. Some compounds are affected to a greater degree in relation to others, enough to change the selectivity of the chromatography. In a case like this, it is possible the cause is an additional interaction mechanism that exists between the organic modifier and the analyte itself based on the analyte’s chemistry.
In our first example, we did not have any selective interaction affecting only one or two of the compounds based on organic modifier, so we could match elution strengths and adjust retention accordingly. As seen in the cannabinoids example, that will not be the case when there are compound-specific retention characteristics. In those circumstances, you cannot simply match elution strength and expect to produce similar chromatographic results.
It is true that the effect of organic solvent on selectivity may not compromise performance if you have the correct peak identification, but in our example, the change in selectivity causes coelutions that aren’t acceptable for this LC-UV cannabis potency analysis.
But, does this hold true for other reversed-phase column chemistries? So far, we have only used a C18 column, but what about another popular reversed-phase stationary phase, such as biphenyl? To explore further, let’s look at the first example shown on the C18-type column, the analysis of uracil, toluene, naphthalene, and biphenyl (Figure 1). This time, we will recreate the conditions on the biphenyl stationary phase instead.
Figure 4: Comparison of the effect of organic solvent on selectivity in LC on two different reversed-phase stationary phases using a matching percentage of organic modifier. All other things being equal, note how the last compound, biphenyl, moves when using acetonitrile compared to methanol. Columns: Raptor ARC-18 (2.7 µm, 50 x 2.1 mm), Raptor Biphenyl (2.7 µm, 50 x 2.1 mm); Peaks: 1. uracil, 2. toluene, 3. naphthalene, 4. biphenyl.
In Figure 4, we see as expected, an overall increase in retention time for all of the hydrophobic compounds (toluene, naphthalene, and biphenyl) when using methanol compared to acetonitrile (the first peak to elute is uracil, which is hydrophilic and is used as a void time marker in this reversed-phase analysis). What is not as immediately apparent, however, are the subtle retention time differences caused by the different solvents when compared to the C18 data. Take, for example, the final peak to elute in this example, which is the compound biphenyl (not the stationary phase biphenyl, but the analyte). When using a mobile phase with 55% acetonitrile on the C18 column, the biphenyl peak elutes at 1.7 minutes, but on the biphenyl column when using the same 55% acetonitrile mobile phase that same peak elutes at 1.176 minutes. So, there is less retention on the biphenyl column when using acetonitrile. If you compare the data using methanol as the organic modifier at 55%, you see on the C18 column the biphenyl peak elutes at 5.995 minutes compared to a retention time of 7.44 minutes of the biphenyl column. So, in this case, there is more retention on the biphenyl column when using methanol. Why is that?
This difference that we see in retention and selectivity between the C18 and biphenyl phases when changing organic modifiers in the mobile phase has to do with the different retention mechanisms of the stationary phases themselves. The biphenyl stationary phase differs in selectivity from a C18 phase because its aromatic phenyl rings are capable of pi-pi interactions with the analytes of interest. The polarizability of the biphenyl phase allows it to change its electron distribution in the presence of an analyte and induce a dipole interaction. These interactions are most commonly seen in dipolar, unsaturated, or conjugated compounds. The unique selectivity of the biphenyl phase is affected differently than a C18 phase is by the choice of organic modifier.
Structurally, acetonitrile has a central carbon atom that has a triple bond with a nitrogen atom. The triple bond is also capable of contributing to or interfering with pi interactions. So, when looking at the retention time of the biphenyl peak when using acetonitrile as the organic modifier, this results in less retention on the biphenyl column compared to the C18. However, when you switch to methanol, an organic solvent that does not have pi bonds, this results in an increased pi-pi interaction between the stationary phase and the analytes, resulting in a different selectivity and increased retention on the biphenyl column compared to the C18 for the biphenyl analyte, and to a lesser extent naphthalene. Acetonitrile has the capacity to have an additional interaction with the stationary phase that isn’t present with methanol, and that additional interaction can fundamentally affect how the analytes interact with the phase, further enhancing the elution strength of acetonitrile beyond what the experience with the C18 would predict.
How does this affect our ability to match elution strengths and get comparable chromatography? In Figure 5, we first evaluated the same conditions that gave a good match with the C18 (Figure 2). As we might expect, when we try to get the same performance on the biphenyl column using the same “matched-elution-strength” mobile phase conditions, we still have more retention when using methanol. With a greater degree of interaction between the biphenyl stationary phase and the organic mobile phase modifier, the effective elution strength of acetonitrile is greater on the biphenyl column than on the C18 column. To account for the difference in effective elution strength, we increased the methanol content beyond what is theoretically considered a matched percentage for the original acetonitrile conditions and produced very similar results in a similar amount of time.
Figure 5: Results of attempting to match the chromatographic separation on a biphenyl stationary phase using different mobile phases: (a) acetonitrile mobile phase, (b) methanol mobile phase using the same percent methanol that matched well on a C18-type column, and (c), a higher percent methanol mobile phase to account for the enhanced elution strength of acetonitrile on the biphenyl stationary phase. Columns: Raptor Biphenyl (2.7 µm, 50 x 2.1 mm); Peaks: 1. uracil, 2. toluene, 3. naphthalene, 4. biphenyl.
So, in the case of the biphenyl column compared to the C18, it was still possible to account for acetonitrile’s enhanced elution strength, which stems from its interaction with the biphenyl phase itself, by increasing the amount of methanol beyond the matched elution strength percentage predicted for C18 columns.
To further explore the effect of organic solvent on selectivity in LC, we can look at the example of oxycodone’s metabolite, noroxycodone, and the painkiller dihydrocodeine on the biphenyl column in Figure 6. These two compounds are structurally similar, and both have an aromatic ring. We see the expected difference in elution strength with more methanol being required to get similar retention to a given amount of acetonitrile. However, as we try to match elution strength we see a switch in elution order. In this case, the choice of organic modifier affects the interactions between the phase and the analytes, causing a change in selectivity.
Figure 6: The effect of organic modifier choice and organic modifier percent composition on selectivity for two pain-panel compounds.
| Peaks | Conc. (µg/mL) | |
|---|---|---|
| 1. | Dihydrocodeine | 0.5 |
| 2. | Noroxycodone | 0.5 |
| Column | Raptor Biphenyl (cat.# 9309A52) |
|---|---|
| Dimensions: | 50 mm x 2.1 mm ID |
| Particle Size: | 2.7 µm |
| Pore Size: | 90 Å |
| Temp.: | Ambient |
| Standard/Sample | |
| Inj. Vol.: | 10 µL |
| Mobile Phase | |
| Flow: | 0.8 mL/min |
| Detector | PDA @ 254 nm |
|---|---|
| Instrument | Waters ACQUITY UPLC H-Class |
| Notes | Mobile Phase Details Acetonitrile (top) Analysis run using 5, 10, 15, and 25% acetonitrile mobile phase. A: Water, 0.1% formic acid B: Acetonitrile, 0.1% formic acid Methanol (bottom) Analysis run using 15 and 25% methanol mobile phase. A: Water, 0.1% formic acid B: Methanol, 0.1% formic acid |
In the end, switching the organic modifier and expecting the same chromatographic performance is not always a reasonable expectation to have. In cases where the switch in organic modifier only affects the overall retention of your compounds of interest, there is a good chance that adjusting the organic modifier composition to match the elution strength of the original mobile phase will get you close to your original performance. But, if the relative retentions of compounds—the selectivity of the chromatographic system—show significant differences, especially in cases where peaks are shown to change elution order, you have an excellent indication that trying to simply match elution strengths will not produce similar chromatography. Sometimes, the organic modifier plays specific roles in specific circumstances, having a greater affinity for some classes of compounds because of a particular mode of interaction that another organic modifier does not possess (or possess to the same degree).
So, if you are interested in finding out if a switch in organic modifier is a viable option for you, first try keeping the percent composition the same before you do any method development. If you have similar selectivity, but with greater overall retention, you may be able to empirically determine a conversion between organic modifier percent composition and elution strength, make the adjustment, and get very close to your target.
If you see not only a change in overall retention due to elution strength but you also observe changes in selectivity, you can be sure that by changing the organic modifier you are also fundamentally changing the way your method is separating your compounds of interest.
Don’t shy away from trying to change organic mobile phase modifiers; if a change is what is right for your lab, just be aware of the effect of organic solvent on selectivity in LC and understand that the choice of organic solvent is a lever that can be used as you develop your methods.
References
- R.E. Majors, The continuing acetonitrile shortage: how to combat it or live with it, LCGC North America (2009) 27(6) 458-471. http://www.chromatographyonline.com/continuing-acetonitrile-shortage-how-combat-it-or-live-it?id=&pageID=1&sk=&date=