HPLC starting method for an acidic compound: which gradient, and why TFA
Author: Dr. Oliver Kraft, Dr. rer. nat., Organische Chemie · Updated 2026-08-03
Short and direct
For ibuprofen the tool proposes a C18 column, 150 by 4.6 millimetres, eluent A water with 0.1 percent trifluoroacetic acid, eluent B acetonitrile, a gradient from 20 to 80 percent B over twelve minutes with a hold, 1.0 millilitre per minute flow and a predicted retention time of 5.0 minutes.
That is a starting point for the first injection, not a method. The acid additive is the actual technical statement: ibuprofen is a carboxylic acid, and without suppressing dissociation the compound runs as a mixture of charged and uncharged form, which produces split peaks and shifting retention times. The rest of the method is convention.
The computed result
| Parameter | Proposal |
|---|---|
| Saeule | C18 ODS 150 x 4,6 mm (Bewertung 0,90) |
| Eluent A | Wasser + 0,1 % Trifluoressigsaeure |
| Eluent B | Acetonitril |
| Gradient | 0 min 20 % B, 2 min 30 % B, 12 min 80 % B, 15 min 80 % B |
| Fluss | 1,0 mL/min |
| Vorhergesagte Retentionszeit | 5,0 min |
Computed 2026-08-03 with chromaopt_suggest_method and chromaopt_column_select, SMILES CC(C)Cc1ccc(cc1)C(C)C(=O)O. Verbatim tool output.
What in the proposal is negotiable?
Almost everything except the acid additive. The column dimension follows the available equipment. Acetonitrile can be replaced by methanol, which changes selectivity and backpressure and is cheaper to buy. The gradient is deliberately broad so the first injection shows where the compound elutes at all, and is narrowed afterwards.
Not negotiable is whether the method later goes into a mass spectrometric coupling. Trifluoroacetic acid suppresses the signal in electrospray ionisation considerably. If that is the plan, replace it with formic acid from the start rather than validating the method twice.
How reliable is the predicted retention time?
As an order of magnitude, not as a number. A predicted retention time depends on dead volume, column age, temperature and the specific batch of stationary phase, and the model knows none of these. The value lies in knowing whether a compound comes after one minute or after twenty, and planning the run accordingly.
Frequent questions
- Why 0.1 percent acid and not more?
- 0.1 percent is enough to push the pH far enough below the pKa of a typical carboxylic acid. More acid does not improve peak shape but stresses the column and the instrument and raises the baseline in the UV region below 220 nanometres.
- Acetonitrile or methanol?
- Acetonitrile has lower viscosity, so less backpressure, and better UV transmission at low wavelengths. Methanol is cheaper and often gives different selectivity, which is the fastest lever when two peaks overlap. Both are acceptable; the choice belongs in the documentation.
- Is this suitable for a purity method under ICH Q2?
- As a starting point yes, as a method no. A purity method must show specificity against the known degradation products, and that needs stressed samples and peak purity evidence. The proposal says nothing about whether a degradation product hides under the main peak.
- What changes for a basic compound?
- The direction of the pH. For a base you either work well below the pKa so the compound stays fully protonated, or well above it with a column rated for that range. The region in between is where peaks broaden.
Sources
- ICH Q2(R2): Validation of analytical procedures. Frames what a method proposal is not yet.
The same calculation on your structures
The results above come from the same tools your language model can call over the Model Context Protocol. Test them free of charge, at your own pace.