(T1430-02-09) Cracking the Chromatographic Code: Innovative Solutions to Address Keto-Enol Tautomerism in LC Method Development

Purpose: Keto-enol tautomerism, the conversion between a ketone and its enol isomer, can occur in analytes with a carbonyl group and an adjacent α-hydrogen, causing interconversion in solution and distorted peaks during chromatography, complicating quantitation. During the process development of a new drug substance, it was found that a starting material underwent on-column keto-enol tautomerization during High-performance liquid chromatography (HPLC) analysis of an in-process control (IPC) sample, causing significant peak tailing and splitting. This abstract describes the investigation into on-column tautomerization, the method development to mitigate it, and the optimization to ensure robust final chromatographic conditions.

Methods: Initial IPC

Method: The initial IPC utilized a Waters XBridge C18, 4.6 x 150 mm, 3.5 μm column at 50 °C. The mobile phase (MP) flow rate was 1.2 mL/min, injection volume was 5 µL, detection wavelength was 222 nm, and sample temperature was 5 °C. The MP composition was A: 0.05 % (v/v) Trifluoroacetic acid (TFA) in water and B: 0.05 % (v/v) TFA in Acetonitrile. Achiral Column and Mobile Phase Screen: Column screening was performed using a UPLC with 15 UPLC columns and acidic, neutral, and basic MPs. All columns were screened with acidic (0.05% TFA) and neutral (10 mM Ammonium Acetate) MP pairs, while only select columns were screened with the basic MP pair (0.05% Ammonium Hydroxide). The screening was conducted at 40 °C with flow rate of 0.35 mL/min, injection volume of 1 µL, detection wavelength 220 nm, and sample temperature of 10 °C. Varying Acid Concentration in the Mobile Phase: An HPLC with a quaternary pump was used with the following MPs: A: water/acetonitrile (95/5), B: water/acetonitrile (5/95), C: 0.4% TFA in water/acetonitrile (95/5), and D: 0.4% TFA in water/acetonitrile (5/95). The experiment utilized a Waters XBridge BEH C8 Column (130Å, 3.5 µm, 4.6 mm x 50 mm) at 25 °C. The flow rate was 2.5 mL/min, injection volume 2 µL, detection wavelength 222 nm, and sample temperature 5 °C. The experiment was conducted by varying amounts of acidic modifiers to achieve the desired acid concentrations. DFT Calculations: All DFT geometry optimization and frequency calculations were conducted with Gaussian 16, revision C.01 (Reference 1). Geometry optimizations were performed in the gas phase with the B3LYP functional and 6-31g* basis sets. All stationary points were characterized by frequency calculations to confirm local minima for ground state optimizations (zero imaginary frequencies) and all energy values are reported in kcal/mol and are the sum of electronic and thermal free energies.

Results: Initial IPC

Method: Analysis of an IPC sample showed significant peak broadening and splitting (Figure 1a), likely due to keto-enol tautomerism (Figure 1b). Optimization of flow rate, column temperature, injection volume, and sample composition had minimal effect. Therefore, a comprehensive screening of columns and mobile phases was conducted. Achiral Column and Mobile Phase Screen: The data indicated evidence of tautomerization with most mobile phases (Table 1). Neutral and basic mobile phases were avoided due to peak broadening. Therefore, a Waters Acquity BEH C8 stationary phase and acidic mobile phase conditions were selected for further method development. Varying Acid Concentration in the Mobile Phase: The results of increasing acid concentration in the mobile phase are shown in Figure 2a. It was found that by increasing the acid concentration, the keto-enol tautomerization could be fully inhibited during the analysis. It was decided to go with 0.4% TFA additive in the final MP composition. As a proof of concept an IPC sample was analyzed using the final method conditions with 0.4% TFA additive in the MP, it was found that keto-enol tautomerization was fully inhibited during the analysis, resulting in the intermediate peak appearing as a single sharp peak (Figure 2b). DFT Calculations: To understand the effect of acid on keto-enol tautomerization, DFT geometry optimization and frequency calculations were conducted using Gaussian 16. The results, shown in Figure 2c, indicated that the addition of acid helped stabilize the keto form over the enol form. The calculations revealed that the keto form was 9.1 kcal/mol lower in energy due to the stabilization of a more favored chair structure, compared to the distorted structure of the enol form.

Conclusion: Our results demonstrate that acid concentration in the mobile phase significantly impacts keto-enol tautomerization, with higher acid concentrations favoring the keto over the enol tautomer. Our DFT calculations revealed that higher acid concentrations stabilize the keto tautomer due to the stabilization of a more favored chair structure, compared to the distorted structure of the enol. This study underscores the necessity of considering tautomerization in HPLC method development and provides a framework for predicting and an option for controlling tautomeric behavior.


Figure 1: a. Initial IPC Sample Chromatogram, b. Keto-enol Tautomers


Table 1: Achiral HPLC column and mobile phase screening results were categorized by tautomerization. For column screening, 2.1×100 mm column dimensions were used with sub-2 µm particles if available.


Figure 2: a. The Effect of Acid Concentration on the Keto-enol Tautomerization, b. IPC Sample Chromatogram Using the Final Method Conditions, c. DFT Geometry Optimization Results for the Keto-enol Tautomerization