(M1030-06-40) Optimizing Detector Settings for Method Scaling-Meeting System Suitability Criteria

Purpose: : For many LC applications, the detector sampling rate can impact signal-to-noise, a critical system suitability criteria. When there are a wide range of analytes at different concentrations it can be challenging to select the appropriate sampling rate. Using the default value for applications may not provide the results for optimum signal-to-noise, sensitivity and LOD. In addition, a minimum number of peaks are needed for quantification, having less than adequate points across a peak will lead to the loss of detail and reproducibility and too many points across a peak will introduce noise [1]. However, the impact of sampling rate would be different in HPLC and UPLC^TM system separations, given the peak width will vary with column and system dispersion. Thus, when scaling methods from HPLC to UPLC system independent analysis may be needed to identify the appropriate sampling rate.

Methods: The USP assay for Diclazuril USP monograph was analyzed on both HPLC and UPLC system. The method was scaled to the appropriate column keeping L/dp constant (-25% to 50%): HPLC - 4.6 × 100mm (3.5µm) and UPLC system - 2.1 × 50mm (1.8µm). Each column was tested with a sample containing both the API and organic impurities [3]. Geometric scaling was performed following USP 621 recommendations and included adjustment of flow rate, injection volume and gradient table to maintain constant column volumes [4].

Results: For signal-to-noise, the impact of sampling rate on chromatographic data varied across HPLC and UPLC system separations. A range of sampling rates were tested for both HPLC and UPLC system separations (Figure 1). For UPLC system separations, with very low peak widths and narrow system dispersion, the impact of sampling rate was more pronounced, with up to a 4-10x increase in sensitivity depending on the peak. In contrast, the HPLC separation was significantly less impacted by the sampling rate, with no significant change or improvement observed over the range of sampling rates (Table 1).

Conclusion: Scaling a method from HPLC to UPLC system provides a comparable separation with a shorter run time and lower solvent consumption; however, depending on the separation, other parameters may need to be adjusted including the detector sampling rate. The lower dispersion and smaller particle size column for the UPLC system separation produced a higher efficiency separation with narrower peaks. This results in a higher optimum sampling rate, or one that provides at least 30 points across the peak, for UPLC system as compared to an HPLC separation. In addition, very low sampling rates (eg. 1 Hz) may not provide sufficient points across the peak for UPLC system separations to quantify the analyte (Figure 1). For both HPLC and UPLC system, the signal-to-noise improves by increasing the sampling frequency to achieve the minimum points need across the peak, but too high of a sampling rate can increase noise thereby producing lower sensitivity. In this example, the UPLC system separation exhibits the highest signal-to-noise at 10 points per second, while the HPLC separation uses 5 points per second. In performing this exercise, the sampling rate is evaluated and selected for the narrowest peak, ensuring the minimum number of points across all peaks in the chromatogram while meeting signal-to-noise criteria.

References: [1] M.F. Wahab et al. / Analytica Chimica Acta 907 (2016) 31-44
[2] https://www.waters.com/webassets/cms/library/docs/720003869en.pdf
[3] https://online.uspnf.com/uspnf/document/1_GUID-766E7DF1-67F0-4AE7-B5C9-80E66660AC7D_2_en-US
[4] https://www.usp.org/sites/default/files/usp/document/harmonization/gen-chapter/harmonization-november-2021-m99380.pdf

Fig. 1 A. Comparison of the chromatograms obtained at sampling rates 1 pps (red) and sampling rate 20 pps (black) on the HPLC System. B. Comparison of the chromatograms obtained at sampling rates 1 pps (red) and sampling rate 20 pps (black) on the UPLC System.

Table 1 Comparing the signal-to-noise and points across peak for 6-Carboxylic acid and 4-amino derivative in the chromatogram on HPLC and UPLC system at different sampling rates. The results are an average of three replicates.