• Application Note

2996 Photodiode Array Detector: Optimization for Intelligent Speed (IS) Columns

2996 Photodiode Array Detector: Optimization for Intelligent Speed (IS) Columns

  • Waters Corporation

This is an Application Brief and does not contain a detailed Experimental section.

Abstract

This application brief focuses on optimizing the 2996 PDA detector settings for use with the IS columns.

Introduction

The Waters Intelligent Speed (IS) line of columns have a 20 mm packed bed and optimized hardware that allow for fast, reproducible separations which generate peaks that are often only 2 to 4 seconds wide. This technical note will focus on optimizing the 2996 PDA detector settings for use with the IS columns. The effects of sampling rate on reproducibility, peak shape, and resolution are examined along with the effects of digital filtering on signal to noise, resolution, and peak capacity. Additional information on Waters Instruments with IS columns can be found in technical note 720000723EN titled: Alliance 2695 Separations Module: Optimization and Performance with 4.6 mm i.d. Intelligent Speed (IS) Columns.

Experimental

System:

Waters Alliance 2695 Separations Module

Waters 2996 PDA Detector

Waters Empower Software

Column:

Xterra MS C18, 4.6 x 20 mm, IS 2.5 μm at 25 °C

Detection:

Wavelength range 195–285 nm

Extracted channel at 220 nm

Mobile phase:

A = 0.1 % TFA in water

B = acetonitrile

Flow rate:

3.0 mL/min

Gradient:

0 % to 50 % over 1 minute

Sample:

10 μL injection of beta blockers (0.1 μg/mL atenolol, 0.1 μg/mL metoprolol, 0.05 μg/mL pindolol)

Results and Discussion

Narrow Peaks require higher sampling rates

Figure 1 shows the same one minute gradient separation performed at different sampling rates. At sampling rates below 5 pts/s, there is a visible effect on peak shape. Table 1 lists the reproducibility of the peak area and retention times for 10 injections, along with the points across the peaks at the different sampling rates. The reproducibility of the peak area is above 0.5% RSD (Alliance System specification for injection precision) at sampling rates below 5 pts/s. In addition, when working with peaks which are only 3 seconds wide, the number of points across the peak fall below the minimum number needed for reproducible quantitation (15) at sampling rates below 5 pts/s.

Figure 1. Effect of decreasing the sampling rate.
Table 1. Reproducibility and points across the peak for metoprolol as a function of sampling rate.

Digital filtering affects resolution and peak capacity

Digital filtering is performed on data to improve the signal-to-noise ratios. The filter calculates a data point that is a modified rolling average for a wavelength over a number of readings. Figure 2 shows that as the filter response for the digital filter increases (more averaging) the peak width increases. At a filter response of 1.0, baseline resolution is lost. This increased peak width results in reduced peak capacity. Table 2 lists the peak capacities obtained at the different filter response levels along with the signal-to-noise ratios. The data show that there is a minimal increase in the signal to noise for a 2 fold reduction in resolution. At a filter response of 1.0 there is a significant increase in the signal to noise, however at the cost of a loss in resolution and diminished peak capacities.

Figure 2. Effect of increasing the filter response of the digital filter.
Table 2. Effect of filter response level on chromatographic parameters.

Conclusion

Slower sampling rates can dramatically affect peak shape, resulting in poor area and retention time reproducability. For most fast gradients 5-10 pts/s will be required. For very narrow peaks, such as those resulting from separations on the IS line of columns, use digital filtering with caution. Digital filtering does not need to be used except in separations where increased sensitivity is much more important than resolution and high peak capacity.

720000722, August 2003

Back To Top Back To Top