(W0930-02-08) Optimization of UV Detector Parameters of Compendial Methods

Purpose: Within the pharmaceutical industry, compendial LC methods are used to assess product safety and efficacy.  These methods often use UV detection for analysis. Aside from the detector type, the methods specify the UV wavelength(s) required for the analysis but do not provide any further information regarding other detector parameters. Due to this, default UV detector settings are frequently used. While these default settings might produce acceptable chromatography, for challenging methods modifications may be needed to meet the system suitability requirements defined in the method. An understanding of available UV detector parameters, including settable and default values is important to produce optimal chromatography. This study describes a process to evaluate and optimize the detector parameters used for a USP method.

Methods: The USP method for organic impurities in ibuprofen tablets was studied. The method’s system suitability requirements include a signal-to-noise ratio (s/n) of not less than (NLT) 10 for a 5-ppm solution of ibuprofen. The signal-to-noise ratio, a measure of the UV signal for the ibuprofen peak relative to the background noise, was used to define method sensitivity. The system suitability solution was initially analyzed using a PDA detector having all parameters (except for wavelength) set at default values. Individual detector parameters were changed and optimized based on their impact to the signal-to-noise ratio. The system suitability solution was re-analyzed using the optimized settings and the results compared against those obtained using default values.

Results: Initial testing of the 5-ppm solution of ibuprofen used a wavelength of 254 nm with all other detector parameters at their default values. The signal-to-noise value obtained under these conditions failed to meet the s/n system suitability requirement of not less than 10. Subsequently, the data rate, filter response, slit width, optical resolution and absorbance compensation were varied one factor at a time from their default values with a focus on the impact on the signal-to-noise ratio. The data rate defines the rate that the detector collects data. The optimum data rate was 2 Hz. At this data rate the ibuprofen peak was well defined with 31 points across the peak (within the desired range of 25 – 50 points) and less noise was observed than that seen from the default setting. The s/n was 25, which meets the system suitability criteria. With the data rate set at 2 Hz, the detector’s filter response was examined. This parameter controls a high frequency noise filter. Faster filter responses result in narrower peaks and remove less baseline noise while slower values produce broader peaks with decreased noise. The optimized filter response was 2 seconds. By changing the filter response from the default setting of 1 sec (Normal) to 2 sec (Slow) the signal-to-noise ratio increased from 25 to 33. The optical resolution, which determines the number of diode responses that are averaged in calculating absorbance at the specified wavelength, was also examined. Changing the optical resolution from its default value of 4 nm had little impact on the signal-to-noise ratio, so the default value was retained. The slit width setting controls the amount of light that reaches the photodiode array sensor. In general, larger slit widths produce less noise and increased sensitivity, and a reduction in spectral resolution. For this method, altering the slit with from the default (50 µm) did not significantly change the signal-to-noise ratio therefore, the default value was kept. Absorbance compensation was also evaluated for its impact on the signal-to-noise ratio. With this feature, non-wavelength dependent noise was reduced by collecting absorbance data over a wavelength range where there was no absorption and subtracting the average from the absorbance of the ibuprofen peak. With absorbance compensation enabled, the s/n increased from 33 to 51 - a 1.5 x increase in sensitivity. Finally, the optimized settings were combined and the signal-to-noise ratio re-measured. The result was compared against that obtained using default values. The signal-to-noise ratio increased from 7 to 51 which exceeds the requirement of NLT 10.

Conclusion: While default UV detector settings may be suitable for routine HPLC analysis, for challenging methods they may need to be adjusted to meet system suitability requirements This study utilized the USP method for organic impurities in ibuprofen tablets to show the impact of detector settings on method performance. When the method was run on a PDA detector using default detector settings, the system suitability requirement for the signal-to-noise ratio was not met. The data rate, filter response, slit width, optical resolution and absorbance compensation settings were examined and optimized. Using the optimized settings, the system suitability requirement was easily met, with an overall 7-fold increase in sensitivity over that obtained using the default values. The study demonstrated the importance of reviewing, understanding, and optimization of UV detector parameters when running challenging methods.

References: Monograph: USP. Ibuprofen Tablets. In USP-NF. Rockville, MD: USP; Dec 1,2016. DOI: https://doi.org/10.31003/USPNF_M39890_01_01

Impact of Data Rate

Comparison of results using default and optimized settings

USP s/n with default and optimized detector settings