GUIDANCE for Spectral Analysis using

Waters' Empower/Millennium

for Photo Diode Array (PDA) Detection in HPLC

Definitions:

• The spectral contrast angle measures the shape difference between two spectra. It is the basis for all spectral comparisons in the software:

Bij =Spectrum at each point in the peak Ai = Spectrum at apex s   = normalization factor 0 <= <= 1 0 <=  <= 90

• The Purity angle is a measure of the spectral heterogeniety of a peak based on the comparison of spectra over all the peak, using the spectral contrast angle
• The non-ideal effects are quantified and provided as a value of the Threshold angle.
• When the peak is pure, the Purity angle is lower than the Threshold angle.

GENERAL SPECTRAL ANALYSIS

(Not for validation of stability indicating chromatographic method!).

Instrumental Considerations:

1. Choose the spectral resolution preferred for the particular compounds. Preferably use 1.2 nm resolution for poor UV spectral shaped molecules. When there are limitations in memory, select 2.4 or 3.6 nm, not higher.
2. Choose the spectral range for the acquisition (nm).  Make sure that it includes at least 5 x wavelength resolution values below the assay's wavelength in the short wavelengths or 5x wavelength resolution values above the assay's wavelength in the longer wavelengths.
3. Check the mobile phase's UV cutoff wavelengths, and make sure to work above it.
4. Select the appropriate acquisition rate (Number of spectra per minute) to make sure that there will be at least 15 points across the peaks. Generally 1 spec/sec is sufficient, however, when fast chromatography is used 2 specs/sec might be preferable (note the double storage memory needed).

Experimental Considerations

1. If available, make sure to have a standard injected at concentrations that will give 0.5-1 AU in the MaxPlot chromatogram, make sure that the sample gives peaks at similar responses if available. The standards and the unknowns should be injected in the same session under the same conditions exactly, to be able to evaluate the values of Threshold angle and Purity Angle.
2. Make sure to include a blank run, to be able to subtract the 3D data in case of problematic results.

Software Considerations

1. Create a method set that uses Maxplot derived channel and apply it for all the channels under study.
2. Check carefully the entire chromatogram for quiet time interval in the chromatogram, minimum of 0.5 minutes long, to assign as the Noise Interval.
3. Create an appropriate PDA Processing Method that enables Peak Purity, set the appropriate wavelength range (cut the noisy low UV range or too quiet high UV range). Set the Noise Interval and the criteria for Threshold Determination. The Auto-threshold will be applicable to most cases.

Results can be observed in the following places:

• Peaks tab at the bottom table in the main window of the Review.
• Results window, the top table, use the Peak Purity tab. Also, select the Purity Plot at the bottom, to observe the results graphically.
• Using the Result view use the Preview tool, select an existing PDA report or create one with the report wizard, in which you can select the PDA Default Report.
 

SPECTRAL ANALYSIS

(FOR VALIDATION OF STABILITY INDICATING METHOD)

Instrumental Considerations:

1. Choose the spectral resolution preferred for the particular compounds. Preferably use 1.2 nm resolution for poor UV spectral shaped molecules. When there are limitations in memory, select 2.4 or 3.6 nm, not higher.
2. Choose the spectral range for the acquisition (nm). Make sure that it includes at least 5 x wavelength resolution values below the stability method's wavelength in the low wavelengths and 5x resolution values above the method wavelength in the longer wavelengths.
3. Check the solvent UV cutoff wavelenth and make sure to work above it.
4. Select the appropriate acquisition rate (Number of spectra per minute) to make sure that there will be at least 15 points across the peaks. Generally 1 spec/sec is sufficient, however, when fast chromatography is used 2 specs/sec might be preferable (note the double storage memory needed).

Experimental Considerations

1. Make sure to have 5-6 replicates of the standard for the Peak Purity analysis and replicates of the standard at 25%, 50%, 75%, 100% 120% or others of the assay's concentration for the spectral matching..
2. The unknowns are stressed samples of the main component's raw material as well as the formulation. It is recommended to inject duplicates. The samples stressed by heat, light, acid, base and oxidizing agent. Make sure to include their blank medium.
3. The standards and the unknowns should be injected in the same session with the unknowns, under the same conditions exactly, to be able to evaluate the values of Threshold angle from the Purity Angles of the standards.
4. Make sure to include a blank run (diluent, mobile phase and medium of the stress) as well as the placebo, to be able to subtract the 3D data in case of problematic results.

Software Considerations

1. Create a method set that uses Maxplot derived channel and apply it for all the channels under study.
2. Apply the method set on all the runs, superimpose the chromatograms in the Results view window, and check carefully the entire chromatogram for quiet baseline interval in all the chromatograms, minimum of 0.5 minutes long, to assign as the Noise Interval for all the runs.

Procedure: Peak Purity

Create an appropriate PDA Processing Method that enables Peak Purity, set the appropriate wavelength range (cut the noisy low UV range or too quiet high UV range). Set the Noise Interval and the criteria for Threshold Determination (Auto-threshold or solvent+ noise). The Auto-threshold might not be applicable to all cases (because it makes assumptions regarded to the signal). For method validation in a regulated environment it is recommended to use the manual sensitive Threshold angle determination:

1. Select the 6 standard replicates and process them, using the method set. Check the results for Purity Angle of all 6 runs. Select the maximum value of the Purity Angle and set it in the processing method under Solvent and Noise criteria for Threshold determination.

2. Process the 6 replicates again and verify that all standards came out pure (Purity Angle < Purity Threshold Angle).

3. Process the unknowns and use an appropriate PDA report to report their results. You can use the Purity Plots in the report. If the Purity angle is smaller than the Threshold angle, the peak is pure.
 
 

Procedure: Spectral Matching

1. Select one of each standard concentration (25%, 50%, 75% and 100%), and create a new spectral library. Make sure to indicate the standard source and level or concentration in its name for information.
2. Check the library name in the Processing Method at the PDA Spectral match window. Select in this window the depth of search (1-3) and the degree of match (in deg). Indicate if you are interested in RT presearch.
3. Select the second channel in the duplicates from each standard concentration and process it using the method set with the updated processing method. Check the Match angle of all 4 channels and select the maximum value from all match angles (3 results each channel), and use it in the Processing Method as the Threshold angle for the Spectral Matching.
4. Process the 4 Channels again, using the method set with the updated Processing Method. Verify that the results indicate that they all are identified appropriately (Match angle<Threshold angle).
5. Process all the unknowns using the same method set. When the Match angle is smaller than the Threshold angle the match is OK.

Results can be observed in the following places:

Peaks tab at the bottom table in the Main window of the Review.

Results window, the top table, use the Library match tab. Also, select the Match Plot at the bottom, to observe the results graphically.

Using the Result view use the Preview tool, select an existing PDA report or create one with the report wizard, in which you can select the PDA Default Report.
 
 

BEWARE OF THE FOLLOWING:

• The Noise's time interval must be peak-less in all the chromatograms.
• The noise's time interval in gradients must be near the studied peak.
• Non linearity of Beer's law will result in wrong results and unusual behavior.
• Integration events can effect the results, make sure to use them appropriately.

See also: 

Peak Purity I; Peak Purity II; Peak Purity III; Library Matching.

   Instrumental Deviation from Beer's Law 

(Based on Waters' help in empower)  

Back to "Supporting Waters' Customers in Israel"  Page

 

Copyright (C)1997 Dr. Shulamit Levin
ABOUT ME: PERSONAL PAGE