Diode Array Detectors
Photodiode Array
Advanced Detection Technologies for Compound Identification
1
Diode Array Detectors
Photodiode Array
PDA Optics Diagram
Lamp
Mirrors Lens
Flow Cell Slit
Grating
Diodes
Advanced Detection Technologies for Compound Identification
Waters 21,661
How the Diode Array Works The of light striking the diode is determined by its position relative to the stationary grating
Diodes discharged
Diodes recharging
Liquid from column
Dr. Shulamit Levin,
1-4
Diode Array Detectors
Extraction of 3D Data
The Data is 3D
nm
Abs
nm
Abs
nm
200 0.00 201 0.01 202 0.02 203 0.03 -
200 201 202 203 -
0.00 0.01 0.02 0.03
200 0.00 201 0.01 202 0.02 203 0.03 -
200 0.00 201 0.01 202 0.02 203 0.03 -
-
-
-
-
Abs
nm
Absorbance
Absorbance
Chromatogram
Abs
2
Spectrum Time
Wavelength
996 PDA Spectrum Index Plot
Coelution of 2 Peaks
DNPH Derivatives 0.25 ng Each Peak Millennium PDA Spectrum Index Plot - SampleWeight 0.25 ng - PDA 360.0 nm 440.00 420.00 400.00 380.00
440.00 420.00 400.00 380.00
360.00 340.00 320.00 300.00
360.00 nm 340.00 320.00 300.00
280.00 260.00
280.00 260.00
A
B
Coelution detection at a single wavelength Coelution is the sum of absorbance of 2 peaks A and B
AU
nm
1
0.0006 0.0006 0.0004 0.0004
AU
AU
0.0002
0.0002
Elution Time
0.0000
0.0000 4.00
Dr. Shulamit Levin,
6.00
8.00
10.00
12.00 Minutes
14.00
16.00
18.00
5-8
Diode Array Detectors
Spectrum Index Plot 300.00
Chromatographic Resolution & Coelution Detection
Spectra at apex and inflection points are displayed
tR(1)
t0 w1
nm
250.00
tR(2)
Rs = t R(2) - tR(1) 1/2 (w1 + w 2)
w2
Spectrumat maximum impurity is different
200.00 0.40
R=0
Maximum Impurity 0.20
R=0.3
R=0.7
R>1.0
AU
R=0 Purity Angle not effective; Match Angle useful R=0.3to R=0.7 Purity & Match Angle useful
0.00 2.20
2.40
2.60
2.80
3.00
R>0.7 Match Angle not useful
Minutes
Photodiode Array Technology Spectral Analyses Library Matching Compound identification Coelution detection Peak Purity Analysis Peak purity/peak homogeneity Coelution detection
Dr. Shulamit Levin,
Importance of Spectral Analyses Library Matching Compound identification Coelution detection Peak Purity Analysis Peak purity/peak homogeneity Coelution detection
9-12
Diode Array Detectors
Major Peak and Minor Peaks Analyte
PROBLEM: Locate all impurities
sin j =
Purity verification
AU
0.004 0.002
Bij2
Unknown
Absorbance
Impurities Resolved Coeluting
0.006
?
sj A)i 2
N
Library identification Standard
0.008
=
i =1
Absorbance
0.010
?i 1 (Bij N
ood quality spectral formation is important for:
0.000 Time
-0.002 0.00
1.00
2.00
3.00
4.00
5.00
6.00
Peak Purity analyzes all spectra (minimum 15) within a peak Apex spectrum is the reference spectrum
7.00
Minutes
240.00
280.00
nm
Dr. Shulamit Levin,
320.00
53 degrees is a large spectral difference
Theophylline Dyphylline
Similar spectra for structurally related compounds
Absorbance
Absorbance
200.00
Time
Spectral Contrast 10 Degrees
Spectral Contrast 53 Degrees Ethylparaben EthylPaba
Time
Matching compares the unknown apex spectrum of the peak with a reference spectrum in a library
230.00
250.00
270.00
290.00
310.00
nm
13-16
Diode Array Detectors
Spectral Contrast 0.5 Degrees
Very Similar Spectra Very similar spectra, CH2 difference
Methylparaben Ethylparaben
Analyte and 2 impurities
Absorbance
Spectral Contrast can differentiate these spectra
200.00
240.00
280.00
320.00
210.00
nm
230.00
250.00
270.00
290.00
nm
Spectral Comparison Spectrum A (apex)
Spectra from 200 to 300 nm
The shapes of Spectrum Aand
Peak Purity Plot
Choose one reference spectrum,
Spectrum Bare represented by vectors is the Spectral Contrast Angle, the difference between
compare to it all other spectraon thepeak. Thecomparisonisdone by linear regretion to a straight line.
Another option
spectral shapes
260
60.00
Purity Angle at every data point
50.00 40.00
250
Spectra B(i)
240
Increases in Purity Angle are spectral differences
30.00
230 220
sin
20.00
210 200
Impurity
10.00
N
? sin
j
=
( B ij s j Ai)
i= 1
0.00 2.40
N
?
2
B ij2
2.50
2.60
Minutes
2.70
Absorbance Purity Angle Noise Angle Baseline
i= 1
Dr. Shulamit Levin,
17-20
Diode Array Detectors
Purity Plot Chemically Pure Compound
Interpretation of Peak Purity Plots
5
Degrees
Peak Purity Plots can indicate Peak homogeneity Spectral homogeneity Coeluting impurities Spectral differences due to artifacts
Absorbance Noise Angle Purity Angle Baseline
0
Minutes
Purity Angle less than Noise Angle, ideal situation
Purity Plot: Mixture of 2 Compounds
Purity Plot: Mixture of 2 Compounds
Absorbance Noise Angle Purity Angle Baseline
10 Impurity
Impurity 10
Absorbance Noise Angle Purity Angle Baseline
Degrees
Degrees
20
0
0
Minutes
Purity Angle is greater than Noise Angle coelution on the front of the peak Dr. Shulamit Levin,
Minutes
Purity Angle is greater than Noise Angle - coelution near the peak apex
21-24
Diode Array Detectors
Purity Plot Chemically Pure Compound 2
Absorbance Noise Angle Purity Angle
Pure Benzoic Acid
2.0
Baseline
Peak Apex Non-linear
Different benzoic acid concentrations
1.5
Absorbance
Degrees
Effect of Concentration on Spectra
1
1.0
Spectral shape changes
0.5
0 0
Minutes
190.00
210.00
230.00
Purity Angle greater than Noise Angle Absorbance out of linear range at some wavelengths
Compound Confirmation
Fail
Fail
Fail
290.00
Chromatographic and Spectral Sensitivity: large bandwidth low resolution
Fail
Pass
270.00
Conflicts in Instrument design
Peak Homogeneity Pass
250.00
nm
Pass
Library Match Fail
Dr. Shulamit Levin,
Linearity: narrow bandwidth
Spectral performances: high resolution
25-28
Diode Array Detectors
Optical vs. Diode Resolution
Optical vs. Diode (Numeric) Resolution
IDEAL High optical resolution is 1 nm Light 1 nm
1 nm
Slit
Optical resolution affects linearity
Slit
1.2 nm Diode Resolution
Diodes
3 nm Optical Resolution
Diodes
996 Spectral Performance 1.2 nm Bandwidth
1 nm Diode
1 nm light beams
No bunching
Brand X Spectral Performance
Slit width, # diodes, and WL range (hardware) determine optical resolution: 1.2 nm
1nm slit, 2 nm 'bunch' nm ptical Resolution m light on >1 diode
0.7 nm diode resolution
3 diodes bunching
1 nm
1 nm ight beam
nm/diode (hardware) determine diode resolution, 1.2 nm
Slit 50 M allowing 1.2 nm bandwidith
Diodes (800-190) / 512 or1.2 nm per diode
Dr. Shulamit Levin,
Diode resolution (numeric resolution) is equal to wavelength coverage divided by diode number Hardware determines optical resolution,3 nm Optical resolution, determines the quality of spectra
Overall: 1.2 nm Spectral Performance
Diode 'bunching' (software) determines the overall spectral performance, 1.2 nm
light beam
nm/diode(hardware) determine diode resolution, 0.7 nm
?? µm slit, allowing1nm bandwidth
Digital resolution, ncorrectly called "Spectral Resolution"
Slit width, # diodes, and WL range (hardware) determine optical resolution: 1 nm
1024 diodes covering 190-950nm 0.7 nm/diode
Overall: 2 nm Spectral Performance
Diode 'bunching' (software) determinesthe spectral performance, 2 nm
29-32
Diode Array Detectors
Photodiode Array Technology
Importance of Detector Linearity
Optical Performance
Quantitation Major peaks Minor peaks
Linearity Optical Resolution Sensitivity
Spectral Analyses Library Matching Peak Purity/Peak Homogeneity
Effects of Optical Resolution on Linearity
Optical vs. Diode Resolution
3 nm Optical Resolution
1 nm per diode is 1 nm diode resolution
3 3
1.3 nm resolution is more linear than 5 or 10 nm
2 2 AU
1 nm Diode
ight nm
Slit width determines optical resolution, 3 nm
1
1.3 nm 5 nm 10 nm
1 0
Slit
Diodes
0
20 10
Dr. Shulamit Levin,
40 30 Concentration
60
Wide bandwidth is non-linear
50
33-36
Diode Array Detectors
Benzoic Acid Spectrum
Technical approaches to gain in linearity
1.60 1.40 1.20
Increase optical resolution
AU
214 nm is on a spectral slope
227.4 nm
1.00 0.80 0.60
One forbidden technical approach: using a prism in place of a grating, since prisms are non linear
0.40 0.20 0.00
214 nm
220.00 240.00 260.00 280.00
Linearity requires good optical resolution
nm
Linearity 214 nm Benzoic acid
Other Causes of Non-Linearity
3
Linearity greater than 2 AU
2
AU
2 1
Second order effects Stray light Chemical interactions
1.2 nm resolution
1 0 0
20 10
40 30
60 50
80 70
100 90
Concentration
Dr. Shulamit Levin,
37-40
Diode Array Detectors
Photodiode Array Technology Optical Performance
Resolution Resolution can be improved by: 1) using a small slit
Linearity Optical Resolution Sensitivity
2) selecting a narrow bandwidth 3) Reducing the wavelength covering (nm/diode) 4) Enlarging the number of diodes Overall quality of optics design and manufacturing is a crucial factor
Resolution
Importance of Optical Resolution
Drawbacks: 1) Small slit: less energy means more noise 1) Reduce the wavelength range: lack of information in the visible 2) More diodes: smaller diodes means noisier signal (less energy on each diode) Quality of optics design and manufacturing: means important R&D plus QC efforts from the supplier
Dr. Shulamit Levin,
Differentiation of Spectral Differences Similar spectra Spectral fine structure Spectral Analyses Library matching Peak purity / peak homogeneity
41-44
Diode Array Detectors
Common Perceptions
Factors Affecting Spectral Resolution
Most UV spectra have very broad spectral peaks
Optical resolution Diode or digital resolution
Good optical resolution is only required when there is spectral fine structure
Spectral Resolution - 1.2 nm vs. 3.6 nm Analyte and one impurity spectra from 245 to 275 nm
Absorbance
1.2 nm resolution
Benzene spectra Absorbance
Spectral Fine Structure
1.2 nm
Slit width and bandwidth
Less resolution at 3.6 nm vs. 1.2 nm UV maxima shifted
246.00
254.00
262.00
nm
Dr. Shulamit Levin,
270.00
230.00
250.00
270.00
nm
45-48
Diode Array Detectors
UV spectra of Triazines with the Waters 996 detector C
C N 2H5 NH
CH 3
N
C
C
NH
Simazine
C2H5
N
CH 3
N
N CH NH C
CH 3 CH CH 3
C NH N
Propazine
Features and Advantages Optical Resolution
C CH3 CH3
CH NH
N
N C
C
NH
C2H5
Atrazine
FEATURES Less than 2 nm optical and dioderesolution
N
221.6 nm
220.5 nm 221.6 nm 221.6 nm
221.6 nm 220.5 nm
260.6 nm 261.8 nm
ADVANTAGES Differentiation of similar spectra Visualizing spectral fine structure Linearity 190 to 800 nm to 2 AU
260.6 nm
220.00
240.00
260.00
280.00
300.00
216.00
220.00
224.00
228.00
nm
nm
Benefits of Good Optical Resolution
Photodiode Array Technology Optical Performance
Peak confirmation Confidence in compound identification Confidence in peak homogeneity with good peak purity analysis Good detector linearity Quantitation at high and low concentrations Spectral analyses Identification of major and minor compounds
Dr. Shulamit Levin,
Linearity Optical Resolution Sensitivity
49-52
Diode Array Detectors
Signal-to-Noise Ratio
Importance of Sensitivity Detection of Low Concentrations of Analytes Detection of impurities, metabolites, by-products and degradation products Quantitation
Signal-to-noise (S/N) is peak height to noise
6:1
3:1
8:1
Detection of Spectra at Low Concentrations Peak identification Peak purity / peak homogeneity
Chromatographic Sensitivity
Perceptions
Signal-to-Noise Ratio
Photodiode array detectors (PDA) are much less sensitive than variable wavelength detectors 0.001 AU
0.2 AU
PDA detectors are noisy
No apparent noise 2.8
3.0
3.2
Minutes
Dr. Shulamit Levin,
3.4
Noise
2.00
3.00
PDA detectors can not be used for quantitation of minor peaks
4.00
Minutes
53-56
Diode Array Detectors
Technological Advances in PDA Detectors
Waters 996 Chromatographic Sensitivity
New designs to improve signal-to- noise performance
0.0003
Waters 996 PDA Peak = 0.0001 AU
0.0002
Increased chromatographic sensitivity
AU
Waters 486 tunable UV Peak = 0.0001 AU
0.0001
Increased spectral sensitivity 0.0000
Enhanced software for improved performance 1.5
2.0
2.5
3.0
3.5
4.0
Minutes
High Sensitivity Chromatogram
Chromatographic Sensitivity Triazine herbicides at detection limit 0.0010
0.00006
0.0008
Simazine
0.00004
0.0006
0.00002
Peak height = 0.00007 AU 257 nm 1 sec filter
0.00000
AU
-0.00002 -0.00004 -0.00006 -0.00008 1.60
2.00
2.40
2.80
Minutes
Dr. Shulamit Levin,
3.20
3.60
Desethylatrazine
0.0004 0.0002
10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00 Minutes
Conditions: Gradient: Phosphate-Acetonitrile Column Novapak2x 300mm Sample: 2 ppb each pesticide Injection: 150 µl (0.3 ng on column)
PDA Resolution: 1.2 nm, Acquisition: 200 to 350 nm, 2 spectra per second. Chromatogram extracted at 220 nm No smoothing or bunching
57-60
Diode Array Detectors
High Sensitivity Spectrum
UV spectrum for Simazine at 2 ppb 0.00024
220.5 nm
0.00020
Absorbance
AU 0.00010
268.9 nm
0.00007 AU 0.530 AU
339.0 nm
306.8 nm
0.00000 -0.00002
210.00 220.00
240.00
260.00
280.00 nm
300.00
320.00
340.00
230.00
250.00
270.00
290.00
nm
Millennium PDA Spectrum Review Plot - SampleName Mel2ppb, Vial 44, Inj. 1
Increase Signal-to-Noise Ratio Signal-to-noise (S/N) is peak height to noise
6:1
3:1
8:1
Increase S/N by increasing peak height
Factors Increasing Signal Increase sample concentration Increase injection volume Wavelength Low volume flow cell
Increase S/N by decreasing noise
Dr. Shulamit Levin,
61-64
Diode Array Detectors
PDA Optics Diagram
Lamp
Mirrors Lens
Flow Cell Slit
Grating
Factors Affecting Noise Optics bench design Lamp energy Wavelengths Mobile phase Resolution Filter
Diodes
Each component in the optics path will affect noise
Technical approaches to gain in chromatographic sensitivity
AU 0.10
0.00 5.00
Sophisticated approaches: optimize the optics design: minimum dispersion, good focus of light on the diodes lamp optimization software (eliminate the need for different slits)
Dr. Shulamit Levin,
PROBLEM 4 Compounds 3 Peaks
7.247
0.20
0.730
enlarge slit width (decrease resolution) change optical resolution (affects spectrum) diode bunching (affects spectrum) noise smoothing (affects peak shape and height) reference wavelength substraction (loss of information in the subtracted band)
1.397
Traditional approaches:
Method Development #1
#
RetTi me (min)
Area (uV*sec)
Minutes
10.00
Match SpectrumNam e
Matc h Angl e
Matc h Thre sh.
1
0.730
651471
PeakA
0.096
1.163
2
1.397
655846
PeakB
0.071
1.284
3
7.247
1019807
PeakC
0.883
1.640
65-68
Diode Array Detectors
Peak tracking
Standards run only once for library
0.00
10.00
5.00
D
AU 0.10
0.00
Mobile phase changed to shorten run time
0.743
0.20
5.627 5.910
9.023
9.773
AU 0.10
Mobile phase changed 4 peaks identified
D
1.590
0.20
Method Development #3 1.277
0.740
Method Development #2
Minutes
2.00
4.00
6.00
8.00
Minutes #
RetTi me (min)
Area (uV*sec)
Match SpectrumName
Matc h Angl e
Matc h Thre sh.
1
0.740
660273
PeakA
0.042
1.203
2
1.590
666849
PeakB
0.026
1.347
3
9.023
560464
PeakC
0.079
1.839
4
9.773
434562
PeakD
0.516
2.747
# 1 2 3 4
Ret Time (min)
Area (uV*sec)
0.743 1.277 5.627 5.910
Match Spectrum Name
652303 654077 366935
PeakA PeakB PeakD PeakC
682223
Match Angle
Match Thresh.
0.139 0.125 1.455
1.161 1.274 2.366
0.369
1.649
Stability Test at 12 Weeks
Stability Test at 8 Weeks
Impurity 1
Impurity 2 Analyte
4.416 0.085
*
0.05
0.03
0.01 0.00 2.0
4.0
6.0
8.0
10.0
0.03 ANALYTE
Unknown
0.02
Impurity2
AU
0.04
12.0
Unknown
210
230
250
270
290
nm Unknown
AU 0.02
14.0
0.01
Minutes
Minutes
ANALYTE
1.021 6.424
Impurity2
Impurity 1 Unknown
Absorbance
Purity Angle Impurity 1
0.04
Impurity 1
0.05
Peak Flag
0.00 2.00
Dr. Shulamit Levin,
4.00
6.00
8.00
10.00
12.00
14.00
69-72
Diode Array Detectors
Stability Data at 12 Weeks Ret #
Match
Time
Area
Spectrum
Match
(Min)
%
Name
Angle
Flag
0.792
*
1
1.777
4.95
2
1.960
0.27
3
2.743
028
4
3.143
0.06
5
9.927
0.93
6
Peak Purity Using Photodiode Array Detection
11.193 93.50
Impurity 1
Purity Angle
Flag
33.261
Spectral homogeneity or peak homogeneity
* NEVER Chemical Purity Impurity has absorbance Impurity is present in high enough concentration Impurity is spectrally different from the analyte
6.461 Impurity 2
6.542
12.879 25.868
Analyte
0.154
*
0.092
Positive Compound Identification and Monitoring Integrated PDA with Mass Detector enables: PDA peak purity to investigate peak homogeneity UV and mass spectral information to be used from the same run for positive compound identification UV monitoring of separation for diagnostic purposes and quantitation
Dr. Shulamit Levin,
73-76