Title: New Raman Instrumentation and Applications for QAQC and PAT
1New Raman Instrumentation and Applications for
QA/QC and PAT
2History of Raman Spectroscopy
- In 1928, Sir C.V. Raman discovered that radiation
scattered by materials had an inelastic
component. - Only 1 in 1,000,000 (0.0001) photons are
scattered inelastically. - By the 1930s, Raman had become the principal
means of non-destructive chemical analysis. - After WWII, advances in electronics and detectors
allowed IR to surpass Raman in this role. - Advent of lasers in the 1960s revived interest in
Raman to some extent.
3History of Raman Spectroscopy
- In 1986, the first commercial FT-Raman instrument
became available. - In the late 1980s, notch filters improved the
viability of Raman spectra. - 1980s and 1990s, dispersive Raman technology
improved including the development of holographic
instruments, refinement of triple spectrographs,
etc.
4History of Raman Spectroscopy
- In 1992, the first holographic transmission
grating is introduced. - In 1993, the first 532-nm Raman spectrometer
based on volume holographics is introduced. - First KOSI on-line system is introduced in 1994.
- In 1996, the first non-contact fiber optic probe
is introduced.
5What is the Raman Effect?
6What is Raman Spectroscopy?
Raman is a scattering technique
LASER
Rayleigh scattering Elastic scatter
1400
1200
1000
Raman Stokes Anti-Stokes Inelastic
scatter
800
Raman Intensity
600
400
200
-400
-200
0
200
400
Raman Shift (cm-1)
7Polarizability...
...is a measure of the ease with which an an
electron cloud can be induced to slosh up and
down a molecule under the influence of an applied
field
Therefore, application of an Electric Field can
induce a Dipole
P ? E
8Stokes Scattering
- Stokes scattering is, by convention,
positive-shifted Raman scatter. Most Analytical
work is done in this region. - Represents inelastic scattering to a region of
lower energy. This means that the energy of the
detected radiation is higher in wavelength
relative to the laser. - The scattered spectrum appears similar to an IR
spectrum and is interpreted similarly. - The Principle of Mutual Exclusion often makes
Raman spectrum complementary to the IR spectrum.
9Rayleigh and Anti-Stokes Scattering
- Rayleigh scatter is the elastic scatter, i.e.
same wavelength, as the incident laser. It is
filtered and is excluded from the spectrum. - Anti-Stokes scattering is inelastic scattering to
a region of higher energy. This means that the
energy of the detected radiation is lower in
wavelength relative to the laser (therefore,
higher frequency). By convention, it represents a
negative Raman shift. - The scattered spectrum is the mirror image to the
spectrum from the Stokes scatter but is less
intense. Therefore, only the more intense Raman
bands are seen in the Anti-Stokes spectrum.
10Scattering Intensities
- Scattering in general is dependent on the
frequency of the excitation radiation to the
fourth power - Rayleigh scattering is considerably stronger than
Raman scattering - Stokes and Anti-Stokes scattering are related to
the population in the ground state and the first
excited vibrational level - It may take as many as 108 excitation photons to
give rise to one Raman photon
11The Paradigm Factor
- The reason we dont use newer technologies is
that we fear that FDA wont accept them because
of the agencys unfamiliarity with them. But the
reason FDA is unfamiliar with newer technologies
is that companies dont submit themWe owe it to
our customers to use the best and fastest
techniques to analyze our products - Emil Ciurczak, Pharm Tech, Analytical
- Validation Issue 1999
12The Purpose of PAT
- To understand processes (manage variability)
- To improve processes (time, automation, etc.)
- To correct processes
- To identify where routine monitoring is needed
- Facilitate continuous processing
- Facilitate mechanistic-based regulatory
specifications - Real-time release (?)
- To save money by eliminating non-optimal
practices which lead, in worst-case scenarios, to
product rejection
13Why is Spectroscopy of interest for PAT?
- Spectroscopy can be used for real-time
quantification. - Spectroscopy can be used to understand the
chemistry of each process. - Spectroscopy can be used to optimize the timing
of the process stages and the endpoint. - Spectroscopy can be used for routine monitoring
by non-experts. - Spectroscopy can be used for remote monitoring.
- Safety advantages
- Also implies non-contact monitoring in the case
of Raman. Process is undisturbed which is
advantageous because of minimal impact on process
14Why is Raman being used for PAT?
improvements in instrumentation have allowed
users to focus on their applications rather than
on the operation and limitations of the
instrument
M. Pelletier, Quantitative Analysis Using Raman
Spectrometry, Appl. Spectrosc., 57 (1), 2003
15Why is Raman being used for PAT?
- Raman is a specific and selective technique
providing well resolved information leading
toBETTER PROCESS UNDERSTANDING - Flexible sampling (remote sampling)
- Confocal optics
- Sampling through containers
- Measurement of various types of samples (liquids,
slurries, pastes, solids, powders, etc.) - Ease of use no longer requires an expert
16PAT Uses of Raman
- To detect quantitative trends
- To understand the chemistry of the process
- Crystallinity
- Particle characteristics
- Reaction intermediates
- Reaction equilibria
- Optimization of timing
- To detect end-points, i.e., desired product status
17Fiber Optic Coupling Remote Process Monitoring
18Reaction Monitoring Hydrogenation of Cyclohexene
to Cyclohexane
H2
Process Understanding and Optimization
19Experimental Setup
Instrument Kaiser RamanRxn1 Analyzer Spectral
Range 100 to 3500 cm-1 Laser 785
nm Sampling Immersion probe Collection Time 10
x 6 seconds
20Spectra from 200 1800 cm-1
0 min
6000
17 min
28 min
39 min
59 min
5000
70 min
80 min
91 min
491 min
4000
1122 min
Raman Inensitty
3000
2000
1000
400
600
800
1000
1200
1400
1600
1800
Raman shift (cm-1)
21Expansion of CH2 Deformation Ring Mode Bands
0 min
17 min
5.0
28 min
Cyclohexene 823 cm-1
Cyclohexane 802 cm-1
39 min
59 min
70 min
4.0
80 min
91 min
491 min
1122 min
3.0
Raman Intensity
2.0
1.0
0.0
790
800
810
820
830
840
850
Raman shift (cm-1)
22CC Stretch Also Shows Appropriate Trend
0 min
3.5
17 min
28 min
39 min
59 min
3.3
70 min
80 min
91 min
3.1
491 min
1122 min
Raman Intensity
2.9
2.7
2.5
2.3
1635
1640
1645
1650
1655
1660
1665
1670
Raman shift (cm-1)
23Trend Using Baseline Integrated Bands in 800
850 cm-1 Region
1.2
Cyclohexane
1
0.8
0.6
Relative Intensity
0.4
0.2
Cyclohexene
0
-0.2
0
0.5
1
1.5
2
2.5
3
3.5
Time (hours)
24MCR (Multivariate Curve Resolution) Shows Some
Interesting Results
25Cyclohexene Pd/C
0.6
y -0.3102x 0.5026
0.5
2
R
0.9748
0.4
0.3
Cyclohexene Molarity
0.2
0.1
0
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Time (hrs)
26Cyclohexene Pt/C
0.5
y -0.9466x 0.4296
0.45
2
R
0.9865
0.4
0.35
0.3
0.25
Cyclohexene Molarity
0.2
0.15
0.1
0.05
0
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Time after 2nd H2 addition (hrs.)
27Cyclohexene Rh Profile
0.6
y -0.5043x 0.4965
0.5
2
R
0.995
0.4
0.3
Cyclohexene Molarity
0.2
0.1
0
0
0.2
0.4
0.6
0.8
1
1.2
Time (hrs.)
28Kinetics Data for Cyclohexene
Catalyst Zero-Order Constant Pd/C 0.310 Pt/C
0.947 Rh 0.504
Understand and Optimize the Process!
29Hydrogenation of Carvone
Process Understanding and Optimization
30Carvone Peak Disappearance
Complex Spectral Changes
31Kinetic Profile for Full Data Set
500
Dihydrocarvone
400
Carvone
300
Tetrahydrocarvone
200
arb. units
100
0
-100
0
2
4
6
8
10
12
14
16
18
Time (hours)
Note Analysis uses 452 cm-1 498 cm-1 region,
MCR 3 factor data reduction used for trend
plotting
32Catalytic Hydrogenation Reaction
- Intermediate (hydroxylamine) is a potential
thermal safety hazard - Preferred pathway excludes the intermediate
species
33Catalytic Hydrogenation Reaction
Reactant Intermediate Product
34Catalytic Hydrogenation Reaction
Reactant Intermediate Product
35Analytical Raman Spectroscopy
IlsLCI Il Raman intensity s Raman
cross section L Pathlength C Concentration I
Instrument parameters
Sample
36Univariate Quantitative Analysis
- Sometimes univariate techniques will do
- Peak heights or areas if isolated peaks are
available - Peak positions
- Peak ratios (internal standard or total area)
- These techniques better fit the mold for a
traditional validation approach
37Example of Peak Intensity - Clotrimazole Films
PEO Only
Clotrimazole E - 7.5
Clotrimazole A - 1
Clotrimazole F - 10
20000
Clotrimazole B - 2
Clotrimazole G - 12.5
Clotrimazole C - 4
Clotrimazole H - 15
Clotrimazole D - 5
Clotrimazole I - 20
Pure Clotrimazole
10000
2nd Derivative
0
-10000
2nd Derivative Spectra
-20000
1550
1560
1570
1580
1590
1600
Raman shift (cm-1)
38Calibration Results for Clotrimazole
r2 0.9969 SEC 0.493 SECV 0.722 MLR using
band at 1585 cm-1
39Raman Spectra of Progesterone
Crystal Forms I and II
- For this Study the CO Stretching Vibration was
used to Quantitate Form I and Form II Polymorphs.
Form I _at_ 1662 cm-1. Form II _at_ 1667 cm-1.
40Calibration Form I in Slurry
90
70
50
Form I Conc. (wt)
30
10
1661
1663
1665
1667
CO Peak Position (cm-1)
41Polymorphic Transformation at 45?C
Form I
Form II
_at_1662
_at_1667
- Crystallizations were monitored over the
temperature range from 5 to 45? C . - Slurry 2 grams Progesterone (25ml Organic Sol.)
added to 500ml H2O . - Temperature control and stirring were provided by
a LabMax automated lab reactor. - Polymorph concentration was determined from the
CO stretch band center position. - Raman measurements were made in-situ with the
RamanRxn1.
1680
1660
1670
Raman Shift (cm 1)
42Transformation Thermodynamics
Crystal Form II to I
43The Analysis of Extruded Films by Raman
Spectroscopy
Process Understanding, Optimization and Monitoring
44Hot-Melt Extrusions
45Raman Spectra for Extruded Films Containing
Clotrimazole
PEO Only
3.0
Clotrimazole A - 1
Clotrimazole E - 7.5
Clotrimazole I - 20
2.6
Pure Clotrimazole
2.2
1.8
Intensity (x10-5)
1.4
1.0
0.6
0.2
500
1000
1500
2000
2500
3000
Raman shift (cm-1)
462nd Derivative Spectra
10000
0
2nd Derivative
-5000
-15000
-25000
PEO Only
Clotrimazole A - 1
Clotrimazole E - 7.5
Clotrimazole I - 20
-35000
Pure Clotrimazole
800
1000
1200
1400
1600
Raman shift (cm-1)
472nd Derivative in Area of Interest
PEO Only
Clotrimazole E - 7.5
Clotrimazole A - 1
Clotrimazole F - 10
20000
Clotrimazole B - 2
Clotrimazole G - 12.5
Clotrimazole C - 4
Clotrimazole H - 15
Clotrimazole D - 5
Clotrimazole I - 20
Pure Clotrimazole
10000
2nd Derivative
0
-10000
-20000
1550
1560
1570
1580
1590
1600
Raman shift (cm-1)
48Calibration Results
r2 0.9969 SEC 0.493 SECV 0.722 MLR using
band at 1585 cm-1
49Ketoprofen Films
5.0
4.5
4.0
PEO
ketoD-5
ketoG-15
3.5
Pure ketoprofen
3.0
Raman Intensity (x10-5)
2.5
2.0
1.5
1.0
0.5
500
1000
1500
2000
2500
3000
Raman shift (cm-1)
502nd Derivative
0.8
0.4
0.0
-0.4
2nd Derivative
-0.8
PEO
-1.2
ketoD-5
ketoG-15
Pure ketoprofen
-1.6
-2.0
200
400
600
800
1000
1200
1400
1600
1800
Raman shift (cm-1)
512nd Derivative in Area of Interest
PEO
ketoD-5
ketoA-1
ketoE-7.5
0.8
ketoB-2
ketoF-10
ketoC-4
ketoG-15
Pure ketoprofen
0.4
2nd Derivative
0.0
-0.4
-0.8
985
990
995
1000
1005
1010
1015
1020
Raman shift (cm-1)
52Calibration Results
r2 0.9979 SEC 0.307 SECV 0.657 MLR using
band at 998/886 cm-1
53SAMPLING LARGER SPOT
54Representative Measurement - Area
PhAT System 6 mm Diameter
Traditional Dispersive and FT-Raman lt100 microns
55Relationship between MCC thickness and API
Intensity (675cm-1)
- API signal is observed when 1.88mm of MCC present
- No API signal is observed when 2.42mm of MCC
Present
56Quantification of Furosemide Degradation
57Furosemide Tablet Spectra
1.60
1.50
1.40
1.30
Intensity
Degradant
1.20
1.10
1.00
0.90
0.80
Degradant
6
10
200
400
600
800
1000
1200
1400
1600
1800
Raman shift (cm-1)
582nd Derivative Expansion
0.03
0.01
-0.01
Intensity
-0.03
-0.05
-0.07
-0.09
1540
1560
1580
1600
1620
1640
1660
Raman shift (cm-1)
59Averaged Data Five Replicates
Note Cross Validation suggests extremes (T0 and
T9) introduce non-linearity
60For Process Too!!!
61Nitrofurantoin Granulation- Small Spot Size
100
-6
y x - 410
2
0.9732
R
80
60
Observed
40
Linearity OK but not Outstanding
20
0
0
20
40
60
80
100
Predicted
62Nitrofurantoin Granulation- Large Spot Size
100
-6
y x - 210
2
R
0.9992
80
60
Observed
40
Much Improved Results Linearity excellent
20
0
0
20
40
60
80
100
Predicted
63Conclusions
- Raman is compatible with PAT goals
- Raman facilitates process understanding
- Raman facilitates process monitoring
- Raman can facilitate risk mitigation
- Raman can facilitate the development of
mechanistic-based (rational) regulatory
specifications
64Everybody Loves Raman!!
65Acknowledgements
- Pfizer
- Steve Arrivo
- Fred LaPlant
- Joep Timmermans
- Purdue University
- Lynne Taylor
- Solvias
- Arne Zillian
- Michigan State Univ.
- Kris Berglund
- FDA
- Robbe Lyon
- Ajaz Hussein
- Chris Ellison
- Everett Jefferson
- University of Mississippi
- Bonnie Avery
- Mitch Avery
- Mike Repka
- Venkat Tumuluri
- Suneela Prodduturi