-
Page 285 of 305
CHAPTER 6
Method Development and validation for Polymorphic purity
determination of Sertraline Hydrochloride by NIR
Near-infrared spectroscopy
The IR region is divided into three regions: the near IR, mid
IR, and far IR. The mid
IR region is used mostly in chemical analysis. This is the
region of wavelengths
between 3 x 10-4
and 3 x 10-3
cm (figure 1)1. Practically it becomes easy to work with
numbers which are easy to write; therefore IR spectra are
sometimes reported in µm,
although another unit , (nu bar or wavenumber), is currently
preferred. A
wavenumber is the inverse of the wavelength in cm. The mid IR
range is 4000–400
cm–1
in wavenumbers. An increase in wavenumber corresponds to an
increase in
energy.
Figure 6.1: Near infrared region
William Herschel discovered the near-IR region2 in the 19th
century, but the first
industrial application began in the 1950s. In the first
applications, NIR spectroscopy
was used only as an add-on unit to other optical devices that
used other wavelengths
such as ultraviolet (UV), visible (Vis), or mid-infrared (MIR)
spectrometers.
-
Page 286 of 305
Near-infrared spectroscopy is a spectroscopic method which uses
the near-infrared
region of the electromagnetic spectrum. Near-infrared
spectroscopy is based on
molecular overtone and combination vibrations3. This is
energetic enough to excite
overtones and combinations of molecular vibrations to higher
energy levels. NIR
spectroscopy is typically used for quantitative measurement of
organic functional
groups, especially O-H, N-H, and C=O. Detection limits are
typically 0.1% and
applications include pharmaceutical, agricultural, polymer, and
clinical analysis. The
molecular overtone and combination bands seen in the near IR are
typically very
broad, leading to complex spectra; it can be difficult to assign
specific features to
specific chemical components. Multivariate (multiple variables)
calibration
techniques (e.g., principal components analysis, partial least
squares, or artificial
neural networks) are often employed to extract the desired
chemical information.
Careful development of a set of calibration samples and
application of multivariate
calibration techniques is essential for near-infrared analytical
methods. In contrast to
sharp absorption peaks in the MIR region, NIR spectra show less
intensity and broad
bands. An assignment of peaks to individual vibrations is thus
not possible.
Near-infrared spectroscopy (NIRS) is a fast and nondestructive
technique that
provides multi-constituent analysis in virtually any matrix. In
recent years, NIR
spectroscopy has gained wide acceptance within the
pharmaceutical industry for raw
material testing, product quality control and process
monitoring.
The growing pharmaceutical interest in NIR spectroscopy is
probably because it
meets the criteria of being accurate, reliable, rapid,
non-destructive, and inexpensive
and its advantages over other analytical techniques, namely, an
easy sample
preparation without any pretreatments, the possibility of
separating the sample
measurement position and spectrometer by use of fiber optic
probes, and the
prediction of chemical and physical sample parameters from one
single spectrum.
NIR absorption bands are typically broad, overlapping and 10–100
times weaker than
their corresponding fundamental mid-IR absorption bands. These
characteristics
severely restrict sensitivity in the classical spectroscopic
sense and call for
chemometric data processing to relate spectral information to
sample properties. The
low absorption coefficient, however, permits high penetration
depth and, thus, an
adjustment of sample thickness. This aspect is actually an
analytical advantage, since
-
Page 287 of 305
it allows direct analysis of strongly absorbing and even highly
scattering samples,
such as turbid liquids or solids in either transmittance or
reflectance mode without
further pretreatments.
The NIR spectrum of sucrose is given below as an example (
Figure 6.2).
Figure 6.2: Typical NIR spectrum
As seen from the above spectrum, unlike an IR spectrum, there
are no characteristic
peaks representing the chemical bonds or representing the
characteristic stretching
and bending thus making it difficult to interpret or draw any
conclusion. In order to
make it applicable, there needs to be a strong data collection
system coupled with a
software capable of multivariate analysis and chemometrics.
Instrumentation
NIR spectroscopy instrumentation is similar to those of other
spectrophotometers such
as IR or UV. This typically contains a source, a detector, and a
dispersive element
(such as a prism, or, more commonly, a diffraction grating) to
allow the intensity at
different wavelengths to be recorded. Fourier transform NIR
instruments using an
interferometer are also common, especially for wavelengths above
~1000 nm.
Depending on the sample, the spectrum can be measured in either
reflection or
transmission.
-
Page 288 of 305
The radiation is generated usually by means of a quartz halogen
or incandescent lamp
used as broadband sources of near-infrared radiation.
There are a variety of sample analysis units based on the type
of application. The most
important aspect of this analysis is that there is no need of
sample preparation, i.e., the
sample can be analysed as such.
Measurement
Some of the types of analysis types include diffuse reflectance,
transmission and
transflection. Based on the sample, the measurement mode can
either be absorbance
or reflectance. The sample for pharmaceutical application can
range from powders i.e.
APIs or Excepient or liquids, semi finished inprocess sample or
liquid samples or
finished products like tablets. Based on the type of sample, the
analysis mode can be
selected, for e.g. for analysis of powders or liquids, either
representative sample can
be analysed in a sample compartment or an optical fiber probe
can be directly placed
in contact with the sample for analysis. In order to analyse the
formulation e.g. tablets,
the same can be placed in tablet holders and analysed
directly.
Pretreatment of spectra
Because the NIR spectral bands are broad, data pretreatment is
typically necessary to
convert the raw data into useful spectral signature information.
This can be performed
in several ways. The most common type is smoothening and
derivative ( 1st order, 2
nd
order etc.)
Figure 6.3: Typical untreated NIR Spectrum
-
Page 289 of 305
The above figure is a representative NIR spectrum (Figure 6.3).
As we can see, the
spectrum does not let us understand any details nor can it be
used to differentiate
between similar spectra with different particle size or
polymorphs. This happens in
most of the cases, thus in order to make it more elaborative and
to bring out minute
details, the same needs to be treated, and the most common
treatment is its derivative.
Given below is a sample picture ( Figure 6.4) of how the same
would be changed if
derivitized to 1st order.
Figure 6.4: Example of Derivative NIR spectra
As we can see, the spectrum signals have become more predominant
and thus now the
spectrum is in a position to differentiate itself from materials
of similar chemical as
well as physical properties. There are other means of sample
treatment depending
upon the instrument software capabilities e.g. Bruker Mpa NIR
equipped with OPUS
software uses a calculation called a Conformity Index. In this
calculation, the spectra
of the pure compound is considered as a base and the
differentiation is calculated by a
specific number by virtue of its dissimilarity with that of the
base spectra by means of
assigning a number i.e. the CI number.
Chemometrics
Chemometrics is the science of extracting information from
chemical systems by
data-driven means. It is a highly interfacial discipline, using
methods frequently
employed in core data-analytic disciplines such as multivariate
statistics, applied
mathematics, and computer science, in order to address some
complex issues. Such
-
Page 290 of 305
multivariate data has traditionally been analyzed using one or
two variables at a time.
We must process all of the data simultaneously to understand the
relations.
In the context of near infrared spectra, several data points are
collected from the start
to the end and during the data processing, all the data points
are considered. Minor
changes in spectrum is thus differentiated as numbers and
interpreted as variations
both quantitatively as well as quantitatively.
Calibration curve
A calibration curve is a plot of increment of one variable
relatively with that of
another variable. In this scenario this relationship curve is
drawn between responses
with respect to concentration. A linear relationship is
considered when a small
increment of concentration reflects in the increment of
analytical signal of an
instrument i.e. response (absorbance or reflectance) obtained by
a spectrophotometer.
The data - the concentrations of the analyte and the instrument
response for each
standard - can be fit to a straight line, using linear
regression analysis. This yields a
model described by the equation y = mx + C, where y is the
instrument response, m
represents the sensitivity, and C is a constant that describes
the background. The
analyte concentration (x) of unknown samples may be calculated
from this equation.
Once the calibration curve is drawn, the concentration of an
unknown sample can be
identified based on the response obtained from it by the above
mentioned formula.
A calibration curve can also be drawn with concentration of an
impure compound by
means of conformity index (CI) values also.
Validation:
Once the calibration curve is obtained, the same can be
validated to ascertain the
reliability. Samples with known concentrations of samples are
prepared and the
concentration is calculated by using the calibration curve. The
true value is then
compared with that of the calculated value. The agreement or
disagreement of both
these values determines the validity of the calibration
curve.
-
Page 291 of 305
Sertraline Hydrochloride
Sertraline Hcl is chemically
(1S,4S)-4-(3,4-dichlorophenyl)-N-methyl-1,2,3,4-
tetrahydronaphthalen-1-amine. This belongs to a class of drugs
known as
antidepressants4. This can further be classified as selective
serotonin reuptake
inhibitor (SSRI) . It was discovered by Pfizer Pharmaceuticals.
The chemical details
are given below
CAS No. : 79559-97-0
Molecular Formula : C17H18Cl3N
Formula Weight : 342.69
Form : solid
Color : white crystalline powder
Sertraline is a widely used antidepressant belonging to the
selective serotonin
reuptake inhibitor class; its efficacy has been demonstrated not
only in the treatment
of major depression, obsessive compulsive and panic disorders,
but also for eating,
premenstrual dysphoric and post-traumatic stress disorders. The
antidepressant effect
of Sertraline is presumed to be linked to its ability to inhibit
the neuronal reuptake of
serotonin5. It has only very weak effects on norepinephrine and
dopamine neuronal
reuptake. At clinical doses, Sertraline blocks the uptake of
serotonin into human
platelets6. Like most clinically effective antidepressants,
Sertraline down regulates
brain norepinephrine and serotonin receptors in animals7.
In the solid state, Sertraline hydrochloride exists in various
crystalline forms having
different physical properties8.Various claims have been made
emphasizing the
differences in bioavailability between different polymorphic
forms of Sertraline HCl.
Two different polymorphic forms were selected for this
experimentation and have
been named as Form I and Form II for ease of differentiation
between the forms.
-
Page 292 of 305
Several works have been published on polymorphic purity
determinations by XRD9,
10. A few publications are also found for polymorphic purity
determinations by NIR
11.
However no reference is available to the best of our knowledge
which describes the
polymorphic determination of Sertraline HCl by NIR. This chapter
describes this
novel work to determine the polymorphic impurity in Sertraline
HCl.
Sertraline Polymorphic purity Method by NIR
Polymorphism is an important phenomenon in the drug development
and
manufacturing process since different polymorphs of compound
show variations in
physicochemical properties such as density, morphology,
solubility, dissolution rate,
stability, and hygroscopicity. As a result, different polymorphs
of the same drug
exhibit differences in bioavailability, efficacy, and drug
product performance. In order
to control polymorphism in the drug development and
manufacturing processes, it is
critical to identify, characterize, and quantitate the presence
of the various
polymorphs of a pharmaceutical compound
There are various techniques to identify the polymorphic form in
pharmaceutical drug
substances and drug products. The most common are the X-Ray
powder diffraction
and Differential scanning calorimetry. The other techniques such
as the Mid Infra Red
Spectroscopy, Near Infra Red Spectroscopy, Raman Spectroscopy
and the optical
Microscopy are also of great importance and have proved to be of
great importance in
this regard. The NIR and Raman spectroscopy have come in to
picture in the recent
times. NIR rays have an ability to penetrate much deeper in to
the compounds thus
can identify the physical variations in a compound. We have
tried to use NIR
spectroscopy (Bruker MPA) to Identify (By means of conformity
index and
quantitative estimation mode) and quantitatively estimate a
polymorphic impurity in
the other pure polymorphic form.
The sample analysis of Sertraline form I( Figure 5) and Form II
( Figure 6)were
analyzed by using Bruker NIR and spectra were generated by using
OPUS software.
The qualitative spectra are given below for reference
-
Page 293 of 305
Figure 6.5: NIR Spectrum of Sertraline Form I
Figure 6.6: NIR Spectrum of Sertraline Form II
For this experiment, Form II was considered as the API and Form
I was considered
as polymorphic impurity. The Aim of the experiment is to develop
a method to
determine the content of form I in form II by using NIR.
-
Page 294 of 305
Procedure:
To develop the method, along with pure samples of Form I and
Form II, solid spiking
of Form I in Form II were performed with serial dilution method
and scanned .By
this technique, individual samples were prepared containing
0.5%, 1.0%, 2.0%,
5.0%, 10.0% and 20% w/w of form I in Form II and analysed by
means of NIR (Solid
probe) . The NIR spectra were recorded from 4000cm-1
to 12500 cm-1
at 8.0 cm-1
resolution. All the spectra which were obtained in this manner
have been overlaid as a
single figure as given below for reference.
Figure 6.7: Overlaid NIR Spectrum of Sertraline Form I, II and
spiked samples
As seen from the above figure, there are no major differences.
On careful observation, however there
is slight variation in absorbance in the region between 8750 and
8800cm-1
(Figure 8).
-
Page 295 of 305
Figure 6.8: Overlaid NIR Spectrum of Sertraline Form I, II and
spiked samples
depicting the visible variation
Figure 6.9: Zoomed section of the variation in absorbance within
the sample
spectra
The selected elaboration in spectra gives a clear representation
of this difference.
spectra showed minute differences in certain regions.
-
Page 296 of 305
Conformity Index method
One of a special feature of the OPUS software is called as
conformity index. For this
type of analysis, pure Form II material was analysed six times
and individual baseline
numerical values were generated to it by carefully selecting the
regions of
differentiations in the spectra. In the similar fashion, spectra
were generated in
triplicate for each of the spiked materials. The software
generates average numerical
values to each of the levels. The conformity index values is
plotted graphically by the
software as shown below( Figure 10).
Figure 6.10: Conformity index graph for spiked and pure
sample
The spots on the conformity index plot have been indicated for
pure form I ( green
spots) and 0.5% spike (blue spots). Similarly the other sets of
the spots are
representations for 1.0%, 2.0%, 5.0%, 10.0% and 20%. The values
obtained for these
values have been tabulated below.
0.5% spike
Pure Form II
-
Page 297 of 305
Table 6.1: Table for conformity index values
S.No %spike CI value
1 0.5 4.49
2 0.5 4.56
3 0.5 5.37
4 1.0 6.01
5 1.0 6.66
6 1.0 6.86
7 2.0 12.09
8 2.0 11.89
9 2.0 13.3
10 5.0 28.52
11 5.0 28.45
12 5.0 27.92
13 10.0 64.55
14 10.0 65.31
15 10.0 63.00
16 20.0 124.24
17 20.0 129.52
18 20.0 127.16
Correl coef (r) 0.999062 ~ 0.9991
The correlation coefficient value thus obtained is 0.999(Table
1) thus shows that there
is a linear relationship between both the variables. Thus this
method of determining
the percentage of polymorphic impurity in form II is accurate
and can be used to
determining the polymorphic impurity.
Quantitative model
The spectra generated earlier were calculated using the software
and by means of
statistical analysis, numerous data points on the spectra are
utilized to define the
spectrum value. This type of analysis is called as Quant model.
A calibration curve
was drawn by the software with all the values against the true
spiking percentage
values of each spiked value(Figure 11). The calibration curve is
stored in the
software.
-
Page 298 of 305
Figure 6. 11: Quant Model development by means of calibration
curve.
A random mixture of Form I in Form II is prepared and analysed
by NIR. IN this case
the true value is determined by means of the software. The
closeness of the values is
used to validate the utility of the calibration curve.
The table presented below shows the true values i.e. intentional
amount of spiking
against the predicted values. As seen from this table, there is
a close agreement
between the true and that of the predictions. The correlation
coefficient obtained is
more than 0.9999, thus is an accurate method for determining the
polymorphic
impurity. A graph has also been plotted with the true and
predicted values. This data
has been tabulated containing both the pure forms I as well as
that of pure form II
(Table2). The r2 value of above 0.999 shows that the method can
be applied in the
total range and not restricted to the experimental
conditions.
-
Page 299 of 305
Table 6.2: Table for validation of the calibration curve
S.No Sample name TRUE Prediction Difference
1 F2.3 0 0.09293 -0.0929
2 F2.4 0 0.08121 -0.0812
3 F2.5 0 0.1436 -0.144
4 0.5% F1 inF2.0 0.5 0.3714 0.129
5 0.5% F1 inF2.1 0.5 0.3593 0.141
6 0.5% F1 inF2.2 0.5 0.3602 0.14
7 1% F1 inF2.0 1 0.8608 0.139
8 1% F1 inF2.1 1 0.9692 0.0308
9 1% F1 inF2.2 1 0.9415 0.0585
10 2% F1 inF2.0 2 2.122 -0.122
11 2% F1 inF2.1 2 2.217 -0.217
12 2% F1 inF2.2 2 2.359 -0.359
13 5% F1 inF2.0 5 4.818 0.182
14 5% F1 inF2.1 5 4.642 0.358
15 5% F1 inF2.2 5 4.707 0.293
16 10% F1 inF2.0 10 10.27 -0.268
17 10% F1 inF2.1 10 10.34 -0.339
18 10% F1 inF2.2 10 10.06 -0.0592
19 20% F1 inF2.0 20 19.31 0.69
20 20% F1 inF2.1 20 20.17 -0.17
21 20% F1 inF2.2 20 20.21 -0.206
22 F1.0 100 99.94 0.0578
23 F1.1 100 99.98 0.017
24 F1.2 100 100.1 -0.0806
Correl Coef (r) 0.999973
Figure 6.12: Regression curve for the predicted and true
values
y = 1x - 0.003R² = 0.999
-20
0
20
40
60
80
100
120
0 20 40 60 80 100 120
-
Page 300 of 305
All these spiked samples were also analysed by Differential
scanning calorimeter
(DSC) and attempts were made to correlate the results with those
obtained form NIR.
The thermograms are provided below( Figure 6.13 to 6.19).
Figure 6.13: DSC Thermogram for 0.5% spiked sample of form I in
Form II
-
Page 301 of 305
Figure 6.14: DSC Thermogram for 1.0% spiked sample of form I in
Form II
Figure 6.15: DSC Thermogram for 2.0% spiked sample of form I in
Form II
-
Page 302 of 305
Figure 6.16: DSC Thermogram for 5.0% spiked sample of form I in
Form II
Figure 6.17: DSC Thermogram for 10.0% spiked sample of form I in
Form II
-
Page 303 of 305
Figure 6.18: DSC Thermogram for 20.0% spiked sample of form I in
Form II
Figure 6.19: DSC Thermogram for sample of form I
-
Page 304 of 305
Observation
The DSC spectra could not yield any significant differentiation
between the samples
and thus is not found to be a suitable technique determination
of polymorphic purity
for Sertraline HCL.
Conclusion
The polymorphic purity determination method by NIR spectroscopy
is thus developed
as an accurate method to determine even very low levels of
polymorphic impurity
i.e.as low as 0.5%
Both the Conformity index method as well as the quant model can
be used as efficient
modes to determine the polymorphic purity for Sertraline Hcl by
NIR.
-
Page 305 of 305
References
1. Online edition for students of organic chemistry lab courses
at the University of
Colorado, Boulder, Dept of Chem and Biochem. (2002)
2. W. Fred McClure Anal. Chem., 66 (1), pp 42A–53A (1994)
3. F. Westad, A. Schimdt and M Kermit, J.Near Infrared
Spectrosc. 16, 265-273 (2008)
4. W. M Welch, Discovery and preclinical development of the
serotonin reuptake
inhibitor Sertraline, Advances in Medicinal Chemistry.3,
113-148, (1995).
5. D. Healy, The Antidepressant Era. Cambridge, Massachusetts:
Harvard University
Press.168, (1999).
6. J. Couzin, The Brains Behind Blockbusters. Science. 309, 728,
(2005).
7. R. Sarges.; JR Tretter.; SS Tenen, A. Weissman, Journal of
Medicinal Chemistry, 16,
1003, (1973)
8. US patent No 6872853, Polymorphic forms of sertraline
hydrochloride.
9. M. Varasteh.; Z. Deng.; H. Hwang.; Y.J. Kim.; B Geoffrey.;
International Journal of
Pharmaceutics 366 74–81(2009)
10. Y Xie.; W Tao.; H Morrison.; R Chiu.; J Jona.; J Fang.; N
Cauchonl.; International
Journal of Pharmaceutics 362 29–36, (2008)
11. M Blanco.; J Coello.; H. Iturriaga.; S Maspoch.; C.
Pérez-Maseda.; Analytica Chimica Acta 407 247–254(2000)