Application Note
Quantification of Tablets Containing Multiple APIs
Using Transmission Raman
The ability to predict multiple constituents of a final dosage form
in one fast, non-destructive measurement can reduce analysis time.
This is especially important when quantification of multiple APIs
is required for tests such as content uniformity, assay and ID.
This example, based on a common cold and flu product, demonstrates the
quantification of 5 components (3 APIs and 2 excipients) using a 9 second
measurement. Nominal concentration ranged from 1 to 85 % w/w.
Introduction
Typical solid dose forms of a drug product are often complex and contain at least one active pharmaceutical
ingredient (API) and a range of excipients at varying levels. The most widely-used means of quantitative analysis
is high performance liquid chromatography (HPLC), which is slow, laborious, prone to manual errors and expensive.
Transmission Raman spectroscopy (TRS) is a regulatory-acceptable alternative technique for content uniformity,
assay and drug product ID, enabling fast, non-destructive analysis of tablets and capsules without chemical
preparation or skilled analytical chemists.
Usually only APIs are quantified, but monitoring excipients as well as API’s may be advantageous. For example,
if a particular batch of drug product differs in terms of one of its excipients, it may affect one of the quality critical
attributes such as dissolution. This extra information comes with almost no extra cost or complexity from
transmission Raman methods.
Experimental
The drug product in this example, a common cold and flu formulation, contained 5 constituents (3 API’s and 2
Excipients) at the nominal concentrations shown below. This product is expensive to analyse by HPLC as multiple
separations are required to measure all 3 APIs, taking several days to complete.
Phenylephrine
Caffeine
Paracetamol
Mag Stearate
Tablettose
1%
4%
85%
1%
9%
7
In order to use transmission Raman spectroscopy
to quantify multiple components a thorough
calibration set is required. The calibration design in
this example included an orthogonal 5 level design
consisting of 20 samples. In order to validate the
models, 5 independent validation samples with
varying concentrations were tested.
3
2
6
1
0
5
Powdered materials were weighed, ground, mixed
and compressed into tablets. Samples were loaded
into a sample tray and then into the TRS100
instrument for automated analysis. Two tablets per
sample were scanned using 1.0 W laser power
(830 nm) for 9 seconds. Spectra shown in Figure 1
indicates regions of variation due to constituents.
Intensity (Counts)
400
500
600
700
0.6
4
0.4
0.2
3
0
−0.2
−0.4
2
−0.6
950
1
1000
1050
3
2
0
1
0
−1
600
800
1000 1200 1400
Wavenumber (cm−1)
1600
1800
1500 1550 1600 1650 1700 1750
Figure 1: Calibration spectra, baselined and normalised to show spectra variation.
www.cobaltlight.com
CAN025-1-EN
Application Note
Results and Analysis
R2 = 0.929
RMSECV = 0.156
1
0.5
0.5
1
1.5
2
2.5
8
6
R2 = 0.981
RMSECV = 0.337
4
2
2
3
4
5
6
7
8
9
10
2
1.5
R2 = 0.893
RMSECV = 0.199
1
0.5
Tablettose
Mag Stearate
1
1
1.5
2
2.5
15
2
R = 0.985
RMSECV = 0.335
10
5
Paracetamol
0.5
6
8
10
12
14
16
90
85
R = 0.980
RMSECV = 0.551
80
85
90
Figure 2. PLS Calibration Models and statistics for each constituent.
Summary
It was demonstrated that multiple APIs in a single
intact tablet can be quantified easily by transmission
Raman spectroscopy. The time taken for CU, assay
and ID was reduced from approximately 2 days (by
chromatographic methods) to <5 minutes for a batch
of 10 tablets, and does not require skilled analytical
resource for routine testing.
With little extra effort the concentrations of excipients
are also determined in the method, which can provide
useful attributes for process monitoring applications.
Transmission Raman is an effective and powerful tool
for quantifying multiple actives quickly and effectively,
even for low concentration components.
References:
J. Griffen, A. Owen, P. Matousek, Comprehensive quantification
of tablets with multiple active pharmaceutical ingredients using
transmission Raman spectroscopy – A proof of concept study;
Journal of Pharmaceutical and Biomedical Analysis, 2015, 115, 227-282.
www.cobaltlight.com
2
1.5
RMSEP = 0.115
1
0.8
2
80
75
75
Validation of the model, and prediction of independent
samples, are shown in Figure 3. Results indicate that all
substances are predicted well with higher performance
for the stronger Raman scatterers (the APIs and caffeine,
as the highest content excipient).
1
1.2
1.4
1.6
1.8
2
2.2
10
Caffeine
Caffeine
10
The results indicate that PLS models can be built for
all constituents, with good model parameters for
the APIs. Of all constituents magnesium stearate is
the worst performing model. This is due to the poor
Raman scattering ability of the constituent and the
low concentration at 1% w/w. Of greatest importance
is the success of the phenylephrine model, which
gives acceptable model performance in a single
measurement, even though the spectra are dominated
by the acetaminophen component.
8
6
4
RMSEP = 0.332
2
1
Mag Stearate
0
2
3
4
5
6
7
8
9
10
2
1.5
1
RMSEP = 0.141
0.5
0.5
Tablettose
0
Calibration spectra were used to build predictive
models for each constituent of the mixture (Fig. 2).
1
1.5
2
15
10
5
RMSEP = 0.271
6
8
10
12
14
16
90
85
80
75
75
RMSEP = 0.537
80
85
90
Figure 3. PLS Validation results for each constituent.
CAN025-1-EN
© Cobalt Light Systems Ltd. This information is subject to change without notice.
Cobalt and TRS100 are trademarks of Cobalt Light Systems Ltd.
1.5
Phenylephrine
2
Paracetamol
Phenylephrine
2.5