Forensic Science International 226 (2013) 282–289

Contents lists available at SciVerse ScienceDirect

Forensic Science International

journal homepage: www.elsevier.com/locate/forsciint

Counterfeit Cialis and Viagra fingerprinting by ATR-FTIR spectroscopy with

chemometry: Can the same pharmaceutical powder mixture be used to falsify
two medicines?
Rafael S. Ortiz a,b,*, Kristiane de Cássia Mariotti b, Bruna Fank a, Renata P. Limberger b,
Michel J. Anzanello c, Paulo Mayorga b
a
Rio Grande do Sul Technical and Scientifical Division, Brazilian Federal Police, 90160-093 Porto Alegre, RS, Brazil
b
Department of Pharmacy, Federal University of Rio Grande do Sul (UFRGS), 90610-000 Porto Alegre, RS, Brazil
c
Department of Industrial Engineering, Federal University of Rio Grande do Sul (UFRGS), Porto Alegre, CEP 90.035-190, Rio Grande do Sul, Brazil

A R T I C L E I N F O A B S T R A C T

Article history: This paper proposes a direct and efficient method to discriminate between counterfeit and authentic
Received 19 September 2012 Cialis and Viagra samples by combining attenuated total reflection Fourier transform infrared (ATR-FTIR)
Received in revised form 9 January 2013 spectroscopy with multivariate techniques. The chemical profile of 53 commercial samples (Viagra1,
Accepted 27 January 2013
Cialis1) and 104 counterfeit samples (Viagra and Cialis) from distinct seizures were obtained from ATR-
Available online 17 February 2013
FTIR data derived from 10 mg of crushed core tablets. Principal component analysis (PCA) technique was
employed to classify samples based on the fingerprint region mid-infrared spectra (1800–525 cm 1)
Keywords:
using OMNIC v.7.2 software; PCA enabled categorizing samples in groups with different chemical
Fourier transform infrared spectroscopy
Attenuated total reflectance
profiles, successfully distinguishing between authentic and counterfeits samples in forensic routine. The
Chemical profile existence of active pharmaceutical ingredients (API) and technological adjuvant others than specified on
Counterfeit medicines the medicine package were also detected in counterfeit samples. In addition, we applied the similarity
Viagra1 match (SM) method to demonstrate that a mixture of pharmaceutical powders deriving from a common
Cialis1 origin may have been used to manufacture both counterfeit Cialis and Viagra tablets from distinct
seizures.
ß 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
The increasing commerce and use of counterfeit medicines
offers serious risks to public health worldwide, and the repression
to that commerce mobilizes drug manufacturers, health surveil-
lance agents, police and forensic forces. Phosphodiesterase type-5
(PDE-5) inhibitors for treating erectile dysfunction are among the
most falsified drugs due to their market success, high cost, and
embarrassment associated with the underlying condition. Coun-
terfeit PDE-5 inhibitors are normally purchased from fraudulent
websites, making it hard for police to control its commerce [1]. In
developed countries, PDE-5 inhibitors like Viagra1 (sildenafil
citrate, Pfizer), Cialis1 (tadalafil, Eli Lilly), and more recently
Levitra1 (vardenafil hydrochloride, Bayer)1 are among the most
popular counterfeited medicines [2]. In Brazil, Viagra1 and Cialis1
* Corresponding author at: Rio Grande do Sul Technical and Scientifical Division,
Brazilian Federal Police, 90160-093 Porto Alegre, RS, Brazil. Tel.: +55 5132359079;
fax: +55 5132359066.
E-mail addresses: rafaelortiz.rso@dpf.gov.br, ortiz.rs@gmail.com (R.S. Ortiz).
1
The symbol 1 was used only in reference to authentic products.
are also deemed the most falsified medicines [3], accounting for
nearly 80% of drug counterfeiting analyzed by the Brazilian federal
police (BFP) from 2007 to 2010 [4].
Several approaches relying on different analytical techniques
have been recently developed to identify unauthentic sildenafil
(SLD), tadalafil (TAD) and analogues; such studies aim at
elucidating structural aspects of those substances in pure form,
dosage forms, and as adulterants in dietary supplements [5–8].
From the analytical point of view, an interesting strategy consists
of applying methods that minimize or require no sample
preparation. For that matter, spectroscopy methods have proven
to be very efficient since they allow characterizing samples by
directly measuring substances in solid state; spectroscopy
methods also rely on straightforward operational steps, and
provide reliable, fast results. Maurin et al. [9] proposed the use of
X-ray pattern diffraction focused on two purposes: the discrimi-
nation of counterfeit Viagra tablets, and the identification of
differences in drug composition. Trefi et al. [10] applied Raman
spectroscopy, 1H NMR, and 2D DOSY (diffusion-ordered spectros-
copy) NMR to analyze genuine and illegal formulations of Cialis1
purchased via internet, while SLD in pure form and Viagra1
formulations were easily and unequivocally characterized by

0379-0738/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.forsciint.2013.01.043
 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289
multinuclear nuclear magnetic resonance (NMR) spectroscopy
including active principle ingredient (API) quantification in Wawer
et al. [11]. Raman spectroscopy was also stated by Veij et al. [12] as
a fast and reliable method for detecting counterfeit Viagra1 tablets
with no sample preparation. With similar purposes, Vredenbregt
et al. [13] described a nondestructive near-infrared (NIR)
spectroscopy method for fast screening of suspect Viagra samples;
practical implications include checking the homogeneity of a
batch, distinguishing between authentic and unauthentic Viagra1,
screening for SLD presence, and detecting whether similar samples
have been previously analyzed. Finally, Sacré et al. [14] compared
Raman spectroscopy, NIR spectroscopy and Fourier transform
infrared spectroscopy (FTIR) ability to discriminate between
counterfeit and genuine samples of Viagra1 and Cialis1.
Attenuated total reflectance (ATR) is a FTIR sampling technique
that became popular because it does not require samples to be
added in KBr pellets or undergo pretreatment; the sample is
measured through direct contact with the crystal surface to obtain
spectroscopic data. ATR-FTIR has been successfully applied to
analyze polymorphic contents of bulk pharmaceutical materials
[15] and powder mixtures [16], to assess dentifrice adulteration
[17], and to differentiate wines from different geographical origins
[18]. More aligned with the propositions of this paper, Champagne
and Emmel [19] used FTIR spectroscopy with ATR sampling to
detect the presence of SLD and TAD as adulterants in dietary
ingredients.
As stated by Muehlethaler et al. [20], ATR-FTIR spectroscopic
data rely on a massive number of correlated variables that may
jeopardize the performance of many statistical-based techniques.
For that matter, multivariate data analysis (MDA) techniques are
deemed powerful tools for interpreting forensic scenarios that
require sample classification. Among those techniques, principal
component analysis (PCA) has been successfully applied as an
exploratory tool to provide information on data structure, detect
outliers, and distinguish between test and standard samples. Also
aligned with the classification purpose, the similarity match (SM)
technique can be applied to numerically express the degree of
similarity between the samples analyzed. Several recent publica-
tions have used the aforementioned techniques to analyze
spectroscopy data in forensic cases, including blood stain
estimation age in [21], counterfeit medicines in [16,22–24],
comparative analysis of automotive paints in [20,25], determina-
tion of genotypes in [26], and differentiation of papers in [27].
In this study, ATR FTIR has provided fast (less than 30 s) and
reliable results in forensic analyses of PDE-5 inhibitors samples.
We analyzed the fingerprint region mid-infrared spectra of 53
commercial and 104 counterfeit samples of Viagra1 and Cialis1
from distinct seizures. The combination of PCA and SM techniques
enabled detecting all counterfeit medicines and grouping different
seizures of illegal products. We also detected high similarity
between some fraudulent samples of Viagra and Cialis, which may
imply that the same mixture of pharmaceutical powders was used
to counterfeit both medicines.
2. Materials and methods
2.1. Samples
Sildenafil citrate (99.9%) and six Viagra1 authentic tablets
containing 50 mg of SLD were supplied by Pfizer Ltda Laboratories.
Tadalafil (99.8%) and 8 Cialis1 authentic tablets containing 20 mg
of TAD were supplied by Eli Lilly do Brasil Ltda Laboratories.
Twenty authentic Cialis1 tablets (TAD, 20 mg) from eight distinct
batches and nineteen authentic Viagra1 tablets (SLD, 50 mg) from
six distinct batches were purchased in local pharmacies. As for the
counterfeit samples, one hundred and four tablets were sent to BPF
283
(Porto Alegre, Rio Grande do Sul State) for forensic analysis through
ATR-FTIR. Information on samples and seizure codes are depicted
in Table 1.
2.2. ATR-FTIR analyses
A Nicolet 380 FTIR Spectrometer (Nicolet Instrument Co.,
Madison, USA) equipped with DTGS (Deuterated Triglycine
Sulphate) detector and smart orbit single reflection diamond
ATR sampling accessory was employed for all experiments.
Measurements were made in absorption mode on a small amount
of sample deposited on the ATR crystal. No sample treatment was
necessary for measurement; tablets without their coats were
homogenized by milling. Next, a sample portion was directly
placed on the ATR element. Each mixture was sampled 3 times, i.e.,
in triplicate. Identical pressure was used for all measurements.
Each spectrum consists of 16 co-added scans measured at a
spectral resolution of 4 cm 1 in the 4000–525 cm 1 range. Spectral
data were acquired with EZ OMNIC software, version 7.2a (Nicolet
Instrument Co.).
After measurement, the crystal was cleaned with acetone and
let to dry in air ambient. An hourly background spectrum was
obtained against air with clean and dry ATR element, using the
same instrumental conditions as the samples. No spectra
pretreatments were performed. Preliminaries studies using SLD,
TAD, Viagra1 and Cialis1 standards show that the baseline
correction yields significant sample misclassification. All spectra
were saved in SPA format for works in EZ OMNIC and TQ Analyst EZ
edition (Nicolet Instrument Co.) software. The most informative
infrared spectra region, the fingerprint region 1800–525 cm 1, was
saved in CSV format, which is a typical format for subsequent input
into statistics software. The raw spectra yielded a 612 (cases, in
triplicate) by 661 (variables, wavelength) data matrix.
3. Results and discussion
This study is mainly focused on the 1800–525 cm 1 spectral
region, where the ‘‘fingerprint’’ region is included and characteristic
groups show absorption, thus enabling the detection of differences
in the spectra [19]. Furthermore, this spectral range includes the
region between 1720 and 1150 cm 1, which highlights major
differences in PDE-5 inhibitors. According to Champagne e Emmel
[15], SLD, TAD and Vardenafil substances (including analogues and
homologues, like Thiosildenafil, Thiohomosildenafil, Hydroxythio-
homosildenafil, Aildenafil, Phentolamine, Sulfoaildenafil, Amino-
tadalafil and KF31327) present absorption peaks in this region,
enabling the detection of multiple PDE-5 inhibitors with a single
test.
Fig. 1 shows the ATR-FTIR spectra of TAD, authentic Cialis1,
SLD, and authentic Viagra1 in the 1800–525 cm 1 spectroscopic
region. Briefly, the most important peaks for TAD (Fig. 1A) can be
Table 1
Description of seizures and tested samples.
Viagra Cialis
Units Code Units Code
Autenthic 25 S Autenthic 28 T
Counterfeit 10 A Counterfeit 1 H
1 B 1 I
1 C 20 J
7 D 16 K
3 E 14 L
4 F 2 M
8 G 4 N
12 O
284 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289

0,32

0,30 A
0,28

0,26

0,24

0,22

0,20
Ab s or ba nc e

0,18

0,16

0,14

0,12

0,10

0,08

0,06

0,04

0,02

- 0,00
18 0 0 16 0 0 14 0 0 12 0 0 10 0 0 80 0 60 0
W av en u mbe r s ( c m- 1)

58 7,94
0,30

55 5,40
0,28
B
0,26
10 25,86

61 8,90
0,24
10 57,04

0,22

73 4,93
93 9,79
11 71,56

0,20

65 3,75
16 98,77

10 94,50

0,18
Ab s or ba nc e

0,16

0,14

0,12

0,10

0,08

0,06

0,04

0,02

18 0 0 16 0 0 14 0 0 12 0 0 10 0 0 80 0 60 0
W av en u mbe r s ( c m- 1)

Fig. 1. ATR FTIR spectra in 1800–525 cm 1 region for TAD, in red, and authentic Cialis1 (A); and SLD, in red, and authentic Viagra1 (B), showing all replicates. (For
interpretation of the references to color in the artwork, the reader is referred to the web version of the article.)

associated to C–O bonds in the 1700 cm 1 band, and C–C bonds
from the ketone group in the 1280 and 1172 cm 1 bands. As for the
SLD (Fig. 1B), the 1676 cm 1 peak can be correlated to C–N
stretching (1690–1640 cm 1); N–H bending appears at
1647 cm 1; the 1490 cm 1 band results from C–C bonds in a ring;
C–N bonds in the O–C–N functional group absorb in the 1400 cm 1
band, accounting for the 1402 cm 1 absorbance; finally, aryl C–N
bonds are responsible for the 1269 cm 1 peak [5,8,15].
Associated with forensic scientists’ expertise, MDA techniques
lead to more consistent and conclusive results in ATR-FTIR spectra
analysis; as stated by Rajalahti and Kvalheim [28]: ‘‘measured data
are not the same as information’’. In our propositions, MDA
techniques are tailored to a very clear purpose in the forensic
filed: assessing the authenticity of a suspect product seized by
police forces. Such results can materialize the occurrence of a crime
and raise a judicial punishment to the offender.
Principal Component Analysis (PCA), a well known MDA
method, decomposes the data matrix into a small number of
principal components (PCs) that maximize the explained variance
in the retained components. PCA outputs include scores and
loadings parameters. Such outputs enable two fronts of analysis: (i)
loading plots reveal covariance among variables and enable
 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289 285

identifying the most relevant variables for variance explanation in
the retained components [28], and (ii) score plots reveal patterns in
the data, including clusters, trends and outliers. In such plots, each
score represents one spectrum (sample); scores closely located
denote similar spectra, i.e., samples with similar characteristics,
while distant scores suggest different spectra [29].
In our propositions, PCA was first applied to sildenafil, tadalafil,
authentic Viagra1 and authentic Cialis1 spectra illustrated in
Fig. 1, yielding the PCA scores in Fig. 2. PCA scores allow a better
visualization than the fuzzy images from infrared spectra,
grouping samples according to the type of solid that constitutes
those samples. Fig. 2A depicts PC1  PC2  PC3 scores; the
retained PCs account for 97.71% of the total variance
(PC1 = 79.44%, PC2 = 10.66%, PC3 = 7.61%). Such figure suggests
four types of samples in different quadrants according to score
signals, i.e., positive or negative: (i) sildenafil (SLD; PC1 < 0;
PC2 < 0; PC3 > 0), (ii) tadalafil (TAD, PC1 < 0; PC2 > 0; PC3 > 0),
(iii) authentic Viagra1 (samples 86–92, PC1 < 0; PC2 < 0;
PC3 < 0), and (iv) authentic Cialis1 (samples 80–85, PC1 < 0;
PC2 > 0; PC3 < 0). Alternatively, one could perform a two
dimension analysis on scores (Fig. 2B–D), enabling a better
visualization of groups. In addition, we decided to retain three
PC due to two main reasons: (1) the first three PCs explain nearly
all the variance total (97.71%), and PC2 is not significantly different
from PC3; and (2) it is possible to distinguish four types of samples
in different quadrants in accordance with positive or negative sign
of the three PCs; that analysis would not be possible using only two
PCs. In our understanding, such a representation yields a very clear
grouping formation, in which replicates or similar samples are
closely linked.
Although loading plots often enable the identification of the
variables that significantly influence the PC’s that was not clearly
observed in this experiment (see Fig. 3). The impact of peaks on PC1
and PC2 are virtually the same. That happens with the most
influential variable on PC1 in 1698 cm 1, which is related to the C–
O of tadalafil. This variable negatively influences PC1 (absorbance
0.08) – ranking all samples according to their tadalafil
concentrations – and positively influences PC2 (absor-
bance + 0.05), explaining that tadalafil and authentic Cialis1
present PC2 > 0. We also see high positive loading values for
PC3 in 1672 cm 1 (abs + 0.11), 1643 cm 1 (abs + 0.14), and
1487 cm 1 (abs + 0.08) bands. These signals are identified in the
spectrum of pure sildenafil (Fig. 1B).
In pharmaceutical powder mixtures, the discrimination
obtained via PCA is related not only to drug presence in the
samples, but also to the various technological adjuvants present in
commercial tablets. This occurs because the FTIR spectrum of
commercial tablets corresponds to a mixture, i.e., drug + adjuvants.
As the adjuvants are generally in higher quantity, e.g., each Cialis1
tablet contains 20 mg of TAD and 245 mg of lactose monohydrate,
the resulting spectrum is very similar to that of the pure adjuvant.
As a consequence, infrared absorption peaks on PC1 observed in

Fig. 2. Scores plots for PCA on SLD, TAD, authentic Viagra1 and authentic Cialis1 spectrum. In (A), PC1  PC2  PC3, the first three PC’s explains 97.71% of the variation in the
experimental data (PC1 = 79.44%; PC2 = 10.66%; PC3 = 7.61%); in (B), PC1  PC2, there is a reasonable data clustering in four groups; in (C), PC1  PC3, there is an
inappropriate grouping, inserting genuine Viagra1 and Cialis1 in the indistinguishable cluster; in (D), PC2  PC3, there is a clear grouping formation, with replicates and
similar samples tightly grouped.
286 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289

Fig. 3. PC1, PC2 and PC3 loading plots for SLD, TAD, authentic Viagra1 and authentic Cialis1 spectrum.

1048 cm 1 (abs 0.05), 909 cm 1 (abs 0.05) and 890 cm 1 (abs
0.05) may be associated to lactose characteristics peaks. Note
that such a result does not imply limitations in the proposed
approach, but emphasizes the occurrence of spectral overlapping
due to the combination of various substances that affect several
analytical techniques.
In the next stage of our propositions, we apply PCA to identify
the fraudulent samples, as in [30,31]. Plots of PC2  PC3 are
deemed appropriate projections to illustrate the experimental
results, as depicted in Fig. 2D. Following the confrontational
forensic routine exemplified in the work of Sacré et al. [14], we
compared counterfeit and genuine Cialis1 samples. Fig. 4 displays
samples split into six groups, with authentic Cialis1 (encoded as T)
inserted in Group 1. Counterfeit Cialis samples were inserted into
the remaining groups (in counter-clockwise): Group 2 (seizures J,
H and K), Group 3 (seizure O), Group 4 (seizure L), Group 5
(seizures M and N), and Group 6 (seizure I).
Similarly, authentic Viagra samples were allocated to Group 1
(S), as depicted in Fig. 5. The counterfeit samples can be separated
according to the seizure (in counter-clockwise): Group 2 (seizures
B, D and F), Group 3 (seizures G and E), Group 4 (seizure C), and
Group 5 (seizure A and one counterfeit sample from seizure D).
In a previous study, we obtained pharmacological active
ingredients profile (via ultra-performance liquid chromatography
with diode array detection coupled mass spectrometry) from the
same samples used in this study. Those results suggested that
some counterfeit Cialis samples contained only SLD, although the
medicine label informed the presence of TAD. We then decided to
abandon the confrontation routine, i.e., genuine Cialis1 versus
counterfeit Cialis or authentic Viagra1 versus counterfeit Viagra,
by applying a new PCA analysis only to ATR-FTIR spectra from
counterfeit samples of Cialis and Viagra. The new results are
depicted in Fig. 6. In the following paragraph, we refer to samples
of Viagra from seizure A as Viagra A for shortness; the same for
other seizures.
All samples of Viagra A and one sample of Viagra D (referred to
as Group 1) are located in the region where PC2 < 0 and PC3 < 0. In
counterclockwise, we have Group 2 (Viagra C), Group 3 (Viagra E,

0,5
D
B B
0,4 DDDDDD D
F
F F F
F DFB
0,3 D D DF DF
F
D
D D D
SSS S S
0,2 SS S
SSSS
SSSS
S S
SSSS
SS
S S
S
SS
S
S S
S
SS
S
S SSS
SSS
S
S S
S
S
SSS SSS
0,1
Factor 2

0,0

-0,1 A
AAA A
AA A A
-0,2 AA A A A
AAAA
G E C A AA
E DAA
D A
-0,3 G EEGGGGGG GEE C A AA A A
GG GG
GG C A
GEE
-0,4 G E

-0,5
-0,4 -0,3 -0,2 -0,1 0,0 0,1 0,2 0,3
Factor 3

Fig. 4. PC2 versus PC3 scores plot for Cialis samples. Authentic Cialis1 (T) in Group 1 Fig. 5. PC2 versus PC3 scores plot for Viagra samples. Authentic Viagra1 (S) is
and counterfeit samples (in counterclockwise direction) are grouped according to inserted in Group 1. Counterfeit samples (in counterclockwise direction) are
their seizures in Group 2 (J, H and K), Group 3 (O), Group 4 (L), Group 5 (M and N) grouped according to their seizures in Group 2 (B, D and F), Group 3 (G and E), Group
and Group 6 (I). 4 (C) and Group 5 (A and one counterfeit sample from seizure D).
 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289 287

Fig. 6. PC2 versus PC3 scores plots for both Cialis and Viagra counterfeits samples. Group 1, consisting of Viagra A and one sample of Viagra D, is located in the region where
PC2 < 0 and PC3 < 0. In counterclockwise, we have Group 2 (Viagra C), Group 3 (Viagra E, Viagra G and some Cialis K, in detail), Group 4 (some Cialis K, Cialis J, Cialis H), Group
5 (Cialis O), Group 6 (Viagra B, Viagra D, Viagra F), Group 7 (Cialis M, Cialis N), Group 8 (Cialis L), and Group 9 (Cialis I).

Viagra G and some Cialis K, in detail in Fig. 6), Group 4 (some Cialis
K, Cialis J, Cialis H), Group 5 (Cialis O), Group 6 (Viagra B, Viagra D,
Viagra F), Group 7 (Cialis M, Cialis N), Group 8 (Cialis L), and
GroCup 9 (Cialis I). This new classification of samples into nine
groups appears to be mainly associated to SLD and TAD
concentrations, since variables related to those substances present
high PC loadings. Samples consisting of the same APIs and
presenting similar concentrations are grouped together: Group 5,
which consists of counterfeit samples presenting only TAD, is
located in PC2 > 0 and PC3 > 0 (as in Fig. 2D for tadalafil in pure
form); Groups 4, 7 and 9 denote the presence of TAD and SLD in
different proportions, and are distributed in PC2 > 0 due to a
predominance of TAD; Groups 2, 3 and 8 are formed by counterfeit
samples consisting of SLD at different levels; finally, Groups 1 and 6
are comprise of forgeries presenting sildenafil and homosildenafil
(HSD, an analogue pharmacological contaminant) in different
proportions. Table 2 depicts the date and city of seizures as well as
APIs identified in seized tablets, suggesting similarities among
seized samples.
We then applied the similarity match (SM) – QCheck spectral
correlation – to the average of three spectra (Viagra E, Viagra G and
Cialis K) in 1800–525 cm 1 region for samples inserted in Group 3
[19]. The SM was obtained using the OMNIC software. QCheck
Spectral Correlation is typically employed for routine quality testing
and day-to-day material verification. It yields a correlation value
measuring the similarity between two spectra; two very similar
spectra, e.g., with the same number of peaks appearing in the same
region, will have a high correlation threshold, i.e., >0.95; two
different spectra, e.g., presenting additional peaks or absorbance
shifts, will have a lower correlation threshold (i.e., <0.80) [14].
The correlations yielded by the SM are presented in Table 3;
correlations higher than 0.95 are highlighted in bold and values
smaller than 0.80 in italics. A brief and preliminary assessment on
results suggests that SMs for authentic Viagra1 and Cialis1 coming
from the same batch are generally greater than 0.95, while
0.85 < SM < 0.95 denotes Viagra and Cialis from different batches
(see Tables 4 and 5, supplementary material). Moreover, SMs for
representative samples of each of the nine counterfeit clusters
(groups) depicted in Fig. 6 assume low values (see Table 6,
supplementary material).
These results suggest that samples inserted in a same group can
be associated with a common origin in terms of the substance used
for its production. That can be derived from (1) a unique mixture
and compressed cycle or (2) repetition of the same standard
formula at different batches. In an adequate mixing and
compressing process, the concentration for each component (w/
w) in the pre-compression mixture and in the tablets made from
such a mixture should be the same. In other words, the core of any
tablet is a representative sample of the API quantitative composi-
tion and technological adjuvant that originated the tablets.
Although the proposed approach yielded satisfactory results,
there are some noteworthy limitations. In the classification of
pharmaceutical tablets by FTIR-ATR, most of the separation is
typically due to differences on tablets chemical composition,

Table 2
Different counterfeit seizures inserted in a common group according to the PCA applied to FTIR-ATR data.

Cluster Seizure Active pharmacological

ingredient
Code Date (dd/mm/year) City

Group 3 Viagra E 06/October/10 Caxias do Sul SLD

Viagra G 21/June/11 Santo Ângelo SLD
Cialis K 02/June/10 Santa Cruz do Sul SLD

Group 4 Cialis H 06/August/08 Porto Alegre SLD/TAD

Cialis J 15/April/10 Gravataı́ SLD/TAD
Cialis K 02/June/10 Santa Cruz do Sul SLD/TAD

Group 6 Viagra B 16/February/09 Porto Alegre SLD/HSD

Viagra D 15/April/10 Gravataı́ SLD/HSD
Viagra F 05/November/10 Passo Fundo SLD/HSD

Group 7 Cialis M 26/November/10 Santo Ângelo SLD/HSD

Cialis N 13/June/11 Santo Ângelo SLD/HSD

HSD: homosildenafil.
288 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289

Table 3
Correlation values for Group 3 samples (counterfeits Viagra E, Viagra G and Cialis K) by QCheck spectral correlation in 1800–525 nm region.

E1 E2 E3 G1 G2 G3 G4 G5 G6 G7 G8 K4 K5 K7 K8 K10 K14 K16

E1 1
E2 0.95 1
E3 0.96 0.90 1
G1 0.97 0.89 0.98 1
G2 0.91 0.85 0.97 0.94 1
G3 0.96 0.93 0.99 0.98 0.96 1
G4 0.95 0.90 0.99 0.98 0.97 0.99 1
G5 0.92 0.83 0.97 0.96 0.98 0.96 0.97 1
G6 0.97 0.92 0.98 0.99 0.94 0.98 0.98 0.94 1
G7 0.95 0.95 0.97 0.95 0.96 0.98 0.98 0.94 0.98 1
G8 0.97 0.91 0.97 0.99 0.92 0.97 0.97 0.94 0.99 0.96 1
K4 0.90 0.90 0.95 0.91 0.96 0.95 0.95 0.94 0.92 0.97 0.90 1
K5 0.88 0.92 0.91 0.86 0.89 0.92 0.89 0.87 0.88 0.94 0.87 0.97 1
K7 0.85 0.91 0.89 0.83 0.91 0.90 0.88 0.86 0.86 0.94 0.83 0.98 0.97 1
K8 0.92 0.93 0.96 0.91 0.95 0.96 0.94 0.92 0.93 0.97 0.83 0.98 0.98 0.97 1
K10 0.93 0.90 0.98 0.94 0.97 0.97 0.96 0.96 0.94 0.96 0.93 0.98 0.96 0.93 0.99 1
K14 0.91 0.82 0.96 0.91 0.91 0.93 0.92 0.94 0.90 0.90 0.89 0.92 0.90 0.85 0.94 0.97 1
K16 0.89 0.75 0.93 0.91 0.88 0.89 0.89 0.92 0.88 0.84 0.87 0.85 0.80 0.74 0.87 0.92 0.97 1

including different APIs, excipients, and contaminants [14]. The References

quantitative analysis of the solid phase using spectroscopy,
however, is often affected by complex spectra, overlapping of
characteristic peaks, and presence of various excipients [28]. In
addition, non-linear responses, i.e., disobedience to Beer’s law, no
absorption of inorganic adjuvants, and instrumental limitations on
ATR sampling may also decrease the performance of this approach.
In order to overcome such limitations, we recommend assuring
that samples are properly homogenized, particles are sufficiently
small and measuring instruments offer reasonable reproducibility
[16,17,28,29].
4. Conclusions
ATR-FTIR spectra provide relatively simple and fast screening
tools in forensic investigations aimed to characterize and classify
authentic (Viagra1, Cialis1) from counterfeit Cialis and Viagra.
In our propositions, ATR-FTIR provides the fingerprinting region
of authentic and counterfeit samples of Viagra and Cialis. The PCA
applied to ATR-FTIR data allowed grouping samples according to
their different chemical profiles, distinguishing successfully
between authentic and counterfeits samples. In addition, PCA
scores inserted counterfeit drugs from different seizures in the
same cluster, suggesting a common illicit source for these
medicines. There was even the formation of a cluster containing
two seizures of counterfeit Viagra and one counterfeit Cialis; for
such a group, the SM suggested a significant correlation between
the samples. In addition, our results suggest that a single
pharmaceutical powder mixture containing only sildenafil was
compressed into tablets of counterfeits Viagra and Cialis.
It is noteworthy that this study was carried out with a limited
number of samples. In scenarios characterized by the frequent
usage of this technique, it is recommended to create a library of
samples for better comparison and classification of new samples.
Furthermore, we understand the proposed method as a technique
applicable in conjunction with other forensic techniques; spectra
pre-treatment, statistical validation of the database, and variable
selection approaches are also promising ways to improve the
interpretation and consolidation of this type of analysis.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.forsciint.2013.01.043.
[1] U. Holzgrabe, M. Malet-Martino, Analytical challenges in drug counterfeiting and
falsification-the NMR approach, J. Pharmaceut. Biomed. Anal. 55 (2011) 679–687.
[2] B.J. Venhuis, D.M. Barends, M.E. Zwaagstra, D. de Kaste, Recent Developments in
Counterfeits and Imitations of Viagra Cialis and Levitra: A 2005–2006 Update,
National Institute for Public Health and the Environment (RIVM), Bilthoven, The
Netherlands, 2007 http://www.rivm.nl/bibliotheek/rapporten/370030001.pdf.
[3] Surveillance Agency (ANVISA), Counterfeit Medicines, Ministry of Health, Brasilia,
2008, http://www.anvisa.gov.br/medicamentos/falsificados/index.htm.
[4] C.R. Jung, R.S. Ortiz, R. Limberger, P. Mayorga, A new methodology for detection of
counterfeit coated tablets by image processing and statistical analysis: Applica-
tion to Viagra1 and Cialis1 tablets, Forensic Sci. Int. 216 (2012) 92–96.
[5] H. Inoue, T. Kanamori, K. Kuwayama, Y. Iwata, T. Satoh, A. Yanagihori, K. Mat-
sushima, Chemical profile of sildenafil and related compounds, Jpn. J. Forensic Sci.
Technol. 13 (2008) 73–82.
[6] S. Singh, B. Prasad, A. Savaliya, R. Shah, V. Gohil, A. Kaur, Strategies for character-
izing sildenafil, vardenafil, tadalafil and their analogues in herbal dietary supple-
ments, and detecting counterfeit products containing these drugs, Trends Anal.
Chem. 28 (2009) 13–28.
[7] P. Zou, S.S. Oh, P. Hou, M. Low, H. Koh, Simultaneous determination of synthetic
phosphodiesterase-5 inhibitors found in a dietary supplement and pre-mixed
bulk powders for dietary supplements using high-performance liquid chroma-
tography with diode array detection and liquid chromatography–electrospray
ionization tandem mass spectrometry, J. Chromatogr. A 1104 (2006) 113–122.
[8] L. Blok-Tip, B. Zomer, F. Bakker, K.D. Hartog, M. Hamzink, J. Ten Hove, M.
Vredenbregt, D. de Kaste, Structure elucidation of sildenafil analogues in herbal
products, Food Addit. Contam. 21 (2004) 737–748.
[9] J. Maurin, F. Plucinski, A. Mazurek, Z. Fijalek, The usefulness of simple X-ray
powder diffraction analysis for counterfeit control – the Viagra example, J. Pharm.
Biomed. Anal. 43 (2007) 1514–1518.
[10] S. Trefi, C. Routaboul, S. Hamieh, V. Gilard, M. Malet-Martino, R. Martino, Analysis
of illegally manufactured formulations of tadalafil (Cialis) by 1H NMR, 2D DOSY
1H NMR and Raman spectroscopy, J. Pharm. Biomed. Anal. 47 (2008) 103–113.
[11] I. Wawer, M. Pisklak, Z. Chilmonczyk, 1H, 13C, 15N NMR analysis of sildenafil base
and citrate (Viagra) in solution, solid state and pharmaceutical dosage forms, J.
Pharm. Biomed. Anal. 38 (2005) 865–870.
[12] M. de Veij, A. Deneckere, P. Vandenabeele, D. de Kaste, L. Moens, Detection of
counterfeit Viagra1 with Raman spectroscopy, J. Pharm. Biomed. Anal. 46 (2008)
303–309.
[13] M.J. Vredenbregt, L. Blok-Tip, R. Hoogerbrugge, D.M. Barends, D. de Kaste,
Screening suspected counterfeit Viagra and imitations of Viagra with near-
infrared spectroscopy, J. Pharm. Biomed. Anal. 40 (2006) 840–849.
[14] P.Y. Sacré, E. Deconinck, T. de Beer, P. Courselle, R. Vancauwenberghe, P. Chiap, J.
Crommen, J.O. de Beer, Comparison and combination of spectroscopic techniques
for the detection of counterfeit medicines, J. Pharm. Biomed. Anal. 53 (2010)
445–453.
[15] A. Salari, R.E. Young, Application of attenuated total reflectance FTIR spectroscopy
to the analysis of mixtures of pharmaceutical polymorphs, Int. J. Pharm. 163
(1998) 157–166.
[16] O. Planinsek, D. Planinsek, A. Zega, M. Breznik, S. Srcic, Surface analysis of powder
binary mixtures with ATR-FTIR spectroscopy, Int. J. Pharm. 319 (2006) 13–19.
[17] M. López-Sánchez, A. Domı́nguez-Vidal, M.J. Ayora-Cañada, A. Molina-Dı́az, As-
sessment of dentifrice adulteration with diethylene glycol by means of ATR-FTIR
spectroscopy and chemometrics, Anal. Chim. Acta 620 (2008) 113–119.
[18] P.A. Tarantilis, V.E. Troianou, C.S. Pappas, Y.S. Kotseridis, M.G. Polissiou, Differentiation
of Greek red wines on the basis of grape variety using attenuated total reflectance
Fourier transform infrared spectroscopy, Food Chem. 111 (2008) 192–196.
 R.S. Ortiz et al. / Forensic Science International 226 (2013) 282–289
[19] A.B. Champagne, K.V. Emmel, Rapid screening test for adulteration in raw
materials of dietary supplements, Vib. Spectrosc. 55 (2011) 216–223.
[20] C. Muehlethaler, G. Massonnet, P. Esseiva, The application of chemometrics on
infrared and Raman spectra as a tool for the forensic analysis of paints, Forensic
Sci. Int. 209 (2011) 173–182.
[21] B. Li, P. Beveridge, W.T. O’Hare, M. Islam, The estimation of the age of a blood stain
using reflectance spectroscopy with a microspectrophotometer, spectral pre-
processing and linear discriminant analysis, Forensic Sci. Int. 212 (2011) 198–204.
[22] F. Been, Y. Roggo, K. Degardin, P. Esseiva, P. Margot, Profiling of counterfeit
medicines by vibrational spectroscopy, Forensic Sci. Int. 211 (2011) 83–100.
[23] K. Dégardin, Y. Roggo, F. Been, P. Margot, Detection and chemical profiling of
medicine counterfeits by Raman spectroscopy and chemometrics, Anal. Chim.
Acta 705 (2011) 334–341.
[24] W.J. Olds, E. Jaatinen, P. Fredericks, B. Cletus, H. Panayiotou, E.L. Izake, Spatially
offset Raman spectroscopy (SORS) for the analysis and detection of packaged
pharmaceuticals and concealed drugs, Forensic Sci. Int. 212 (2011) 69–77.
[25] E. McIntee, E. Viglino, C. Rinke, S. Kumor, L. Ni, M.E. Sigman, Comparative analysis
of automotive paints by laser induced breakdown spectroscopy and nonpara-
metric permutation tests, Spectrochim. Acta B 65 (2010) 542–548.
289
[26] L. Ren, Y.-Z. Gao, J.-H. Yin, W.-X. Shi, X.-W. Yu, H.-P. Xie, W.-P. Wang, The
determination for the three genotypes of D16S539 locus based on near-infrared
spectroscopy and chemical pattern recognition, Anal. Chim. Acta 638 (2009)
202–208.
[27] V. Causin, C. Marega, A. Marigo, R. Casamassima, G. Peluso, L. Ripani, Forensic
differentiation of paper by X-ray diffraction and infrared spectroscopy, Forensic
Sci. Int. 197 (2010) 70–74.
[28] T. Rajalahti, O.M. Kvalheim, Multivariate data analysis in pharmaceutics: a
tutorial review, Int. J. Pharm. 417 (2011) 280–290.
[29] J. Aaltonen, K.C. Gordon, C.J. Strachan, T. Rades, Perspectives in the use of
spectroscopy to characterise pharmaceutical solids, Int. J. Pharm. 364 (2008)
159–169.
[30] R.S. Ortiz, K.C. Mariotti, N.V. Schwab, G.P. Sabin, W.F.C. Rocha, E.V.R. Castro, R.P.
Limberger, P. Mayorga, M.I.M.S. Bueno, W. Romão, Fingerprinting of sildenafil
citrate and tadalafil tablets in pharmaceutical formulations via X-Ray fluores-
cence spectrometry (XRF), J. Pharm. Biomed. Anal. 58 (2012) 7–11.
[31] R.S. Ortiz, K.C. Mariotti, W. Romão, M.N. Eberlin, R.P. Limberger, P. Mayorga,
Chemical fingerprinting of counterfeits of Viagra and Cialis tablets and analogues
via electrospray ionization mass spectrometry, Am. J. Anal. Chem. 2 (2011) 919–928.