Photodegradation of flupentixol in aqueous solution under irradiation at 254 nm: Identification of the photoproducts generated

Article abstract of DOI:10.1016/j.jphotochem.2010.06.036

After irradiation at 254 nm of aqueous solutions of the antipsychotic drug flupentixol, the structures of the photodegradation products were determined by ultra high performance liquid-chromatography linked to mass spectrometry. Fragmentation patterns of the parent ions were established on a hybrid linear ion trap-orbitrap mass spectrometer allowing accurate mass measurements of both parent and daughter ions. This allowed to propose plausible structures for the main photolysis products of flupentixol. A total of nine photoproducts were detected after irradiation of the drug. The main photoproduct is generated following the addition of a hydroxyl group on the double bond adjacent to the thioxanthene ring. Secondary photoproducts were also observed.

Full text of DOI:10.1016/j.jphotochem.2010.06.036

Journal of Photochemistry and Photobiology A: Chemistry 214 (2010) 224–229
Contents lists available at ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Photodegradation of flupentixol in aqueous solution under irradiation
at 254 nm: Identification of the photoproducts generated
Aubert Maquille, Sybiline Salembier, Marie-France Hérent, Jean-Louis Habib Jiwan∗
Louvain Drug Research Institute, Université catholique de Louvain (UCL), 1200 Brussels, Belgium
a r t i c l e i n f o
a b s t r a c t
Article history:
After irradiation at 254 nm of aqueous solutions of the antipsychotic drug flupentixol, the structures
of the photodegradation products were determined by ultra high performance liquid-chromatography
linked to mass spectrometry. Fragmentation patterns of the parent ions were established on a hybrid
linear ion trap–orbitrap mass spectrometer allowing accurate mass measurements of both parent and
daughter ions. This allowed to propose plausible structures for the main photolysis products of flupen-
tixol. A total of nine photoproducts were detected after irradiation of the drug. The main photoproduct
is generated following the addition of a hydroxyl group on the double bond adjacent to the thioxanthene
ring. Secondary photoproducts were also observed.
Received 11 March 2010
Received in revised form 28 June 2010
Accepted 30 June 2010
Available online 8 July 2010
Keywords:
Flupentixol
Photolysis products
UHPLC
© 2010 Elsevier B.V. All rights reserved.
Mass spectrometry
MSn
1. Introduction
exhibit biological activity which could affect aquatic biosystems
and impact drinking water supplies. Photochemical reactions
Many pharmaceuticals are sensitive to light, leading to a
decrease in their purity and to the formation of photodegradation
products. In addition, after the absorption of drugs, exposure to UV-
A or UV-B irradiation in superficial areas in the body could lead to
phototoxicity (caused by direct damage to the tissue induced by a
photogenerated chemical agent) or photoallergic reaction (forma-
tion of an allergen such as a drug protein adduct) [1]. Stability tests
on drugs, including light stress tests are required by the European
Agency for the Evaluation of Medicinal Products [2] to assess the
stability of the active substance.
Flupentixol is a thioxanthene drug, a class of typical antipsy-
chotic used for the treatment of schizophrenic patients. Thioxan-
thene neuroleptics are photosensitive and are often responsible for
photo-induced skin eruptions in patient treated with these drugs
[3]. Detailed investigation on the identification of the photolysis
products of this drug has not yet been carried on. The study of pho-
todegradation products of such drugs may give insight onto the
mechanisms leading to their formation.
account for the degradation of pharmaceutical pollutants. Direct
photolysis can also be used during drinking water treatment [6].
To some extent, photolysis at 254 nm can be extrapolated to envi-
ronmental conditions [7,8], particularly concerning the structure of
photoproducts. To this regard, the identification of the photolysis
products of pharmaceuticals could improve the understanding of
their environmental fate [9,10].
Hyphenated MS techniques have proven to be valuable tools
for identifying drugs photolysis products [11,12]. The aim of this
work is to identify degradation of flupentixol in aqueous solu-
tions exposed to UV irradiation at 254 nm in order to define
its photodegradation process. The UV spectrum of flupentixol in
aqueous solution recorded in the wavelength region between 200
and 400 nm shows a maximum around 265 nm, which is close to
the chosen irradiation wavelength. The main photolysis products
were analyzed by LC–(MS)n. Quantification of the photoproducts
was performed. Since standard compounds are not available this
quantification is relative to the parent compound. This method
of quantification is the recommended method by the European
Agency for the Evaluation of Medicinal Products [13]. Structural
characterizations of the main photolysis products were based on
changes in molecular masses and spectral patterns of product ions.
Accurate mass measurements and collision induced dissociation
were carried on in a hybrid linear trap/orbitrap. Accurate mass mea-
surement provides a higher grade of confidence for the assignment
of the possible structure of the photodegradation products.
Pharmaceuticals are released in the environment as a result
from their use in human clinical practice and are commonly
detected in the aquatic environment [4,5]. These compounds
∗
Corresponding author at: Université catholique de Louvain, Avenue E. Mounier,
72 bte 7230, B-1200 Brussels, Belgium. Tel.: +32 2 764 73 68; fax: +32 2 764 72 96.
E-mail address: j-l.habib@uclouvain.be (J.-L. Habib Jiwan).
1010-6030/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2010.06.036

A. Maquille et al. / Journal of Photochemistry and Photobiology A: Chemistry 214 (2010) 224–229
225
2. Materials and methods
plotting the peak areas (measured in nominal units) against the cor-
responding flupentixol concentration. The obtained relationship
was the following: y = 8167x − 7, r2 = 0.998. Standard deviation of
the slope was 16 whilst standard deviation of the intercept was
1581. The limits of detection (LOD) and quantification (LOQ), con-
sidered, respectively, as 3 and 10 times the signal-to-noise ratio
[14], were evaluated by injecting progressive dilutions of flupen-
tixol into the LC system. LOQ was 2 ␮M.
Repeatability and reproducibility were evaluated by analyzing
the same standard solution three times on the same day and prepar-
ing and analyzing standard solutions at similar concentration on
three different days. The observed RSD for intraday variation was
about 1.5%. The inter-day RSD was about 2%.
2.1. Materials
Cis-Z-flupentixol dihydrochloride (2-[4-[3-[2-(trifluoromethyl)
thioxanthen-9-ylidene] propyl]piperazin-1-yl]ethanol) (>98%
purity) was purchased from Sigma–Aldrich (St. Louis, MO, USA)
and was stored in the dark. HPLC-MS grade methanol was supplied
from Biosolve (Valkenswaard, Netherlands). Glacial formic acid
was provided by Merck. Deionized water from a Millipore Milli-Q
water purification system was used.
2.2. Samples
Three milliliters of flupentixol dihydrochloride 0.5 mg ml−1
solutions prepared with deionized water were put in quartz cells
(1 cm path length) and irradiated in a Rayonet photochemical
reactor equipped with four UV fluorescent lamps RPR 2537 Å. Air
equilibrated solutions were exposed to UV irradiation for time
intervals of 10, 15, 30, 45, 60, 90 and 180 min. Unirradiated con-
trol samples consisted in flupentixol solutions filled in quartz
cells wrapped in aluminum foil and therefore, protected from
light exposure. Experiments were performed at room temperature
(293 4 K). All samples were protected from further light exposure
after irradiation.
3. Results and discussion
3.1. Photodegradation of flupentixol
Fig. 1 represents the remaining flupentixol concentration as a
function of the irradiation time. The logarithmic plot of the flu-
pentixol as a function of time up to 60 min of irradiation inserted
in Fig. 1 shows that the photolysis degradation of flupentixol at
254 nm follows first-order kinetics up to that time of irradiation.
Regression analysis gave a high correlation coefficient, demonstrat-
ing the validity of the first-order kinetic. Under these conditions, a
rate constant, k, of 0.0096 min−1 and a half-time, T1/2, of 72.2 min
are obtained. Fast degradation of flupentixol occurs following irra-
diation. Only 55% of the initial drug concentration remains after
a 60 min irradiation. However, for longest irradiation times, the
remaining concentration is higher than expected (e.g. about 40%
of the drug is still present after a 180 min irradiation) which may
be due to a quenching of the photolysis products.
2.3. UHPLC-DAD-MS system
The LC MS/MS system consisted in a Thermo Accela pump,
autosampler and photodiode array detector. Separation was per-
formed on a Thermo Hypersil Gold C18 column, 50 mm × 2.1 mm
packed with 1.9 ␮m particle at a constant flow rate of 500 ␮l/min
using a binary solvent system: Solvent A, HPLC grade water with
0.1% formic acid and Solvent B, methanol. The initial HPLC gradient
system was held at 15% B for 0.5 min then linearly increased to 90%
B in 8 min, followed by a return to initial conditions for column re-
equilibration. Analyses were performed in triplicate. The injection
volume was 5 ␮l. The absorbance was measured between 200 and
600 nm. The quantification wavelength was 265 nm. The MS system
consisted of an LTQ-Orbitrap XL hybrid mass spectrometer (Thermo
Fisher Scientific, Bremen, Germany) equipped with an electrospray
ionization source. Mass spectrometry analyses were carried out in
positive ion mode using full-scan MS with a mass range of 100–800
m/z. The orbitrap operated at 30,000 resolution (full width at half
height maximum, FWHM). All experimental data were acquired
using external calibration prior to data acquisition. Data acquisi-
tion and processing was carried out with Xcalibur 2.0.7 software
(Thermo). The mass parameters were as follows: capillary temper-
ature: 250 ◦C; sheath and auxiliary gas flow 40 and 20 arbitrary
units, respectively; source voltage: 5 kV; capillary voltage: 8 V; tube
lens voltage: 40 V. Collision induced dissociation experiments were
carried out at 40% normalized collision energy.
3.2. Photodegradation products
An overlay of the UV chromatograms of photo-irradiated flu-
pentixol is shown in Fig. 2. The quantification results for the main
photolysis products expressed as percentages of the initial drug
concentration are reported in Table 1. The degradation peaks were
numbered by increasing retention times.
Peak 4 was already present in control unirradiated samples and
did not increase significantly following UV irradiation. A total of
10 degradation peaks are detected after irradiation of flupentixol.
2.4. Validation studies
Quantification of all peaks was based on UV detection. The per-
centage of flupentixol remaining after UV exposure was calculated
from the ratio of the areas under the curve of the drug peak between
irradiated and control samples [14].
The photolysis products were quantified as a percentage of the
initial drug concentration. As no standards were available for these
products, and according to the guidelines [15], they were quantified
as flupentixol, assuming they have similar response factors as the
parent drug. Flupentixol solutions in concentrations ranging from 2
to 200 ␮M were injected in order to establish the calibration curve
(six calibration points). The calibration curve was constructed by
Fig. 1. Evolution of the concentration of flupentixol as a function of irradiation time.
Insert: first-order kinetic plot for the photo-induced degradation of flupentixol up
to 60 min.

226
A. Maquille et al. / Journal of Photochemistry and Photobiology A: Chemistry 214 (2010) 224–229
Fig. 2. Overlay of UHPLC-UV chromatograms of flupentixol solutions 0.5 mg ml−1 obtained at increasing irradiation times.
The most intense degradation peak (7) accounts for about 18% of
the initial parent drug concentration after 60 min irradiation. The
concentration of all photolysis products increased up to 90 min irra-
diation although at 60 min, a decrease is observed for peak 7 which
is the major degradation peak. This decrease may be attributed
the photolysis of this compound, leading to secondary photolysis
products.
tion. Mass accuracy measurements allowed the assignment of the
molecular formula of product ions.
Fig. 3 shows the collision induced dissociation mass spectrum
of flupentixol together with its major fragments. CID of flupentixol
yields to three fragments ions at mass-to-charge ratios (m/z) of 390
(resulting from the loss of C2H5O from the lateral chain), 307 orig-
inating from the loss of the hydroxyethylpiperazine ring and 362
(corresponding to the breakage of the piperazine ring). A m/z 265
ion (C14H8F3S) corresponds to the loss of the whole lateral chain.
Consecutive losses of hydrofluoric acid are observed and lead to
m/z 415, 395 and 375 daughter ions. A m/z 169 fragment that corre-
sponds to the whole lateral chain is also present. All the proposed
structures of the fragment ions were consistent with the results
from the accurate mass measurement.
The exact masses of flupentixol and its photolysis products as
well as the corresponding formula are listed in Table 2. The main
CID fragments for photolysis products and their daughter ions are
reported in Table 3. The structures of the photolysis products are
shown in Fig. 4.
3.3. Characterization of photolysis products
Identification of photolysis products was performed by liquid-
chromatography coupled to
a linear ion trap–orbitrap mass
spectrometer that allows collision induced dissociation (CID) and
accurate mass determination on both product ions and precursors.
This enables to determine the empirical formula for the unknown
photolysis products. The relative mass errors were below 2 ppm
which guarantees the correct assignation in all the cases. Using
an electrospray ionization source and detection in the positive
mode, pseudo-molecular ions [M+H]+ have been obtained for each
compound. Collision induced dissociation was performed on each
photolysis product in order to confirm the proposed structure. A
comparison of the fragmentation pattern of the photoproducts with
that of flupentixol allowed to assign the location of the modifica-
Three isobaric compounds with m/z 451 are detected at 4.4, 4.5
and 6.8 min (products 3b, 4 and 8). These compounds correspond
to the addition of an oxygen atom on the parent drug. Slight dif-
ferences are observed in their MS–MS fragmentation patterns. The
Table 1
Quantification of the degradation peaks generated following UV irradiation of flupentixol 0.5 mg ml−1 solution at different times.
Peak
Products
R.T. (min)
Mean percentages of impurity peaks relative to the initial drug concentration
0 min
10 min
15 min
30 min
45 min
60 min
90 min
1
2
3
4
5
6
7
8
9
(1)
(2)
(3a) (3b)
(4)
(5)
(6)
(7)
(8)
(9)
3.35
4.15
4.35
4.49
4.61
5.12
5.27
6.82
6.26
7.43
<LOD
<LOD
<LOD
0.77
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOQ
0.70
0.34
0.67
1.77
0.56
7.76
0.35
<LOQ
1.64
0.29
0.81
0.79
0.64
2.57
0.75
10.22
0.48
0.31
2.43
0.37
1.54
1.69
0.67
2.94
1.03
14.34
0.75
0.36
3.62
0.52
2.26
2.37
0.66
3.58
1.24
16.29
1.04
0.42
4.90
0.65
2.80
3.15
0.74
4.54
1.39
18.32
1.42
0.64
6.98
1.55
4.04
6.96
0.75
7.78
1.50
14.72
2.46
0.95
13.38
10
(10)

A. Maquille et al. / Journal of Photochemistry and Photobiology A: Chemistry 214 (2010) 224–229
227
Fig. 3. MS–MS spectra and structure of flupentixol and its fragment ions.
first product (3b) shows fragments such as 169 and 143 that cor-
respond to an intact lateral chain. In addition it exhibits a m/z 281
fragment that corresponds to the ring containing an oxygen. There-
fore, this product might be formed following oxidation of the sulfur
into a S-oxide such as it is the case for many sulfide containing drugs
such as diltiazem [16].
ion amongst the daughter ion of product (4) shows that the oxygen
is located on the fist part of the lateral chain. Product (4) does not
increase following irradiation so that it could not be considered
as a photoproduct. For product (8), the m/z 421 fragment results
from the loss of a CH2O moiety, which differs from the fragmen-
tation pattern of flupentixol. The m/z 307 fragment shows that the
modification is located on the piperazine part. This product might
result from the cleavage of the piperazine ring according to a mech-
The fragmentation of the two other products does not indicate
the presence of an oxygen on the ring. The presence of a m/z 293
Table 2
List of experimentally measured accurate mass of flupentixol and its photodegradation products with the predicted mass-to-charge values for the proposed chemical formula
and the relative mass error.
Product no.
Chemical formula
Experimental mass
Theoretical mass
Mass error (ppm)
1
2
3a
3b
4
5
6
7
C21H24O2N2F3S
C23H29O4N2S
425.15005
429.18378
427.16569
451.16577
451.16578
469.17602
409.15512
453.18101
451.16552
307.03962
281.02411
435.17044
425.15051
429.18425
427.16616
451.16616
451.16616
469.17672
409.1556
453.18181
451.16616
307.0399
−1.08
−1.11
−1.1
C
C
21H26O2N2F3S
23H26O2N2F3S
−0.86
−0.84
−1.5
C23H26O2N2F3S
C
C
23H28O3N2F3S
21H24ON2F3S
−1.16
−1.77
−1.42
−0.9
C23H28O2N2F3S
C23H26O2N2F3S
C16H10OF3S
C14H8OF3S
C23H26ON2F3S
8
9
10
281.02425
435.17125
−0.49
−1.85
Flupentixol
Table 3
List of MS–MS fragments and daughter ion fragments for the main photoproducts of flupentixol.
Product no.
1
Experimental mass
425.15005
Theoretical mass
425.15051
Mass error (ppm)
Fragmentation
409.1556
−1.08
364.0982
338.0820
281.0245
293.0244
MS3 of 409
411.1734
409.1555
364.0983
299.0736
391.1449
322.0871
143.1176
281.0245
2
3a
429.18378
427.16569
429.18425
427.16616
−1.11
−1.1
MS3 of 409
433.1555
433.1555
451.1659
391.1449
321.0562
406.1321
433.1556
348.1028
281.0244
293.0244
339.0661
322.0871
169.1335
265.0294
157.1334
307.0760
143.1178
143.1177
265.0293
3b
4
5
451.16577
451.16578
469.17602
451.16616
451.16616
469.17672
−0.86
−0.84
−1.5
MS3 of 339
391.1449
435.1711
433.1555
287.0337
261.0179
321.0556
348.1028
323.0712
421.1555
279.0450
233.0230
297.0193
322.0871
295.0393
364.09774
265.0293
212.0288
281.0242
307.0762
281.0243
307.0763
238.0446
184.0339
6
7
8
9
409.15512
453.18101
451.16552
307.03962
281.02411
409.1556
453.18181
451.16616
307.0399
281.02425
−1.16
−1.77
−1.42
−0.9
265.0293
265.0294
265.0294
10
−0.49

228
A. Maquille et al. / Journal of Photochemistry and Photobiology A: Chemistry 214 (2010) 224–229
Fig. 4. Structures of flupentixol and its main photolysis products (protonated).
anism proposed by Mella et al. [18]. The m/z 364 ion corresponds
to C19H17F3NOS.
The formula C21H24ON2F3S was assigned to product (6), which
differs from the parent compound by the loss of acetylene. The
265 m/z fragment indicates that the thioxanthene ring was not
modified. Comparison of the MS–MS spectra of this photoprod-
uct with that of flupentixol determined that the modification was
located on the piperazine ring as the presence of m/z 307 amongst
the fragments indicated that the first part of the lateral chain is
intact. In addition, the loss of 61.0523 mass units to yield m/z
348 product ion corresponds to the loss of 2-aminomethanol from
the lateral chain. Therefore, the modification corresponds to N-
dealkylation of the ring following UV irradiation. The breakage of
the piperazine ring has also been reported for other drugs such as
ciprofloxacin [17,18].
Product (10) with m/z 281 corresponds to the formula
C14H8OF3S. Its main fragments fragment with m/z 261 results from
the loss of hydrofluoric acid whereas m/z 212 originates from the
loss of the CF3 group. The m/z 233 fragment is due to the losses
of both HF and CO which may indicate the presence of a ketone
within this molecule. This product would result from oxidation of
flupentixol into a thioxanthone ring.
Product (9), with m/z 307 corresponds to the formula
C16H10OF3S. CID this product leads to the losses of HF (m/z 287),
CF3 (m/z 238) and CO (m/z 279). The presence of the m/z 265 ion
indicates that the thioxanthene ring is intact. Therefore, the modi-
fication is located on the lateral chain which is consistent with the
proposed structure.
Fragmentation of product 3a (m/z 427) gives a m/z 409 fragment.
Fragments originating from MS3 fragmentation of this daughter

A. Maquille et al. / Journal of Photochemistry and Photobiology A: Chemistry 214 (2010) 224–229
229
ion are identical to those obtained after CID of product (6). There-
fore, these two products only differ by the addition of a hydroxyl
together with the reduction of a double bond. This product might
be a secondary photoproduct originating from the photolysis of
product (7) since a significant increase of its corresponding peak
is observed at long irradiation times.
The phototoxicity and the photoallergic reaction of the drug
could be explained by the formation of photolysis products such
as products (9) or product (10) with aldehyde and ketones that
could readily react with proteins following degradation of the drug
in the superficial layer of the body.
The m/z 453 product originates from the addition of a hydroxyl
on the molecule and from the reduction of a double bond. CID of this
product leads to losses of water (m/z 435), the m/z 323 fragment
correspond to the breakage between the carbon and the piperazine
ring, indicated no changes on this part of the lateral chain. The m/z
281 product, with a formula of C14H8OF3S, indicates the presence of
an oxygen on the thioxanthene ring although the fragment m/z 265
results from the loss of that oxygen together with the whole lateral
chain. Therefore, this product could correspond to the addition of
a hydroxyl radical on the double bond from the lateral chain. Due
to its higher electrophilicity, the addition is favored on the ternary
carbon from the thioxanthene ring to give product (7).
4. Conclusions
Flupentixol is rapidly degraded upon UV irradiation of solutions
up to ca. 60%. At that point, the concentration of the photoprod-
ucts in the solution is certainly sufficiently high to compete with
flupentixol for light absorption and photoreaction. The major pho-
todegradation pathway occurs through hydroxyl addition on the
double bond. For the majority of the photoproducts, the thioxan-
thene ring remains intact although oxidation into thioxanthone
was observed for three minor products.
Acknowledgments
Product (5) corresponds to the addition of two oxygen atoms
on the parent drug together with the reduction of a double bond.
Fragmentation of this product into a m/z 339 ion indicates that the
piperazineethanol part was not modified. The m/z 297 fragment
corresponds to the loss of C9H20N2O, which is the whole lateral
chain. Therefore, the modification has occurred on the ring and may
correspond to addition of a hydroxyl on the double bond adjacent
to the ring together with oxidation of the sulfur into a S-oxide.
The exact structure of product (2) could not be determined. It
contains no fluoride and its fragmentation pattern indicates that
the lateral chain is the same as the parent drug. The position of the
three oxygen atoms on the ring could not be determined by mass
spectrometry.
The authors would like to thank the Belgian National Fund for
Scientific Research (FNRS) (FRFC 2.4555.08), the Special Fund for
Research (FSR) and the faculty of medicine of UCL for their financial
support for the acquisition of the LTQ-Orbitrap.
References
[1] G. Cosa, Photodegradation and photosensitization in pharmaceutical products:
assessing drug phototoxicity, Pure Appl. Chem. 76 (2004) 263–275.
[2] European Agency for the Evaluation of Medicinal Products (EMEA), Photosta-
bility testing of new active substances and medicinal products, ICH Topic Q1B,
step 5 (CPMP/ICH/279/95), EMEA, London, 1995.
[3] B. Eberlein-König, A. Bindl, B. Przybilla, Phototoxic properties of neuroleptic
drugs, Dermatology 194 (1997) 131–135.
The fragmentation of product (1) with a m/z of 425 fragments
results in the formation of daughter ions with m/z of 409 (loss
of water), 364 (C19H17F3NOS), 338 and 281. This later fragment
indicates the presence of an oxygen on the ring. The fragmenta-
tion pattern is concordant with the breakage of the piperazine ring
together with oxidation of the sulfur from the thioxanthene ring.
[4] M. Petrovic´, M.D. Hernando, M.S. Díaz-Cruz, D. Barceló, Liquid chromato-
graphy–tandem mass spectrometry for the analysis of pharmaceutical residues
in environmental samples: a review, J. Chromatogr. A 1067 (2005) 1–14.
[5] O.A. Jones, N. Voulvoulis, J.N. Lester, Human pharmaceuticals in the aquatic
environment a review, Environ. Technol. 22 (2001) 1383–1394.
[6] O. González, C. Sans, S. Espulgas, S. Malato, Application of solar advanced
oxidation processes to the degradation of the antibiotic sulfamethoxazole, Pho-
tochem. Photobiol. Sci. 8 (2009) 1032–1039.
[7] A.L. Boreen, W.A. Arnold, K. McNeill, Photochemical fate of sulfa drugs in
the aquatic environment: sulfa drugs containing five-membered heterocyclic
groups, Environ. Sci. Technol. 38 (2004) 3933–3940.
3.4. Photolysis mechanisms
[8] D. Błe˛dzka, D. Gryglik, J.S. Miller, Photodegradation of butylparaben in aqueous
solutions by 254 nm irradiation, J. Photochem. Photobiol. A: Chem. 203 (2009)
131–136.
[9] A. Piram, A. Salvador, C. Verne, B. Herbreteau, R. Faure, Photolysis of beta-
blockers in environmental waters, Chemosphere 73 (2008) 1265–1271.
[10] B.L. Edhlund, W.A. Arnold, K. McNeill, Aquatic photochemistry of nitrofuran
antibiotics, Environ. Sci. Technol. 40 (2006) 5422–5427.
[11] C. Sirtori, A. Zapata, S. Malato, W. Gernjak, A.R. Fernández-Alba, A. Agüera, Solar
photocatalytic treatment of quinolones: intermediates and toxicity evaluation,
Photochem. Photobiol. Sci. 8 (2009) 644–651.
The main photodegradation mechanism of flupentixol appears
to be due to the addition of a hydroxyl group on the double bond
adjacent to the ring to give product (7). The decrease in the con-
centration of this product after long irradiation times indicates
that it also undergo photolysis into secondary photoproducts such
as products (3a) and (5) whose corresponding peaks appears to
increase.
In the proposed structures of the photolysis products, no break-
down of the thioxanthene ring is observed which suggest that
the photolytic process do not lead to complete photomineraliza-
tion of the drug. The attack of oxygen on the sulfur is commonly
observed following UV irradiation although in the case of flupen-
tixol it accounts only for a small portion of the drug degradation,
contrary to other sulfur containing cycles such as dilthiazem for
which the S-oxide was the major degradation product [16].
The breakage of the piperazine ring following UV irradiation has
already been described for ciprofloxacine by Mella et al. [18]. This
reaction necessitates hydrogen abstraction from the ring followed
by the attack of water on the radical and hydrogen transfer to give
an aldehyde that may correspond to product (8). According to this
mechanism, further photolysis of product (8) leads to product (6),
which is a secondary photoproduct.
[12] A. Maquille, J.-L. Habib-Jiwan, LC–MS characterization of photolysis products
from UV irradiated metoclopramide solutions, J. Photochem. A: Chem. 205
(2009) 197–202.
[13] European Agency for the Evaluation of Medicinal Products (EMEA), Note
for guidance on impurities in new medicinal products, Topic Q3B, step 4
(CPMP/ICH/2738/99), London, 2006.
[14] European Pharmacopoeia, 5th edition, Council of Europe, Strasbourg, 2005.
[15] European Agency for the Evaluation of Medicinal Products (EMEA), Note
for guidance on impurities in new medicinal products, Topic Q3B, step 4
(CPMP/ICH/282/95), London, 1996.
[16] V. Andrisano, P. Hrelia, R. Gotti, A. Leoni, V. Cavrini, Photostability and photo-
toxicity studies on diltiazem, J. Pharm. Biomed. Anal. 25 (2001) 589–597.
[17] T.G. Vasconcelos, D.M. Henriques, A. König, F.A. Martins, K. Kümmerer, Photo-
degradation of the antimicrobial ciprofloxacin at high pH: identification and
biodegradability assessment of the primary by-products, Chemosphere 76
(2009) 487–493.
[18] M. Mella, E. Fasani, A. Albini, Photochemistry of 1-cyclopropyl-6-fluoro-1,4-
dihydro-4-oxo-7-(piperazin-1-yl)quinoline-3-carboxylic acid (=ciprofloxacin)
in aqueous solution, Helv. Chim. Acta 84 (2001) 2508–2519.