Photochemistry of flutamide in various media: Investigation of the reaction mechanism as revealed by external magnetic field effects on product yields

Article abstract of DOI:10.1016/j.jphotochem.2011.10.015

Photochemistry of an anticancer drug flutamide, 2-methyl-N-[4-nitro-3- (trifluoromethyl) phenyl] propanamide has been investigated in various media in the absence and presence of 0.1 T magnetic field. On photoexcitation flutamide gives rise to a phenol de

Full text of DOI:10.1016/j.jphotochem.2011.10.015

Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
Contents lists available at SciVerse ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Photochemistry of flutamide in various media: Investigation of the reaction
mechanism as revealed by external magnetic field effects on product yields
Chikako Udagawaa,b, Shuichi Fukuyoshia, Shotaro Morimotob,
Yoshifumi Tanimotob, Ryoichi Nakagakia,∗
a Graduate School of Natural Science & Technology, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Japan
b Faculty of Pharmacy, Osaka Ohtani University, Nishikiori-kita, Tondabayashi, Osaka 584-8540, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Photochemistry of an anticancer drug flutamide, 2-methyl-N-[4-nitro-3-(trifluoromethyl) phenyl]
propanamide has been investigated in various media in the absence and presence of 0.1 T magnetic
field. On photoexcitation flutamide gives rise to a phenol derivative through nitro-to-nitrite rearrange-
ment in homogeneous as well as in inhomogeneous solutions. Magnetic field effects on the product yields
revealed that a radical pair is involved in the photoreduction, where the excited triplet state of flutamide
is the precursor of the pair. It undergoes hydrogen abstraction reaction in inhomogeneous solutions
where solvent molecules act not only as component of solvent cage but also as hydrogen donor.
© 2011 Elsevier B.V. All rights reserved.
Received 18 August 2011
Received in revised form
30 September 2011
Accepted 20 October 2011
Available online 29 October 2011
Keywords:
Flutamide
Photochemistry
Magnetic field effects
Solvent cage
1. Introduction
Since a radical pair is composed of two radicals, it has two different
spin states, i.e., singlet (S) and triplet (T). In the absence of mag-
Flutamide,
2-methyl-N-[4-nitro-3-(trifluoromethyl)phenyl]
netic fields, S and three triplet sublevels (T+, T , T ) of a radical pair
−
0
propanamide, is a non-steroidal anticancer drug, commonly used
in advanced prostate cancer [1–3]. Some reports have shown the
capability of flutamide to induce in vivo phototoxic and photo-
allergic effects [4–9]. To clarify the mechanism of the drug-induced
disorders, it is necessary to investigate its photophysical and pho-
tochemical properties. Some of preceding studies are concerned
with the photochemistry of flutamide in various media [10–14].
Flutamide gave rise exclusively to a phenol derivative through the
nitro-to-nitrite photo-rearrangement followed by cleavage of the
O–N bond [10,12–14] not only in homogeneous solution such as
phosphate buffer but also in inhomogeneous solutions containing
cage-forming compounds, ␤-cyclodextrin or vesicles. Only in
inhomogeneous solution, flutamide gave forth photo-reduction
product [10,12–14].
Magnetic field effects (MFEs) can be used as a simple and use-
ful tool to reveal mechanism and microenvironment for chemical
reactions involving radical pair [15]. The MFEs on reaction rates
and product yields have been rationalized in terms of the radical
pair model developed for interpretation of CIDNP [16,17]. The MFEs
observed for biradicals can be dealt with in a similar manner [17].
are degenerate and the intersystem crossing (ISC) between S and
three triplet sublevels is induced by the electron–nuclear hyper-
fine coupling. The singlet radical pair recombines, resulting in a
cage product, whereas the triplet radical pair cannot recombine
to form stable species. Because almost all stable organic com-
pounds are in the ground singlet state, the triplet radical pair
cannot be transformed into stable compounds and separates into
two free radicals to yield escape products. In the presence of mag-
netic fields the ISC rate is reduced because the degeneracy is
removed due to the Zeeman splitting of sublevels. Consequently,
product yields can be modulated on application of a magnetic
field.
The spin multiplicity of the radical pair is the same as that of
its precursor, for example, excited singlet or triplet state. If the
excited triplet state acts as the precursor, the cage product yield
decreases due to the reduced intersystem crossing between the
triplet and singlet radical pair induced by hyperfine coupling. The
decrease in the cage product yield is concomitant with the increase
in the escape product yield by applying a magnetic field. Thus, we
can easily determine the initial spin multiplicity of the radical pair
precursor by observing and analyzing MFEs.
From the analysis of MFEs we can speculate about the nature of
microenvironment surrounding the solute molecules. The lifetime
of the radical pair is controlled by the environmental properties,
∗
Corresponding author.
E-mail address: nakagaki@p.kanazawa-u.ac.jp (R. Nakagaki).
1010-6030/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2011.10.015

58
C. Udagawa et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
for instance, viscosity. Since the radical pair is short-lived in non-
viscous homogeneous solutions, no MFEs or very small MFEs are
expected to be observed on reaction rates or yields. On the other
hand, the radical pair lifetime is long in inhomogeneous media, and
therefore appreciable or significant MFEs are observable.
In order to clarify the reaction mechanism in detail, we elu-
cidated the photochemistry of flutamide by comparing reactivity
in various media, such as phosphate buffer (pH 7.4), micelle such
as sodium dodecyl sulfate or polyoxyethylene(23)lauryl ether (Brij
35), ␤-cyclodextrin (␤-CD), and bovine serum albumin (BSA) solu-
tion in the absence and presence of a magnetic field (ca. 0.1 T).
These media were selected for the sake of modeling biologi-
cal environments with increasing complexity and physiological
pH.
photolysis at 313 nm, below the 15% conversion of the starting
species, on the basis of the following relation [10]:
ꢀ
ꢁꢀ
ꢁ
d[X]
dt
v
FI
˚ =
(1)
where d[X]/dt is the initial rate of disappearance of flutamide or the
initial rate of formation of the single photoproducts, v is the vol-
ume of the irradiation sample, F = 1–10−A is the fraction of photons
absorbed by flutamide at the excitation wavelength and I is the light
intensity (mol of photons min−1). F is evaluated from absorbance
(A) at 313 nm.
2.4. Magnetic field effects on the photochemistry of flutamide
A pair of rare-earth permanent magnets (30 mm × 20 mm ×
8 mm, Tokin LM-30) was placed on the side surface of a cell holder
to generate 0.1 T.
2. Materials and methods
2.1. Chemicals
2.5. Synthesis and characterization of the photoproducts
Flutamide was purchased from Tokyo Chemical Industry Co. and
used as received. Phosphate buffer 10−2 M pH 7.4 (PB) was pre-
pared from reagent grade products. The pH values were measured
for buffered solutions with a glass electrode. The following solu-
tions were prepared from reagent grade products by dissolving in
the buffer solution: 10−2 M ␤-cyclodextrin (␤-CD), 10−1 M sodium
dodecyl sulfate (SDS), 1.6 × 10−2 M polyoxyethylene (23)lauryl
ether (Brij35) and 1 mg/mL bovine serum albumin (BSA). Solvents
of HPLC and analytical grade were used in the photochemical exper-
iments.
After irradiation, irradiated solutions (2 mL × 3) were extracted
with diethyl ether. Extracted solutions were evaporated and dis-
solved in an eluting solution for preparative scale HPLC. Product 1
and Product 2 were obtained on a preparative scale HPLC. On the
other hand, the Product 1 and Product 2 were synthesized accord-
ing to the method described below. The photoproducts separated
by HPLC were identified in comparison with absorption spectra,
HPLC retention times and GC–MS fragmentation patterns obtained
for the synthesized authentic samples.
2.5.1. Product 1: N-[4-hydroxy-3-(trifluoromethyl)phenyl]
isobutyramide
2.2. Instrumentation
2.5.1.1. Preparation of 4-nitro-2-(trifluoromethyl)phenol. To 4-
nitro-2-(trifluoromethyl) aniline (502 mg, 2.43 mmol) in concen-
trated sulfuric acid (2.5 mL), a solution of sodium nitrite (190 mg of
sodium nitrite in 1.3 mL of concentrated sulfuric acid) was added
gradually and stirred for 1 h. The reaction mixture was poured
into ice-cold water and vigorously stirred for 5 min. The solution
was heated gradually to 100 ◦C for 1 h. After cooling at room
temperature, aqueous ammonia was added for neutralization
and extracted with ethyl acetate. The extract was dried over
anhydrous Na2SO4 and evaporated, affording the corresponding
phenol (46.5 mg, 9%) as an orange solid. 1H-NMR (400 MHz, CDCl3)
ı: 8.49 (d, J = 2.6 Hz, 1H), 8.34 (dd, J = 2.7, 7.9 Hz, 1H), 7.26 (brs, 1H),
7.09 (d, J = 9.0 Hz, 1H).
Absorption spectra were recorded on a spectrophotometer
(USB4000 and DT-NIMI-2-GS, Ocean Optics Inc.). Proton magnetic
resonance (1H-NMR) spectra were recorded on a JEOL ECS-400.
Gas chromatography–Mass spectrometry (GC–MS) was performed
on a Hewlett-Packard HP6890/HP5973. High performance liquid
chromatograms were obtained on a Shimadzu LC-20 series. The
quantitative analysis of flutamide and its photoproducts were
achieved on a Hiber LiChrosorb RP-18 eluting with a mixed sol-
vent of CH3CN and 0.01 M phosphate buffer (pH 7.4) of a volume
ratio of 1–1. Preparative scale HPLC was achieved on a Wakosil-
II 5C18 HG-Prep using eluting with a mixed solvent consisting of
CH3CN and water of a volume ratio of 1 to 1. The retention times of
the starting species, Product 1 (N- [4-hydroxy-3-(trifluoromethyl)
phenyl]isobutylamide) and Product 2 were found to be 12, 4.8
and 13.5 min, respectively. These retention times in HPLC anal-
ysis were essentially the same as those observed in various
solvents.
2.5.1.2. Preparation of 4-amino-2-(trifluoromethyl)phenol. To mix-
ture of 4-nitro-2-(trifluoromethyl)phenol (25 mg, 0.12 mmol) and
ammonium chloride (50 mg, 0.94 mmol) in methanol/water 1:1
solution, zinc powder (933 mg, 14.3 mmol) was added and stirred
for 1 h. The mixture was filtered off, washed with hot water, and
extracted three times with ethyl acetate. The extract was dried over
anhydrous Na2SO4 and evaporated, affording the corresponding
amine (18.2 mg, 85%) as an orange solid.
2.3. Irradiation conditions
Irradiation of flutamide was performed by using a Xe arc (Ushio
Optical Modulex UXL-300SX) filtered by a glass filter UTVAF-50S-
34U (Sigma-Koki) and a solution filter K2CrO4 (7.15 mg/100 mL;
path length, 2 cm), its peak wavelength being 313 nm. The incident
photon flux on a 2.2 mL quartz cuvettes was determined to be ca.
1 × 1016 quanta s−1 by means of ferric oxalate actinometry [18,19].
Flutamide solutions of 10−4 M in various media were degassed by
freeze–pump–thaw cycles and irradiated for periods up to 2 h. The
absorption spectra were measured at regular time intervals. After
irradiation, reaction mixtures were analyzed by HPLC.
2.5.1.3. Preparation of N-[4-hydroxy-3-(trifluoromethyl)phenyl]
isobutyramide. The amine 4-amino-2-(trifluoromethyl)phenol
(5.7 mg, 32.2 ␮mol) was treated with an excess of isobutyryl
chloride (27.5 ␮L, 177 ␮mol) and pyridine at room temperature
for 19 h. The reaction mixture was extracted with ethyl acetate
and washed with aqueous cupric sulfate. After concentration, the
crude (9.9 mg) was purified by column chromatography (Al2O3,
hexane–ethyl acetate = 3:2) to give almost pure hydroxy form as a
white solid. HR–MS (EI) m/z: 247.0819 (Calcd. for C11H12F3NO2:
247.0820). 1H-NMR (400 MHz, CD3OD) ı: 7.72 (d, J = 2.7 Hz, 1H),
The quantum yields related to the flutamide photo-degradation
and to the formation of its photoproducts were calculated for

C. Udagawa et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
59
1
0.8
0.6
0.4
0.2
0
7.54 (dd, J = 8.8, 2.7 Hz, 1H), 6.88 (d, J = 8.8 Hz, 1H), 2.59(sep,
J = 6.8 Hz, 1H), 1.09 (d, J = 6.8 Hz, 6H).
initial
60 min
120 min
2.5.2. Product 2: N-[4-nitroso-3-(trifluoromethyl)phenyl]
isobutyramide
2.5.2.1. Preparation
of
N-[4-amino-3-(trifluoromethyl)phenyl]
isobutyramide. To mixture of flutamide (100.7 mg, 365 ␮mol) and
ammonium chloride (0.15 g, 2.80 mmol) in methanol/water 1:1
solution, zinc dust (3.76 g, 57.43 mmol) was added and stirred for
1.5 h. The mixture was filtered off, washed with hot water, and
extracted three times with ethyl acetate. The extract was dried
over anhydrous Na2SO4 and evaporated, affording the correspond-
ing amine (79.9 mg, 84%) as an orange–yellow solid. 1H-NMR
(400 MHz, CDCl3) ı: 7.55 (d, J = 2.4 Hz, 1H), 7.45 (dd, J = 8.7, 2.4 Hz,
1H), 7.35 (brs, 1H), 6.68 (d, J = 8.7 Hz, 1H), 4.08 (brs, 2H), 2.51 (sep,
J = 6.9 Hz, 1H), 1.22 (d, J = 6.9 Hz, 6H).
230
280
330
380
430
wavelength (nm)
Fig. 1. Spectral changes observed for degassed 10−4 M flutamide solution in phos-
phate buffer (pH 7.4) upon regular irradiation time intervals of 60 min.
2.5.2.2. Preparation
isobutyramide. To ice-cooled dichloromethane solution of N-
[4-amino-3-(trifluoromethyl)phenyl]isobutyramide (40.2 mg,
of
N-(4-nitroso-3-(trifluoromethyl)phenyl)
Reaction mixtures were analyzed by HPLC on detecting absorbance
at 254 nm. All chromatograms recorded for photolyzed solutions
(PB, SDS, Brij35, ␤-CD and BSA) had a peak with retention time of
4.8 min, whose intensity increased with irradiation time. This peak
was assigned to Product 1 by comparing UV absorption spectrum,
HPLC chromatogram, and GC–MS fragmentation pattern with the
separately synthesized authentic sample. Except for the PB solu-
tion, a HPLC peak with a retention time of 13.5 min was detected
by monitoring absorbance at 360 nm, whose intensity augmented
on prolonged irradiation. This peak corresponded to Product 2
by comparing the results of absorption spectra, HPLC and GC–MS
obtained for the synthesized authentic Product 2. In both of the
homogeneous and inhomogeneous media, photolysis of flutamide
gave forth the phenolic Product 1 through the photo-induced nitro-
to-nitrite rearrangement. Flutamide underwent photoreduction in
SDS, Brij35, ␤-CD and BSA solutions, yielding Product 2.
0.163 mmol) was added dropwise with stirring two molar amount
of meta-chloroperoxybenzoic acid (56.4 mg) in the same solvent.
After being kept at 0 ◦C overnight, the reaction mixture was shaken
with aqueous sodium carbonate to remove acid. The organic layer
was dried over anhydrous Na2SO4, and then the solvent was
evaporated. The crude product was purified by means of column
chromatography (SiO2; EtOAc/hexane, 4:6) to afford nitroso
compound as a greenish yellow solid (21.4 mg, 50%). HR–MS (EI)
m/z: 260.07726 (Calcd. for C11H11F3N2O2: 260.07726). 1H-NMR
(400 MHz, CDCl3) ı: 8.25 (d, J = 2.2 Hz, 1H), 7.77 (dd, J = 8.9, 2.3 Hz,
1H), 7.26 (brs, 1H), 6.38 (d, J = 8.9 Hz, 1H), 2.59 (sep, J = 6.9 Hz, 1H),
1.30 (d, J = 6.9 Hz, 6H).
3. Results and discussion
3.1. Photoreactions in various media
3.2. Photoreactivity in various media
In 0.1 M SDS and 1.6 × 10−2 M Brij35 solutions, micelles were
formed as their critical micellar concentrations in phosphate buffer
solution were 4.5 × 10−3 and 4 × 10−5 M, respectively. From solu-
bility measurements of flutamide in SDS and Brij35 solutions, the
binding constants of flutamide to SDS and Brij35 micelles were
estimated to be about 1.9 × 104 and 5.8 × 104 M−1, respectively
[20]. The ratio of flutamide bound to the respective micelle to
that in water was 30:1 under the present experimental conditions,
indicating that the photoreaction of flutamide takes place almost
exclusively in micelle cages and not in buffer solution. In the case of
10−2 M ␤-CD solution, the ratio of flutamide bound to ␤-CD to that
in water was estimated as 3:1 from the reported binding constant
(300 50 M−1) [10], indicating photoreaction takes place mainly
flutamide bound to ␤-CD. BSA is used as biological environment
and flutamide is supposed to bound partly to BSA.
Fig. 1 shows the spectral changes observed for a degassed solu-
tion of flutamide 10−4 M in phosphate buffer pH 7.4 at different
irradiation times upon 313 nm light excitation. The absorption
spectrum of flutamide in phosphate buffer solution is character-
ized by a broad band with maximum at 300 nm. The peak intensity
at 300 nm decreased with irradiation time, while the absorbance at
260 nm increased.
The quantum yields for photodegradation (˚−FM), photore-
arrangement (˚1) and photoreduction (˚2) of flutamide were
evaluated through HPLC analysis (Table 1).
Photolysis of flutamide in phosphate buffer solution yielded
exclusively Product 1. No other photoproducts were identified
by HPLC. The photo-degradation quantum yield of flutamide
was ˚−FM = 1.1 × 10−4 and photo-production quantum yield of
Product
1
and Product
2
were ˚1 = 4.0 × 10−5 and ˚2 = 0,
respectively. In 0.1 M SDS solution, the photo-degradation
1
initial
30 min
60 min
90 min
0.8
0.6
0.4
0.2
0
Fig. 2 shows the spectral changes observed for flutamide in
0.1 M SDS solution. The photo-induced spectral changes observed
for SDS solution were entirely different from those observed for
the phosphate buffer solution. The spectral changes observed for
the solutions containing SDS, Brij35, ␤-CD and BSA exhibit isos-
bestic points in the wavelength region 350–370 nm, suggesting
formation of species with an absorption maximum around 360 nm.
230
280
330
380
430
wavelength (nm)
Fig. 2. Spectral changes observed for degassed 10−4 M flutamide solution in phos-
phate buffer (pH 7.4) in the presence of 10−1 M SDS upon regular irradiation time
intervals of 30 min.

60
C. Udagawa et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
Table 1
Quantum yields for the photodegradation of flutamide (˚−FM) and the formation of
photoproducts (˚1 and ꢀ2) in various media.
homogeneous
inhomogeneous
Product 1
˚
−FM
˚
1
˚
2
PBa
1.1 × 10−4
5.5 × 10−4
5.1 × 10−3
6.0 × 10−4
1.0 × 10−3
4.0 × 10−5
2.0 × 10−4
5.8 × 10−4
2.5 × 10−4
8.8 × 10−5
0
0.1 M SDS/PB
0.016 M Brij35/PB
0.01 M ␤-CD/PB
1 mg BSA/mL PB
7.0 × 10−5
2.1 × 10−3
7.6 × 10−5
8.8 × 10−5
hν
Flutamide
inhomogeneous
a
PB: phosphate buffer solution (pH 7.4).
homogeneous media : PB
inhomogeneous media : β-CD, SDS, Brij35, BSA
quantum yield of flutamide was ˚−FM = 5.5 × 10−4 and photo-
Product 2
production quantum yield of Product 1 and Product 2 were
˚1 = 2.0 × 10−4 and ˚2 = 7.0 × 10−5, respectively. The value of ˚
−FM
Scheme 1. Photoproducts formed on photolysis of flutamide.
in the SDS solution was about 5 times the value of ˚
in phos-
−FM
phate buffer solution. The nitro-to-nitrite photo-rearrangement
was accelerated in the SDS solution compared with in the phos-
phate buffer solution. In addition, the photo-reduction took place
in the SDS solution. The value of ˚2 was about 1/3 of the value of
phosphate buffer solution to miceller media, the n,␲* state of flu-
tamide, having less polar electronic structure, is stabilized. This
leads to the increase in the n, ␲* nature in the lowest ␲,␲*
triplet state. Thus, the reactivity of flutamide triplet presumably
increases in SDS and Briji35 micellar solutions. Similarly, reactivity
of flutamide bound to ␤-CD and BSA increases in a hydrophobic
environment compared to that in phosphate buffer.
˚
1
in the SDS solution. The nitro-to-nitrite photo-rearrangement
is dominant compared to the photo-reduction in the SDS solution.
SDS micelles provide a hydrophobic environment suitable for the
photo-induced rearrangement.
We propose Scheme 1 for the photochemistry of flutamide on
the basis of above-mentioned results. In conclusion, at least two dif-
ferent photoreactions (the photoinduced rearrangement and redox
process) compete in inhomogeneous media. Two different reaction
pathways in this scheme can be further rationalized by analyzing
MFEs on photochemical yields for Product 1 and Product 2 (vide
infra).
The value of ˚
times higher than that of ˚
in the 0.016 M Brij35 solution was about 50
−FM
in phosphate buffer solution. In
−FM
the Brij35 solution, the value of ˚2 was larger than the value of
˚1. Since the value of ˚1 in the Brij35 solution was about 3 times
the value in the SDS solution and the value of ˚2 in the Brij35
solution was about 30 times the value in the SDS solution, the
photo-reduction was overwhelmingly accelerated compared with
the photo-rearrangement. This is in accordance with the fact that
the polyoxyethylene chain of Brij35 can act as a better reductant
than the methylene groups in SDS molecule on supplying hydrogen
atoms to the nitro group in the excited flutamide.
The values of the ˚−FM, ˚1 and ˚2 in the ␤-CD solution were
almost the same as those in the SDS solution, suggesting that the
macrocyclic environment of the ␤-CD was similar to that of the SDS
micelle for the photoreaction of flutamide. At the present stage of
investigation we have no reasonable explanation for the similarity
observed in two environments.
3.3. Magnetic field effects on the photoreaction of flutamide
MFEs on the product yields were studied in order to identify
the excited state of flutamide responsible for the present photo-
chemical reaction. Flutamide was irradiated with UV light in the
absence or in the presence of a magnetic field (0.1 T). Fig. 3 shows
the UV spectral changes induced by photolysis of flutamide solu-
tions in Brij35, (a) in the absence and (b) in the presence of 0.1 T. Eq.
(2) is introduced and employed for evaluating the relative quantum
yield (Q) of disappearance or formation for photoinduced reduction
of the nitro-aromatic species [21,22]:
The value of ˚
in the BSA solution was about 10 times the
−FM
value of ˚
in phosphate buffer solution. The value of ˚1 in the
−FM
BSA was about 2 times larger than the value in the phosphate buffer.
The value of ˚2 in the BSA was almost the same as those in the SDS
and ␤-CD solutions. These facts indicate that flutamide was bound
to BSA. The quantum yield for flutamide disappearance (˚−FM) is
much larger than those for product formation (˚1 and ˚2). The
discrepancy in the magnitude of quantum yields suggests that
other unknown reaction pathways become dominant in the BSA
solution.
In SDS, Brij35, ␤-CD and BSA solutions the degradation yield
of flutamide becomes larger than that in phosphate buffer solu-
tion. The photochemistry of flutamide in ␤-CD and BSA solution is
assumed to be the reaction of flutamide-␤-CD and flutamide-BSA
complexes. However, the photoreaction in ␤-CD and BSA solutions
is similar to that in micellar solutions, implying that photochemi-
cal behavior of flutamide complexes is analogous to that in micellar
solutions. It seems that macromolecules, ␤-CD (mol. wt., 1135) and
BSA (mol. wt., 66,000) provide locally hydrophobic environment
for flutamide bound to them, as interior of SDS and Brij35 micelle
is hydrophobic or nonpolar.
ln {exp (A) − 1} − ln {exp (A0) − 1} = Qt
(2)
where the A and A0 values are absorbance at the time t = t and
t = 0, respectively. The yield for the disappearance of flutamide was
determined by the use of Eq. (2) and absorbance changes at 306 nm
shown in Fig. 3. The formation yield of Product 2 was calculated in
a similar manner by analyzing the absorbance increase at 380 nm
shown in Fig. 3. Fig. 4 illustrates the plots of absorbance changes
based on Eq. (2) at (a) 306 nm and (b) 380 nm. While the disappear-
ance of flutamide exhibits no MFEs, the formation yield of Product
2 is remarkably lowered on application of 0.1 T. The deviation from
a linear relationship observed for the absorbance change at 380 nm
may be related to subsequent processes involving Product 2 [23].
There were no MFEs on the absorbance change at 260 nm. However
it was difficult to conclude that there were no MFE on the yield of
Product 1, as absorption bands of Product 1, flutamide and Product
2 were overlapped in this wavelength region.
Table 2 lists the magnitude of MFEs on the relative yield of
photoreaction calculated from absorbance at the time of 10% degra-
dation of the starting species in SDS, Brij35 and ␤-CD solutions,
except for a BSA solution. The reaction mixture obtained after
15 min irradiation was analyzed for the BSA solution. The mag-
netic field of 0.1 T suppressed appreciably the yield of Product 2
in inhomogeneous media.
According to semi-empirical quantum mechanical calculations
[10], the lowest excited triplet state and second lowest triplet state
of flutamide are characterized as ␲,␲* and n,␲* transition, respec-
tively. Probably the n,␲* triplet state is mixed to some extent with
the lowest ␲,␲* triplet state by vibronic coupling. On from the

C. Udagawa et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
61
Table 2
Table 3
Magnetic field effects on relative yield QH/Q0, where QH and Q0 were calculated on
the basis of equation (2) for the absorbance changes in the absence and presence of
magnetic field (0.1 T). Photolysis time is set at 10% decrease in the peak absorbance
for ␤-CD, SDS, Brij35 solutions or 15 min for BSA solution.
Magnetic field effects on relative yield, AH/A0, where AH and A0 refer to the integrated
peak intensity of flutamide and photoproducts (Product 1 or Product 2) in HPLC
chromatograms for sample solutions photolyzed in the absence and presence of
magnetic field (0.1 T), respectively. Photolysis time is set at 10% decrease in the peak
absorbance for ␤-CD, SDS, Brij35 solutions, 180 min for PB solution, and 15 min for
BSA solution.
Q
H/Q0 of 300–306 nmb
Q
H/Q0 of 380–400 nmb
PBa
–
–
A
H/A0 for
A
H/A0 for
A
H/A0 for
Product 2
0.1 M SDS/PBa
0.016 M Brij35/PBa
0.01 M ␤ − CD/PBa
1 mg BSA/mL PBa
0.95
0.97
0.94
(0.82)
0.75
0.54
0.67
(0.68)
flutamide
Product 1
PBa
1.01 0.02
1.01 0.03
1.02 0.01
0.98 0.05
1.00 0.01
1.02 0.02
1.01 0.03
0.99 0.06
0.94 0.04
1.10 0.32
No product
0.51 0.04
0.70 0.18
0.89 0.05
0.66 0.07
0.1 M SDS/PB
0.016 M Brij35/PB
0.01 M ␤-CD/PB
1 mg BSA/mL PB
a
PB: phosphate buffer solution (pH 7.4).
Qt = ln {exp (A) − 1} − ln {exp (A0) − 1}, where A and A0 refer to the absorbance
b
at the time t = t and t = 0, respectively.
PB: phosphate buffer solution (pH 7.4).
HPLC analysis of the irradiated solution revealed the magnitude
of MFEs on the photoreaction of flutamide in a more accurate man-
ner. Table 3 shows the MFEs on the degradation of flutamide and
on the formation yield of Product 1 and Product 2 in various media.
The yield of Product 2 was suppressed by the external magnetic
field in all media. In contrast, neither the consumption of flutamide
nor the formation of Product 1 was affected in the presence of 0.1 T.
The magnitude of MFEs is related to the lifetime of radical pair,
since the lifetime is the inverse of the sum of rate constants for
ISC (magnetic field dependent) and the escape process (magnetic-
field independent) of radicals from a micelle in the case of micellar
solution. The observed MFEs in Brij35 solutions are smaller than
that in SDS solution, whereas the yield of hydrogen abstraction in
Brij35 solution is larger than that in SDS solution (Table 1). These
facts suggest that in SDS micellar solution flutamide was presum-
ably solubilized in viscous hydrocarbon core of the micelle, whereas
in Briji35 (polyoxyethylene(23)lauryl ether) solution it was solubi-
lized in fluid polyoxyethylene outer core and not in hydrocarbon
inner one. Similarly, in the case of photoreaction of flutamide-␤-
CD and flutamide–BSA complexes the binding of flutamide to ␤-CD
does not seem to be strong compared to that of BSA. Radicals formed
in a cage provided by ␤-CD (mol. wt. 1135) could easily escape from
the cage, as a ␤-CD cage with a doughnut-shape is small compared
with micelle cages and has two open exits to bulk solution. In the
case of BSA, flutamide is bound presumably in an inner pocket of
giant BSA molecule (mol. wt. 66,000), though no detailed informa-
tion of binding cite is available at present.
1.2
iniꢀal
2 min
4 min
6 min
8 min
H = 0 T
(a)
0.8
0.4
0
wavelength (nm)
1.2
0.8
0.4
0
iniꢀal
2 min
4 min
6 min
8 min
H = 0.1 T
(b)
As has been described in Introduction, when the radical pair is
involved in the photochemistry, the product yields may be affected
by applying a magnetic field. The fact that no MFEs were observed
on the degradation of flutamide indicates that no flutamide was
regenerated via radical pair intermediate. Since no MFEs were
observed on Product 1 formation, the photoinduced rearrangement
into Product 1 does not occur through the radical pair mechanism.
wavelength (nm)
Fig. 3. Magnetic field effects on the UV-spectral changes induced by the photolysis
of the 0.016 M Brij35 solution of 10−4 M flutamide. (a) In the absence of magnetic
field (b) In the presence of magnetic field (0.1 T).
(a)
(b)
λ = 306nm
λ = 380nm
0
-0.05
-0.1
0.4
0.3
0.2
-0.15
0 T
0.1 T
0 T
0.1
0
-0.2
0.1 T
-0.25
Time (min)
Time (min)
Fig. 4. Plots based on Eq. (2) for the absorbance changes (a) at 306 nm and (b) at 380 nm in Fig. 3.

62
C. Udagawa et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
Scheme 2. Photoreduction mechanism of flutamide in micellar solutions.
If the triplet radical pair acts as the precursor, the yield of cage prod-
uct decreases in the presence of a magnetic field of 0.1 T, since T–S
ISC of the radical pair, which is induced by electron–nuclear hyper-
fine interaction, is reduced on application of the magnetic field.
Product 2 can be characterized as cage product, which is formed
from the singlet radical pair derived from hydrogen-acceptor (aro-
matic nitro group) and hydrogen-donor (cage-forming compounds,
i.e., SDS, Brij35, ␤-CD and BSA). From the fact that the magnetic field
loweredthe formationyield of Product 2, it is concludedthat the ini-
tial spin state of radical pair is triplet and that hydrogen abstraction
reaction takes place at the nitro group in the excited triplet state
of flutamide, because the spin multiplicity is conserved during the
photochemical reaction.
The radical pair involved in photoredox reaction of the nitro
group is not a direct precursor of photoinduced nitro-to-nitrite
rearrangement. In other words, the photo-rearrangement reaction
occurs in the upstream region of radical pair formation from flu-
tamide in the excited state. Flutamide is non-fluorescent or weakly
fluorescent, though its degradation quantum yield is as small as
1.1 × 10−4 in phosphate buffer solution. This fact suggests that
ISC from the excited singlet state to the triplet state may be very
rapid for the excited flutamide and therefore the photoinduced
rearrangement reaction takes place in the triplet manifold of the
nitro-aromatic species, which has been generally discussed in the
literature [24,25].
We propose Scheme 2 for the photoreduction of flutamide in
micellar solutions. SDS and Brij35 surfactant molecules can act as
a good hydrogen-donor. The nitro group of aromatic species in the
triplet state abstracts hydrogen atom from the cage-forming com-
pound and produce a triplet radical pair. The triplet radical pair
undergoes the ISC to the singlet radical pair, which can be trans-
formed into stable photoredox end-products in the ground singlet
state. The precursor of stable singlet products is composed of an
aromatic-ring-N-O-CH moiety (see Scheme 2) where CH group is
a part of cage-forming compound and subsequently decomposes
into the nitroso-aromatic cage product (Product 2) and the oxidized
cage molecule containing a carbonyl group [23]. The magnetic field
modulates the magnitude of ISC between the triplet and singlet
radical pairs. The presence of the cage-forming compounds such as
SDS and Brij35 enhances the lifetime of the triplet radical pair and
the transition probability of the ISC to the singlet radical pair may
become comparable to or larger than the probability of diffusion
from the solvent cage. The singlet radical pair readily recombines
to form the direct precursor, which yields cage products, namely,

C. Udagawa et al. / Journal of Photochemistry and Photobiology A: Chemistry 226 (2011) 57–63
63
the nitroso-aromatic species (Product 2) and the oxidation product
derived from the solvent molecule. In the absence of cage-forming
compounds, the triplet radical pair diffuses away immediately after
its generation by hydrogen abstraction. The photoreaction of flu-
tamide bound to ␤-CD and BSA which are cage-forming compound,
can be explained in a similar manner. On the other hand, in a homo-
geneous medium such as phosphate buffer solution, the triplet
radical pair may not undergo efficiently ISC due to rapid diffusion
from the solvent cage, and therefore no cage products are formed in
the buffer solution. The role of cage-forming compounds is to pro-
vide hydrogen atoms as reductant and to prevent the diffusion of
the triplet radical pair. In other words, micellar media slows down
the escape process and therefore ISC induced by HFC mechanism
can compete with the diffusion.
In some of preceding papers [10,12–14] the photoreactivity of
flutamide is accounted for in terms of the structural changes caused
by “steric constraints” in the solvent cage or environment. Sortino
and co-workers explained that the increased planarity of flutamide
in inhomogeneous media decreases the overlap of the p atomic
orbital of nitro-oxygen with that of aromatic ring carbon necessary
for the nitro-to-nitrite rearrangement [10,12–14]. They argued that
the increase in torsional angle of the nitro group with respect to the
aromatic ring favors the nitro-to-nitrite rearrangement and at the
same time disfavors photo-reduction. The results obtained in the
present study clearly shows that the nitro-to-nitrite rearrangement
is accelerated in inhomogeneous media. It can be interpreted that
the hydrophobic environment provided by the solvent cage is more
suitable for the nitro-to-nitrite rearrangement.
[2] D.K. Ornstein, Combined finasteride and flutamide therapy in men with
advanced prostate cancer, Urology 48 (1998) 901–905.
[3] C. Mahler, J. Verhelst, L. Denism, Clinical pharmacokinetics of the antiandro-
gens and their efficacy in prostate cancer, Clin. Pharmacokinet. 34 (1998)
405–417.
[4] D. Leroy, A. Dompmartin, C. Szczurko, Flutamide photosensitivity, Photoder-
matol. Photoimmunol. Photomed. 12 (1996) 216–218.
[5] R. Zabala, J. Gardeazabal, D. Manzano, Photosensibidad por flutamida, Acta
Dermosifiliogr. 86 (1995) 323–325.
[6] J. Vilaplana, C. Romaguera, A. Azon, M. Lecha, Flutamide photosensitivity –
residual vitiliginous lesions, Contact Dermatitis 38 (1998) 68–70.
[7] M.B. Reid, L.M. Glode, Flutamide induces lupus, J. Urol. 159 (1998) 2098–2101.
[8] R. Yokote, Y. Tokura, N. Igarashi, O. Ishikawa, Y. Miyachi, Photosensitive drug
eruption induced by flutamide, Eur. J. Dermatol. 8 (1998) 427–429.
[9] G. Borroni, V. Brazzelli, F. Baldini, M.R. Gaviglio, B. Beltrami, G. Nollo, Flutamide-
induced pseudoporphyria, Br. J. Dermatol. 138 (1998) 711–712.
[10] S. Sortino, S. Giuffrida, G. De Guidi, R. Chillemi, S. Petralia, G. Marconi, G. Con-
dorelli, S. Sciuto, The photochemistry of flutamide and its inclusion complex
with ␤-cyclodextrin. Dramatic effect of the microenvironment on the nature
and on the efficiency of the photodegradation pathways, Photochem. Photobiol.
73 (2001) 6–13.
[11] F. Vargas, C. Rivas, H. Méndez, A. Fuentes, G. Fraile, M. Velásquez, Photochem-
istry and phototoxicity studies of flutamide, a phototoxic anti-cancer drug, J.
Photochem. Photobiol. B: Biol. 58 (2000) 108–114.
[12] S. Sortino, G. Marconi, G. Condorelli, New insight on the photoreactivity of the
phototoxic anti-cancer flutamide: photochemical pathways selectively locked
and unlocked by structural changes upon drug compartmentalization in phos-
pholipid bilayer vesicles, Chem. Commun. (2001) 1226–1227.
[13] S. Sortino, G. Marconi, S. Petralia, G. Condorelli, Photobinding of flutamide
to phospholipid vesicles: Additional evidence for photoprocess unexpectedly
triggerd by conformational changes in the bilayer, Helv. Chim. Acta 85 (2002)
1407–1414.
[14] S. Sortino, S. Petralia, G. Condorelli, G. Marconi, Direct spectroscopic evi-
dence that the photochemical outcome of flutamide in a protein environment
is tuned by modification of the molecular geometry: A comparison with
the photobehavior in cyclodextrin and vesicles, Helv. Chim. Acta 86 (2003)
266–273.
[15] B. Chakraborty, S. Basu, Study of interaction of proflavin with triethylamine in
homogeneous and micellar media: Photoinduced electron transfer probed by
magnetic field effect, Chem. Phys. Lett. 477 (2009) 382–387;
4. Conclusion
M. Wakasa, T. Yago, A. Hamasaki, Nanoscale heterogeneous structure of ionic
liquid as revealed by magnetic field effects, J. Phys. Chem. B 113 (2009)
10559–10561.
The mechanism of flutamide photochemistry can be summa-
rized below.
[16] U. Steiner, T. Ulrich, Magnetic field effects in chemical kinetics and related
phenomena, Chem. Rev. 89 (1989) 51–147.
[17] H. Hayashi, Chapter 2, Magnetic field and magnetic isotope effects on reac-
tions of radical pairs, in: S. Nagakura, H. Hayashi, T. Azumi (Eds.), Dynamic
Spin Chemistry–Magnetic Controls and Spin Dynamics of Chemical Reactions,
Kodansha, Tokyo, 1998, pp. 7–47;
The nitro group in excited flutamide undergoes the nitro-to-
nitrite rearrangement both in homogeneous and inhomogeneous
media. The photo-induced rearrangement is accelerated in inho-
mogeneous media (SDS, Brij35, ␤-CD and BSA solutions), which
provide hydrophobic environment. The solvent cage plays an
important role in photo-reduction with SDS, Brij35, ␤-CD and BSA.
The photo-reduction is significantly accelerated in the presence
of hydrogen donors, thereby increasing the yield of Product 2.
From the analysis of MFEs it is clearly demonstrated that hydrogen
abstraction reaction takes place in the triplet excited state of flu-
tamide. In homogeneous buffer solution, the hydrogen abstraction
reaction does not occur because no hydrogen donors are dissolved
in the solution.
R. Nakagaki, Y. Tanimoto, Y. Fujiwara, Chapter 3, Magnetic field and mag-
netic isotope effects on biradical reactions, in: S. Nagakura, H. Hayashi, T.
Azumi (Eds.), Dynamic Spin Chemistry–Magnetic Controls and Spin Dynamics
of Chemical Reactions, Kodansha, Tokyo, 1998, pp. 49–81.
[18] C.A. Parker, A new sensitive chemical actinometer I. Some trials with potassium
ferrioxalate, Proc. R. Soc. (Lond.) A 220 (1953) 104–116.
[19] C.G. Hatchard, C.A. Parker, A new sensitive chemical actinometer potassium
ferrioxalate as a standard chemical actinometer, Proc. R. Soc. (Lond.) A 235
(1956) 518–536.
[20] A. Yekta, M. Aikawa, N.J. Turro, Photoluminescence methods for evalu-
ation of solubilization parameters and dynamics of micellar aggregates.
Limiting cases which allow estimation of partition coefficients, aggre-
gation numbers, entrance and exit rates, Chem. Phys. Lett. 63 (1979)
543–548.
[21] S. Kato, S. Minagawa, M. Koizumi, Photoreduction of acridine in deaer-
ated and aerated ethanol solutions, Bull. Chem. Soc. Jpn. 34 (1961)
1026–1031.
The results obtained in the present study may contribute to a
better understanding of photoreactions of flutamide in the biologi-
cal condition by taking account of the similarity of inhomogeneous
media to biological environment.
[22] Chemical Soceity of Japan, Jikken-Kagaku-Kouza (Zoku) 11, Denshi Supekutoru,
6.2 Kagakuhannou-Jousuu-no-Sokutei, (Electronic Spectra, 6.2 Determination
of Reaction Rate Constants), Maruzen, Tokyo, 1965, pp. 552–557.
[23] R. Nakagaki, K. Mutai, Photophysical properties and photosubstitution and pho-
toredox reactions of aromatic nitro compounds, Bull. Chem. Soc. Jpn. 69 (1996)
261–274.
[24] Y.L. Chow, Chapter 6, Photochemistry of nitro and nitroso compounds,
in: S. Patai (Ed.), The Chemistry of Amino, Nitroso, and Nitro Com-
pounds and their Derivative, Part 1, Supplement F, Wiley, New York, 1982,
pp. 181–290.
Acknowledgement
We thank Mr. M. Saito of our laboratory for his synthesis of
Product 1 and Product 2.
References
[1] P. Schelhammer, R. Sharifi, N. Block, A controlled trial of bicalutamide versus
flutamide, each in combination with luteinizing hormone analogue therapy, in
patients with advanced prostate cancer, Urology 45 (1995) 745–752.
[25] W.M. Horspool, P.-S. Song (Eds.), CRC Handbook of Organic Photochemistry and
Photobiology, D. Doepp, 81 Photochemical Reactivity of the Nitro Group, CRC
Press, Boca Raton, 1995, pp. 1019–1062.