Photochemical outcome modification of diflunisal by a novel cationic amphiphilic cyclodextrin

Article abstract of DOI:10.1039/b209157g

The effects of a novel cationic amphiphilic cyclodextrin (SC6CDNH2) on the photoprocesses of the phototoxic non-steroidal anti-inflammatory drug diflunisal (DF) have been investigated in aqueous medium. Association between DF and SC6CDNH2 was indicated by steady-state absorption, induced circular dichroism and fluorescence spectroscopy. Laser flash photolysis and steady-state photolysis experiments revealed a particular sensitivity of the DF photochemistry to the new microenvironment. Reduction of the triplet state generation efficiency, lengthening of its lifetime, changes in the photoionisation pathways, besides remarkable drug photostabilization and modification of the stable photoproducts distribution, were observed. A rationale for these modifications to the photochemistry of DF based on the multifaceted role of SC6CDNH2 in influencing the efficiency of the primary photochemical act and in interfering with secondary radical reactions is proposed. Relation of the overall results to the phototoxic effects displayed by the drug is also commented upon.

Full text of DOI:10.1039/b209157g

View Article Online / Journal Homepage / Table of Contents for this issue
Photochemical outcome modification of diflunisal by a novel cationic
amphiphilic cyclodextrin
Salvatore Sortino,*a Salvatore Petralia,a Raphael Darcy,b Ruth Donohueb and
Antonino Mazzaglia*c
a
`
Dipartimento di Scienze Chimiche, Universita degli Studi di Catania, Viale Andrea Doria 8,
I-95125, Catania, Italy. E-mail: ssortino@mbox.unict.it
Centre for Synthesis and Chemical Biology of the Conway Institute, Department of Chemistry,
National University of Ireland, University College Dublin, Belfield, Dublin 4, Ireland
b
c
`
Istituto per lo Studio dei Materiali Nanostrutturati (ISMN-CNR), Universita di Messina, Salita
Sperone 31, I-98166, Messina, Italy, and INFM, Unita di Messina, I-98166, Messina, Italy.
`
E-mail: mazzaglia@chem.unime.it
Received (in Montpellier, France) 18th September 2002, Accepted 21st November 2002
First published as an Advance Article on the web 28th January 2003
The effects of a novel cationic amphiphilic cyclodextrin (SC6CDNH2) on the photoprocesses of the phototoxic
non-steroidal anti-inflammatory drug diflunisal (DF) have been investigated in aqueous medium. Association
between DF and SC6CDNH2 was indicated by steady-state absorption, induced circular dichroism and
fluorescence spectroscopy. Laser flash photolysis and steady-state photolysis experiments revealed a particular
sensitivity of the DF photochemistry to the new microenvironment. Reduction of the triplet state generation
efficiency, lengthening of its lifetime, changes in the photoionisation pathways, besides remarkable drug
photostabilization and modification of the stable photoproducts distribution, were observed. A rationale for
these modifications to the photochemistry of DF based on the multifaceted role of SC6CDNH2 in influencing
the efficiency of the primary photochemical act and in interfering with secondary radical reactions is proposed.
Relation of the overall results to the phototoxic effects displayed by the drug is also commented upon.
entirely composed of neutral amphiphilic CDs. In these macro-
Introduction
cycles the non-polar moiety is represented by alkylthio chains
of varying length whereas the polar moiety is an oligo(ethylene
oxide), which is expected to reduce immunogenicity as well as
increase the overall colloidal stability of these CD aggregates,
much like ‘‘stealth’’ liposomes.18 Recently, we have synthe-
sised (from precursor heptakis[2-o-bromo-O-oligo(ethylene
Cyclodextrins (CDs) as drug complexing agents have been the
object of intense interest for both fundamental aspects and
practical purposes for a long time.1 Recently, this attention
has turned to the problem of biological photosensitisation by
drugs.2 Indeed, despite their excellent therapeutic activity,
many pharmacologically important chemicals such as antibac-
terials, antimicotics and non-steroidal anti-inflammatory drugs
(NSAIDs) can induce phototoxic, photoallergic and photomu-
tagenic phenomena strictly related to the drug photochemical
reactivity.3 It has been reported that in some cases such effects
can be substantially decreased in the presence of CDs with
model cellular systems.4–8 Application of CDs was, therefore,
suggested as a useful strategy to minimise the biological
damage induced by drugs and increase drug photostability.
However, is should be stressed that drug–CD complexes
usually dissociate once introduced into the body, where there
is also exposure to a wide range of endogenous species.7,9 Poly-
meric or supramolecular aggregates of CDs have the potential
to overcome these drawbacks. Indeed, CD aggregates can com-
bine macrocyclic and liposomal features to have release prop-
erties different from those of single CD molecules and bear a
close structural relationship to naturally occurring membranes.
It has been shown that amphiphilic CDs can display lyotropic
and thermotropic mesophases, form monolayers at the air–
water interface10 and also form micelles.11 Amphiphilic CDs
have been admixed with phospholipid monolayers,12 as well
as liposomes,13 and were dispersed in nanoparticles of pharma-
ceutical importance.14 This has opened up new possibilities for
their application as drug delivery systems. Recently, we have
prepared novel vesicles15,16 and micellar aggregates16,17
oxide)-6-hexylthio]-b-CD)
a
novel cationic amphiphilic
cyclodextrin, heptakis[2-o-amino-O-oligo(ethylene oxide)-6-
hexylthio]-b-CD, SC6CDNH2 (Chart 1).19a Dynamic light
scattering measurements showed that SC6CDNH2 forms
nanoaggregates in aqueous solutions within the size range of
120 to 1000 nm, at concentrations between 1 ꢀ 10ꢁ6 and
6 ꢀ 10ꢁ4 M.19b,c
This paper centres on the effects of SC6CDNH2 nanoparti-
cles on the photoprocesses of diflunisal (DF), 20,40-difluoro-4-
hydroxy-[1,10-biphenyl]-3-carboxylic acid. DF is one of the
most common NSAIDs used in the market and is known to
Chart 1
602
New J. Chem., 2003, 27, 602–608
DOI: 10.1039/b209157g
This journal is # The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2003

View Article Online
Absorption and induced circular dichroism spectra (ICD)
were recorded with a Beckman 650 DU spectrophotometer
and a Jasco J-715 dichrograph, respectively. Fluorescence
emission spectra were recorded with a Spex Fluorolog-2
(mod. F-111) spectrofluorimeter.
The reaction mixtures were analysed with a Hewlett-Pack-
ard LC-ESI-MS system equipped with an on-line photodiode
array detector (DAD) and a LiChroCart RP-18 column (5
mm packing, 4 ꢀ 250 mm, Hewlett Packard). The gradient used
for elution was CH3CN in 0.01 M phosphate buffer (pH 7)
from 0 to 75% in 25 min, at a flow rate of 1 mL minꢁ1
.
Irradiation experiments were performed using a Rayonet
photochemical reactor equipped with ‘‘black light’’ phosphor
lamps emitting in the 310–390 nm range with a maximum at
350 nm. The incident photon flux on 3 mL quartz cuvettes
was ca. 1 ꢀ 1016 quanta sꢁ1. The experimental procedures of
irradiation and the light intensity measurements have been
described previously.25
Scheme 1 Photodegradation mechanism of DF in neutral aqueous
solution.
The photodegradation of DF was determined by LC-ESI-
MS analysis, from the disappearance of the starting com-
pound, up to ca. 25% conversion. The retention times for
DF and photoproduct 1 were 13.7 and 12.8 min, respectively.
A quantitative evaluation of injected material in the presence
of SC6CDNH2 ensured that the column retained no inclusion
products.
induce phototoxic effects both in vitro20,21 and in vivo22 upon
sunlight absorption. The present investigation, carried out by
coupling steady-state and time-resolved photochemical techni-
ques, aimed to shed light on the potential of the SC6CDNH2
nanoparticles to be used as a photoprotective drug-carrier
systems.
For a better understanding of the results, we consider it use-
ful to mention briefly the main photochemical features of DF
in aqueous medium.
Nanosecond laser flash photolysis
The samples were excited with the fourth harmonic of a Nd-
YAG Continuum Surelite II-10 laser system (pulse width 6
ns FWHM, at l ¼ 266 nm) and the excited solutions were ana-
lysed at a right angle geometry using a mini mLFP-111 appa-
ratus developed by Luzchem Research. Briefly, the monitoring
beam was supplied by a ceramic xenon lamp and delivered
through quartz fibre optical cables. The laser pulse was probed
by a fibre that synchronised the mLFP system with a Tektronix
TDS 3032 digitiser operating in the pre-trigger mode. The sig-
nals from a compact Hamamatsu photomultiplier were initi-
ally captured by the digitiser and then transferred to a
personal computer that controlled the experiment with Luz-
chem software developed in the LabView 5.1 environment
from National Instruments. The energy of the laser pulse
was measured at each laser shot by a SPHD25 Scientech pyro-
electric energy monitor. Oxygen was removed by vigorously
bubbling the solutions with a constant flux of argon previously
passed through a water trap. The solution (in a flow cell of 1
cm pathlength) was renewed after each laser shot. The sample
temperature was 295 ꢃ 2 K.
Our preceding study concerning the steady-state photo-
chemistry of DF at neutral pH20 has shown that under UVA
excitation DF undergoes photodefluorination, an uncommon
reaction in fluoroaromatics due to the strength of the C–F
bond (dissociation energy ca. 523 kJ molꢁ1). The main
photoproduct isolated in anaerobic conditions was the 20-
(2000,4000-difluoro-300-carboxy-[100,1000-biphenyl]-400-oxy)-40-fluoro-
4-hydroxy-[1,10-biphenyl]-3 carboxylic acid, 1 (see Scheme 1).
The high toxicity displayed by 1 itself towards cell mem-
branes has led to this compound being identified as one of
the main species responsible for photoinduced damage.20
Our recent detailed report concerning the transient photo-
chemistry of DF in neutral aqueous solution has shed light
on the molecular mechanism of the drug’s photodegradation.23
As summarised in Scheme 1, the primary photochemical pro-
cess is photoionisation. Electron photoejection takes place
via mixed one- and two-photon pathways mediated by the
excited singlet state and generating a phenoxy radical. Defluor-
ination occurs via the radical anion, formed by a diffusion-con-
trolled trapping of the hydrated electrons by the DF ground
state. A s-aryl radical is produced after loss of the fluoride
anion. Formation of 1 is fully consistent with the occurrence
of a cross-combination reaction, almost exclusively controlled
by the Fisher–Ingold persistent radical effect,24 between the
phenoxy and the s radical centres.
Results and discussion
Association of DF with SC6CDNH2
The interaction between DF and SC6CDNH2 was charac-
terised by UV-VIS absorption, induced circular dichroism
(ICD) and fluorescence spectroscopy. Fig. 1 shows the absorp-
tion spectrum of DF recorded in the presence of increasing
amounts of SC6CDNH2. A remarkable decrease in the molar
absorption coefficient of the 260 nm band, corresponding to
transitions localised on the phenyl rings,26 accompanied by
an 8 nm red shift of the maximum upon addition of
SC6CDNH2, was observed. Lesser spectral modifications were
also noticed in the lower energy absorption band. Such beha-
viour compares with that exhibited by DF in the presence of
either b-cyclodextrin27 or micelles28 and provides a first indica-
tion of the association between drug and SC6CDNH2.
Experimental
Materials and methods
DF was purchased from Sigma Chemical Company (Milan,
Italy) and was used as received. Water was purified through
a Millipore Milli-Q system. All the experiments were per-
formed in 10ꢁ2 M phosphate buffer at pH 7.4. The pH of solu-
tions was measured with a glass electrode.
SC6CDNH2 solutions were prepared as follows: weighed
amounts of SC6CDNH2 (average molecular weight 2702) were
dissolved in chloroform and dried under nitrogen flux. To the
resulting film was added phosphate buffer before sonication at
40 ꢂC for 30 min.
A more direct proof of the host-guest complexation is
afforded by induced ICD measurements. DF is not optically
active. Addition of SC6CDNH2 induces optical activity.
A positive and a negative band around 260 and 310 nm,
New J. Chem., 2003, 27, 602–608
603

View Article Online
Fig. 3 Fluorescence spectra of 5 ꢀ 10ꢁ5 M DF observed in 10ꢁ2
M
phosphate buffer solution, pH 7.4 (a) in the absence and (b) in the pre-
sence of 0.25 mg mlꢁ1 SC6CDNH2. lexc ¼ 295 nm.
Fig. 1 Absorption spectra of 5 ꢀ 10ꢁ5 M DF in 10ꢁ2 M phosphate
buffer solution, pH 7.4 in the presence of 0.0, 0.03, 0.06, 0.12 and
0.25 mg mlꢁ1 SC6CDNH2. Cell path 1 cm. The spectra in the presence
of SC6CDNH2 were recorded using a reference solution containing the
same concentration of SC6CDNH2.
two different phenomena coming into play upon light excita-
tion (see Scheme 1): (1) a change in the dihedral angle
between the two planes of the phenyl rings (twisted in the
ground state and coplanar in the excited state), typical for
biphenyl derivatives;30 (2) an intramolecular proton transfer
between the hydroxyl group and the carboxyl moiety, typical
for salicylic acid derivatives.31,32 In view of this, it could be
suggested that the influence of the SC6CDNH2 microenviron-
ment on the rotational barrier of the phenyl rings, on the
excited state prototropic shift or on both could play a role
in the blue shift observed. In our recent work on DF-b-CD
complex we established that only the former is affected by
complexation since both the hydroxyl and the carboxyl moie-
tyes of DF are located virtually outside the cavity.27 Never-
theless, we believe that in the present case the above
photoinduced hydrogen shift could be influenced by
SC6CDNH2 complexation owing to a reasonable electrostatic
interaction between the carboxylate head of DF and the pro-
tonated amino groups of SC6CDNH2. With this in mind, a
binding mode consistent with the hydrophilic groups of DF
anchored to the positively charged external surface of the host
and the neutral aromatic chromophore protruding towards
the host interior and very close to the cyclodextrin cavity
can be tentatively proposed. The spectral changes observed
also in the lower energy absorption band (see Fig. 1), as well
as the modification in the photoionisation pathways (see
below), further support this proposal.
respectively, arise indeed as a consequence of complexation
with the chiral macrocycle (Fig. 2). The ICD bands are charac-
terised by maxima corresponding to those of the absorption
bands, ruling out the occurrence of exciton splitting phenom-
ena.29 The presence of the ICD signals affords some hints on
the probable binding mode of DF with the host. Since these
bands are in general strictly dependent on the distance between
the host and the guest, their appearance indicates that the DF
molecule is localised close to the chiral cavity of SC6CDNH2.
The addition of SC6CDNH2 also affects the fluorescence
emission of the drug. As can be seen in Fig. 3, the fluores-
cence intensity of DF is decreased and the emission maximum
blue-shifted ca. 5 nm. This behaviour is also similar to that
observed in the presence of either b-cyclodextrin27 or
micelles.28 While the decrease of the fluorescence intensity
could be simply due to a polarity effect (both solutions are
optically matched at the excitation wavelength), the shift
towards high energy deserves some further comment. A recent
study on the excited-state properties of DF in aqueous med-
ium23 showed a large Stokes shift between the absorption
and the emission maxima, pointing to a large geometry differ-
ence between the ground and the relaxed excited state. It was
demonstrated that such a large Stokes shift can be ascribed to
Laser flash photolysis studies
The transient spectra recorded upon 266 nm laser excitation in
nitrogen-saturated solution containing DF (5 ꢀ 10ꢁ5 M) in the
presence of SC6CDNH2 (0.25 mg mlꢁ1) were very similar to
those observed in its absence, suggesting that no new DF tran-
sient species are generated besides those observed in aqueous
solution. In fact, we monitored the typical band of the DF tri-
plet centred at 380 nm, the broad absorption extending
beyond 700 nm due to the hydrated electrons and the two
bands at 340 and 540 nm of the phenoxy radical originating
upon photoionisation.23 Despite these spectral similarities,
both kinetic behaviour and photogeneration efficiency of the
transients were affected by the presence of the macrocycle.
Fig. 4 shows the triplet decay observed in the presence and
in the absence of SC6CDNH2. The kinetic trace obtained in
the former case was well-fitted by a mono-exponential func-
Fig. 2 ICD spectrum of 5 ꢀ 10ꢁ5 M DF observed in 10ꢁ2 M phos-
phate buffer solution, pH 7.4 in the presence of 0.25 mg mlꢁ1
SC6CDNH2. Cell path 1 cm. The spectrum was recorded using a refer-
ence solution containing the same concentration of SC6CDNH2.
tion with a rate constant of kT(SC6CDNH2) ¼ 5.5 ꢀ 104 sꢁ1
.
Such a value is substantially smaller than that observed in
aqueous medium [kT(w) ¼ 1.6 ꢀ 105
s
ꢁ1],23 ruling out a fast
604
New J. Chem., 2003, 27, 602–608

View Article Online
of the lowest triplet state of DF. Each slope is proportional
to the product FISC ꢀ eT-T , where FISC and eT-T are the ISC
quantum yield and the absorption coefficient of the lowest
excited triplet state of DF, either in the absence or in the pre-
sence of SC6CDNH2. Taking into account that the solutions
are almost optically matched at the excitation wavelength
(the difference in absorbtion by the two samples is less than
3%) and that a large change in eT-T upon SC6CDNH2 com-
plexation is fairly unlikely (no significant differences in the
eT-T of DF were in fact observed upon DF incorporation in
either a CD cavity or micellar cage), the ratio of the slopes
reported in Fig. 5 points to a ca. 50% reduction of the ISC
quantum yield in the DF–SC6CDNH2 complex.
The modifications induced by SC6CDNH2 in the properties
of the DF triplet state compare well with those noticed for
other anti-inflammatory drugs included in b-cyclodextrin cav-
ities2 and are of important photobiological significance. Long-
lived triplet states of several non-steroidal anti-inflammatory
drugs are directly involved in bimolecular energy and electron
transfer processes with oxygen and biomolecule components.
These photoprocesses often generate noxious species such as
singlet oxygen, superoxide anion, charged and neutral radicals,
etc., which are directly responsible for photoinduced disor-
ders.3 On the basis of the confinement of the DF triplet within
the SC6CDNH2 network and the lower ISC yield, a decrease
in the efficiency of some of these bimolecular reactions20,21 is
expected.
As mentioned in the introduction, DF photoionisation is not
only the main photoprocess occurring upon light excitation
but is also responsible for the stimulated DF defluorination
through electron trapping. Fig. 6 shows the kinetic traces for
hydrated electron decay in the presence and in the absence
of SC6CDNH2. It shows that electron decay is faster in the
former case [time constant teaq(SC6CDNH2) ca. 380 ns compared
with teaq(w) ¼ 890 ns]. Although this finding could be attribu-
ted to additional quenching processes involving the participa-
tion of SC6CDNH2 itself, we believe that this might not be
the case. It is well-known that hydrated electrons (i) are not
expected to react at all with the cyclodextrin unit of the host27
and (ii) are characterised by a reactivity with the cationic
amino groups of the SC6CDNH2 periphery about three
orders of magnitude smaller than that observed with the DF
ground state in aqueous solution.34 Rather, the faster electron
capture observed in the presence of SC6CDNH2 seems to be
Fig. 4 Kinetic traces for the decay of the DF triplet state at 380 nm
M
observed in an argon-saturated 5 ꢀ 10ꢁ5 M DF solution in 10ꢁ2
phosphate buffer, pH 7.4 (K) in the absence and (S) in the presence
of 0.25 mg mlꢁ1 SC6CDNH2.
exit of the excited triplet state from the SC6CDNH2 structure.
The lengthening of their decay is not uncommon for triplets
upon incorporation in organised and constrained media such
as micelles and cyclodextrins and is generally related to the
protection exerted by the host cage against external quenching
impurities and to a perturbation of the intersystem crossing
(ISC) process to the ground state.2,33 However, inhibition of
triplet state-ground state interactions due to incorporation of
only one molecule per SC6CDNH2 molecule can also
be responsible for the behaviour observed. The above hypoth-
eses agree with the lack of any effect on the triplet lifetime
observed in organic solvents,28 suggesting that the lengthening
observed in our case is not correlated to changes of the local
micropolarity.
Fig. 5 shows the triplet absorbance as a function of the laser
intensity in the presence and in the absence of SC6CDNH2.
The linearity observed agrees with monophotonic generation
Fig. 5 Laser intensity dependence for the formation of the DF triplet
state observed in an argon-saturated 5 ꢀ 10ꢁ5 M DF solution in 10ꢁ2
M phosphate buffer, pH 7.4 (K) in the absence and (S) in the presence
of 0.25 mg mlꢁ1 SC6CDNH2. DA taken 0.1 ms after the laser pulse.
(No effect of laser power on the triplet lifetimes of the two samples
was observed in the energy range explored).
Fig. 6 Kinetic traces for the decay of the solvated electrons at 720 nm
M
observed in an argon-saturated 5 ꢀ 10ꢁ5 M DF solution in 10ꢁ2
phosphate buffer, pH 7.4 (K) in the absence and (S) in the presence
of 0.25 mg mlꢁ1 SC6CDNH2.
New J. Chem., 2003, 27, 602–608
605

View Article Online
due to its incomplete exposure. This view is consistent with our
proposal of electrostatic interaction between DF and the
charged SC6CDNH2 surface (see above). This result notably
differs from that observed when DF is incorporated in either
b-cyclodextrin or neutral micelles. In those cases, no reduction
in the monophotonic photoionisation yield was observed,
in accordance with localisation of the hydroxyl group of DF
outside the host cavity.27,28
Steady-state photolysis studies
When DF was irradiated in the presence of SC6CDNH2 in an
argon-saturated water solution, the photochemical scenario
was drastically changed compared to that in aqueous solution.
HPLC-ESI-MS analysis of the irradiated mixtures performed
within 25% conversion revealed, in fact, remarkable changes
in both the efficiency and nature of the photochemical out-
come. Fig. 8 shows the transformation percentage observed
in the presence and in the absence of SC6CDNH2.
Since the amount of incident radiation was virtually the
same in both cases (the difference was less than 3%), from
the ratio of the slopes obtained one can readily notice that a
ca. 4-fold reduction of photodegradation is achieved in the
presence of SC6CDNH2. Furthermore, formation of the
cross-combination product 1, the main product in aqueous
solution, was totally suppressed. This result, beyond its photo-
chemical interest, is remarkable in view of the toxicity of 1
mentioned above.20,21
Fig. 7 Laser intensity dependence for the formation of solvated elec-
trons observed in an argon-saturated 5 ꢀ 10ꢁ5 M DF solution in 10ꢁ2
M phosphate buffer, pH 7.4 (K) in the absence and (S) in the presence
of 0.25 mg mlꢁ1 SC6CDNH2. DA taken 0.05 ms after the laser pulse.
(No effect of laser power on the hydrated electron lifetimes of the
two samples was observed in the energy range explored).
gouverned by the typical scavenging mechanism that comes
into play when solutes are incorporated into cationic
micelles.34,35 As proposed in such cases, hydrated electrons
may undergo rapid trapping in the external cationic field such
as that of SC6CDNH2 and subsequent reaction with the
complexed DF.
To gain insight into the influence of the SC6CDNH2 micro-
environment on the photoionisation process, we measured the
laser power effect on the absorbance, monitored at 720 nm
(maximum of the hydrated electrons)36 in the presence and
in the absence of SC6CDNH2. Even in this new environment
photoionisation takes place through a mixture of mono- and
biphotonic mechanisms, similarly to that already observed in
aqueous medium.23 As shown in Fig. 7, the experimental data
fit well the following equation:
Plausible reasons for the enhanced photostability of the
DF-SC6CDNH2 complex, as well as for the suppression of
the persistent radical effects leading to product 1, may lie in
the peculiar role of SC6CDNH2 in (i) reducing the efficiency
of the one-photon photoionisation pathway and (ii) providing
a suitable microenvironment to efficiently trap the two radical
intermediates involved in the cross-combination reaction. A
simplified description of the probable processes is proposed
in Scheme 2 and discussed below.
On the basis of the two-fold reduction of the monophotonic
photoionisation route, which dominates under the low light
intensity of the steady-state irradiation, a similar reduction
of the DF photodegradation would have been expected. Sur-
prisingly, the reduction in DF photodegradation was greater
than that of monophotonic photoionisation. This apparent
DA=E ¼ a þ bE
where E is the laser intensity, a is a coefficient depending on the
quantum yield of the one-photon process and b is a factor
depending on the extinction coefficients and yields of the inter-
mediate steps in the two-photon process.37
From the slopes and the intercepts of the linear plots it is
straightforward to note that accommodation of DF in the
SC6CDNH2 network leads to a 50% drop in both the bipho-
tonic and monophotonic photoionisation efficiencies. How-
ever, it has to be emphasised that the one-photon pathway
dominates under environmental conditions due to the low
intensity of irradiation. In general, quantum yields for the
one-photon photoionisation pathways involving phenol moi-
eties are expected to be particularly sensitive to the surround-
ing environment and the structure of the solvent around the
OH groups is known to play a dominant role. In particular,
the formation of H-bonded supramolecular structures in which
the OH binds two water molecules by acting as both H-donor
and H-acceptor is believed to be the key step in determining
the efficiency of the monophotonic yield.38,39 Elegant studies
performed by Monti et al. highlighted a reduction in the effi-
ciency of the one-photon photoionisation pathway of some
phenol derivatives when the hydroxyl groups are deeply incor-
porated within cyclodextrin cavities.40 Likewise, reduction of
the yield in our case may be due to hindered formation of
the H-bond between water and the hydroxyl of DF, probably
Fig. 8 Percent photodegradation of DF observed in an argon-satu-
rated 5 ꢀ 10ꢁ5 M DF solution in 10ꢁ2 M phosphate buffer, pH 7.4
(K) in the absence and (S) in the presence of 0.25 mg mlꢁ1
SC6CDNH2.
606
New J. Chem., 2003, 27, 602–608

View Article Online
Scheme 2 Proposed photodegradation pathways of DF in neutral aqueous solution in the presence of SC6CDNH2.
discrepancy might be explained on the basis of a H-abstraction
reaction of the phenoxy radical from the SC6CDNH2 struc-
ture. Indeed, such a process would lead back to the starting
compound, accounting for the observed results. This hypoth-
esis is supported by the fact that in view of the more hydropho-
bic character of the phenoxy radical compared to DF, it is
reasonable to assume that it tends to remain within the host
microenvironment. The low yield of the phenoxy radical
observed in our case (data not shown), similarly to the
DF-b-CD complex, is consistent with fast disappearance of
this intermediate within the SC6CDNH2 network. Although
phenoxy radicals are generally characterised by low reactivity
for H-abstraction,41 it is widely reported that they can become
much more reactive in the presence of good hydrogen sources
such as cyclodextrins, with 14 available hydrogen atoms
bonded to secondary carbons and in close proximity to the
radical center.42–44
The total inhibition of the cross-combination product 1 can
also be related to an intermolecular reaction involving the s-
aryl radical generated upon defluorination. Given the high
H-abstraction reactivity of this radical species,45,46 it is
expected to react promptly with SC6CDNH2 before exiting
and to convert to the monofluorinated photoproduct 2. Unfor-
tunately, it was not possible to isolate such a stable product on
a preparative scale nor to gain direct evidence for the s-aryl
radical, due its well-known optical transparency in the moni-
tored wavelength range.45,46 However, the appearance of one
major peak at m/z ¼ 231, in accordance with the structure
of 2, besides that of the starting compound (m/z ¼ 249) in
the LC-ESI-MS analysis of the irradiated mixture, is in agree-
ment with the above proposal.
confined within the SC6CDNH2 and are quenched through
intermolecular H-abstraction The overall photobehaviour of
DF in the SC6CDNH2 microenvironment seems gouverned
by a synergism of effects due to the multifaceted role
of SC6CDNH2 in influencing the primary photochemical
events and secondary radical reactions of this drug. The
overall results make this tailored cyclodextrin an interesting
system for increasing drug photostability and minimising the
biological damage. The results of in vitro photosensitisation
experiments to elucidate this issue will be reported in due
course.
Acknowledgements
Financial supports from MURST ‘‘cofinanziamento di pro-
grammi di ricerca di rilevante interesse nazionale’’ and INCA
(Consorzio Interuniversitario per la Chimica dell’Ambiente)
are gratefully acknowledged.
References
1
(a) J. Szejtli, Chem. Rev., 1998, 98, 1743; (b) V. T. D’Souza and
K. B. Lipkowitz, Chem. Rev., 1998, 98, 1741; (c) K. Uekama,
F. Hirayama and T. Irie, Chem. Rev., 1998, 98, 2045; (d ) J. A.
Arancibia, M. A. Boldrini and G. M. Escandar, Talanta, 2000,
52, 261; (e) J. A. Arancibia and G. M. Escandar, Analyst, 1999,
124, 1833.
2
3
S. Monti and S. Sortino, Chem. Soc. Rev., 2002, 31, 287.
(a) F. Bosca, M. L. Marin and M. A. Miranda, Photochem. Photo-
biol., 2001, 74, 637; (b) G. Condorelli, L. L. Costanzo, G. De
Guidi, S. Giuffrida, P. Miano, S. Sortino and A. Velardita, EPA
´
Newslett., 1996, 58, 60; (c) F. Bosca and M. A. Miranda, J. Photo-
chem. Photobiol., B, 1998, 43, 1.
4
5
K. Uekama, T. Irie and F. Hirayama, Chem. Lett., 1978, 1109.
T. H. Hoshino, K. Ishida, T. Irie, F. Hirayama and M. Yamasaki,
J. Inclusion Phenom., 1988, 6, 415.
Conclusions
6
G. De Guidi, G. Condorelli, S. Giuffrida, G. Puglisi and
G. Giammona, J. Inclusion Phenom. Mol. Recognit. Chem.,
1993, 15, 43.
M. Partyka, B. H. Au and C. H. Evans, J. Photochem. Photobiol.,
A, 2001, 140, 67.
S. Sortino, J. C. Scaiano, G. De Guidi and S. Monti, Photochem.
Photobiol., 1999, 70, 549.
V. J. Stella, V. M. Rao, E. A. Zabbou and V. Zia, Adv. Drug
Delivery Rev., 1999, 36, 3.
This study has provided the first example of how a new catio-
nic amphiphilic cyclodextrin can influence significantly the
photochemistry of a photosensitising drug. The triplet state
of DF, playing an important role in the photosensitising
side-reactions, is generated with lower efficiency and remains
trapped within the SC6CDNH2 network. The photoreactivity
of the associated guest is lowered and the product distribution
is drastically modified. The photoproduct mainly responsible
for the toxic effects is completely suppressed in the presence
of SC6CDNH2. The two radical species generated upon
photoexcitation, the phenoxy and s-aryl radicals, remain
7
8
9
10 (a) Y. Kawabata, M. Matsumoto, M. Tanaka, H. Takahashi,
Y. Irinatsu, S. Tamura, W. Tagaki, N. Nakahara and K. Fukuda,
Chem. Lett., 1986, 1933; (b) H. Parrot-Lopez, C.-C. Ling,
P. Zhang, A. Baszkin, G. Albrecht, C. de Rango and A. W.
New J. Chem., 2003, 27, 602–608
607

View Article Online
Coleman, J. Am. Chem. Soc., 1992, 114, 5479; (c) P. C.
Tschoreloff, M. M. Boissonnade, A. W. Coleman and A. Baszkin,
Langmuir, 1995, 11, 191; (d ) M. H. Greenhall, P. Lukes,
R. Kataky, N. E. Agbor, J. P. S. Badyal, J. Yarwood, D. Parker
and M. C. Petty, Langmuir, 1995, 11, 3997; (e) D. P. Parazak,
A. R. Khan, V. T. D’Souza and K. J. Stine, Langmuir, 1996,
23 S. Sortino, J. C. Scaiano and G. Condorelli, J. Phys. Chem. B,
1999, 103, 9279.
24 (a) H. Fisher, J. Am. Chem. Soc., 1986, 108, 3925; (b) D. Griller
and K. U. Ingold, Acc. Chem. Res., 1976, 9, 13.
25 G. De Guidi, R. Chillemi, L. L. Costanzo, S. Giuffrida, S. Sortino
and G. Condorelli, J. Photochem. Photobiol., B, 1994, 23, 125.
26 I. Berlman, J. Chem. Phys., 1970, 52, 5616.
27 S. Sortino, S. Giuffrida, S. Fazio and S. Monti, New J. Chem.,
2001, 25, 707.
28 S. Sortino, G. De Guidi, S. Giuffrida, A. Belvedere and
G. Condorelli, J. Phys. Chem. A, 2000, 104, 11 918.
29 (a) P. Bortolus, G. Marconi, S. Monti, B. Mayer, G. Kohler and
G. Grabner, Chem.-Eur. J., 2000, 6, 1578; (b) P. Bortolus,
G. Marconi, S. Monti, G. Grabner and B. Mayer, Phys. Chem.
Chem. Phys., 2000, 2, 2943.
´
12, 4046; ( f ) B. Hamelin, L. Jullien, A. Laschewsky and C. Herve
du Penhoat, Chem.-Eur. J., 1999, 5, 546.
´
´
11 R. Auzely-Velty, F. Djeda¨ıni-Pilard, S. Desert, B. Perly and
T. Zemb, Langmuir, 2000, 16, 3727.
12 (a) A. W. Coleman and A. Kasselouri, Supramol. Chem., 1993, 1,
155; (b) A. Kasselouri, A. W. Coleman and A. Baszkin, J. Colloid
Interface Sci., 1996, 180, 384; (c) A. Kasselouri, A. W. Coleman,
G. Albrecht and A. Baszkin, J. Colloid Interface Sci., 1996,
180, 398.
13 (a) L. Jullien, T. Lazrak, J. Canceill, L. Lacombe and J. M. Lehn,
J. Chem. Soc., Perkin Trans. 2, 1993, 1011; (b) J. Lin,
C. Creminon, B. Perly and F. Djeda¨ıni-Pilard, J. Chem. Soc.,
30 E. C. Lim and Y. H. Li, J. Chem. Phys., 1970, 52, 6416.
31 A. Weller, Electrochemistry, 1956, 60, 1144.
32 P. J. Kovi, C. L. Miller and S. G. Schulman, Anal. Chim. Acta,
1972, 61, 7.
´
Perkin Trans. 2, 1998, 2639; (c) R. Auzely-Velty, B. Perly, O.
´
´
Tache, T. Zemb, P. Jehan, P. Guenot, J. P. Dalbiez and F.
Djeda¨ıni-Pilard, Carbohydr. Res., 1999, 318, 82; (d ) S. Lesieur,
D. Charon, P. Lesieur, C. Ringard-Lefebvre, V. Muguet,
D. Duchene and D. Wouessidjewe, Chem. Phys. Lipids, 2000,
106, 127.
33 (a) P. Bortolus and S. Monti, Adv. Photochem., 1996, 21, 1; (b)
K. Kalyanasundaram, Photochemistry in Microheterogeneous Sys-
tems, Academic Press, Orlando, 1987; (c) M. H. Kleinman and C.
Bohne, in Organic Photochemistry: Molecular and Supramolecular
Photochemistry, ed. V. Ramamurthy and K. S. Schanze, Marcel
Dekker Inc., New York, 1997, pp. 391–466; (d ) J. K. Thomas,
Chem. Rev., 1980, 80, 283.
ˆ
14 (a) M. Skiba, C. Morvan, D. Duchene, F. Puisieux and D.
Wouessidjewe, Int. J. Pharm., 1995, 126, 275; (b) M. Skiba, D.
ˆ
Duchene, F. Puisieux and D. Wouessidjewe, Int. J. Pharm.,
34 M. Gra¨tzel, J. K. Thomas and L. K. Patterson, Chem. Phys. Lett.,
1974, 29, 393.
1996, 129, 113; (c) A. Gulik, H. Delacroix, D. Wouessidjewe
and M. Skiba, Langmuir, 1998, 14, 1050; (d ) E. Lemos-Senna,
ˆ
D. Wouessidjewe, D. Duchene and S. Lesieur, Colloids Surf., B,
35 J. H. Fendler and L. K. Patterson, J. Phys. Chem., 1970, 74, 190.
36 G. L. Hug, Optical Spectra of Non-Metallic Inorganic Transient
Species in Aqueous Solution, Natl. Stand. Ref. Data Ser. 69,
National Bureau of Standards, Washington, D.C., 1981.
37 U. Lachish, A. Shaffermann and G. Stein, J. Chem. Phys., 1976,
64, 4205.
38 G. Kohler and N. Getoff, Chem. Phys. Lett., 1974, 26, 525.
39 J. Zechner, G. Kohler, G. Grabner and N. Getoff, Can. J. Chem.,
1980, 58, 2006.
1998, 10, 291.
15 A. Mazzaglia, R. Donohue, B. J. Ravoo and R. Darcy, Eur. J.
Org. Chem., 2001, 1715.
16 B. J. Ravoo and R. Darcy, Angew Chem., Int. Ed., 2000, 39,
4324.
17 A. Mazzaglia, B. J. Ravoo, R. Darcy, P. Gambadauro and
F. Mallamace, Langmuir, 2002, 18, 1945.
18 D. D. Lasic, Angew Chem., Int. Ed. Engl., 1994, 33, 1685.
19 (a) R. Donohue, N. Mazzaglia, B. J. Ravoo and R. Darcy, Chem.
Commun., 2002, 2864; (b) A. Mazzaglia, L. Monsu` Scolaro,
R. Darcy, R. Donohue and B. J. Ravoo, J. Inclusion Phenom.
Macrocycl. Chem. , in press; (c) A. Mazzaglia, N. Angelini,
R. Darcy, R. Donohue, D. Lombardo, N. Micali, M. T. Sciortino,
V. Villari, L. Monsu` Scolaro, submitted.
40 S. Monti, G. Kohler and G. Grabner, J. Phys. Chem., 1993, 97,
13 011 M.
41 V. W. Bowry and K. U. Ingold, Acc. Chem. Res., 1999, 32, 27 and
references therein.
42 S. Symala, B. Nageswer Rao and V. Ramamurthy, Tetrahedron,
1988, 44, 7234.
43 A. V. Veglia, A. M. Sanchez and R. H. de Rossi, J. Org. Chem.,
1990, 55, 4083.
44 A. V. Veglia and R. H. de Rossi, J. Org. Chem., 1993, 58,
4941.
45 J. C. Scaiano and L. C. Stewart, J. Am. Chem. Soc., 1983, 105, 3609.
46 V. Madhavan, R. H. Schuler and R. W. Fessenden, J. Am. Chem.
Soc., 1978, 100, 888.
20 G. De Guidi, R. Chillemi, S. Giuffrida, G. Condorelli and
´
M. Cambria Fama, J. Photochem. Photobiol., B, 1991, 10, 221.
21 L. L. Costanzo, G. De Guidi, S. Giuffrida, P. Miano, A. Velardita
and S. Sortino, Photochem. Photobiol., 1996, 63S, 72.
22 B. Ljunggren and K. Lundberg, Photodermatology, 1985, 2,
377.
608
New J. Chem., 2003, 27, 602–608