Photochemistry and Photobiology, 2002, 75(5): 488–494

The Effect of pH and Surfactant on the Aggregation Behavior of

Chlorin p6: A Fluorescence Spectroscopic Study¶
Anindya Datta*, Alok Dube, Beena Jain, Arjun Tiwari and Pradeep Kumar Gupta
Biomedical Applications Section, Centre for Advanced Technology, Indore, India

Received 23 October 2001; accepted 6 February 2002

ABSTRACT several photosensitizers (6,7). The interplay of photophysical

properties, structure and intracellular distribution has been
Steady state and time-resolved fluorescence properties of
examined in several recent studies (7–10).
chlorin p6, a potential drug for photodynamic therapy,
Chlorin derivatives are attractive photosensitizers for PDT
have been investigated as functions of pH. A decrease in
not only because they have strong absorption in the thera-
pH of the medium has been shown to cause protonation
peutic window but also because of the fact that the singlet
of the ionizable carboxylic acid side chain, leading to an
oxygen yield has been shown to be considerably high for
increase in hydrophobicity and consequent aggregation.
this class of compounds (11). This has led to the synthesis
The aggregates dissociate on further protonation. The
and study of various chlorin derivatives by several groups
dissociation is explained in terms of formation of cations
(7,12–16). Chlorin p6 and purpurin 18 (Fig. 1) are two such
and their mutual repulsion. A synchronous fluorescence
closely related compounds. On hydrolysis in alcoholic
spectroscopic study revealed the presence of two anionic
NaOH, the anhydride ring in purpurin 18 opens to form
forms in equilibrium at physiological pH, with a shift in
chlorin p6, which has the two carboxylic acid groups directly
the equilibrium on slight decrease in the pH. The anionic
bound to the aromatic skeleton (13). In phosphate buffered
nature of chlorin p6 in aqueous solutions at physiological
saline (PBS), at physiological pH, purpurin 18 exists as ag-
pH has been confirmed by complexation with surfactants.
gregates whereas chlorin p6 exists as monomers. This dif-
The nature of the charge on the headgroups of the sur-
ference in hydrophilicity is ascribed to the presence of polar,
factants has been found to govern the formation of chlor-
ionizable –COOH groups attached directly to the ring system
in–surfactant complexes.
of chlorin p6 and the lack of such groups in purpurin 18. It
is important to understand the nature of the species present
INTRODUCTION in aqueous solutions of these compounds at physiological
pH, because it governs hydrophilicity, lipophilicity, aggre-
Photodynamic therapy (PDT) for treatment of cancer has
gation and cellular uptake of the drugs (7,17). With this mo-
been a growing field of interest over the last 25 years (1–3).
tive, we have studied the effect of variation of pH on the
The technique is based on selective localization of photo-
absorption and fluorescence properties of chlorin p6, using
sensitizers in tumor cells and subsequent destruction of the
steady state and time-resolved fluorescence spectroscopy.
tumor cells by photoexcitation of the photosensitizers (2).
The presence of more than one anionic species has been
The desirable features of a photosensitizer for use in PDT
observed around physiological pH. The ionic nature of chlor-
are strong absorption in the so-called therapeutic window of
in p6 has been explored further by studying its interaction
650–900 nm, where penetration in the tissues is the maxi-
with surfactants.
mum (4), tumor selectivity and high quantum efficiency for
singlet oxygen generation. The last is affected by its state of
aggregation and primary photophysics. The intracellular dis- MATERIALS AND METHODS
tribution of the drug is mainly determined by the charge on Synthesis of photosensitizers and preparation of solutions. Purpurin
it (5). The increase in lipophilicity, induced by protonation, 18 was prepared from dry spinach leaves following the procedure
has been reported to result in a higher cellular uptake for of Hoober et al. (13). Chlorin p6 was prepared by alkaline hydrolysis
of purpurin 18 (13), which was purified by silica column and pre-
parative thin-layer chromatography (TLC) (Rf 5 0.56) before hy-
¶Posted on the web site on 22 February 2002. drolysis. The purity of chlorin p6 was tested by TLC, high-perfor-
*To whom correspondence should be addressed at: Biomedical Ap- mance liquid chromatography (HPLC) and superimposability of the
plications Section, Laser R&D Block D, Centre for Advanced absorption and fluorescence excitation spectra. Approximately 20
Technology, Indore 452 013, India. Fax: 91-731-488430; e-mail: mM solutions of chlorin p6 were prepared in UV spectroscopic grade
[email protected] methanol. Aliquots of 50 mL of these solutions were added to 10
Abbreviations: AR, analytical reagent; CMC, critical micellar con- mL of PBS solutions of different pH. HCl and NaOH used to vary
centration; CTAB, cetyl trimethyl ammonium bromide; HPLC, the pH were from E-Merck, Mumbai, India. The ionic strength of
high-performance liquid chromatography; PBS, phosphate buff- these solutions was maintained at 0.2 mM by addition of analytical
ered saline; PDT, photodynamic therapy; SDS, sodium dodecyl reagent (AR) grade NaCl from E-Merck. The AR grade surfactants
sulfate; TLC, thin-layer chromatography. cetyl trimethyl ammonium bromide (CTAB) and sodium dodecyl
q 2002 American Society for Photobiology 0031-8655/01 $5.0010.00 sulfate (SDS), obtained from SD Fine Chemicals, Mumbai, India,

488
 Photochemistry and Photobiology, 2002, 75(5) 489

Figure 1. Structures of purpurin 18 and chlorin p6.

and Spectrochem, Mumbai, India, respectively, were recrystallized

from alcohol–water mixtures before use.
Steady state absorption, fluorescence and synchronous lumines-
cence spectroscopies. The steady state absorption and emission
spectra were recorded on a GBC Cintra 20 absorption spectropho-
tometer and a Spex Fluorog fluorimeter, respectively. The excitation
and emission wavelengths were set at 355 and 725 nm for obtaining
the emission and excitation spectra, respectively. While recording
the synchronous luminescence spectra (18,19), the excitation and
emission monochromators were scanned simultaneously with an off-
set of 5 nm. A band pass of 1 nm was used for all spectral mea- Figure 2. Normalized absorption (—), fluorescence (······, lex 5 355
surements. Fluorescence was collected at a right angle, and a long- nm) and excitation (--, lem 5 725 nm) spectra of 0.7 3 1025 M
pass filter was used in front of the emission monochromator to cut chlorin p6 at different pH values. The absorption peak at 675 nm
off stray excitation light. observed between pH 5.5 and 1.8 is absent in the corresponding
The relative fluorescence quantum yields were calculated by nor- excitation spectra and are assigned to the aggregates of chlorin p6.
malizing the integrated fluorescence spectra with respect to the in- It should be noted that the peak normalization has been performed
tegrated fluorescence spectrum at pH 7.2, after correcting for vari- for the sake of comparison. In the non-normalized spectra, the ab-
ations in absorbance at the excitation wavelength. Spectral compo- sorption band at 675 nm grows as that at 640 nm decreases.
nent resolution was performed on the synchronous luminescence
spectra using a standard data fitting software, which used the sim-
plex algorithm. at pH 7.2 to 405 nm at pH 2.7 (data not shown). A blueshift
Time-resolved fluorescence studies. Time-resolved fluorescence
is observed in the longest wavelength absorption band, ac-
measurements were performed at magic angle using a UV–visible
single shot streak camera system (20). The instrument resolution companied by the development of a new band around 675
used was ;50 ps/channel. The excitation source used was the fre- nm. This band becomes most prominent at pH 2.7. At this
quency tripled output (355 nm, ;21 ps, 5 Hz) of an Nd:YAG laser pH, the absorption spectra show two prominent peaks at 642
(Continuum (Santa Clara, CA), model no. PY61C-10). Appropriate and 675 nm. The species absorbing at 675 nm is nonfluo-
filters were positioned immediately after the sample to avoid any
stray light. A small fraction (;4%) of the excitation pulse was also rescent, as is manifested in its absence in the excitation spec-
coupled to the streak camera slit directly to serve as the reference
for temporal and amplitude measurements of fluorescence from the
sample.

RESULTS AND DISCUSSION

The effect of pH on steady state absorption and
fluorescence properties
The absorption spectra of chlorin p6 exhibit a strong Soret
band near 400 nm and a weaker Q band between 600 and
700 nm, which is similar to other chlorin derivatives (6,13).
A very weak band is obtained around 500 nm. As shown in
Fig. 2, the longest wavelength bands of the absorption and
the excitation spectra have peaks at 655 nm and are super-
imposable for alkaline and neutral pH. The emission peak
remains at 662 nm (61 nm) throughout the alkaline–neutral
Figure 3. Variation of relative quantum yield (frel
f ) and fluorescence
pH range. The fluorescence quantum yield is fairly constant maximum (lem max) of 0.7 3 10
25
M chlorin p6 with pH. The inset
over this pH range (Fig. 3). On decreasing the pH gradually shows the second derivatives of the plots, whose zero-crossing
from 7.2, the Soret band undergoes a redshift from 401 nm points correspond to the points of inflection.
490 Anindya Datta et al.

tra (Fig. 2). On exciting at 675 nm, no fluorescence is ob-

served at any wavelength at any pH value. Excitation spectra
obtained at other emission wavelengths also lack this band.
The fluorescence spectra exhibit a progressive peak shift to
650 nm and decrease in relative quantum yield by an order
of magnitude at pH 2.7 (Fig. 3).
The trend of variation of the steady state spectral prop-
erties undergoes a reversal on decreasing the pH further from
2.7. The absorption band at 675 nm is found to decrease
progressively, and the difference in the absorption and ex-
citation spectra toward longer wavelengths gets considerably
reduced. The longest wavelength absorption peak shifts
gradually to 647 nm at a pH of 0.8, whereas the Soret band
is shifted to 408 nm. The emission peak moves back to 663
nm, and the relative quantum yield increases to 0.25. In or-
der to investigate the possibility of quenching by chloride
Figure 4. The variation of the ratio of absorbances at 690 and 630
ions present in the solutions, absorption and fluorescence nm with increase in concentration of chlorin p6 at pH 5.3.
spectra were recorded at pH 7.4 with addition of NaCl. No
perceptible change was observed for NaCl concentrations as
high as 0.1 M. variation of fluorescence quantum yield and peak position,
At physiological pH, chlorin p6 is expected to exist as an indicates dissociation of the aggregates below pH 2.7. This
anion, because of the presence of three carboxylic acid phenomenon can be explained by considering protonation of
groups. In fact, a mixture of anionic forms is expected to the nitrogen atoms of the chlorin ring. It has been reported
coexist in the physiological pH range, as has been indicated that protonation of the imino nitrogens of porphyrin deriv-
for chlorin e6 (7). The lack of pH-dependence in steady state atives takes place at pH 2–4 (17). The neutral chlorin p6
absorption and fluorescence spectral features between pH 7.2 molecules, on being protonated at the fairly high hydrogen
and 13 suggests that there is no further ionization in this ion concentration, acquire positive charges. The electrostatic
range. We propose that on decreasing the pH to the slightly repulsion between the like charges causes the aggregates to
acidic range (5–7), one or more of the carboxylate groups dissociate into monomeric cations. The cations being formed
get protonated. The decrease in negative charge induces in- at the expense of the aggregates, the fluorescence quantum
creased hydrophobicity, leading to aggregation and giving yield shows an increase at very low pH values. The plots of
rise to the absorption band at 675 nm. Such a band has been relative quantum yield as well as peak position against pH
reported for the closely related purpurin 18, which, lacking show a point of inflection around pH 2 (Fig. 3), which can
ionizable groups, exists exclusively as aggregates in neutral be related to the protonation of the nitrogen atoms in the
aqueous solutions (13). The nonfluorescent nature of the spe- chlorin ring. However, it should not be taken to be the pKa
cies that absorb at 675 nm is a strong indication of aggre- value, for reasons discussed already.
gation. It is common knowledge that aggregation usually In order to verify that the absorption peak at 675 nm is
leads to severe quenching of fluorescence (21,22). The de- indeed caused by the formation of aggregates, concentration
crease in fluorescence quantum yield of chlorin p6 is ascribed dependence studies were carried out at pH 5.3. The absorp-
partly to the decrease in population of its monomers. How- tion peak of the aggregate was found to increase with respect
ever, this does not seem to be the sole cause. The fluores- to the monomer as the concentration was increased, till a
cence spectra undergo a progressive blue shift with decrease concentration of about 5 3 1025 M of chlorin p6. This was
in pH in this range. This might indicate the formation of manifested in the variation of the ratio of the absorbances at
some new fluorescing species. It is very likely that the rel- 690 and 630 nm with chlorin p6 concentration (Fig. 4). Be-
ative population of the different anionic forms undergoes a yond this concentration, the rate of change of the ratio was
change in this pH range. The variations of quantum yield seen to fall off, indicating a saturation in aggregation.
and fluorescence peak positions with pH show points of in-
flection at pH 6 and 5, respectively (Fig. 3, inset). This dif-
Synchronous luminescence spectroscopy
ference is also an indication of the presence of more than
one fluorophore in this pH range. However, as has been The previous section highlights the existence of a mixture
pointed out by C̆underlı́ková et al. (7), it is not possible to of the different anionic forms as well as the aggregate in
determine the individual pK values in this system by simple slightly acidic solutions of chlorin p6. The difficulty of un-
absorption, emission and excitation spectroscopic tech- derstanding such a complex system by simple absorption and
niques. As the involvement of more than one fluorescent fluorescence spectroscopy has prompted us to perform syn-
species is suspected, it would probably be incorrect to iden- chronous luminescence experiments (18,19) in which the ex-
tify these points of inflection with the pKa values of the ionic citation and emission wavelengths are varied simultaneously
species. The fluorescence property of the mixture is affected with a constant offset. The resulting spectra have narrower
not only by a change in concentration of the components but peaks compared with conventional excitation spectroscopy,
also by their intrinsic quantum yield and peak position. allowing discrimination between fluorophores that have wide
The decrease in the characteristic absorption band of the and overlapping spectra. This method has previously been
aggregates, accompanied by the reversal in the trend in the utilized to analyze complex mixtures in as widely separated
 Photochemistry and Photobiology, 2002, 75(5) 491

Figure 6. Fluorescence decays of of 0.7 3 1025 M chlorin p6 at pH

0.8, 3.9, 5.6, 7.2 and 13.2. lex 5 355 nm. The decays are fitted to
a biexponential function. I 5 a1 exp(2t/t1) 1 a2 exp(2t/t2).

more reliable than the result obtained from the plots of fluo-
rescence maxima and quantum yield with pH because in this
case the variation of concentration of a single species has
been monitored. This technique is free from complications
arising out of a contribution from the other fluorescing com-
ponent of the mixture.

Time-resolved fluorescence studies

The pH dependence of the time-resolved fluorescence traces
is shown in Fig. 6. The decays at pH 7.2 and 13.2 are su-
perimposable. Those at pH 5.6 and 3.9 form another set of
superimposable decays, with faster decay constants than the
Figure 5. (a) Synchronous fluorescence spectra of 0.7 3 1025 M
chlorin p6 at pH 5.0, 5.6, 6.5, 7.4 and 9.0. (b) The spectrum at pH previous two. The decay at pH 0.8 is the fastest of all. The
5.6 (solid line), fitted as S 5 a1S1 1 a2S2, where S1 and S2 are the decays are fitted to a double-exponential function. It is well
spectra at 5.0 and 9.0, respectively, and a1 and a2 are the correspond- known that a biexponential fit can often be misleading. A
ing weighting factors. The weighted components are drawn in more complicated distribution of time constants can often be
dashed lines. The dotted line represents the fit. (c) The plot of the
fit satisfactorily by a simple biexponential function (23). A
second derivative of a2 against pH. The zero-crossing point corre-
sponds to the pKa value of chlorin p6. biexponential decay does not always indicate the presence
of two different fluorescent species, as is exemplified in the
case of tryptophan in neutral aqueous solutions, whose fluo-
fields as cancer diagnosis (18) and hydrocarbon analysis in rescence decay is biexponential (24). A biexponential decay
mineral oils (19). has earlier been observed for the porphyrin H2TmPy41 (25).
The synchronous luminescence spectra of chlorin p6 in the The intricate analysis and assignment of the two decay con-
pH range 5–9.2 are shown in Fig. 5a. The spectrum at pH stants in a system like ours is a problem in itself and is
5 shows a peak at 640 nm. The spectrum at pH 9.2 shows beyond the scope of this study. So, in this paper, we con-
a much more intense peak at 655 nm. The spectra of solu- centrate on the general trend, by considering the average
tions having intermediate pH are broad and appear to be lifetime, ^t&, defined as
weighted sums of the spectra at pH 5 and 9.2. Assuming the (a1 t1 1 a 2 t2 )
presence of two fluorophores in this pH range, one of which ^t& 5 ,
(a1 1 a 2 )
dominates at pH 5 and the other at pH 9.2, the spectra are
resolved into the two components. An excellent fit was ob- where a1, a2, t1 and t2 are the amplitudes and time constants
tained in all the cases, as is represented in Fig. 5b. The plots of the biexponential decay. The average lifetimes, listed in
of the partial contributions of the two components against Table 1, are greater in alkaline and neutral media than in
pH show an inflection point at pH 7 (Fig. 5c), which can be acidic media, much like the fluorescence quantum yield.
identified with the pK value of interconversion between two However, this similarity in trend is only qualitative. On de-
anionic forms. This result is consistent with the findings of creasing the pH from 7.2 to 5.6, the fluorescence quantum
C̆underlı́ková et al., who have reported inflection points yield becomes one-fourth, whereas the time constant is de-
around this pH range for chlorin e6, which is a very closely creased to 86% of the value at pH 13. This decrease is as-
related compound (7). In a recent publication, the same cribed partly to the formation of the anion with a lesser
group has reported an inflection point in the same pH range charge and partly to quenching because of aggregation. On
for hematoporphyrin IX, which has carboxylic acid groups further decrease of pH to 3.8, the lifetime remains constant,
in the side chain (17). The pK value obtained from the plot within experimental error, even as the relative quantum yield
of fits to the synchronous fluorescence spectra is probably decreases to 0.10. Thus, in this pH range, the decrease in
492 Anindya Datta et al.

Table 1. Comparison between steady state and time-resolved

emission parameters of chlorin p6 at different pH values

pH fem
rel * a1 t1 (ns)† t2 (ns)† ^t& (ns)‡

0.8 0.21 0.8 1.0 7.8 2.3

3.8 0.10 0.5 1.2 4.1 3.2
5.6 0.46 0.7 1.8 6.5 3.1
7.2 1.00 0.6 2.0 6.5 3.7
13.2 0.99 0.6 1.9 6.5 3.7

*Relative quantum yield.

†I(t) 5 a1t1 1 a2t2, a1 1 a2 5 1.
‡Average lifetime; ^t& 5 a1t1 1 a2t2.

fluorescence is ascribed to the formation of nonfluorescent

aggregates. No new fluorescent species are formed. On de-
creasing the pH to 0.8, however, the lifetime decreases in
spite of an increase in relative quantum yield. This is as-
cribed to the formation of the cations, which, being different
species altogether, fluoresce with a different characteristic
lifetime. Thus, the time-resolved results serve to establish
the hypothesis we had proposed from the steady state data.

Interaction with surfactants

Figure 7. (a) Fluorescence spectra of 0.7 3 1025 M chlorin p6 in
The anionic nature of chlorin p6 is established by studying PBS with and without surfactants. lex 5 355 nm. (b) Variation in
its interaction with surfactants. CTAB, a surfactant with a relative quantum yield of 0.7 3 1025 M chlorin p6 with surfactant
concentration.
positively charged headgroup, has a dramatic effect on the
spectral features of chlorin p6 at pH 7.2. On addition of
CTAB in concentrations as low as 0.02 mM (concentration yield remains more or less the same over the whole range
of chlorin p6 was 0.7 3 1025 M), the Soret and the Q bands of concentration of SDS. The peak-normalized fluorescence
shift from 401 to 405 nm and from 655 to 665 nm, respec- spectra of chlorin p6 in PBS and 40 mM CTAB and SDS
tively. The absorbance ratio of the Soret and the Q bands are shown in Fig. 7a. The variation of quantum yield with
decreases from 5.2 to 3.7. On subsequent addition of the surfactant concentration is illustrated in Fig. 7b.
surfactant, the band positions do not shift. The ratio of ab- The dramatic effect of CTAB on the fluorescence spectra
sorbances at the peaks too does not change any further, even of chlorin p6 can be attributed to the formation of complexes
when the concentration of CTAB is as high as 40 mM. SDS, of negatively charged chlorin p6 with the positively charged
however, does not have any effect on the absorption spectra surfactant molecules. The effect of the charge of the head-
at such low concentrations. The Soret band shifts to 404 nm group of the surfactant is evident in the fact that only CTAB
and the Q-band to 659 nm when the concentration of SDS and not SDS can produce the spectral changes. This phe-
is 9 mM. Thus, it is seen that the electrostatic interaction nomenon is qualitatively similar to the formation of porphy-
that is observed between chlorin p6 and CTAB is absent for rin-surfactant complexes reported by Periasamy and cowork-
SDS. However, there is definitely a marked hydrophobic in- ers (25,26). The variation of relative quantum yield with the
teraction between SDS and chlorin p6. This gives rise to the concentration of CTAB follows the trend of the variation
shift in the peak of the fluorescence spectrum. observed for H2TMPy41 on addition of SDS and that ob-
The effect of CTAB on the emission properties of chlorin served for the addition of CTAB to H2TPPS42 (25). It is
p6 is even more pronounced. In the presence of 0.02 mM possible that the complexes of chlorin p6 with CTAB might
CTAB, the emission peak gets broadened, and the relative consist of one or more forms of aggregates possible for such
quantum yield drops sharply to 0.2. The emission peak is systems, namely random, J- and H- aggregates (25–28).
shifted to 665 nm for the 0.04 mM CTAB solution, with the However, we refrain from commenting on the exact com-
quantum yield dropping further to 0.15. In the 0.06 mM position of the complexes at this point. The increase in fluo-
CTAB solution, the peak shifts to 667 nm, and the relative rescence quantum yield on addition of CTAB in concentra-
quantum yield hits a low of 0.1. On subsequent addition of tions more than 0.06 mM might be attributed to the micel-
CTAB, the peak position remains the same, within experi- lization of the chlorin p6–CTAB complexes. It is well known
mental error. However, the relative fluorescence quantum that many dyes, on incorporation in micelles, show a marked
yield of chlorin p6 increases till it reaches a plateau of 1.2 increase in quantum yield and fluorescence lifetime (29–31).
at the CTAB concentration of 0.65 mM. SDS does not have However, a close look at Fig. 7b reveals that the variation
such remarkable effects on the fluorescence spectral prop- in quantum yield reaches a plateau at a concentration that is
erties. The fluorescence peak does not shift to 665 nm until lower than the critical micellar concentration (CMC) of
the concentration of SDS reaches 1.4 mM. It shifts further CTAB (0.92 mM). Usually, any property that marks the for-
to 669 nm at higher concentrations of SDS. The quantum mation of micelles shows a sharp change around the CMC
 Photochemistry and Photobiology, 2002, 75(5) 493

(30,32). The increase in quantum yield at CTAB concentra- their excited states. The ground state and excited state pKa
tions higher than 0.06 mM clearly indicates that the chlorin– values are generally widely separated. For example, the ex-
surfactant complexes undergo some additional interaction cited state pKa value for 2-naphthol is 2, as compared with
with more surfactant molecules on increasing the surfactant the value of 9.2 for the ground state pKa (37). However, it
concentration. As the complexes are broken away from each is quite possible that for a system like ours, in which several
other by incorporation in these premicellar aggregates, there acidic and basic groups are present, closely spaced ground
is a drastic reduction in collisional quenching, and thus, the state and excited state acid–base equilibria might be in play.
quantum yield increases with increasing surfactant concen- The investigation of the extent of involvement of the excited
tration. The saturation in quantum yield that is observed state requires rigorous pump-probe and fluorescence upcon-
even before reaching the reported CMC for CTAB can be version studies in the femtosecond–picosecond time domain.
explained in two different ways. Firstly, it is possible that This is beyond the scope of the present study. At the mo-
chlorin p6 affects the process of micellization of CTAB and ment, though we cannot rule out the involvement of excited
decreases it. There have been reports of modifications in the state acid–base reactions, it is apparent that the ground state
apparent CMC by external agents like b-cyclodextrin (33). equilibria play a very major role in the pH dependence of
However, the modification of the CMC by a fluorophore, the electronic spectra of chlorin p6.
present at a concentration that is a couple of orders of mag-
nitudes lower than the surfactant concentration, is somewhat Acknowledgements A.D. thanks the Department of Atomic Energy
unlikely, especially to the extent observed in the present ex- for a Visiting Scientist position at CAT, Indore. Mr. S. K. Majumder
periment. It is more plausible that the premicellar aggregates and Mr. N. Ghosh are thanked for useful suggestions and discus-
sions. The spectral resolution software was kindly provided by Dr.
attenuate the collisional quenching to the maximal level even Kaustuv Das. The authors thank Dr. Mrinalini Sharma for help with
before the formation of the micelles, and so, the micelliza- the HPLC measurements.
tion process is not sensed by the fluorescence properties of
chlorin p6. Additional experiments involving light scattering REFERENCES
and steady state and time-resolved fluorescence anisotropies
are required before one can conclusively comment on this 1. Henderson, B. W. and T. J. Dougherty, Editors.(1992) Photo-
phenomenon. dynamic Therapy. Marcel Dekker Inc., New York.
2. Bellnier, D. A., Y.-K. Ho, R. K. Pandey, J. R. Missert and T.
J. Dougherty (1989) Distribution and elimination of photofrin
CONCLUSIONS II in mice. Photochem. Photobiol. 50, 221–228.
3. Oschner, M. (1997) Photophysical and photobiological process-
The formation and dissociation of aggregates of chlorin p6 es in the photodynamic therapy of tumors. J. Photochem. Pho-
with change in pH has been explained in terms of progres- tobiol. B Biol. 39, 1–18.
4. Troy, T. L. and E. M. Sevick-Muraca (1999) Fluorescence life-
sive protonation of the ionizable groups. The presence of time imaging and spectroscopy in random media. In Applied
more than one anionic form at physiological pH has been Fluorescence in Chemistry, Biology and Medicine (Edited by
established by steady state and time-resolved fluorescence W. Rettig, B. Strehmel, S. Schrader and H. Seifert), pp. 3–22.
experiments. The complexation of these anions with CTAB Springer, New York.
5. Modica-Napolitano, J. S., J. L. Goyal, G. Ara, A. R. Oscroff
has been demonstrated. The efficiency of synchronous lu-
and J. R. Aprille (1990) Mitochondrial toxicity of cationic pho-
minescence spectroscopy as a technique for analyzing a mix- tosensitizers for photochemotherapy. Cancer Res. 50, 7876–
ture of species with overlapping fluorescence spectra has 7881.
been exploited to analyze the mixture of ionic forms at phys- 6. Böhmer, R. M. and G. Morstyn (1985) Uptake of hematopor-
iological pH. It has been unambiguously shown that on de- phyrin derivative by normal and malignant cells: effect of se-
rum, pH, temperature and cell size. Cancer Res. 45, 5328–5334.
creasing the pH slightly from physiological pH the anion 7. Cunderlı́ková, B., L. Gangeskat and J. Moan (1999) Acid-base
with smaller charge and consequently higher hydrophobicity properties of chlorin e6: relation to cellular uptake. J. Photo-
predominates in the mixture. This shift in equilibrium ex- chem. Photobiol. B Biol. 53, 81–90.
plains, at least in part, the selective uptake of hydrophilic 8. Boyle, R. W. and D. Dolphin (1996) Structure and biodistri-
bution relationships of photodynamic sensitizers. Photochem.
chlorin-type molecules in tumor cells, which have been re-
Photobiol. 64, 469–485.
ported to have a lower intracellular pH in the range 5.8–7.2 9. Spizziri, P. G., J. S. Hill, S. B. Kahl and K. P. Ghiggino (1996)
(34–36). Photophysics and intracellular distribution of a boronated por-
It can be argued that the changes in the fluorescence spec- phyrin phototherapeutic agent. Photochem. Photobiol. 64, 975–
tral properties can be because of the excited state acid–base 983.
10. Weitman, H., S. Schatz, H. E. Gottelib, N. Kobayashi and B.
chemistry, rather than the ground state phenomena. How- Ehrenberg (2001) Spectroscopic probing of the acid-base prop-
ever, in our experiment, pH was observed to have a profound erties and photosensitization of a fluorinated phthalocyanine in
effect on the absorption spectra. This would suggest that organic solutions. Photochem. Photobiol. 73, 473–481.
ground state acid–base equilibria are very important in our 11. Krasnovsky, A. Jr. (1979) Photoluminescence of singlet oxygen
in pigment solution. Photochem. Photobiol. 29, 29–36.
case. If only the acid–base chemistry of the excited state
12. Nelson, J. S., W. G. Roberts and M. W. Berns (1987) In vivo
were involved, then we should have not seen much change studies on the utilization of mono-L-aspartyl chlorin (Npe6) for
in the absorption spectra but should have seen dramatic photodynamic therapy. Cancer Res. 47, 4681–4685.
changes in fluorescence. C̆underlı́ková et al. have earlier ex- 13. Hoober, J. K., T. W. Sery and N. Yamamoto (1988) Photody-
plained the pH variation of the absorption and fluorescence namic sensitizers from chlorophyll: purpurin-18 and chlorin p6.
Photochem. Photobiol. 48, 579–582.
spectra of chlorin e6 and other porphyrin derivatives by con- 14. Mironov, A. F., O. A. Efremova, R. I. Yakubovskaya, N. I.
sidering ground state acid–base reactions (7,17). It is well Kazachkina, G. I. Fomina, L. A. Shitova and N. N. Zharkova
known that acidic groups are usually even more acidic in (1996) Novel effective sensitizers for photodynamic cancer ther-
494 Anindya Datta et al.

apy on the base of chlorin p6 and its derivatives. SPIE Proc. aggregates of porphyrin-surfactant complexes: time-resolved
2924, 102–108. fluorescence and other spectroscopic studies. J. Phys. Chem. B
15. Brault, D., B. Aveline, O. Delgado and M. T. Martin (2001) 102, 1528–1538.
Chlorin-type photosensitizers photochemically derived from vi- 26. Maiti, N. C., S. Mazumdar and N. Periasamy (1996) Controlled
nyl porphyrins. Photochem. Photobiol. 73, 331–338. J-aggregation of porphyrins by cationic surfactants. Curr. Sci.
16. Di Stefano, A., A. Ettorre, S. Sbarna, C. Giovani and P. Neri 70, 997–999.
(2001) Purpurin-18 in combination with light leads to apoptosis 27. Maiti, N. C., M. Ravikanth and N. Periasamy (1995) Fluores-
or necrosis in HL60 leukemia cells. Photochem. Photobiol. 73, cence dynamics of noncovalently linked porphyrin dimers and
290–296. aggregates. J. Phys. Chem. 99, 17192–17197.
17. C̆underlı́ková, B., E. G. Bjørklund, E. O. Pettersen and J. Moan 28. Pasternack, R. F. and P. Collings (1995) Resonance light scat-
(2001) pH-Dependent Spectral Properties of HpIX, TPPS2a, tering: a new technique for studying chromophore aggregation.
mTHPP and mTHPC. Photochem. Photobiol. 74, 246–252. Science 269, 935–939.
18. Majumder, S. K. and P. K. Gupta (2000) Synchronous lumi- 29. Pal, S. K., A. Datta, D. Mandal and K. Bhattacharyya (1998)
nescence spectroscopy for oral cancer diagnosis. Lasers Life Sci. Photoisomerisation of diethyloxadicarbocyanine iodide in mi-
9, 143–152. celles. Chem. Phys. Lett. 288, 793–798.
19. Patra, D. and A. K. Mishra (2000) Effect of sample geometry 30. Sarkar, N., K. Das, S. Das, A. Datta, D. N. Nath and K. Bhat-
on synchronous fluorimetric analysis of petrol, diesel, kerosene tacharyya (1995) Excited-state intramolecular proton transfer of
and their mixtures at higher concentration. Analyst 125, 1383– 2-(29-hydroxyphenyl)benzimidazole in micelles. J. Phys. Chem.
1386. 99, 17711–17714.
20. Jain, B., P. K. Gupta, V. A. Podzyavnikov and V. K. Chevokin 31. Saroja, G. and A. Samanta (1995) Polarity of the micelle-water
(1996) Development and characterization of a UV-visible streak interface as seen by 4-aminophthalimide, a solvent sensitive
camera and its use for time resolved fluorescence studies on fluorescence probe. Chem. Phys. Lett. 246, 506–512.
human tissues. In Proceedings of the National Laser Sympo- 32. Moulik, S. P. (1996) Micelles: self-organized surfactant assem-
sium, B.A.R.C., Mumbai, India, January 17–19, 1996, National blies. Curr. Sci. 71, 368–376.
Laser Program, Department of Atomic Energy, Government of 33. Datta, A., D. Mandal, S. K. Pal, S. Das and K. Bhattacharyya
India, pp. C3–C4. (1998) Interaction of Triton X-100 with Cyclodextrins. A Fluo-
21. Mandal, D., S. K. Pal, D. Sukul and K. Bhattacharyya (1999) rescence Study. J. Chem. Soc. Faraday Trans. 94, 3471–3475.
Photophysical processes of merocyanine 540 in solutions and in 34. Newell, K., A. Franchi, J. Pouyssegur and I. Tannock (1993)
organized media. J. Phys. Chem. A 103, 8156–8159. Studies with glycolysis-deficient cells suggest that production of
22. Verkman, A. S. (1987) Mechanism and kinetics of merocyanine lactic acid is not the only cause of tumor acidity. Proc. Natl.
540 binding to phospholipid membranes. Biochemistry 26, Acad. Sci. USA 90, 1127–1131.
4050–4056. 35. Fuchs, C., R. Riesenberg, J. Siegert and R. Baumgartner (1997)
23. Ware, W. R. (1992) Recovery of fluorescence lifetime distri- pH-Dependent formation of 5-aminolaevulinic acid-induced
bution in heterogeneous system. In Photochemistry in Orga- protoporphyrin IX in fibrosarcoma cells. J. Photochem. Photo-
nized and Constrained Media (Edited by V. Ramamurthy), pp. biol. B Biol. 40, 49–54.
10010–10027. VCH Publisher Inc., New York. 36. Kennedy, K. A., J. D. McGurl, L. Leondaridis and O. Alabaster
24. Petrich, J. W., M. C. Chang, D. B. McDonald and G. Fleming (1985) pH-Dependence of mitomycin C-induced cross-linking
(1983) On the origin of nonexponential fluorescence decay of activity in EMT6 tumor cells. Cancer Res. 45, 3541–3547.
aqueous tryptophan and its derivatives. J. Am. Chem. Soc. 105, 37. Lakowicz, J. R. (1999) Excited-State reactions. In Principles of
3824–3832. Fluorescence Spectroscopy, second edn., pp. 516–517. Plenum
25. Maiti, N. C., S. Mazumdar and N. Periasamy (1998) J- and H- Publishers, New York.