Indole substituted zinc phthalocyanine: Improved photosensitizing ability and modified photooxidation mechanism

Article abstract of DOI:10.1016/j.jphotochem.2011.10.008

The photophysical and photochemical processes within a novel photosensitizer (PS), zinc phthalocyanine (ZnPc) modified by indole units, were explored. The properties related to photodynamic therapy of tumor (PDT) were studied by time-resolved transient UV

Full text of DOI:10.1016/j.jphotochem.2011.10.008

Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
Contents lists available at SciVerse ScienceDirect
Journal of Photochemistry and Photobiology A:
Chemistry
journal homepage: www.elsevier.com/locate/jphotochem
Indole substituted zinc phthalocyanine: Improved photosensitizing ability and
modified photooxidation mechanism
Xian-Fu Zhanga,b,∗, Wenfeng Guoa
a Chemistry Department & Center of Instrumental Analysis, Hebei Normal University of Science and Technology, Qinghuangdao, Hebei Province 066004, China
b MPC Technologies, Hamilton, Ontario, Canada L8S 3H4
a r t i c l e i n f o
a b s t r a c t
Article history:
The photophysical and photochemical processes within a novel photosensitizer (PS), zinc phthalocyanine
(ZnPc) modified by indole units, were explored. The properties related to photodynamic therapy of tumor
(PDT) were studied by time-resolved transient UV–vis absorption spectra, steady state and time-resolved
fluorescence spectra, and chemical trapping of singlet oxygen by diphenylisobenzofuran (DPBF). Intra-
molecular photoinduced electron transfer (PET) within the conjugate from the indole subunits (donor
A), to S1 (excited singlet state) of ZnPc moiety (acceptor D), is featured by the significant decrease of
fluorescence quantum yield and lifetime of ZnPc moiety, and the occurrence of transient absorption bands
of ZnPc•− at 570 and 630 nm. The triplet state, on the other hand, was not quenched by indole units. The
kinetics and thermodynamics of PET were analyzed quantitatively, and the quantum efficiency of PET is
computed to be 38%, almost double of the emission efficiency (20%). The quantum efficiency of triplet
(T1) formation is 0.50 and the quantum yield of DPBF photooxidation is 0.72. Both are larger than the
expected value of 0.32. The evolution of transient absorption spectra showed that the charge separation
state (ZnPc•−–indole•+) recombined to triplet state ZnPc(T1)–indole, which is responsible for the high
yield of T1 formation. In the presence of oxygen, both T1 and ZnPc•− were quenched efficiently, which
forms singlet oxygen and superoxide anion, respectively. DPBF is therefore photo-oxidized by both singlet
oxygen (Type II reaction, 46%) and superoxide anion radical (Type I reaction, 54%), which led to the high
yield of photooxidation. This is in contrast to free ZnPc PS, in which only singlet oxygen is responsible
for the photooxidation. The result suggests that the reaction mechanism is changed upon conjugation so
that the importance of Type I reaction is greatly enhanced, and the indole-conjugated ZnPc is an even
better PS than the free ZnPc.
Received 8 June 2011
Received in revised form
16 September 2011
Accepted 3 October 2011
Available online 10 October 2011
Keywords:
Phthalocyanine
Indole
Photosensitizer
Photophysics
Singlet oxygen
Fluorescence
Electron transfer
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
are probably quite different from that in the solid tumor tissue.
The localization of PS within tumor tissues brings various aro-
Photodynamic therapy of tumor (PDT) is under intensively
developing to address the lack of selectivity in removing and
destroying tumor tissues by traditional cancer therapies, such as
surgery, radiotherapy, and chemotherapy [1–3]. The interaction of
a photosensitizer (PS) with biomolecules in the presence of light
plays the key role to understand the mechanism for PDT and design
the new generation drugs for it [3]. There have been a lot of mech-
anism studies in which aqueous solution was used as the medium
and amino acids were employed as substrates for photosensitized
electron transfer or singlet oxygen quenching [4–9]. Intermolecu-
lar interactions in aqueous solution between a PS and amino acids
matic electron donor moieties, such as N-containing amino acid
residues, into its much closer vicinity than that in aqueous solu-
tion, hence the interaction is more similar to that of intramolecular
electron donor–acceptor case (interaction through bonding). The
intermolecular photoinduced electron transfer (PET) between free
phthalocyanine PS and amino acid donors has been evidenced
previously [4–9]. The PS functionalized by biomolecules can be con-
sidered as intra-molecular electron donor–acceptor pairs, which
usually make PET or energy transfer process from S1 (lowest
excited singlet state) of PS much faster than that of intermolecular
cases. PET competes with intersystem crossing (ISC) and therefore
reduces the efficiency of triplet formation, which further decreases
the yield of singlet oxygen and its subsequent photooxidation (Type
II reaction). PET also forms charge separated state (CSS) radical
pairs that can reduce oxygen to superoxide anion radicals, both CSS
∗
Corresponding author at: Chemistry Department & Center of Instrumental Anal-
ysis, Hebei Normal University of Science and Technology, Qinghuangdao, Hebei
Province 066004, China. Tel.: +86 3358357040; fax: +86 3358357040.
E-mail address: zhangxianfu@tsinghua.org.cn (X.-F. Zhang).
•−
and O2 may also participate photooxidations of biological sub-
strates, and hence enhances Type I reaction. It is therefore helpful
1010-6030/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.jphotochem.2011.10.008

118
X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
Fig. 1. Chemical Structure of ZnPc(␤-indole)4.
to evaluate how the bioconjugation of PS affects the photosensitiz-
ing process of PDT, whereas the past focus of related investigations
has been placed on the synthesis of PS bioconjugates and their bio-
logical activities. Proteins or peptides contain up to twenty three
amino acid residues, all of these could act as electron donors which
make the clear elucidation of photo events within protein/peptide
conjugated PS difficult. Tryptopan and tyrosine, however, are gen-
erally considered as the most active components in protein for
PDT. To simplify the problem, indole, the main reactive component
in tryptophan, was selected and attached to zinc phthalocyanine
(ZnPc) in this study (Fig. 1). ZnPc was selected since its derivatives
have been tested both in vitro [7,10–19] and in vivo [20,21] to be
effective photosensitizers for PDT. We show in this report that the
indole-modification of ZnPc significantly changes the mechanism
of photosensitized oxidation while improves the photosensitizing
ability.
NMR (CDCl3, ppm): ı 8.05–8.09 (d, 1H), 7. 97 (s, 1H), 7.88–7.95
(m, 2H), 7.70–7.72 (d, 1H), 7.63–7.66 (d, 1H), 7.33–7.35 (m, 2H),
6.81–6.82 (d, 1H). MS, m/z: 243 [M+].
2.2.2. Synthesis of tetra[ˇ-(indol-1-yl)] ZnPc, ZnPc(ˇ-indole)4
The two-step procedure for the synthesis of ZnPc(␤-indole)4 is
shown in Fig. 2. It is a well established procedure which has been
extensively discussed in literature [27].
Zinc acetate (0.070 g, 0.33 mmol), 4-(indol-1-yl)phthalonitrile
(0.31 g, 1.3 mmol), and 8 ml dried n-pentanol were mixed and
stirred at 135 ◦C for 4 h under argon atmosphere in the presence
of two drops of DBU as catalyst. After cooling down, the green pre-
cipitate was collected by filtration and washed with n-pentanol
and n-hexane. The dried crude product was dissolved in CH2Cl2 and
purified by column chromatography (silica gel) using CH2Cl2 as the
mobile phase. Yield: 39%. UV–vis (DMF): ꢁmax nm (log ε) 366 (4.66),
620 (4.41), 686 (5.10). IR [(KBr) ꢀmax/cm−1]: 3053 (Ar–H), 1608,
1519, 1490, 1332, 1137, 1049, 742 (Pc ring C C etc.), 895(N–Zn).
1H NMR (CDCl3): ı, ppm 9.10–9.16 (4H, m), 8.88–8.91 (4H, m),
7.89–8.31 (16H, m), 7.27–7.55 (8H, m), 6.97–7.05 (4H, d). Anal.
(C64H36N12Zn): C, H, N. MALDI-TOF-MS m/z: calculated 1036.3;
found 1037.6 (M+1).
2. Experimental
2.1. Reagents and apparatus
All reagents for synthesis were analytical grade and used as
received. Dimethylformide (DMF) was dried and redistilled before
use. 1H NMR spectra were recorded at room temperature on a
Bruker dmx 300 MHz NMR spectrometer. MS spectra were recorded
either on a Bruker APEX II or Autoflex III Maldi-TOF spectrometer.
IR spectra were recorded at room temperature on a Shimadzu FTIR-
8900 spectrometer. UV–vis spectra were recorded on a Shimadzu
4500 spectrophotometer using 1 cm matched quartz cuvettes.
2.3. Photophysical measurements
The absorption and fluorescence spectra, fluorescence quantum
yields and excited singlet-state lifetimes, as well as triplet proper-
ties were investigated at room temperature ca 22 ◦C. Steady-state
fluorescence spectra were acquired on a FLS 920 instrument. All
spectra were corrected for the sensitivity of the photo-multiplier
tube. The fluorescence quantum yield (˚f) was measured by using
2.2. Synthesis
2.2.1. Synthesis of 4-(indol-1-yl)phthalonitrile
Fs A0 n2s
1.73 g (0.01 mol) 4-nitrophthalonitrile and 1.17 g (0.01 mol)
indole were added to 20 ml DMF in argon atmosphere, the resulted
solution was stirred at room temperature under argon for 48 h,
during which 4.1 g (0.03 mol) anhydrous potassium carbonate was
added in four equal portions every 2 h. The product solution was
then poured into 100 ml ice-water to precipitate, the filtered
solid was washed with water neutral pH and dried under vac-
uum. Slightly yellow solid was obtained after recrystallization with
hexane–methanol solvent. Yield: 43%. m.p. 142–143 ◦C. IR(KBr),
ꢀ(cm−1): 3027, 1597, 1456 (Ar–H), 2233 (C N), 1070 (Ar–N). 1H
˚f = ˚0
·
·
f
2
0
F0 As n
in which F is the integrated fluorescence intensity, A is the
absorbance at excitation wavelength, n is the refractive index of the
solvent used, the subscript 0 stands for a reference compound and
s represents samples. ZnPc was used as the reference (˚0f = 0.30)
[22,23]. Excitation wavelengths of 610 nm corresponding to the
vibronic band of S0 to S1 transitions were employed. The sample
and reference solutions were prepared with the same absorbance

X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
119
Fig. 2. Synthetic path for ZnPc(␤-indole)4.
(Ai) at the excitation wavelength (near 0.09) in the same solvent
DMF. All solutions were air saturated for ˚f measurements.
Fluorescence lifetime of S1 was measured by time-correlated
single photon counting method (Edinburgh FLS920 spectropho-
tometer) with excitation at 672 nm diode laser (50 ps FWHM) and
emission was monitored at 690 nm and 730 nm.
Transient absorption spectra were recorded in degassed DMF
(prepared by bubbling with argon for 20 min) with an Edinburgh
LP920 laser flash photolysis system. A Nd:YAG laser (Brio, 355 nm
and 4 ns FWHM) was used as excitation source. The analyzing light
was from a pulsed xenon lamp. The laser and analyzing light beams
perpendicularly passed through a quartz cell with an optical path
length of 1 cm. The signal was displayed and recorded on a Tek-
tronix TDS 3012B oscilloscope and an R928B detector. The laser
energy incident at the sample was attenuated to ca. 20 mJ per pulse.
Time profiles at a series of wavelengths from which point by-point
spectra were assembled were recorded with the aid of a Pc con-
trolled kinetic absorption spectrometer. The concentrations of the
target compounds were typically 15 ␮M providing A355 = 0.40 in a
10 mm cuvette.
the irradiation wavelength of 660 nm by the samples and standard,
respectively. Their ratio can be obtained by Eq. (4).
ref
Iaref
1 − 10−A
670
=
.
(4)
Ia
1 − 10−A
670
To avoid chain reactions induced by DPBF in the presence
of singlet oxygen, the concentration of DPBF was lowered to
∼3 × 10−5 mol dm−3. A solution of sensitizer (absorbance ∼0.70 at
the irradiation wavelength) that contained DPBF was prepared in
the dark and irradiated in the Q-band region. DPBF degradation was
monitored by UV–vis absorption spectrum. The error in the deter-
mination of ˚ꢂ was ∼10% (determined from several ˚ꢂ values).
3. Results and discussion
The main concern of this study is how the conjugated indole
moieties affect the behavior of the excited state of ZnPc, includ-
ing S1 (lowest excited singlet state) and T1 (lowest excited triplet
state), and its capability as PS. All the related data is collected in
Table 1, together with that of the unconjugated ZnPc for compari-
son. The S1 properties of the conjugate were revealed by absorption
(Fig. 3) and fluorescencemeasurements, includingthe spectraldata,
the fluorescence quantum yield (˚f), and fluorescence lifetime (ꢃf).
˚f of ZnPc(␤-indole)4 is decreased one-third compared to that of
ZnPc. The ꢃf of the major emitting component is also shortened to
one-third, matching the ˚f decrease. This fluorescence quenching
shows that the indole conjugation significantly affects the behavior
of S1 of ZnPc, and that the interaction between S1 of ZnPc and indole
units does occur, which may be photoinduced electron transfer
(PET), or energy transfer (PeT), etc.
The triplet–triplet absorption coefficients (ꢂεT) of the samples
were obtained using the singlet depletion method [24], and the
following equation was used to calculate the ꢂεT:
ꢂAT
ꢂεT = εS
(1)
ꢂAS
where ꢂAS and ꢂAT are the absorbance change of the triplet
transient difference absorption spectrum at the minimum of the
bleaching band and the maximum of the positive band, respec-
tively, and εS is the ground-state molar absorption coefficient at
the UV–vis absorption band maximum. Both ꢂAS and ꢂAT were
obtained from the triplet transient difference absorption spectra.
The triplet quantum yield ˚T was obtained by comparing the
ꢂAT of the optically matched sample solution at 355 nm in a 1 cm
cuvette to that of the reference using the equation [24]:
To understand the above observation, measurements on absorp-
tion spectra for T1 and other transient species were carried out.
Studies on ␮s transient absorption spectra (TAS) in Fig. 4 gave the T1
properties (Table 1), while nano second TAS (Fig. 5) led to the iden-
tification of ZnPc•− formed by PET from indole units (donor D) to
ꢂεZTnPc
ꢂAT
˚
T
= ˚ZnPc
·
·
,
(2)
T
ꢂAZTnPc
ꢂεT
where the superscript represents the reference, ꢂAT is the
absorbance of the triplet transient difference absorption spectrum
at the selected wavelength, and ꢂεT is the triplet state molar
absorption coefficient.
Singlet oxygen quantum yield (˚ꢂ) determinations were car-
ried out using the chemical trapping method [25]. Typically,
a 3 ml portion of the respective PS solutions that contained
diphenylisobenzofuran (DPBF) was irradiated at 660 nm in air sat-
urated DMF. ˚ꢂ value was obtained by the relative method using
ZnPc as the reference (Eq. (3)):
k
kref
Iaref
˚
= ˚ref
,
(3)
ꢂ
ꢂ
Ia
where ˚rꢂef is the singlet oxygen quantum yield for the standard
(0.60 for ZnPc in DMF) [26], k and kref are the DPBF photo-bleaching
rate constants in the presence of the respective samples and stan-
dard, respectively; Ia and Iaref are the rates of light absorption at
Fig. 3. Comparison of the UV–vis absorption spectrum of ZnPc (4 ␮M) with that of
ZnPc(␤-indole)4 (7.9 ␮M) in DMF.

120
X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
Table 1
Photophysical properties in DMF.a
T−T
ꢁ
(nm) ꢂεT (M−1cm−1
max
abs
em
ꢁ
(nm) log ε
ꢁ
(nm) ꢂꢀ¯ (cm−1
)
Es (eV)
˚
f
ꢃf (ns)
ꢄ2
)
˚
T
ꢃT, (␮s)
˚
ꢂ
˚
PET
max
max
ZnPc
668
5.40 680
5.10 697
264.2
230.1
1.85
1.83
0.30 3.60
0.20 2.22(87%), 4.91(13%)
1.19 470
1.11 520
35,620
31,715
0.60 113
0.50 107
0.60 n.a.
0.72 0.38
ZnPc(␤–indole)4 686
a
abs
em
ꢁ
is the UV–vis absorption maximum. ꢁ
is the fluorescence emission maximum. ꢂꢀ¯ is the Stoke’s shift. Es is the energy of lowest excited state (S1). ˚f is the
max
max
fluorescence quantum yield, ꢃf is the fluorescence lifetime. ꢄ2 is the Chi squared value for the fitting of emission decay.
hv
spectrum for ZnPc(␤-indole)4, a band around 280 nm due to indole
chromophore appears, which is another evidence that indoles are
indeed linked to ZnPc.
−
S1 state of ZnPc moiety (acceptor A): D–A−→ D–A(S1) → D•+–A•
.
The details will be discussed in later sections.
3.1. Ground state absorption
3.2. ꢅs scaled transient absorption spectra
Fig. 3 compares the UV–vis absorption spectrum of ZnPc with
that of ZnPc(␤-indole)4. In red region, the spectral shape of the two
spectra is almost identical, indicating that the geometry of the ␲
skeleton of ZnPc moiety within ZnPc(␤-indole)4 is the same as that
of free ZnPc. In other words, it is not altered by the linked indoles.
Fig. 4 displays the ␮s-scale transient absorption spectra (TAS)
for ZnPc(␤-indole)4 and ZnPc. These spectra were recorded in argon
saturated DMF with excitation at 355 nm by a 4 ns pulsed laser. TAS
of ZnPc (Fig. 4b) is the typical of triplet–triplet (T1–Tn) absorptions
for Pcs and in good agreement with that reported for ZnPc [24]. TAS
of ZnPc(␤-indole)4 (Fig. 4a) exhibits very similar spectral shape to
that of ZnPc, except for a red shift of the absorption maximum,
which is then assigned to T1 of ZnPc(␤-indole)4 by analogy and
also based on the following analysis.
The minimum of negative absorptions in the B- and Q-band
region showed peaks matching the absorption maximum of the
corresponding ground state absorption; the positive bands are
separated from the ground state bleaching with well defined isos-
bestic points, and the bleaching recovery are synchronous to the
Upon the attachment of four indole units, the absorption maximum
abs
(ꢁ
) is only red shifted by 18 nm. This small shift suggests that the
max
indole aromatic system is not ␲-conjugated with the ZnPc aromatic
ring, even though the indole’s nitrogen atom is directly connected
to ZnPc. The molecular dynamic calculation by MM2 suggests that
the dihedral angle is around 30◦ between the two types of aromatic
rings in the ground state.
Due to the mutual independence of the two types of aromatic
system, an indole sub-unit behaves just like an usual electron-
abs
donating group with respect to ꢁ
of ZnPc. In the UV region of the
max
Fig. 4. ␮s scaled transient absorption spectra of ZnPc(␤-indole)4 (a) and ZnPc (b) in argon purged DMF, 15 ␮M, excitation wavelength 355 nm. Inset of ZnPc shows the decay
at 480 nm. (c) Decay of positive transient signal and recovery of negative ground state absorption in argon saturated DMF for ZnPc(␤-indole)4, excitation with 355 nm laser
pulse. Inset is the decay at 520 nm in air saturated DMF (c).

X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
121
Fig. 5. ns scaled transient absorption spectra of ZnPc(␤-indole)4 in argon purged DMF, 15 ␮M, excitation wavelength 355 nm. Inset (black) shows CR of CSS to T1 state. Inset
(right) shows the absorption rise, decay and rise again at 570 nm.
6 ns
5 ns
PET : ZnPc(S1)–indole−→ ZnPc• –indole•+−→ ZnPc–indole (III)
positive absorption decay, indicating concomitant behavior, i.e., as
the positive T1–Tn absorption decays, the ground state is repopu-
lated. The transient decay at 520 nm or 480 nm is also given in Fig. 4
for ZnPc(␤-indole)4 and ZnPc, the concomitant bleaching recovery
at ground state absorption minimum is also included. These curves
can all be well fit by the mono exponential function.
The triplet lifetime (ꢃT) thus obtained is collected in Table 1.
The ꢃT values are 113 ␮s for ZnPc, and 107 ␮s for ZnPc(␤-indole),
respectively. The result suggests that attached indole units do not
quench T1 of ZnPc, i.e. PET or PeT from indoles to ZnPc does not
occur. This is in contrast with the case of S1 state, from which
PET does proceed. The ꢃT values are comparable to those of other
Pcs [28,29], and are all sufficiently long for photosensitizing the
production of singlet oxygen.
−
PET
CR
ISC generates T1 which is 107 ␮s long-lived and absorbs light
in a broad range of wavelength, which may interfere with the
observation of other transient species. Charge separated state (CSS),
however, exhibits short lifetime of 5 ns (obtained in Section 3.3)
due to easy charge recombination (CR), as shown by process (III).
By recording spectra in ␮s scale we can therefore easily identify
the transient absorptions of T1, but not CSS since it is disappeared
already.
3.3. Nano second transient absorption spectra
Fig. 5 shows the evolution of the TAS of ZnPc(␤-indole)4 during
the first 23 ns. Absorption bands at 570 and 630 nm due to ZnPc•−
are now clearly resolvable in additional to T1–Tn absorption band at
520 nm. The signal of ZnPc•− rises almost at the same speed as that
of T1–Tn absorption in the first 15 s, since the rate constant of elec-
tron transfer (kET) and of triplet formation (kisc) are equal (Table 1,
also see process (II) and (III) above). However, the triplet signal
became higher after 20 ns of excitation, due to the faster decay of
CSS by CR. The formation and decay at 570 nm are illustrated in
the inset of Fig. 5, which shows that CSS state has a lifetime of ca.
5 ns. The presence of ZnPc•− absorption unambiguously indicates
the presence PET from indole units to ZnPc moiety. We would then
quantitatively evaluate the importance of PET relative to radiation
decay (fluorescence) and ISC in the S1 decay of ZnPc moiety. In air
saturated DMF solutions, the absorption bands at 570 and 630 nm
In air saturated DMF solution, however, the ꢃT of ZnPc(␤-
indole)4 is shortened dramatically to 0.37 ␮s (inset of Fig. 4c),
while that of ZnPc is reduced to 0.30 ␮s. The rate constant by oxy-
gen quenching can be then evaluated to be 1.35 × 109 M−1 s−1 for
ZnPc(␤-indole)4 and 1.66 × 109 M−1 s−1 for ZnPc, which are close
to diffusion rate constant. This effective oxygen quenching also
suggests that the positive absorptions are indeed due to T1–Tn
triplet absorptions. The triplet absorption coefficient (ꢂεT) was
determined to be 31,715 M−1 cm−1 for ZnPc(␤-indole)4. The triplet
formation quantum yield was measured to be 0.50, slightly smaller
than 0.60 for ZnPc.
In this micro second time scale, no transient species other than
T1 was found from TAS in Fig. 4. This also shows that no triplet
quenching is present, and hence no PET originates from T1 state.
Within this time range, ZnPc•−–Indole•+ due to PET from S1 is also
not observed, because the CSS is generated and decayed in nano
seconds. Since two-thirds of S1 states of ZnPc are not quenched by
the conjugated indoles, ISC (intersystem crossing, i.e. process (II)
below) from S1 can still effectively proceed, which will compete
with the process (III), i.e. S1 quenching by indoles as below.
were quenched, indicating molecular oxygen can effectively react
with ZnPc•− by capturing the electron: ZnPc− + O2 → ZnPc + O2
.
•−
The formed O2•− can be trapped by DMPO and recorded by ESR, for
which we have reported previously [30].
3.4. Fluorescence studies
hꢀ
Absorption : ZnPc–indole−→ ZnPc(S1)–indole
(I)
The fluorescence emission spectra of ZnPc(␤-indole)4 in DMF
were recorded under the same condition as that of ZnPc with exci-
tation at 610 nm. These spectra are compared in Fig. 6. In addition to
the intensity quenching mentioned previously, the spectral shape
fs
107 ␮s
6 ns
ISC : ZnPc(S1)–indole−→ ZnPc(T1)–indole −→ ZnPc–indole (II)
ISC

122
X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
Fig. 6. (Top) Normalized fluorescence emission spectra with excitation at 610 nm
(absorbance 0.090), (bottom) normalized absorption, fluorescence excitation
(emission at 730 nm) and emission spectra (with excitation at 610 nm) for ZnPc(␤-
indole)4.
Fig. 7. (Top) time profile of fluorescence decay of with excitation at 672 nm diode
laser (50 ps), the emission was monitored at 690 nm, the concentration of dyes is ca.
2.0 ␮M. Middle and bottom: fitting residues.
also becomes broader, the width of half height is increased from
24.5 nm of ZnPc to 29.5 nm of ZnPc(␤-indole)4. The spectral shape
in the region above 720 nm also shows difference. The band broad-
ening and the shape change indicate the presence of a new minor
emission band. ˚f of ZnPc(␤-indole)4 is 0.20, decreased from 0.30
of ZnPc. This fluorescence quenching is reasonable, because indole
units contain a nitrogen atom and can act as electron donors for
PET, so do tryptophan residues in peptides or proteins.
The transient decay of fluorescence emission at 690 nm is
illustrated in Fig. 7. A visual examination already tells that the fluo-
rescence lifetime (ꢃf) of ZnPc(␤-indole)4 is shortened with respect
to that of ZnPc. While ZnPc exhibits a typical mono-exponential
decay with ꢃf of 3.60 ns, ZnPc(␤-indole)4 shows bi-exponential
decaying behavior, suggesting the presence of two emitting species,
matching the observation from emission spectrum. The short-lived
component has a ꢃf of 2.22 ns and is responsible for 87% emission
while the long one has a ꢃf of 4.91 ns and takes the possession of
13% emission. The ꢃf of major emitting species is significantly short-
ened to 61.6% of the value for free ZnPc, apparently due to PET from
indole moieties to the excited singlet state (S1) of ZnPc sub-unit in
the conjugate. The de-excitation of S1 is often bi-exponential when
PET is involved in the process. Two emitting species are usually
responsible for the bi-exponential behavior: (i) CSS emission due
to CR (charge recombination process (IV)), and (ii) part of D–A (S1)
with the unfavourable conformation for PET.
bi-exponential fitting is also necessary but the percentage of long-
lived component is increased to 39% from 13% at 690 nm, since
emission maximum of CSS is located in this region.
In summary, the fluorescence investigation reveals that PET
does occur from indole units to ZnPc fragment within the conju-
gate. With excitation at 610 nm, only ZnPc moiety is populated to
S1, energy transfer from S1 of ZnPc to indole is not allowed, since
the emitted photons have much lower energy than that of indoles
can absorb. CSS charge recombination also emits fluorescence with
the lifetime of 4.91 ns. This ꢃf value is actually the lifetime of CSS
itself, which is in good agreement with the value of 5.0 ns obtained
from the decay of transient absorption of ZnPc•− at 570 nm.
3.5. Efficiency and kinetics of PET
The rate constant of PET (ket) can be calculated from the fluo-
rescence lifetime data using Eq. (5), in which ꢃf0 is the fluorescence
lifetime of ZnPc while ꢃf is the value of short-lived component for
ZnPc(␤-indole)4. ket is thus obtained as 0.17 × 109 s−1
.
ket = ꢃ−1 − (ꢃf0)−1
(5)
f
Efficiency for PET (˚et) is calculated to be 38% by using the
ket value (˚et = ketꢃf), which indicates that PET efficiency is almost
double of its fluorescence emission (˚f = 0.20) in the system. kf (the
rate constant of fluorescence emission from S1) is 0.90 × 108 s−1
and 0.83 × 108 s−1 for ZnPc(␤-indole)4 and ZnPc, respectively (com-
puted by kf = ˚f/ꢃf), showing no remarkable change.
The summation of ˚f, ˚et, and ˚T is 1.08 (=0.20 + 0.38 + 0.5) for
ZnPc(␤-indole)4, larger than 0.90 (=˚f + T = 0.30 + 0.60) for ZnPc.
The internal conversion (IC) efficiency of ZnPc can be estimated as
0.10 (˚ic = 1 − ˚f − ˚T). The substitution of H by indole is expected
to enhance IC, such that ˚ic of ZnPc(␤-indole)4 is likely greater
than 0.10. Therefore the summation of ˚f, ˚et, ˚ic and ˚T for
ZnPc(␤-indole)4 would be larger than 1.18. This large value caused
our attention. While the repeated measurements for ˚f, ˚et and
5 ns
CR fluorescence : ZnPc• –indole•+−→ ZnPc–indole + hꢀ
(IV)
−
CR
For CR fluorescence, a red-shifted, weak and broader emis-
sion usually occurs due to the lower energy level of D•+–A•− than
D–A(S1). For D–A(S1) with the unfavourable conformation, the ꢃf of
minor component is close to that for free ZnPc due to the absence
of PET, and no new emission band is expected in this case. In
this particular case, the minor component is due to CSS emission
since its lifetime is 4.91 ns, quite longer than 3.60 ns of free ZnPc.
The broader emission spectrum also implicates the contribution
from other emitting species. Monitoring the decay at 730 nm, a

X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
123
Fig. 8. The formation of CSS (D•+–A•−) by PET, and CR of CSS to generate T1 and
emission.
Fig. 9. The change of absorption spectrum of upon irradiation time in air saturated
DMF containing 5 ␮M DPBF and 11 ␮M photosensitizer with irradiation at 660 nm.
Inset is the linear plot of absorbance at 415 nm against irradiation time.
˚
T
showed reproducible values, we noticed ˚T is larger than that
anticipated, because CR of CSS also generate T1 state. Fig. 8 gives the
TAS during 20–40 ns. With the decrease of the band for ZnPc•− at
570 nm, the T1–Tn absorption is concurrently increased, indicating
that CR of CSS also forms triplet state T1 of ZnPc:
˚T, i.e. 0.50, since it is generated by energy transfer from T1 state
to molecular oxygen:
3
1
T1(PS) + O2 → S0 (PS) + O2(1ꢂg)
(VI)
CR−ISC
+
−
CR to T1
:
ZnPc• –indole• −→ ZnPc(T1)–indole
(V)
1
DPBF + O2(1ꢂg) → producttypeII
(VII)
The absorption change at 570 nm (where both ZnPc•− and T1
absorb) with time, inset of Fig. 5, shows all the processes: (1) first
rise stage: both ZnPc•− and T1 are formed after laser excitation, (2)
fast decay of ZnPc•− due to CR of ZnPc•−–Indole•+, (3) CR of CSS
to form T1, (4) T1 decay. These processes are illustrated in Fig. 8.
The presence of CR-ISC is the reason why the observed ˚T (0.50) is
bigger than the expected value of 0.32, which is only due to ISC from
S1 (1 − ˚f − ˚et − ˚ic ≤ 1 − 0.2 −0.38 −0.1 = 0.32). This means that
almost half of CSS ((0.50 − 0.32)/0.38 = 47%) goes to triplet state, i.e.
process (V) plays an important role in CR of CSS.
The measured ˚ꢂ of 0.72 is significantly greater than the max-
imum quantum yield of singlet oxygen (0.50). This is due to the
fact that DPBF traps not only singlet oxygen but also superoxide
anion radical O2•−[33]. Therefore the measured photooxidation
efficiency is the sum of two types of oxidations: (1) DPBF oxi-
dized by singlet oxygen, and (2) DPBF oxidized by superoxide anion
•−
radical O2
.
•−
For the conjugate, superoxide anion radical O2
is produced
by the reaction of ZnPc•− with molecular oxygen (process (VIII)
below), this is evidenced by the oxygen quenching to TA of ZnPc•−
at 570 nm mentioned previously. In contrast, free ZnPc has no this
process.
3.6. Thermodynamics of PET
−
+
−
•
ZnPc• –Indole• + O2 → ZnPc–Indole•+ + O2
(VIII)
(IX)
The oxidation potential of methyl-indole was measured to be
0.65 V [31], while the reduction potential of the phthalocyanine
ring is −0.90 V in DMF [32]. The electron transfer from indole to
ZnPc ring in their ground states is not thermodynamically allowed,
due to the large positive value of free energy change (ꢂG) cal-
culated by ꢂGET = Eox − Ered − C = 0.65 − (−0.85) − 0.06 = 1.44 eV, in
which Eox is the oxidation potential of a donor, Ered represents the
reduction potential of an acceptor, and C is a small constant asso-
ciated with solvent. PET from indole to S1 of ZnPc moiety (by the
photo excitation of the Pc moiety), on the other hand, is thermo-
dynamically favored, since its ꢂG is a negative value obtained by:
ꢂGPET = ꢂGET – E00 (ZnPc) = 1.44 − 1.83 = –0.39 eV. PET from indole
donors to T1 of ZnPc moiety is forbidden, since the energy of the T1
state is 0.99 eV, which suggests a ꢂG of +0.45 eV.
DPBF + O2 − → producttypeI
•
The quantum efficiency for CSS generation is 0.38 (˚ET). In the
presence of oxygen process (V), i.e. CR to T1, is inhibited, such that
all CSS is dissipated in process (VIII), which gives the maximum
•−
quantum efficiency of O2 generation as 0.38. In this case, all T1
is originated from S1 by ISC, the efficiency of which is 0.32, this is
the amount of T1 that can contribute to the generation of singlet
oxygen. Therefore the total efficiency of DPBF photooxidation is
expected to be the sum of process (VII), i.e. Type II reaction, and
process (IX), i.e. Type I reaction. The summation gave the value
0.70 (=0.38 + 0.32). This estimated value is in good agreement with
the measured value of 0.72.
We can also roughly estimate the relative importance of Type
I over Type II reaction. The percentage of Type I reaction is about
54% (0.38/0.70), while the importance of Type II reaction is 46%.
This result shows that the binding of indole can affect the reaction
mechanism significantly, while the photo-oxidizing ability of the
bioconjugated PS is even increased.
3.7. Singlet oxygen formation and DPBF photooxidation
The quantum yield (˚ꢂ) for photooxidation was measured
using the DPBF chemical trapping method with irradiation at
660 nm in air saturated DMF. DPBF degradation was monitored by
UV–vis spectra. Fig. 9 displays the decrease of [DPBF] upon irradi-
ation time, for which the first order kinetics was observed (inset of
Fig. 9). ˚ꢂ is also included in Table 1. The measured value of ˚ꢂ
for ZnPc(␤-indole)4 is 0.72, while the maximum quantum yield of
singlet oxygen 1O2 (1ꢂg) production is expected to be the value of
4. Conclusion
We have synthesized a bioconjugated PS (BCPS) model com-
pound in which four indole units were attached to ZnPc through

124
X.-F. Zhang, W. Guo / Journal of Photochemistry and Photobiology A: Chemistry 225 (2011) 117–124
covalent bonding. The photophysical and photochemical processes
related to PDT were revealed by comparing the transient and steady
state spectra of S1, T1 and CSS with that of the free ZnPc. PET occurs
within the molecule from indole subunits to S1 of ZnPc moiety,
which is evidenced by the quenching of ˚f and ꢃf, the presence of
transient absorption bands for ZnPc•−, the emission due to CSS, the
thermodynamic and kinetic analysis. The indole conjugated ZnPc
shows unusually high yield for triplet formation because CR of CSS
(ZnPc•−–indole•+) is recombined to triplet state ZnPc(T1)–indole.
The yield for DPBF photooxidation is higher than that of ZnPc, due to
the good yield of superoxide anion radical generated by the reaction
of CSS with oxygen (which is absent for free ZnPc). Type I mecha-
nism accounts for 54% of total reaction, which is higher than 46%
by Type II reaction. These results suggest that the indole-modified
ZnPc is a better PS than the free ZnPc, but the reaction mechanism is
altered so that the importance of Type I mechanism is significantly
enhanced.
[9] J. Davila, A. Harriman, Photochem. Photobiol. 50 (1989) 29–35.
[10] W.M. Sharman, J.E. Lier v, Bioconjugate Chem. 16 (2005) 1166–1175.
[11] G. Valduga, S. Nonell, E. Reddi, G. Jori, S.E. Braslavsky, Photochem. Photobiol.
48 (1988) 1–5.
[12] R.M. Negri, A. Zalts, E.A. RomáN, P.F. AramendíA, S.E. Braslavsky, Photochem.
Photobiol. 53 (1991) 317–322.
[13] N. Cauchon, H. Tian, R. Langlois, C. La Madeleine, S. Martin, H. Ali, D. Hunting,
J.E. van Lier, Bioconjugate Chem. 16 (2005) 80–89.
[14] X.F. Zhang, H.J. Xu, J. Chem. Soc., Faraday Trans. 89 (1993) 3347–3351.
[15] C.F. Choi, P.T. Tsang, J.D. Huang, E.Y.M. Chan, W.H. Ko, W.P. Fong, D.K.P. Ng,
Chem. Commun. 2004 (2004) 2236–2237.
[16] T. Nyokong, Coord. Chem. Rev. 251 (2007) 1707–1722.
[17] S.-i. Ogura, K. Tabata, K. Fukushima, T. Kamachi, I. Okura, J. Porphyrins Phthalo-
cyanines 10 (2006) 1116–1124.
[18] D. Woehrle, O. Suvorovab, R. Gerdesa, O. Bartelsa, L. Lapoka, N. Baziakinab, S.
Makarovb, A. Slodeka, J. Porphyrins Phthalocyanines 8 (2004) 1020–1041.
[19] V. Mantareva, V. Kussovski, I. Angelov, E. Borisova, L. Avramov, G. Schnurpfeil,
D. Woehrle, Bioorg. Med. Chem. 15 (2007) 4829–4835.
[20] N. Yusuf, S.K. Katiyar, C.A. Elmets, Photochem. Photobiol. 84 (2008) 366–370.
[21] J.D. Miller, E.D. Baron, H. Scull, A. Hsia, J.C. Berlin, T. McCormick, V. Colussi, M.E.
Kenney, K.D. Cooper, Toxicol. Appl. Pharmacol. 224 (2007) 290–299.
[22] M. Montalti, A. Credi, L. Prodi, M.T. Gandolfi, Photophysical properties of organic
compounds, in: Handbook of Photochemistry, Taylor & Francis Group, LLC,
London, 2006, 255.
[23] K. Ishii, N. Kobayashi, The photophysical properties of phthalocyanines and
related compounds, in: K.M. Kadish, K.M. Smith, R. Guilard (Eds.), The Porphyrin
Handbook, vol. 16, Academic Press, New York, Amsterdam, 2003, pp. 1–42.
[24] I. Carmichael, G.L. Hug, J. Phys. Chem. Ref. Data 15 (1986) 1–250.
[25] M.G. Lagorio, L.E. Dicelio, E.A.S. Roman, S.E. Braslavsky, J. Photochem. Photobiol.
B 3 (1989) 615–624.
Acknowledgements
This work has been supported by Hebei Provincial Science Foun-
dation (Contract B2010001518) and HBUST.
[26] N. Kuznetsova, N. Gretsova, E. Kalmykova, E. Makarova, S. Dashkevich, V. Neg-
rimovskii, O. Kaliya, E. Luk’yanets, Russ. J. Gen. Chem. 70 (2000) 133–139.
[27] N.B. McKeown, The synthesis of symmetrical phthalocyanines, in: K.M. Kadish,
K.M. Smith, R. Guilard (Eds.), The Porphyrin Handbook, vol. 15, Academic Press,
New York, 2003, p. 61.
[28] X.-F. Zhang, Y. Chang, Y. Peng, F. Zhang, Aust. J. Chem. 62 (2009) 434–440.
[29] X.F. Zhang, J. Huang, H. Zhao, X. Zheng, Z. Junzhong, J. Photochem. Photobiol.
A: Chem. 215 (2010) 96–102.
[30] X.F. Zhang, H.J. Xu, D.W. Chen, J. Photochem. Photobiol. B: Biol. 22 (1994)
235–239.
[31] Y.A. Udum, M. Düdükcüa, F. Kölelia, React. Funct. Polym. 68 (2008) 861–867.
[32] M. L’her, A. Pondaven, Electrochemistry of Phthalocyanines, in: K.M.S.K.M.
Kadish, R. Guilard (Eds.), The Porphyrin Handbook, Academic Press, San Diego,
2003, pp. 117–170.
References
[1] C.A. Robertson, D.H. Evans, H. Abrahamse, J. Photochem. Photobiol. B 96 (2009)
1–8.
[2] N.V. Kudinova, T.T. Berezov, Biochem. (Moscow) Suppl. Ser. B: Biomed. Chem.
4 (2010) 95–103.
[3] K. Plaetzer, B. Krammer, J. Berlanda, F. Berr, T. Kiesslich, Lasers Med. Sci. 24
(2009) 259–268.
[4] J.J. Lopez, M.A.G. Carter, Y.P. Tsentalovich, O.B. Morozova, A.V. Yurkovskaya, P.J.
Hore, Photochem. Photobiol. 75 (2002) 6–10.
[5] X.-F. Zhang, H. Xu, T. Shen, Sci. China B 38 (1995) 641–648.
[6] G. Ferraudi, G.A. Arguello, H. Ali, J.E. Lier, Photochem. Photobiol. 47 (1988)
657–660.
[7] X.F. Zhang, H.J. Xu, J. Photochem. Photobiol. B 24 (1994) 109–116.
[8] A. Segalla, C.D. Borsarelli, S.E. Braslavsky, J.D. Spikes, G. Roncucci, D. Dei, G. Chiti,
G. Jori, E. Reddi, Photochem. Photobiol. Sci 1 (2002) 641–648.
[33] T. Ohyashiki, M. Nunomura, T. Katoh, Biochim. Biophys. Acta (BBA): Biomem.
1421 (1999) 131–139.