Synthesis, structure and properties of a new two-photon photopolymerization initiator

Article abstract of DOI:10.1039/b206578a

A new two-photon free-radical photopolymerization initiator, (E,E)-4-{2-[p′-(N,N-di-n-butylamino)stilben-p-yl]vinyl} pyridine (abbreviated to DBASVP), has been synthesized. Quantum chemistry calculations showed that the new initiator possesses a large delocalized π electron system, a large change in dipole moment on transition to the excited state anda large transition moment. The calculated two-photon absorption cross-section is as high as 881.34 × 10^-50 cm⁴ s photon^-1. The single-photon and two-photon absorption and fluorescence properties in various solvents have been investigated carefully. The new initiator exhibits outstanding solvent-sensitivity, which experimentally interprets the excellent electron delocalized properties of the molecule. A microstructure has been fabricated under irradiation at 800 nm using a 200 fs, 76 MHz Ti:sapphire femtosecond laser.

Full text of DOI:10.1039/b206578a

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis, structure and properties of a new two-photon
photopolymerization initiator
Yan Ren,*a Xiao-Qiang Yu,a Dong-Ju Zhang,b Dong Wang,a Ming-Liang Zhang,c
Gui-Bao Xu,a Xian Zhao,a Yu-Peng Tian,c Zong-Shu Shaoa and Min-Hua Jianga
aState Key Laboratory of Crystal Materials, Shandong University, Jinan, 250100, P. R. China.
E-mail: ry@icm.sdu.edu.cn
bSchool of Chemistry and Chemical Engineering, Shandong University, Jinan, 250100,
P. R. China
cSchool of Chemistry and Chemical Engineering, Anhui University, Hefei, 230039, P. R. China
Received 8th July 2002, Accepted 19th September 2002
First published as an Advance Article on the web 23rd October 2002
A new two-photon free-radical photopolymerization initiator, (E,E)-4-{2-[p’-(N,N-di-n-butylamino)stilben-p-
yl]vinyl}pyridine (abbreviated to DBASVP), has been synthesized. Quantum chemistry calculations showed that
the new initiator possesses a large delocalized p electron system, a large change in dipole moment on transition
to the excited state and a large transition moment. The calculated two-photon absorption cross-section is as
high as 881.34 6 10250 cm4 s photon21. The single-photon and two-photon absorption and fluorescence
properties in various solvents have been investigated carefully. The new initiator exhibits outstanding
solvent-sensitivity, which experimentally interprets the excellent electron delocalized properties of the molecule.
A microstructure has been fabricated under irradiation at 800 nm using a 200 fs, 76 MHz Ti:sapphire
femtosecond laser.
It was proved be an effective two-photon photopolymeri-
Introduction
zation initiator excited with an 800 nm, 200 fs, 76 MHz
Ti:sapphire laser source. Because of its intriguing properties
as a new near-IR two-photon free-radical photoinitiator, we
have also conducted a thorough study of its structure by quan-
tum chemistry calculations and investigated its spectroscopic
properties including single-photon and two-photon fluores-
cence in different solvents in order to get an understanding of
the structure–spectroscopy–property relationship and to pro-
mote the design of new effective two-photon photoinitiators.
In two-photon absorption (TPA), the excitation is confined
to a small focal volume due to the quadratic dependence of
absorption on the incident intensity. This characteristic of
two-photon processes has made two-photon absorption based
photopolymerization an effective tool for processing photonic
devices, micromachines, zero-threshold lasers and integrated
optical waveguides in three dimensions with high spatial
resolution.1–4
For two-photon free radical photopolymerization, the
polymerization rate and photoinitiation threshold is closely
related to the absorption behavior of the initiator at the excita-
tion wavelength (generally twice its single-photon absorption
wavelength). At present, effective initiators are still rare
because the design strategy, or in other words, the relationship
of the initiator structures and their properties for initiating
two-photon photopolymerization, remained unclear. Conse-
quently, commercially available UV photopolymer initiators
are often used as substitutes in two-photon 3D microfabri-
cation. Unfortunately, these initiators always exhibit low
photosensitivity due to their small two-photon absorption
cross-section. Therefore, the search for effective and high
photosensitive two-photon photopolymerization initiators is
urgently required.
Some approaches to improving the photosensitivity of
photoinitiator molecules that have been reported are based
on the idea of increasing their two-photon absorption cross-
sections.5 According to Albota et al.6 and Reinhardt et al.,7 the
conjugation length, donor and acceptor strength, and planarity
of the p center are the important parameters for enhancing
the TPA. On the basis of this design strategy, in this article,
(E,E)-4-{2-[p’-(N,N-di-n-butylamino)stilben-p-yl]vinyl}pyridine
(abbreviated to DBASVP), which possesses a large p con-
jugated system and a strong donor and acceptor group, has
been synthesized using a relatively simple route. This com-
pound exhibits a strong two-photon absorption at y830 nm.
A microstructure has been fabricated using this new initiator.
Experimental
1 Chemicals
4-Vinylpyridine (95%), palladium(II) acetate (47.5% Pd) and
4-bromobenzyl bromide (98%) were products of Acros Organics.
N,N-Di-n-butylaniline (w98%) and tris(2-methylphenyl)phos-
phine (w95%) were purchased from Tokyo Kasei Kogyo Co.,
Ltd. Triphenylphosphine was AR grade. The above chemicals
were used without further purification. N,N-Dimethylforma-
mide, phosphorus trichloride and triethylamine and all solvents
were purified before use. A tert-butoxide–tert-butyl alcohol
solution was prepared immediately before use.
Photoluminescence and absorption measurements were
conducted in various solvents including toluene, chloroform,
THF, diethyl ether, ethanol, methanol, benzyl alcohol, acetone,
acetonitrile and N,N-dimethylformamide (DMF). All of the
above solvents were anhydrous grade after further purification.
Moreover, each of the solutions used for the single-photon
fluorescence measurements was freshly prepared and kept in
the dark before the measurements. The solution concentration
was 1 6 1025 mol L21
.
2 Instruments
The 300 MHz 1H NMR spectra were obtained on an INOVA-
300 spectrometer. Elemental analysis was performed using a
DOI: 10.1039/b206578a
J. Mater. Chem., 2002, 12, 3431–3437
This journal is # The Royal Society of Chemistry 2002
3431

View Article Online
PE 2400 elemental analyzer. The electrospray mass spectrum
(ES-MS) was determined on a Finnigan LCQ mass spectro-
20 g of sodium acetate in 200 mL of cold water was added
slowly with stirring. The reaction mixture was then stirred for
an additional hour and the resulting solution was extracted
with diethyl ether. The combined extract was washed with
saturated anhydrous sodium bicarbonate and then washed
with water and the organic layer was dried over anhydrous
magnesium sulfate. Then the solvent was removed under
reduced pressure, and the residue was distilled in vacuum at
y148 uC, 1 mmHg. 24 g 4-(N,N-Di-n-butylamino)benzalde-
hyde was obtained as a yellow liquid.
graph; the concentration of the sample was about 1.0 mmol L21
.
The diluted solution was electrosprayed at a flow rate of
5 6 1026 L min21 with a needle voltage of 1 4.5 kV. The
mobile phase was an aqueous solution of methanol (v/v, 1 : 1)
acidified with acetic acid. The sample was run in the positive-
ion mode.
UV-vis-near-IR spectra were measured on a Hitachi U-3500
recording spectrophotometer. The steady-state fluorescence
spectra measurement was performed using an Edinburgh
FLS920 spectrofluorimeter. A 450 W Xe arc lamp provides
the y400 nm excitation source. Spectra were recorded between
420 nm and 780 nm using a photomultiplier tube as detector,
which is operated in the single photon counting mode. The
spectral resolution is 0.1 nm. The quartz cuvettes used were
of 1 cm path length.
4-Bromobenzyl(triphenyl)phosphonium bromide (2). To
a
250 mL flask with a stirrer, 25 g (0.10 mol) 4-bromobenzyl
bromide, 30 g (0.11 mol) triphenylphosphine and 200 mL
xylene were added. The reaction mixture was refluxed for
6 hours, then cooled to room temperature and filtered. The
resulting white solid was washed with diethyl ether three times
and re-crystallized in a 3 : 1 mixture of chloroform and ethanol.
23 g colorless crystals were obtained. Yield: 45%.
3 Synthesis
The synthetic route is shown in Scheme 1.
4-Bromo-4’-(N,N-di-n-butylamino)stilbene (3). To a 250 mL
four-necked flask with a magnetic stirrer, a dropping funnel, a
nitrogen input tube and an adapter with a stopcock, 15.36 g
(0.03 mol) 4-bromobenzyl(triphenyl)phosphonium bromide
were added under a nitrogen atmosphere. Then 4.66 g
(0.02 mol) 4-(N,N-di-n-butylamino)benzaldehyde–THF solu-
tion were injected into the flask through the stopcock of the
adapter. The reaction mixture was cooled to 0 uC in an ice bath,
0.15 mol potassium tert-butoxide–tert-butyl alcohol solution
was added to the dropping funnel and was dropped into the
flask at 0 uC over one and a half hours. After the reaction
mixture had been stirred for two hours at 0 uC, the ice bath was
removed and the mixture stirred at room temperature
overnight and then for another two hours at 60 uC in an oil
bath. After being cooled to room temperature, the product was
poured into 200 mL distilled water, neutralized with dilute
hydrochloric acid, and extracted with dichloromethane. The
organic layer obtained was dried over anhydrous magnesium
sulfate. After filtration, the solvent was removed from the
solution at reduced pressure and the residue dissolved in a
minimum of DMF. The solution obtained was chromato-
graphed over 200 g of silica gel. Elution with 1 : 5 diethyl
ether–petroleum ether and then recrystallization three times in
ethanol produced light green crystals. Calcd for C22H28BrN2:
C, 68.39; H, 7.25; N, 3.63. Found: C, 68.56; H, 7.77; N, 3.76%.
1H NMR (CDCl3) (300MHz; TMS) dH: 0.95 (3H, t, J 7.2 Hz),
1.35 (2H, m), 1.51 (2H, m), 3.28 (2H, t, J 7.56), 6.62 (2H, d,
J 8.59 Hz), 6.69 (1H, d, J 16.15 Hz), 7.01 (1H, d, J 16.15 Hz ),
7.32 (2H, d, J 8.59 Hz), 7.36 (2H, d, J 8.94 Hz), 7.42 (2H, d,
J 8.59 Hz).
4-(N,N-Di-n-butylamino)benzaldehyde (1). Dimethylforma-
mide (100 mL) was cooled to 0 uC and treated dropwise with
phosphorus oxychloride (21 mL, 225 mmol). The solution was
stirred at 0 uC for 1 hour and at room temperature for another
1 hour. To the red solution was added dropwise N,N-di-n-
butylaniline (45.3 mL, 200 mmol). The mixture was heated at
60 uC for 10 hours and then cooled to 0 uC. Then a solution of
(E,E)-4-{2-[p’-(N,N-Di-n-butylamino)stilben-p-yl)vinyl]pyridine
(4). 1.5 g (3.75 mmol) 4-Bromo-4’-(N,N-di-n-butylamino)-
stilbene, 0.49 g (1.61 mmol) tri-o-tolylphosphine, 0.87 mL
(8.06 mmol) vinylpyridine, 0.045 g palladium(II) acetate
(0.2 mmol) and 80 mL redistilled triethylamine under nitrogen,
were added to a three-necked flask equipped with a mag-
netic stirrer, a reflux condenser, and a nitrogen input tube. The
reaction mixture was refluxed in an oil bath under nitrogen.
An orange product was obtained after heating and stirring for
16 h. Then the solvent was removed under reduced pressure
and the residue was dissolved in methylene chloride, washed
three times with distilled water, and dried with anhydrous
magnesium sulfate. Then it was filtered and concentrated. The
resulting solution was chromatographed over 200 g of silica gel.
Elution with 3 : 2 ethyl acetate–petroleum ether and then
recrystallization twice from diethyl ether produced light yellow
crystals. ES-MS, m/z (%): 411.4 ([M1H]1, 100), 206.2 (33).
Calcd for C29H34N2: C, 84.87; H, 8.29; N, 6.83. Found: C,
84.40; H, 8.74; N, 7.60%.
Scheme 1
3432
J. Mater. Chem., 2002, 12, 3431–3437

View Article Online
Fig. 1 Equilibrium geometry of DBASVP in a vacuum.
4 Computation
Fig. 2 Orbital energy and electron density distribution of frontier
molecular orbitals.
The restricted Hartree–Fock (RHF) theory and quantum che-
mical parametrization AM1 (a modified neglect of differential
overlap) in the program package MOPAC 6.0 were employed
in the structural optimization.8,9 Minimum-energy geometry
was used for ZINDO CI calculations using the CNDO/1
method with the highest 15 occupied orbitals and the lowest
15 unoccupied (virtual) orbitals to determine the distribution of
the frontier molecular orbitals, the corresponding electronic
spectrum,10,11 the state dipole and transition dipole moments,
and oscillator strength. All parameters used in our calcula-
tions were the default values in the programs. The state dipole
moment and transition moment were directly used to evaluate
the two-photon absorption cross-section based on the two-
state model. All computations were accomplished on an O2
workstation.
show that transition between the ground state S0 and the first
excited state S1 is mainly attributed to the promotion of an
electron from the HOMO to the LUMO. The transition energy
is 28489.1 cm21 with the an oscillator strength (f) of 2.081.
Fig. 2 illustrates the electron density distribution of the frontier
molecular orbitals. In the HOMO, the electrons are mainly
concentrated on the nitrogen atom of the amino and its
adjacent benzene ring. While in the LUMO, the electrons are
spread over the molecule along the molecular axis. Obviously,
the electron density is shifted from the N,N-di-n-butylamino
to the vinylpyridine upon excitation. The calculation results
also show that this transition causes a large change in the
ge
molecular dipole moment (Dmtotal ~ 6.600 D) and a large
ge
transition moment (Mtotal ~ 12.454 D). The conjugated
Results and discussion
geometric configuration and the electron density distribution
of the frontier molecular orbitals reveal that the molecule has
a highly delocalized p-electron system. We consider that these
structural features are necessary conditions for two-photon
absorption and for initiating a photopolymerization reaction.
The two-photon absorption cross-section was evaluated
based on the two-state model.12,13 Generally, from the sum-
over-state formulae, the two-photon matrix elements for the
resonant absorption of two-photons with identical energy are
expressed as
1 Structural features
Fig. 1 depicts the equilibrium geometry of this compound in
a vacuum. Table 1 lists the selected bond lengths, bond angles
and dihedral angles. From Fig. 1, one can see that the molecule
is not perfectly planar. According to Table 1, the dihedral
angle of C11–C10–C5–C4, C12–C11–C10–C5 and C13–C12–
C11–C10 are 215.9u, 179.8u and 15.9u respectively. Therefore,
the two benzene rings are actually in same plane. The dihedral
angles of C21–C20–C15–C14, C22–C21–C20–C15 and C23–
C22–C21–C20 are 2160.0u, 2179.5u and 14.9u respectively.
The angle between the pyridine ring and the benzene ring
is y34.4u. The bond lengths of the two benzene rings and the
pyridinium ring are characteristic of aromatic bonds. The
bridge bond lengths of C5C10, C10C11, C11C12 and C15C20,
C20C21, C21C22 indicate that the bridge bonds are also highly
conjugated.
ꢀ
ꢁ
ꢀa
ꢀb
vi { vf =2
ꢀb
ꢀa
X
Soj jiTSij jfT Soj jiTSij jfT
Sab
~
z
(1)
vi { vf =2
i
where vf denotes the excitation energy of the excited state |f m,
m is the electronic dipole operator, a and b denote the different
orientation of x, y and z, o, i and f represent the ground,
intermediate and final state, respectively; vi is the excitation
energy for the intermediate state. For molecules in the gas
phase and in solution, the TPA cross-section is given by
The excited state calculation (ZINDO CI) was performed
based on the trans orientation of the compound. The results
˚
Table 1 The selected bond lengths (A), bond angles (u) and dihedral angles (u) optimized by AM1
N1–C2
C2–C3
C3–C4
C4–C5
C5–C6
C5–C10
C10–C11
1.405
1.418
1.387
1.401
1.404
1.450
1.344
C11–C12
C12–C13
C13–C14
C14–C15
C15–C16
C17–C12
C20–C15
1.451
1.403
1.390
1.404
1.403
1.405
1.452
C20–C21
C22–C23
C23–C24
C24–N25
N25–C26
C27–C22
1.344
1.404
1.406
1.346
1.348
1.406
N1–C2–C3
C2–C3–C4
C3–C4–C5
C4–C5–C6
C7–C2–C1
C10–C5–C4
C11–C10–C5
C11–C10–C5–C4
C13–C12–C11–C10
121.01
121.19
121.40
117.81
121.84
123.00
125.02
215.92
15.85
C12–C11–C10
C14–C13–C12
C15–C14–C13
C16–C15–C14
C17–C12–C11
C20–C15–C14
C21–C20–C15
C12–C11–C10–C5
C21–C20–C15–C14
124.88
120.82
120.79
118.47
119.14
119.15
124.55
179.83
2160.02
C22–C21–C20
C23–C22–C21
C24–C23–C22
N25–C24–C23
C26–N25–C24
C27–C22–C21
124.88
122.97
118.91
123.76
117.00
119.38
C23–C21–C20–C15
C23–C22–C21–C20
2179.50
14.92
J. Mater. Chem., 2002, 12, 3431–3437
3433

View Article Online
orientational averaging over the two-photon absorption
probability
408 nm in ethanol (emax ~ 53364), 405 nm in acetonitrile
(emax ~ 48431) and 411 nm in DMF (emax ~ 49947)
respectively. Above 525 nm, the solutions are completely
transparent. The absorption deviation in solution can be attri-
buted to the interaction of the solute with the local molecular
environment including solvent molecules and other surround-
ing solute molecules. One can see that the peak position and
peak broadening are different for different solutions, which
further identify the effect of solvent on absorption behavior.
Some interesting results can be observed from Fig. 3. A slight
red shift in absorption is observed in low polar solvents, while
the absorption peaks are found to be blue shifted in high polar
solvents such as acetone, ethanol, and acetonitrile. In order to
avoid any possible error, the solutions were freshly prepared
and the spectra measured again, but the same curves were
obtained. The slight red shift in low polar solvents can be
explained by the fact that the energy level of the excited state is
reduced more than that of the ground state by solvent action
since a relatively high polarity of DBASVP predominates in
the excited state and a lower polarity predominates in the
ground state (the values of mg, me, are 5.604 D and 12.204 D
respectively). When a D–A structural molecule is in a high
polar solvent, the molecule in the ground state tends to be
polarized to form a charge-separated molecule by the action of
an electric field formed by the surrounding solvent molecules.
We consider that the absorption blue shift of DBASVP in
such a polar environment is like the blue shift behavior of
zwitterionic hemicyanine dyes and pyridinium dyes. Compared
to DBASVP,14,15 generally, they show a large negative solva-
tochromism because of an opposite polar form in the ground
and excited states. The corresponding quantum chemistry
calculation is in progress in order to obtain a reasonable
explanation for the solvatochromism of DBASVP.
h
i
X
ꢀ
ꢀ
zH
dTPA
~
F
zG
(2)
S
aa Sbb
S
ab Sab
S
ab Sbꢀa
i
where the coefficients F, H and G are related to the polarization
of the radiation source. The values of F, H and G are 2, 2 and
2 for linear polarized light and 22, 3 and 3 for the circular one.
For one dimensional molecules, however, some simplifications
can be made according to the two-state model in which only the
ground and first low-lying charge transfer excited state are
taken into account and the TPA cross-section is believed to be
completely dominated by the component along the z molecular
axis, Szz,
2Dꢀgze
Mzge
Szz~
(3)
DE
where Dmzge and Mzge are the change in dipole moment and the
transition moment respectively between the ground state and
the first excited state, and DE ~ ˆıv, where v is the fundamental
frequency of the light. Thus the TPA cross-section is given by
X
The TPA cross-section directly comparable with experiment
ꢂ
ꢀ ꢃ
ꢀ
ꢀ
zH
dTPA
~
F
zG
(4)
S
zz Szz
S
zz Szz
S
zz Szz
zz
is defined as
3 a5
4 p 0 a v2
sTPA
~
dTPA
(5)
15c Cf
Based on the above model, the two-photon absorption
cross-section sTPA of DBASVP was calculated. The state
dipole and transition dipole moments obtained from ZINDO
were directly used during the calculations. Assuming linear
polarized light was used, the value calculated is as high as
881.34 6 10250 cm4 s photon21, which indicates that a strong
two-photon absorption occurs upon excitation.
3 Single-photon fluorescence
Steady state fluorescence. The single-photon fluorescence
spectra of DBASVP show sensitivity to the polarity of the
environment around DBASVP (see Fig. 4). There is a remar-
kable shift in the emission spectra on varying the solvent
polarity. As shown in Fig. 4, the fluorescence peak is at 508 nm
in toluene, the solvent of least polarity (0.36 D). In DMF, the
solvent of highest polarity (3.82 D), the peak is located at
616 nm, which is red-shifted by 108 nm due to the changed
molecular environment. Such behavior indicates significant
charge redistribution upon excitation, which is also in agree-
ment with the statement in the above Structural feature section.
Besides, one can see that the fluorescence intensity of DBASVP
in aprotic solvents decreases gradually with increasing solvent
2 Linear absorption
According to the above quantum chemistry calculation results,
the single-photon absorption peak in a vacuum is at 351 nm
(28489.1 cm21, the oscillator strength f ~ 2.081). In fact, the
single-photon absorption peak in solution was observed to
deviate from that in a vacuum. Fig. 3 shows the resulting
spectra observed experimentally. The absorption peak appears
at 402 nm in cyclohexane (with the corresponding molar absorp-
tion coefficient emax ~ 27288), 408 nm in toluene (emax ~
36707), 411 nm in chloroform (emax ~ 44052), 422 nm in benzyl
alcohol (emax ~ 40485), 406 nm in acetone (emax ~ 48206),
Fig. 4 Single-photon (of 450 W Xe arc lamp as excitation source)
fluorescence spectra of DBASVP in different polarity solvents at a
Fig. 3 Linear absorption spectra of DBASVP in solvents of different
polarity at a concentration of 1 6 1025 mol L21
.
concentration of 1 6 1025 mol L21
.
3434
J. Mater. Chem., 2002, 12, 3431–3437

View Article Online
Table 2 Stokes’ shift (Dv) of DBASVP in solvents with different
¯
orientational polarizability (Df) (e is the value at 25 uC unless specified
otherwise)
are found in protic solvents, even though hydrogen bonds exist
between the DBASVP molecules and the protic solvents. The
specific solvent effect caused by protic solvents manifests itself
in a decrease in fluorescence intensity (Fig. 4).
Based on eqn. (6) and Table 2, one can evaluate the charge
separation of DBASVP upon excitation in aprotic solvents.
˚
Assuming that the cavity radius is 8.7 A, which is comparable
to the radius of DBASVP (the distance calculated between
˚
two nitrogen atoms is 17.4 A), on the basis of eqn. (6), the
labs-max
nm
/
lem-max
nm
/
Dv/
¯
Solvents
n
e
cm21 Df
Cyclohexane
Toluene
1.424 2.023b 402
1.494 2.379 408
1.444 4.806b 411
1.352 4.34b 398
488
508
546
518
602
550
616
590
606
612
4384 20.0005
4825
6016
5821
7085
6509
8097
7681
8008
8351
0.0140
0.1487
0.1672
0.2100
0.2107
0.2760
0.2849
0.2894
0.3062
Chloroform
Diethyl ether
Benzyl alcohol 1.538 13.1b
422
405
411
406
408
405
difference in the calculated dipole moment between the ground
and excited states in solution is 29.4 D, which is larger than
that in a vacuum. It is known that a dipole moment of
4.8 D results from a charge separation of one unit charge
THF
DMF
Acetone
Ethanol
Acetonitrile
1.404 7.58
1.427 37.6
1.357 20.7
1.359 24.3
1.342 37.4
(4.8 6 10210 esu) by 1 A. Hence, for DBASVP, 29.4 D
˚
˚
corresponds to an intramolecular charge separation of 6.125 A
aThe values of the refractive index (n) and relative permittivity are
upon excitation. This fact indicates that DBASVP is highly
dependent on solvent polarity. The redistribution of electrons
occurs within the DBASVP molecule, which again proves that
DBASVP possesses a highly delocalized conjugated system.
taken from ref. 23 . bRelative permittivities measured at 20 uC.
polarity except in ether, in which DBASVP exhibits the highest
fluorescence intensity of all the solutions. In protic solvents,
such as in ethanol, methanol and benzyl alcohol, DBASVP
shows relatively low fluorescence intensity.
Fluorescence lifetime and quantum yield. The fluorescence
lifetime measurement of DBASVP in solvents of different
polarity was performed on the same Edinburgh FLS920
spectrofluorometer with a Hydrogen flash lamp (pulse duration
v1 ns) as the excitation source. During the measurements,
DBASVP was excited with the optimal excitation wavelength
(lmaxex) in order to give the largest relative fluorescence inten-
sity in each solution. Emission peaks were observed irrespec-
tive of the excitation wavelength. The time distribution of
the hydrogen lamp pulse is measured immediately before the
measurement of the fluorescence decay. The observed decay of
fluorescence is given by convolution of the lamp pulse with
the impulse response of the sample. The decays were analyzed
by the ‘‘least-squares’’ and iterative reconvolution method. The
quality of the exponential fits was evaluated by the goodness of
fit (x2). For all the decays in this experiment, x2 is less than
1.2, which indicates a good fit. The results are shown in Fig. 6.
The fluorescence lifetime of DBASVP in aprotic solvents
increases with increases in solvent polarity. In protic solvents,
DBASVP exhibits reduced lifetimes, especially in methanol, in
which the value is less than 0.6 ns.
In order to further demonstrate the influence of solvent
on fluorescence, Table 2 lists the Stokes’ shift of DBASVP in
solvents of different polarity. The Stokes’ shift is defined as the
loss of energy between absorption and reemission of light,
which is a result of several dynamic processes. These processes
include energy losses due to dissipation of vibrational energy,
redistribution of electrons in the surrounding solvent mole-
cules induced by the altered dipole moment of the excited
chromophore, reorientation of the solvent molecules around
the excited state dipole, and specific interactions between the
fluorophore and the solvent or solutes. The Lippert equation
is the most widely used equation to describe the effects of the
physical properties of the solvent on the emission spectra of
fluorophore.16
ꢄ
ꢅ
2
n2
ꢀ
ꢀ
)
g
2
e{1
{1
n2
z1
(
{
a3
e
ꢀ
ꢀ
na { nf %
{
zconst (6)
hc 2ez1
2
In this equation, h is Planck’s constant, c is the speed of light,
and a is the radius of the cavity in which the fluorophore
resides. The wavenumbers of the absorption and emission are
v and v (in cm21). e and n are the relative permittivity and the
The photoluminescence quantum efficiencies in different
solvents were measured on a Perkin Elmer LS-55B fluoro-
spectrometer by comparison of the fluorescence intensities with
that of compounds of known quantum yield under identical
experimental conditions. The Rh6G ethanol solution was used
as the quantum-yield standard (W ~ 0.97), and a RhB ethanol
¯
¯
f
a
refractive index respectively. The terms in the first brackets are
called the orientation polarizability (Df). Fig. 5 shows a plot of
v 2 v versus the orientation polarizability. In aprotic solvents,
f
¯
¯
a
Stokes’ shifts are approximately proportional to the orienta-
tional polarizability. But no obvious shifts larger than expected
Fig. 6 Fluorescence lifetime of 1 6 1025 mol L21 DBASVP solutions
in solvents of different polarity. 1 Toluene; 2 diethyl ether; 3
chloroform; 4 THF; 5 benzyl alcohol; 6 acetone; 7 ethanol; 8 methanol;
9 acetonitrile; 10 DMF.
Fig. 5 Stokes’ shift of DBASVP in various solvents versus orientation
polarizability of the solvents. The concentration of solution measured
was 1 6 1025 mol L21
.
J. Mater. Chem., 2002, 12, 3431–3437
3435

View Article Online
Table 3 Fluorescence lifetime tf, quantum yield Wf, radiative rate
constants kf ~Wf/tf and nonradiative rate knr ~ (1 2 Wf)/tf of
DBASVP in solvents of different polarity
Solvents
tf/ns
Wf
kf/107 s21
knr/107 s21
Toluene
1.50
0.93
0.59
2.67
0.39
0.05
0.06
0.04
25.8
4.9
10.3
1.5
40.8
102.7
159.2
35.9
Benzyl alcohol
Methanol
DMF
solution was used to calibrate the system. 2 6 1027 mol L21
DBASVP dilute solutions were prepared in order to avoid
possible aggregation and self-absorption. The quantum yield
was calculated according to the following equation.17
Fig. 7 Experimental setup for the measurement of the two-photon
fluorescence intensity at different excitation wavelengths.
Ð
ꢄ
ꢅꢄ
ꢅꢄ
ꢅ
ns2
nr2
ðl Þ IðlrÞ
Fs
Fr
Ar
As
r
Ð
Ws ~ Wr
(7)
ðl Þ IðlsÞ
s
Here, W is the quantum yield, n is the refractive index, I(l) is
the relative intensity of the exciting light at wavelength l, A(l)
is the absorbance of the solution at the exciting wavelength
l, and bF is the integrated area under the corrected emission
spectrum. Subscripts s and r refer to the sample and reference
solutions, respectively. Radiative and nonradiative rate con-
stants were also calculated and the results are listed in Table 3.
One can see that the quantum yield and fluorescence lifetime of
DBASVP in protic solvents are strongly reduced (the reduced
fluorescence lifetime is also illustrated in Fig. 6). The very
large nonradiative rates indicate the presence of an additional
radiationless channel. We attribute the outstanding behavior of
DBASVP in benzyl alcohol and methanol to the specific solvent
effect of hydrogen bonding. Because the absorption spectrum
remains unchanged in protic solvents (see Fig. 3), the ground-
state complexation between DBASVP and protic solvents
can be ruled out. Therefore, the formation of a nonfluorescent
charge transfer excited state engaged in an excited hydrogen
bonded pair is the very reason for the strongly reduced
fluorescence yield.18 Besides, the behavior of the DBASVP–
DMF solution is noticeable. It possesses an equally small
quantum yield as that shown by DBASVP in protic solution.
Decreased quantum yields have frequently been reported for
molecules with a similar structure to that of DBASVP in
which the rotation of the C–N bond connecting the amino
group to the chromophore forms a twisted intramolecular
charge transfer (TICT) geometry.19–21 According to the TICT
model, the increase in polarity of the solvents will lower the
energy level of the CT state, and thus the fluorescence emission
peak will red shift. In addition, the TICT state will rapidly
undergo nonradiative transition to triplets, and it results in
a marked decrease in the fluorescence quantum yield with
increasing solvent polarity. The red-shifted fluorescence peak
shown in Fig. 4 and the markedly decreased fluorescence
yield listed in Table 3 suggest that the TICT mechanism is
also suitable for interpreting fluorescence behavior of the
DBASVP–DMF solution.
Fig. 8 Two-photon (from
a 200 fs, 76 MHz Ti:sapphire laser)
fluorescence intensity (peak) of DBASVP in toluene (circle) and
DMF (square) versus excitation wavelengths of identical energy of
0.380 W. Insert: two-photon fluorescence spectra excited at 800 nm.
The concentration of the solutions measured was 0.01 mol L21
.
were recorded in the excitation wavelength range of 710 nm
to 900 nm with a step size of 5 nm. The results are plotted
in Fig. 8. Each datum expresses the maximum fluorescence
intensity excited at each wavelength. One can see that the
two-photon fluorescence intensity gradually increases with the
red shift of the excited wavelength, and after a maximum, it
decreases rapidly with the red shift of the excited wavelength.
The maximum two-photon fluorescence emission was found to
be excited by an 830 nm laser beam in toluene solution and an
855 nm laser beam in DMF solution, which indicates that the
optimal absorption wavelengths of two-photon excitation for
two DBASVP solutions are located at 830 nm and 855 nm
respectively. Obviously, the two-photon absorption peak is
red-shifted with increasing solvent polarity. The two-photon
fluorescence spectra of DBASVP in these two solvents, excited
at 800 nm, are also shown in Fig. 8 (inset). The fluorescence
peaks at 508 nm in toluene, and is red-shifted to 600 nm in the
more polar solvent of DMF. Therefore, for DBASVP, the
solvatochromism of the two-photon process is similar to that
of the single-photon process.
4 Two-photon fluorescence
The experimental setup for two-photon fluorescence measure-
ment is shown in Fig. 7. The laser beam from a mode-locked
Ti:sapphire laser (Coherent Mira 900F) is the pump source
with a pulse duration of 200 fs, a repetition rate of 76 MHz, and
a single-scan streak camera (Hamamatsu Model C5680-01)
together with a monochromator as the recorder. Using this
experimental setup, the maximum two-photon fluorescence
intensities at different excitation wavelengths were measured at
identical pump energy levels of 0.380 W. The samples mea-
sured were 0.01 mol L21 DBASVP–toluene and DBASVP–
DMF solutions. A polarizer was used to keep the incident
beam energy constant at different excited wavelength. The data
5 Two-photon photopolymerization
A grating was created by two-photon polymerization of a
poly(urethane acrylate) oligomer using DBASVP as initiator.
A film of the mixture of initiator and oligomer with a weight
ratio of y2 to y3% (a little DMF solvent was added to con-
trol the viscosity) was prepared by spin-coating onto glass
plates. The experimental setup is shown in Fig. 9. The same
mode-locked Ti:sapphire laser as that used in the two-photon
3436
J. Mater. Chem., 2002, 12, 3431–3437

View Article Online
there is significant charge redistribution upon excitation. A
microstructure was fabricated by two-photon photopoly-
merization using this new initiator using the 800 nm irradiation
of a Ti:sapphire femtosecond laser.
Acknowledgements
Fig. 9 Experimental setup for two-photon photopolymerization. M~
mirror.
The authors would like to acknowledge Professor Chuan-Kui
Wang for useful discussions on the theoretical calculations of a
two-photon absorption cross-section. This work was supported
by a grant for the State Key Program of China (G1998061402),
partly supported by the National Natural Science Foundation
of China (Grant No. 50173015) and the Scientific Research
Foundation for Outstanding Young Scientists of Shandong
Province of China (Grant No. 01BS24).
References
1
2
3
S. Kawata, H. Sun, T. Tanaka and K. Takada, Nature, 2001, 412,
597–598.
H. Sun, V. Mizeikis, Y. Xu, S. Juodkazis, J. Ye, S. Matsuo and
H. Misawa, Appl. Phys. Lett., 2001, 79, 1–3.
Fig. 10 Optical micrograph of the grating fabricated via two-photon
polymerization of a poly(urethane acrylate) oligomer using DBASVP
as initiator.
M. P. Joshi, H. E. Pudavar, J. Swiatkiewicz, P. N. Prasad and
B. A. Reianhardt, Appl. Phys. Lett., 1999, 74, 170–172.
P. Galajda and P. Ormos, Appl. Phys. Lett., 2001, 78, 249–251.
S. M. Kuebler, B. H. Cumpson, S. Ananthavel, S. Barlow,
J. E. Ehrlich, L. L. Etsine, A. A. Heikal, D. McCord-Maughon,
J. Qin, H. Rocel, M. Rumi, S. R. Marder and J. W. Perry,
SPIE-Int. Soc. Opt. Eng., 2000, 3937, 97–105.
M. Albota, D. Beljonne, J. L. Bredas, J. E. Ehrlich, J. Y. Fu,
A. A. Heikal, S. E. Hess, T. Kogej, M. D. Levin, S. R. Marder,
D. McCord-Maughon, J. W. Perry, H. Ro¨ckel, M. Rumi,
G. Subramaniam, W. W. Webb, X. L. Wu and C. Xu, Science,
1998, 281, 1653–1656.
B. A. Reinhardt, L. L. Brott, S. J. Clarson, A. G. Dillard,
J. C. Bhatt, R. Kannan, L. X. Yuan, G. S. He and P. N. Prasad,
Chem. Mater., 1998, 10, 1863–1874.
J. J. P. Stewart, Quantum Chem. Program Exch., 1990, 10, 86.
J. J. P. Stewart, MOPAC 93.00 Manual, Fujitsu Limited, Tokyo,
1993.
4
5
fluorescence measurements was used for two-photon micro-
fabrication. It provided the 800 nm lasing source which was
tightly focused via an objective lens (640, NA ~ 0.65), and the
focal point was focussed on the sample film on the xy-step
monitorized stage controlled by computer. The pulse energy
before being focused by the objective lens was y1.2 W. The
polymerized solid skeleton was obtained after any unreacted
liquid mixture had been washed out. The grating fabricated
was observed through a polarization microscope (Opton,
Germany). Its photograph is illustrated in Fig. 10.
The photopolymerization mechanism of this new initiator
is still unknown. According to Cumpston et al.,22 strong donor
substituents would make the conjugated system electron rich,
and after one- or two-photon photoexcitation, these chromo-
phores would be able to transfer an electron even to relatively
weak acceptors, and that this process could be used to activate
the polymerization reaction. DBASVP possesses strong donor
and acceptor substituents, and its molecule is electron rich. As
the two-photon absorption cross-section of DBASVP is large,
the two-photon-induced electron transfer within the molecule
should be efficient. DBASVP also has the virtue of excellent
electron delocalizataion upon excitation, which is believed to
be another necessary condition for it to easily initiate a two-
photon photopolymerization reaction. The exact mechanism of
photoinitiation is currently being investigated. The systematic
study of the microstructure created by two-photon free-radical
photopolymerization using DBASVP as the initiator will be
reported in detail elsewhere.
6
7
8
9
10 A. Klamt, J. Phys. Chem., 1995, 99, 2224–2235.
11 M. Karelson, K. Pihlaja and M. C. Zerner, J. Photochem.
Photobiol. A, 1995, 85, 119.
12 P. Macak, Y. Luo and H. Agren, J. Chem. Phys., 2001, 114, 9813–
9820.
˚
13 Y. Luo, P. Macak, P. Norman, C. K. Wang and H. Agren,
Nonlinear Opt., 2001, 27, 33–46.
14 P. Fromherz, J. Phys. Chem., 1995, 99, 7188–7192.
15 U. Narang, C. F. Zhao, J. D. Bhawalkar, F. V. Bright and
P. N. Prasad, J. Phys. Chem., 1996, 100, 4521–4525.
16 J. R. Lakowicz, Principles of Fluorescence Spectroscopy, Plenum
Press, New York, 1983, p. 190.
17 J. N. Demas and G. A. Crosby, J. Phys. Chem., 1971, 75, 991–
1024.
18 M. Maus, W. Rettig, D. Bonafoux and R. Lapouyade, J. Phys.
Chem. A, 1999, 103, 3388–3401.
19 J. D. Simon and S. G. Su, J Chem. Phys., 1987, 87, 7016–7023.
20 R. K. Guo, N. Kitamura and S. Tazuke, J. Phys. Chem., 1990, 94,
1404–1408.
˚
21 G. S. He, L. Yuan, Y. Cui, M. Li and P. N. Prasad, J. Appl. Phys.,
1997, 81, 2529–2537.
Conclusion
22 B. H. Cumpston, S. P. Ananthavel, S. Barlow, D. L. Dyer,
J. E. Ehrlich, L. L. Erskine, A. A. Heikal, S. M. Kuebler,
I.-Y. Sandy Lee, D. McCord-Maughon, J. Qin, H. Ro¨ckel,
M. Rumi, X. L. Wu, S. R. Marder and J. W. Perry, Nature, 1999,
398, 51–54.
A new two-photon photopolymerization initiator of DBASVP
has been synthesized. Its two-photon absorption cross-section
determined by quantum chemistry calculations is as high
as 881.34 6 10250 cm4 s photon21. The experimental results
showed that DBASVP is solvent-sensitive, which indicates
23 CRC Handbook of Chemistry and Physics, ed. D. R. Lide, CRC
Press, Boca Raton, FL, 73rd edn., 1992–1993.
J. Mater. Chem., 2002, 12, 3431–3437
3437