Non-ionic photo-acid generators for applications in two-photon lithography

Article abstract of DOI:10.1039/b816434g

Non-ionic photoacid generators (PAGs) have been designed and synthesized for use in two-photon lithography (TPL). The chromophores in these new PAGs are covalently linked to the photocleavable group by a flexible joint. Their thermal stability, solubility

Full text of DOI:10.1039/b816434g

www.rsc.org/materials
Volume 19 | Number 4 | 28 January 2009 | Pages 437–552
ISSN 0959-9428
PAPER
Rudolf Zentel,
Christopher K. Ober et al.
Non-ionic photo-acid generators
for applications in two-photon
lithography
HIGHLIGHT
Koen Binnemans
Luminescence of metallomesogens in
the liquid crystal state

PAPER
www.rsc.org/materials | Journal of Materials Chemistry
Non-ionic photo-acid generators for applications in two-photon lithography†
Lorenz Steidl,‡ad Shalin J. Jhaveri,‡ab Ramakrishnan Ayothi,a Jing Sha,a Jesse D. McMullen,c
a
Sin Yee Cindy Ng, Warren R. Zipfel, Rudolf Zentel and Christopher K. Ober
c
d
a
*
*
Received 19th September 2008, Accepted 30th October 2008
First published as an Advance Article on the web 8th December 2008
DOI: 10.1039/b816434g
Non-ionic photoacid generators (PAGs) have been designed and synthesized for use in two-photon
lithography (TPL). The chromophores in these new PAGs are covalently linked to the photocleavable
group by a flexible joint. Their thermal stability, solubility and efficiency to produce acid under both
one- and two-photon excitation were characterized. The potential of these PAGs for TPL was tested in
two negative-tone resist systems relying on different mechanisms: free-radical/cationic polymerization
or a cationically initiated cross-linking reaction. These PAGs needed lower threshold power for
polymerization compared to a commercially available photoinitiator, isopropylthioxanthone, and
a photoacid generator, N-hydroxynaphthalimide triflate. Microstructures with a resolution of 0.6 mm
were fabricated and the threshold power for polymerization was found to be below 2 mW.
The success of two-photon absorption (TPA) depends on the
1. Introduction
use of optical materials, which are capable of TPA. The TPA
The theory of two-photon absorption (TPA) processes dates
back to 1931 when Maria Goeppert-Mayer first predicted that an
atom or a molecule can absorb two photons simultaneously in
the same quantum event.1 The advantage of this process is that
outside the focal point, the incident light is below the absorbance
threshold of most materials. Hence, if one tightly focuses
a femtosecond laser beam into a polymerizable resin, then the
polymerization would occur only in close proximity to the focal
point. Denk et al. first reported the use of a laser scanning
microscope for fluorescence imaging using two-photon excita-
tion.2 By using a laser scanning microscope and by tightly
focusing a femtosecond laser into a polymerizable resin, one can
generate three-dimensional structures by controlling the laser in
the x, y and z directions. The extent of TPA depends quadrati-
cally on excitation intensity; therefore, TPA is confined to a focal
volume of the order of the cube of the excitation wavelength.
Much recent work is related to biological imaging.3–5 In the area
of materials science, work has been mostly carried out in areas
of three-dimensional microfabrication,6–8 although researchers
have also used this physical process for ultra-high density optical
data storage,9–11 and two-photon pumped frequency up-con-
verted lasing,12,13 to name but two examples.
sensitivity depends on the TPA cross-section (d), expressed as
.
GM units, where 1 GM ¼ 1 ꢀ 10ꢁ50 cm4 s photonꢁ1 moleculeꢁ1
Most of the commercially available materials have low two-
photon sensitivity (<30 GM).14 It is thus usually necessary to
sensitize commercial photoactive compounds for TPA use by the
addition of two photon chromophores. Several research groups
have worked to design and synthesize molecules having a large
d value.15 In addition most of the dyes, which are usually formed
for multi-photon purposes, are used for polymerizing acrylates
and other unsaturated imaging materials via a free-radical
mechanism. There are a few additional reports that use non-
acrylate based systems.16,17
Here we describe the synthesis of new efficient two-photon
photoacid generators, which combine both a chromophore and
a photocleavable group in one unit with the goal of improving
energy transfer. In addition, two of these photosensitive units
were combined in one molecule to increase its overall sensi-
tivity. Photoacid generators (PAGs) were selected as they
primarily split into two radicals, which form a strong acid after
chemical reactions with each other or the medium.18 They can
thus be used to initiate reactions, which proceed via either
a cationic or a radical pathway. This widens the range of resins
that can be employed with these PAGs for 3D micro-
fabrication. This is particularly important, because although
most of the commercially available resins are based on cross-
linking of vinyl groups, it is very difficult to obtain high quality
3D structures with them as they exhibit high shrinkage during
both the exposure and development process steps.19 In
designing these PAGs we concentrated on non-ionic structures
and we linked the chromophores via a flexible joint to achieve
a better solubility in apolar solvents. We used the new non-
ionic PAGs for three-dimensional microfabrication of both
a radical and a cationic polymerizable resin. The PAGs were
also tested for their solubility, thermal properties and their
ability to produce acid under one and two-photon excitation
conditions.
aDepartment of Materials Science and Engineering, Cornell University,
Ithaca, NY, 14853, USA. E-mail: cober@ccmr.cornell.edu; Fax: +1
(607) 255-2365; Tel: (+607) 255-8417
bDepartment of Chemistry and Chemical Biology, Cornell University,
Ithaca, NY, 14853, USA
cDepartment of Applied and Engineering Physics, Cornell University,
Ithaca, NY, 14853, USA
dInstitute of Organic Chemistry, Department of Chemistry and Pharmacy,
University of Mainz, Duesbergweg 10-14, 55128 Mainz, Germany. E-mail:
zentel@uni-mainz.de; Fax: +49 (6131) 39-24778; Tel: +49 (6131) 39-
20361
† Electronic supplementary information (ESI) available: NMR spectra of
the prepared PAGs and UV-Vis, as well as FT-IR spectra of the detection
of photoacid formation in solution and polymer film. See DOI:
10.1039/b816434g
‡ These authors contributed equally to this work.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 505–513 | 505

suspension was stirred for 30 minutes and then heated to 150 ꢂC
for 15 hours. The suspension was cooled to room temperature
and poured into crushed ice. The yellow precipitate was filtered,
dried under vacuum (12 h at 60 ꢂC) and recrystallized with DMF
to give pale yellow crystals (4.90 g ¼ 54%). 1H-NMR (300 MHz,
DMSO), d ¼ 8.78 (dd, 2H), 8.61 (dd, 2H), 8.50 (d, 2H), 7.97 (dd,
2H), 7.74 (t, 1H), 7.40 (t, 1H), 7.34 (dd, 2H), 7.21 (d, 2H) ppm.
2. Experimental section
2.1 Materials
Norland 63 was purchased from Norland Products (Cranbury,
NJ), and 2,6-difluorobenzene sulfonyl chloride from Fluorochem
Ltd. Poly-[4-(tert-butoxycarbonyl)oxy]styrene was synthesized
by free-radical polymerization using AIBN as initiator. All other
chemicals and solvents were purchased from Sigma-Aldrich and
Lancaster Synthesis and were used without further purification.
N,N0-Dihydroxy-4,40-(phenylene-1,3-dioxy)-dinaphthalene-1,8;
10,80-tetracarboxylic imide (V). To 4 g (7.96 mmol) of compound
(II) and 1.35 g (19.64 mmol) of hydroxylamine hydrochloride in
a three neck flask, 65 mL of pyridine was added. The flask was
2.2 Synthesis
ꢂ
flushed with nitrogen and the mixture was heated to 90 C for
The synthesis of the four new non-ionic PAGs was carried out
according to the reaction scheme presented in Fig. 1. Synthesis
starts with an etherification to give compounds II–IV, which
were then converted to the corresponding N-hydroxynaph-
thalimides V–VII. Reaction of these compounds with the
respective sulfonyl chlorides yielded the photoacid generators
PAG 1, PAG 2, PAG 3 and PAG 4. Compound (I) was prepared
via monoalkylation of phloroglucinol.
3 hours. After cooling, the suspension was poured into 60 mL of
ice-cooled water; the yellow precipitate obtained was filtered,
washed with water and dried under vacuum for 20 hours at
ꢂ
1
60 C. A yield of 4 g (95%) was obtained. H-NMR (300 MHz,
DMSO), d ¼ 10.72 (s, 2H), 8.68 (dd, 2H), 8.57 (dd, 2H), 8.47
(d, 2H), 7.92 (dd, 2H), 7.69 (t, 1H), 7.36 (t, 1H), 7.29 (dd, 2H),
7.19 (d, 2H) ppm.
N,N0-Di-(2-Nitro-4-(trifluoromethyl)benzenesulfonyloxy)-4,40-
(phenylene-1,3-dioxy)dinaphthalene-1,8;10,80-tetracarboxylic imide
(PAG 1). In the dark, 2 g (3.76 mmol) of the hydroxyimide (V),
2.18 g (7.52 mmol) of 2-nitro-4-trifluoromethylbenzene sulfonyl
chloride and 60 mL of cold tetrahydrofuran (THF) were
combined in a three neck flask. The flask was cooled to ꢁ5 ꢂC
and flushed with nitrogen. Over a period of 30 min., 0.76 g
Bis-(4-oxy-1,8-naphthalic anhydride)phenylene (II). 1.99
g
(18.05 mmol) of resorcinol, 4.97 g (18.05 mmol) of potassium
carbonate and 60 mL of dimethylformamide (DMF) were placed
in a three neck flask. The flask was flushed with nitrogen and the
mixture was heated to 70 ꢂC. 10 g (36.1 mmol) of 4-bromo-
naphthalic anhydride and 15 mL of DMF were added, the
Fig. 1 Reaction scheme for the synthesis of the non-ionic photoacid generators (PAG 1, PAG 2, PAG 3 and PAG 4).
506 | J. Mater. Chem., 2009, 19, 505–513 This journal is ª The Royal Society of Chemistry 2009

(7.52 mmol) of triethylamine in 20 mL of cold THF was
added dropwise. The mixture was stirred for 24 hours at room
temperature and filtered. The filtrate was collected, the THF
removed under vacuum and the yellow residue was dissolved
in ethyl acetate and washed with water and sodium bicar-
bonate solution (1%). The ethyl acetate was removed and the
crude product was recrystallized with dichloromethane/diethyl
ether yielding 1.17 g (30%) of a yellow powder (PAG 1).
1H-NMR (300 MHz, acetone), d ¼ 8.87 (dd, 2H), 8.65–8.62
(m, 4H), 8.54–8.46 (m, 4H), 8.41 (dd, 2H), 8.00 (dd, 2H), 7.78
(t, 1H), 7.42–7.38 (m, 3H), 7.29 (d, 2H) ppm. 19F NMR (282
MHz, acetone), d ¼ ꢁ64.3 ppm. Elem. Anal.: Calcd: C, 50.87;
H, 1.94; N, 5.39; F, 10.97. Found: C, 50.67; H, 1.74; N, 5.31;
F, 11.07. UV (acetonitrile): lmax ¼ 362 nm (3 ¼ 3170 m2/mol).
N,N0-Di-(2-Nitro-4-(trifluoromethyl)benzenesulfonylthio)-4,40-
(phenylene-1,3-dioxy)dinaphthalene-1,8;10,80-tetracarboxylic imide
(PAG 3). To 1 g (1.77 mmol) of hydroxyimide (VI) and 1.03 g
(3.55 mmol) of 2-nitro-4-trifluoromethylbenzene sulfonyl chloride
were added 50 ml of dichloromethane in the dark and the solution
was cooled to ꢁ5 ꢂC. Under nitrogen 359 mg (3.55 mmol) of
triethylamine in 15 mL of dichloromethane were added, the
mixture was stirred for 24 hours at room temperature and filtered.
The filtrate was washed with water and sodium bicarbonate
solution (1%) and the crude product was recrystallized with
dichloromethane/diethylether. 1.45 g (76%) of PAG 3 were
obtained.
1H-NMR (300 MHz, CDCl3), d ¼ 8.71 (dd, 2H), 8.63 (dd, 2H),
8.39 (d, 2H), 8.35 (d, 2H), 8.24 (d, 2H), 8.11 (dd, 2H), 7.83 (dd,
2H), 7.61–7.57 (m, 4H), 7.36 (d, 2H) ppm. 19F NMR (282 MHz,
CDCl3), d ¼ ꢁ63.7 ppm. Elem. Anal.: Calcd: C, 49.35; H, 1.88;
N, 5.23; F, 10.64. Found: C, 49.25; H, 1.43; N, 5.12; F, 10.44. UV
(acetonitrile): lmax ¼ 392 nm (3 ¼ 2990 m2/mol).
N,N0-Di-(2,6-difluorobenzenesulfonyloxy)-4,40-(phenylene-1,3-
dioxy)dinaphthalene-1,8;10,80-tetracarboxylic imide (PAG 2). In
the dark, 1 g (1.88 mmol) of the hydroxyimide (V) and 0.38 g
(3.76 mmol) of triethylamine were dissolved in 40 mL of cold
dichloromethane (CH2Cl2). The solution was cooled and 0.8 g
(3.76 mmol) of 2,6-difluorobenzene sulfonyl chloride in 10 mL
of CH2Cl2 were added in 30 min. The mixture was stirred for
24 hours at room temperature and filtered. The filtered solid
was collected, stirred with dichloromethane for one hour and
again filtered. The combined filtrates were washed with water
and sodium bicarbonate solution (1%) and dried with magne-
sium sulfate. After recrystallization with dichloromethane/
diethylether, 383 mg (23%) of PAG 2 were obtained.
1H NMR (300 MHz, acetone), d ¼ 8.85 (dd, 2H), 8.62 (dd,
2H), 8.48 (d, 2H), 8.01–7.94 (m, 4H), 7.76 (t, 1H), 7.42–7.33 (m,
7H), 7.28 (d, 2H) ppm. 19F NMR (282 MHz, acetone), d ¼
ꢁ106.8 (dd) ppm. Elem. Anal.: Calcd: C, 57.02; H, 2.28; N, 3.17;
F, 8.59. Found: C, 56.91; H, 2.16; N, 3.15; F, 8.60. UV (aceto-
nitrile): lmax ¼ 362 nm (3 ¼ 2950 m2/mol).
1-Tetradecyloxy-3,5-dihydroxybenzene (I). 2.38 g (79.5 mmol)
of sodium hydride were washed with 10 mL of hexane to remove
fats and suspended in 10 mL DMF. The colourless suspension
was cooled to 0 ꢂC. Under nitrogen a solution of 5 g (39.5 mmol)
pholoroglucinol in 15 mL of DMF were added to this suspen-
sion. After stirring the reaction mixture for two hours at room
temperature, the same was again cooled to 0 ꢂC and 13.17 g
(47.5 mmol) of 1-bromo-tetradecane in 5 mL DMF were added.
The mixture was stirred for three weeks at room temperature,
200 ml water was added and the product was extracted with ethyl
acetate. After drying the organic phase with magnesium sulfate
the solvent was removed in vacuum and the oily residue was
purified by column chromatography using ethyl acetate/
petroleum ether as eluent. Yield: 2.2 g (17%). Rf (ethyl acetate/
petroleum ether, 1/3 by volume): 0.58. 1H-NMR (300 MHz,
CDCl3), d ¼ 6.00 (d, 2H), 5.94 (t, 1H), 5.01 (s, 2H), 3.87 (t, 2H),
1.75 (quint, 2H), 1.48–1.18 (m, 22H), 0.88 (t, 3H) ppm.
Bis-(4-thio-1,8-naphthalic anhydride)phenylene (III). Under
nitrogen, 1 g (7.03 mmol) of 1,3-benzenedithiol and 1.95 g
(14.06 mmol) of potassium carbonate were suspended in 20 mL
of dimethylacetamide (DMAc) and heated to 70 ꢂC. 3.9 g
(14.06 mmol) of 4-bromonaphthalic anhydride in 50 mL DMAc
were added and the suspension was heated to 150 ꢂC for 12 hours.
The mixture was cooled and the precipitate was filtered and
washed with chloroform. The combined chloroform phases were
washed with water and dried and the product was precipitated
with hexane and dried under vacuum at 60 ꢂC for 12 hours. 2.01 g
1,3-Bis-(4-oxy-1,8-naphthalic
anhydride)-5-(tetradecyloxy)-
phenylene (IV). 1.5 g (4.65 mmol) of (I) and 1.28 g (9.03 mmol) of
potassium caꢂrbonate were suspended in 20 mL of DMF and
heated to 70 C under nitrogen. 2.58 g (9.3 mmol) of 4-bromo-
naphthalic anhydride were added together with 15 mL of DMF
and the suspension was heated to 150 ꢂC for 20 hours. The
mixture was cooled and the solvent was removed under vacuum.
Water (500 mL) was added to the residue and the product was
extracted with dichloromethane. After the addition of methanol
(50 mL) the product precipitated. The water phase was acidified
with hydrochloric acid and again extracted with dichloro-
methane. Methanol was added and the mixture was stored in
a refrigerator for 2 weeks. The combined precipitates were
filtered and dried yielding 710 mg (21%) of (IV).
1
of a yellow, fluorescent powder was obtained (54%). H-NMR
(300 MHz, DMSO), d ¼ 8.67 (dd, 2H), 8.57 (dd, 2H), 8.39 (d,
2H), 7.96 (dd, 2H), 7.70–7.63 (m, 4H), 7.50 (d, 2H) ppm.
N,N0-Dihydroxy-4,40-(phenylene-1,3-dithio)dinaphthalene-1,8;10,
80-tetracarboxylic imide (VI). 2 g (3.74 mmol) of compound (III)
and 0.6 g (8.73 mmol) of hydroxylamine hydrochloride were
suspended in 30 mL of pyridine. The mixture was heated for 3
hours at 90 ꢂC, cooled and most of the solvent was removed under
reduced pressure. The product was precipitated with distilled
water and dried. Yield: 1.75 g (83%).
1H-NMR (300 MHz, CDCl3), d ¼ 8.76 (dd, 2H), 8.68 (dd, 2H),
8.54 (d, 2H), 7.85 (dd, 2H), 7.14 (d, 2H), 6.71 (d, 2H), 6.64
(t, 1H), 3.97 (t, 2H), 1.79 (quint, 2H), 1.50–1.20 (m, 22H), 0.87
(t, 3H) ppm.
1H-NMR (300 MHz, DMSO), d ¼ 10.78 (s, 2H), 8.60 (dd, 2H),
8.54 (dd, 2H), 8.38 (d, 2H), 7.91 (dd, 2H), 7.62–7.58 (m, 4H), 7.52
(d, 2H) ppm.
N,N0-Dihydroxy-4,40-(phenylene-5-tetradecyloxy-1,3-dioxy)-
dinaphthalene-1,8;10,80-tetracarboxylic imide (VII). 600 mg (0.84
mmol) of (IV) and 250 mg (3.68 mmol) of hydroxylamine
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 505–513 | 507

hydrochloride were suspended in 30 mL of pyridine and heated
at 90 ꢂC for 4 hours. The mixture was cooled and poured into ice
water (100 mL). The orange precipitate was filtered, washed
with water and dried under vacuum for 6 hours at 60 ꢂC. Yield:
600 mg (92%).
1H-NMR (300 MHz, DMSO), d ¼ 10.67 (bs, 2H), 8.66 (dd,
2H), 8.57 (dd, 2H), 8.47 (d, 2H), 7.92 (dd, 2H), 7.23 (d, 2H), 6.86
(d, 2H), 6.84 (t, 1H), 4.00 (t, 2H), 1.68 (quint, 2H), 1.45–1.15
(m, 22H), 0.82 (t, 3H) ppm.
where Sx, Cx, s*x, and xx are the slope of the fit, the concentration,
the action cross-section, and the instrument collection efficiency
(comprised of the dye emission characteristics, filter response
and detector quantum efficiencies) for either the unknown (unk)
dye or the standard (ref). These values were recorded from 700
to 1000 nm in 10 nm increments.
2.3.3 Measurement of the quantum yield. To extract the
absorption cross-section from the action cross-section it was
necessary to determine the dye’s quantum yield. As it was diffi-
cult to assess the quantum yield for the case of two-photon
absorption, the assumption was made that the two-photon
and one-photon quantum yields are the same, and the latter
value was measured following the procedure outlined in ref. 21.
Samples of each dye were prepared at six different concentrations
in acetone. Two fluorescence emission curves were collected
using 360 and 375 nm excitation using a PTI (Photon Tech-
nology International, Birmingham, NJ) fluorimeter. The
absorption of each of these concentrations was measured
between 300 and 700 nm using a Varian Cary 300 spectropho-
tometer. To correct for the bandwidth of the fluorescence
N,N0-Di-(2-Nitro-4-(trifluoromethyl)benzenesulfonyloxy)-4,40-
(phenylene-5-tetradecyloxy-1,3-dioxy)dinaphthalene-1,8;10,80-tet-
racarboxylic imide (PAG 4). In the dark, 250 mg (0.33 mmol) of
the hydroxyimide (VII), 186 mg (0.66 mmol) of 2-nitro-4-
trifluoromethylbenzene sulfonyl chloride and 20 mL of cold
dichloromethane were combined in a three neck flask and the
ꢂ
mixture was cooled to ꢁ5 C. Over a period of 15 min 65.6 mg
(0.66 mmol) of triethylamine in 10 mL of cold dichloromethane
were added dropwise. The mixture was stirred for 24 hours at
room temperature and filtered. The filtrate was collected,
washed with water and sodium bicarbonate solution (1%) and
dried with magnesium sulfate. The solvent was removed and the
crude product was recrystallized with dichloromethane/hexane
yielding 170 mg (41%) of a brownish powder (PAG 4). 1H-NMR
(300 MHz, CDCl3), d ¼ 8.76 (dd, 2H), 8.64 (dd, 2H), 8.49 (d,
2H), 8.37 (d, 2H), 8.23 (d, 2H), 8.07 (dd, 2H), 7.83 (dd, 2H), 7.13
(d, 2H), 6.69 (d, 2H), 6.59 (t,1H), 3.97 (t, 2H), 1.79 (quint, 2H),
1.45–1.20 (m, 22H), 0.87 (t, 3H) ppm. 19F NMR (282 MHz,
CDCl3), d ¼ ꢁ63.7 ppm. Elem. Anal.: Calcd: C, 55.68; H, 3.87;
N, 4.48; Found: C, 55.56; H, 3.86; N, 4.47. UV (acetonitrile):
lmax ¼ 365 nm (3 ¼ 3320 m2/mol).
excitation beam, the absorption was taken as:
ð
IexcðlÞAðlÞ dl
ð
hAi ¼
IexcðlÞ dl
where Iexc(l) was the profile of the excitation beam and A(l) the
measured absorption curve. Plots of the absorption versus inte-
grated fluorescence were made for each dye and the quantum
yield (Fx) determined as the ratio of the slope of each dye relative
to that of the standard anthracene in ethanol times the quantum
yield of anthracene (FAnth ¼ 0.2721):
2.3 Methods
2.3.1 Characterization. 1H-NMR spectra and 19F-NMR
spectra were recorded at a Bruker AC 300 spectrometer. UV-Vis
spectra were measured using a Shimadzu UV-2102 PC spec-
trometer. Fluorescence measurements were undertaken using
a Perkin Elmer LS50B unless otherwise noted. Elemental anal-
ysis was performed at the analysis lab of the University of Mainz.
Thermal analysis was performed on a Perkin Elmer Pyris 6 TGA
and a Perkin Elmer DSC 7 under nitrogen atmosphere at
Slopex
Fx ¼ FAnth
SlopeAnth
As ethanol and acetone have the same refractive index of
1.37,22 solvent index effects were ignored. To ensure that the
degradation of the PAGs does not affect the results of the fluo-
rescence measurements, the PAG solutions were exposed to UV
irradiation (Oriel 68910 Arc Lamp (500 W) using a 365 nm filter)
and the fluorescence spectra were measured. No significant
change in the fluorescence spectra were observed up to a radia-
tion dose of ꢃ400 mJ/cm2. Hence, we believe that there is no
significant effect of PAG degradation on the fluorescence
measurements.
ꢂ
a heating rate of 20 C per minute.
2.3.2 Measurement of the two-photon action cross-section. To
determine the two-photon action cross-section, the beam from
a mode-locked Ti:Sapphire laser (Spectra Physics Mai Tai) was
focused into
a cuvette containing the sample dyes at
2.3.4 Two-photon microfabrication. For two-photon micro-
fabrication a radical (A) and a cationic (B) curable negative
photoresist system was used. Resin (A) was prepared by mixing
Norland optical adhesive 63 (NOA 63) as polymerizable
monomer and the respective photoinitiator (0.5 wt% of PAG 2
and the same molar ratio of the other initiators with respect to
NOA 63). Norland NOA 63 resin is reportedly an unsaturated
polyurethane oligomer and is cross-linked using a mercapto-
ester compound. A small amount of acetone was used for better
solubility. A droplet of the clear mixture was put on a glass-
slide for two-photon microfabrication. For the preparation of
a concentration of 10 mM. The fluorescence vs. excitation
intensity was quantified by photon counting with a GaAsP
PMT (Hamamatsu) at different laser intensities. According to
the method detailed in ref. 20, the slope of a linear fit of the
observed photon counts versus the intensity squared is ratioed
against that of a known reference (10 mM Fluorescein in 0.1N
NaOH) to yield the action cross-section at a particular wave-
length.
ꢀ
ꢁꢀ
ꢁꢀ
ꢁ
SunkðlÞ
Sref ðlÞ
xref ðlÞ
Cref
s*2;unkðlÞ ¼
s*2;ref ðlÞ
xunkðlÞ Cunk
508 | J. Mater. Chem., 2009, 19, 505–513
This journal is ª The Royal Society of Chemistry 2009

Table 1 Solubility in various solvents and thermal properties of the
prepared PAGs
resin (B) MG2-OH, 15 wt% Powderlink cross-linker TMMGU,
5 wt% of PAG and 1 wt% of AF-69 (wt % with respect
to molecular glass) were dissolved in PGMEA. After spin-
coating the solution on a glass-slide the film was baked at
Solubility (wt%)
Thermal properties
TM (ꢂC)
TD (ꢂC)
ꢂ
PAG
MeCN
DCM
Toluene
115 C for 60 s.
For two-photon excitation the resins were exposed using
an Olympus IX70 confocal laser scanning microscope with
Lambda Physik titanium:sapphire laser operating mode-locked
(ꢃ80 MHz, ꢃ100 fs pulse width) at a wavelength of 780 nm. The
samples were mounted on a computer-controlled 3D translation
stage. The incident beam was focused through a water objective
(40 ꢀ ꢃ1.15 NA) and the 3D pattern was defined by scanning the
focal point of the laser beam inside the resin voxel by voxel with
each voxel separated by 0.5 mm in x,y-plane and 0.7–1.0 mm in the
z-plane. Exposure times were 100 ms per voxel. The laser power
at the entrance of the objective was measured with a Spectra
Physics 407A power meter and the percentage of the total power
reaching the surface of the resin was controlled through COMOS
software. After two photon exposure, the samples prepared with
Norland 63 were developed in ethyl acetate for 30 sec. Devel-
opment of the molecular glass photoresist was done in a 1:20 (by
volume) mixture of tetramethylammonium hydroxide (TMAH)
PAG 1
PAG 2
PAG 3
PAG 4
3
0.5
0.8
3
0.1
<0.1
0.4
10
158
154
253
92
312
328
311
294
0.1
0.5
0.5
10
3.2 Solubility and thermal properties
Important properties which determine the application and the
quality of photoacid generators are solubility and thermal
stability. Solubility measurements have been investigated using
solvents ranging from high (acetonitrile) to low polarity
(toluene). The results are shown in Table 1. PAG 1 showed good
solubility in polar solvents such as acetone, acetonitrile or ethyl
acetate, which have been used for development in the micro-
fabrication processes. Dichloromethane and tetrahydrofuran
proved to be good solvents for PAG 3. The solubility for the
PAG 2 in all solvents was only moderate but was sufficient for
the conducted experiments. On the other hand, the solubility of
the PAG 1 could be increased by adding an alkyl chain to the
PAG core. The resulting PAG 4 had a solubility of over 10 wt%
in solvents of low and medium polarity.
ꢂ
in water after a postbake of 120 s at 90 C.
The samples prepared for imaging were sputtered for 30 s with
Au/Pd and examined through a LEICA 440 scanning electron
microscope (SEM) at 15 kV and 1.0 nA.
To analyze the thermal stability of the compounds, TGA and
DSC measurements were undertaken. The results of these
experiments are also summarized in Table 1. All compounds
show high thermal stability with decomposition temperatures
3. Results and discussion
ꢂ
ꢂ
ranging from ꢃ300 C to 330 C. These TD values confirm the
high thermal stability which has been observed for other PAGs
based on N-hydroxy-sulfonamides.23 The incorporation of an
alkyl chain (PAG 4) doesn’t have a large influence on the
decomposition temperature, but decreases the melting point.
Concerning general stability, the PAGs didn’t show a change
in their properties (1H-NMR analysis) over a period of six
months, if stored at room temperature in a dark and dry place.
This demonstrates that they do not decompose and are stable
over a long period of time.
3.1 Design and preparation of non-ionic photoacid generators
A non-ionic photoacid generator advantageous for two-photon
microfabrication should have both a high two-photon absorp-
tion cross-section at the wavelength of the laser used and the
ability to produce a strong acid upon irradiation. In addition the
structure of such a PAG should also ensure moderate solubility
and high thermal stability. Donor–acceptor substituted naph-
thalene systems were developed as TPA ‘‘chromophores’’ in
which two units are linked via their donor-groups. The four new
PAGs (PAG 1–4) were synthesized as shown in the reaction
scheme (Fig. 1). In order to increase the polarizability of the
resulting PAG structure and thereby shift the absorption
towards longer wavelength the linking oxygen atom (PAG 1) was
replaced with sulfur (PAG 3) using similar reaction conditions.
In all of the PAGs the sulfonyl group serves as acceptor group as
well as anion of a strong acid. Therefore it was also of interest to
determine the effect of changing the sulfonyl group (PAG 1 and
2) on the optical properties as well as the efficiency in the
formation of acid under one and two-photon irradiation. The
chosen scheme is a compromise between solubility and absorp-
tion strength and the resulting PAGs offer some advantages
compared with more extended conjugated systems. The flexible
structure provides a better solubility and one molecule of the
PAG contains two potential acid groups, which helps to improve
its efficiency. By the introduction of an alkyl chain (PAG 4), the
solubility of the PAG based on resorcinol could be increased
further.
3.3 Optical properties
The optical characterization of the prepared PAGs in terms of
linear and two-photon absorption was relevant for the estima-
tion of their lithographic performance. Therefore the PAGs were
compared to the commercial available initiator ITX as well as the
sensitizing dye AF-69, which both are widely used systems in
two-photon lithography.24,25 ITX and AF-69 were also used for
comparison in the two-photon microfabrication experiments.
The PAG performance in two-photon microfabrication and in
photoacid formation upon UV-irradiation was furthermore
compared to the commercial available non-ionic PAG N-
hydroxynaphthalimide triflate (NHNTf). In Fig. 2 the molecular
structures of these systems are shown.
3.3.1 Linear absorption. In Fig. 3a the absorption spectra in
acetonitrile of the prepared compounds as well as photoinitiator
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 505–513 | 509

394 nm, respectively, as well as the broad absorption bands,
a strong absorption at a wavelength of 780 nm can be expected,
which is the wavelength used for the two-photon experiments.
The lmax in the absorption spectra of PAG 3 has a red-shift of
30 nm in comparison with the other compounds. This results
from the higher donor strength of the sulfur atom. Furthermore,
PAG 3 has a broader absorption band showing three maxima in
contrast to the other compounds, which indicates a higher
number of allowable excited states. The alkoxy group in PAG 4
increases the donor strength slightly and leads to a red-shift
of the absorption maximum compared to PAG 1 and PAG 2.
However, the shape of the absorption band doesn’t change. The
different arylsulfonyl moieties used for PAG 1 and PAG 2 do not
have much influence on the optical properties of the compounds.
Although, PAG 1 has a slightly stronger absorption than PAG 2
the absorption maxima are at the same wavelength of 362 nm in
both cases. Generally speaking, the absorption properties are
controlled by the core. Indeed, the spectra of the corresponding
anhydrides II–IV had the same shape but were blue shifted by
about 10 nm.
Fig. 2 Chemical structure of the compounds which have been used for
comparison: Nonionic PAG N-hydroxynaphthalimide triflate (NHNTf),
two-photon initiator isopropylthioxanthone oxime (ITX) and two-
photon sensitizer AF-69.
Fluorescence quantum yields (see Experimental section) have
been measured for the compounds PAG 1 and PAG 2 in
acetone, and compared to ITX and the two-photon dye AF-69.
The values are reported in Table 2. The quantum yield for PAG
2 is more than twice as high as compared to PAG 1. Fluores-
cence is a process competing with other de-activation pathways
such as relaxation. Relaxation is relevant for the efficiency of
PAGs as it leads to bond breaking and the formation of radi-
cals. Thus it can be assumed that the higher fluorescence of
PAG 2 compared to that of PAG 1 is associated with a lower
efficiency in acid formation which was found in subsequent
experiments. Presumably the radicals of this compound are
stabilized by the nitro-group in 2-position of the anion similar
to the photochemical decomposition of 2-nitrobenzylesters.26
Hence the light intensity which is needed for the bond breaking
is lowered.
3.3.2 Two-photon absorption. To demonstrate the applica-
bility of the new PAGs for TPL we have measured TPA cross-
sections for compounds PAG 1 and PAG 2 and compared them
to the photoinitiator ITX and the two-photon dye AF-69. In
Fig. 4, the corresponding TPA spectra are shown. The cross-
section values are listed in Table 2. Details for the determination
of these values are given in the Experimental section.
Broad bands for two-photon absorption are detected for the
PAGs as well as the dye AF-69. The absorption maxima of the
new PAGs are observed at about 720 nm, which is approximately
twice the wavelength of linear absorption (362 nm). The cross-
section decreases from the high values at 720 nm (ꢃ35 GM) on
increasing the wavelength and gets relatively low at wavelengths
above 840 nm. At 780 nm, which is the wavelength mostly used
for the two-photon microfabrication experiments, the cross-
section for TPA is, however, still acceptable and strongly
improved with respect to the two-photon initiator ITX. The dye
AF-69 has a larger cross-section for two-photon absorption.
However, it is neither a radical initiator nor a PAG itself. For use
in an initiator system it has to be combined as a sensitizer with
PAGs with low cross-section. Problems in energy transfer may
limit the efficiency of such systems.
Fig. 3 Absorption spectra (A) and fluorescence spectra (B) of the
prepared PAGs are shown and compared to the commercially available
photoinitiator ITX. Fluorescence was not detectable for PAG 4 in
acetonitrile.
ITX are shown. The results of the absorption and fluorescence
measurements are summarized in Table 2. The corresponding
fluorescence spectra are shown in Fig. 3b. It can be assumed that
the absorption maximum for two photon absorption is about
twice the absorption maximum of a single photon process. Due
to the position of the absorption maxima lmax at 362, 365 and
510 | J. Mater. Chem., 2009, 19, 505–513
This journal is ª The Royal Society of Chemistry 2009

Table 2 Optical properties of the synthesized PAGs as compared to commercially available ITX and AF-69a b
,
Compound
labs.max (nm)
3 (m2/mol)
l
em,max (nm)
Dl (nm)
F
d (GM)
PAG 1
PAG 2
PAG 3
PAG 4
ITX
362
362
394
365
382
390
3170
2950
2990
3320
530
436
436
482
n.m.
415
490
74
74
90
n.m.
33
100
0.008
0.025
n.m.
n.m.
0.004
0.708
12 ꢄ 1
10 ꢄ 1
n.m.
n.m.
1.9 ꢄ 0.3
AF-69c
n.m.
179 ꢄ 18
a
Values for absorption and fluorescence are measured in acetonitrile, quantum yields F and TPA cross-sections d measured in acetone. The listed TPA
cross-sections are measured at a wavelength of 780 nm. Uncertainties for the determination of the quantum yields F and the absorption coefficients 3
were estimated as 10 and 3%, respectively. n.m. ¼ not measured. c Absorption and fluorescence maxima in CH2Cl2 (ref. 27).
b
rely on different mechanisms. Here we have used two different
resist formulations to (1) demonstrate the applicability of the
PAGs in microfabrication processes, which proceed either by
a cationic or a radical pathways and (2) show the PAG efficiency
in these experiments in comparison with commercial available
photo initiator systems. As a model for a radical curable system
the photoresist Norland 63 (NOA 63) together with either the
PAGs (PAG 1 or PAG 2) or the photoinitiator ITX was tested.
This resist consists of urethane oligomers with C]C double
bonds and mercapto-ester oligomers as cross-linking agents.
Although cross-linking may proceed through a free-radical as
well as a cationic pathway we assumed that the radical mecha-
nism plays the major role since ITX works with this resist but not
with the molecular glass which was used as cationic cross-
linkable system (resin (II)). A ratio of 0.5 weight% of PAG 2 and
Fig. 4 TPA spectra of PAGs (PAG 1 and 2) as compared to AF-69 and
the same mol% of PAG 1 was used for the microfabrication.
ITX. The size of each symbol approximately represents the uncertainty of
With this concentration a significant difference in the threshold
the data.
power for polymerization (Pth) was observed when compared to
resin formulations with the same molar ratio of the other initi-
ators. Pth is defined as the local absorbed energy density below
which no polymerization occurs. In Table 3, the Pth required to
obtain microstructures with NOA 63 using the different PAGs/
initiators is shown.
3.4 Determination of photoacid formation
The ability of the new PAGs to generate acid upon irradiation in
solution as well as in a polymer film was investigated using UV-
and IR-spectroscopy.28 To detect acid generation in solution we
used the pH-dependent optical properties of Rhodamine B,
whose protonated form absorbs strongly at 555 nm. Solutions of
the four new PAGs with a concentration of 85 mM in acetonitrile
were prepared containing Rhodamine B base in a concentration
of 8 mM. The samples were irradiated with a broadband UV-Vis
source. Thereby the formerly colourless solutions turned red
because of the protonation of the Rhodamine B lactone ring.
Photoacid formation in a film was investigated by measuring
the acid catalyzed deprotection of poly[4-(tert-butoxy-
carbonyl)oxy]styrene (PTBOCS). Films of this polymer con-
taining the respective PAG were spin-coated on a silicon wafer
and exposed with the UV lamp and postbaked. IR-measure-
ments confirmed the deprotection of the tert-butoxycarbonyl-
groups and the formation of polyhydroxystyrene (PHS) which
was visible from the decrease of the carbonyl band at 1763 cmꢁ1
Both resin containing the PAG 1 or PAG 2 feature a higher
sensitivity than the formulation containing the same molar ratio
of ITX. PAG 1 has the highest efficiency well above that of ITX.
This observation is in line with the TPA cross-section measure-
ments, fluorescence measurements and the results of the
photolysis experiments. In Fig. 5, 3D structures, which have been
written within this type of resist using PAG 2 and PAG 1 are
shown. The smallest features obtained using a 40ꢀ 1.15 NA
water objective was 0.6 mm. All patterns were fabricated at
a power of 1.5 mW, which is below the minimum power neces-
sary for the fabrication with ITX.
The use of a resin, which is cured via a cationic pathway,
finally should show the efficiency of the prepared compounds
as pure two-photon photoacid generators. For these
experiments a negative-tone system was used composed of
and the appearance of a hydroxyl band centered at 3398 cmꢁ1
.
Details for both experiments can be found in the ESI†.
Table 3 Pth required for the microstructuring of Norland 63 using the
prepared PAGs and ITX
3.5 Two-photon microfabrication
Compound
PAG 1
1
PAG 2
1.5
ITX
2
None
3.0
The potential of the synthesized PAGs in TPL was demonstrated
by preparing 3D microstructures with two kinds of resins, which
Power (mW)
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 505–513 | 511

that have been produced in the molecular glass resist are shown.
The lines have a width of 2 mm with a spacing of about 6 mm. The
P
th for this system was 1.4 mW. It is noteworthy that with PAG
NHNTf instead of PAG 1 no structures could be obtained using
TPL. AF-69 alone gave very poor structures and required very
high power.
4. Conclusion
Four new non-ionic photoacid generators based on naph-
thalimide moieties have been prepared. Their properties for
one-photon absorption are comparable to commonly used
UV-photoacid generator systems in terms of thermal stability
and acid generation efficiency. For one of the PAGs (PAG 4) the
solubility in apolar matrices was strongly improved. Enhanced
two-photon absorption of these PAGs as compared to
commercially available photoinitiating systems results in a higher
sensitivity for two photon lithography of acrylate- and non-
acrylate-based resist formulations. Hence, the PAGs proved to
be more efficient in TPL for radical as well as cationic induced
microfabrication. The development of sufficiently sensitive, non-
ionic two-photon absorbing photoacid generators have the
potential to overcome the shortcomings, particularly the energy
transfer process, of a two-component system involving an initi-
ator and a photosensitizer. This development of non-ionic TPA
PAGs is expected to increase the potential of TPL as a broadly
applicable patterning process.
Fig. 5 Microstructures fabricated with Norland 63 and 0.5 wt% of PAG
2 (a–d) and the same molar ratio of PAG 1 (e–f).
a hydroxyl-containing molecular glass (MG2-OH) and a cross-
linker (TMMGU). The respective molecular structures are shown
in Fig. 6a. An acid-catalyzed mechanism is proposed for this
system, which leads to cross-linked material in the exposed
area.29 Patterning of the molecular glass was possible using PAG
1 (5 wt%) by single-photon absorption at 365 nm. For two-
photon lithography, PAG 1 in combination with AF-69 has been
used (see Experimental section). In Fig. 6b grid-type patterns
Acknowledgements
L.S. thanks the DAAD for support for an exchange stay at the
Cornell University, where this work started. S. J. acknowledges
the support of NIH NSR01-044287 and the Nanobiotechnology
Centre (NBTC), an STC program of the National Science
Foundation under Agreement Number ECS-9876771. In addi-
tion, we thank Loon-Seng Tan and the Air Force Research Labs
for supply of AF-69 and for valuable, partial support of this
work. The authors wish to acknowledge the use of the micros-
copy and imaging facility at Cornell University and its facility
manager Carol Bayles. The authors also acknowledge the use of
the electron microscope, housed in the Cornell Center for
Materials Research Shared Experimental Facilities, supported
through the National Science Foundation Materials Research
Science and Engineering Center Program (DMR-0520404).
References
1 M. Goeppert-Mayer, Ann. Phys., 1931, 9, 273.
2 W. Denk, J. H. Strickler and W. W. Webb, Science, 1990, 248, 73.
3 W. Denk, PNAS, 1994, 91, 6629.
4 P. T. C. So, C. Y. Dong, B. R. Masters and K. M. Berland, Ann. Rev.
Biomed. Eng., 2000, 2, 399.
5 K. A. Kasischke, H. D. Vishwasrao, P. J. Fisher, W. R. Zipfel and
W. W. Webb, Science, 2004, 305, 99.
6 D. Yang, S. J. Jhaveri and C. K. Ober, MRS Bulletin, 2005, 30, 976.
7 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. S. Lee, D. McCord-Maughon, J. Qin, H. Ruckel, M. Rumi,
¨
X.-L. Wu, S. R. Marder and J. W. Perry, Nature, 1999, 398, 51.
8 S. M. Kuebler, M. Rumi, T. Watanabe, K. Braun, B. H. Cumpston,
A. A. Heikal, L. L. Erskine, S. Thayumanavan, S. Barlow,
S. R. Marder and J. W. Perry, J. Photopolym. Sci. Technol., 2001,
14, 657.
Fig. 6 (a) Molecular glass system used for TPL via a cationic-cross-
linking mechanism. (b) SEM micrographs of structures fabricated at
a power of 1.4 mW in molecular glass photoresist using PAG 1 and
AF-69.
512 | J. Mater. Chem., 2009, 19, 505–513
This journal is ª The Royal Society of Chemistry 2009

9 D. A. Parthenopoulos and P. M. Rentzepis, Science, 1989, 245, 843.
10 H. E. Pudavar, M. P. Joshi, P. N. Prasad and B. A. Reinhardt, Appl.
Phys. Lett., 1999, 74, 1338.
11 S. Kawata and Y. Kawata, Chem. Rev., 2000, 100, 1777.
12 G. S. He, J. D. Bhawalkar, C. F. Zhao, C. K. Park and P. N. Prasad,
Opt. Lett., 1995, 20, 2393.
13 A. Abbotto, L. Beverina, R. Bozio, S. Bradamante, C. Ferrante,
G. A. Pagani and R. Signorini, Adv. Mater., 2000, 12, 1963.
14 K. J. Schafer, J. M. Hales, M. Balu, K. D. Belfield, E. W. Van
Stryland and D. J. Hagan, J. Photochem. Photobiol. A, 2004, 162, 497.
15 G. S. He, L.-S. Tan, Q. Zheng and P. N. Prasad, Chem.Rev., 2008,
108, 1245.
16 W. H. Zhou, S. M. Kuebler, K. L. Braun, T. Y. Yu, J. K. Cammack,
C. K. Ober, J. W. Perry and S. R. Marder, Science, 2002, 296, 1106.
17 T. Yu, C. K. Ober, S. M. Kuebler, W. Zhou, S. R. Marder and
J. W. Perry, Adv. Mater., 2003, 15, 517.
D. E. Bergbreiter and C. R. Martin, Plenum Press, New York,
1989, pp. 192–200.
20 C. Xu and W. W. Webb, J. Opt. Soc. Am. B, 1996, 13, 481.
21 Jobin Yvon Ltd., Guide to Recording Fluorescence Quantum Yields,
http://www.jobinyvon.com.
22 American Institute of Physics Handbook, ed. D. E. Gray, 2nd edn,
McGraw-Hill, New York, 1963, Table 6e–1.
23 T. Asakura, H. Yamato, A. Matsumoto, P. Murer and M. Ohka,
J. Photopolym. Sci. Technol., 2003, 16, 335.
24 K. D. Belfield, K. J. Schafer, Y. Liu, X. Ren and E. W. Van Stryland,
J. Phys. Org. Chem., 2000, 13, 837.
25 C. A. Coenjarts and C. K. Ober, Chem. Mater., 2004, 16, 5556.
26 M. Shirai and M. Tsunooka, Progress of Polymer Science, Pergamon
Press, 1996, vol. 21, pp. 1–45.
27 K. D. Belfield and K. J. Schafer, Chem. Mater., 2002, 14, 3656.
28 R. Ayothi, Y. Yi, H. B. Cao, W. Yueh, S. Putna and C. K. Ober,
Chem. Mater., 2007, 19, 1434.
18 M. Shirai and M. Tsunooka, Bull. Chem. Soc. Jpn., 1998, 71, 2483.
´
19 J. M. J. Frechet, F. M. Houlihan, F. Bouchard, E. Eichler,
S. MacDonald, A. Hult and G. Willson, in Functional Polymers, ed.
29 Q. Lin, A. Katnani and C. G. Willson, Proc. SPIE, 1997, 3049,
974.
This journal is ª The Royal Society of Chemistry 2009
J. Mater. Chem., 2009, 19, 505–513 | 513