Effect of head group size on the photoswitching applications of azobenzene Disperse Red 1 analogues

Article abstract of DOI:10.1039/c4tc00996g

We investigate the effect of the increased molecular bulk in the 'head' group for a class of newly synthesized azobenzene chromophores with a clickable ethynyl group para and a nitro group ortho to the azo bond on the distal benzene ring. This 'variable-head' functionalization provides a family of dyes with photophysical characteristics very similar to those of Disperse Red 1, one of the most commonly used azo dyes in materials science. Phenyl, naphthyl, and anthracyl derivatives were synthesized as small molecules, monomers, homopolymers, and copolymers in a rapid and facile manner using click chemistry, confirming the versatility of this parent chromophore. Photochemical and spectral studies indicate that this strategy is suitable to build a 'bulkiness series' of stimuli-responsive materials, as the various material derivatives retain the absorption and kinetic characteristics of the parent chromophore necessary for all optical patterning applications that DR1 dyes have been optimized for. In thin films, larger head group size was found to increase the stability of light-induced birefringence in copolymers. The homopolymers formed stable surface-relief gratings upon interference irradiation, whose grating depths correlate with head group size, demonstrating that this new class of polymers can also undergo tailored macroscopic photoinduced motions, which could have applications in all optical nano-patterning. the Partner Organisations 2014.

Full text of DOI:10.1039/c4tc00996g

Volume 2 Number 36 28 September 2014 Pages 7439–7696
Journal of
Materials Chemistry C
Materials for optical, magnetic and electronic devices
www.rsc.org/MaterialsC
ISSN 2050-7526
PAPER
Christopher J. Barrett et al.
Effect of head group size on the photoswitching applications of
azobenzene Disperse Red 1 analogues

Journal of
Materials Chemistry C
View Article Online
View Journal | View Issue
PAPER
Effect of head group size on the photoswitching
applications of azobenzene Disperse Red 1
analogues†
Cite this: J. Mater. Chem. C, 2014, 2,
7505
Alexis Goulet-Hanssens,‡a T. Christopher Corkery,a Arri Priimagib
a
*
and Christopher J. Barrett
We investigate the effect of the increased molecular bulk in the ‘head’ group for a class of newly synthesized
azobenzene chromophores with a clickable ethynyl group para and a nitro group ortho to the azo bond on
the distal benzene ring. This ‘variable-head’ functionalization provides a family of dyes with photophysical
characteristics very similar to those of Disperse Red 1, one of the most commonly used azo dyes in materials
science. Phenyl, naphthyl, and anthracyl derivatives were synthesized as small molecules, monomers,
homopolymers, and copolymers in a rapid and facile manner using click chemistry, confirming the
versatility of this parent chromophore. Photochemical and spectral studies indicate that this strategy is
suitable to build a ‘bulkiness series’ of stimuli-responsive materials, as the various material derivatives
retain the absorption and kinetic characteristics of the parent chromophore necessary for all optical
patterning applications that DR1 dyes have been optimized for. In thin films, larger head group size was
found to increase the stability of light-induced birefringence in copolymers. The homopolymers formed
stable surface-relief gratings upon interference irradiation, whose grating depths correlate with head
group size, demonstrating that this new class of polymers can also undergo tailored macroscopic
photoinduced motions, which could have applications in all optical nano-patterning.
Received 13th May 2014
Accepted 13th July 2014
DOI: 10.1039/c4tc00996g
www.rsc.org/MaterialsC
groups across the azo bond in DR1 (1, Fig. 1) which provides a
red-shied absorbance far into the visible region as well as a far
faster thermal reconversion of the cis- to the trans-state.6 These
AONP light-driven phenomena are now best-thought to be
caused by free volume changes of the azobenzene as it
undergoes its trans–cis–trans isomerization cycle,7–9 which
motivates our current study.
The ability to increase or tune the degree of free-volume
change during isomerization would allow for improved
design of current materials. To explore this optimization we
designed and prepared analogues (Fig. 1) that maintain the
features that make DR1 the main azo-workhorse in the
Introduction
Light-responsive applications requiring reversible, easy and
rapid actuation that is critical to the materials' success oen
rely on azobenzene as the working photoswitch. The light-trig-
gered isomerization of the azobenzene (azo) chromophore from
the trans- to the cis-geometries when irradiated with visible light
drives material response, and Disperse Red 1 (DR1)-based azo
materials can impart large changes to surface properties and
macromolecular structures.1,2 Overlapping cis- and trans-
absorption bands make this class of chromophores well suited
for All Optical Nano-Patterning (AONP) such as surface relief
grating (SRG) formation, photoinduced birefringence studies
and photomechanics3–5 since many reconversion cycles can be
triggered by the same wavelength of light. This is possible due
to the push–pull feature of the amino and nitro substituted
aDepartment of Chemistry, McGill University, 801 Sherbrooke St. West, Montreal, QC,
H3A 0B8, Canada. E-mail: christopher.barrett@mcgill.ca
bAalto University, Optics and Photonics Group, Department of Applied Physics, P.O.
BOX 13500, FI-00076 Aalto, Finland
† Electronic supplementary information (ESI) available: Synthesis details, 1H, 13C,
HRMS and GPC analyses of materials where appropriate and possible. Additional
UV/vis as well as kinetic data from the optical testing of solutions and lms. See
DOI: 10.1039/c4tc00996g
¨
‡ Current address: Department of Chemistry, Humboldt-Universitat zu Berlin,
Fig. 1 The chromophores DR1 (1), diethylaniline-DR1 (2), ortho-DR1
(3), and ethynyl ortho-DR1 (4).
Brook-Taylor-Str. 2, 12489 Berlin, Germany.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 7505–7512 | 7505

View Article Online
Journal of Materials Chemistry C
Paper
materials literature, while providing an easily functionaliz-
able ‘head’ group. The pseudo-stilbene chromophore, 2,
features a 4-40 push–pull amino-nitro structure across the azo
bond that makes its absorption well suited for rapid reversible
actuation by lamp or laser irradiation. Relocating the nitro
group from the para position to an ortho position retains the
push–pull across the azo bond while freeing the para position,
considered to be the most important in terms of effecting
photoinduced changes in free volume,10,11 for further func-
tionalization. Although the ortho structure (3) has been
known for decades, it has only been studied in depth in the
commercial dye literature,12,13 and not in the context of
functional materials. Placing an ethynyl group in the para
position (4) allows for rapid functionalization using now-
common copper-catalyzed alkyne azide cycloaddition
(CuAAC, or ‘click’) chemistry.14 As we reported in a previous
publication, this approach is an easy and versatile way to
rapidly generate diverse azo structures while maintaining the
switching properties of the unfunctionalized dye.15
A possible reason for these o-nitro structures being absent in
the materials literature can be attributed to evidence that
functionalization of azo dyes ortho to the azo bond can decrease
their photostability.16–18 For investigating this issue, photo-
stability has been studied for several diethylaniline dyes
including DR1 (1) and its ortho analogue (3). It was reported that
both of these dyes photodecompose, with ortho-DR1 bleaching
faster than DR1 in acetone solutions in the presence of oxygen;
however, in deoxygenated acetone solutions or crystalline forms
the ortho species is more stable than DR1.19 Given the impact
DR1 has had in the literature, with proper engineering and
selection of handling parameters, ortho-DR1 could also act as a
viable tool in materials science.
To determine the effect of the head group on the dye prop-
erties, structure 4 was reacted with aryl azides to form new
‘head’ groups bearing an increasingly larger functionality
(Fig. 2). These were used to investigate the effects free volume
changes had on the photoisomerization of these structures. To
help elucidate the results observed in these materials, small
molecule analogues, monomers, and homopolymers of the
three derivatized azos were synthesized. As a proof of concept of
the variable preparation of these structures, copolymers were
prepared by diazotization as a mechanism of macromolecular
post-functionalization20 which could then be ‘clicked’ to the
nal chromophore structure. All of these structures can be
easily attained using click chemistry, providing derivatives with
similar optical properties.
Experimental
Material syntheses
With the goal of keeping these syntheses as facile as possible,
using the mildest conditions allowable, all reactions except for
Sonogashira couplings and polymerizations (dry solvent,
nitrogen atmosphere) were conducted under atmospheric
conditions at room temperature. 1H NMR spectra were acquired
at 298 K, on a Varian-Mercury 300 MHz or 400 MHz spectrometer
while 13C NMR spectra were acquired on a Varian-Mercury 300
MHz NMR spectrometer. Chemical shis are reported in ppm on
the d-scale referenced to either the solvent signal or the internal
TMS standard. High resolution mass spectrometry (HR-MS) was
acquired on a Thermo Scientic Exactive Plus Orbitrap. Samples
were ionized using either atmospheric-pressure chemical ioni-
zation (APCI) or electrospray ionization (ESI). All observed ions in
positive and negative ionization modes are reported. All chem-
icals were obtained from Sigma-Aldrich Corporation (St. Louis,
MO, USA), with the exception of trimethylsilylacetylene which
was obtained from Oakwood Chemicals (West Columbia, SC,
USA). For synthesis details, please see the ESI le.†
Polymer synthetic approach
Homopolymer samples were prepared by free-radical polymeri-
zation of pure monomers in THF initiated thermally by AIBN. For
the copolymer an approach to functionalize a common batch of
azo precursor polymer through a diazotization reaction with an
ethynyl-bearing head group (akin to 4) which could then be
‘clicked’ for further functionalization was targeted. To maintain
high solubility of the azo side chains a target functionalization of
2 mol% was attempted. Absolute characterization of the mono-
mer dye content was impossible within a reasonable margin of
error using NMR. The absorption coefficient of the copolymer
dyes is expressed as a mass extinction coefficient (ESI), and the
dye content for all copolymers was found to be between 0.8 and
2.1 mol%. Homopolymers were found to have molecular weights
ranging between 2490 and 3560 g molꢀ1 while the copolymers
had molecular weights ranging from 9350 to 10 390 g molꢀ1. Low
azo content is critical for optimal processing abilities in these
higher molecular weight polymers, and this low dye content
allows for efficient photoisomerization in the solid state and the
inscription of photoinduced birefringence.21,22 Further details
can be found in the ESI.†
Optical characterization in solution
Solution characterization of the small molecules and polymers
was conducted in dry spectral grade THF on a Cary Bio 300
UV/vis spectrophotometer to measure the absorption spectra
and lmax values. The isomerization kinetics were measured with
a Melles Griot Series 43 Ar+-ion laser with emission at 488 nm
that was set up in a pump–probe arrangement, as shown in
Fig. S15.† Thermal decay of the cis- to trans-state was monitored
by examining the transmittance of the samples over time: upon
the pump cycle, the absorbance of the sample solutions at
488 nm (which is used for both pumping and probing)
decreases due to trans-to-cis isomerization, and the thermal cis–
Fig. 2 Structures of the four model dyes (4–7) used in this study.
7506 | J. Mater. Chem. C, 2014, 2, 7505–7512
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
trans reconversion shows up as an intensity decrease (absor- absorbance is blue-shied, with a lmax of 452 nm, as opposed to
bance increase) over time. The power of the pump beam was the 489 nm lmax in 2. This 37 nm shi in the absorption
ꢁ300 mW cmꢀ2 with a pump cycle between 5 and 10 seconds to maximum is on the order of generally observed shis seen in
induce isomerization and ensure sample saturation back to the polyenes upon addition of an extra p system, and may be due to
¨
trans state. The probe beam was chopped mechanically at 1410 a shorter Huckel conjugation length in the donor–acceptor ends
Hz, attenuated, and passed through the sample into a photo- of the azobenzene.24,25 In terms of their thermal cis–trans
diode detector, where the intensity was recorded as a function isomerization, the ortho-DR1 species relaxes signicantly faster
of time. The pump irradiation time (ꢁ100–200 ms) was short than the native DR1 dye in solution, with a half-life of 0.13
enough to avoid heating the sample; so the kinetic measure- seconds compared to 1.1 seconds. This order of magnitude
ments can be considered to have been acquired at room increase in speed is possibly due in part to a more sterically
temperature.23
hindered cis state due to the ortho-nitro group, which would
accelerate reconversion.26 Overall, with the change to the ortho
species, the desirable absorbance is retained and the rapid
reconversion is improved. This latter fact permits more
switching cycles per unit time, increasing the effective work
done by photoresponsive materials as more time is spent in the
active switching portion of the isomerization cycle. In all of our
solution-based optical studies we have observed successful
switching for 100's of cycles without a decrease in switching
behaviour.
Optical characterization of thin lms
The photoresponsive properties of the homopolymers and
copolymers as thin lms were also studied. The polymers were
dissolved in THF and spin coated onto clean glass substrates to
yield lms with thicknesses in the range of 500 nm (homopoly-
mers) to 1.5 mm (copolymers). A thin lm (150 nm) of the
commercially available poly(DR1 acrylate) (Sigma-Aldrich) was
used as a reference. All lms were homogeneous and isotropic
as veried with polarized optical microscopy. The lms were
characterized with photoinduced spectral changes, birefrin-
gence, and surface-relief grating formation, using a collimated
and spatially ltered 488 nm Ar+-ion laser as a pump. The
excitation was carried out either with a circularly polarized
(photoinduced spectral changes, 50 mW cmꢀ2) or a horizontally
polarized (photoinduced birefringence, 150 mW cmꢀ2; SRG
formation, 300 mW cmꢀ2) light beam.
The absorption spectra in the dark and during irradiation
(488 nm, 50 mW cmꢀ2) were measured with a ber-optic spec-
trometer (Ocean Optics USB2000+) using a deuterium tungsten
halogen lamp as a light source (Ocean Optics DH-2000-BAL).
The photoinduced birefringence was measured with a normally
incident probe beam from a diode laser (l ¼ 820 nm). The
transmitted intensity of the probe beam through a polarizer/
sample/analyzer combination was monitored with a photo-
diode. The orientation of the polarizer/analyzer was set to ꢂ45ꢃ
with respect to the polarization of the excitation beam, and the
birefringence |Dn| was calculated from the transmission data I
as |Dn| ¼ l(pd)ꢀ1(I/I0)1/2, where I0 is the photodiode signal for a
parallel polarizer/analyzer orientation (in the absence of the
sample) and d is the lm thickness. The SRGs were inscribed in
a Lloyd mirror conguration by directing half of the expanded
incident beam directly onto the sample and reecting the other
half from a mirror set at right angle with the sample. The period
of the resulting interference pattern was set to ca. 1 um. The
build-up of SRGs was monitored in real time by measuring
diffraction from a l ¼ 820 nm probe beam. The height proles
of the inscribed SRGs were determined by atomic force
microscopy (Veeco Dimension 5000 SPM).
Molecular dyes
Model compounds with different para-substituents were
prepared by functionalizing 4 using ‘click’ chemistry (Fig. 2). As
we sought to explore the architectures that provided little or no
electronic effect on the azo chromophore and varied solely in its
molecular size, phenyl-(5), 2-naphtyl-(6), and 1-anthracyl-(7)
functionalized dyes were prepared. In these dyes, the electron
donation of the amine is balanced slightly with the 10 amine of
the triazole, resulting in a slight decrease in the push–pull effect
from the nitro group to the diethyl aniline amine. This leads to a
consistent hypsochromic shi in the absorbance wavelength
with respect to the lmax of the parent compound (4), as seen in
Fig. 3. Additionally some of the anthracene p / p* transitions
between 300 and 400 nm are clearly evident as ne features.27
The kinetic decays of the model dyes are of rst order (Fig. 4)
and the time constants are given in Table 1. The half-life
Results and discussion
Comparing o-DR1 to DR1
The absorption spectra and extinction coefficients for 2 and 3
Fig. 3 UV/vis spectra of ‘clickable’ (4) and ‘clicked’ (5–7) small mole-
are detailed in the ESI (Fig. S1†). The ortho-nitro species cules in THF.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 7505–7512 | 7507

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 5 Structures of the four model monomer dyes (8–11) used in this
study.
explanation for this observation is that there is no longer one
side that preferentially moves while the other remains still in
the isomerization event. Similar to the decay rate being
decreased with higher molecular head size, all of the monomers
now show a slower thermal relaxation to the trans state than
their corresponding non-monomer forms, since all possess a
larger ‘tail’ group.
Fig. 4 Absorbance normalized pump–probe cycles for small mole-
cules 4–7 in THF, showing the increase in the cis state upon irradiation
and thermal reconversion to the stable trans state.
increases systematically with increasing size of the head group,
from 0.18 s for the native dye (4) to 0.74 s for the naphthyl (6)
and anthracyl (7) derivatives. As the diethylamino “tail” is
identical in all instances, it indicates that the larger molecular
size of the head group leads to slower thermal relaxation. This
nding suggests that the para functionalization should not
cause prohibitive steric hindrance in the cis state, which would
cause these dyes to perform poorly as switches.
Aer validation of the approach using small molecules,
model monomer compounds, shown in Fig. 5, were studied as
an intermediate step between the model dyes and polymers to
elucidate the effect of further functionalization on the azo-
benzene tail to the photoresponse of these chromophores. As
with the model compounds, the ‘clicked’ monomers'
(compounds 9–11) lmax values blue shi from the ‘unclicked’
form (compound 8), as seen in Table 1, and also show a similar
slowing of the thermal relaxation rate with increasing size of the
clicked moiety, since the single bond character of the azo bond
decreases with the competing intra-ring electron donation. The
addition of the methacrylate tail causes a signicant increase in
the t1/2 when compared to the non-monomer dyes. A possible
Macromolecular dyes
Homopolymers (compounds 12–14, Fig. 6) were prepared by
free radical polymerization of pure monomers (9–11), to verify
the behaviour of these novel azos in the pure polymeric form.
The absorption prole (Fig. S9†) of the three homopoly-
mers in solution remains the same as for the corresponding
monomers, showing that polymerization has no effect on the
absorption of the azo dyes. The decay kinetics (Table 2) of
homopolymers 12–14 shows trends similar to that of the
small molecule dyes 5–7 in terms of half-lives. As the molec-
ular size increases, the thermal reconversion slows down, due
to a larger molecular ‘arm’ that has to be moved. However,
overall we generally observe faster isomerization rates than
that in the monomer form. This is likely due to the more
sterically congested cis states and suggests that the molecular
architecture has a large effect on determining the baseline
isomerization rate, with the trend in ‘head’ size determining
the relative rates.
Copolymers were synthesized by polymer post-functionali-
zation as opposed to copolymerization (Fig. 7). This was done as
a proof of concept, since diazotization has grown in prominence
as an easy way to generate amphiphilic photoresponsive poly-
mers,20 and could also be used to generate complex functional
Table 1 lmax and t1/2 for model diethyl-aniline compounds 4–7 and
monomer compounds 8–11 in THF
Compound
lmax (nm)
t1/2 (seconds) polymers. As mentioned above, azo copolymers synthesized
from monomers can suffer from low molecular weights and
2
3
4
5
6
7
8
9
489
452
482
477
476
473
471
465
467
468
1.10
high chromophore aggregation, two factors that are non-desir-
0.13
able in creating polymers for use in materials applications. The
0.18
approach employed here28 solves both of these issues as all
0.57
0.74
0.74
0.36
1.45
1.61
diversied polymers have the same ꢁ10 000 g molꢀ1 molecular
weight and spaced chromophore placement.
The spectra of the functionalized polymers (Fig. S11†) show a
bathochromic shi in lmax to 473 nm, as compared with 464 nm
10
11
for the starting material (15). These polymers exhibit thermal
reconversion that is best tted by a biexponential decay
0.93
7508 | J. Mater. Chem. C, 2014, 2, 7505–7512
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Fig. 6 Structures of the phenyl, naphthyl and anthracyl homopoly-
mers (12–14) used in this study.
Table 2 lmax and t1/2 (for biexponential decays: fast ¼ ta, slow ¼ tb) for
homopolymers 12–14 and for copolymers 15–18 in THF
Compound
lmax (nm)
ta (s)
tb (s)
12
13
14
15
16
17
18
466
466
466
464
473
472
454a
0.77
0.65
1.1
0.51
0.47
0.23
0.34
—
—
—
—
2.49
1.26
1.53
Fig. 7 Strategy for achieving diversified azo copolymers. An N-ethyl
anilino ethanol methacrylate homopolymer is diazotized in step A to
form an azo copolymer (15). This copolymer can be diversified in
step B with any head group to create a functionalized azo copolymer
(16–18).
a
The absolute absorbance peak for the anthracyl derivative is an
anthracene feature atop the azo absorption, and the azo maxima itself
is comparable to the other two derivatives at ꢁ470 nm.
measured (Table 3): the drops in absorbance are 11%, 8%,
and 5% for the phenyl, naphthyl, and anthracyl-functional-
ized copolymers, respectively. For the homopolymers 12 and
13 (high-quality lms of 14 could not be formed) the light-
induced decrease in absorbance is quite negligible, meaning
that only a very small fraction of the molecules converts to the
cis state. This indicates that in homopolymers, smaller co-
monomers are critical to permitting a reasonably high cis
state in the bulk material (Fig. 8b).
Fig. 9 presents the photoinduced birefringence curves for
the copolymers 15–18. Due to the low azobenzene content, the
absolute birefringence is relatively low for all samples, but
comparable to those of copolymers22 and doped polymers29
containing low concentration of DR1. For all samples, the
birefringence can be erased with circularly polarized light and
function. The fast component, as shown in Table 2, increases in
speed as the monomer head size increases. The observed data
indicate a behaviour similar to that of the monomer species,
and most likely does not mimic the homopolymer model as
there is less steric hindrance in these farther-spaced chromo-
phores than that in the homopolymer. As with the homopoly-
mer series, the polymeric forms of the dyes are more sterically
crowded leading to increased rates in the fast relaxation
component, which is evidenced in all derivatives.
Thin lms
While solution data yield important information on the rewritten on the same spot without fatigue. Like the isomer-
trends between different molecular structures, the photo- ization efficiency, the birefringence decreases systematically
responsive behaviour within thin azo-polymer lms is of with increasing size of the head group, suggesting that it is
greater practical interest. Spun-cast lms were investigated more difficult for larger molecular moieties to move within
rst for their isomerization potential as the chromophores are the polymer matrix in response to irradiation with polarized
in a much more tightly packed environment. Recording UV/ light. The decreased mobility is further supported by the fact
vis spectra of thin lms under constant irradiation allowed us that the residual birefringence increases with the head group
to measure the photostationary state of the chromophores. size, indicating that the free-volume change during isomeri-
The copolymer 15 can achieve a relatively large cis-population zation drives a more stable reorganization of the polymer
upon irradiation, as evidenced by the 15% decrease in lm. No SRGs could be inscribed into the copolymers, which
absorbance (Fig. 8a). As the head group is increased in size in can be attributed to the low content of the azobenzene
the copolymer family a notable drop in cis-content is moieties.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 7505–7512 | 7509

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 9 Photoinduced birefringence in copolymer samples 15–18,
demonstrating reduced birefringence with increased head group size.
The writing beam is turned on at 30 s, and the irradiation is ceased at
500 s.
for the smaller phenyl-substituted dye (12). Hence macroscopic
molecular movements seem to be enhanced by employing a
molecular switch requiring larger free-volume change upon
photoisomerization. Yet we note that compared to the widely
used and commercially available homopolymer poly(Disperse
Red 1 acrylate), which can be used as a gold standard for SRG
applications, the grating formation efficiency is low (Fig. 10a).
Another explanation for the higher diffraction efficiency of
13 as compared to 12 may be less chromophore packing
between neighbouring azobenzene units, due to the larger head
group size of 13 and the rotational symmetry of 12. This is
Fig. 8 (a) Isomerization extent of a film of polymer 15 spun cast on a
clean glass substrate from THF, demonstrating an absorbance
decrease of 15% under constant irradiation, and (b) the blue-shift and
decrease in absorbance change when homopolymer 12 is compared supported by the observation that when comparing the corre-
to copolymer 16 under constant irradiation. Irradiation in both figures
sponding homo- and copolymers (12 vs. 16; 13 vs. 17) the blue-
is performed with a 488 nm Ar+ laser at 50 mW cmꢀ2 intensity.
shi in the absorption spectrum – indicative of chromophore–
chromophore intermolecular interactions22,30
–
is more
pronounced for the phenyl-containing systems (ca. 14 nm; see
Fig. 10b) than for the napthyl-containing systems (ca. 8 nm).
The difference is small yet clearly evident, providing a plausible
explanation for the lower isomerization efficiency of 12 with
Table 3 lmax and absorbance decrease in the polymer films when
excited at 488 nm
Isomerization
lmax
Polymer
extent in lm
(nm) respect to 13, in contrast to the copolymers, for which 16
displays larger decrease in absorbance upon irradiation
than 17.
#12
#13
#15
#16
#17
#18
0.5%
1%
15%
11%
8%
471
475
472
485
483
468
The existence of stable SRGs on thin lms of 12 and 13 was
veried by imaging the sample surfaces with an atomic-force
microscope. The surface-modulation depths, as shown in
Fig. 10b, are ca. 15 nm and 27 nm for 12 and 13, respectively,
and a 3D view of an SRG inscribed on 13 is shown in Fig. 10c.
The modulation depths are in line with the diffraction curves
5%
Even though homopolymers 12 and 13 exhibit very low shown in Fig. 10a, since at the small-modulation-depth regime
isomerization efficiency, we successfully inscribed SRGs onto the diffraction efficiency is expected to scale quadratically with
lms of both polymers using interference irradiation. The the modulation depth.31 Even if the achieved modulation
build-up of the gratings was followed by monitoring the rst- depths are modest given that depths of hundreds of nm can be
order diffraction in real time, as shown in Fig. 10a. In contrast obtained, we note that only a vanishingly small fraction of the
to the photoinduced birefringence, the SRG formation effi- azobenzene molecular motors participates in driving the
ciency is higher for the napthyl-substituted dye (13) than that macroscopic motions in these systems. Hence we believe that
7510 | J. Mater. Chem. C, 2014, 2, 7505–7512
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Conclusions
We synthesized and tested a variety of complex azo architec-
tures based on a novel ortho-DR1 dye with a ‘clickable’ para-
ethynyl group that retains actuation and relaxation properties
while allowing for rapid structural variation. This model was
validated in small molecule, monomer, homopolymer as well as
in post-functionalized copolymer schemes. With the exception
of the copolymer species, the lmax of these chromophores shows
a modest blue shi aer ‘click’ functionalization. Post-func-
tionalization accelerates the decay kinetics of the dyes in solu-
tion, a fact that may seem counter-intuitive if considering only
molecular size and not steric hindrance in the cis state.
Homopolymers of these structures exhibit only a 0.5–1%
decrease in absorbance upon irradiation, demonstrating
limited cis content, yet could form relatively large depth surface
relief gratings. Copolymers made using a post-functionalization
strategy exhibited stable photoinduced birefringence, with the
long-term stability being determined by the ‘head’ group size.
Together these ndings indicate that the free-volume change
that azobenzenes undergo during trans–cis photo-isomerization
drives these macromolecular phenomena. We additionally note
that optimal response in terms of both SRG formation and
birefringence is yet to be achieved using a dye concentration
outside of the current study. Future work will focus on optimal
architectures where a fraction of chromophores may be func-
tionalized to maximize both mobility and mass motion in the
same material, allowing for more effective AONP. Macromo-
lecular post-functionalization using CuAAC ‘click’ chemistry of
this class of azos provides a rapid and facile way to generate
complex architectures, where the end groups may be given
additional functionality.
Acknowledgements
AGH would like to acknowledge FQRNT for a B2 Doctoral
Scholarship as well as a B3 Post-Doctoral Fellowship. TCC is
funded through an NSERC-CREATE grant. The authors would
like to thank Petr Fiurasek for help in polymer characterization
as well as the members of Zamboni Chemical Solutions: RZ, HZ,
JC, and MT for helpful discussions during synthesis.
Notes and references
1 M. Poprawa-Smoluch, J. Baggerman, H. Zhang,
H. P. A. Maas, L. De Cola and A. M. Brouwer, J. Phys. Chem.
A, 2006, 110, 11926–11937.
2 A. Natansohn and P. Rochon, Chem. Rev., 2002, 102, 4139–
4175.
3 Z. Mahimwalla, K. G. Yager, J. Mamiya, A. Shishido,
A. Priimagi and C. J. Barrett, J. Polym. Sci., Part A: Polym.
Chem., 2012, 69, 967–1006.
Fig. 10 (a) First order diffraction efficiency as a function of irradiation
time for pDR1A as a reference and homo-polymers 12 and 13, (b)
measured modulation depth in the films for polymers 12 and 13, and (c)
3-D AFM plot of surface relief grating formed by 300 mW cmꢀ2 488
nm Ar+ laser light in a thin film of polymer 13.
4 H. Koerner, T. J. White, N. V. Tabiryan, T. J. Bunning and
R. A. Vaia, Mater. Today, 2008, 11, 34–42.
5 K. G. Yager and C. J. Barrett, Azobenzene Polymers for Photonic
Applications, John Wiley & Sons Ltd, Boca Raton, 2009.
once the material's composition is properly optimized, the
photoinduced motions driven by the azo moieties reported in
this work can be greatly enhanced.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 7505–7512 | 7511

View Article Online
Journal of Materials Chemistry C
Paper
6 H. M. D. Bandara and S. C. Burdette, Chem. Soc. Rev., 2012, 19 T. Caronna, F. Fontana, B. Marcandalli and E. Selli, Dyes
41, 1809–1825. Pigm., 2001, 49, 127–133.
7 K. Uchida, S. Yamaguchi, H. Yamada, M. Akazawa, 20 Y. He, W. He, R. Wei, Z. Chen and X. Wang, Chem. Commun.,
T. Katayama, Y. Ishibashi and H. Miyasaka, Chem.
Commun., 2009, 4420–4422.
8 O. M. Tanchak and C. J. Barrett, Macromolecules, 2005, 38,
10566–10570.
9 A. Priimagi and A. Shevchenko, J. Polym. Sci., Part B: Polym.
Phys., 2014, 52, 163–182.
10 N. M. Ahmad, X. Lu and C. J. Barrett, J. Mater. Chem., 2010,
20, 244–247.
11 A. A. Beharry and G. A. Woolley, Chem. Soc. Rev., 2011, 40,
4422–4437.
12 D. W. Jenkins, K. El-Tahlawy, A. El-Shafei, H. S. Freeman and
S. M. Hudson, Color. Technol., 2006, 122, 345–349.
13 H. S. Freeman and S. A. McIntosh, Text. Res. J., 1989, 59, 343–
349.
2012, 48, 1036–1038.
21 J. Vapaavuori, A. Priimagi and M. Kaivola, J. Mater. Chem.,
2010, 20, 5260–5264.
22 D. Brown, A. Natansohn and P. Rochon, Macromolecules,
1995, 28, 6116–6123.
23 K. G. Yager and C. J. Barrett, J. Chem. Phys., 2004, 120, 1089–
1096.
24 B. Hudson and B. Kohler, Annu. Rev. Phys. Chem., 1974, 25,
437–460.
25 B. E. Kohler and I. D. W. Samuel, J. Chem. Phys., 1995, 103,
6248–6252.
26 H. Nishioka, X. Liang and H. Asanuma, Chem.–Eur. J., 2010,
16, 2054–2062.
27 R. N. Jones, Chem. Rev., 1947, 41, 353–371.
14 M. G. Finn and V. V. Fokin, Chem. Soc. Rev., 2010, 39, 1231– 28 X. Wang, J. Kumar, S. K. Tripathy, L. Li, J.-I. Chen and
1232. S. Marturunkakul, Macromolecules, 1997, 30, 219–225.
15 A. Goulet-Hanssens and C. J. Barrett, J. Photochem. 29 A. Priimagi, A. Shevchenko, M. Kaivola, F. J. Rodriguez,
Photobiol., A, 2014, DOI: 10.1016/j.jphotochem.2014.07.013.
16 H. A. Wegner, Angew. Chem., Int. Ed., 2012, 51, 4787–4788.
17 O. Sadovski, A. Beharry, F. Zhang and G. A. Woolley, Angew.
Chem., Int. Ed., 2009, 48, 1484–1486.
M. Kauranen and P. Rochon, Opt. Lett., 2010, 35, 1813–1815.
30 A. Priimagi, M. Kaivola, M. Virkki, F. J. Rodriguez and
M. Kauranen, J. Nonlinear Opt. Phys. Mater., 2010, 19, 57–73.
31 U. Pietsch and P. Rochon, J. Appl. Phys., 2003, 94, 963.
18 E. Merino, Chem. Soc. Rev., 2011, 40, 3835–3853.
7512 | J. Mater. Chem. C, 2014, 2, 7505–7512
This journal is © The Royal Society of Chemistry 2014