Impact of backbone rigidity on the photomechanical response of glassy, azobenzene-functionalized polyimides

Article abstract of DOI:10.1021/ma402178z

Azobenzene-functionalized polyimide materials can directly transduce light into mechanical force. Here, we examine the impact of polymer backbone rigidity on the photomechanical response in a series of linear, azobenzene-functionalized polymers. The rigidity of the backbone was varied by the polymerization of five dianhydride monomers with a newly synthesized diamine (azoBPA-diamine). The azobenzene-functionalized linear polymers exhibit glass transition temperatures (T_g) ranging from 276 to 307 C and maintain excellent thermal stability. The photomechanical response of these materials was characterized by photoinduced cantilever bending as well as direct measurement of photogenerated stress upon exposure to linearly polarized, 445 nm light. Increasing the rigidity of the polymer backbone increases the magnitude of stress that is generated but decreases the angle of cantilever deflection.

Full text of DOI:10.1021/ma402178z

Article
pubs.acs.org/Macromolecules
Impact of Backbone Rigidity on the Photomechanical Response of
Glassy, Azobenzene-Functionalized Polyimides
David H. Wang, Jeong Jae Wie, Kyung Min Lee, Timothy J. White,* and Loon-Seng Tan*
Air Force Research Laboratory, Materials & Manufacturing Directorate, Wright-Patterson Air Force Base, Ohio 45433-7750,
United States
S
* Supporting Information
ABSTRACT: Azobenzene-functionalized polyimide materials can
directly transduce light into mechanical force. Here, we examine the
impact of polymer backbone rigidity on the photomechanical
response in a series of linear, azobenzene-functionalized polymers.
The rigidity of the backbone was varied by the polymerization
of five dianhydride monomers with a newly synthesized diamine
(azoBPA-diamine). The azobenzene-functionalized linear polymers
exhibit glass transition temperatures (Tg) ranging from 276 to 307 °C
and maintain excellent thermal stability. The photomechanical response
of these materials was characterized by photoinduced cantilever bending
as well as direct measurement of photogenerated stress upon exposure
to linearly polarized, 445 nm light. Increasing the rigidity of the polymer
backbone increases the magnitude of stress that is generated but
decreases the angle of cantilever deflection.
The two order of magnitude improvement (compared to azo-
LCNs of similar azobenzene concentration) is attributable to the
high performance characteristics of polyimides. Our prior work has
primarily focused on the characterization of cross-linked
azobenzene-containing aromatic polyimides and investigated the
influence of structural and morphological factors on the
photomechanical response of these materials.5−8,17 Subsequent
to this initial report, we have reported on enhancement of
photogenerated stress by prestraining (hot drawing),7 polarization-
controlled flexural−torsional (bend and twist) deflections,8 and the
role of free volume on the resulting photomechanical response.17
We have also recently reported on the preparation and
characterization of linear polyimide materials containing main-
chain azobenzene units.5 Through systematic preparation of
(co)polyimides prepared from flexible (6FDA) and rigid (PMDA)
dianhydrides and a structurally rigid diamine (4,4′-diaminoazo-
benzene) we have shown that crystallinity decreases the magnitude
of bending and photogenerated stress.25
A number of recent reports have theoretically examined
photomechanical effects in both liquid crystalline26−31 and
amorphous32−34 polymers. Of relevance to the work presented
here, Toshchevikov et al.32−34 correlate descriptors of the local
macromolecular environment (such as the Kuhn segment and
chromophore number density) to the resulting photomechan-
ical output. The goal of this work is to experimentally elucidate
the contribution of the backbone rigidity of a series of
INTRODUCTION
■
Photomechanical effects in polymers have been pursued for
nearly 50 years.1 Photoinitiated shape adaptivity or force
generation (actuation) are particularly intriguing due to the salient
features of light: remote and wireless (contactless) triggering with
ease of spatial, temporal, directional (through polarization), and
magnitude (with intensity) control. Much of the recent literature
has focused on the preparation and characterization of
azobenzene-functionalized liquid crystalline polymer networks
(azo-LCNs). Azo-LCNs are distinguished in that the local director
orientation can be manipulated to generate monolithic, engineered
materials ultimately designed for a particular shape or surface
response.2−4 However, higher performance materials such as
polyimides may be advantageous, particularly in applications of the
materials in load-bearing or extreme environments.5−8
Polyimides represent an important class of heat-resistant
polymers useful in a variety of applications deriving from their
excellent combination of physical properties, thermal stability,
and processability.9−11 Azobenzene-functionalized polyimides
have been investigated for photoinduced alignment of liquid
crystal devices and nonlinear optical (NLO) materials.12−16
Recent examinations from our groups5−8,17 have extended
upon the work of Agolini and Gay18 in studying photo-
mechanical effects in azobenzene-functionalized polyimides
(azo-PIs). Photomechanical effects based on azobenzene and
other photochemical units and processes have also been subject
to considerable recent attention in polymeric19−21 as well as
crystalline materials.22−24
Received: October 22, 2013
Revised: December 30, 2013
Published: January 13, 2014
We have reported that up to 1.3 MPa of stress can be wirelessly
photogenerated in azobenzene-functionalized polyimide materials.
© 2014 American Chemical Society
659
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
carbonate (24.5 g,180 mmol), and DMF (300 mL). The mixture was
stirred at room temperature for 24 h and filtered. The filtrate was
diluted with ethyl acetate (1200 mL), and the organic layer was
separated. The organic layer was washed three times with water, then
dried with magnesium sulfate, and filtered. The filtrate was evaporated
to afford yellowish crystals, which recrystallized from methanol/ethyl
acetate to yield 18.33 g (61%) of off-white crystals; mp 120.0−121.2 °C.
IR (KBr, cm−1): 3446, 3072, 2971, 2877, 2451, 1904, 1610 (NO2), 1344
(NO2), 1249 (ether). NMR (DMSO-d6, δ in ppm) 1.70 (s, 6H, CH3),
7.11−7.14 (m, 8H, Ar−H), 7.35−7.38 (m, 4H, Ar−H), 8.23−8.27 (m,
4H, Ar−H). MS (m/z) 470 (M+). Anal. Calcd for C27H23N2O6: C,
68.80%; H, 4.89%; N, 6.01%. Found: C, 69.52%; H, 4.76%; N, 5.70%.
2,2-Bis[4-(4-aminophenoxy)phenyl]propane (4). 2,2-Bis[4-(4-
nitrophenoxy)phenyl]propane (3; 5.0 g, 10.6 mmol) dissolved in ethyl
acetate (100 mL) and palladium on activated carbon (0.30 g) was
placed in a hydrogenation bottle. The bottle was tightly secured on a
Parr hydrogenation apparatus, flushed four times with hydrogen, and
pressurized to 55 psi. After the mixture had been agitated at room
temperature for 24 h under the hydrogen pressure of 55 psi, it was
filtered through Celite. The filter cake was washed with ethyl acetate,
and then the filtrate was evaporated to dryness on a rotary evaporator
to afford 4.214 g (95%) of off-white crystals; mp 126.5−127.5 °C. IR
(KBr, cm−1): 3423, 3402, 3333, 3235 (amine), 3038, 2964, 2869, 1879,
amorphous linear polyimides to the photomechanical output.
Toward this end, we synthesized a new, bis(azobenzene)-
diamine monomer and used it to prepare five, amorphous
azobenzene-functionalized linear polyimides through polymer-
ization with structurally distinctive dianhydride monomers. The
thermal and optical properties of the materials were
characterized with dynamic mechanical analysis, differential
scanning calorimetry, thermogravimetric analysis, X-ray, and
UV−vis spectroscopy while the photomechanical response was
visualized with cantilever bending and measured in a tensile
experiment. Understanding the impact of the local molecular
environment on photomechanical output provides information
critical to enabling the continued optimization and develop-
ment of the materials chemistry to improve the effectiveness
and efficiency of these materials as energy transducers.
EXPERIMENTAL SECTION
■
Materials. 1,3-Bis(3-aminophenoxy)benzene (ABP) (99% min.),
pyromellitic dianhydride (PMDA), 3,3′,4,4′-biphenyltetracarboxylic
dianhydride (BPDA), 3,3′,4,4′-benzophenonetetracarboxylic dianhy-
dride (BTDA), and oxy-4,4′-di(phthalic anhydride) (ODPA) were
purchased from Chriskev Company, Inc. 1,1,1,3,3,3-Hexafluoro-2,2-
bis(4-phthalic anhydride)propane (6FDA) was purchased from Akron
Polymer Systems. All five dianhydrides were sublimed before use. All
other reagents and solvents were purchased from Aldrich Chemical
Inc. and used as received, unless otherwise noted.
1
1733, 1610, 1498, 1222 (ether), 872. H NMR (DMSO-d6, δ in ppm):
1.55 (s, 6H, CH3), 4.94 (s, 4H, NH2), 6.55−6.71(m, 4H, Ar−H), 6.71−
6.77 (m, 8H, Ar−H), 7.11−7.13 (m, 4H, Ar−H). Anal. Calcd for
C27H26N2O2: C, 79.00%, H, 6.38%, N, 6.82%, Found: C, 78.27%, H,
6.65%, N, 6.45%.
Instrumentation. The mechanical properties and photogenerated
stress of polyimide films (6 mm (L) × 2 mm (W) × 0.02 mm (T))
were measured by a strain-controlled dynamic mechanical analyzer
(DMA, TA Instruments RSA III) in tension. In a transient mode,
stress−strain curves were obtained with a 0.01 s−1 Hencky strain (or
true strain) rate, and tensile modulus was determined from logarithmic
modulus−strain plots by extrapolating the modulus plateau to zero
strain. The glass transition temperatures of the polyimides are reported
from the maximum tan δ (loss modulus/storage modulus) by a stress-
controlled DMA (TA Instruments DMA Q800) with a heating rate of
4 °C/min in a nitrogen atmosphere. Infrared (IR) spectra were
recorded on a Nicolet Nexus 470 spectrometer, and attenuated total
reflectance IR (ATR-IR) spectra were collected on a Bruker Alpha-R
spectrometer. Proton and carbon nuclear magnetic resonance (NMR)
spectra were measured at 300 MHz with a Bruker AVANCE 300
spectrometer. Thermogravimetric analysis (TGA) was conducted in
either nitrogen (N2) or air atmosphere at a heating rate of 10 °C/min
using a TA Hi-Res TGA 2950 thermogravimetric analyzer. Gas
chromatography/mass spectroscopy (GC/MS) was then performed
using a Varian 1200 Series. Melting points were obtained from Buchi
melting point apparatus B-545 with a heating rate of 2 °C/min. Wide-
angle X-ray experiments were carried out on a Statton box camera at
53 mm sample-to-image plate distances in transmission mode using
Cu Kα generated by a Rigaku Ultrax18 system.
Photomechanical effects in the polyimide materials were charac-
terized with irradiation of linearly polarized 445 nm light at room
temperature. All measurements were conducted with light polarization
parallel to long axis of the cantilevers (E||x). The cantilever dimensions
were 6 mm (L) × 0.1 mm (W) × 0.02 mm (T) for all the experiments
reported here. Photogenerated stress was measured upon illumination
with linearly polarized (E||x) blue light from a 445 nm light-emitting
diode (LED) to polyimide films held in DMA at minimal prestrain
(4 × 10−5%). Time-resolved evolution of both cantilever bending and
photogenerated stress were monitored with light exposure for 1 h.
UV−vis absorption spectra of spin-coated polyimide were captured
by a Cary 5000 UV−vis−NIR spectrometer. The time-dependent
absorbance of the polyimides was monitored for 1 h upon exposure to
linearly polarized (E||x) 445 nm light at 60 mW/cm2.
2,2-Bis{4-[4-(4-acetamidophenyldiazenyl)phenoxy]phenyl}-
propane (6). 2,2-Bis[4-(4-aminophenoxy)phenyl]propane (4; 0.821
g, 2.00 mmol), N-(4-nitrosophenyl)acetamide (5; 1.312 g, 8 mmol),
and acetic acid (40 mL) were charged into a 150 mL round-bottomed
flask equipped with a magnetic stir bar. The mixture was stirred at
room temperature for 48 h. The mixture at first turned into a greenish
solution, and then yellow particles started to precipitate out of the
solution. The mixture was diluted by deionized water (100 mL). Solids
were collected and washed with water (500 mL) followed by of
ethanol (200 mL) to remove most of the unreacted nitroso reagent.
The raw product was slurried in hot ethanol (50 mL) and filtered after
being cooled to room temperature twice to give 0.91 g (65%) of
yellow solids; mp 267.2−269.1 °C (dec.). IR (KBr, cm−1): 3320
(NHCO), 3062, 2964, 1674 (CO), 1594, 1503, 1489, 1238, 1115,
1
849. MS (m/e): 702 (M+). H NMR (d6-DMSO, δ in ppm): 1.67 (s,
6H, COCH3), 2.08 (s, 6H, CCH3) 7.05−7.13 (dd, 8H, Ar−H), 7.30−
7.32 (d, 4H, Ar−H), 7.77−7.88 (m, 12H, Ar−H), 10.27 (s, 2H,
NHCO). 13C NMR (d6-DMSO, δ in ppm): 24.12, 30.57, 41.84,
118.23, 119.08, 123.43, 124.36, 128.26, 142.10, 146.16, 147.37, 147.65,
153.32, 159.37, 168.72. Anal. Calcd for C43H38N6O4: C, 73.49%, H,
5.45%, N, 11.96%, Found: C, 73.33%, H, 5.37%, N, 11.70%.
2,2-Bis{4-[4-(4-aminophenyldiazenyl)phenoxy]phenyl}-
propane (7a). To a 500 mL round-bottomed flask with a stir bar and
a condenser, 2,2-bis{4-[4-(4-acetamidophenyldiazenyl)phenoxy]-
phenyl}propane (6; 5.00 g, 7.11 mmol), 6 M HCl (200 mL), and
95% ethanol (20 mL) were charged and heated to 105 °C for 48 h.
After it was allowed to cool to room temperature, water (600 mL) was
added. The resulting red solid was collected by filtration and washed
with 1 N sodium hydroxide solution, followed by deionized water
(300 mL). After being air-dried, the crude product was purified by
column chromatography (silica gel, ethyl acetate as eluent). The
solvent was removed by a rotary evaporator to afford 2.95 g (67%) of
orange red solid; mp 162.0−163.1 °C. IR (KBr, cm−1): 3467, 3381
(NH2), 3036, 2964, 1619, 1598, 1505, 1489, 1238, 1145, 834. MS (m/e):
618 (M+). 1H NMR (d6-DMSO, δ in ppm): 1.66 (s, 6H, CCH3), 6.03 (s,
4H, NH2), 6.64−6.66 (d, 4H, Ar−H), 7.01−7.03 (d, 4H, Ar−H), 7.07−
7.09 (d, 4H, Ar−H), 7.28−7.30 (d, 4H, Ar−H), 7.61−7.63 (d, 4H, Ar−
H), 7.75−7.77 (d, 4H, Ar−H).13C NMR (d6-DMSO, δ in ppm): 30.54,
41.73, 113.31, 118.41, 118.64, 123.45, 124.85, 128.12, 142.72, 145.77,
148.15, 152.49, 153.73, 157.87. Anal. Calcd for C39H34N6O2: C, 75.71%;
H, 5.54%; N, 13.58%. Found: C, 75.62%; H, 5.39%; N, 13.48%.
2,2-Bis[4-(4-nitrophenoxy)phenyl]propane (3). Into a 1 L
three-necked flask equipped with a magnetic stir bar and nitrogen inlet
and outlet were placed 2,2-bis(4-hydroxyphenyl)propane (1; 18.0 g,
79.0 mmol), 4-fluoronitrobenzene (2; 25.0 g, 180 mmol), potassium
660
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
Model Compound Synthesis (2,2-Bis{4-[4-(4-
phthalimidophenyldiazenyl)phenoxy]phenyl}propane, 7b).
Into a 50 mL three-necked flask equipped with a magnetic stir bar
and nitrogen inlet and outlet were placed 2,2-bis{4-[4-(4-
aminophenyldiazenyl)phenoxy]phenyl}propane (0.33 g, 0.53 mmol),
phthalic anhydride (0.24 g, 1.6 mmol), and acetic acid (10 mL). The
mixture was stirred under refluxing 14 h and allowed to cool to room
temperature. The precipitate was collected by filtration and dried in
oven to afford 0.96 g (81%) of orange powder; mp 247.7−249.7 °C.
IR (KBr, cm−1): 3063, 2964, 1788, 1720, 1592, 1493, 1386, 1256,
1114, 841, 715. MS (m/e): 878.29 (M+). Anal. Calcd for C55H38N6O6:
Scheme 1. Syntheses of Bis(azobenzeneamine) Monomer
(7a) and Model Compound (7b)
1
C, 75.16%; H, 4.36%; N, 9.56%. Found: C, %; H, %; N, %. H NMR
(d6-DMSO, δ in ppm): 1.69 (s, 6H, CCH3), 7.08−7.10 (d, 4H, Ar−H),
7.16−7.18 (d, 4H, Ar−H), 7.33−7.35 (d, 4H, Ar−H), 7.67−7.70 (d, 4H,
Ar−H), 7.91−8.02 (m, 16H, Ar−H).
Procedure for the Preparation of PMDA Polyimide (10a,
AzoBPA-PMDA). 2,2-Bis{4-[4-(4-aminophenyldiazenyl)phenoxy]-
phenyl}propane (7a; 0.3184 g, 0.5246 mmol) and DMAc (4 mL)
were added to a 25 mL three-necked flask equipped with a magnetic
stirrer, nitrogen inlet, and outlet and stirred under dry nitrogen at
room temperature for 30 min. PMDA (8a; 0.1122 g, 0.5246 mmol)
was then charged. The dark red solution was agitated at room
temperature for 24 h to afford a viscous poly(amic acid) solution. This
solution was diluted with DMAc (3 mL), poured into a glass dish,
followed by vacuum evaporation of DMAc at 50 °C, and heat-treated
at 100 °C/2 h, 150 °C/2 h, 175 °C/1 h, 200 °C/2 h, 250 °C/1 h, and
300 °C/1 h to form imidized polymers. The film thickness was
approximately 20−100 μm. FT-IR (KBr, cm−1): 3049 (Ar−H), 2965
(CH3), 1777, 1719 (imide), 1587, 1488, 1356, 1232, 1169, 1113, 1080,
1012, 821, 722, 546.
Procedure for the Preparation of BPDA Polyimide (10b,
AzoBPA-BPDA). The OPDA polyimide (10b) was prepared from
2,2-bis{4-[4-(4-aminophenyldiazenyl)phenoxy]phenyl}propane (7a;
0.2849 g, 0.461 mmol), OPDA (8b; 0.1355 g, 0.461 mmol), and
DMAc (5.0 mL) using the same procedures as 10a to form imidized
polymers. The film thickness was approximately 20−100 μm. FT-IR
(KBr, cm−1): 3053 (Ar−H), 2964 (CH3), 1775, 1714 (imide), 1588,
1489, 1417, 1359, 1233, 1077, 1012, 832, 736, 548.
Procedure for the Preparation of BTDA Polyimide (10c,
AzoBPA-BTDA). The BTDA polyimide (10c) was prepared from
2,2-bis{4-[4-(4-aminophenyldiazenyl)phenoxy]phenyl}propane (7a;
0.3583 g, 0.579 mmol), BTDA (8c; 0.1866 g, 0.679 mmol), and
DMAc (5.0 mL) using the same procedures as 10a to form imidized
polymers. The film thickness was approximately 20−100 μm. FT-IR
(KBr, cm−1): 3054 (Ar−H), 2964 (CH3), 1779, 1714 (imide), 1588,
1489, 1361, 1235, 1080, 1012, 831, 714, 549.
Procedure for the Preparation of 6FDA Polyimide (10d,
AzoBPA-6FDA). The 6FDA polyimide (10d) was prepared from 2,2-
bis{4-[4-(4-aminophenyldiazenyl)phenoxy]phenyl}propane (7a;
0.5087 g, 0.822 mmol), 6FDA (8d; 0.3652 g, 0.822 mmol), and
DMAc (5.5 mL) using the same procedures as 10a to form imidized
polymers. The film thickness was approximately 20−100 μm. FT-IR
(KBr, cm−1): 3052 (Ar−H), 2966 (CH3), 1785, 1722 (imide), 1589,
1489, 1411, 1232, 1207, 1191, 1138, 1080, 1012, 834, 720, 548.
Procedure for the Preparation of OPDA Polyimide (10e,
AzoBPA-OPDA). The OPDA polyimide (10e) was prepared from
2,2-bis{4-[4-(4-aminophenyldiazenyl)phenoxy]phenyl}propane (7a;
0.4770 g, 0.771 mmol), OPDA (8e; 0.2392 g, 0.771 mmol), and
DMAc (5.0 mL) using the same procedures as 10a to form imidized
polymers. The film thickness was approximately 20−100 μm. FT-IR
(KBr, cm−1): 3056 (Ar−H), 2964 (CH3), 1778, 1714 (imide), 1589,
1489, 1230, 1078, 1012, 834, 705, 549.
to photomechanical effects in polymeric materials. Toward this
end, a series of five azobenzene-functionalized polyimides were
prepared in which the rigidity of the imide unit varies widely as a
function of the dianhydride structure. The azo-PIs examined here
were prepared with a new diamine containing two azobenzenes
per molecule (7a, azoBPA-diamine). The four-step synthesis of 7a
is shown in Scheme 1. To prepare 7a, 2,2-bis(4-hydroxyphenyl)-
propane (i.e., bisphenol A, 1) was treated with 1-fluoro-4-
nitrobenzene (2) in the presence of potassium carbonate to yield
2,2-bis[4-(4-nitrophenoxy)phenyl]propane (3). 3 was then
reduced to 2,2-bis[4-(4-aminophenoxy)phenyl]propane (4) by
catalytic hydrogenation. The condensation reaction of 4 and
4-nitrosoacetanilide35 (5) in acetic acid yielded 6, a precursor
containing two azobenzene units. The azoBPA-diamine monomer
(7a) was generated after the deprotection (deacetylation) of 6 via
alkaline hydrolysis. To facilitate examination of photoisomerization
of the chromophore unit, the model compound diphthalimidobis-
(azobenzene) 7b was synthesized by end-capping 6 with phthalic
anhydride as shown in Scheme 1. The photoisomerization of 7b
was examined in solution (Figures S1 and S2 of the Supporting
Information). The two-stage synthesis of the azobenzene-
containing polyimides (azoBPA-XXDA, where XX stands for the
generic dianhydride with the specific chemical structure and
associated code as indicated in Scheme 2) was conducted as
follows: (i) a dianhydride (8) and azoBPA-diamine (7a) were
dissolved temperature for 24 h to generate poly(amic acid)
(PAA, 9); (ii) the resulting PAA precursor was poured onto glass
slides and cured in an oven set to 300 °C to imidize the polymer
RESULTS AND DISCUSSION
■
Our prior work has demonstrated that irradiation of linear and
cross-linked azobenzene-functionalized polyimides (azo-PIs) can
generate considerable photogenerated stress (relative to azoben-
zene-functionalized liquid crystal networks, azo-LCN).5−8,17 The
goal of this work is to isolate the contribution of backbone rigidity
661
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
Scheme 2. Synthesis of Linear Azobenzene-Containing Polyimides 10a−e
films. The formulation, density, Tg, Td, and absorption coefficient
of the materials are summarized in Table 1.
The thermomechanical properties of the azobenzene-
functionalized polyimides (azo-PIs) were characterized. The
glass transition temperatures (Tg) of the materials were
measured by DMA from the peak of the tan δ and range
from 276 to 307 °C (Table 1). The tensile properties of
polyimide films were characterized by DMA with 0.01 s−1
Hencky strain (or true strain). The raw data are shown in
Figure 1, and the values of the tensile modulus, yield stress,
ultimate tensile strength, strain to failure, and toughness for the
five polymers are summarized in Table 2. In the elastic regime,
azoBPA-PMDA exhibited the largest tensile modulus (2.51 GPa)
and azoBPA-OPDA had the lowest tensile modulus (1.34 GPa)
(see Figure S5 for linear regime in stress−strain curves). AzoBPA-
PMDA and azoBPA-OPDA had the largest and the lowest yield
stress, respectively. After the yield point, significant strain
hardening was observed only from azoBPA-PMDA. Generally,
polymers showing larger strain hardening have ductile deformation
and larger toughness due to strain delocalization.36 As expected,
azoBPA-PMDA had a significantly greater toughness (or energy to
break) as well as extensibility (rigid−ductile) compared to other
polyimides. AzoBPA-OPDA exhibits comparable toughness to
azoBPA-PMDA due to the significant elongation after yield as
evident in Figure 2. The large value of the strain to failure of
azoBPA-OPDA is attributed to chain flexibility promoted by the
rotational freedom of the ether linkage in the oxydiphthalic imide
(OPDI) unit.12
Figure 1. Representative stress−strain responses of polyimide films in
uniaxial tension: (i) azoBPA-PMDA, (ii) azoBPA-BPDA, (iii) azoBPA-
BTDA, (iv) azoBPA-6FDA, and (v) azoBPA-OPDA.
Classically, the conformational rigidity of a polymer can be
described by the Kuhn length or persistence length.37−40
Experimental examinations41−43 have correlated these theoreti-
cal descriptions to measurable physical properties such as glass
transition temperature42,44 or modulus.45 At the molecular
level, the conformational rigidity of a single chain can be rea-
sonably assumed to be dictated by the rigidity of the repeating
unit. Thus, in the polyimides presented here the molecular
rigidity of the repeat unit is varied by the employment of five
662
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
sp3-swivel moiety between the phenylene rings, and the
⟩C(CF3)2 moiety in 6FDA is expected to be more rigid than
the oxygen linkage in OPDA. Thus, based on this qualitative
correlation that is supported by prior literature46 as well as the
thermomechanical properties reported in Figure 1, Table 2, and
the Supporting Informationthe molecular rigidity of the
polyimides are expected to rank as PMDA > BPDA > BTDA >
6FDA > OPDA.
The morphology of the materials was characterized with
wide-angle X-ray diffraction (WAXD). As evident in Figure 2,
azoBPA-OPDA, azoBPA-6FDA, and azoBPA-BTDA were
completely amorphous evident in the featureless diffraction
patterns of these materials. The diffraction patterns of azoBPA-
PMDA and azoBPA-BPDA revealed very slight crystalline
characteristics with calculated crystallinity indices of 1.5% and
0.4%, respectively. The influence of the very low crystalline
content on the photomechanical response of these materials is
expected to be negligible.5
The azobenzene chromophores in the polyimide materials
examined here absorb 445 nm light, and a small portion of
these chromophores reorient in the direction perpendicular to
light polarization through trans−cis−trans reorientation mech-
anism.47 Because of the large absorption coefficient of the
materials (summarized in Table 1), light is absorbed
nonuniformly across the sample thickness. The absorption
gradient is then mirrored by a strain gradient which causes
deflection of the film in the cantilever geometry. The
magnitude of the deflections can be quantified by reporting
Figure 2. Wide-angle X-ray diffraction patterns of the materials examined
here: (i) azoBPA-PMDA, (ii) azoBPA-BPDA, (iii) azoBPA-BTDA,
(iv) azoBPA-6FDA, and (v) azoBPA-OPDA.
anhydrides of varying composition and rigidity. In the case of
PMDA, BPDA, and BTDA, all of which are composed of
skeletal sp2-carbons, the rigidity can be qualitatively assessed
based on the number of single C−C bonds between the two
phenylene rings. Accordingly, the rigidity of the repeat unit
should vary as PMDA > BPDA > BTDA. In the case of the
polyimides based on 6FDA and OPDA, the rigidity will be
strongly influenced by the steric limitations imparted by the
Figure 3. Photoinduced bending of cantilevers composed of (i) azoBPA-PMDA, (ii) azoBPA-BPDA, (iii) azoBPA-BTDA, (iv) azoBPA-6FDA, and
(v) azoBPA-OPDA upon exposure to linearly polarized 445 nm light (E||x) at 120 mW/cm2 for 1 h. The retention or relaxation of the materials is
apparent in the images of the cantilevers collected 10 days after irradiation (i′) azoBPA-PMDA, (ii′) azoBPA-BPDA, (iii′) azoBPA-BTDA, (iv′)
azoBPA-6FDA, and (v′) azoBPA-OPDA.
663
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
Table 1. Composition, Thermal, and Optical Properties of Polyimide Films
c
density
α445 nm
a
b
b
sample
dianhydride (mol %) 7a (mol %) APB (mol %)
(g/cm3)
Tg (°C) Td5% (°C) in air Td5% (°C) in nitrogen
(cm−1
)
azoBPA-PMDA
azoBPA-BPDA
azoBPA-BTDA
azoBPA-6FDA
azoBPA-OPDA
100
100
100
100
100
100
100
100
100
100
0
0
0
0
0
1.363
1.355
1.347
1.339
1.304
307
302
294
284
276
451
458
456
474
451
424
427
422
421
420
331
525
570
723
801
a
b
Tg measured from the peak of tan δ (DMA) as an average value taken from four measurements. Temperature at which 5% weight loss as recorded
c
on TGA thermogram with a heating rate of 10 °C/min (raw data plotted in Figures S3 and S4, Supporting Information). Measured absorption
coefficient for the polyimide films at 445 nm.
Table 2. Tensile Properties of Polyimide Films
sample
tensile modulus (GPa)
yield stress (MPa)
ultimate tensile strength (MPa)
strain to failure (%)
toughness (MJ/m3)
azoBPA-PMDA
azoBPA-BPDA
azoBPA-BTDA
azoBPA-6FDA
azoBPA-OPDA
2.51 0.16
1.99 0.42
1.66 0.14
1.37 0.17
1.34 0.10
99.1 8.3
86.5 6.6
92.6 12.3
75.0 4.7
63.7 3.0
147.0 11.0
95.5 5.6
94.1 10.3
75.1 4.7
81.7 1.2
147.5 17.9
127.5 18.2
82.4 20.9
23.3 4.1
213.7 57.7
121.5 10.6
63.5 12.1
19.0 4.1
257.6 5.8
228.4 3.5
the bending angle between the mounting and tip points of the
cantilever. In Figure 3, photoinduced bending (i−v) and the
retention/relaxation (i′−v′) of the polyimide films were moni-
tored in the cantilever geometry upon exposure to 445 nm
light linearly polarized parallel to the long axis of the cantilever
(E||x) at 120 mW/cm2 for 1 h as well as after storage of the
materials in the dark (10 days), respectively. Polyimide films
prepared with PMDA and BPDA (tensile modulus ≥2 GPa)
had smaller magnitude of bending than the other polymers.
The response of the polymers after dark storage was also
dependent on the local molecular rigidity. Partial optically
fixable shape memory48 was observed in the azo-PI materials
with more rigid backbone chemistries while the shape
restoration in the dark was observed in azo-PI with more
flexible backbones chemistries.
Photoinduced bending and unbending (relaxation) were
further investigated in the materials with the most rigid
backbone (azoBPA-PMDA) and contrasted to the response
observed in the material with the most flexible backbone
(azoBPA-OPDA). As summarized in Figure 4a, the magnitude
of the photomechanical response (derived from cantilever ben-
ding) was measured both during and immediately after irra-
diation. Both polyimides reached near-equilibrium bending
angle within 30 min. The bending angle was then monitored
over the course of 10 days after light was removed. As evident in
Figure 4a, the photogenerated strain in azoBPA-OPDA relaxed,
and the sample fully recovered to the original vertical position.
Interestingly, after the irradiation of the azoBPA-PMDA cantilever
was taken away, a slight reduction in bending was observed, but a
nonzero bending angle was retained over the course of 10 days.
The photomechanical responses of azoBPA-PMDA and
azoBPA-OPDA were measured by DMA (Figure 4b) with
concomitant irradiation of linearly polarized (E||x) blue light.
Irradiation of the materials in these conditions generated
contractile stress, which was measured as positive stress by
DMA in tension. The photogenerated stress measured by the
DMA also reached equilibrium in 30 min for both polyimides.
Irradiation of azoBPA-PMDA generated larger magnitude of
stress. After the removal of light, ca. 250 kPa of stress is
retained within azoBPA-PMDA while full stress relaxation was
observed within azoBPA-OPDA.49 We attribute the negative
◇
Figure 4. Temporal response of azoBPA-PMDA ( ) and azoBPA-
○
OPDA ( ) in (a) cantilever bending and (b) tension upon exposure
to linearly polarized (E||x) 445 nm light.
stress sign after the relaxation to the convolution of prestrain,
photoinduced creep, and photosoftening.50,51
664
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
Figure 5. Absorption spectra of (a) azoBPA-PMDA and (b) azoBPA-OPDA upon the irradiation of linear polarized (E||x) 445 nm light exposure at
2
◇
○
60 mW/cm for the duration of 1 h. (c) Normalized absorption peak intensity of azoBPA-PMDA ( ) and azoBPA-OPDA ( ) as a function of time.
The photochemical responses of azobenzene materials are
known to be strongly impacted by the local molecular
environment. In polymers, the isomerization and/or reorienta-
tion of the azobenzene chromophores requires sufficient free
volume. Density measurements, summarized in Table 1,
indicate that azoBPA-PMDA has a higher density (and likely
reduced free volume) than azoBPA-OPDA. Less clear in the
prior literature is the influence of the rigidity of the polymer
backbone on the extent of azobenzene isomerization. To isolate
the impact of backbone rigidity on the photochemical response
of the materials examined here, absorption spectra were
collected for azoBPA-PMDA (most rigid backbone chemistry)
and azoBPA-OPDA (least rigid backbone chemistry) during
irradiation. The absorbance spectra (normalized to account for
slight differences in sample thickness) were collected from thin,
spin-cast polyimide films (on glass substrates) before and
during 60 min of exposure to linearly polarized (E||x) 445 nm
light at 60 mW/cm2. As illustrated in the data in Figure 5a−c,
the less rigid and dense materials (azoBPA-OPDA) exhibits a
considerable increase in the photochemical response, as
indicated by the reduction in absorbance of the material at
365 nm upon irradiation with 445 nm light.
summarized in Figure 6 for both cantilever bending and tensile
experiments as a function of modulus. The photomechanical
response when visualized in cantilever bending experiments
decreases as modulus increased. The opposite trend is observed
Figure 6. Dependence of the photomechanical response measured as
cantilever bending (open symbols) and photogenerated stress (filled
symbols).
As discussed, the photomechanical effects in polymeric
materials are interesting from the perspective of applications
relating to shape change or actuation. The goal of this study
was to elucidate the contribution of polymer backbone rigidity
to photomechanical effects in azobenzene-functionalized
polyimides examined here.52 As discussed earlier, the rigidity
of the backbone can be strongly correlated to modulus. Thus,
the photomechanical response of the materials examined here is
in the tensile experiment. These two experiments are measuring
sticks for the potential utility of these materialsshape change
(cantilever bending) or actuation (tensile experiment). From
Figure 6 it is clear that in applications relating to shape change
665
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
more flexible backbone chemistries (lower modulus) allow the
identical input energy stimulus to be transduced into a larger
deflection. However, in the employment of these materials in
applications in actuation (leveraging appropriate mechanical
design), the more rigid (higher modulus) materials are able to
more effectively and efficiently translate the photogenerated
stress to potentially generate much larger force. Thus, future
implementations of these materials as energy transducers in
mechanical designs should select the appropriate backbone
chemistry to facilitate the optimal output desired.
REFERENCES
■
(1) Lovrien, R. Proc. Natl. Acad. Sci. U. S. A. 1967, 57 (2), 236−42.
(2) McConney, M. E.; Martinez, A.; Tondiglia, V. P.; Lee, K. M.;
Langley, D.; Smalyukh, I. I.; White, T. J. Adv. Mater. 2013, in press.
́
(3) de Haan, L. T.; Sanchez-Somolinos, C.; Bastiaansen, C. M. W.;
Schenning, A. P. H. J.; Broer, D. J. Angew. Chem., Int. Ed. 2012, 51
(50), 12469−12472.
(4) Lee, K. M.; Bunning, T. J.; White, T. J. Adv. Mater. 2012, 24 (21),
2839−2843.
(5) Lee, K. M.; Wang, D. H.; Koerner, H.; Vaia, R. A.; Tan, L. S.;
White, T. J. Angew. Chem., Int. Ed. 2012, 51 (17), 4117−4121.
(6) Wang, D. H.; Lee, K. M.; Yu, Z. N.; Koerner, H.; Vaia, R. A.;
White, T. J.; Tan, L. S. Macromolecules 2011, 44 (10), 3840−3846.
(7) Lee, K. M.; Wang, D. H.; Koerner, H.; Vaia, R. A.; Tan, L. S.;
White, T. J. Macromol. Chem. Phys. 2013, 214 (11), 1189−1194.
(8) Wang, D. H.; Lee, K. M.; Koerner, H.; Yu, Z. N.; Vaia, R. A.;
White, T. J.; Tan, L. S. Macromol. Mater. Eng. 2012, 297 (12), 1167−
1174.
CONCLUSIONS
■
A new diamine (azoBPA-diamine) was synthesized and
subsequently polymerized with five selected dianhydrides to
prepare a series of heat-resistant (high degradation temperature
and Tg ranging from 276 to 307 °C) and photoresponsive
linear polyimides. Morphology characterization indicated that
all the polyimides were amorphous or contained only a small
amount of crystallinity which had negligible impact in
photoinduced deformation. The rigidity of the polymer
backbone (controlled by the dianhydride subunit) directly
affected the mechanical properties and photomechanical
response both during and after irradiation with 445 nm light.
Irradiation of polyimides prepared with more rigid backbone
chemistry was shown to generate comparatively small
deflections in cantilever experiments and comparatively large
photogenerated stresses in tensile experiments. Conversely,
irradiation of polyimides with more flexible backbone
chemistries was shown to generate comparatively large
deflections in cantilever experiments and comparatively
reduced photogenerated stresses in tensile experiments.
Interestingly, shape memory (stress retention) was observed
in polyimides with more rigid backbone chemistries while shape
recovery (stress relaxation) was observed in the polyimides
with more flexible backbone chemistries. These results attest to
the importance of molecular engineering the polymer structure
to tailor a material for specific and optimized photomechanical
outputs.
(9) Wilson, D. S.; Stenzenberger, H. D.; Hergenrother, P. M.
Polyimides. Chapman and Hall: New York, 1990, 1−78.
(10) Abadie, M. J. M.; Sillion, B. Polyimides and Other High-
Temperature Polymers: Elsevier: Amsterdam, 1991; pp 1−18.
(11) Sroog, C. E. Prog. Polym. Sci. 1991, 16 (4), 561−694.
(12) Usami, K.; Sakamoto, K.; Tamura, N.; Sugimura, A. Thin Solid
Films 2009, 518 (2), 729−734.
(13) Park, B.; Jung, Y.; Choi, H. H.; Hwang, H. K.; Kim, Y.; Lee, S.;
Jang, S. H.; Kakimoto, M.; Takezoe, H. Jpn. J. Appl. Phys., Part 1: Regul.
Pap. Short Notes Rev. Pap. 1998, 37 (10), 5663−5668.
(14) Chigrinov, V. G. K., Kozenkov, V. M.; Kwok, H. S.
Photoalignment of Liquid Crystalline Materials: Physics and Applications
2008, John Wiley & Sons, Chichester, UK.
(15) Koshiba, Y.; Yamamoto, M.; Kinashi, K.; Misaki, M.; Ishida, K.;
Oguchi, Y.; Ueda, Y. Thin Solid Films 2009, 518 (2), 805−809.
(16) Delaire, J. A.; Nakatani, K. Chem. Rev. 2000, 100 (5), 1817−
1845.
(17) Lee, K. M.; Koerner, H.; Wang, D. H.; Tan, L. S.; White, T. J.;
Vaia, R. A. Macromolecules 2012, 45 (18), 7527−7534.
(18) Agolini, F.; Gay, F. P. Macromolecules 1970, 3 (3), 349−351.
(19) Jin, Y. H.; Paris, S. I. M.; Rack, J. J. Adv. Mater. 2011, 23 (37),
4312−4317.
̌
(20) Ugur, G.; Chang, J.; Xiang, S.; Lin, L.; Lu, J. Adv. Mater. 2012,
24 (20), 2685−2690.
(21) Kloxin, C. J.; Scott, T. F.; Park, H. Y.; Bowman, C. N. Adv.
Mater. 2011, 23 (17), 1977−1981.
ASSOCIATED CONTENT
■
S
* Supporting Information
(22) Terao, F.; Morimoto, M.; Irie, M. Angew. Chem., Int. Ed. 2012,
51 (4), 901−904.
Figures S1−S5. This material is available free of charge via the
(23) Naumov, P.; Kowalik, J.; Solntsev, K. M.; Baldridge, A.; Moon, J.
S.; Kranz, C.; Tolbert, L. M. J. Am. Chem. Soc. 2010, 132 (16), 5845−
5857.
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
*E-mail Timothy.White.24@us.af.mil, Tel 937-255-9551
(T.J.W.).
■
(24) Bushuyev, O. S.; Singleton, T. A.; Barrett, C. J. Adv. Mater.
2013, 25 (12), 1796−1800.
(25) White, T. J.; Tabiryan, N. V.; Serak, S. V.; Hrozhyk, U. A.;
Tondiglia, V. P.; Koerner, H.; Vaia, R. A.; Bunning, T. J. Soft Matter
2008, 4 (9), 1796−1798.
*E-mail Loon.Tan@us.af.mil, Tel 937-255-9153 (L.-S.T.).
Author Contributions
D.H.W. and J.J.W. contributed equally to this work.
Notes
The authors declare no competing financial interest.
D.H.W.: Also with UES Inc.
J.J.W. and K.M.L.: Also with Azimuth Inc.
(26) Warner, M.; Mahadevan, L. Phys. Rev. Lett. 2004, 92 (13),
134302/1−134302/4.
(27) Corbett, D.; Warner, M. Phys. Rev. Lett. 2007, 99 (17), 174302/
1−174302/4.
(28) Corbett, D.; Warner, M. Phys. Rev. E: Stat., Nonlinear, Soft
Matter Phys. 2008, 77 (5−1), 051710/1−051710/11.
(29) Corbett, D.; Warner, M. Phys. Rev. E: Stat., Nonlinear, Soft
Matter Phys. 2008, 78 (6−1), 061701/1−061701/13.
(30) van Oosten, C. L.; Corbett, D.; Davies, D.; Warner, M.;
Bastiaansen, C. W. M.; Broer, D. J. Macromolecules 2008, 41 (22),
8592−8596.
ACKNOWLEDGMENTS
■
This work was completed at the Air Force Research Laboratory
(AFRL) with funding provided by the Materials and Manu-
facturing Directorate as well as Air Force Office of Scientific
Research. We are grateful to Marlene Houtz (University of
Dayton Research Institute) for TGA and DMA data.
(31) Warner, M.; Modes, C. D.; Corbett, D. Proc. R. Soc. London, A
2010, 466 (2122), 2975−2989.
(32) Toshchevikov, V. P.; Saphiannikova, M.; Heinrich, G. J. Chem.
Phys. 2012, 137 (2), 024903.
666
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667

Macromolecules
Article
(33) Toshchevikov, V.; Saphiannikova, M.; Heinrich, G. J. Phys.
Chem. B 2011, 116 (3), 913−924.
(34) Toshchevikov, V.; Saphiannikova, M.; Heinrich, G. J. Phys.
Chem. B 2009, 113 (15), 5032−5045.
(35) Cain, J. C. J. Chem. Soc., Trans. 1908, 93, 681−684.
(36) Hoy, R. S.; Robbins, M. O. J. Polym. Sci., Part B: Polym. Phys.
2006, 44 (24), 3487−3500.
(37) Erman, B.; Flory, P. J.; Hummel, J. P. Macromolecules 1980, 13
(3), 484−91.
(38) Sperling, L. H Introduction to Physical Polymer Science, 4th ed.,
Wiley-Interscience, New York, 2006; p 212−213.
(39) Rubinstein, M.; Colby, R. H. Polymer Physics; Oxford University
Press: Oxford, 2003; p 454.
(40) Doi, M.; Edwards, S. F. The Theory of Polymer Dynamics;
Clarendon: Oxford, 1986; p 11.
(41) Bacosca, I.; Hamciuc, E.; Bruma, M.; Ronova, I. A. J. Iranian
Chem. Soc. 2012, 9 (6), 901−910.
(42) Hamciuc, E.; Hamciuc, C.; Ronova, I. A. Rev. Roum. Chim. 2012,
57 (4−5), 407−413.
(43) Abadie, M. J. M. High Performance Polymers - Polyimides Based -
From Chemistry to Applications; InTech: Rijeka, Croatia, 2012; p 7.
(44) Rusu, R. D.; Damaceanu, M. D.; Bruma, M.; Ronova, I. A. Iran.
Polym. J. 2011, 20 (1), 29−40.
(45) Gray, E. W.; Winsor, P. A. Liquid Crystals and Plastic Crystals;
Ellis Horwood: Chichester, 1974; Vol. 2, pp 231−235.
(46) Ree, M.; Kim, K.; Woo, S. H.; Chang, H. J. J. Appl. Phys. 1997,
81 (2), 698−708.
(47) Lee, K. M.; Tabiryan, N. V.; Bunning, T. J.; White, T. J. J. Mater.
Chem. 2012, 22 (2), 691−698.
(48) Mol, G. N.; Harris, K. D.; Bastiaansen, C. W. M.; Broer, D. J.
Adv. Funct. Mater. 2005, 15 (7), 1155−1159.
(49) Notably, after the light had been turned off, an artificial jump in
measured stress was observed. This perturbation may be related to the
utilization of a dynamic mode to maintain the film at a given prestrain
in the tensile experiment. Upon removal of the light, the material is no
longer subjecting the tensile grips to force, and accordingly the
instrument adjusts to this new condition.
(50) Verploegen, E.; Soulages, J.; Kozberg, M.; Zhang, T.; McKinley,
G.; Hammond, P. Angew. Chem., Int. Ed. 2009, 48 (19), 3494−3498.
(51) Petr, M.; Helgeson, M. E.; Soulages, J.; McKinley, G. H.;
Hammond, P. T. Polymer 2013, 54 (12), 2850−2856.
(52) Because of differences in the molecular weight of the
dianhydride dianhydride monomers, the molar concentration of the
azobenzene diamine was varied and the corresponding increase in
absorption coefficent (Table 1) is unavoidable in this study. However,
based on prior reports (refs 31−33), it is expected that if the number
density of azobenzene chromophores (e.g., absorption coefficient)
took precedence over the contribution of the backbone rigidity
photogenerated stress measurements would decrease as a function of
modulus.
667
dx.doi.org/10.1021/ma402178z | Macromolecules 2014, 47, 659−667