Biobased polyimides from 4-aminocinnamic acid photodimer

Article abstract of DOI:10.1021/ma402499m

The development of high-performance biobased polymers such as polyimides (PIs) is indispensable to establish a sustainable green society, but it is very difficult due to the incompatibility of their monomeric aromatic diamines with microorganisms. Here, we developed biobased PIs from bioavailable aromatic diamines, which were photodimers of 4-aminocinnamic acid (4ACA) derived from genetically manipulated Escherichia coli. These biobased PI films showed ultrahigh thermal resistance with T₁₀ values over 425 °C and no T_g values under 350 °C, which is the highest value of all biobased plastics reported thus far. The PI films also showed high tensile strength, high Young's moduli, good cell compatibility, excellent transparency, and high refractive indices.

Full text of DOI:10.1021/ma402499m

Article
pubs.acs.org/Macromolecules
Biobased Polyimides from 4‑Aminocinnamic Acid Photodimer
Phruetchika Suvannasara,†,§ Seiji Tateyama,† Akio Miyasato,† Kazuaki Matsumura,† Tatsuya Shimoda,†
Takashi Ito,⊥ Yukiho Yamagata,⊥ Tomoya Fujita,⊥ Naoki Takaya,⊥ and Tatsuo Kaneko*,†
†School of Materials Science, Japan Advanced Institute of Science and Technology (JAIST), 1-1 Asahidai, Nomi, Ishikawa 923-1292,
Japan
⊥Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan
§Program of Petrochemistry and Polymer Science, Faculty of Science, Chulalongkorn University, Patumwan, Bangkok, 10330
Thailand
S
* Supporting Information
ABSTRACT: The development of high-performance biobased polymers such as
polyimides (PIs) is indispensable to establish a sustainable green society, but it is
very difficult due to the incompatibility of their monomeric aromatic diamines with
microorganisms. Here, we developed biobased PIs from bioavailable aromatic
diamines, which were photodimers of 4-aminocinnamic acid (4ACA) derived from
genetically manipulated Escherichia coli. These biobased PI films showed ultrahigh
thermal resistance with T10 values over 425 °C and no Tg values under 350 °C,
which is the highest value of all biobased plastics reported thus far. The PI films
also showed high tensile strength, high Young’s moduli, good cell compatibility,
excellent transparency, and high refractive indices.
monoamines have rarely been produced by microorgan-
isms.13−15
INTRODUCTION
■
The development of biobased polymers is indispensable for the
establishment of a sustainable low-carbon society. A number of
aliphatic biobased polymers such as polyesters; poly(lactic
acid),1 poly(hydroxyalkanote)s,2 and poly(butylenes succi-
nate),3 and polyamides (polyamide 11 and polyamide 66)
have been developed, but their low glass transition temperature,
Tg, and only a small percentage of their substitutes limited their
use for various applications as superengineering plastic. The
introduction of an aromatic component into the thermoplastic
polymer backbone is an efficient method to intrinsically
improve the material performance.4,5 Although we developed
aromatic biobased polyesters from hydroxycinnamate deriva-
tives such as p-coumaric acid (4-hydroxycinnamic acid,
4HCA),6,7 caffeic acid, (3,4-dihydroxycinnamic acid, DHCA),8
and ferulic acid (3-methoxy-4-hydroxycinnamic acid, MHCA),9
their Tg ranged to 169 °C at most.10 In order to apply them to
the advanced fields of electronics, transportation vehicles, or
aviation materials, a very high value of Tg over 200 °C in neat
polymer materials without any additives is required. From this
point of view, a high mechanical strength and Young’s modulus
are also necessary. One of the most advanced classes of high-
performance polymeric materials is the polyimides (PIs), with
excellent thermo-mechanical performance, high chemical
stability, low coefficient of thermal expansion, etc.11,12 However
bioderived PIs were very difficult to prepare since the aromatic
diamines cannot be made using biosynthesis because they are
incompatible with microorganisms and plant cells, presumably
due to the combined interactions of ionic, hydrophobic, and π-
electron-related factors with cell constituents. Even aromatic
Here, we used the photodimer of a microorganism-derived
aromatic monoamine, 4-aminocinnamic acid (4ACA), to
prepare bioderived PIs based on its reaction with various
tetracarboxylic dianhydrides, and obtained high performance
PIs with high thermomechanical properties, high transparency,
and good cell-compatibility.
EXPERIMENTAL SECTION
■
Materials. Dianhydrides such as pyromellitic dianhydride (PMDA:
from TCI), 1,2,3,4-tetracarboxycyclobutane dianhydride (CBDA: from
Aldrich), 3,3′,4,4′-benzophenone tetracarboxylic dianhydride (BTDA:
from Aldrich), 4,4′-oxidiphthalic anhydride (OPDA: from Aldrich),
3,4,3′,4′-biphenyltetracarboxylic dianhydride (BPDA: from Aldrich),
and 3,3′,4,4′-diphenylsulfone tetracarboxylic dianhydride (DSDA:
from Aldrich) were recrystallized in acetic anhydride by refluxing for
5 h and then cooling to 0−5 °C. The crystals were carefully collected
by filtration, washed in hot dioxane, and dried in vacuo. 4-
aminocinnamic acid (4ACA: from TCI), trimethylsilyl chloride
(TMSCl: from Aldrich), and N,N-dimethylacetamide (DMAc: 99.8%
anhydrous from Kanto chemical) were used as received.
Production of 4ACA. The Rhizobium etli gxrA gene (accession
number, ACO35311.1) was amplified by PCR using the primers (5′-
CCGGATCCATGTCAGTTCGTCCTCCCGTCC-3′ and 5′-
GCGAATTCCTAATAACCGGCGGCGCGATCG-3′), digested
with BamHI and EcoRI, and then cloned into BamHI + EcoRI-
digested pCWfoxy vectors.16 The 2.5-kb DNA fragments containing
Received: December 6, 2013
Revised: January 15, 2014
Published: February 18, 2014
© 2014 American Chemical Society
1586
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
lacI and gxrA were amplified using the primers (5′-GCGAATTC-
CAGTCGGGAAACCTGTCGTGCC-3′ and 5′-CCGTATGCTAA-
TAACCGGCGGCGCGATCG-3′), and the resulting plasmid was
digested with EcoRI and SphI and cloned into EcoRI + SphI-digested
pHSG298 (Takara Bio, Kyoto, Japan) to generate pHSGgxrA. The
AtPal4 cDNA (NP_187645.1) was amplified using the primers (5′-
CCGGATCCATGGAGCTATGCAATCAAAACAATC-3′ and 5′-
CCGCATGCTCAACAGATTGAAACCGGAGCTCCG-3′), digested
with BamHI and SphI, and then cloned into BamHI + SphI-digested
pHSGgxrA to give rise to pHSG-Atpal4. Escherichia coli NST3717
transformed with pHSG-Atpal4 was cultured in 100 mL of Luria broth
(10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl) containing 50 mg/
L kanamycin sulfate at 37 °C and 120 rpm. After the optical density of
the culture reached 3, 1 mM IPTG was added and incubated for
another 12 h under the same conditions. The cells were collected by
centrifugation, washed with 0.1 M potassium phosphate (pH 8.0), and
incubated in the same buffer containing 10 mM 4-APhe at 37 °C and
120 rpm for 12 h. The reactions were analyzed by HPLC (HP-1100,
Hewlett-Packard, CA, USA) using a packed silica gel column
(Purospher star RP-18e 5 μm, 4.6 × 150 mm, Merck, Germany).
Methanol:20 mM phosphate (6:4 v/v) was used as the eluant at a flow
rate of 0.8 mL/min. Phenylalanine ammonia lyase activity was
measured as described previously.18
Poly(amic acid) and PI Syntheses. A typical polymerization
procedure for poly(amic acid) (PAA) is shown below. A diamine of
4ATA dimethyl ester, (0.20 g, 0.5647 mmol) was dissolved in DMAc
(0.56 mL, 0.6 M) in a 10 mL test tube under a nitrogen atmosphere.
Dianhydrides such as CBDA (0.11 g, 0.5647 mmol), PMDA (0.11 g,
0.5235 mmol), BTDA (0.18 g, 0.5650 mmol), OPDA (0.18 g, 0.5648
mmol), BPDA (0.17 g, 0.5642 mmol), and DSDA (0.20 g, 0.5660
mmol) were then added into the diamine solution. The reaction
mixture was vigorously stirred by an overhead stirrer at room
temperature to produce a pale yellow solution, and was further stirred
for 48 h to yield a viscous PAA solution. The PAA solution was diluted
in DMAc, and then added dropwise into ethanol to precipitate PAA
fibrils, which were collected by filtration, thoroughly washed with
water and dried in a vacuum oven for 12 h. The PAA film was obtained
by casting a DMAc yellow solution onto a silicon wafer and heating at
60−70 °C. The PAA specimen for the UV−vis measurements was
prepared by the spin-casting method (spin rate = 1000 rpm, MS-A100
Spincoater, Misaka Co., Ltd.) to adjust the thickness to 9−15 μm.
The PI films were obtained by a thermal imidization of the PAAs in
an oven under vacuum by stepwise heating at 100, 150, 200, and 250
°C for 1 h at each step
Measurements. NMR measurements such as 1H NMR, 13C NMR,
1
1H−13C HSQC, and H−13C HMBC were performed by a Bruker
Monomer Syntheses: Synthesis of 4,4′-Diamino-α-truxillic
Acid Dihydrochloride. 4,4′-Diamino-α-truxillic acid (4ATA) dihy-
drochloride was synthesized by the procedure shown below. Then 12
N hydrochloric acid (2 mL) was added dropwise into an acetone
solution (80 mL) of 4ACA (3.26 g, 19.99 mmol) to produce 4-
aminocinnamic acid hydrochloride (4ACA hydrochloride). The 4ACA
hydrochloride powders were then irradiated by a 100-W high pressure
Hg-lamp (Omni Cure S1000, EXFO Photonic Solution Inc.) with a
250−450 nm band-pass filter and an intensity of 2.7 mW/cm3 for 30−
40 h to induce [2 + 2] photocycloaddition. As a result, 4ATA
dihydrochloride was formed with the following specifications. Mp:
178.0 °C. 1H NMR (400 MHz, DMSO-d6, δ, ppm): 3.82 (dd, 2H, J =
7.7, 9.6 Hz), 4.30 (dd, 2H, J = 7.7, 9.6 Hz), 7.33 (d, 4H, J = 7.7 Hz),
7.45 (d, 4H, J = 7.7 Hz), 10.37 (s, 6H), 12.07 (s, 2H). 13C NMR (400
MHz, DMSO-d6, δ, ppm): 40.4, 46.0, 122.8, 128.9, 130.6, 139.1, 172.6.
ATR-FTIR: 2876, 2576, 1702, 1502, 1178, 822 cm−1. FT-ICR MS
(ESI): calcd for [M + H, C18H17N2O4]+, 325.1188; found, 325.1190.
Synthesis of 4,4′-Diamino-α-truxillic Acid Dimethyl Ester.
4ATA dimethyl ester was synthesized as follows. 4ATA dihydro-
chloride (3.00 g, 7.54 mmol) was mixed with trimethylsilyl chloride
(TMSCl) (1.63 g, 5.08 mmol) and methanol (12.20 mL, 0.32 mol) in
a two-necked round-bottomed flask under a nitrogen atmosphere at
room temperature. The reaction proceeded in a dispersion state at
first, but as a result of ongoing esterification, the products became
dissolved in the methanol. The reaction mixture was precipitated and
dissolved into water, and extracted by ethyl acetate. The collecting
ethyl acetate organic layer was dried over MgSO4, and it was then
evaporated and neutralized using 1 N NaOH. The specifications were
as follows. Mp: 193 °C. 1H NMR (400 MHz, DMSO-d6, δ, ppm): 3.27
(s, 6H), 3.72 (dd, 2H, J = 7.4, 10.2 Hz), 4.11 (dd, 2H, J = 7.4 Hz, 10.1
Hz), 5.00 (s, 4H), 6.51 (d, 4H, J = 8.4 Hz), 6.95 (d, 4H, J = 8.4 Hz).
13C NMR (400 MHz, DMSO-d6, δ, ppm): δ 40.4, 46.6, 51.0, 113.7,
Biospin AG 400 MHz, 54 mm spectrometer using DMSO-d6 as the
solvent. The FT-IR spectra were recorded with a Perkin-Elmer
Spectrum One spectrometer between 4000 and 600 cm−1 using a
diamond-attenuated total reflection (ATR) accessory. The mass
spectra were measured using a Fourier transform ion cyclotron
resonance mass spectrometer (FT-ICR MS, Solarix), and scanned
from m/z 150 to m/z 1000. X-ray diffraction (XRD, RINT 2000 and
Rigaku) was used to determine the degree of crystallization using the
Gram AI program. The inherent viscosities were measured using an
Ubbelohde viscometer at 30 °C in DMAc at a polymer concentration
of 0.5 g/dL. The tensile measurements were carried out at an
elongation speed of 0.5 mm/min on a tensiometer, the Instron 3365 at
room temperature. The number-average molecular weight (Mn),
weight-average molecular weight (Mw) and the molecular weight
distribution (PDI) were determined by gel permeation chromatog-
raphy (GPC, concentration 5 g/L, DMF eluant) after calibration with
pullulan standards. The ultraviolet−visible (UV−vis) optical absorp-
tion spectra were recorded on a Perkin-Elmer, Lambda 25 UV/vis
spectrophotometer at room temperature over the range of 200−800
nm. The refractive indices of thin PI films were evaluated by an Abbe
refractometer (Atago, NRA, 1 T) employing α-bromonapthalene as a
contact liquid at room temperature. The solubility of the polymers was
investigated using 0.001 g of sample in 1 mL of solvent at room
temperature and at 60 °C. Thermogravimetric analysis (TGA) and
different scanning calorimetry (DSC) were performed on Seiko
Instruments SII, SSC/5200 and Seiko Instruments SII, X-DSC7000T,
respectively, at a heating rate of 10 °C/min under a nitrogen
atmosphere. The polymer specimens were dried at 100 °C for 1 h to
remove any absorbed moisture before both TGA and DSC.
Surface Free Energy Estimation. The surface wettability and
hydrophobic properties of the polymers were assessed by static contact
angle (θ) measurements using the sessile drop method. Three different
liquids were used: ultrapure water (polar), ethylene glycol, and
diiodomethane (nonpolar). The contact angle between a droplet of
liquid and the polymer surface was measured by a contact angle
measurement apparatus (Drop-Master DM 300, Kyowa Interface
Science Co). The total surface energies and their dispersion, polar and
hydrogen bonding components, were calculated using the following
equation. The presented data are the average of three measurements
125.7, 128.0, 147.3, 172.1. ATR-FTIR: 3400, 3314, 3046, 2952, 1714,
1516, 1172, 804 cm−1. FT-ICR MS (ESI): calcd for [M + H,
C20H23N2O4]+, 355.1658; found, 355.1663.
Synthesis of 4,4′-Diamino-α-truxillic Acid Diethyl Ester.
4ATA diethyl ester was prepared by an analogous procedure to
4ATA dimethyl ester. However, instead of methanol, ethanol was used
as the reactant and solvent under a nitrogen atmosphere at reflux
temperatures. The specifications were as follows. Mp: 137 °C. 1H
NMR (400 MHz, DMSO-d6, δ, ppm): 0.84 (t, 6H, J = 7.0 Hz), 3.68
(dd, 2H, J = 7.2, 7.2 Hz), 3.73 (q, 4H, J = 6.8, 7.6 Hz), 4.10 (dd, 2H, J
= 7.2, 7.6 Hz), 4.96 (S, 4H), 6.49 (d, 4H, J = 8.4 Hz), 6.95 (d, 4H, J =
8.4 Hz). 13C NMR (400 MHz, DMSO-d6, δ, ppm): 13.7, 40.3, 46.5,
59.6, 113.5, 125.7, 128.1, 147.3, 171.6. ATR-FTIR: 3456, 3364, 3032,
2964, 1716, 1518, 1190, 824 cm−1. FT-ICR MS (ESI): calcd for [M +
H, C22H27N2O4]+, 383.4680; found, 383.1300.
γtot = γLW + 2 γ+γ−
(1)
where γLW is the surface energy from van der Waals interactions or
dispersion, γ+ is the surface energy from acceptor interactions, γ− is the
surface energy from donor interactions, and γtot is the total surface
energies.
1587
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
In Vitro Cell Culture. Cell Culture. A mouse fibroblast-like cell line
(L929) was selected for all biological assays in order to evaluate the
effect of the various PIs on cell adhesion. The L929 fibroblast cell line
was obtained from the American Type Culture collection (Manassas,
VA). The cells were cultured in Dulbecco’s Modified Eagle’s Medium
(DMEM, Sigma, USA) supplemented with 10% heat-inactivated fetal
bovine serum (FBS, Biochrom AG, Germany) incubated at 37 °C in a
humidified atmosphere with 5% CO2.
Cell Adhesion. The polymer films were casted on 15 mm glass
microcoverslips, and were seeded with 100 μL of a cell suspension (2.5
× 104 cells·mL−1) and cultured for 1, 2, and 4 days at 37 °C. Prior to
culturing, all of the polymers were sterilized with ethanol. A bare well
plate was used as the control. After each incubation period, the
samples were rinsed with a buffer saline (PBS, Sigma-Aldrich, USA).
We can distinguish whether the cells changed from a circular to an
elongated shape after adhesion.19 The number of cells adhering to the
surface of the polymers was then counted.
Figure 1. High-performance liquid chromatography (HPLC) separa-
tion of 4ACA bioconverted from 4APhe. The AtPal4 producing
recombinant E. coli was incubated with 10 mM 4APhe at 37 °C for 0 h
(line 1) and 12 h (line 2).
RESULTS AND DISCUSSION
■
Monomer Syntheses. The cinnamate photonic [2 + 2]
cycloaddition is one of the most efficient chemical dimeriza-
tions if the reaction conditions are well arranged. We then
made a strategy to prepare bioderived aromatic diamines using
the photoreaction of cinnamoyl monoamine. In terms of a
functional substitute group for cinnamates, the phenylene
hydroxyl group has mainly been reported,10 and there have
been a few studies on aminocinnamate derivatives.15 One
report showed that some bacteria produced 4-aminophenyla-
lanine (4APhe) as an intermediate of antibiotics.20 There are
established systems for fermenting glucose biomass to produce
4APhe.21,22 4ACA could be obtained from 4APhe under a kind
of phenylalanine ammonia lyases, in which the α-amino residue
of phenylalanine (or its derivatives) could be deaminated to
generate cinnamic acid (or its derivatives). However, the
bioavailability of 4ACA was unclear. We therefore investigated
the deamination of 4APhe to synthesize biomass-derived ACA
by using recombinant phenylalanine ammonia lyases. One
isozyme, phenylalanine ammonia lyase 4 (AtPal4) from
Arabidopsis thaliana, was functionally produced in recombinant
E. coli. The strain produced 1.4 × 10−2 and 3.8 × 10−4 μmol
min−1 mg−1 of phenylalanine- and 4APhe-ammonia lyase
activity, respectively, indicating that AtPal4 used 4APhe as a
substrate. Incubating the bacterial cells with 4APhe decreased
the amount of 4APhe produced, and generated 0.13 g/L of
4ACA (Figure 1). This indicated that the combination of this
process and the 4APhe fermenting system produced 4ACA
from the biomass.
appeared, while the proton signals assigned to carboxylic acids,
amines, and aromatics shifted slightly. It also confirmed a
photodimer cyclobutane with only trans-configuration showing
simple dd signals with coupling constant J = 7.7, 9.6 Hz. When
4ACA in a nonsalt state was photoirradiated, no reaction
occurred. The salt state of 4ACA was therefore very important
for a quantitative reaction. 4ATA has two amine groups but also
two carboxylic acids, and was difficult to use as a monomer for
polyimides. Next, 4ATA was esterified in the dispersion state of
reactant alcohol solvents such as methanol and ethanol in the
presence of TMSCl. If 4ACA was esterified, then the products
would be dissolved in the solvents to react smoothly. As a
result, the 1H NMR signals of these alkyl esters were confirmed
as shown in Figure S1c (4ATA dimethyl) and S1d (4ATA
diethyl). The IR spectrum of a broad hydroxyl peak at 2865
cm−1 disappeared and the carbonyls at 1700 cm−1 shifted
toward the higher wavenumber at 1713 cm−1 as a result of the
carboxylic acid reaction to become esters (Figure S4).
Moreover, the expected structures of both 4ATA esters
dihydrochloride were also confirmed by 13C NMR (Figure
S2), and the 4ATA methyl ester dihydrochloride was confirmed
1
by its H−13C NMR HSQC and HMBC spectra (Figure S3).
Polymer Syntheses. PAAs, which are the precursors of PIs,
were prepared by the polycondensation of the prepared
diamines of 4ATA esters with stoichiometric amounts of the
dianhydrides CBDA, PMDA, BTDA, OPDA, BPDA, and
DSDA (Scheme 2). The resulting PAAs were abbreviated as
PAA-1m, PAA-1e, PAA-2m, PAA-2e, PAA-3m, PAA-4m, PAA-
5m, and PAA-6m where suffixes m and e means methly and
ethyl side chains, respectively, as shown in Table 1. Among
these dianhydrides, we confirmed that CBDA was converted by
Hg-lamp irradiation (λ = 250−450 nm) to maleic acid
dimethyl, which was a bioavailable derivative, and could be
successively cyclized and treated with acetic anhydride. The
corresponding PAA-1m and PAA-1e were fully biobased
Recently, a [2 + 2] photocycloaddition reaction of trans-
cinnamic acid to α-truxillic acid has been reported under
illumination with ultraviolet light source with wavelengths
above 260 nm which was stable at elevated temperature and
under a wide range of wavelengths of UV lights.23 Single crystal
X-ray analyses of photoirradiated trans-cinnamic acid had also
revealed the formation of a centrosymmetic in α-truxillic acid
photodimer.24 Our strategy, starting from 4ATA dihydro-
chloride, was synthesized by a perfect [2 + 2] photo-
cycloaddition of 4ACA hydrochloride salt (Scheme 1). The
1
polymers, whereas the others were partially biobased. The H
NMR spectra and FT-IR spectra of all PAAs are shown in
1
1
photodimerized structure was confirmed by H NMR (Figure
Figure S5 and S7, respectively. In the H NMR spectra for
S1 in the Supporting Information) and FT-IR (Figure S4 in the
Supporting Information). Figure S1a (4ACA) and S1b (4ATA
dihydrochloride) in the Supporting Information showed that
photoirradiation resulted in the disappearance of the vinylene
proton signals around 7.44−7.40 and 6.15−6.12 ppm, and
instead a cyclobutane proton signal around 4.33−3.81 ppm
PAA-1m, the main chain proton signals for amides, aromatics,
and cyclobutane appeared around 10.6−10.1, 7.6−7.3, and
4.3−3.3 ppm, respectively. Other PAAs showed signals from
dianhydride-derived aromatic protons around 8.5−7.3 ppm in
addition to the above-mentioned signals. The methyl protons
of the PAAs overlapped with water at about 3.3 ppm. However,
1588
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
Scheme 1. Syntheses of Biobased Aromatic Diamines, 4,4′-Diamino-α-truxillic Acid Diester
Scheme 2. Syntheses of Biobased Aromatic Poly(amic acid)s and Polyimides from 4ACA (R is Methyl (m) or Ethyl (e) Side
Chain)
no loss of signal of the integral values of the methyl proton was
confirmed by the integral values of the benzene/methyl protons
in d7-DMF (Figure S6). Similarly, NMR studies confirmed the
formation of PAAs derived from the aromatic diamine of 4ATA
ethyl ester with dianhydrides.
Furthermore, other peaks at 1518 cm−1 (C−C stretching of
aromatic), 1441 cm−1 (CC stretching of p-substituted
benzene) 1376 cm−1, (C−N stretching of imide), and 1175
cm−1 (imide ring deformation) appeared, which indicated a
complete imidization.25 The PIs were abbreviated PI-1m, PI-1e,
PI-2m, PI-2e, PI-3m, PI-4m, PI-5m, and PI-6m, whose
numerals correspond with the PAAs shown in Table 1. PI-4m
from OPDA showed IR peaks assigned to the C−O stretching
of ether group at 1238 cm−1,26 and PI-6m from DSDA showed
peaks assigned to asymmetric and symmetric SO stretching
at 1323 and 1148 cm−1, respectively. These results clearly
indicated the formation of the expected PIs, which were clear,
flexible and strong.
The weight-average molecular weight of (Mw), the number-
average molecular weight (Mn) and the molecular weight
distribution (PDI) were determined using the PAAs, and are
summarized in Table 1. PAAs had high Mw and Mn values in
the range of 2.09−3.99 × 105 and 1.68−4.61 × 105,
respectively, and the PDI ranged from 1.25 to 1.51. PAA
solutions with a concentration of 0.5 g/dL have inherent
viscosities of 0.34−1.23 dL/g at 30 °C, indicating a moderate to
All of the PAAs were soluble in DMAc solvent, and colorless
or yellowish films with high transparency were easily prepared
by casting at 60−70 °C. The films were subsequently imidized
to anneal stepwise at 100, 150, 200, and 250 °C for 1 h at each
step in an oven. The films became light to deep yellow. The
imidization was confirmed by IR spectroscopy. The FT-IR
spectra of the PAAs (Figure S7) showed broad absorption
bands of the hydroxyls in the range of 2500 to 3500 cm−1 (O−
H stretching), two different carbonyl peaks at 1719 cm−1 (C
O stretching, carboxylic) and 1668 (CO stretching, amide),
and aromatic peaks at 1520 and 1436 cm−1 (C−H overtone
aromatic), in all samples. On the other hand, all of the annealed
films (Figure S8) prepared here showed double carbonyl
absorption: a small peak at 1785 cm−1 (CO asymmetric
stretching) and a big peak at 1716 cm−1 (CO symmetric
stretching), which were characteristic to PI structures.
1589
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
Table 1. Molecular Weights of PAAs and PIs and Transparency of their Films
a
b
b
b
c
c
polymers
ηinh (dL/g)
Mw
Mn
PDI
λ0 (nm)
T450 nm (%)
PAA-1m
PI-1m
1.23
ND
1.00
ND
0.50
ND
0.74
ND
0.72
ND
0.34
ND
0.53
ND
0.56
ND
ND
3.99 × 105
ND
3.57 × 105
ND
3.19 × 105
ND
2.09 × 105
ND
3.06 × 105
ND
3.15 × 105
ND
2.56 × 105
ND
2.78 × 105
ND
2.78 × 105
ND
2.54 × 105
ND
4.61 × 105
ND
1.68 × 105
ND
2.25 × 105
ND
2.20 × 105
ND
1.70 × 105
ND
1.97 × 105
ND
1.43
ND
1.46
ND
1.45
ND
1.25
ND
1.36
ND
1.44
ND
1.51
ND
1.41
ND
ND
284
270
290
282
345
357
329
355
329
326
326
363
351
383
346
366
384
99.5
88.2
96.9
86.9
95.8
80.1
91.8
50.7
95.7
79.1
87.9
68.9
89.5
65.4
90.1
82.2
0.7
PAA-1e
PI-1e
PAA-2m
PI-2m
PAA-2e
PI-2e
PAA-3m
PI-3m
PAA-4m
PI-4m
PAA-5m
PI-5m
PAA-6m
PI-6m
Kapton
ND
ND
a
b
ηinh: Inherent viscosities of PAA were measured at a concentration of 0.5 dL/g in DMAc at 30 °C. The weight-average molecular weight, Mw, the
number-average molecular weight, Mn, and the distribution of polymer molecular weight, PDI, of PAA were measured by GPC. ND refers to not
c
determined. UV cutoff wavelength, λ0, and transmittance at 450 nm, T450 nm, were measured by UV/vis spectrophotometer.
Table 2. Thermomechanical Properties and Refractive Indices of PI Films
TGA
film
T
degree of
crystallization (%)
T
T
tensile strength
moduli
d
elongation
refractive
g a
b
5 c
10 c
d
d
e
polymers property (°C)
(°C)
(°C)
density (g·cm−3
)
(MPa)
(GPa)
(%)
index
PI-1m
PI-1e
flexible
flexible
flexible
flexible
flexible
flexible
flexible
flexible
flexible
ND
ND
ND
ND
260
250
240
275
30
11
−
−
−
−
−
−
−
365
380
410
380
400
400
395
410
440
390
394
425
394
420
410
410
425
590
1.20
1.26
1.15
1.21
1.32
1.23
1.24
1.31
1.43
75 6.6
81 7.7
89 9.2
96 3.4
48 0.8
98 5.7
71 2.1
90 5.3
135 12.5
10.01 3.68
4.33 1.13
8.02 1.19
5.65 0.98
4.24 0.18
13.39 3.03
4.36 0.55
4.77 0.75
4.60 0.79
1.82 0.28
2.07 0.01
2.48 0.12
4.59 0.66
1.72 0.33
4.49 0.49
2.42 0.43
3.31 0.32
13.27 0.56
1.60
1.65
1.65
1.65
1.65
1.64
1.65
1.64
1.65
PI-2m
PI-2e
PI-3m
PI-4m
PI-5m
PI-6m
Kapton
f
180
a
Glass transition temperature, Tg, which was measured by DSC thermogram scanned at a heating rate of 10 °C/min under nitrogen atmosphere.
b
c
Degree of crystallization was determined from XRD. 5% and 10% weight loss temperatures, T5 and T10, were obtained from TGA curve scanned at
d
a heating rate of 10 °C/min under nitrogen atmosphere. Tensile strength, moduli, and elongation were obtained from a tensiometer at room
temperature. Refractive index was measured by Abbe refractometer. Tg of commercial Kapton was measured by DSC by ourselves and was lower
e
f
than expected.
high molecular weight of the polymers (Table 1). Although the
inherent viscosity of PI-4m was relatively low, it gave a self-
standing film similar to the other PIs prepared here. Fully
biobased PAA-1m exhibited the highest molecular weight and
the highest viscosity of all PAAs investigated here.
toluene, hexane, chloroform, diethyl ether, and dichloro-
methane, (2) polar protic solvent such as distilled water,
methanol, ethanol, and concentrated sulfuric acid, and (3) polar
aprotic solvent such as acetone, acetonitrile, DMF, DMAc,
NMP, DMSO, THF, and ethyl acetate. The PAAs were soluble
in polar solvents such as concentrated sulfuric acid, NMP,
DMAc, DMF, and DMSO at room temperature. On the other
hand, all of the PIs were insoluble in any solvents, even after
heating, except for concentrated sulfuric acid, probably due to
their coplanar rings of aromatic-imide structure. These results
indicated that the PIs had high chemical resistance.
Thermo-Mechanical Properties. The thermal degrada-
tion of the PIs was investigated by thermogravimetric analysis
(TGA) in a nitrogen atmosphere at a heating rate of 10 °C/
min, and the 5% and 10% weight-loss temperatures, T5 and T10,
were determined (Figure S10). As shown in Table 2, all of the
PIs exhibited a T10 range of 390−425 °C, meaning good
resistance against thermal decomposition; especially PI-2m
from PMDA exhibited the highest thermal stability with a T10
of 425 °C. Nevertheless, PI-2e with ethyl side chains showed a
The crystalline structures of the PIs were investigated by X-
ray diffraction (XRD) (Figure S9). The XRD pattern of PI-1m
from CBDA gave two diffraction peaks at 2θ = 6° and 17°
overlapping with a broad amorphous halo, whereas the other
PIs showed no crystalline peaks. These results suggest that the
PI-1m structures with two cyclobutanes and two phenylenes, as
well as the imide rings in the main chains, induced a partial
crystallization. If the methyl ester side group was replaced by an
ethyl, the crystallization degree decreased from 30% to 11%
because of possible steric hindrance from the bulky ethyl
groups against the crystalline packing structure. The other PIs
were amorphous, which has typically been observed in other PI
films.27
The solubility of the polymers was tested by dissolving them
in three groups of solvent: (1) nonpolar solvent such as
1590
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
Figure 2. UV−vis spectra of PAAs, PIs, and Kapton films (thickness ∼9−15 μm).
T10 of only 395 °C, presumably due to the steric effect of the
ethyl groups. PI-1m with a methyl side group and PI-1e with an
ethyl one derived from cyclobutane dianhydride showed lower
degradation temperatures than 400 °C due to the lower
amount of benzene rings than the others.
The densities of the polymer films, which are crucial factors
for polymer applications, were evaluated to be 1.15−1.32
g.cm−3. This is lower than that of commercial Kapton provided
from Du Pont-Toray Co., Ltd. in commercial grade containing
some additives such as plasticizers. Commercial Kapton showed
a higher mechanical strength, better mechanical failure and a
higher T10, but a lower tensile modulus and lower Tg, and were
heavier than PI-4m and the fully biobased PI-1m.
The Tg values of the PIs were determined by differential
scanning calorimetry (DSC, Figure S11) under a nitrogen
atmosphere and summarized in Table 2. PI-3m, PI-4m, PI-5m,
and PI-6m showed Tg values of 260, 250, 240, and 275 °C,
respectively. The Tg values of these PIs depended on the
backbone structure of the dianhydride component.25,26 PI-1m
and PI-2m exhibited no Tg below 350 °C, which is the
measuring limit of the DSC machine, meaning that they could
not determine Tg values due to their decomposition temper-
ature lower than Tg phenomena. Although the sulfone group of
DSDA should give flexibility to the PI-6m backbone, the PI had
a higher Tg than PI-3m, PI-4m, or PI-5m. It could be speculated
that an intermolecular charge transfer complex (CTC) of the
electron-rich diamine phenylene with the electron-poor
dianhydride phenylene connecting to the sulfone induced
good interchain packing in PI-6m. The relatively higher Tg of
PI-5m to PI-3m and PI-4m may also be attributed to an
interchain CTC of the electron-rich diamine phenylene with
the electron-poor dianhydride phenylene connecting to a
ketone. Overall, the Tg values of all PIs prepared here were
much higher than those of conventional biobased polymers of
poly(lactic acid)s, polyamide-11, and other aromatic biobased
polymers reported thus far. The stability of the planar 5-
membered ring of the imide linkage resulting from the n−π
conjugation between the carbonyl group with the nitrogen
atom without a lone electron pair could increase the Tg. Such
high Tg properties may be suitable for applications in
superengineering plastics.
Optical Properties. Figure 2 shows the UV−vis trans-
mittance dependence of PAA and PI films, where the curves
were normalized to the transmittance of a film with a thickness
of 9−15 μm. The optical transparency at 450 nm (T450) and
the cutoff wavelengths (λ0) of the UV−vis spectra are
summarized in Table 1. The PAA films showed no color or
pale-yellow colors, and the T450 values of most polymers were
88−100%. After imidization, the films became darker yellow
and showed lower T450 values. PI-1m, PI-1e, PI-3m, and PI-6m
kept their high transparency over a T450 range of 79−88%,
whereas PI-2e, PI-4m, and PI-5m showed T450 ranges of 51−
69% but were much more transparent than commercial Kapton.
The imdization caused ring-closing reactions, resulting in more
efficient charge transfer between the benzene from the
diamines and the one from the dianhydrides, leading to the
color change.28 Kapton has a dark brown color due to the
characteristic absorption tailing from UV to the visible region,
caused by strong charge transfer (CT) interactions in the
electron-rich oxidianiline component with dianhydride moi-
eties. The much higher optical transparency of the present
PAAs and PIs than that of Kapton should be due to the
cyclobutane moiety in the 4ATA component, which has fewer
electron-rich benzenes to the weaker CT interactions with the
dianhydride component than the oxidianiline in Kapton. In
almost colorless PI-1m films derived from CBDA, which is one
of the most simple alicyclic dianhydride monomers, non-
conjugated aromatic structures in the dianhydride component
gave a weak electron accepting ability to the imide moieties.29
The refractive indices of PI films ranged from 1.60 to 1.65 as
listed in Table 2. The high refractive indices are primarily due
to the relatively dense molecular packing originating from the
rigid structure composed of aromatic groups with high electron
density, as compared to other transparent plastics such as
polycarbonate (1.58) and poly(ethylene terephthalate) (1.57−
The mechanical properties of the PI films were measured by
a tensile test (Table 2). The PI films had tensile strength values
in the range of 48−98 MPa, tensile moduli in the range of 4.2−
13.4 GPa, and gross mechanical failure at strains in the range of
1.7−4.6%. The tensile strength, tensile modulus, and
mechanical failure of PI-4m were at 98 MPa, 13.4 GPa, and
4.5%, respectively, which were indicated that the PI-4m film
was the strongest and toughest.
1591
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
Figure 3. Time dependence of cell numbers of L929 fibroblasts adhered on various PIs.
Figure 4. Representative photos of L929 fibroblasts cells adhered on PI-4m film after incubated for 1 day and 4 days.
1.58). The lower refractive indices of PI-1m and PI-1e derived
from aliphatic CBDA dianhydrides than other the PIs was due
to the lack of aromatic moieties in the dianhydride.
containing no benzene ring in their dianhydride component
exhibited higher contact angles in water than PI-2m and PI-2e,
presumably due to the lower polarity of the dianhydride
cyclobutane rings. Furthermore, all of the PI films may be
considered monopolar basic owing to the zero values of their
basic component (γ+) < 0.3 mJ·m−2.
The number of the L929 fibroblast cells adhered onto various
PI film surfaces was counted (Figure 3). Commercial Kapton
was used as a positive control, because good cell compatibility
with Kapton has already been accepted.31,32 On the first and
second days of growth, the number of adhered cells on the PI
films did not increase, but the number increased on the fourth
day of incubation, especially PI-4m as representative PI (Figure
4). All PI films did not differ very much in the number of cells
adhered, and exhibited good cell compatibility although Kapton
showed a slightly higher number of cells adhered than these PI
films.
Surface Properties. Contact angles were measured on the
PI films using the three test solvents of distilled water, ethylene
glycol, and diiodomethane with the sessile drop technique
(Table S1 in the Supporting Information). In general, the
dispersive component (nonpolar) and donor/acceptor compo-
nent (polar) had great influence on the contact angle.30 The
contact angle measurements, especially for water, increased
with increased temperatures, indicating that the surface of the
PI films became less polar or more hydrophobic (80−88°).
Moreover, they had a slightly larger dispersive component
(γLW) and a slightly smaller nondispersive component (γ+ and
γ−). Therefore, van der Waals forces (dispersive force) played a
crucial role in the PI films. PI-3m, PI-4m, PI-5m, and PI-6m
with two benzene rings in the dianhydride component in the
repeated unit showed higher contact angles in water (85−88°),
whereas PI-2m and PI-2e with only one benzene ring showed
lower contact angles (80−81°). However, PI-1m and PI-1e
1592
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593

Macromolecules
Article
(13) Suzuki, H.; Ohnishi, Y.; Furusho, Y.; Sakuda, S.; Horinouchi, S.
J. Biol. Chem. 2006, 281, 36944−36951.
CONCLUSION
■
The biobased aromatic diamine 4,4′-diamino-α-truxilic acid
(4ATA) was prepared as a photodimer of 4-aminocinnamic
acid, which was bioavailable from genetically manipulated E.
coli, even though the direct biosynthesis of aromatic diamines
has never been reported. 4ATA reacted with various anhydrides
to create polyamic acid, and then successive heat treatments
yielded the polyimides (PI). When we used 1,2,3,4-
tetracarboxycyclobutane dianhydride (CBDA) as a counter
monomer, the PIs were fully biobased because CBDA is also a
photodimer of biomolecular maleic acid derivatives. All of the
biobased PI films prepared here showed ultrahigh thermal
resistance, T10 values over 425 °C and Tg values not determined
under 350 °C, which is the highest value of all biobased plastics
reported thus far. The PI films also showed high tensile
strength and high Young’s moduli, especially PI-4m which
showed the values of 98 MPa and 13.4%, respectively, and good
cell compatibility. These characteristics of biobased PI can lead
to the development of safe yet high-performance biomedical
materials. The additional properties of excellent transparency
and high refractive indices may lead to the development of
ophthalmological materials.
(14) Noguchi, A.; Kitamura, T.; Onaka, H.; Horinouchi, S.; Ohnishi,
Y. Nat. Chem. Biol. 2010, 6, 641−643.
(15) Johnson, O. H.; Green, D. E.; Pauli, R. J. Biol. Chem. 1994, 153,
37−47.
(16) Kitazume, T.; Tanaka, A.; Takaya, N.; Nakamura, A.;
Matsuyama, S.; Suzuki, T.; Shoun, H. Eur. J. Biochem. 2002, 269,
2075−2082.
(17) Tribe, D. E. Novel microorganism and method. U.S. Patent
Number: 4,681,852, 1987.
(18) Kane, F. K.; Fiske, M. J. J. Bacteriol. 1985, 161, 963−966.
(19) Yokota, T.; Terai, T.; Kobayashi, T.; Meguro, T.; Iwaki, M. Surf.
Coat. Technol. 2007, 201, 8048−8051.
(20) He, J.; Magarvey, N.; Piraee, M.; Vining, L. C. Microbiology
2001, 147, 2817−2829.
(21) Mehl, R. A.; Anderson, J. C.; Santoro, S. W.; Wang, L.; Martin,
A. B.; King, D. S.; Horn, D. M.; Schultz, P. G.; et al. J. Am. Chem. Soc.
2003, 125, 935−939.
(22) Yanai, K.; Sumida, N.; Okakura, K.; Moriya, T.; Watanabe, M.;
Murakami, T. Nat. Biotechnol. 2004, 22, 848−855.
(23) Bernstein, H. I.; Quimby, W. C. J. Am. Chem. Soc. 1943, 65,
1845−1846.
(24) Chi, Y.-M.; Nakamura, M.; Yoshizawa, T.; Zhao, X.-Y.; Yan, W.-
M.; Hashimoto, F.; Kinjo, J.; Nohara, T.; Sakurada, S. Biol. Pharm. Bull.
2005, 28, 1776−1778.
ASSOCIATED CONTENT
(25) Khan, M.; Brunklaus, G.; Enkelmann, V.; Spiess, H.-W. 2008,
130, 1741−1748.
(26) Zhang, S.; Li, Y.; Yin, D.; Wang, X.; Zhao, X.; Shao, Y.; Yang, S.
Eur. Polym. J. 2005, 41, 1097−1107.
■
S
* Supporting Information
Details of characterization including 1H NMR, 13C NMR,
1
1H−13C HSQC, H−13C HSQC, FT-IR and XRD, thermal
(27) Shao, Y.; Li, Y.-F.; Zhao, X.; Wang, X.-L.; Ma, T.; Yang, F.-C.
Polym. Sci., Part A: Polym. Chem. 2006, 44, 6836−6846.
(28) Mathews, A. S.; Kim, I.; Ha, C.-S. Macromol. Res. 2007, 15,
114−128.
analysis, TGA and DSC, and contact angles and surface energy.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(29) Lu, Y.; Hu, Z.; Wang, Y.; Fang, Q.-X. J. Appl. Polym. Sci. 2012,
125, 1371−1376.
AUTHOR INFORMATION
Corresponding Author
*(T.K.) E-mail: kaneko@jaist.ac.jp. Telephone: +81-761-51-
1633. Fax: +81-761-51-1635.
■
(30) Hasegawa, M.; Horie, K. Prog. Polym. Sci. 2001, 26, 259−335.
(31) Harnett, E. M.; Alderman, J.; Wood, T. J. Colloid Surf. B:
Biointerfaces 2007, 55, 90−97.
(32) Ball, M. D.; Prendergast, U.; O’ Connell, C.; Sherlock, R. Exp.
Mol. Pathol. 2007, 82, 130−134.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The research was financially supported by Advanced Low
Carbon Technology Research and Development Program (JST
ALCA, 5100270), Tokyo, Japan.
REFERENCES
■
(1) Okano, K.; Kondo, A.; Noda, H. Eco. Ind. 2006, 11, 43.
(2) Vert, M. Biomacromolecules 2005, 6, 538−546.
(3) Taniguchi, I.; Kimura, Y. Biopolymers 2001, 3b, 431.
(4) Thielemans, W.; Wool, R. P. Biomacromolecules 2005, 6, 1895−
1905.
(5) Madhavamoorthi, P. Synth. Fibers 2004, 33, 16−28.
(6) Kaneko, T.; Matsusaki, M. T.; Hang, M.; Akashi, M. Macromol.
Rapid Commun. 2004, 25, 673−677.
(7) Kaneko, T.; Tran, H. T.; Matsusaki, M.; Akashi, M. Chem. Mater.
2006, 18, 6220−6226.
(8) Kaneko, D.; Kinugawa, S.; Matsumoto, K.; Kaneko, T. Plant
Biotechnol. 2010, 27, 293−296.
(9) Wang, S.; Kaneko, D.; Kan, K.; Jin, X.; Kaneko, T. Pure Appl.
Chem. 2012, 84, 2559−2568.
(10) Kaneko, T.; Thi, T. H.; Shi, D. J.; Akashi, M. Nat. Mater. 2006,
5, 966−970.
(11) Ronava, I. A.; Bruma, M. Struct. Chem. 2010, 21, 1013−1020.
(12) Choi, M.-C.; Wakita, J.; Ha, C.-S.; Ando, S. Macromolecules
2009, 42, 5112−5120.
1593
dx.doi.org/10.1021/ma402499m | Macromolecules 2014, 47, 1586−1593