Multi-functional polyimides containing tetraphenyl fluorene moieties: Fluorescence and resistive switching behaviors

Article abstract of DOI:10.1039/c7tc01807j

Two novel aromatic polyimides (PIs), TPF27OPI and TPF99OPI, with a tetraphenyl fluorene (TPF) unit as the electron donor (D) and a 4,4′-oxydiphthalic dianhydride (ODPA) unit as the electron acceptor (A) were synthesized and characterized. Both of the as-synthesized PIs showed obvious photoluminescence properties, and compared with TPF27OPI, where the fluorene unit is located in the polymer main chain, TPF99OPI with a fluorene unit in the side chain showed stronger fluorescence and a fluorescence blue shift in the film state. The memory device with a sandwich configuration of ITO/TPF27OPI/Al showed the nonvolatile write once, read many (WORM) switching behavior, while that of ITO/TPF99OPI/Al exhibited static random access memory (SRAM) switching behavior. Both of the devices exhibited a stable ON/OFF current ratio above 10⁴ and a long operation time, following 10⁴ pulse reading cycles with no obvious degradation in current. The mechanism of the resistive switching behaviors was discussed using the experimental and molecular simulation results. These multi-functional polyimides showed attractive potential applications in the field of polymer optoelectronic devices and memory devices.

Full text of DOI:10.1039/c7tc01807j

Journal of
Materials Chemistry C
View Article Online
View Journal
PAPER
Multi-functional polyimides containing
tetraphenyl fluorene moieties: fluorescence
Cite this:DOI: 10.1039/c7tc01807j
and resistive switching behaviors†
Lunjun Qu, Shida Huang, Yi Zhang, * Zhenguo Chi,
and Jiarui Xu
Siwei Liu, Xudong Chen
Two novel aromatic polyimides (PIs), TPF27OPI and TPF99OPI, with a tetraphenyl fluorene (TPF) unit as
the electron donor (D) and a 4,40-oxydiphthalic dianhydride (ODPA) unit as the electron acceptor (A)
were synthesized and characterized. Both of the as-synthesized PIs showed obvious photoluminescence
properties, and compared with TPF27OPI, where the fluorene unit is located in the polymer main chain,
TPF99OPI with a fluorene unit in the side chain showed stronger fluorescence and a fluorescence blue
shift in the film state. The memory device with a sandwich configuration of ITO/TPF27OPI/Al showed
the nonvolatile ‘‘write once, read many’’ (WORM) switching behavior, while that of ITO/TPF99OPI/Al
exhibited ‘‘static random access memory’’ (SRAM) switching behavior. Both of the devices exhibited a
stable ON/OFF current ratio above 104 and a long operation time, following 104 pulse reading cycles
with no obvious degradation in current. The mechanism of the resistive switching behaviors was discussed
using the experimental and molecular simulation results. These multi-functional polyimides showed
attractive potential applications in the field of polymer optoelectronic devices and memory devices.
Received 25th April 2017,
Accepted 5th June 2017
DOI: 10.1039/c7tc01807j
rsc.li/materials-c
with low glass transition temperature.9 PIs not only exhibit
outstanding thermal stability, high mechanical strength, good
Introduction
With the increasing development of information technology in solubility and flexibility, but also excellent structural designability.
the past decade, low turn-on voltage, fast reading speed and Moreover, unlike encoding ‘‘0’’ (charge OFF) and ‘‘1’’ (charge ON)
high data storage density have been the indispensable conditions in a memory cell in silicon-based memory devices, polymeric
of the semiconducting memory devices.1,2 Compared with the memory devices are characterized by bistable resistive switching
inorganic materials, organic and polymeric materials with low- behaviors with a low applied voltage and a high current ‘‘ON/OFF’’
cost potential, easy processability and large capacity for data ratio.10 Recently, functional PI-based memory devices have been
storage have been promising materials and will be applied in demonstrated to exhibit different kinds of resistive switching
the next generation of the semiconductor industry.3 The reported behaviors from non-volatile memory properties, such as write
organic or polymeric materials for electronic memory device once-read many memory (WORM)11 and FLASH memory,12 to
applications include conjugated organic molecules,4 complex volatile memory properties, such as dynamic random access
organic molecules,5 copolymers,6 polymer nanocomposites,7 memory (DRAM)13 and static random access memory (SRAM).14
and functional polyimides,8 etc.
Although various PI-based memory devices with different
Polyimides (PIs) with optoelectronic functions have attracted resistive switching behaviors have been prepared, the structure–
more and more attention as new functional materials compared property relationship of PI and their electronic memory
with the traditional inorganic materials and polymeric materials mechanisms are still unclear.15 Currently, the intramolecular
and intermolecular charge-transfer (CT) interactions between
the electron donor (D, diamine moiety) and acceptor (A, dian-
PCFM Lab, GD HPPC Lab, Guangdong Engineering Technology Research Centre for
hydride moiety) in polyimide moieties play an important role in
the memory performances and have been considered as a main
factor to influence the resistive switching behaviors under the
applied voltage.16 However, most previously reported PI-based
memory devices possess different molecular structures, and
a fine structure–property relationship is rarely systematically
established.
High-performance Organic and Polymer Photoelectric Functional Films, State Key
Laboratory of Optoelectronic Materials and Technologies, School of Chemistry,
Sun Yat-sen University, Guangzhou 510275, China. E-mail: ceszy@mail.sysu.edu.cn;
Fax: +86 20 84112222; Tel: +86 20 84112222
† Electronic supplementary information (ESI) available: NMR spectra of the
monomers and the polyimides; FTIR spectra, WAXD patterns, thermal properties
and transmittance of polyimides; SEM and fitted curves of the memory device.
See DOI: 10.1039/c7tc01807j
This journal is ©The Royal Society of Chemistry 2017
J. Mater. Chem. C

View Article Online
Paper
Herein, we report two novel diamines, 4,40-(9,9-diphenyl-9H-
Journal of Materials Chemistry C
fluorene-2,7-diyl)dianiline (TPF27DA) and 4,40-(2,7-diphenyl-
9H-fluorene-9,9-diyl)dianiline (TPF99DA), with the same molecular
weight containing a tetraphenyl fluorene (TPF) unit core. Based
on the two diamines, two aromatic polyimides, TPF27OPI and
TPF99OPI, were prepared and characterized. Both of the as-prepared
PIs showed obvious photoluminescence properties, and compared
with TPF27OPI (maximum emission wavelength, EMmax = 506 nm),
where the fluorene unit is located in the polymer main chain,
TPF99OPI with a fluorene unit in the side chain showed stronger
fluorescence with a blue shift in the film state (EMmax = 492 nm).
And the resistive switching behaviors, conducted by a simple
sandwich device configuration consisting of spin-coated PI films
between the ITO bottom electrode and the Al top electrode, were
also characterized. The ITO/TPF27OPI/Al device showed WORM
switching behavior, while the ITO/TPF99OPI/Al device exhibited
SRAM switching behavior. Both of the PI-based memory devices
exhibited a stable ON/OFF current ratio above 104 and a long
operation time, following 104 pulse reading cycles with no obvious
degradation in current. In order to clarify the switching mechanism
of the devices, theoretical calculations performed using the density
functional theory (DFT) method at the B3LYP level with the 6-31G(d)
basic set were also applied to analyze the geometry effect and
electronic transitions of the studied polyimides.
Fig. 1 1H NMR spectra of the diamines.
TPF99DA (4.95 ppm), which is due to the conjugation of the
amino group with the TPF unit, and will have an obvious effect
on the photoluminescence of the diamines and the polyimides.
The absorption and photoluminescence spectra of the diamines
were characterized both in solution and in the solid state, as shown
in Fig. 2. The absorption peaks at 355 nm and 310 nm of TPF27DA
and TPF99DA, respectively, resulted from the p–p* transition of the
conjugated fluorene with a benzene ring, which had a red-shift in
polar solvents (Fig. 2(a) and Fig. S4, ESI†). However, the absorption
peak at 336 nm of TPF99DA resulted from intramolecular charge
transfer, which was weakened by increasing the solvent polarity
(Fig. S5, ESI†). When excited at the maximum excitation wave-
length of 362 nm, TPF27DA showed a strong royal purple emission
peak at 421 nm, while that of TPF99DA was red-shifted and had a
bright blue-green emission peak at 473 nm when excited at 341 nm
in THF solution (Fig. 2(b)). In the solid state, TPF27DA exhibited a
strong blue emission peak at 432 nm (Fig. 2(c)), while that of
TPF99DA exhibited a strong emission peak at 420 nm. It is
interesting to find that the Stokes shift of TPF99DA is much
larger than that of TPF27DA. Thus we calculated the frontier
orbitals for the two diamine molecules (Fig. S6, ESI†). The
results show that the HOMO and LUMO of TPF27DA have
higher overlapping integrals, while those of TPF99DA are totally
separated. These phenomena indicate that the special large
conjugated and non-coplanar structure of TPF99DA forms a
typical D–p–A molecule with a strong internal charge transfer
(ICT) nature, according to our previous reports.17,18 In other
words, ICT is much easier to be formed in the case of TPF99DA.
In fact, the Stokes shift of TPF99DA (Fig. S8, ESI†) is much larger
than that of TPD27DA (Fig. S7, ESI†) in different solutions,
indicating that TPF99DA tends to have stronger ICT properties.
Results and discussion
Synthesis and characterization of the diamines
The diamine TPF27DA was synthesized via the Suzuki–Miyaura
reaction, and the diamine TPF99DA was synthesized according
to our previous work,17 as shown in Scheme 1. 1H NMR, 13C
NMR, mass spectra, IR spectra and elemental analysis techniques
were used to confirm the chemical structure of the intermediates
and the diamines, as shown in the Experimental section and
Fig. S1 and S2 (ESI†). The proton signal of TPF27DA is at around
5.25 ppm, and that of TPF99D is at around 4.95 ppm (Fig. 1). The
infrared absorptions at around 3345 cmꢀ1 and 3420 cmꢀ1, and
3370 cmꢀ1 and 3460 cmꢀ1, respectively, which are characteristic
of the amino groups, indicating that the diamines were successfully
synthesized, as shown in Fig. S3 (ESI†). The resonance of the proton
signals at 5.25 ppm for TPF27DA is much higher than that for
Synthesis and characterization of the polyimides
The polyimide films (TPF27OPI and TPF99OPI) were prepared
by reaction of the as-synthesized diamine and the commercial
4,4-oxydiphthalic anhydride (ODPA) through traditional poly-
condensation, as shown in Scheme 2. Mixing of the monomers
produced a viscous polyamic acid (PAA) solution in DMF, then
the PAA solution was casted on a clean dry glass plate followed
by programmed thermal imidization. In Fig. 3, the infrared
absorptions at 1780 cmꢀ1 (asymmetric CQO stretching) and
1720 cmꢀ1 (symmetric CQO stretching) for TPF27OPI and at
1782 cmꢀ1 and 1724 cmꢀ1 for TPF99OPI, respectively, correspond
Scheme 1 Synthesis routes of diamine monomers.
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2017

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 2 The optical spectra of the diamines in THF solution and in the solid state.
Scheme 2 Synthesis of the polyimides.
Fig. 3 FT-ATR spectra of the polyimides.
to the imide ring, and no amide bands are found. TPF27OPI was
insoluble in traditional solvents, such as NMP, DMSO, DMF, etc.,
even upon heating. And TPF99OPI was partially soluble upon
heating (90 1C). The thermal properties of the polyimides were emission wavelength (EMmax) is at 506 nm, and it emits a
measured using TGA and DSC, as shown in Fig. S9 and S10 considerably low intensity yellow-green light, when excited by
(ESI†). Both of the polyimides had a high glass transition the maximum excitation wavelength (EXmax) of 416 nm. However,
temperature (Tg), for TPF27OPI at 340 1C and for TPF99OPI at the TPF99OPI emits a considerably high intensity blue-green light
338 1C, respectively. And they also showed good thermal stability with the EMmax at 492 nm, when excited at 388 nm (EXmax).
with a 5% weight loss temperature (Td5%) for TPF27OPI at 602 1C Fluorene is an efficient fluorescence emission group; for
and for TPF99OPI at 586 1C, respectively. Their excellent thermal TPF99OPI, the sp3 hybridized carbon atom (C-9) separated the
properties could ensure the electronic memory devices can be conjugated fluorene structure from the polymer main chain, so it
prepared at high temperature and maintain high stability in shows a non-conjugated structure in the main chain, which may
applications.
enhance the luminescence, while for TPF27OPI, the conjugated
The optical properties of the polyimide films were studied by fluorene structure is at the main chain, which may quench the
UV-vis and photoluminescence spectroscopy. The absorption fluorescence of the conjugated fluorene structure because of the
spectra of the polyimide films are similar to those of the strong CT effect between the diamine moiety (electron donor)
diamines, as shown in Fig. 4(a) and Fig. S11 (ESI†). For the and the dianhydride moiety (electron acceptor).
yellow TPF27OPI film, the absorption edge wavelength (ledge
)
The electrochemical behaviors of the as-prepared polyimides
and the maximum absorption peak (lambasx) are 377 nm and 340 nm, were investigated by cyclic voltammetry (CV) measurements, as
respectively, while for the light-colored TPF99OPI film, the ledge is shown in Fig. 4(c). The PI film casted on the ITO-coated glass
362 nm, and there are two distinct absorption peaks located at slide acts as the working electrode, tetrabutylammonium per-
abs
max
333 nm and 318 nm, respectively. The l
of TPF27OPI and chlorate (TBAP, 0.1 mol mLꢀ1) in CH3CN as the supporting
TPF99OPI at 340 nm and 333 nm is suggested to be related to the electrode and Ag/AgCl as the reference electrode. TPF27OPI
OX
imide ring centered n–p* transitions. However, the absorption shows a maximum oxidation peak (E ) at 1.90 V and the onset
max
OX
onset
OX
max
peak at 318 nm is suggested to be related to the p–p* transition oxidation voltage (E
) is 1.48 V, while the E
of TPF99OPI is
is 1.54 V. Note that the oxidation voltage of
OX
centered at the big conjugated fluorene structure. In addition, the 1.86 V and the E
onset
optical energy band gaps (Eg) of the TPF27OPI and TPF99OPI are ferrocene (external standard) is 0.38 V, and thus the estimated
estimated to be 3.29 eV and 3.43 eV, respectively, according to the highest occupied molecular orbital (HOMO) energy levels of
Planck equation (Eg = 1240/ledge).
TPF27OPI and TPF99OPI are calculated as ꢀ5.90 V and ꢀ5.96 V,
The photoluminescence of the two polyimides is obviously respectively, based on the reference energy level of ferrocene
different, as shown in Fig. 4(b). For TPF27OPI, the maximum (4.80 eV below the vacuum level). Meanwhile, the lowest unoccupied
This journal is ©The Royal Society of Chemistry 2017
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
Fig. 4 (a) Absorption spectrum of the polyimide films; (b) fluorescence spectroscopy of the polyimide films; (c) cyclic voltammetry (CV) of PI cast films
on an ITO-coated glass slide in CH3CN (scan rate: 50 mV sꢀ1).
molecular orbital (LUMO) energy levels of the PIs are ꢀ2.61 eV and
ꢀ2.53 eV, respectively. Both the optical and the electrochemical
properties of the polyimides are summarized in Table 1. These
results indicate that the location of the tetraphenyl fluorene unit in
the macromolecular main chain has a certain impact on the HOMO
and LUMO energy levels of the polyimides, which in turn may have
an impact on the switching behaviors of the memory devices.
Memory device characteristics of the polyimides
Scheme 3 Schematic diagram of the memory device.
The configurations of sandwich memory devices with the
as-synthesized PIs as the active layer were fabricated successfully,
as shown in Scheme 3. The resistive switching behaviors of the
the ‘‘writing’’ process (‘‘1’’ signal in memory data storage). After
this transition, the device could remain in the ON state,
observed during the subsequent positive scan from 5 V to 0 V
(the second sweep). The next dual sweeps were performed after
turning the power off for usually less than 3 minutes, and the
subsequent negative scan from 0 V to ꢀ5 V (the third negative
sweep) did not turn the device from the ON state to the OFF
state, and the device could also retain the ON state in the fourth
sweep, indicating that the device was non-erasable. The long
retention ability of the ON state determines the ‘‘write once,
read many’’ (WORM) functionality of the ITO/TPF27OPI/Al
device.
In addition, other parameters such as stability and read
cycles are also important for the performance of the memory
device. Fig. 5(b) shows the effects of the operation time on the
memory device. At the constant voltage of 1 V, both the ON state
and the OFF state of the device were observed to have no
obvious degradation in current in 1 hour. The ON/OFF current
ratio is roughly as high as 104 at 1 V. As shown in Fig. 5(c), the
ON and OFF states were also stable up to 10 000 read pulses of
1 V (1 ms in period, 0.5 ms in duration width), indicating the
excellent stability of the memory device.
ITO/TPF27OPI/Al and ITO/TPF99OPI/Al devices, with Al as
the anode and ITO as the cathode, were investigated using
the current–voltage (I–V) characteristics.
For the ITO/TPF27OPI/Al device, the thicknesses of the
polyimide layer and the Al electrode were estimated to be
around 60 nm and 100 nm (Fig. S12–S14, ESI†), respectively.
Fig. 5(a) exhibits the typical I–V curves of the device (recorded
device units of 0.5 ꢁ 0.5 mm2) at a scan rate of 0.05 V sꢀ1. In the
first positive voltage sweep from 0 V to 5 V, the device was
initially in the OFF state (‘‘0’’ signal in memory data storage)
with a low current in the range of 10ꢀ14–10ꢀ9 A, until an abrupt
increase is observed in the current range at around 1.7 V with a
higher conductivity (10ꢀ5–10ꢀ4 A, ON state), which was defined
as the switching threshold voltage. This electronic transition
from the OFF state to the ON state in the first sweep serves as
Table 1 Optical and electrochemical properties of the polyimides
c
UV-vis (nm)
HOMOa (eV) LUMOb (eV) Eg (eV)
OX
onset
E
abs
Polyimides l
ledge (eV) Exp. Calc.d Exp. Calc.d Exp. Calc.d
max
TPF27OPI 340 377 1.48 ꢀ5.90 ꢀ5.42 ꢀ2.61 ꢀ2.51 3.29 2.91
TPF99OPI 333 362 1.54 ꢀ5.96 ꢀ5.44 ꢀ2.53 ꢀ2.54 3.43 2.90
However, the ITO/TPF99OPI/Al sandwich device exhibits
completely different switching behavior, as shown in Fig. 6(a).
During the first positive voltage sweep from 0 V to 5 V, the
a
The HOMO energy levels were determined from the CV measurement
OX
onset oxidation voltage (E
(4.8 eV below the vacuum level): HOMO = ꢀ[E
) using ferrocene as the external reference
onset
OX
ꢀ Eferrocene + 4.8] (eV).
onset
b
Eferrocene is determined to be 0.38 V vs. Ag/AgCl. Determined from the device is also initially at a low current, but a sharp increase in
equation: LUMO = Eg + HOMO. Estimated from the UV-vis absorption
current was observed at a switching threshold voltage around
c
edge wavelength (ledge) using the Planck equation: Eg = 1240/ledge
The data were obtained by quantum calculation of the basic units of
.
d
2.0 V, indicating the transition of the device from the OFF state
the PIs based on DFT/B3LYP/6-31G(d) using the Gaussian 09 program. to the ON state. In the subsequent second sweep from 5 V to 0 V,
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2017

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 5 (a) Current–voltage (I–V) characteristics of the ITO/TPF27OPI/Al sandwich device with an initial positive voltage sweep. (b) Effects of operation
time on the ON and OFF states of the memory devices under a constant 1 V bias voltage. (c) Effects of 1 V read pulse (1 ms in period, 0.5 ms in duration
width) on the ON and OFF states of the sandwich device.
the ON state could still be retained. However, the device relaxed of the synthesized polyimides were investigated by molecular
to the OFF state after the power was turned off for about simulations. The basic units (BUs) of TPF27OPI and TPF99OPI
3 minutes. In the third sweep from 0 V to ꢀ5 V, a sharp increase were selected as the model compounds, for the sake of model
in current was also observed at a switching threshold voltage of simplicity and practical availability for simulations.
about ꢀ2.0 V, indicating that the transition of the device from
Although XRD patterns (Fig. S15, ESI†) indicate that the two
the OFF state to the ON state occurs again. In the next sweep polyimides are amorphous, according to the simulation results,
(the fourth sweep from ꢀ5 V to 0 V), the ON state was also stable. the BU of TPF27OPI is more planar than that of TPF99OPI, as
It was exhibited that the ON and OFF states were randomly shown in Fig. 7. The twisted conformation of the ether bond on
accessible and volatile in the memory device, which was similar the dianhydride and the dihedral angle between the imide
to the data storage behavior of a static random access memory rings are similar. The most different dihedral angles between
(SRAM) functionality.
fluorene and the connected benzene rings of the BUs of two PIs
The stability of the device and the ON/OFF current ratio were for TPF27OPI are 371 and 411, but those for TPF99OPI are 911
measured as a function of the operation time, as shown in and 1451. These indicate that TPF99OPI is more twisted than
Fig. 6(b). No obvious degradation in current was observed for TPF27OPI in spatial structures. The rigid planar structure of
the OFF state and the ON state during the testing process, and TPF27OPI is more conducive to electron transport than the
the device maintained an ON/OFF current ratio as high as 105 at twisted structure of TPF99OPI.
1 V. Fig. 6(c) shows the stimulus effect of the read pulse on the
Fig. 8 shows the simulation results for the BUs of TPF27OPI
OFF state and the ON state. The PI-based memory device was and TPF99OPI based on DFT/B3LYP/6-31G(d) using the Gaussian
stable up to at 10 000 continuous read pulses of 1 V at least, and 09 program. The HOMOs of the model compounds are located on
the OFF state and the ON state were not changed significantly. the diamine side (as the donor, D), whereas the LUMOs are
located on the ODPA side (as the acceptor, A). These features will
Molecular simulations and the proposed electronic transition
mechanism
facilitate the charge transfer from the diamine side to the ODPA
side to form charge-transfer (CT) complexes in the polyimide
system under the applied voltage. For TPF27OPI, the calculated
HOMO and LUMO energies of the model compound are ꢀ5.42 eV
and ꢀ2.51 eV, and the Eg is 2.91 eV, which are basically consistent
with the experimental results obtained from the UV-vis and CV
To have a clear understanding of the charge transfer process
and the corresponding switching behavior of the present
polyimide-based memory devices, the electronic structure,
molecular orbitals, electrostatic potentials, and energy levels
Fig. 6 (a) Current–voltage (I–V) characteristics of the ITO/TPF99OPI/Al sandwich device with an initial positive voltage sweep. (b) Effects of operation
time on the ON and OFF states of the memory devices under a constant 1 V bias voltage. (c) Effects of 1 V read pulse (1 ms in period, 0.5 ms in duration
width) on the ON and OFF states of the sandwich device.
This journal is ©The Royal Society of Chemistry 2017
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
Table 2 The oscillator strengths (OS) and the assignment of S0 to Si
transitions in the BUs of PIsa
PI
Si
OS
Orbitals
Contribution
TPF27OPI
4
0.9577
HOMO - LUMO+2
HOMO - LUMO+3
HOMO - LUMO+3
HOMO - LUMO+4
HOMO - LUMO+3
HOMO - LUMO+4
HOMO - LUMO+5
0.91
0.09
0.14
0.03
0.05
0.53
0.10
9
0.2536
0.5186
10
TPF99OPI
1
2
8
0.0001
0.0005
0.1170
HOMO - LUMO
HOMO - LUMO+1
HOMO - LUMO
HOMO - LUMO+1
HOMO - LUMO+2
HOMO - LUMO+3
HOMO - LUMO+4
0.81
0.19
0.19
0.81
0.03
0.04
0.93
Fig. 7 The optimized geometries for the BUs of PIs.
a
Prohibited transitions are not listed.
electrons are more inclined to be excited from the HOMO
(ꢀ5.44 eV) to LUMO+4 (ꢀ1.27 eV) with an oscillator strength
of 0.1170 rather than the LUMO (ꢀ2.44 eV) with an oscillator
strength of 0.0001. The energy for electron transition of the BUs
of TPF27OPI (3.79 eV, from the HOMO to the LUMO+2) is less
than TPF99OPI (4.17 eV, from the HOMO to the LUMO+4),
which is consistent with the experimental results of the threshold
voltage (1.7 V o 2.0 V) of PI-based memory devices.
The energy level diagrams of the polyimides and the memory
devices are also shown in Fig. 9. Distinctly lower energy barriers
between the two electrodes (work function: ITO = ꢀ4.80 eV and
Al = 4.28 eV) and the HOMO energy levels of the polyimides
(TPF27OPI: ꢀ5.90 eV and TPF99OPI: ꢀ5.96 eV) are observed.
Theoretically, an applied electronic field exceeding 1.10 eV (the
barrier between the ITO electrode and the HOMO of TPF27OPI)
and 1.16 eV (the barrier between the ITO electrode and the
HOMO of TPF99OPI) will provide lower energy for hole injection
than the barrier between the Al electrode and the LUMO of the
PIs. This suggests that TPF27OPI and TPF99OPI function more
likely as p-type materials, and indicates that the hole injection
from the ITO electrode into the HOMO is more favorable than
the electron injection from the Al electrode into the LUMO
during the charge injection process.
Fig. 8 Calculated frontier molecular orbitals of the basic units of PIs.
measurements (HOMO: ꢀ5.90 eV, LUMO: ꢀ2.61 eV, Eg: 3.29 eV),
while for the TPF99OPI model, the HOMO energy is ꢀ5.44 eV and
the LUMO energy is ꢀ2.54 eV, respectively, and the Eg is 2.91 eV,
basically consistent with experimental results (HOMO: ꢀ5.96 eV,
LUMO: ꢀ2.53 eV, Eg: 3.43 eV). Both of them indicate the
rationality of the theoretical quantum calculations and the
corresponding experimental results.
In addition, to better illustrate the process from the OFF
state to the ON state, the I–V curves in both states are analyzed
Calculated oscillator strength (OS) and transition assignment
of the BUs of PIs are also summarized, as shown in Table 2,
which indicates the slight differences in the processes of the
electronic transition and might afford to explain the different
switching behaviors of the two PI-based memory devices. For
the TPF27OPI system, the HOMO (ꢀ5.42 eV) to the LUMO
(ꢀ2.51 eV) transition is almost impossible with an oscillator
strength of 0.0006, but LUMO+2 (0.9577), LUMO+3 (0.2536) and
LUMO+4 (0.5186) are more likely to form transitions with a
larger value of the oscillator strength. This indicates that the
accumulated electrons are more inclined to be excited to LUMO+2
(ꢀ1.63 eV) rather than the LUMO (ꢀ2.51 eV) under sufficient
energy. However, for the TPF99OPI system, the accumulated Fig. 9 Energy level diagrams of the PIs and the memory devices.
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2017

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 11 The plausible electronic transition processes occurring in TPF27OPI
and TPF99OPI.
Fig. 10 Experimental (dots) and fitted (solid line) I–V curves for (a) the OFF
state and (b) the ON state of the ITO/TPF27OPI/Al device.
LUMO+2, LUMO+3, and LUMO+4 orbitals tend to mainly locate
on the TPF27DA unit (donor, D), while the LUMO and the
LUMO+1 orbitals tend to distribute completely on the ODPA unit
(acceptor, A). The electronic transition corresponds to promotion
of the electrons from the ground state to the excited state. This
suggests1–3 that some electrons at the HOMO gain enough energy
and transit to the LUMO+2 orbital to give rise to an excited state
(combining the oscillator strength and their contribution), as the
applied bias reaches the threshold voltage. However, the excited
state which is highly energetic and unstable tends to relax to the
lowest excited state (the LUMO) with internal conversion. Thus
charge transfer can occur to form conductive charge-transfer (CT)
complexes. The electronic transition occurring in the TPF99OPI
system is similar to the higher LUMO+4 orbital through the
intermediate LUMO+3, LUMO+2, and LUMO+1 orbitals, and then
to the LUMO to form charge-transfer (CT) complexes. Once the
charge-transfer (CT) state is formed under the applied voltage, an
abrupt increase in current is observed and the device is transited
to the conductivity state (ON state).
in detail using theoretical models. The I–V curves of TPF27OPI
in the OFF state under a low positive voltage (0–1.6 V), as shown
in Fig. 10(a), follow the Schottky emission model,18 described
h
ꢀ
ꢁ
i
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
by the equation: J ¼ AT2 exp ꢀq fB ꢀ qV=4peid =kT . This
suggests that a Schottky barrier is formed at the ITO interface
due to the work function differences between the ITO electrode
(ꢀ4.80 eV) and the HOMO (ꢀ5.90 eV) of TPF27OPI. When the
applied electric field reaches the threshold voltage, electric
field-induced charge transfer occurs. Electrons transit from
the HOMO to the LUMO of TPF27OPI, leaving holes that spread
over the imide moieties. These holes give rise to an open
channel for charge carrier migration. Therefore, an abrupt
increase of the current in the device could be observed at the
threshold voltage, indicating the formation of a high conductivity
state (the ON state). During the subsequent positive sweeps of the
ON state, as shown in Fig. 10(b), the I–V curve of TPF27OPI
follows the Ohmic conduction model19 described by the equation:
J = qnmV/d. The relationship between current density ( J) and voltage
(V) is a linear correlation, which is similar to the conductive
mechanism of metal. The change from OFF state to ON state
indicates that the switching behavior of the device dependent
on the current through the TPF27OPI layer is switched on. Both
the Schottky emission model and the Ohmic conduction model
are suitable for the OFF state and the ON state of TPF99OPI, as
shown in Fig. S16 (ESI†).
Conclusions
In summary, we have successfully synthesized two novel tetra-
phenyl fluorene-based functional polyimides, TPF27OPI and
TPF99OPI. Both of the polyimides show excellent thermal properties
and obvious photoluminescence properties. TPF99OPI shows a
strong fluorescence with a blue shift compared to TPF27OPI.
The PI-based memory device, ITO/TPF27OPI/Al shows nonvolatile
Combining the calculated frontier molecular orbitals and WORM type switching behavior, while the ITO/TPF99OPI/Al one
the oscillator strengths from S0 to Si transitions of the simulation shows volatile SRAM type switching behavior. These results are
models, the possible electronic transition processes occurring in
the TPF27OPI and the TPF99OPI systems are described, as shown
in Fig. 11. For the TPF27OPI system, the HOMO and the
related to the fluorene structure located in the polymer main
chain and side chain, which affects the conjugation of polyimides.
In addition, molecular simulations, and the optical and electronic
This journal is ©The Royal Society of Chemistry 2017
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
chemical properties were used to clarify the CT processes and and a few drops of Aliquat336 were added. Then a few drops of
explain the resistive switching mechanisms. These results suggest the Pd(PPh3)4 catalyst were added and the reaction mixture was
that the memory switching behaviors of the PI-based devices can be stirred at 80 1C for 24 h under N2. After cooling to room temperature,
affected and adjusted by the chain structure and spatial structural the product was concentrated and purified by silica gel column
design of the polyimides.
chromatography using dichloromethane/n-hexane (v/v = 2/1). The
yield of the product is about 86% as a yellow powder.
1H NMR (400 MHz, DMSO-d6) d 7.90 (d, J = 8.0 Hz, 2H), 7.58
(d, J = 8.1 Hz, 2H), 7.51 (s, 2H), 7.27 (dt, J = 26.0, 8.1 Hz, 14H),
6.62 (d, J = 8.2 Hz, 4H), 5.24 (s, 4H). 13C NMR (101 MHz, DMSO-
d6) d 153.22, 150.14, 147.66, 141.90, 139.04, 130.24, 129.62,
129.22, 129.01, 128.43, 126.90, 124.41, 122.44, 116.14, 66.93.
FT-IR (KBr, cmꢀ1): 3462, 3371 (N–H, stretching), 1278 (C–N,
stretching). MS (m/z): [M]+ calc. for C37H28N2, 500.63. Found,
500.23.
Synthesis of 2,7-dibromo-9,9-bis(4-nitrophenyl)-9H-fluorene
(WuBrDN). The intermediate WuBrDN (3) was synthesized according
to a previously reported procedure.20 2,7-Dibromofluorene (2)
(16.21 g, 0.05 mol), 4-fluoronitrobenzene (14.14 g, 0.10 mol) and
potassium tertbutoxide (5.61 g, 0.01 mol) were dissolved in DMF
(100 mL) and stirred for 48 h under N2 at 120 1C. After reaction,
the product (3) was purified by silica gel column chromatography
using dichloromethane/n-hexane (v/v = 1/2). The yield of the
product was about 80% as a white powder.
Experimental
Materials
4,40-Oxydiphthalic dianhydride (ODPA), tetrabutylammonium
perchlorate (TBAP), hydrazine hydrate and ferrocene were
purchased from Aladdin. 2,7-Dibromo-9,9-diphenylfluororene,
2,7-dibromofluorene and 4-fluoronitrobenzene were purchased
from TCI. Palladium on activated carbon (10% Pd (Pd/C)) and
tetrakis(triphenylphosphine) palladium (Pd(PPh3)4) were purchased
from Alfa Aesar. Phenylboronic acid and 4-aminophenylboronic
acid were purchased from Soochiral Chemical Science & Technology
Co., Ltd. Dimethylformamide (DMF), dimethyl sulfoxide (DMSO),
1-methyl-2-pyrrolidinone (NMP), tetrahydrofuran (THF), and
methanol were obtained from Guangzhou Chemical Works. All
other reagents were of analytical grade and used as received from
commercial sources, unless otherwise mentioned.
1H NMR (400 MHz, DMSO-d6) d 8.19 (d, J = 8.6 Hz, 4H), 8.04
(d, J = 8.1 Hz, 2H), 7.73 (s, 4H), 7.45 (d, J = 8.7 Hz, 4H).
Synthesis of 9,9-bis(4-nitrophenyl)-2,7-diphenyl-9H-fluorene
(TPF99DN). Phenylboronic acid (6.10 g, 0.05 mol) and inter-
mediate compound (3) (11.12 g, 0.02 mol) were dissolved in 200 mL
of THF. Then 2 mol mLꢀ1 aqueous K2CO3 solution and a few drops
of Aliquat336 were added. Then a few drops of the Pd(PPh3)4 catalyst
were added and the reaction mixture was stirred at 80 1C for 24 h
under N2. After cooling to room temperature, the product (4) was
concentrated and purified by silica gel column chromatography
using dichloromethane/n-hexane (v/v = 1/4). The yield of the product
is about 72% as a light white powder.
Instrumentation
Cyclic voltammetry (CV) measurements were obtained using a
Bio-Logic VMP-300 multi-channel electro-chemical workstation
using a three-electrode cell at a rate of 50 mV sꢀ1 in CH3CN
solution (TBAP, 0.1 mol mLꢀ1). A PI thin film coated onto an
ITO glass substrate was used as the working electrode, Ag/AgCl
and KCl (sat.) as the reference electrode and a platinum disk as
the auxiliary electrode. The current–voltage (I–V) measurements
of the memory device were obtained using a Keithley 4200-SCS
semiconductor with a Keithley 4205-PG2 pulse generator. Wide-
angle X-ray diffraction (WAXD) measurements were performed
on a SmartLab X-ray diffractometer with a sweeping rate of
1H NMR (400 MHz, DMSO-d6) d 8.20 (d, J = 8.5 Hz, 4H), 8.16
(d, J = 7.9 Hz, 2H), 7.84 (d, J = 8.0 Hz, 2H), 7.80 (s, 2H), 7.68 (d,
J = 7.6 Hz, 4H), 7.57 (d, J = 8.6 Hz, 4H), 7.47 (t, J = 7.6 Hz, 4H),
7.37 (t, J = 7.4 Hz, 2H).
Synthesis of 4,40-(2,7-diphenyl-9H-fluorene-9,9-diyl)dianiline
(TPF99DA). A mixture of intermediate compound (4) (5.61 g,
0.01 mol), the 10% Pd/C catalyst (0.5 g), hydrazine monohydrate
(16 mL), and ethanol (200 mL) was placed in a three-necked flask
and heated at 80 1C for 24 h. The mixture was then filtered to
remove the Pd/C catalyst. After cooling to room temperature, the
precipitated crystals were isolated by filtration, recrystallized from
ethanol twice, ground into powder, and dried under vacuum. The
yield of the product is about 88% as a white powder.
101 minꢀ1 1H NMR and 13C NMR spectra were recorded on a
.
Bruker 400 MHz spectrometer. Infrared spectra (IR) were recorded
using a Bruker spectrometer. Mass spectra (MS) were recorded on
a Thermo EI mass spectrometer. Thermal gravimetric analysis
(TGA) was performed on a TA Q50 and differential scanning
calorimetry (DSC) was carried out with a NETZSCH DSC 204 at a
rate of 20 1C minꢀ1 under nitrogen. Ultraviolet-visible (UV-vis)
absorption spectra were recorded on a Hitachi U-3900. Fluorescence
spectra were recorded using a Shimadzu RF-5301PC spectrometer
with a slit width of 5 nm for polyimides and 3 nm for monomers,
respectively. The thickness of the vacuum evaporated Al electrodes
and the spin-coated PI thin films was measured using a Kosaka
Surfcorder ET150 and a Hitachi S-4800 field emission scanning
electron microscope (SEM).
1H NMR (400 MHz, DMSO-d6) d 7.99 (d, J = 7.9 Hz, 2H), 7.68
(d, J = 8.0 Hz, 2H), 7.62 (d, J = 7.7 Hz, 4H), 7.58 (s, 2H), 7.46 (t, J =
7.7 Hz, 4H), 7.36 (t, J = 7.3 Hz, 2H), 6.88 (d, J = 8.1 Hz, 4H), 6.45
(d, J = 8.1 Hz, 4H), 4.94 (s, 4H). 13C NMR (101 MHz, DMSO-d6) d
Synthesis of the monomers
Synthesis of 4,40-(9,9-diphenyl-9H-fluorene-2,7-diyl)dianiline 155.36, 148.96, 142.10, 141.32, 140.17, 134.51, 130.81, 130.13,
(TPF27DA). 4-Aminophenyl-boronic acid (2.74 g, 0.02 mol) and 129.18, 128.46, 127.74, 125.71, 122.71, 115.57, 65.73. FT-IR
2,7-dibromo-9,9-diphenylfluororene (1) (4.76 g, 0.01 mol) were (KBr, cmꢀ1): 3422, 3341 (N–H, stretching), 1285 (C–N, stretching).
dissolved in 200 mL of THF. Then 2 mol mLꢀ1 K2CO3 solution MS (m/z): [M]+ calc. for C37H28N2, 500.63. Found, 500.23.
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2017

View Article Online
Journal of Materials Chemistry C
Paper
Preparation of the polyimides
Notes and references
The synthesis of TPF27OPI was used as an example to illustrate
the general synthetic procedure, as shown in Scheme 2. To a
solution of TPF27DA (0.5006 g, 1 mmol) in 4 mL of freshly
distilled anhydrous DMF, the dianhydride ODPA (0.3102 g,
1 mmol) was added in one portion. The mixture was stirred
at 20 1C under argon for 24 h to form poly(amic acid) (PAA)
solution. The PAA solution was subsequently coated on a clean
dry glass plate, followed by thermal imidization at 80 1C/1 h,
160 1C/1 h, 240 1C/1 h and 350 1C/1 h in a vacuum. After
cooling, it was peeled with hot water. Due to the insolubility of
the PI, the degree of polymerization of the PAA was estimated
by inherent viscosity (Zinh). The Zinh value measured using an
Ostwald viscometer was 0.87 dL gꢀ1. FT-IR (air, cmꢀ1): 1780
(asym, CQO), 1720 (sym, CQO), 1367 (C–N) and 740 (imide ring).
TPF99OPI was prepared by a similar method to that used for
TPF27OPI. And the Ostwald viscosity of TPF99OPAA was
0.51 dL gꢀ1. FT-IR (air, cmꢀ1): 1782 (asym, CQO), 1724 (sym,
CQO), 1371 (C–N) and 742 (imide ring).
1 (a) Q. D. Ling, D. J. Liaw, C. X. Zhu, D. S. H. Chan, E. T. Kang
and K. G. Neoh, Prog. Polym. Sci., 2008, 33, 917–978;
(b) Q. D. Ling, F. C. Chang, Y. Song, C. X. Zhu, D. J. Liaw,
D. S. H. Chan, E. T. Kang and K. G. Neoh, J. Am. Chem. Soc.,
2006, 128, 8732–8733.
2 (a) A. R. Pease, J. O. Jeppesen, J. F. Stoddart, Y. Luo,
C. P. Collier and J. R. Heath, Acc. Chem. Res., 2001, 34,
433–444; (b) W. Y. Lee, T. Kurosawa, S. T. Lin, T. Higashihara,
M. Ueda and W. C. Chen, Chem. Mater., 2011, 23, 4487–4497.
3 (a) S. Song, B. Cho, T. W. Kim, Y. Ji, M. Jo, G. Wang, M. Choe,
Y. H. Kahng, H. Hwang and T. Lee, Adv. Mater., 2010, 22,
5048–5052; (b) B. M. Dhar, R. Ozgun, T. Dawidczyk,
A. Andreou and H. E. Katz, Mater. Sci. Eng., R, 2011, 72, 49–80.
4 (a) L. X. Wu, P. Wang, C. Y. Zhang, J. H. He, D. Y. Chen,
J. Jun, Q. F. Xu and J. M. Lu, Chem. – Asian J., 2016, 11,
102–111; (b) H. Zhuang, Q. H. Zhou, Q. J. Zhang, H. Li,
N. J. Li, Q. F. Xu and J. M. Lu, J. Mater. Chem. C, 2015, 3,
416–422; (c) Q. F. Gu, J. H. He, D. Y. Chen, H. L. Dong,
Y. Y. Li, H. Li, Q. F. Xu and J. M. Lu, Adv. Mater., 2015, 27,
5968–5973.
Fabrication and characterization of the memory devices
ITO glass substrates were cleaned by ultrasonication with
water, acetone, and ethanol each for 10 min, and then cleaned
by plasma for 10 min. Homogenous PAA solutions were prepared
in DMF and filtered. The PAA films were prepared by spin-
coating onto the ITO glass substrate at a speed rate of 2500 rpm
for 60 s. Then the PAA film was heated to 350 1C at a rate of
about 10 1C minꢀ1 under N2. The thickness of the smooth PI thin
films was 60 nm. Finally, an aluminum (Al) top electrode with a
thickness of 100 nm was deposited onto the polyimide film surface
through a shadow mask (each electrode at 0.5 ꢁ 0.5 mm2) by
thermal evaporation under vacuum. The sandwich memory devices
were eventually prepared with the ITO/PI/Al configuration.
5 (a) E. Y. H. Hong, C. T. Poon and V. W. W. Yam, J. Am. Chem.
Soc., 2016, 138, 6368–6371; (b) L. Liang, K. Li, C. Xiao,
S. J. Fan, J. Liu, W. S. Zhang, W. H. Xu, W. Tong,
J. Y. Liao, Y. Y. Zhou, B. J. Ye and Y. Xie, J. Am. Chem.
Soc., 2015, 137, 3102–3108; (c) N. D. Paul, U. Rana,
S. Goswami, T. K. Mondal and S. Goswami, J. Am. Chem.
Soc., 2012, 134, 6520–6523.
6 (a) S. Jung, W. Kwon, D. Wi, J. Kim, B. J. Ree, Y. Y. Kim,
W. J. Kim and M. Ree, Macromolecules, 2016, 49, 1369–1382;
(b) A. D. Yu, T. Kurosawa, Y. H. Chou, K. Aoyagi, Y. Shoji,
T. Higashihara, M. Ueda, C. L. Liu and W. C. Chen, ACS
Appl. Mater. Interfaces, 2013, 5, 4921–4929; (c) T. Kurosawa,
Y.-C. Lai, T. Higashihara, M. Ueda, C.-L. Liu and W.-C. Chen,
Macromolecules, 2012, 45, 4556–4563.
Molecular simulations
The molecular simulations, molecular geometry, molecular
orbitals and the BUs of polyimide were calculated and optimized
in the Gaussian 09 program package by means of density
functional theory (DFT) using Becke’s three-parameter hybrid
density functional method with Lee–Yang–Parr’s correction
functional (B3LYP) method, and the 6-31G(d) basis set. Time-
dependent DFT (TD-DFT) was additionally applied in oscillator
strength calculations and transition contributions.21
7 (a) Y. P. Hsiao, W. L. Yang, L. M. Lin, F. T. Chin, Y. H. Lin,
K. L. Yang and C. C. Wu, Microelectron. Reliab., 2015, 55,
2188–2197; (b) D. J. Choi, J. K. Kim, H. Seong, M. S. Jang and
Y. H. Kim, Org. Electron., 2015, 27, 65–71; (c) Y. Song,
H. Jeong, J. Jang, T. Y. Kim, D. Yoo, Y. Kim, H. Jeong and
T. Lee, ACS Nano, 2015, 9, 7697–7703.
8 (a) Y. P. Hsiao, W. L. Yang, C. C. Wu, L. M. Lin, F. T. Chin,
Y. H. Lin and K. L. Yang, Jpn. J. Appl. Phys., 2016, 55, 01AA09;
(b) Z. X. Zhou, L. J. Qu, T. T. Yang, J. L. Wen, Y. Zhang,
Z. G. Chi, S. W. Liu, X. D. Chen and J. R. Xu, RSC Adv., 2016,
6, 52798–52809.
Acknowledgements
9 (a) F. Pan, S. Gao, C. Chen, C. Song and F. Zeng, Mater. Sci.
Eng., R, 2014, 83, 1–59; (b) A. Bousquet, H. Awada,
R. C. Hiorns, C. Dagron-Lartigau and L. Billon, Prog. Polym.
Sci., 2014, 39, 1847–1877.
10 (a) T. Kurosawa, A. D. Yu, T. Higashihara, W. C. Chen and
M. Ueda, Eur. Polym. J., 2013, 49, 3377–3386; (b) C. J. Chen,
Y. C. Hu and G. S. Liou, Chem. Commun., 2013, 49, 2536–2538;
(c) C. J. Chen, H. J. Yen, Y. C. Hu and G. S. Liou, J. Mater.
Chem. C, 2013, 1, 7623–7634; (d) B. J. Ree, W. Kwon, K. Kim,
The financial support from the National 973 Program of China
(No. 2014CB643605), the National Natural Science Foundation
of China (No. 51373204, 51173214, 51233008), the National 863
Program of China (No. 2015AA033408), the Science and Technology
Project of Guangdong Province (No. 2015B090915003 and
2015B090913003), the Leading Scientific, Technical and
Innovation Talents of Guangdong Special Support Program
(No. 2016TX03C295), and the Fundamental Research Funds for
the Central Universities (No. 161gzd08) is gratefully acknowledged.
This journal is ©The Royal Society of Chemistry 2017
J. Mater. Chem. C

View Article Online
Paper
Journal of Materials Chemistry C
Y. G. Ko, Y. Y. Kim, H. Lee and M. Ree, ACS Appl. Mater. 17 (a) Y. W. Liu, Y. Zhang, X. H. Wu, Q. Lan, C. S. Chen,
Interfaces, 2014, 6, 21692–21701.
S. W. Liu, Z. G. Chi, L. Jiang, X. D. Chen and J. R. Xu, J. Mater.
Chem. C, 2014, 2, 1068–1075; (b) Y. W. Liu, Z. X. Zhou,
L. J. Qu, Z. Q. Chen, Y. Zhang, S. W. Liu, Z. G. Chi, X. D. Chen
and J. R. Xu, Mater. Chem. Front., 2017, 1, 326–337.
11 (a) B. J. Ree, W. Kwon, K. Kim, Y. G. Ko, Y. Y. Kim, H. Lee and
M. Ree, ACS Appl. Mater. Interfaces, 2014, 6, 21692–21701;
(b) L. Shi, H. Ye, W. Liu, G. Tian, S. Qi and D. Wu, J. Mater.
Chem. C, 2013, 1, 7387; (c) J. Jang, W. Park, C. A. Kyungjune, 18 A. J. Campbell, D. D. C. Bradley and D. G. Lidzey, J. Appl.
S. B. Hyunwook and T. Lee, Curr. Appl. Phys., 2013, 13,
1237–1240.
12 (a) G. F. Tian, D. Z. Wu, L. Shi, S. L. Qi and Z. P. Wu, RSC
Phys., 1997, 82, 6326–6342.
19 S. Menzel, U. Bottger, M. Wimmer and M. Salinga, Adv.
Funct. Mater., 2015, 25, 6306–6325.
Adv., 2012, 2, 9846–9850; (b) B. L. Hu, F. Zhuge, X. J. Zhu, 20 Y. W. Liu, Z. X. Zhou, L. J. Qu, B. Zou, Z. Q. Chen, Y. Zhang,
S. S. Peng, X. X. Chen, L. Pan, Q. Yan and R. W. Li, J. Mater.
Chem., 2012, 22, 520–526; (c) A. D. Yu, T. Kurosawa,
S. W. Liu, Z. G. Chi, L. Jiang, X. D. Chen and J. R. Xu, Mater.
Chem. Front., 2017, 1, 326–337.
Y. C. Lai, T. Higashihara, M. Ueda, C. L. Liu and 21 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
W. C. Chen, J. Mater. Chem., 2012, 22, 20754–20763.
13 (a) Y. C. Hu, C. J. Chen, H. J. Yen, K. Y. Lin, J. M. Yeh, W. C. Chen
and G. S. Liou, J. Mater. Chem., 2012, 22, 20394–20402;
(b) C. L. Liu, T. Kurosawa, A. D. Yu, T. Higashihara, M. Ueda
and W. C. Chen, J. Phys. Chem. C, 2011, 115, 5930–5939;
(c) G. F. Tian, D. Z. Wu, S. L. Qi, Z. P. Wu and X. D. Wang,
Macromol. Rapid Commun., 2011, 32, 384–389.
14 (a) H. J. Yen, C. J. Chen and G. S. Liou, Adv. Funct. Mater., 2013,
23, 5307–5316; (b) Y. W. Liu, Y. Zhang, Q. Lan, S. W. Liu,
Z. X. Qin, L. H. Chen, C. Y. Zhao, Z. G. Chi, J. R. Xu and
J. Economy, Chem. Mater., 2012, 24, 1212–1222; (c) L. Shi,
G. F. Tian, H. B. Ye, S. L. Qi and D. Z. Wu, Polymer, 2014, 55,
1150–1159.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,
V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari,
A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega,
J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken,
C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,
15 T. Kurosawa, T. Higashihara and M. Ueda, Polym. Chem.,
2013, 4, 16–30.
¨
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and
16 C. L. Tan, Z. D. Liu, W. Huang and H. Zhang, Chem. Soc.
Rev., 2015, 44, 2615–2628.
D. J. Fox, Gaussian 09, Revision E.01, Gaussian, Inc., Wallingford
CT, 2009.
J. Mater. Chem. C
This journal is ©The Royal Society of Chemistry 2017