New random copolymers with pendant carbazole donor and 1,3,4-oxadiazole acceptor for high performance memory device applications

Article abstract of DOI:10.1039/c0jm02535f

New non-conjugated random copolymers containing pendent electron-donating 9-(4-vinylphenyl)carbazole (VPK) and electron-accepting 2-phenyl-5-(4- vinylphenyl)-1,3,4-oxadiazole (OXD) or 2-(4-vinylbiphenyl)-5-(4-phenyl)-1,3,4- oxadiazole (BOXD) were successf

Full text of DOI:10.1039/c0jm02535f

As featured in:
Showcasing research from Prof. Wen-Chang Chen’s group
at Department of Chemical Engineering and Institute
of Polymer Science and Engineering, National Taiwan
University, Taiwan, and Dr. Cheng-Liang Liu at Department
of Organic Device Engineering, Yamagata University, Japan
Title: New random copolymers with pendant carbazole donor
and 1,3,4-oxadiazole acceptor for high performance memory
device applications
Non-conjugated random copolymers with pendant carbazole donor
and oxadiazole acceptor exhibit electrical characteristics of diode,
volatile memory and insulator depending on donor/acceptor ratio and
acceptor strength. The excellent memory effects make them promising
candidates for low-cost reproducible SRAM memory devices.
See Yi-Kai Fang, Cheng-Liang Liu and
Wen-Chang Chen, J. Mater.
Chem., 2011, 21, 4778.
www.rsc.org/materials
Registered Charity Number 207890

Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2011, 21, 4778
www.rsc.org/materials
PAPER
New random copolymers with pendant carbazole donor and 1,3,4-oxadiazole
acceptor for high performance memory device applications†
a
Yi-Kai Fang,‡ Cheng-Liang Liu‡ and Wen-Chang Chen
b
ac
*
Received 4th August 2010, Accepted 16th December 2010
DOI: 10.1039/c0jm02535f
New non-conjugated random copolymers containing pendent electron-donating 9-(4-
vinylphenyl)carbazole (VPK) and electron-accepting 2-phenyl-5-(4-vinylphenyl)-1,3,4-oxadiazole
(OXD) or 2-(4-vinylbiphenyl)-5-(4-phenyl)-1,3,4-oxadiazole (BOXD) were successfully synthesized by
nitroxide-mediated free radical polymerization (NMRP) method. The prepared random copolymers
are denoted as P(VPKxOXDy) or P(VPKxBOXDy) with three different electron donor/acceptor (x/y)
ratios of 8/2, 5/5, and 2/8. The electrical switching behavior based on the ITO/polymer/Al device
configuration could be tuned through the donor/acceptor ratio or acceptor trapping ability. Both
experimental and theoretical results indicated that the charge transfer between the pendant donor and
acceptor was relatively weak without significant orbital hybridization. In addition, the low-lying
HOMO energy level of OXD or BOXD units as compared to VPK units created the trapping
environment. Therefore, distinct electrical current–voltage (I–V) characteristics changed between the
diode, the volatile memory, and the insulator depending on the relative donor/acceptor ratios of 10/0, 8/
2, and (5/5, 2/8 and 0/10), respectively. The memory device based on P(VPK8OXD2) or
P(VPK8BOXD2) copolymers exhibited volatile static random access memory (SRAM) behavior with
an ON/OFF current ratio of approximately 104–105, up to 107 read pulses, and retention time of more
than 1 h. The unstable ON state in the device was due to the shallow trapped holes with spontaneously
back transferring of charge carriers when the electric field was removed and thus exhibited a volatile
nature. The slightly lower HOMO level of OXD moieties than that of BOXD led to the P(VPK8OXD2)
device storing the charge for a longer period of time. The present study suggested the high performance
polymer memory devices could be achieved by changing the donor/acceptor ratio or chemical structure.
polymer-based memory devices with electrically bistable
Introduction
behavior attracted significant scientific interest recently due to
the advantages of rich structure flexibility, low cost, solution
processability, and three-dimensional stacking capability.19–22
The reported D–A polymeric materials for memory device
applications included conjugated polymers,23–26 non-conjugated
polymers with donor/acceptor chromophores (main chain poly-
mers27–29 or pendent polymers30–35) and polymer nanocomposites
(metal nanoparticle,36,37 fullerene,38 carbon nanotube39 or gra-
phene oxide40 embedded). The reported polymer memory types
included non-volatile memory (write-once-read-many times
(WORM) memory,26,33–35,39 flash (rewritable) memory25,28,30–34,37–40
or negative differential resistance (NDR) behavior36,41) and
volatile memory (dynamic random access memory
(DRAM)24,27,29 or static random access memory (SRAM)29). The
memory devices with two different states of conductance were
realized through several proposed mechanisms such as charge
Donor–acceptor (D–A) type polymers have been widely explored
for electronic and optoelectronic device applications, including
light emitting diodes,1–3 photovoltaic cells,4–10 field-effect tran-
sistors,10–18 and polymer memory devices.19–41 In particular, D–A
aInstitute of Polymer Science and Engineering, National Taiwan
University, Taipei, Taiwan, 106. E-mail: chenwc@ntu.edu.tw; Fax:
+886-2-3366-5237; Tel: +886-2-3366-5236
bDepartment of Organic Device Engineering, Yamagata University,
Yonezawa, Yamagata, 992-8510, Japan
cDepartment of Chemical Engineering, National Taiwan University,
Taipei, Taiwan, 106
† Electronic supplementary information (ESI) available: 1H NMR
spectra of P(VPK5BOXD5), P(VPK2BOXD8), P(VPK8OXD2),
P(VPK5OXD5) and P(VPK2OXD8) in CD2Cl2. DSC and TGA
thermograms of all homopolymers and copolymers. Retention time test
on the ON and OFF states of the ITO/P(VPK8OXD2)/Al device under
a continuous readout voltage. Stimulus effect of read pulses on the ON
and OFF states of the ITO/P(VPK8OXD2)/Al device. See DOI:
10.1039/c0jm02535f
transfer
effect,7–33,35–38,40
trapping/de-trapping
of
charges,24,26,34,39,41 and filamentary conduction.25 It was believed
that the polymer structure and morphology significantly affected
the origin of the switching mode.
‡ Y. K. Fang and C. L. Liu equally contribute to this work.
4778 | J. Mater. Chem., 2011, 21, 4778–4786
This journal is ª The Royal Society of Chemistry 2011

Electronic resistive-type memory devices using organic poly-
mers as active elements were shown to efficiently store the data
based on the high- and low-conductance response to an applied
voltage. Non-conjugated polymers with specific pendent chro-
mospheres were widely used in the fabrications of memory device
to obtain the switching characteristics.30–35,42–48 Polymers with
side chain carbazole-containing units exhibited hole-transporting
ability that dominated the conduction. Conformation-induced
electrical bistability in the pendant polymers with carbazole units
exhibited the WORM and SRAM switching effect.44,45 The
carbazole moieties as electron donors were expected to enhance
charge transfer ability with suitable strong acceptors such as
organic acceptors,31,35,43 Eu complex,30,31 or C60.32 Therefore, the
devices showed two distinguished conductivity states through the
formation of the charge transfer complex when the voltage bias
was applied. However, the effects of pendant donor/acceptor
ratios or chemical structure on the non-conjugated polymer
based memory devices have not been explored yet, to the best of
our knowledge.
Experimental section
Materials
4-Vinylphenylboronic acid, sodium carbonate and Pd(PPh3)4
were purchased from Aldrich (Milwaukee, WI), Showa (Tokyo,
Japan), and Strem Chemicals (Newburyport, MA), respectively.
Benzoyl peroxide (BPO) and 2,2,6,6-tetramethylpiperidine-1-
oxyl (TEMPO) was obtained from Acros (Geel, Begium).
Anhydrous toluene and dimethyl formamide (DMF) were
purchased from TEDIA (Fairfield, CT). All the chemicals were
used as received and without further purification. Tetrabuty-
lammonium perchlorate (TBAP, TCI) was recrystallized twice
from ethyl acetate and then dried under vacuum prior to use. The
monomers of 9-(4-vinylphenyl)carbazole (VPK) and 2-phenyl-5-
(4-vinylphenyl)-1,3,4-oxadiazole (OXD) were synthesized
according to the literature method.49 The synthesis routes of 2-(4-
vinylbiphenyl)-5-(4-phenyl)-1,3,4-oxadiazole (BOXD) monomer
and PVPK, POXD and PBOXD homopolymers are described in
the ESI†.
In this work, we report the synthesis and memory device
characteristics of new non-conjugated random copolymers with
pendant electron-donating 9-(4-vinylphenyl)carbazole (VPK)
and electron-withdrawing 2-phenyl-5-(4-vinylphenyl)-1,3,4-oxa-
diazole (OXD) or 2-(4-vinylbiphenyl)-5-(4-phenyl)-1,3,4-oxa-
diazole (BOXD) moieties. The random copolymers are denoted
as P(VPKxOXDy) and P(VPKxBOXDy), in which the x/y is the
donor/acceptor ratio as shown in Scheme 1. Copolymers with
three different donor/acceptor (x/y) ratios of 8/2, 5/5, and 2/8
were synthesized through the nitroxide-mediated free radical
polymerization (NMRP) method. The thermal, optical, and
electrochemical properties of the prepared copolymers were
characterized. The low-lying HOMO energy level of OXD or
BOXD electron acceptors can block the transport of charge
carriers. The ITO/polymer/Al sandwiched device configuration
was employed to explore the energy levels on the memory
characteristics and the volatility of the memory effect was
determined by the ability of charge trapping/back transferring of
trapped charges. The experimental results suggested that the
electrical switching could be efficiently tuned by donor/acceptor
ratio or acceptor ability in the prepared copolymers as compared
to the homopolymers.
General polymerization procedure of random copolymers
The random copolymers, poly(9-(4-vinylphenyl)carbazole-g-2-
phenyl-5-(4-vinylphenyl)-1,3,4-oxadiazole) P(VPKxOXDy) and poly
(9-(4-vinylphenyl)-carbazole-g-2-(4-vinylbiphenyl)-5-(4-phenyl)-
1,3,4-oxadiazole) P(VPKxBOXDy) with three different VPK/
OXD or BOXD ratios (8/2, 5/5, 2/8), were synthesized via
nitroxide-mediated free radical polymerization (NMRP) and
further purified by Soxhlet extractor in acetone, as illustrated in
Scheme 1. The copolymers were finally dried under vacuum at 60
ꢀC overnight. The weight-average molecular weights (Mw) of the
copolymers, obtained by GPC using THF as the eluent and
polystyrene as standards, were 12 300–22 300 g molꢁ1 with the
polydispersity index (PDI) of 1.19–1.43, and the relative ratio of
electron donor/acceptor segments in the random copolymers
were estimated from elemental analysis.
P(VPK8OXD2). VPK (539 mg, 2.0 mmol), 124 mg of OXD (0.5
mmol), 6.0 mg of BPO (0.025 mmol), 5.1 mg of TEMPO (0.0325
mmol), and DMF (1 mL) were used to afford 273 mg yellow solid
(41.2%). 1H NMR (300 MHz, CDCl3) d 1.99–1.49 (br, 3H,
–CH2–CH–), 6.33–8.17 (br, 21H, Ar–H); anal. calcd. for
[(C20H15N)0.8+(C16H12N2O)0.2]: C 86.90, H 5.43, N 6.34; found
C 84.87, H 5.46, N 6.21. Weight-average molecular weight (Mw)
and polydispersity index (PDI) estimated from gel permeation
chromatographic (GPC) are 12 300 and 1.19, respectively.
P(VPK5OXD5). VPK (337 mg, 1.25 mmol), 310 mg of OXD
(1.25 mmol), 6.0 mg of BPO (0.025 mmol), 5.1 mg of TEMPO
(0.0325 mmol), and DMF (1 mL) were used to afford 325 mg
white solid (50.2%). 1H NMR (300 MHz, CDCl3) d 1.35–2.08 (br,
3H, –CH2–CH–), 6.37–8.21 (br, 21H, Ar–H); anal. calcd. for
[(C20H15N)0.5+(C16H12N2O)0.5]: C 83.46, H 5.22, N 8.11; found
C 82.14, H 5.70, N 8.68. Mw and PDI estimated from GPC are
12 400 and 1.21, respectively.
P(VPK2OXD8). VPK (135 mg, 0.5 mmol), 310 mg of OXD (2.0
mmol), 6.0 mg of BPO (0.025 mmol), 5.1 mg of TEMPO (0.0325
mmol), and DMF (1 mL) were used to afford 382 mg white solid
(60.5%). 1H NMR (300 MHz, CDCl3) d 1.31–2.09 (br, 3H,
–CH2–CH–), 6.30–8.24 (br, 21H, Ar–H); anal. calcd. for
[(C20H15N)0.2+(C16H12N2O)0.8]: C 79.84, H 4.99, N 9.98; found:
Scheme 1 Synthesis of P(VPKxOXDy) and P(VPKxBOXDy) random
copolymers.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 4778–4786 | 4779

C 78.79, H 5.27, N 10.05. Mw and PDI estimated from GPC are
12 600 and 1.30, respectively.
Before the fabrication of the polymer layer, the glass was pre-
cleaned by ultrasonication with water, isopropanol, and acetone
each for 15 min, respectively. Then, 20 mg mLꢁ1 of polymer
solution in chlorobenzene was first filtered through 0.45 mm pore
size of a PTFE membrane syringe filter. The filtered solution was
spin-coated onto the pre-cleaned ITO glass at a speed rate of
1000 rpm for 60 s and annealed at 100 ꢀC for 10 min under
vacuum. The polymer film thickness was determined to be
around 70 nm. Finally, a 300 nm thick Al top electrode (recorded
device units of 0.5 ꢂ 0.5 mm2 in size) was thermally evaporated
through the shadow mask at a pressure of 10ꢁ7 torr with
P(VPK8BOXD2). VPK (539 mg, 2.0 mmol), 162 mg of OXD
(0.5 mmol), 6.0 mg of BPO (0.025 mmol), 5.1 mg of TEMPO
(0.0325 mmol), and DMF (1 mL) were used to afford 219 mg
1
yellow solid (31.3%). H NMR (300 MHz, CDCl3) d 1.38–2.08
(br, 3H, –CH2–CH–), 6.39–8.21 (br, 25H, Ar–H); anal. calcd. for
[(C20H15N)0.8+(C22H16N2O)0.2]: C 87.32, H 5.42, N 5.99; found:
C 86.09, H 5.60, N 6.18. Mw and PDI estimated from GPC are
21 100 and 1.43, respectively.
P(VPK5BOXD5). VPK (337 mg, 1.25 mmol), 405 mg of OXD
(1.25 mmol), 6.0 mg of BPO (0.025 mmol), 5.1 mg of TEMPO
(0.0325 mmol), and DMF (1 mL) were used to afford 299 mg
a uniform depositing rate of 3–5 A sꢁ1. The electrical charac-
ꢀ
terization of the memory device was performed by a Keithley
4200-SCS semiconductor parameter analyzer equipped with
a Keithely 4205-PG2 arbitrary waveform pulse generator. ITO
was used as the anode and Al was set as the cathode (maintained
as common) during the voltage sweep with a step of 0.1 V. The
probe tip used 10 mm diameter tungsten wire attached to a tinned
copper shaft with a point radius < 0.1 mm (GGB Industries, Inc.).
All of the electronic measurements were performed in a glove
box.
1
white solid (40.3%). H NMR (300 MHz, CDCl3) d 1.38–2.03
(3H, –CH2–CH–), 6.41–8.22 (br, 25H, Ar–H); anal. calcd. for
[(C20H15N)0.5+(C22H16N2O)0.5]: C 84.89, H 5.22, N 7.07; found:
C 83.91, H 5.35, N 7.28. Mw and PDI estimated from GPC are
22 300 and 1.40, respectively.
P(VPK2BOXD8). VPK (135 mg, 0.5 mmol), 657 mg of OXD
(2.0 mmol), 6.0 mg of BPO (0.025 mmol), 5.1 mg of TEMPO
(0.0325 mmol), and DMF (1 mL) were used to afford 390 mg
1
white solid (49.3%). H NMR (300 MHz, CDCl3) d 1.35–2.06
(3H, –CH2–CH–), 6.35–8.18 (br, 25H, Ar–H). Mw and PDI
Computational methodology
estimated from GPC are 16 800 and 1.30, respectively.
Molecular calculations studied in this work have been performed
with Gaussian 03 program package.50 Equilibrium ground state
geometry and electronic properties were optimized by means of
the density functional theory (DFT) method at the B3LYP level
of theory (Beckes-style three-parameter density functional theory
using the Lee–Yang–Parr correlation functional) with the 6-
31G(d) basic set.
Characterization
1H NMR spectra were measured on a Bruker Avance 300 MHz
FT-NMR and Bruker Avance DRX 500 MHz FT-NMR spec-
trometer. Gel permeation chromatographic (GPC) analysis was
performed on a Lab Alliance RI2000 instrument (one column,
MIXED-D from Polymer Laboratories) connected with one
refractive index detector from Schambeck SFD Gmbh. All GPC
analyses were performed on polymer/THF solution at a flow rate
of 1 mL minꢁ1 at 40 ꢀC and calibrated with polystyrene stan-
dards. Elemental analyses were performed with a Heraeus Var-
ioEL-III-NCSH instrument.
Results and discussion
Polymer structure
The chemical structures of the synthesized homopolymers and
1
random copolymers were confirmed by H NMR and elemental
1
analysis. Fig. 1 shows the H NMR spectra of PVPK, PBOXD
Thermogravimetric analysis (TGA) was conducted with
a PerkinElmer Pyris 1 TGA at a heating rate of 20 ꢀC minꢁ1
.
and P(VPK8BOXD2) in CD2Cl2, respectively. P(VPK8BOXD2)
random copolymer has vinyl protons (1.3 to 2.1 ppm) and
aromatic protons (6.3 to 8.3 ppm), similar to those of the
Differential scanning calorimetry (DSC) measurements were
performed under a nitrogꢀen atmosphere at a heating rate of 20 ꢀC
minꢁ1 from ꢁ50 to 250 C using a TA instrument DSC-Q100.
Electrochemistry was performed with a CHI 611B electro-
chemical analyzer. Voltammograms are presented with the
positive potential pointing to the left and with increasing anodic
currents pointing downwards. Cyclic voltammetry was per-
formed by a three-electrode cell in which ITO (polymer films area
about 0.7 ꢂ 0.5 cm2) was used as a working electrode. A platinum
wire was used as an auxiliary electrode. All cell potentials were
taken with the use of a homemade Ag/AgCl, KCl (sat.) reference
electrode. Absorption spectra were measured with a Hitachi
U4100 UV–Vis–NIR spectrophotometer. The thickness of the
polymer film was measured with a Microfigure Measuring
Instrument (Surfcorder ET3000, Kosaka Laboratory Ltd).
1
homopolymers, PVPK and PBOXD. The H NMR spectra of
the other copolymers also show similar peak signals and are
consistent with the proposed structure, as shown in Fig. S1 and
S2 of the ESI†. Due to the mixed proton signals of the random
copolymers, it is difficult to estimate the relative ratio of donor
(VPK)/acceptor (OXD or BOXD) from the peak integration of
NMR spectra. Therefore, the pendent donor/acceptor ratios of
the prepared copolymers are further confirmed by elemental
analysis. As described in the synthesis part of the experimental
section, the carbon, hydrogen, and nitrogen experimental
contents of the synthesized copolymers are all in a good agree-
ment with the theoretical contents. It suggests that the pendent
donor/acceptor ratios of the synthesized random copolymers are
close to the original design. The weight-average molecular
weights (Mw) of the copolymers are from 12 300 to 22 300 with
the polydispersity index (PDI) in the range of 1.19–1.43, deter-
mined by GPC analysis with polystyrene as standards. All the
polymers were readily soluble in several common organic
Fabrication and characterization of polymer memory devices
The memory devices were fabricated on the indium-tin oxide
(ITO) coated glass, with the configuration of ITO/polymer/Al.
4780 | J. Mater. Chem., 2011, 21, 4778–4786
This journal is ª The Royal Society of Chemistry 2011

Fig.
1
1H NMR spectra of (a) PVPK, (b) PBOXD and (c)
P(VPK8BOXD2) in CD2Cl2.
solvents such as tetrahydrofuran (THF), chlorobenzene and
chloroform. Thus, uniform thin films could be obtained by spin-
coating for electronic device applications.
Fig. 2 Normalized UV–Vis absorption spectra of the polymer films on
quartz plate: (a) PVPK, POXD and P(VPKxOXDy) and (b) PVPK,
PBOXD and P(VPKxBOXDy).
Optical properties
Electrochemical properties
UV–Vis absorption spectra of the synthesized polymer films are
abs
shown in Fig. 2 and the corresponding absorption maxima (l
)
The electrochemical properties of the polymers were investigated
by cyclic voltammetry (CV), as shown in Fig. 3 and summarized
in Table 1. The oxidation and reduction behaviors of the
prepared film were measured on an ITO-coated glass substrate in
dry acetonitrile or DMF containing 0.1 M of TBAP under
nitrogen. The HOMO and LUMO energy levels of the random
max
in thin film state are summarized in Table 1. The PVPK film
shows three strong absorption bands at 294, 330 and 342 nm,
which is consistent with spectral features of the poly(N-vinyl-
carbazole) (PVK) film.32 The spectra of POXD or PBOXD film
has one main absorption peak at 293 or 311 nm, respectively, due
to the p–p* transition of the aromatic ring. The slightly red-
shifted spectrum of PBOXD compared to that of POXD is
probably due to the additional conjugated phenylene ring of the
former. For the synthesized random copolymers of
P(VPKxOXDy) (or P(VPKxBOXDy)), the absorption band
reveals a superposition rule of PVPK and POXD (or PBOXD)
spectra. For example, the maximum peak of P(VPK8BOXD2)
shows two bands centered at around 294 and 311 nm, which are
mainly contributed to from the p–p* transition of PVPK and
PBOXD, respectively. The optical band gaps of PVPK, POXD,
PBOXD, P(VPKxOXDy), and P(VPKxBOXDy) estimated from
the absorption edges are 3.48, 3.68 and 3.40, 3.44–3.52, and 3.42–
3.47 eV, respectively. Besides, the copolymers with different
pendent donor/acceptor ratios exhibit no obvious significant
shift to a lower energy transition. It suggests that the intra-
molecular charge transfer between the pendent carbazole donor
and the 1,3,4-oxadiazole acceptor is relatively weak.
copolymers were estimated from CV with reference to ferrocene
onset
(4.8 eV) by the following equations: LUMO ¼ (E
vs. Ag/AgCl
red
onset
red
+ 4.8 eV ꢁ E1/2, ferrocene); HOMO ¼ (E
vs. Ag/AgCl + 4.8
eV ꢁ E1/2, ferrocene). In the CV, the PVPK and POXD (or
PBOXD) homopolymer only show quasi-reversible anodic and
cathodic peaks at 1.23 and ꢁ1.89 V (or ꢁ1.74 V), respectively,
which corresponds to the HOMO level of ꢁ5.55 eV and LUMO
level of ꢁ6.08 eV (or ꢁ5.95 eV), respectively. The cathodic peak
is not detected in the CV of PVPK and thus the LUMO level of
ꢁ2.07 eV is determined from the difference between the HOMO
level and the optical band gap. Similar estimation is used to
determine the HOMO levels of POXD and PBOXD, which are
ꢁ6.08 eV and ꢁ5.95 eV, respectively. The LUMO energy level of
PBOXD (ꢁ2.55 eV) is lower than that of POXD (ꢁ2.40 eV),
which is attributed to the incorporation of one more conjugated
phenyl moieties. The CV of P(VPKxOXDy) or P(VPKxBOXDy)
random copolymers show both p-doping and n-doping behavior.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 4778–4786 | 4781

Table 1 Thermal, optical and electrochemical properties of the synthesized homopolymers and random copolymers
E/V (vs. Ag/AgCl
in MeCN or
DMF)c
film
lmax
Index
Tg (ꢀC)
Td (ꢀC)a
Egb (eV)
Eox
Ered
HOMOd (eV)
LUMOd (eV)
PVPK
POXD
PBOXD
P(VPK8OXD2)
P(VPK5OXD5)
P(VPK2OXD8)
P(VPK8BOXD2)
P(VPK5BOXD5)
P(VPK2BOXD8)
206
173
192
193
187
177
202
197
193
405
367
381
376
373
379
386
404
406
294, 330, 342
293
311
295, 342
294
292
294, 311
311
311
3.48
3.68
3.40
3.44
3.48
3.52
3.42
3.44
3.47
1.23
NA
NA
1.24
1.25
1.22
1.23
1.25
1.22
NA
ꢁ5.55
ꢁ6.08
ꢁ5.95
ꢁ5.56
ꢁ5.57
ꢁ5.54
ꢁ5.55
ꢁ5.57
ꢁ5.54
ꢁ2.07
ꢁ2.40
ꢁ2.55
ꢁ2.41
ꢁ2.38
ꢁ2.39
ꢁ2.57
ꢁ2.54
ꢁ2.53
ꢁ1.89
ꢁ1.74
ꢁ1.88
ꢁ1.91
ꢁ1.90
ꢁ1.72
ꢁ1.75
ꢁ1.76
a
Thermal decomposition temperature (5% weight-loss). b The data were calculated by the equation: gap ¼ 1240/lonset of monomer and polymer film.
c
vs. Ag/AgCl in CH3CN (oxidation), vs. Ag/AgCl in DMF (reduction). d The HOMO energy levels were calculated from cyclic voltammetry and were
referenced to ferrocene (4.8 eV).
The polymer memory devices store data based on the high- (ON)
and low-conductivity (OFF) response to the external applied
voltages. Fig. 4 shows the typical I–V characteristics of the
memory devices fabricated with P(VPK8BOXD2) and
Fig. 3 Cyclic voltammograms of the homopolymer and copolymer thin
films spin-coated onto an indium-tin oxide (ITO)-coated glass substrate
in CH3CN (oxidation) and DMF (reduction) containing 0.1 M TBAP.
Scan rate ¼ 0.1 V sꢁ1. The inset shows HOMO and LUMO energy level of
PVPK, POXD, PBOXD, P(VPKxOXDy) and P(VPKxBOXDy) and work
function of Al and ITO.
The HOMO energy levels of the copolymers with different
donor/acceptor ratio are almost located at ꢁ5.55 eV from the
electron-donating tendency of PVPK, whereas the LUMO
energy levels are located at ꢁ2.40 (or 2.55 eV) from the electron-
accepting tendency of POXD (or PBOXD). The HOMO and
LUMO energy levels of the random copolymers are similar to
their corresponding donor and acceptor moieties, indicating the
relatively weak interaction between them.
Memory device characteristics of random copolymers
The effects of donor/acceptor ratio and acceptor moieties of the
random polymers on memory device performance were system-
atically investigated. The memory behavior on P(VPKxOXDy)
and P(VPKxBOXDy) random copolymers were tested by the
current–voltage (I–V) characteristics of ITO/polymer/Al device
structures. ITO was used as the anode and Al was set as the
cathode (maintained as common) in all electrical measurements.
Fig. 4 Current–volatge (I–V) characteristics of (a) P(VPK8BOXD2), (b)
P(VPK8OXD2) memory device.
4782 | J. Mater. Chem., 2011, 21, 4778–4786
This journal is ª The Royal Society of Chemistry 2011

P(VPK8OXD2) in steps of 0.1 V. In the case of P(VPK8BOXD2)
(see Fig. 4(a)), the device is initially in the OFF state (‘‘0’’ signal
in data storage) with a current in the range of 10ꢁ11–10ꢁ12 A as the
voltage sweeps from 0 to 6 V. When the voltage increases further,
an abrupt current jump occurs at a threshold voltage of about 6
V indicating that the transition from low-conductivity OFF state
to the high-conductivity ON state (‘‘1’’ signal in data storage).
This electronic transition from the OFF state to ON state in the
first sweep serves as the ‘‘writing’’ process. The electrical
switching of the P(VPK8BOXD2) device shows the distinctly
bistable conductance state with the ON/OFF current ratio up to
105. In the second sweep, the memory device can be kept in the
ON state during a subsequent sweep from 0 to 8.0 V. The third
sweep is conducted after turning off the power for about 8 min.
The device can be reprogrammed from OFF state to ON state
when the voltage bias increases to 6.1 V and the current is kept in
the ON state (the third and fourth sweeps). From the continuous
repeated sweep cycle, it indicates that the ON state could be
maintained for approximately a period of 5–10 min after the
removal of power and would gradually relax back to the OFF
state. This electrical switching concludes that the
P(VPK8BOXD2) memory device still exhibits the volatile nature
of a static random access memory (SRAM) because the stored
data are finally lost. It is also found that the transition to the
ON state is reversible with a variation on threshold voltage,
probably due to the delay on conformation change of the pendent
aromatic ring under the voltage bias. Moreover, the volatile
P(VPK8BOXD2) device can be repeated many times for at least 10
different devices with reproducible data. The unstable ON state of
the volatile memory device can also be retained by a refreshing
voltage pulse of 1 V within 1 ms duration every 30 s (named the rf
trace). Further increase in the BOXD content, P(VPK5BOXD5) or
P(VPK2BOXD8) shows almost an insulating state in a low current
variation of 10ꢁ11ꢁ10ꢁ12 A, as shown in Fig. 5(a).
Fig. 5 Current–volatge (I–V) characteristics of the memory device: (a)
P(VPK5OXD5), P(VPK2OXD8), (P(VPK5BOXD5) and
P(VPK2BOXD8), and (b) PVPK, POXD and PBOXD.
The electrical behavior of the P(VPK8OXD2) device was also
studied, as shown in Fig. 4(b). The I–V characteristics define the
two electrical conductance state of P(VPK8OXD2) device and
also reveal the volatile nature of the memory effect. As compared
to the P(VPK8BOXD2) device, the P(VPK8OXD2) memory
device performance shows a slightly longer period for the
retained ON state. The transition from the OFF (10ꢁ11–10ꢁ12 A)
state to the ON (10ꢁ6–10ꢁ7A) state is observed with an abrupt
current increase (the first and third sweep). Besides, the ON/OFF
current ratios are about 104 with a threshold voltage of 6.3 V. The
third sweep is performed after turning off the power, whereas the
device returns back to the OFF state after more than 10 min
(usually 15 min). The OFF state can be further recovered to
a stored state again with a re-applied switching voltage bias
indicating that this device is also rewritable. Thus, the
P(VPK8OXD2) device still reveals the common characteristics as
a SRAM feature since it exhibits a temporary data remanence
behavior. However, for the higher content of POXD moieties,
the I–V curves of P(VPK5OXD5) and P(VPK2OXD8) device
display the insulator behavior (Fig. 5(a)) which is similar to the
P(VPK5BOXD5) and P(VPK2BOXD8) devices. Besides, testing
on the device based on homopolymers (PVPK, POXD and
PBOXD) was also carried out as a reference (Fig. 5(b)). The
device fabricated with PVPK material as the active memory layer
shows only single moderately high conductance state without
a conductance switching behavior whereas the POXD or
PBOXD device maintains in a low conductance state even if the
voltage across the device is swept from 0 to 8 V.
The stability of the memory effect was also evaluated under the
same atmosphere. Fig. 6 shows the retention time tests under
a constant stress of 1 V for both the ON and OFF states of the
P(VPK8BOXD2) device, respectively. As shown in the figure, an
ON/OFF current ratio of 104–105 can be maintained and no
obvious degradation in current is observed for both the ON and
OFF states after periods of at least 1 h. In addition, the stimulus
effect of read pulse on the ON and OFF state was also investi-
gated as shown in Fig. 7. The insert shows the pulse generation
used in the measurement with a pulse period and width of 3 and 2
ms, respectively. The P(VPK8BOXD2) memory device is quite
stable for over 107 continuous read pluses of 1 V. Therefore, both
ON and OFF states are conducted under the voltage stress and
continuous pulse cycles.
Operating mechanism of random copolymers memory device
The volatile memory behavior can be rationalized by evaluating
the electronic structures of basic units and random copolymers.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 4778–4786 | 4783

lying HOMO level of OXD or BOXD moieties embedded in the
polymer thin film layer, a poor (or limited) charge transport
occurred through the neighboring VPK moieties. As a result, the
incorporation of the electron-accepting OXD or BOXD moieties
could serve as the hole-blocking layer in the active memory layer.
For the case of the larger acceptor content in the copolymers, the
I–V characteristic curves of P(VPK5OXD5), P(VPK5BOXD5),
P(VPK2OXD8), and P(VPK2BOXD8) devices exhibit a low
current state between 10ꢁ12 to 10ꢁ11 A as shown in Fig. 5(a). This
insulating behavior is almost the same as the pure POXD or
PBOXD device (Fig. 5(b)). On the other hand, the PVPK
homopolymer is a hole transport material and the formation of
pathways for hole hopping with neighboring carbazole units
under the electric field contribute to the ON state conductance
(Fig. 5(b)).31,32,44,45 Therefore, the electrical switching between the
top and bottom electrodes depends on the trapping amounts of
POXD or PBOXD moieties in the random copolymers. The
higher electron acceptor ratio (5/5 or 2/8) of random copolymers
may prevent the positive charge from the direct hopping process
and maintain the OFF state.
Fig. 6 Retention time test on the ON and OFF states of the ITO/
P(VPK8BOXD2)/Al device under a continuous readout voltage.
For the case of a device with relatively higher donor ratio of the
copolymers, P(VPK8OXD2) and P(VPK8BOXD2), the electrical
switching exhibits the volatile memory behavior. Under a low
voltage bias(0–6 V),holes donothave sufficientenergy toovercome
the charge injection barrier between VPK/OXD or VPK/BOXD
moieties and are blocked by the electron-accepting OXD or BOXD
moieties in the film. Thus, the device is on the low conductance state
(OFF state). As sweeping up to the threshold voltage, the generated
amounts of holes with sufficient activation energy inject into
memory layer and fill the charge trapping environment. The holes
are able to flow continuously and more effectively in the copolymer
film, switching the device from the OFF to ON state. The increase of
current probably results from the efficient interchain hopping of the
chargecarriersintheneighboringcarbazolechain. TheONstatecan
not be maintained without a continuous voltage bias and it returns
to the original state after the removal of the applied electric field for
a short period of time. This is probably due to the shallow depth of
trap, which is 0.59 eV for P(VPK8BOXD2) and 0.74 eV for
P(VPK8OXD2), respectively. Note that the depth is calculated from
the difference of HOMO level of VPK and BOXD or OXD units in
Fig. 8. The shallow traps contribute to the unstable ON state of the
volatile nature.24 The reported PVK–C60 copolymer memory device
exhibited a non-volatile memory characteristic since the formed
charge transfer complex switching to ON state could be retained,
due to the strong dipole moment.32 However, our synthesized
P(VPK8OXD2) or P(VPK8BOXD2) copolymers show no obvious
interaction between donor and acceptor moieties and the sponta-
neously back transferring of shallow trapped charges lead to the
volatile memory performance as the electric field is removed.24
Besides, the higher HOMO energy level of P(VPK8BOXD2)
relative to that of P(VPK8OXD2) allows a shorter holding period
of the ON state, due to the relative rapid back transferring of
shallow trapped holes process. The slightly smaller threshold
voltage of P(VPK8BOXD2) as compared to P(VPK8OXD2)
may attribute to the lower hole injection barrier between
ITO/polymer interface. However, poly(9-(2-((4-vinylbenzyl)
oxy)ethyl)-9H-carbazole) (PVBCz) reported in the literature45
exhibited a much smaller threshold voltage (ꢁ2 V) since the
P(VPK8OXD2) and P(VPK8BOXD2) devices need to overcome
Fig. 7 Stimulus effect of read pulses on the ON and OFF states of the
ITO/P(VPK8BOXD2)/Al device. The insert shows the pulse shapes in the
measurement.
Molecular simulation on the VPK, OXD and BOXD moieties
and basic units of P(VPKxOXDy) and P(VPKxBOXDy) were
carried out by DFT/B3LYP/6-31G(d) with Gaussian 03 soft-
ware. The results of optical absorption and CV indicate that
there is no obvious charge transfer interaction between the
pendent donor (VPK) and acceptor (OXD or BOXD) moieties,
which is also verified by theoretical calculation. The relative
energy diagram is summarized in Fig. 8. The electronic structures
of the basic units through the simulation give a reasonable
qualitative result and are consistent with our trend of experi-
mental data. In the random copolymers, the HOMO energy level
is associated with PVPK units whereas the LUMO energy level is
associated with the corresponding POXD or PBOXD units.
With
a
configuration
of
ITO/P(VPKxOXDy)
or
P(VPKxBOXDy)/Al device, the barrier height of the ITO/poly-
mer (HOMO level) contact (0.74–0.76 eV) is quite smaller than
that of Al/polymer (LUMO level) contact (1.73–1.92 eV) when
the polymers are at the ground state. It suggests that the hole
injection from ITO into the HOMO level of random copolymers
is the dominated process and all the random copolymers can be
served as hole-transporting materials. However, due to the low-
4784 | J. Mater. Chem., 2011, 21, 4778–4786
This journal is ª The Royal Society of Chemistry 2011

helpful discussion on memory devices. The technical assistance
of Mr. An-Dih Yu, Mr. Jhao-Chen Chen, and Ms. Shiang-Lin
Lian are highly appreciated.
References
1 J. Liu, X. Guo, L. J. Bu, Z. Y. Xie, Y. X. Cheng, Y. H. Geng,
L. X. Wang, X. B. Jing and F. S. Wang, Adv. Funct. Mater., 2007,
17, 1917.
2 W. C. Wu, C. L. Liu and W. C. Chen, Polymer, 2006, 47, 527.
3 Q. Peng, J. B. Peng, E. T. Kang, K. G. Neoh and Y. Cao,
Macromolecules, 2005, 38, 7292.
4 H. Xin, X. G. Guo, F. S. Kim, G. Q. Ren, M. D. Watson and
S. A. Jenekhe, J. Mater. Chem., 2009, 19, 5303.
Fig. 8 Molecular simulation of HOMO and LUMO energy levels for
VPK, OXD and BOXD segments and basic units of P(VPKxOXDy) and
P(VPKxBOXDy) along with the work function of the electrodes.
5 N. S. Baek, S. K. Hau, H. L. Yip, O. Acton, K. S. Chen and
A. K. Y. Jen, Chem. Mater., 2008, 20, 5734.
6 J. H. Tsai, C. C. Chueh, M. H. Lai, C. F. Wang, W. C. Chen, B. T. Ko
and C. Ting, Macromolecules, 2009, 42, 1897.
7 M. H. Lai, C. C. Chueh, W. C. Chen, J. L. Wu and F. C. Chen, J.
Polym. Sci., Part A: Polym. Chem., 2009, 47, 973.
8 J. C. Bijleveld, A. P. Zoombelt, S. G. J. Mathijssen, M. M. Wienk,
M. Turbiez, D. M. de Leeuw and R. A. J. Janssen, J. Am. Chem.
Soc., 2009, 131, 16616.
9 L. Huo, J. Hou, S. Zhang, H. Y. Chen and Y. Yang, Angew. Chem.,
Int. Ed., 2010, 49, 1500.
the percolation threshold of the trapping site for effective charge
carrier transport. It concludes that electric-field induced memory
switching based on P(VPK8OXD2) or P(VPK8BOXD2) random
copolymers depend on the charge hopping conduction between
the carbazole units with coexisting charge trapping environment
and the volatility of the memory effect is determined from the
ability of charge trapping/back transferring of trapped charge.
10 J. H. Tsai, W. Y. Lee, W. C. Chen, C. Y. Yu, G. W. Hwang and
C. Ting, Chem. Mater., 2010, 22, 3290.
11 R. D. Champion, K. F. Cheng, C. L. Pai, W. C. Chen and
S. A. Jenekhe, Macromol. Rapid Commun., 2005, 26, 1835.
12 A. Babel, Y. Zhu, K. F. Cheng, W. C. Chen and S. A. Jenekhe, Adv.
Funct. Mater., 2007, 17, 2542.
13 K. F. Cheng, C. L. Liu and W. C. Chen, J. Polym. Sci., Part A: Polym.
Chem., 2007, 45, 5872.
14 W. Y. Lee, K. F. Cheng, T. F. Wang, C. C. Chueh, W. C. Chen,
C. S. Tuan and J. L. Lin, Macromol. Chem. Phys., 2007, 208, 1919.
15 C. L. Liu, J. H. Tsai, W. Y. Lee, W. C. Chen and S. A. Jenekhe,
Macromolecules, 2008, 41, 6952.
16 H. A. Becerril, N. Miyaki, M. L. Tang, R. Mondal, Y. S. Sun,
A. C. Mayer, J. E. Parmer, M. D. McGehee and Z. Bao, J. Mater.
Chem., 2009, 19, 591.
17 H. Yan, Z. H. Chen, Y. Zheng, C. Newman, J. R. Quinn, F. Dotz,
M. Kastler and A. Facchetti, Nature, 2009, 457, 679.
18 J. H. Tsai, C. C. Chueh, W. C. Chen, C. Y. Yu, G. W. Hwang,
C. Ting, E. C. Chen and H. F. Meng, J. Polym. Sci., Part A:
Polym. Chem., 2010, 48, 2351.
19 Q. D. Ling, D. J. Liaw, C. Zhu, D. S. H. Chan, E. T. Kang and
K. G. Neoh, Prog. Polym. Sci., 2008, 33, 917.
20 Q. D. Ling, D. J. Liaw, E. Y. H. Teo, C. Zhu, D. S. H. Chan,
E. T. Kang and K. G. Neoh, Polymer, 2007, 48, 5182.
21 Y. Yang, J. Ouyang, L. Ma, R. J. H. Tseng and C. W. Chu, Adv.
Funct. Mater., 2006, 16, 1001.
22 J. C. Scott and L. D. Bozano, Adv. Mater., 2007, 19, 1452.
23 T. Ouisse and O. Stephan, Org. Electron., 2004, 5, 251.
24 Q. D. Ling, Y. Song, S. L. Lim, E. Y. H. Teo, Y. P. Tan, C. Zhu,
E. T. Kang and K. G. Neoh, Angew. Chem., Int. Ed., 2006, 45, 2947.
25 S. Baek, D. Lee, J. Kim, S. H. Hong, O. Kim and M. Ree, Adv. Funct.
Mater., 2007, 17, 2637.
26 X. D. Zhuang, Y. Chen, B. X. Li, D. G. Ma, B. Zhang and Y. Li,
Chem. Mater., 2010, 22, 4455.
27 Q. D. Ling, F. C. Chang, Y. Sang, C. X. Zhu, D. J. Liaw,
D. S. H. Chan, E. T. Kang and K. G. Neoh, J. Am. Chem. Soc.,
2006, 128, 8732.
Conclusions
We have successfully synthesized new non-conjugated random
copolymers containing pendant electron-donating carbazole and
electron-accepting 1,3,4-oxadiazole by nitroxide-mediated free
radical polymerization. The random copolymers exhibited good
solubility in common organic solvents and good thermal
stability. The optical band gaps of PVPK, POXD, PBOXD,
P(VPKxOXDy) and P(VPKxBOXDy) estimated from the
absorption edges were 3.48, 3.68, 3.40, 3.44–3.52, 3.42–3.47 eV,
respectively, indicating that there is no strong electronic inter-
action between donor and acceptor segments. The low-lying
HOMO energy level of OXD or BOXD compared to VPK could
serve as hole-blocking moieties and provide the charge trapping
environment. Thus, the switching effect of the ITO/random
polymers/Al device was found to be dependent on the donor/
acceptor ratio or acceptor ability. The P(VPK8OXD2) or
P(VPK8BOXD2) memory device exhibited volatile SRAM
features, and ON/OFF current ratio reached up to 104–105. Both
ON and OFF states could endure 107 cycles under a pulse read
voltage and be stable for at least 1 h under a constant voltage
stress. The electrical volatile nature is mainly attributed to the
back transferring of shallow trapped charges in the pendent
functional moieties. Our results provided the strategies to
manipulate the memory characteristics on non-conjugated
copolymers with pendent donor/acceptor moieties.
28 N. H. You, C. C. Chueh, C. L. Liu, M. Ueda and W. C. Chen,
Macromolecules, 2009, 42, 4456.
Acknowledgements
29 T. Kuorosawa, C. C. Chueh, C. L. Liu, T. Higashihara, M. Ueda and
W. C. Chen, Macromolecules, 2010, 43, 1236.
Financial support from the National Science Council of Taiwan,
the Ministry of Economic Affairs, and the Excellence Research
Program of National Taiwan University are highly appreciated.
C. L. Liu also thanks the support of the Promotion of Envi-
ronmental Improvement for Independence of Young
Researchers Program of Ministry of Education, Culture, Sports,
Science and Technology-Japan. We thank Prof. Moonhor Ree of
Pohang University of Science & Technology (POSTECH) for
30 Q. Ling, Y. Song, S. J. Ding, C. Zhu, D. S. H. Chan, D. L. Kwong,
E. T. Kang and K. G. Neoh, Adv. Mater., 2005, 17, 455.
31 Q. Ling, W. Wang, Y. Song, C. Zhu, D. S. H. Chan, D. L. Kwong,
E. T. Kang and K. G. Neoh, J. Phys. Chem. B, 2006, 110, 23995.
32 Q. D. Ling, S. L. Lin, C. X. Zhu, D. S. H. Chan, E. T. Kang and
K. G. Neoh, Langmuir, 2007, 23, 312.
33 S. L. Lim, N. J. Li, J. M. Lu, Q. D. Ling, C. X. Zhu, E. T. Kang and
K. G. Neoh, ACS Appl. Mater. Interfaces, 2009, 1, 60.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 4778–4786 | 4785

34 H. Li, N. Li, H. Gu, Q. Xu, F. Yan, J. Lu, X. Xia, J. Ge and L. Wang,
J. Phys. Chem. C, 2010, 114, 6117.
42 Y. S. Lai, C. H. Tu, D. L. Kwong and J. S. Chen, Appl. Phys. Lett.,
2005, 87, 122101.
35 X. D. Zhuang, Y. Chen, G. Liu, B. Zhang, K. G. Neoh, E. T. Kang,
C. X. Zhu, Y. X. Li and L. J. Niu, Adv. Funct. Mater., 2010, 20, 2916.
36 J. Quyang, C. W. Chu, C. R. Szmanda, L. Ma and Y. Yang, Nano
Lett., 2005, 5, 1077.
37 R. J. Tseng, J. Huang, J. Quyang, R. B. Kaner and Y. Yang, Nat.
Mater., 2004, 3, 918.
43 J. S. Choi, J. H. Kim, S. H. Kim and D. H. Suh, Appl. Phys. Lett.,
2006, 89, 15211.
44 E. Y. H. Teo, Q. Ling, Y. P. Tan, W. Wang, E. T. Kang,
D. S. H. Chan and C. Zhu, Org. Electron., 2006, 7, 173.
45 S. L. Lim, Q. Ling, E. Y. H. Teo, C. Zhu, D. S. H. Chan, E. T. Kang
and K. G. Neoh, Chem. Mater., 2007, 19, 5148.
38 C. W. Chu, J. Quyang, J. H. Tseng and Y. Yang, Adv. Mater., 2005,
17, 1440.
46 J. Lin and D. Ma, Appl. Phys. Lett., 2008, 93, 093505.
47 C. M. Huang, Y. S. Liu, C. C. Chen, K. H. Wei and J. T. Sheu, Appl.
Phys. Lett., 2008, 93, 203303.
48 C. L. Liu, J. C. Hsu, W. C. Chen, K. Sugiyama and A. Hirao, ACS
Appl. Mater. Interfaces, 2009, 1, 1974.
49 L. Deng, P. T. Furuta, S. Garon, J. Li, D. Kavulak, M. E. Thompson
39 G. Liu, Q. D. Ling, E. Y. H. Teo, C. X. Zhu, D. S. H. Chan,
K. G. Neoh and E. T. Kang, ACS Nano, 2009, 3, 1929.
40 X. D. Zhang, Y. Chen, G. Liu, P. P. Li, C. X. Zhu, E. T. Kang,
K. G. Neoh, B. Zhang, J. H. Zhu and Y. X. Li, Adv. Mater., 2010,
22, 1731.
ꢁ
and J. M. J. Frechet, Chem. Mater., 2006, 18, 386.
50 M. J. Frisch et al., Gaussian 03, Revision B.04 Gaussian, Inc.,
Wallingford, CT, 2004.
41 Y. K. Fang, C. L. Liu, C. Li, C. J. Lin, R. Mezzenga, W. C. Chen,
Adv. Funct. Mater., 20, p. 3012.
4786 | J. Mater. Chem., 2011, 21, 4778–4786
This journal is ª The Royal Society of Chemistry 2011