Effects of terminal electron acceptor strength on film morphology and ternary memory performance of triphenylamine donor based devices

Article abstract of DOI:10.1039/c3tc30228h

This study reports the syntheses, photophysical and electrochemical properties and memory characteristics of triphenylamine (TPA) donor based molecules with progressively weaker terminal acceptor strength (i.e., nitro, acetyl and bromine). The influence of the terminal electron acceptor strength on the film morphology and the devices storage performances was investigated. Nonvolatile ternary ( 0 , 1 and 2 states) memory devices for high-density data storage could be achieved with a simple ITO/D-A molecule/Al sandwich configuration for TPA-NAP and TPA-AAP. It is noteworthy that the memory device based on TPA-AAP exhibited a better reproducibility and stability with lower operation voltages than that based on TPA-NAP, promising low-power consumption data-storage. These obtained results demonstrate that altering the terminal electron accepting strength in D-A molecules can adjust the film morphology and the device performances for the design of future advanced organic electronic devices. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3tc30228h

Journal of
Materials Chemistry C
View Article Online
View Journal | View Issue
PAPER
Effects of terminal electron acceptor strength on film
morphology and ternary memory performance of
triphenylamine donor based devices†
Cite this: J. Mater. Chem. C, 2013, 1,
3816
Hao Zhuang,a Qijian Zhang,a Yongxiang Zhu,a Xufeng Xu,a Haifeng Liu,a Najun Li,a
a
a
ab
and Lihua Wanga
*
*
Qingfeng Xu, Hua Li, Jianmei Lu
This study reports the syntheses, photophysical and electrochemical properties and memory characteristics
of triphenylamine (TPA) donor based molecules with progressively weaker terminal acceptor strength (i.e.,
nitro, acetyl and bromine). The influence of the terminal electron acceptor strength on the film
morphology and the devices storage performances was investigated. Nonvolatile ternary (“0”, “1” and
“2” states) memory devices for high-density data storage could be achieved with a simple ITO/D–A
molecule/Al sandwich configuration for TPA-NAP and TPA-AAP. It is noteworthy that the memory
device based on TPA-AAP exhibited a better reproducibility and stability with lower operation voltages
than that based on TPA-NAP, promising low-power consumption data-storage. These obtained
results demonstrate that altering the terminal electron accepting strength in D–A molecules can
adjust the film morphology and the device performances for the design of future advanced organic
electronic devices.
Received 4th February 2013
Accepted 15th April 2013
DOI: 10.1039/c3tc30228h
www.rsc.org/MaterialsC
studied. A better understanding of this type of effect is essential
for the rational molecular design of high performance devices
Introduction
Resistive memory devices are very attractive nowadays because
of their advantages of low-cost potential, good scalability and
tunable performance through molecular tailoring.1,2 Recently,
there have been several reports on elaborately designed devices
able to exhibit tristable conductance states under an external
electric eld, leading to an exponential increase in data-storage
density relative to binary memory systems, which is considered
to be an effective route to deal with the ever-growing demand for
future ultra-high density data-storage (UHDDS).3,4
Although a number of conjugated molecules containing
electron-accepting (A) and electron-donating (D) moieties have
been investigated in binary devices as electrically active mate-
rials,5–12 the effects of the terminal electron-accepting strength
in D–A small molecules on the lm morphology and nanolm-
based devices' ternary memory performances have rarely been
for future high-density data storage.
In this work, we focused on adjusting the morphology of
organic nanolms and tuning the nanolm-based device
performances through adjusting the acceptor strength. Consid-
ering the strong electron-donating ability of the triphenylamine
(TPA) moiety,5,10,11 herein, we designed and synthesized three
TPA-based D–A molecules containing the same azo chromo-
phore, but different terminal acceptors with various strengths
(Fig. 1a). The effects ofthe terminal electron acceptor strength on
the charge transfer extent, redox properties, HOMO/LUMO
energy levels and the memory performance of the small mole-
cules were fully investigated through measurements and theo-
retical simulations. A regular change in lm morphology could
be noted for D–A small molecules with progressively weaker
terminal acceptor strength, demonstrating a new, effective route
to improve the morphology and the interface between the lm
and electrode. The fabricated sandwich devices (ITO/D–A small
molecule/Al) were found to exhibit different ON-switching volt-
ages, state currents (i.e., different current levels for the OFF- and
ON1-states) and device stability, which could have a great impact
on the devices' power consumption and practical operation
process. The difference in performance could mainly be attrib-
uted to the difference in terminal acceptors and the resulting
variation in lm morphology and molecular energy levels. We
believe that this approach will provide new strategies for the
structural design of D–A molecules for high performance devices
in organic electronics.
aCollege of Chemistry, Chemical Engineering and Materials Science, China Petroleum
and Chemical Industry key laboratory of organic wastewater adsorption treatment &
resource, Soochow University, Suzhou 215123, P.R. China. E-mail: lujm@suda.edu.
cn; Fax: +86 512 65880367; Tel: +86 512 65880368
bKey Lab of Lithium Ion Battery Materials of Jiangsu Province, Institute of Chemical
Power Sources, Soochow University, Suzhou 215006, P.R. China
† Electronic supplementary information (ESI) available: NMR spectra for all the
products, UV-vis absorption spectra of the small molecules in solution, 3D-AFM
topography images, atomic force microscopy (AFM) image of the TPA-NAP lm
with prolonged annealing time, X-ray diffraction pattern, I–V characteristics of
the ITO/TPA-NAP/Al devices with various lm thicknesses and of the
ITO/TPA-BAP/Al device. See DOI: 10.1039/c3tc30228h
3816 | J. Mater. Chem. C, 2013, 1, 3816–3824
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry C
DMF solution (50 mL) of triphenylamine (2.45 g, 10.0 mmol),
and stirred at 0 ꢀC to room temperature overnight. Aer the
reaction was completed, the mixture was neutralized with
aqueous sodium hydroxide. The resulting precipitate was
ltered and puried by silica-gel column chromatography
(dichloromethane : petroleum ether ¼ 2 : 1). A dark red powder
1
was nally obtained. Yield: 2.05 g (52%). H NMR (400 MHz,
DMSO-d6): 8.40 (d, J ¼ 9.0 Hz, 2H, ArH), 7.99 (d, J ¼ 9.0 Hz, 2H,
ArH), 7.87 (d, J ¼ 9.0 Hz, 2H, ArH), 7.44 (t, J ¼ 7.8 Hz, 4H, ArH),
7.18–7.27 (m, 6H, ArH), 6.98 (d, J ¼ 9.0 Hz, 2H, ArH). 13C NMR
(75 MHz, DMSO-d6) d 156.08, 152.08, 148.02, 146.14, 146.05,
130.45, 126.67, 125.84, 125.70, 125.46, 123.36, 119.51. HRMS
(ESI): calcd for C24H18N4O2: 395.1506 (M + H+). Found: 395.1497
(M + H+). Anal. calcd for C24H18N4O2: C, 73.08; H, 4.60; N, 14.20.
Found: C, 72.52; H, 4.37; N, 14.08%.
Fig. 1 (a) Molecular structures of the D–A molecules and (b) scheme of the ITO/
D–A molecule/Al memory device.
Synthesis of (E)-1-(4-((4-(diphenylamino)phenyl)diazenyl)
phenyl)ethanone (TPA-AAP). TPA-AAP was prepared under
similar conditions, with 4-acetylaniline (1.35 g, 10.0 mmol) as
the reactant. A red powder was nally obtained. Yield: 0.89 g
Experimental
Materials
1
(23%). H NMR (300 MHz, DMSO-d6): 8.11 (d, J ¼ 8.2 Hz, 2H,
Triphenylamine, 4-bromoaniline, 4-acetylaniline, 4-nitroani-
line, and sodium nitrite were all purchased from commercial
sources (TCI, Alfa Aesar, and Sigma-Aldrich). All solvents were
purchased from Sinopharm Chemical Reagent Co., Ltd. All the
materials were used as received.
ArH), 7.98–7.76 (m, 4H, ArH), 7.49–7.32 (m, 4H, ArH), 7.26–7.11
(m, 6H, ArH), 6.97 (d, J ¼ 8.9 Hz, 2H, ArH), 2.63 (s, 3H, CH3). 13
C
NMR (75 MHz, DMSO-d6) d 197.74, 155.09, 151.52, 146.29,
137.91, 130.40, 129.97, 126.44, 125.54, 125.27, 122.62, 119.99,
27.36. HRMS (ESI): calcd for C26H21N3O: 392.1757 (M + H+).
Found: 392.1764 (M + H+). Anal. calcd for C26H21N3O: C, 79.77;
H, 5.41; N, 10.73. Found: C, 79.21; H, 5.23; N, 10.67%.
Instrumentation and characterization
Synthesis of (E)-4-((4-bromophenyl)diazenyl)-N,N-diphenyl-
aniline (TPA-BAP). TPA-BAP was prepared under similar
conditions, with 4-bromoaniline (1.71 g, 10.0 mmol) as the
reactant. An orange powder was nally obtained. Yield: 0.76 g
NMR spectra were measured on a NMR system 300 MHz and
INOVA 400 MHz spectrometer. Elemental analysis was per-
formed on a Carlo Erba-MOD1106 elemental analyzer. High-
resolution mass spectra (HRMS) were determined on a Micro-
mass GCT-TOF mass spectrometer with an ESI source. UV-vis
absorption spectra were carried out at room temperature with a
Shimadzu UV-3600 spectrophotometer. Cyclic voltammetry (CV)
measurements were carried out in 0.1 M acetonitrile solution of
tetrabutylammonium perchlorate (TBAP) using a CorrTest CS
Electrochemical Workstation equipped with a platinum gauze
auxiliary electrode and an Ag/AgCl reference electrode. Atomic
force microscopy (AFM) measurements were performed using a
MFP-3DTM (Digital Instruments/Asylum Research) AFM
instrument. Thermal properties were estimated from a PE TGA-
7 thermogravimetric analysis system (TGA) and a Pyris-1 DSC
differential scanning calorimeter (DSC) under a nitrogen
1
(18%). H NMR (400 MHz, DMSO-d6): 7.81 (d, J ¼ 8.8 Hz, 2H,
ArH), 7.70–7.78 (m, 4H, ArH), 7.41 (t, J ¼ 7.8 Hz, 4H, ArH), 7.20
(m, 6H, ArH), 6.99 (d, J ¼ 8.8 Hz, 2H, ArH). 13C NMR (75 MHz,
DMSO-d6) d 151.53, 151.15, 146.40, 146.15, 132.82, 130.36,
126.27, 125.35, 124.96, 124.44, 124.16, 120.30. HRMS (ESI):
calcd for C24H18BrN3: 428.0718, 430.0697 (M + H+). Found:
428.0756, 430.0729 (M + H+). Anal. calcd for C24H18BrN3: C,
67.30; H, 4.24; Br, 18.66; N, 9.81. Found: C, 66.85; H, 4.07; Br,
18.39; N, 9.67%.
Fabrication and measurements of the memory devices
The indium-tin oxide (ITO) glass was precleaned by sonication
with deionized water, acetone and ethanol, in that order, for 15
min each. Then the D–A molecule was deposited onto the
surface of the ITO under a pressure of 10ꢁ6 Torr. The thickness
of the lm was typically about 90 nm, which was traced by a
calibrated quartz crystal monitor. Finally, Al was thermally
evaporated onto the lm surface at 5 ꢂ 10ꢁ6 Torr through a
shadow mask to yield top electrodes with a thickness of around
100 nm and area of 0.0314 mm2. The scheme of the fabricated
sandwich memory device is shown in Fig. 1b.
atmosphere at a heating rate of 20 C minꢁ1, respectively. The
ꢀ
devices were characterized under ambient conditions, using a
Hewlett-Packard 4145B semiconductor parameter analyzer with
a Hewlett-Packard 8110A pulse generator.
Synthesis of the donor–acceptor molecules
Synthesis of (E)-4-((4-nitrophenyl)diazenyl)-N,N-diphenyl-
aniline (TPA-NAP).13 To concentrated sulfuric acid (6 mL) was
added sodium nitrite (0.69 g, 10.0 mmol). The mixture was
heated to 70 ꢀC and cooled to 0 ꢀC. To the mixture was added a
propionic acid–acetic acid–DMF mixed solution (2.5 mL : 12.5
Molecular simulation
mL : 10 mL) of 4-nitroaniline (1.38 g, 10.0 mmol), and the Calculations on the D–A molecules investigated in this work
mixture was stirred at 0 ꢀC for 2 h. To the mixture was added a were carried out with the Gaussian 03 program package.14
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. C, 2013, 1, 3816–3824 | 3817

View Article Online
Journal of Materials Chemistry C
Paper
Electronic properties of TPA-NAP, TPA-AAP and TPA-BAP in the thermal stability with progressively weaker terminal acceptor
ground state, including molecular orbitals and the electrostatic strength. The relatively high Td and Tg values of the donor–
potential (ESP) surface, were calculated from the density func- acceptor molecules particularly suggest their potential in long-
tional theory (DFT): the Becke's three-parameter functional term stability for data-storage utility.
combined with the Lee, Yang, and Parr's correlation functional
(B3LYP),15 along with the 6-31G(d) basis set.
Photophysical properties
Fig. 3 shows the optical absorption spectra of the TPA-NAP, TPA-
Results and discussion
AAP and TPA-BAP thin lms. The high-energy absorption bands
(ꢃ300 nm) can be attributed to the p–p* transition in these D–A
The chemical structures of the three designed small molecules
molecules, mainly arising from the triphenylamine moieties.11
are shown in Fig. 1a and their thermal properties were
The low-energy absorption bands in the region of 440–490 nm
arise from the intramolecular charge transfer (ICT). With the
evaluated.
enhancement of the terminal electron acceptor strength, a
Thermal properties
bathochromic shi can be observed for the ICT absorption
bands of the D–A molecules. The absorption maximum shied
from 443 nm in TPA-BAP to 457 nm in TPA-AAP to 487 nm in
TPA-NAP (the terminal electron acceptor varied from bromine
to acetyl to nitro). Since the electron donor moiety in all the
molecules is the same, the observed bathochromic shi could
thus be assigned to the stronger ICT with increasing electron
acceptor strength. The absorption onset correspondingly red-
shied from ꢃ540 nm in TPA-BAP to ꢃ560 nm in TPA-AAP to
ꢃ610 nm in TPA-NAP, indicating a reduction in the optical band
gap of the D–A molecule with increasing electron-accepting
strength. The optical band gap of TPA-BAP, TPA-AAP and TPA-
NAP can be calculated from the absorption edge to be 2.38 eV,
2.25 eV and 2.08 eV, respectively.
TGA and DSC measurements were carried out to estimate the
thermal properties of the D–A molecules (Fig. 2), and the results
are summarized in Table 1. All three D–A molecules had a glass
transition process, which were at 146 ꢀC, 143 ꢀC and 143 ꢀC for
TPA-BAP, TPA-AAP and TPA-NAP, respectively. The relatively
high glass transition temperatures (Tg) observed for these D–A
molecules suggested a good thermal stability of the deposited
lms, a desired property of memory materials because of the
inevitable joule heat during the course of device operation.
Though the terminal acceptor strength was changed, the onset
decomposition temperatures of the molecules determined by
TGA were in the range of 268–278 ꢀC (Table 1), further
demonstrating the good thermal stability of the materials. It is
noteworthy that both Td and Tg increased slightly with
decreasing the terminal acceptor strength, implying a better
Electrochemical properties
To understand the electronic properties of the triphenylamine
D–A molecules with different terminal electron acceptors we
performed cyclic voltammetry (CV) measurements on solutions
of the molecules (Fig. 4). The obtained cyclic voltammograms in
acetonitrile are all shown in Fig. 3, and the calculated CV data is
summarized in Table 1. As can be observed, each D–A molecule
shows both p-doping and n-doping behavior. The CV curves can
be repeatedly obtained for many cycles without a change in the
redox characteristics. The rst reduction peak of TPA-BAP was
found to arise from the Br atom, according to the reference CV
of 4-bromo-N,N-diphenylaniline (Fig. S8†). For TPA-NAP and
TPA-AAP, the rst reduction peak could be attributed to the
nitro group and the acetyl group, respectively; and the second
reduction peak was caused by the azo moiety with
a
relatively weak acceptor strength.16a The half-wave reduction
potentials (Er1e/2ds) of TPA-BAP, TPA-AAP and TPA-NAP were ꢁ0.86
V, ꢁ0.81 V and ꢁ0.68 V (vs. Ag/AgCl), respectively. Thus, an
Fig. 2 TGA curves of the D–A molecules with a heating rate of 20 ꢀC minꢁ1
under nitrogen. The inset shows the DSC thermograms.
Table 1 Thermal, optical, and electrochemical properties of the prepared D–A compounds
Compound
Td (ꢀC)
Tg (ꢀC)
lmax (nm)
Band gapa (eV)
HOMOb (eV)
LUMOc (eV)
TPA-NAP
TPA-AAP
TPA-BAP
268
270
278
139
139
142
302, 487
301, 457
304, 443
2.06
2.25
2.36
ꢁ5.46
ꢁ5.41
ꢁ5.39
ꢁ3.40
ꢁ3.16
ꢁ3.03
a
Estimated from the onset of absorption. b The HOMO energy levels were calculated from cyclic voltammetry with reference to ferrocene (0.45 eV).
c
LUMO ¼ HOMO + band gap.
3818 | J. Mater. Chem. C, 2013, 1, 3816–3824
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry C
obvious shi of the half-wave reduction potential to more
positive values with increasing terminal electron acceptor
strength could be noted.16b Moreover, the half-wave oxidation
potentials (Eo1/x2s) of the D–A molecules were approximately 1.15
eV, arising from the same electron-donating tendency of the
TPA units. The HOMO levels of TPA-BAP, TPA-AAP and TPA-NAP
were determined to be ꢁ5.39 eV, ꢁ5.41 eV and ꢁ5.46 eV (listed
in Table 1), respectively.
Morphology of the thin lms
To investigate the microstructure of the small-molecule thin
lms, atomic force microscopy (AFM) was carried out. The 3D-
AFM topography images are shown in the ESI (Fig. S9†). As
shown in Fig. 5, the tapping-mode AFM height images clearly
show a twining structure for TPA-NAP (the strongest acceptor
strength), which was found to grow into strand-like microwires
when the annealing time was prolonged by 12 h (Fig. S10†),
indicating that the TPA-NAP molecules underwent a self-orga-
nization process during thermal annealing. This type of
morphology has also been observed on other substrates with
different polarities, such as glass and quartz, indicating that it
is the interaction between the molecules, rather than the
interaction between the molecules and the substrates, that
drives the molecular self-organization during the thermal
annealing process. Additionally, a grain-like morphology was
obtained for TPA-AAP and a nearly amorphous lm was
obtained for TPA-BAP, which might be due to the progressively
weak intermolecular interactions. This is also in accordance
with the UV-vis absorption and CV results. These results
demonstrate that the different terminal acceptor strengths of
the molecules may result in completely different lm
morphologies. The surface root-mean-square (RMS) roughness
was 0.67 nm for TPA-BAP, while TPA-AAP and TPA-NAP have
much rougher surface morphologies with RMS roughnesses of
15.2 nm and 35.2 nm, respectively (Fig. 5). The less rough
Fig. 3 The UV-visible absorption spectra of TPA-NAP, TPA-AAP and TPA-BAP thin
films on a quartz substrate.
Fig. 4 Cyclic voltammograms of TPA-NAP, TPA-AAP and TPA-BAP in acetonitrile
solution. The scan rate was 100 mV sꢁ1
.
Fig. 5 Tapping-mode AFM topography (A–C) of vacuum-deposited nanofilms annealed at 80 ꢀC; typical cross-section profiles (a–c) of AFM topographic images of
TPA-NAP, TPA-AAP and TPA-BAP films on ITO substrates, respectively.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. C, 2013, 1, 3816–3824 | 3819

View Article Online
Journal of Materials Chemistry C
Paper
surface of the TPA-AAP lm compared to the TPA-NAP may be level ꢃ10ꢁ7 A). As the negative bias increased, a sharp transition
benecial for improving the device's performance in decreasing from the low-conductivity (OFF) state to an intermediate-
the injection barrier.
conductivity (ON1) state was observed at ꢁ2.0 V (the switch-ON1
In order to measure the crystalline ordering of the organic voltage), as was indicated by the abrupt increase in the current
thin lms, wide-angle X-ray diffraction (WAXRD) experiments (i.e., from 10ꢁ7 A to 10ꢁ5 A) (sweep 1 and 2). Furthermore, when
were performed on the lms annealed at 80 ꢀC and the patterns the negative bias was even higher (i.e., the fourth sweep from
are displayed in the ESI (Fig. S11†). Both the TPA-NAP and TPA- 0 to ꢁ4.0 V), the current underwent a second leap to 10ꢁ2
A
AAP lm exhibited conspicuous diffraction peaks, indicative of (ON2-state) around ꢁ3.1 V (the switch-ON2 voltage). The device
the microstructural ordering. The reections at 2q ¼ 10.57ꢀ could be switched on progressively under an external electric
ꢀ
˚
˚
(8.36 A) and 2q ¼ 21.24 (4.08 A) of TPA-AAP correspond to a eld, exhibiting three distinctive conductivity states. The tran-
typical intermolecular p–p interaction distance.17a Also, the sition from OFF- to the ON1- and further to the ON2-state could
reection at 2q ¼ 4.97ꢀ observed for TPA-NAP corresponds to a serve as the “writing” process for practical data storage appli-
17b
˚
d-spacing of 17.72 A and indicates long-range order. These cations. The obtained ON1- and ON2-states could be well
results imply that both TPA-NAP and TPA-AAP self-assembled retained, when the power was shut off (sweep 3 and 5) or even
into an ordered crystalline structure and packed closely in the during the reverse voltage sweep, fullling the features of a
thin lm state, which could facilitate the charge transport. In WORM device. The I–V curves of various lm thicknesses are
contrast, no discernible XRD peak could be observed for the displayed in the ESI (Fig. S12†). The OFF-, ON1- and ON2-states
TPA-BAP thin lm, implying an amorphous feature. These can be programmed to digital signals “0”, “1” and “2”, respec-
results indicate the close relationship between the molecular tively, which could increase the storage density dramatically.
terminal substituents and the molecular packing. The differ-
The I–V characteristics of the ITO/TPA-AAP/Al device showed
ences in the molecular microstructures may account for their a similar ternary memory behavior, yet still showed differences
difference in memory performance.
in some aspects (Fig. 6b). The device underwent two abrupt
increases in conductivity at threshold voltages around –1.5 V
(switch-ON1 voltage) and –2.5 V (switch-ON2 voltage), exhibiting
electrically tristable states. But the current levels of the OFF- and
intermediate-state are obviously higher than those of the TPA-
NAP-based device. Also, it is noteworthy that the on-switching
voltages of TPA-AAP were lower than those of TPA-NAP, which
Current–voltage (I–V) characteristics of the memory devices
Fig. 6 shows the typical I–V characteristics of the memory
devices with a simple sandwich structure. Initially, the ITO/TPA-
NAP/Al device was in a low-conductivity (OFF) state (current
Fig. 6 Typical I–V curves of the (a) ITO/TPA-NAP or (b) TPA-AAP/Al memory device: The first sweep was carried out from 0 to ꢁ4 V on storage cell 1; the other voltage
sweeps were carried out on storage cell 2; (c) the statistical data of reproducibility and current distribution (read at ꢁ1 V; 45 cells for each kind of device).
3820 | J. Mater. Chem. C, 2013, 1, 3816–3824
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry C
promises low-power consumption. The relatively higher current storage devices. It is noteworthy that the probability of switch-
level and low threshold voltage of TPA-AAP than TPA-NAP may ing for TPA-BAP, a binary active material, was the highest with a
benet from the facilitated charge transport by the more device yield of 78% (35 out of 45 cells).
ordered microstructure. For the ITO/TPA-BAP/Al device, only
one switching from the OFF- to ON-state was observed at about
ꢁ1.1 V, exhibiting electrically bistable states (Fig. S13†). All the
I–V curves could be reproducibly obtained in more than three
devices fabricated in different batches. Thus, the devices
memory performance (here, mainly the switching voltages and
state current levels) based on D–A small molecules can be
effectively tuned through the introduction of different terminal
substituents.
As the reproducibility of the switching phenomenon is
considered to be an important parameter for memory devices,
135 cells were selected to investigate the cell-to-cell uniformity
for the small molecule based devices. The statistical data,
obtained from 15 cells in three different devices (a total of 45
cells for each small molecule) is depicted in Fig. 6c. The current
level of the OFF- and ON1-states for the three molecules grad-
ually increased with the decreasing strength of the terminal
acceptor moieties, which is consistent with the I–V test results
(Fig. 6a and b and S13†). The probability of obtaining electrical
tristability for TPA-AAP was 28 out of 45 cells (device yield of
62%), which is apparently higher than TPA-NAP with a device
yield of 49%, suggesting that such a rough lm as TPA-NAP
(with a larger calculated RMS roughness) is not benecial for
improving the probability and stability of high-density data-
Stability measurements
The stability performance of the different conductive states
was evaluated for the memory devices based on these D–A
molecules under ambient air conditions. Fig. 7 shows repre-
sentative results of the retention time test under a constant
stress of ꢁ1 V for the OFF-, ON1- (intermediate-state) and ON2-
states of the devices. As can be seen, during the entire test
period of 8000 s, no signicant degradation in the current could
be observed for all the distinctive states. Moreover, the effect of
continuous read pulses of ꢁ1 V on the OFF-, ON1- (interme-
diate-state) and ON2-states was also investigated (Fig. 7b and d).
A pulse with width and duration of 1 ms and 2 ms, respectively,
which are typical values in practical devices, was adopted. No
obvious degradation in the current was observed for the
different states aer more than ten million (107) continuous
read cycles, indicating that the distinctive conductivity states
were all insensitive to read cycles.
Proposed memory mechanisms
To better understand the switching behavior of the memory
devices, we studied the electronic properties of the TPA-based
triad at the ground state by density function theory (DFT). Fig. 8
Fig. 7 Stability tests of the fabricated ITO/TPA-NAP (a and b) or TPA-AAP (c and d)/Al memory device: (a and c) retention time measurement for the ON-, intermediate-
and OFF-states with a constant reading voltage of ꢁ1 V; (b and d) effect of read pulse of ꢁ1 V on the ON-, intermediate- and OFF-states. The inset shows the pulse shape
employed.
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. C, 2013, 1, 3816–3824 | 3821

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 8 HOMOs, LUMOs and ESP surfaces of (a) TPA-NAP, (b) TPA-AAP and (c) TPA-BAP in their optimized ground-state structures.
shows the optimized geometries, molecular orbitals and elec- AAP. The observed lower current level of TPA-NAP for the OFF-
trostatic potential (ESP) surfaces of the triad obtained with the and intermediate-states could be ascribed to the lower injection
B3LYP/6-31G(d) basis set. The HOMOs are mainly localized at efficiency from the ITO to the TPA-NAP layer resulting from the
the donor side (i.e., the triphenylamine moiety), while the relatively high injection barrier.19
LUMOs are mainly localized at the acceptor side (i.e., the azo
Since these D–A molecules are not fully conjugated, the
and nitro/acetyl chromophores). The localization of the LUMOs intermolecular hopping of charge carriers between neighboring
of TPA-NAP and TPA-AAP is stronger on the acceptor moieties molecules dominates the charge transport within the material.
compared with that in TPA-BAP since nitro and acetyl are In Fig. 8, DFT simulations show that the molecular surface has
stronger acceptors than bromine atoms. The localization of continuous positive ESP regions (in purple), which can serve as
HOMO/LUMO orbitals on the donor–acceptor moieties has open channels for charge carrier migration throughout the
been commonly observed in theoretical studies on D–A mole- conjugated backbone. The negative ESP regions (in brown)
cules, indicating the occurrence of signicant charge transfer in arising from the electron-decient groups, such as azo chromo-
the ground-state to the excited-state transition.6,18
phore and nitro/acetyl groups, can function as “traps” with
To gain insight into the switching mechanisms for the different depths to impede the mobility of charge carriers.6,20 In
memory devices, an energy level diagram for the ITO/D–A the TPA-NAP and TPA-AAP lm, the microstructural orderingwas
molecule/Al devices is summarized in Fig. 9. The energy barrier revealed by the XRD diffraction. Once an electric eld was
for hole injection from ITO to the HOMO of the active layer (i.e., applied, the traps were lled progressively, corresponding to the
ꢃ0.5 eV) is much lower than the energy barrier for electron double electrical on-switching behavior. Therefore, the TPA-NAP
injection from Al to the LUMO of the active layer. Therefore, and TPA-AAP based devices exhibited ternary memory behavior
hole injection is a favored process under a negative bias. The (i.e., double switching) while the TPA-BAP lm with an amor-
hole injection barrier for TPA-AAP is smaller than TPA-NAP, phous feature only showed single switching though a Br atom
which could contribute to the lower switch-ON voltages of TPA- which also has the potential to serve as a trap, as indicated by the
CV measurement. The low switch-ON voltages of TPA-AAP are
benecial to the reduction of the power consumption during the
operation course for memory devices. Therefore, high-perfor-
mance low-power-consumption memory devices could be ach-
ieved by manipulating the lm morphology and energy levels of
these TPA-based molecules through altering the strength of
electron acceptors. The trapped charges were stabilized by
intramolecular charge transfer forming a charge-separated state
and could not be easily de-trapped under reverse bias, resulting
in a high-conductivity state retainable for a long time. Accord-
ingly, the devices show WORM-type memory characteristics.
Conclusions
Novel donor–acceptor small molecules have been successfully
synthesized. The effect of the terminal electron acceptor
Fig. 9 Energy level diagram of the HOMO and LUMO for the D–A molecules
along with the work function of the electrodes.
3822 | J. Mater. Chem. C, 2013, 1, 3816–3824
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Paper
Journal of Materials Chemistry C
strength on the optoelectronic properties, lm morphology and
memory device performance with ITO/D–A molecule/Al sand-
wich conguration was investigated. The as-fabricated devices
based on TPA-NAP or TPA-AAP both exhibited ternary memory
behavior, while that based on TPA-BAP only showed binary
memory characteristics. The ITO/TPA-AAP/Al device has a better
ternary performance with higher stability and reproducibility
and a lower operation voltage than ITO/TPA-NAP/Al, which may
arise from the less rough surface and the resulting lower
injection barrier. The low switching voltage and low current
level are highly benecial in practical application to reduce the
6 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.
7 N.-G. Kang, B. Cho, B.-G. Kang, S. Song, T. Lee and J.-S. Lee,
Adv. Mater., 2012, 24, 385.
8 Y. H. Zhang, H. Zhuang, Y. Yang, X. F. Xu, Q. Bao, N. J. Li,
H. Li, Q. F. Xu, J. M. Lu and L. H. Wang, J. Phys. Chem. C,
2012, 116, 22832.
9 S. G. Hahm, S. Choi, S. H. Hong, T. J. Lee, S. Park, D. M. Kim,
W. S. Kwon, K. Kim, K. Kim, O. Kim and M. Ree, Adv. Funct.
Mater., 2008, 18, 3276.
power consumption and the joule heat generated during the 10 (a) Y. Shang, Y. Wen, S. Li, S. Du, X. He, L. Cai, Y. Li, L. Yang,
operation course. These results may provide new strategies to
further improve the performance of organic electronic devices
by altering the electron accepting strength of terminal acceptor
moieties in D–A molecules.
H. Gao and Y. Song, J. Am. Chem. Soc., 2007, 129, 11674; (b)
Q. D. Ling, E. T. Kang, K. G. Neoh, Y. Chen, X. D. Zhuang,
C. X. Zhu and D. S. H. Chan, Appl. Phys. Lett., 2008, 92,
143302; (c) X.-D. Zhuang, Y. Chen, B.-X. Li, D.-G. Ma,
B. Zhang and Y. Li, Chem. Mater., 2010, 22, 4455; (d)
Y.-K. Fang, C.-L. Liu, G.-Y. Yang, P.-C. Chen and
W.-C. Chen, Macromolecules, 2011, 44, 2604.
Acknowledgements
This work was nancially supported by the Chinese Natural 11 W.-Y. Lee, T. Kurosawa, S.-T. Lin, T. Higashihara, M. Ueda
Science Foundation (21076134, 21176164 and 21206102), NSF and W.-C. Chen, Chem. Mater., 2011, 23, 4487.
of Jiangsu Province (BK2010208), a project of the Department of 12 (a) Q. D. Ling, Y. Song, S. L. Lim, E. Y. H. Teo, Y. P. Tan,
Education of Jiangsu Province (12KJB430011), Suzhou Nano-
project (ZXG2012023) and Project supported by the Specialized
Research Fund for the Doctoral Program of Higher Education of
China (Grant no. 20113201130003 and 20123201120005). A
C. X. Zhu, D. S. H. Chan, D. L. Kwong, E. T. Kang and
K. G. Neoh, Angew. Chem., Int. Ed., 2006, 45, 2947; (b)
Y. Liu, Y. Zheng, Q. Lan, S. Liu, Z. Qin, L. Chen, C. Zhao,
Z. Chi, J. Xu and J. Economy, Chem. Mater., 2012, 24, 1212.
Project Funded by the Priority Academic Program Development 13 M. Matsui, M. Suzuki, M. Hayashi, K. Funabiki, Y. Ishigure,
of Jiangsu Higher Education Institutions (PAPD).
Y. Doke and H. Shiozaki, Bull. Chem. Soc. Jpn., 2003, 76, 607.
14 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, T. Vreven
Jr, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar,
J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima,
Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox,
H. P. Hratchian, 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,
P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador,
J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich,
A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,
Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski,
B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-
Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe,
P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong,
C. Gonzalez and J. A. Pople, Gaussian 03, revision B.04,
Gaussian, Inc., Wallingford, CT, 2004.
Notes and references
1 (a) J. C. Scott and L. D. Bozano, Adv. Mater., 2007, 19, 1452;
(b) Q. Ling, D. Liaw, C. Zhu, D. Chan, E. Kang and
K. Neoh, Prog. Polym. Sci., 2008, 33, 917.
2 (a) J. Liu, Z. Yin, X. Cao, F. Zhao, L. Wang, W. Huang and
H. Zhang, Adv. Mater., 2013, 25, 233; (b) J. Fang, H. You,
J. Chen, J. Lin and D. Ma, Inorg. Chem., 2006, 45, 3701; (c)
J. Lin and D. Ma, Appl. Phys. Lett., 2008, 93, 093505; (d)
F. Zhao, J. Liu, X. Huang, X. Zou, G. Lu, P. Sun, S. Wu,
W. Ai, M. Yi, X. Qi, L. Xie, J. Wang, H. Zhang and
W. Huang, ACS Nano, 2012, 6, 3027; (e) J. Liu, Z. Zeng,
X. Cao, G. Lu, L.-H. Wang, Q.-L. Fan, W. Huang and
H. Zhang, Small, 2012, 8, 3517.
3 (a) H. Li, Q. F. Xu, N. J. Li, R. Sun, J. F. Ge, J. M. Lu, H. W. Gu
and F. Yan, J. Am. Chem. Soc., 2010, 132, 5542; (b) S. F. Miao,
H. Li, Q. F. Xu, Y. Y. Li, S. J. Ji, N. J. Li, L. H. Wang, J. W. Zheng
and J. M. Lu, Adv. Mater., 2012, 24, 6210; (c) S. F. Miao, H. Li,
Q. F. Xu, N. J. Li, J. W. Zheng, R. Sun, J. M. Lu and C. M. Li,
J. Mater. Chem., 2012, 22, 16582.
4 (a) C. Q. Ye, Q. Peng, M. Z. Li, J. Luo, Z. M. Tang, J. Pei, 15 (a) C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785;
J. M. Chen, Z. G. Shuai, L. Jiang and Y. L. Song, J. Am.
Chem. Soc., 2012, 134, 20053; (b) S.-J. Liu, P. Wang,
(b) B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys.
Lett., 1989, 157, 200.
Q. Zhao, H.-Y. Yang, J. Wong, H.-B. Sun, X.-C. Dong, 16 (a) A. Bandyopadhyay, S. Sahu and M. Higuchi, J. Am. Chem.
W.-P. Lin and W. Huang, Adv. Mater., 2012, 24, 2901.
5 S. G. Hahm, N.-G. Kang, W. Kwon, K. Kim, Y.-G. Ko, S. Ahn,
Soc., 2011, 133, 1168; (b) Y. Zhu, V. A. P. Kulkarni, P.-T. Wu
and S. A. Jenekhe, Chem. Mater., 2008, 20, 4200.
B.-G. Kang, T. Chang, J.-S. Lee and M. Ree, Adv. Mater., 2012, 17 (a) K. Balakrishnan, A. Datar, W. Zhang, X. M. Yang,
24, 1062.
T. Naddo, J. L. Huang, J. M. Zou, M. Yen, J. S. Moore and
This journal is ª The Royal Society of Chemistry 2013
J. Mater. Chem. C, 2013, 1, 3816–3824 | 3823

View Article Online
Journal of Materials Chemistry C
Paper
L. Zang, J. Am. Chem. Soc., 2006, 128, 6576; (b) S. Loser,
C. J. Bruns, H. Miyauchi, R. P. Ortiz, A. Facchetti,
V. Hernadez and J. T. Lopez Navarrete, Int. J. Quantum
Chem., 2005, 104, 635.
S. I. Stupp and T. J. Marks, J. Am. Chem. Soc., 2011, 133, 19 P. Cias, C. Slugovc and G. Gescheidt, J. Phys. Chem. A, 2011,
8142. 115, 14519.
18 (a) J. A. Marsden, J. J. Miller, L. D. Shirtcliff and M. M. Haley, 20 G. Liu, B. Zhang, Y. Chen, C.-X. Zhu, L. Zeng, D. S.-H. Chan,
J. Am. Chem. Soc., 2005, 127, 2464; (b) R. P. Ortiz,
K.-G. Neoh, J. Che and E.-T. Kang, J. Mater. Chem., 2011, 21,
R. M. Osuna, M. C. R. Delgado, J. Casado, S. A. Jenekhe,
6027.
3824 | J. Mater. Chem. C, 2013, 1, 3816–3824
This journal is ª The Royal Society of Chemistry 2013