Battery-inspired, nonvolatile, and rewritable memory architecture: A radical polymer-based organic device

Article abstract of DOI:10.1021/ja075553p

A nonvolatile, bistable, and rewritable organic memory device based on radical polymers was prepared and tested. The excellent performance of a battery-inspired memory architecture with a configuration of p- and n-type charge-transporting radical polymers sandwiching a dielectric layer was characterized. The ON-OFF ratio was more than 4 orders of magnitude, and retention and endurance cycles of more than 10⁴ and 10³, respectively, were accomplished. Copyright

Full text of DOI:10.1021/ja075553p

Published on Web 10/27/2007
Battery-Inspired, Nonvolatile, and Rewritable Memory Architecture: a Radical
Polymer-Based Organic Device
Yasunori Yonekuta, Kentaro Susuki, Kenichi Oyaizu, Kenji Honda, and Hiroyuki Nishide*
Department of Applied Chemistry, Waseda UniVersity, Tokyo 169-8555, Japan
Received July 25, 2007; E-mail: nishide@waseda.jp
Organic robust polyradicals are establishing their position as a
new class of functional polymers, such as ferromagnetic materials,
based on the spin alignment of unpaired electrons through
conjugated backbones.1 An intriguing aspect of the current research
is the application of radical polymers, i.e., aliphatic polymers
bearing redox-active radical pendant groups, to high capacity
charge-storage materials for secondary batteries.2 The high power
characteristics of the so-called “radical battery” originate from the
large heterogeneous electron-transfer rate of the redox centers and
the efficient mass-transfer process within the polymer layers,
allowing facile accommodation of electrolyte ions to compensate
charges generated from the neutral radicals. Nernstian electrochemi-
cal behaviors have been found for a number of robust radicals,
such as the 1e- oxidation of 2,2,6,6-tetramethylpiperidine-1-oxyl
(TEMPO) to the oxoammonium cation and the 1e- reduction of
galvinoxyl radicals to the galvinolate anion.3 Radical polymers
bearing pendant TEMPO and galvinoxyl groups undergo heavy p-
and n-doping and are used as cathode- and anode-active materials
in the radical battery, respectively, by sandwiching an electrolyte
layer. We anticipated that a similar charge-storage configuration
should develop for a dry system in the absence of the electrolyte
layer, by sandwiching a dielectric material with the radical
polymers. The expected result is an electroconductive bistability,
rather than the power storage. Here we report the excellent
properties of an organic “radical memory”, which provided insight
into the mechanism of charge storage at an electrode interface under
dry conditions.
Figure 1. Radical polymer-based memory architecture and charge injection,
transporting, and trapping mechanism at the PVDF interfaces, and energy
level diagram for the ITO, PTMA, PGSt, and Al.
The organic-based memory was first introduced for FeRAM-
type devices, wherein fluorine-containing polymers were used as
the dielectric material to replace silicon dioxide in combination with
inorganic semiconductor-based transistors.4 All plastic-type memory
systems have attracted recent attention as a new type of low voltage-
driven memory between conventional DRAM/SRAM devices and
magnetic hard discs, in addition to the inherent advantages allowing
for a facile wet-process fabrication. Several types of organic
materials have been examined for this purpose, such as organic
semiconductors,5-7 charge-transfer complexes,8,9 redox-active com-
pounds, and metal nanoparticle-dispersed thin films.10,11 However,
the organic-based devices often fell into write-once-read-many type
memory characteristics and suffered from a low ON-OFF ratio
remaining at less than 2 orders of magnitude. We selected structure-
defined radical polymers which yielded contamination-free, homo-
geneous, and tough organic layers for improvement of the device
behaviors.
The metal-insulator-metal diode-type structure of the fabricated
radical memory is shown in Figure 1. The device was composed
of the thin layers of poly(2,2,6,6-tetramethylpiperidine-1-oxyl
methacrylate) (PTMA) as the p-type redox active material, poly-
vinylidene difluoride (PVDF) as the dielectric material, and poly-
(4-(2,6-di-tert-butyl-R-(3,5-di-tert-butyl-4-oxo-2,5-cyclohexadien-
1-ylidene)-p-tolyloxy)styrene) (PGSt) as the n-type redox activate
Figure 2. (a) I-V characteristics of the device with an ITO/PTMA/PVDF/
PGSt/Al configuration. Inset: Analysis of I-V characteristics for the device
in the ON (b) and OFF (O) states. The ON- and OFF-state responses were
fitted for the Poole-Frenkel emission and Schottky effects, respectively.
(b) Retention cycle tests for current density at -1 V after pulse applications
at the write voltage of -7 V (blue circles) and the erase voltage of 5 V
(red circles). Inset: Endurance test results during the application of write-,
read-, and erase-voltage cycles.
material. These layers were conveniently spin-coated onto an ITO/
glass electrode from solutions of PTMA in ethyl lactate, PVDF in
hexamethylphosphoramide, and PGSt in toluene. The polymer
multilayer was successfully formed without substantial intermixing,
by virtue of the significant difference in solubility. An AFM image
of the PTMA layer on the ITO substrate (Figure S1) revealed a
highly smooth and homogeneous surface with a mean roughness
of 1 nm, although the ITO surface had a larger roughness of ca. 10
nm. The effects of the depletion defects and pinholes are thus likely
to be excluded for the polymer-based devices. The top Al electrode
had a width of 5 mm, so that the device had a measured area of 25
mm2.
The typical I-V characteristics of the device are shown in Figure
2a. When an increasing voltage of 0 to -5 V was applied to the
Al electrode versus the ITO electrode (sweep 1), the state of the
device was precipitously switched to low resistance near -4.5 V
9
14128
J. AM. CHEM. SOC. 2007, 129, 14128-14129
10.1021/ja075553p CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
log(I/V) versus V1/2 relationship, suggesting that the PF emission
was responsible for the charge transfer (see Figure 2a). In the ON
state, charges were trapped at the radical polymer/PVDF interfaces,
which induced accumulation of the opposite charges at the radical
polymer/electrode interfaces and reduced the Schottky barriers,
allowing charges to transfer across the radical polymer/electrode
interfaces even at low voltages.
The SOMO levels of PTMA and PGSt were 5.2 and 4.2 eV,
respectively. The PTMA layer accepts holes from the ITO, and
electrons are injected into the PGSt layer from Al. The injected
holes and electrons are transported by the hopping mechanism and
are stored at the radical polymer/PVDF interfaces. Under open
circuit conditions, the trapped holes and electrons were nonvolatile
due to not only the charge trapping at the radical polymer/PVDF
interfaces but also the blocking of the charge transfer in the PTMA
and PGSt layers having sufficient resistances. The asymmetric
configuration of the electrodes with the different work functions
allows initialization of the device by applying an inverse voltage.
In conclusion, the battery-inspired PTMA/PVDF/PGSt config-
uration led to the unique electroconductive bistability that functions
as the memory device. Our preliminary impedance experiments have
suggested that the device operates at a frequency of up to 105 Hz
indicating a high rate of switching for the ON-OFF transition,
which are the topics of our continuous research.
Figure 3. (a) I-V characteristics of the rectifying device with a config-
uration of ITO/PTMA/PGSt/Al (see Figure S3). (b) Retention cycle tests
by applying the pulses in the order of VW and VR (ON state, b), or VE and
VR (OFF state, O), consecutively. Inset: Energy level diagram for the
rectifying device.
as the threshold voltage. The low-resistance state was maintained
during the reverse sweep of the voltage from -5 to 1.4 V (sweeps
2 and 3 in Figure 2a). The low-resistance or ON state was observed
for repeated sweeps, regardless of the sweep direction (Figure S2a).
When a bias of ca. 1.5 V was applied, the device sharply switched
to the high-resistance or OFF state again (sweep 4). In the hysteretic
curve, the ON-OFF ratio amounted to 4 orders of magnitude. The
optimum write (VW), erase (VE), and read voltages (VR) were
determined to be -7, 5, and -1 V, respectively. The results of the
retention and endurance cycle tests for switching between the ON
and OFF states are shown in Figure 2b. The current for the OFF
state after the VE pulse for 0.1 s, detected by the VR pulse with a
current of 10-3 µA, was significantly smaller than that for the ON
state after the VW pulse for 0.1 s (100 µA). The retention cycles of
the ON and OFF states under open-circuit conditions persisted for
more than 104 times. Furthermore, each state survived for a month
after the corresponding once-time-only pulse as well as consecutive
pulses. Endurance tests were similarly performed by consecutively
applying the pulses in the order of VW, VR, VE, and VR. The slight
decay of the ON-OFF ratio might be caused by slow degradation
of the device. However, the repeatable performance was confirmed
for more than 103 cycles.
Control experiments using a PVDF-free device with a config-
uration of ITO/PTMA/PGSt/Al (Figure 3a) revealed the origin of
the ON-state stability. The negative bias applied to the pristine
highly resistant device induced a similar transition to the low-
resistant state (sweeps 1 and 2). However, the resulting ON state
was quenched by applying any positive bias (sweep 3). The ON-
state current gradually decreased and converged to that of the OFF
state after 50 sweeps (Figure 3b). A plausible mechanism based
on these result is a rectification effect of the p-n junction where
charge consumption by recombination prevails at the PTMA/PGSt
interface (Figure 3b, inset). The lack of any current from the
positively biased device (sweeps 4 and 5) suggests the absence of
the trapped charges in the polymer layers.
Acknowledgment. This work was partially supported by Grants-
in-Aid for Scientific Research (Nos. 19105003, 17067017) and the
Global COE Program from MEXT, Japan.
Supporting Information Available: Materials, measurements,
synthesis, and additional experiments. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) (a) Magnetic Properties of Organic Materials; Lahti, P. M., Ed.; Marcel
Decker: New York, 1999. (b) Rajca, A.; Wongsriratanakul, J.; Rajca, S.
Science 2001, 294, 1503. (c) Carbon-Based Magnetism; Makaroba, T.,
Palacio, F., Ed.; Elsevier: Amsterdam, 2006. (d) Kaneko, T.; Makino,
T.; Miyaji, H.; Teraguchi, M.; Aoki, T.; Miyasaka, M.; Nishide, H. J.
Am. Chem. Soc. 2003, 125, 3554. (e) Fukuzaki, E.; Nishide, H. J. Am.
Chem. Soc. 2006, 128, 996.
(2) (a) Nishide, H.; Suga, T. Electrochem. Soc. Interface 2005, 14, 32. (b)
Suga, T.; Konishi, H.; Nishide, H. Chem. Commun. 2007, 17, 1730. (c)
Suga, T.; Pu, Y.-J.; Kasatori, S.; Nishide, H. Macromolecules 2007, 40,
3167.
(3) Yonekuta, Y.; Oyaizu, K.; Nishide, H. Chem. Lett. 2007, 36, 866.
(4) Stikeman, A. Tech. ReV. 2002, 105, 31.
(5) Mo¨ller, S.; Perlov, C.; Jackson, W.; Taussig, C.; Forrest, S. R. Nature
2003, 426, 166.
(6) Tondelier, D.; Lmimouni, K.; Vuilaume, D. Appl. Phys. Lett. 2004, 85,
5763.
(7) Bozano, L. D.; Kean, B. W.; Deline, V. R.; Salem, J. R.; Scott, J. C.
Appl. Phys. Lett. 2004, 84, 607.
(8) (a) Potember, R. S.; Poehler, T. O.; Cowan, D. O. Appl. Phys. Lett. 1979,
34, 405. (b) Oyamada, T.; Tanaka, H.; Matsushige, K.; Sasabe, H.; Adachi,
C. Appl. Phys. Lett. 2003, 83, 1252.
(9) (a) Ling, Q. D.; Song, Y.; Ding, S. J.; Zhu, C. X.; Chan, D. S. H.; Kwong,
D. L.; Kang, E. T.; Neoh, K. G. AdV. Mater. 2005, 17, 455. (b) Song, Y.;
Ling, Q. D.; Zhu, C. X.; Kang, E. T.; Chan, D. S. H.; Wang, Y. H.; Kwong,
D. L. IEEE Electron. DeVice Lett. 2006, 27, 154. (c) Ling, Q. D.; Song,
Y.; Lim, S. L.; Teo, E. Y. H.; Tan, Y. P.; Zhu, C. X.; Chan, D. S. H.;
Kwong, D. L.; Kang, E. T.; Neoh, K. G. Angew. Chem., Int. Ed. 2006,
45, 2947.
(10) Gofer, Y.; Sarker, H.; Killian, J. G.; Poehler, T. O.; Searson, P. C. Appl.
Phys. Lett. 1997, 71, 1582.
(11) Tseng, R. J.; Huang, J.; Ouyang, J.; Kaner, R. B.; Yang, Y. Nano Lett.
2005, 5, 1077.
The Schottky effect and the Poole-Frenkel (PF) emission are
anticipated for the charge injection at the radical polymer/electrode
interface and the charge transfer within the radical polymer layer,
respectively.12-14 As shown in Figure 2a, the plots of log(I) versus
V1/2 were linear in the high-resistance range of 0 to -4.5 V before
the electrical transition, which indicated that the OFF state was
dominated by the Schottky barrier or the charge-injection process.
On the other hand, the I-V plots for the ON state obeyed a linear
(12) Laurent, C.; Kay, E.; Souag, N. J. Appl. Phys. 1988, 64, 336.
(13) Vollmann, W.; Poll, H.-U. Thin Solid Films 1975, 26, 201.
(14) Chu, C. W.; Ouyang, J.; Tseng, J.-H.; Yang, Y. AdV. Mater. 2005, 17,
1440.
JA075553P
9
J. AM. CHEM. SOC. VOL. 129, NO. 46, 2007 14129