New fullerene-based polymers and their electrical memory characteristics

Article abstract of DOI:10.1021/ma5021402

Covalent incorporations into polymers of fullerene were achieved via the Cu(I)-catalyzed azide-alkyne click polymerizations of a fullerene derivative monomer functionalized with 5-(trimethylsilyl)pent-4-yn-1-yl groups and a comonomer functionalized with a

Full text of DOI:10.1021/ma5021402

Article
pubs.acs.org/Macromolecules
New Fullerene-Based Polymers and Their Electrical Memory
Characteristics
Yong-Gi Ko,† Suk Gyu Hahm,† Kimie Murata,‡ Young Yong Kim,† Brian J. Ree,† Sungjin Song,†
Tsuyoshi Michinobu,*,‡ and Moonhor Ree*,†
†Department of Chemistry, Division of Advanced Materials Science, Center for Electro-Photo Behaviors in Advanced Molecular
Systems, Pohang Accelerator Laboratory, Polymer Research Institute, and BK School of Molecular Science, Pohang University of
Science & Technology, Pohang 790-784, Republic of Korea
‡Department of Organic and Polymeric Materials and Global Edge Institute, Tokyo Institute of Technology, 2-12-1-S8-24 Ookayama,
Meguro-ku, Tokyo 152-8552, Japan
S
* Supporting Information
ABSTRACT: Covalent incorporations into polymers of
fullerene were achieved via the Cu(I)-catalyzed azide−alkyne
click polymerizations of a fullerene derivative monomer
functionalized with 5-(trimethylsilyl)pent-4-yn-1-yl groups
and a comonomer functionalized with azidomethyl groups,
producing the novel fullerene polymers P1-C60 and P2-C60.
Despite their extremely high fullerene loading levels, the
polymers were soluble in common organic solvents and
exhibited no aggregation of fullerene units in films. Moreover,
devices containing these fullerene polymers were easily
fabricated with common coating processes that exhibit
excellent unipolar and bipolar flash memory characteristics as
well as unipolar permanent memory characteristics, with high ON/OFF current ratios, long retention times, and low power
consumption. These electrical switching behaviors were favorably operated by electron injection. Overall, these devices are the
first n-type bipolar and unipolar digital polymer memory devices which can be operated in flash and write-once-read-many-times
modes.
Another possible approach in developing high performance
materials is to covalently incorporate fullerene molecules into
polymers as a side group component and/or a backbone
component and/or an end group component, producing
fullerene polymers.4,9−11 Some fullerene-based polymer
systems have been reported, although a certain level of
difficulties was associated in the syntheses.4,9−11 Most of
fullerene-based polymers, however, were also found to have
solubility problem in solvents; higher fullerene content
aggravated the solubility issue.4 Among the reported fullerene
polymers, only two systems were introduced for applications in
electrical memory devices.10,11 One system is poly(N-vinyl-
carbazole) bearing fullerene covalently (PVK-C60).10 The PVK-
C60 polymer was prepared by the direct coupling reaction of
poly(N-vinylcarbazole) and fullerene with the aid of sodium
hydride, and indeed, the incorporation of fullerene was limited
to only 1 mol % with respect to the total number of repeat units
in the polymer backbone. The films (ca. 50 nm thick) of PVK-
C60 in devices with aluminum (Al) and indium−tin-oxide
(ITO) electrodes were found to reveal bipolar flash memory
INTRODUCTION
■
Since the discovery of fullerene1 in 1985 and synthetic scheme
producing macroscopic quantity2 in 1990, there has been a
huge effort in understanding the properties and structure of the
interesting molecule.3,4 In particular, the versatile electrical and
optical properties of fullerene3,4 led to the development of high
performance materials including fullerene as a component or
solely consisting of fullerene.4−11
One possible approach developing high performance
materials is the use of fullerene as an ingredient in the
production of polymer composites via physical mixing. A
number of fullerene/polymer composite systems have been
reported.4−7 In particular, some of the composite systems were
prepared and investigated for applications in electrical memory
devices.6,7 Severe aggregations of fullerene molecules in the
composite systems, however, caused unusually low stability in
the associated memory devices.6,7 Fullerene aggregation,
attributed to fullerene’s insolubility in most organic solvents
and immiscibility in most polymers, is a common phenomenon
observed among fullerene/polymer composites, even in
composites with low loading levels of fullerene. Thus, to
overcome these drawbacks, improvements in solubility and
miscibility of fullerene have currently been attempted via
chemical modifications.3,4,8
Received: October 20, 2014
Revised: November 10, 2014
Published: November 18, 2014
© 2014 American Chemical Society
8154
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
Figure 1. Synthetic schemes of (a) P1-C60 and (b) P2-C60. Memory devices with top and bottom electrodes fabricated with P1-C60 and P2-C60. (c)
Schematic diagrams of the memory device cells. Optical images of memory device cells with various sizes (0.5 × 0.5, 1.0 × 1.0, and 1.5 × 1.5 mm2)
fabricated on a high quality glass substrate (d).
behavior, which was similar to one of the memory character-
istics (bipolar flash and write-once-read-many-times (WORM)
memories) observed for films of poly(N-vinylcarbazole) itself
(i.e., the mother polymer).12 Overall, its electrical memory
behavior was mainly attributed to the mother polymer rather
than the fullerene; perhaps the incorporated fullerene could
assist the observed memory behavior in a certain level, but its
role could not be understood well. The other polymer system is
poly(2-(N-carbazolyl)ethyl methacrylate) whose one end is
capped with fullerene (PCzMA-C60), which was synthesized via
living anionic polymerization in a well-controlled manner.11
The PCzMA-C60 films (60 nm thick) in devices exhibited
bipolar and unipolar flash memory behaviors as well as WORM
memory behavior. In comparison, the mother polymer films
revealed unipolar flash and WORM memory behaviors. These
results collectively suggest that the fullerene incorporated as a
minor component plays a role in tuning electrical memory
characteristics of the mother polymer in positive ways.
However, the observed nonvolatile memory behaviors of
PCzMA-C60 films seem to be mainly driven by the mother
polymer rather than the fullerene. Hence, the development of
fully fullerene-based polymers with high-performance electrical
memory characteristics remains at early stages.
carboxylate)) (P1-C60) and poly(1,2,3-triazol-1,4-diylmethy-
lene-1,3-(5-methoxycarbonyl-2,4,6-tris(dodecyloxy)-
phenylene)methylene-1,2,3-triazol-1,4-diyl(diethyldipropylene
bismethano[60]fullerene-61,61,62,62-tetracarboxylate)) (P2-
C60) (Figure 1a,b). Despite these extremely high fullerene
loading levels, the polymers were found to reveal good
solubility in common organic solvents as well as no fullerene
aggregation. Moreover, the associated memory devices were
easily fabricated via spin-coating, and exhibited excellent flash-
type memory characteristics as well as permanent memory
characteristics, with high ON/OFF current ratios, long
retention times, and low power consumption. Remarkably,
these devices exhibited both bipolar and unipolar switching
behaviors. Excitingly, these are the first n-type bipolar and
unipolar digital polymer memory devices; their nonvolatile
memory characteristics might be driven by electron injection
rather than hole injection.
EXPERIMENTAL SECTION
■
Synthesis. A fullerene-containing monomer, bis(5-(trimethylsilyl)-
pent-4-yn-1-yl) 1,2-methano[60]fullerene-61,61-dicarboxylate (1), was
prepared according to the method reported previously.13 To a 1000
mL flask, fullerene (C60: 595 mg, 0.826 mmol), bis(5-(trimethylsilyl)-
pent-4-yn-1-yl) malonate (314 mg, 0.825 mmol), iodine (I2: 320 mg,
1.26 mmol), and dry toluene (600 mL) were placed at 20 °C under
argon. To this solution, 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU:
308 μL, 2.06 mmol) was added, and then the reaction solution was
stirred at 20 °C for 22 h. After filtration through a short plug [silicate
(SiO2), methylene chloride (CH2Cl2)], the solvent was evaporated.
In this study, two novel fullerene-based polymers were
elegantly designed and successfully synthesized by using click
chemistry: poly(1,2,3-triazol-1,4-diylmethylene-1,3-(5-methox-
ycarbonyl-2,4,6-tris(dodecyloxy)phenylene)methylene-1,2,3-tri-
azol-1,4-diyl(dipropylene 1,2-methano[60]fullerene-61,61-di-
8155
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
Column chromatography (SiO2, hexane → hexane:CH2Cl2 (= 1:9,
v:v) yielded the desired product as a brown solid (508 mg, 56%). The
product was characterized by means of nuclear magnetic resonance
(NMR) and Fourier-transform infrared (FTIR) spectroscopies and
matrix-assisted laser desorption/ionization time-of-flight mass
(MALDI-TOF MS) spectrometry. NMR spectra were recorded on a
JEOL AL300 spectrometer with proton (1H) and/or carbon (13C)
probe; tetramethylsilane (TMS) with the solvent’s residual signal was
used as an internal reference. IR spectra were measured on a
spectrometer (model FT/IR-4100, JASCO). MALDI-TOF MS
spectroscopy measurements were conducted on a mass spectrometer
(model AXIMA-CFR, Shimadzu) using dithranol as a matrix. A
solution of sample (1 g/L), silver trifluoroacetate (1 g/L), and
dithranol (20 g/L) in tetrahydrofuran (THF) was placed on a sample
filtration, the filtrate was evaporated. Column chromatography (SiO2,
hexane:chloroform (CHCl3) (= 1:1, v:v)) yielded methyl 2,4,6-
tris(dodecyloxy)benzoate as a colorless oil (2.05 g, 99%). To a 20 mL
flask, methyl 2,4,6-tris(dodecyloxy)benzoate (345 mg, 0.500 mmol)
and zinc chloride (ZnCl2: 136 mg, 1.00 mmol) were placed in
nitrogen. Chloromethyl methyl ether (0.52 mL, 6.95 mmol) was
added, and the mixture was stirred at 20 °C for 23 h. After the addition
of ice−water, the mixture was stirred for 30 min. The organic layer was
extracted with ether, washed with water, and dried over Na2SO4. After
filtration, the filtrate was evaporated. Column chromatography (SiO2,
hexane:CHCl3 (= 1:1, v:v) followed by purification using using a
recycling preparative HPLC (Japan Analytical Industry, LC-9204,
CHCl3) yielded methyl 3,5-bis(chloromethyl)-2,4,6-tris(dodecyloxy)-
benzoate as a colorless oil (188 mg, 48% yield). The obtained product
was characterized by NMR and IR spectroscopies. 1H NMR (δ (ppm),
300 MHz, CDCl3): 0.88 (t, J = 6 Hz, 9 H), 1.27−1.57 (m, 54 H),
1.74−1.92 (m, 6 H), 3.91 (s, 3 H), 4.05 (t, J = 4.5 Hz, 4 H), 4.18 (t, J
= 4.5 Hz, 2 H), 4.66 (s, 4 H). IR (neat, cm−1): 2921.6, 2852.2, 1735.6,
1583.3, 1464.7, 1437.7, 1379.8, 1319.1, 1262.2, 1202.4, 1186.0, 1163.8,
1116.6, 1089.6, 998.9, 908.3, 721.2, 666.3.
1
plate, and the solvent was evaporated. H NMR (δ (ppm), 300 MHz,
deuterated chloroform (CDCl3)): 0.17 (s, 18 H), 2.07 (q, J = 7 Hz, 4
H), 2.44 (t, J = 7 Hz, 4 H), 4.60 (t, J = 6 Hz, 4 H), where s, d, t, q, and
m denote singlet, doublet, triplet, quartet, and multiplet, respectively.
IR (neat, cm−1): 2955.4, 2173.4, 1744.3, 1460.8, 1427.1, 1390.4,
1317.1, 1265.1, 1229.4, 1202.4, 1185.0, 1111.8, 1052.0, 1025.9, 901.6,
837.0, 756.9, 728.0, 701.0, 638.3, 575.6. MALDI-TOF MS (dithranol,
m/z): calcd for C79H30O4Si2: 1098.17 g/mol; found: 1098.69 g/mol
[M+].
Another fullerene-containing monomer, bis(ethyl(5-
(trimethylsilyl)pent-4-yn-1-yl) bismethano[60]fullerene-61,61,62,62-
tetracarboxylate (2), was synthesized as follows. To a 500 mL flask,
5-(trimethylsilyl)pent-4-yn-1-ol (1.83 mL, 10 mmol), dehydrated
CH2Cl2 (133 mL), and dehydrated pyridine (0.81 mL) were placed,
and the solution was cooled to 0 °C under a nitrogen atmosphere.
After ethyl 3-chloro-3-oxopropanoate (1.27 mL, 10 mmol) was added,
the solution was stirred at 0 °C for 1 h followed by at 20 °C for 20 h.
The precipitate was filtered off, and the filtrate was evaporated.
Column chromatography (SiO2, CH2Cl2) afforded the desired
compound (2.41 g, 8.91 mmol) in 89.1% yield. The obtained ethyl
5-(trimethylsilyl)pent-4-yn-1-yl propanedioate product was identified.
1H NMR (δ (ppm), 300 MHz, CDCl3): 0.11 (s, 9 H), 1.25 (t, J = 6
Hz, 3 H), 1.79−1.87 (m, 2 H), 2.29 (t, J = 6 Hz, 2 H), 3.33 (s, 2 H),
4.13−4.23 (m, 4 H). 13C NMR (δ (ppm), 75 MHz, CDCl3): 0.12,
14.11, 16.51, 27.60, 41.65, 61.59, 64.10, 85.48, 105.54, 166.56, 166.58.
IR (neat, cm−1): 2960.2, 2902.3, 2175.3, 1735.6, 1411.6, 1369.2,
1328.7, 1249.7, 1184.1, 1147.4, 1070.3, 1033.7, 941.1, 840.8, 759.8,
698.1, 640.3.
The obtained methyl 3,5-bis(chloromethyl)-2,4,6-tris(dodecyloxy)-
benzoate (574 mg, 0.730 mmol) and sodium azide (189.8 mg, 2.92
mmol) were added to dehydrated DMF (3.0 mL) in a 50 mL flask at
20 °C and followed by stirring for 31.5 h. After the reaction was
completed, water was added to the solution, and then the organic layer
was extracted with ethyl acetate. The organic layer was again washed
with water and dried over Na2SO4. After filtration, the filtrate was
evaporated. Column chromatography (SiO2, hexane → hexane:ethyl
acetate (= 100:1, v:v)) afforded the desired compound (561.4 mg,
0.7025 mmol) in 96.2% yield. The product was characterized by NMR
1
and IR spectroscopies as well as MALDI-TOF MS spectrometry. H
NMR (δ (ppm), 300 MHz, CDCl3): 0.88 (t, J = 6 Hz, 9 H), 1.27−1.51
(m, 54 H), 1,70−1,89 (m, 6 H), 3.89−3.98 (m, 9 H), 4.38 (s, 4 H).
13C NMR (δ (ppm), 75 MHz, CDCl3): 14.03, 22.64, 25.85, 25.90,
29.31, 29.35, 29.38, 29.52, 29.54, 29.59, 29.61, 29.66, 30.14, 30.23,
31.88, 44.14, 52.44, 76.09, 118.74, 119.02, 157.48, 160.20, 166.49. IR
(neat, cm−1): 2922.6, 2853.2, 2089.5, 1735.6, 1586.2, 1464.7, 1431.9,
1378.9, 1339.3, 1296.9, 1265.1, 1201.4, 1178.3, 1116.6, 1101.2, 1003.8,
907.3, 722.2. MALDI-TOF MS (dithranol, m/z): calcd for
C46H82N6O5: 798.63 g/mol; found: 821.87 g/mol [M + Na]+.
As shown in Figure 1a, P1-C60 was synthesized from the
polymerization of the monomers 1 and 3. To a 100 mL flask, the
monomer 1 (1.1 g, 1.0 mmol) and THF (25 mL) were placed, and the
solution was stirred at 0 °C. After (nC4H9)4NF solution in THF (1 M,
2.5 mL) was added, the solution was stirred at 0 °C for 3 h. Water was
added, and the organic layer was extracted with CH2Cl2. The organic
layer was washed with water again and dried over Na2SO4. After
filtration, the filtrate was concentrated to ca. 25 mL by evaporation. To
this solution, the comonomer 3 (799.2 mg, 1.0 mmol), water (25 mL),
CuSO4·5H2O (25.0 mg, 1.0 mmol), and sodium ascorbate (59.4 mg,
1.0 mmol) were added, and the mixture was stirred at 20 °C for 24 h
under nitrogen. The solution was concentrated by evaporation and
poured into methanol (CH3OH). The precipitate was collected by
filtration and dried in vacuo. This solid was washed with hexane for 7 h
by using a Soxhlet extractor and then extracted with CHCl3. The
solution was concentrated by evaporation and poured into a mixture of
hexane:CHCl3 (= 1:1, v:v), yielding the target polymer as a brown
solid (420.7 mg). The product was characterized by NMR and IR
spectroscopies. 1H NMR (δ (ppm), 300 MHz, CDCl3): 0.88 (br s, 9n
H), 1.25 (br s, 54n H), 2.04−2.46 (m, 14n H), 3.92 (br s, 6n H), 4.11
(s, 3n H), 4.47 (br s, 4n H), 4.63 (t, J = 6.3 Hz, 4n H). IR (neat,
cm−1): 3296.7, 2920.7, 2849.3, 1742.4, 1578.5, 1458.9, 1427.1, 1376.9,
1292.1, 1227.5, 1112.7, 1052.0, 1020.2, 693.3, 638.3.
The obtained ethyl 5-(trimethylsilyl)pent-4-yn-1-yl propanedioate
(450 mg, 2.50 mmol) was added together with C60 (600 mg, 0.833
mmol) and I2 (646 mg, 2.55 mmol) to toluene (600 mL) in a 1000
mL flask, and the solution was stirred at 20 °C under argon. After
DBU (0.72 mL, 4.17 mmol) was added, the solution was stirred at 20
°C for 22 h. The solution was passed through a plug (SiO2, CH2Cl2),
and the solvents were evaporated. Repeated column chromatography
(SiO2, hexane → CH2Cl2 and hexane → hexane:CH2Cl2 (= 1:1, v:v))
afforded the desired compound in the form of regioisomeric mixtures
(456 mg, 0.363 mmol) in 43.6% yield. The target product was
characterized by NMR and IR spectroscopies as well as MALDI-TOF
1
MS spectrometry. H NMR (δ (ppm), 300 MHz, CDCl3): 0.15−0.20
(m, 18 H), 1.39−1.63 (m, 6 H), 1.98−2.08 (m, 4 H), 2.38−2.55 (m, 4
H), 4.43−4.74 (m, 8 H). IR (neat, cm−1): 2956.3, 2173.4, 1741.4,
1430.9, 1388.5, 1230.4, 1099.2, 1058.7, 1020.2, 837.0, 756.0, 730.9,
702.0. MALDI-TOF MS (dithranol, m/z): calcd for C86H40O8Si2:
1256.23 g/mol; found: 1279.71 g/mol [M + Na]+.
A comonomer, methyl 3,5-bis(azidomethyl)-2,4,6-tris(dodecyloxy)-
benzoate (3), was synthesized as follows. Methyl 3,5-bis-
(chloromethyl)-2,4,6-tris(dodecyloxy)benzoate was prepared as de-
scribed in the literature.8,14 To a 50 mL flask, methyl 2,4,6-
trihydroxybenzoate (553 mg, 3.00 mmol), potassium carbonate
(K2CO3: 2.65 g 19.2 mmol), and dry dimethylformamide (DMF:
3.0 mL) were placed, and the mixture was stirred under nitrogen. To
this suspension, 1-bromododecane (2.3 mL, 9.2 mmol) was added,
and the mixture was heated to 70 °C for 18 h. After cooling to 20 °C,
water was added. The organic layer was extracted with ether, washed
with water and brine, and dried over sodium sulfate (Na2SO4). After
From the monomers 2 and 3, (P2-C60) was prepared (Figure 1b).
To a 100 mL flask, the monomer 2 (1.18 g, 0.938 mmol) and THF
(23 mL) were placed, and the solution was stirred at 0 °C. After (n-
C4H9)4NF solution in THF (1 M, 2.3 mL) was added, the solution
was stirred at 0 °C for 3 h. Water was added, and the organic layer was
extracted with CH2Cl2. The organic layer was washed with water again
and dried over Na2SO4. After filtration, the filtrate was concentrated to
8156
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
ca. 23 mL by evaporation. To this solution, the comonomer 3 (749.6
mg, 0.938 mmol), water (23 mL), CuSO4·5H2O (23.4 mg, 0.938
mmol), and sodium ascorbate (55.8 mg, 0.938 mmol) were added and
then stirred at 20 °C for 24 h under a nitrogen atmosphere. The
solution was concentrated by evaporation and poured into CH3OH.
The precipitate was collected by filtration and dried in vacuo. This
solid was washed with hexane by using a Soxhlet extractor and then
extracted with CHCl3. The solution was concentrated and poured into
a mixture of CH3OH:CHCl3 (= 2:1, v:v), yielding the target polymer
product as a dark brown solid (1.03 g). The product was characterized
by NMR and IR spectroscopies. 1H NMR (δ (ppm), 300 MHz,
CDCl3): 0.88 (br s, 9n H), 1.27 (br s, 54n H), 1.41−1.75 (m, 6n H),
2.00−2.36 (m, 14n H), 3.91−3.96 (m, 9n H), 4.38−4.53 (m, 12n H);
IR (neat, cm−1): 2919.7, 2850.3, 2366.2, 2337.3, 1737.6, 1577.5,
1457.9, 1367.3, 1290.1, 1222.7, 1093.4, 1014.4, 705.8, 669.2, 630.6.
Measurements. Molecular weight analysis was performed with gel
permeation chromatography (GPC) combined with multiangle light
scattering (MALS); a Shodex GPC system equipped with polystyrene
gel columns and a detector miniDAWN Tristar were employed. THF
as an eluent was used at a rate of 1.0 mL/min and 40 °C.
Thermogravimetric analysis (TGA) was conducted with a ramping
rate of 10.0 °C/min using a thermogravimeter (model SII TG/DTA
6200, Seiko). Differential scanning calorimetry (DSC) was carried out
with a ramping rate of 10.0 °C/min using a calorimeter (model SII
DSC 6220, Seiko). During these measurements, dry nitrogen gas was
purged with a flow rate of 100 cm3/min. Ultraviolet−visible (UV−vis)
spectroscopy measurements were conducted using a Scinco
spectrometer (model S-3100). Cyclic voltammetry (CV) measure-
ments were conducted with a scan rate of 100 mV/s in THF
containing 0.1 M (n-C4H9)4NClO4 at 20 °C under an argon
atmosphere using a cell composed of three-electrodes (a platinum
(Pt) working electrode with a diameter of 1.6 mm, a reference
electrode (Ag/Ag+/CH3CN/(n-C4H9)4NClO4), and a Pt auxiliary
electrode); in the measurements, a ferrocene/ferricinium (Fc/Fc+)
couple was used as internal standard.
Device Fabrication. Devices of the polymers with aluminum (Al)
top electrode and ITO or Al or gold (Au) bottom electrode were
fabricated. ITO glasses were used as a substrate with ITO bottom
electrode. Al or Au was deposited with 300 nm thickness onto silicon
substrates with thermally grown oxide layer by electron beam
sputtering; the deposited Al or Au layer was used as a bottom
electrode. The polymer solutions were spin-cast onto ITO glass
substrates or metal-deposited Si substrates and followed by drying in
vacuo at 100 °C for 30 min. Then, the Al top electrodes (300 nm
thick) were deposited onto the polymer film layers through a shadow
mask by thermal evaporation in vacuum, with a size of 0.5 × 0.5, 1.0 ×
1.0, and 1.5 × 1.5 mm2. Current−voltage (I−V) analysis was
conducted under ambient air at room temperature, using a
semiconductor analyzer (model 4200, Keithley).
RESULTS AND DISCUSSION
■
P1-C60 was successfully prepared in THF by the polymerization
of in situ deprotected 1 with 3 via copper(I)-catalyzed azide−
alkyne click polymerization (Figure 1a). In a similar manner,
P2-C60 was synthesized from 2 and 3 (Figure 1b). The obtained
polymer products were characterized with H and 13C NMR
1
and IR spectroscopies as well as with GPC-MALS. P1-C60 was
found to have a weight-average molecular weight ( Mw) of 79
800 and a polydispersity index (PDI) of 1.20. P2-C60 was found
to have Mw = 76 500 and PDI = 1.53.
Both P1-C60 and P2-C60 exhibited good solubilities in THF,
CH2Cl2, CHCl3, and toluene, despite their extremely high
fullerene content. These good solubilities mean that the
polymers exhibit excellent film formability, affording high-
quality nanoscale thin films with conventional solution
processes including simple spin-coating. Soluble polymers
with high fullerene content are generally known to be very
elusive because the strong aggregation tendencies of fullerene
moieties significantly suppress polymer solubility,4,17 so these
good solubility results are a significant achievement. The
solubilities of these polymers are attributed to the positive
contributions of the three n-dodecyloxy groups and the three
ester linkers per polymer repeat unit. In particular, the three n-
dodecyloxy groups significantly improve the polymers’
solubilities because of their high solubility in organic solvents;
they are also good steric stabilizers of the fullerene moieties.
The arrangement of the three n-dodecyloxy groups at the 1,3,5-
benzene unit also effectively prevents the aggregation of the
fullerene moieties.
Thin Film Preparation. Each polymer was dissolved in CHCl3
with a concentration of 20 mg/mL and then filtered through
polytetrafluoroethylene (PTFE) membrane microfilters with a pore
size of 0.45 μm at room temperature. The polymer solutions were
spin-cast on substrates such as silicon (Si) wafers and indium−tin
oxide (ITO) glass substrates, followed by drying in vacuo at 100 °C for
30 min. The film thicknesses were measured by spectroscopic
ellipsometry (model M2000, Woollam). The film surface morphology
was measured in tapping mode by means of atomic force microscopy
(AFM: model Multimode AFM Nanoscope IIIa, Digital Instruments);
silicon cantilevers with a 42 N/m spring constant and 330 kHz
resonance frequency were used.
Synchrotron Grazing Incidence X-ray Scattering (GIXS).
Grazing incidence wide-angle and small-angle X-ray scattering
(GIWAXS and GISAXS) analyses were performed with a two-
dimensional (2D) charge-coupled detector (CCD: MAR USA) at the
3C and 9A beamlines of Pohang Accelerator Laboratory (PAL),
according to the method reported in the literature.15 Measurements
were conducted with an incident angle αi of 0.160°, using an X-ray
radiation source (λ = 0.1380 nm wavelength). The sample-to-detector
distance was 121.1 mm for GIWAXS and 2201.5 mm for GISAXS. All
GIXS measurements with a collection time of 60 s were carried out at
room temperature. The scattering angle correction of the measured
GIXS data was made using the reflected X-ray beam position as well as
polystyrene-b-polyethylene-b-polybutadiene-b-polystyrene block co-
polymer and silver behenate powder standards.15
Synchrotron X-ray Reflectivity (XR). XR analysis was conducted
in θ−2θ scanning mode at the 3D and 8D POSCO beamlines16 of
PAL. An X-ray radiation source (λ = 0.1541 nm) was used, and the
beam size at the sample position was 0.1 × 2.0 mm2. The reflectivity R
was measured down to just above 10−8; here the reflected beam
intensity was normalized to the incident beam intensity. The measured
data had undergone the geometrical correction and background
subtraction procedure described in the literature.16
P1-C60 and P2-C60 were found to start degradation at 210 °C
(= Td) in a nitrogen atmosphere (Figure S1a in Supporting
Information). However, their glass transitions could not be
discernible below 200 °C in DSC analysis (Figure S1b). The
UV−vis spectroscopy analysis found that P1-C60 and P2-C60
reveal a band gap (i.e., the difference between the highest
occupied molecular orbital (HOMO) level and the lowest
unoccupied molecular orbital (LUMO) level) of 3.20 and 3.10
eV, respectively (Figures S2a and S3a). The reduction halfwave
potential ERed of P1-C60 was determined to be as −0.89 V vs Fc/
Fc+; the ERed of P2-C60 was −1.03 V vs Fc/Fc+ (Figures S2b and
S3b). The LUMO level for the Fc/Fc+ standard is assumed to
be −4.80 eV with respect to the zero vacuum level. From the
band gap and CV data, the HOMO and LUMO levels were
estimated to be −7.11 and −3.91 eV for P1-C60 and −6.87 and
−3.77 eV for P2-C60, respectively.
Figure 2 shows representative GISAXS and GIWAXS
patterns of P1-C60 and P2-C60 in thin films (50−60 nm
thick). The GISAXS patterns of both the P1-C60 and P2-C60
films are featureless, indicating that neither of the films of the
8157
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
Figure 3. Representative XR profiles of P1-C60 and P2-C60 in thin films
(50−60 nm thick), which were prepared on silicon substrates with
native oxide layer by solution coating and subsequent drying. The
symbols are the measured data, and the solid line represents the fit
curve assuming a homogeneous electron density distribution within
the film. The inset shows a magnification of the region around the two
critical angles: αc,f and αc,s are the critical angles of the film and the
silicon substrate, respectively.
easily produce high quality films via common solution coating
process.
Memory devices were fabricated on ITO-glass and silicon
substrates, as shown in Figure 1c,d. Surprisingly, the P1-C60
devices with Al top and ITO bottom electrodes exhibit
excellent bipolar switching (i.e., flash type) memory character-
istics, as shown in Figure 4a. The P1-C60 film layer in the device
Figure 2. Synchrotron GIXS data of polymer films (50−60 nm thick):
(a) GISAXS and (b) GIWAXS patterns of P1-C60; (c) GISAXS and
(d) GIWAXS patterns of P2-C60; (e) out-of-plane scattering profiles
extracted along the αf direction at 2θf = 0.000° from the GIWAXS
patterns in (b) and (d); (f) in-of-plane scattering profiles extracted
along the 2θf direction at αf = 0.300° from the GIWAXS patterns in
(b) and (d). The wavelength λ of the X-ray beam was 0.1380 nm; the
incident angle αi of the X-ray beam was set at 0.160°.
polymers contains nanostructures. These scattering patterns
further confirm that aggregation of the fullerene units
incorporated into the polymer chains has been prevented
during the film formation process. The films further show
featureless GIWAXS patterns, again confirming the absence of
ordered phases; only three amorphous scattering rings are
discernible. Their d-spacings are 1.72 nm (3.70°), 1.01 nm
(6.30°), and 0.47 nm (13.60°) for the P1-C60 film and 1.93 nm
(3.30°), 1.10 nm (5.80°), and 0.46 nm (13.90°) for the P2-C60
film. Considering their chemical structures, the relatively strong
ring scatterings at 3.70° (1.72 nm d-spacing) and 3.30° (1.93
nm) for P1-C60 and P2-C60, respectively, might come from the
average interdistances of the fullerene units. The weak ring
scatterings at 6.30° (1.01 nm d-spacing) and 5.80° (1.10 nm)
for P1-C60 and P2-C60, respectively, might result from the
average interdistances of the polymer chains without fullerene
units, whereas the ring scatterings at 13.60° (0.47 nm d-
spacing) and 13.90° (0.46 nm) might originate from the mean
interdistances of the n-dodecyloxy side groups.
Figure 4. Bipolar I−V curves of Al/polymer(60 nm thick)/ITO
devices. I−V measurements were carried out during negative voltage
sweeps and reverse voltage sweeps: (a) P1-C60; (b) P2-C60. The
electrode contact area was 1.0 × 1.0 mm2.
Figure 3 depicts representative specular XR profiles of P1-C60
and P2-C60 in thin films (50−60 nm thick). The XR profiles
could be well analyzed using the algorithm proposed by
Parratt.16,18 The analysis results are summarized in Table S1.
The synchrotron XR analysis found that for the P1-C60 film the
electron density ρe and surface roughness σ are 471.6 nm−3 and
0.1 nm, respectively, whereas for the P2-C60 film ρe = 505.6
nm−3 and σ = 0.2 nm. The P1-C60 film exhibits relatively lower
ρe value than that of P2-C60. The σ values are slightly lower
than the root-mean-square (rms) roughnesses (0.2 nm for the
P1-C60 film and 0.3 nm for the P2-C60 film) measured by
atomic force microscopy (AFM; Figure S4). Collectively, the
low σ values again confirmed that both P1-C60 and P2-C60 can
is initially in an OFF-state (i.e., a low conductivity state). The
polymer device switches on to an ON-state (i.e., a high
conductivity state) at the critical voltage, Vc,ON, of −1.10 V,
during the first voltage sweep ranging from 0 to −3.5 V with a
compliance current of 0.01 A (1st sweep). The ON-state
persists even after the removal of the power supply as well as
during the second voltage sweep from 0 to −3.5 V with the
same compliance current. Additionally, the ON-state could not
be switched off by positive voltage sweeps with a compliance
current of 0.10 A, which is 10 times higher than that of the first
sweep (Figure S5a). The ON-state, however, switches off at
8158
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
+0.90 V (= Vc,OFF, the critical switch-OFF voltage) during the
reverse voltage sweep from −3.5 V to a positive voltage
(second sweep). The OFF-state persists during the positive
voltage sweep; moreover, the OFF-state could not be
transformed to an ON-state by any further positive voltage
sweeps. This whole bipolar switching process is repeatable, as
shown in the third and fourth sweeps. Overall, the P1-C60
device exhibits high ON/OFF current ratio (104−105). P2-C60
devices with Al top and ITO bottom electrodes exhibited
compliance current were observed for the devices whose
operation was initiated with a positive voltage sweep (Figure
5b): Vc,ON = +1.28 V, Vc,OFF = +0.63 V, and the ON/OFF
current ratio ranges from 105 to 109. Similar unipolar WORM
memory and ON−OFF switching behaviors were also
demonstrated for the P2-C60 devices fabricated with Al
electrodes (Figure 5c,d and Figure S6b): Vc,ON = −1.67 or
+1.40 V and Vc,OFF = −0.51 or +0.67 V; the ON/OFF current
ratios are in the range 107−1010.
similar bipolar switching behaviors: Vc,ON = −1.44 V, Vc,OFF
=
Similar unipolar WORM memory behavior and ON−OFF
+0.95 V, and the ON/OFF current ratios ranged from 104 to
switching behavior were demonstrated for the P1-C60 and P2-
107 (Figure 4b and Figure S5b).
C
60 devices with Al top and Au bottom electrodes (Figure S7);
Interestingly, the P1-C60 devices with Al electrodes exhibited
quite different memory behaviors from these bipolar flash
memory characteristics. As shown in Figure 5a, the device in an
P1-C60 devices: Vc,ON = −1.49 or +2.10 V and Vc,OFF = −0.50 or
+0.81 V; P2-C60 devices: Vc,ON = −2.05 or +1.80 V and Vc,OFF
=
−0.50 or +0.79 V. Their ON/OFF current ratios are in the
range 102−106.
To evaluate P1-C60 and P2-C60 devices as bipolar nonvolatile
memory devices, the write-read-erase cycles and the memory
retention times were measured. Figure 6 displays the
Figure 5. Unipolar I−V curves for Al/polymer(60 nm thick)/Al
devices: (a, b) P1-C60 devices; (c, d) P2-C60 devices. The first negative
voltage sweep was performed with a current compliance of 0.01 A
from 0 to −3.0 V to switch the device on, the second forward sweep
was performed with the same current compliance (0.01 A) to test the
ON-state of the device, and the third sweep was carried out with a
current compliance of 0.10 A to switch the device off. The first positive
voltage sweep was performed with a current compliance of 0.01 A
from 0 to +3.0 V to switch the device on, the second forward sweep
was performed with the same current compliance (0.01 A) to test the
ON-state of the device, and the third sweep was carried out with a
current compliance of 0.10 A to switch the device off. The electrode
contact area was 1.0 × 1.0 mm2.
OFF-state switches on to an ON-state at −2.10 V (= Vc,ON
)
Figure 6. Bipolar write-read-erase cycles of Al/polymer(60 nm thick)/
ITO devices: (a) voltages applied in the cycles (each single cycle: −2.5
V (write) → −1.0 V (read) → +2.0 V (erase) → −1.0 V (read)); (b)
current responses of the P1-C60 device during the cycles; (c) current
responses of the P2-C60 device during the cycles. The voltages were
applied during the cycles with a current compliance of 0.01 A. The
electrode contact area was 1.0 × 1.0 mm2.
during a negative voltage sweep with a current compliance of
0.01 A (first sweep). The ON-state remains even after the
elimination of power supply or during further forward voltage
sweep (second sweep). In addition, the ON-state could not be
returned to the OFF-state by any reverse voltage sweeps
(Figure S6a). These results collectively indicate that the P1-C60
devices fabricated with Al electrodes exhibit WORM memory
behavior when operated under a fixed compliance current.
Surprisingly the switched-ON P1-C60 devices can, however, be
undergone to the OFF-state by a voltage sweep with a current
compliance (0.10 A) higher than that set in the first switching-
ON process; Vc,OFF = −0.60 V in a negative voltage sweep (see
the third sweep). The ON/OFF current ratios range 105−109.
Similar WORM memory behavior with a fixed compliance
current and ON−OFF switching behavior with variation in the
characteristics of write-read-erase cycles, with writing voltage
of −3.0 V and erasing voltage of +2.0 V. The ON-state
(memory writing) and OFF-state (memory erasing) were
monitored through small probe voltage pulses of −1.0 V. The
current running through the device shows sensitive response to
continuous write-read-erase cycles for several hours. The write-
8159
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
read-erase cycles in unipolar switching cases were confirmed to
be stable as well (Figure 7).
stabilities (Figure 8a,c); the reading voltage of −1.5 V was used.
Both the bipolar and the unipolar P2-C60 devices also revealed
highly stable ON- and OFF-state over a test period of 1.0 × 104
to 4.0 × 104 s as observed for the P1-C60 devices (Figure 8b,d).
Moreover, it was confirmed that the P1-C60 and P2-C60
memory devices in both the bipolar and the unipolar modes
functioned without error after being kept in ambient air for 1.5
years (Figure 9), demonstrating high quality unipolar and
bipolar flash memory performances and outstanding reliabil-
ities.
Figure 7. Unipolar write-read-erase cycles of Al/polymer(60 nm
thick)/Al devices. P1-C60 devices: (a) unipolar write-read-erase cycles;
(b) switch-ON and OFF voltages (Vc,ON and Vc,OFF) and switch-OFF
(i.e., turn-OFF) currents versus voltage sweep cycles. P2-C60 devices:
(c) unipolar write-read-erase cycles; (d) switch-ON and OFF voltages
(Vc,ON and Vc,OFF) and switch-OFF (i.e., turn-OFF) currents versus
voltage sweep cycles. In every single cycle, the first sweep was
performed with a current compliance of 0.01 A from 0 to −3.0 V to
switch the device on, the second sweep was performed with the same
current compliance from 0 to −3.0 V to read or test the ON-state of
the device, and the third sweep was carried out with a current
compliance of 0.1 A to switch the device off.
The stabilities of the fullerene-based polymer devices were
further evaluated under ambient air at room temperature. As
shown in Figure 8a,c, the ON-states of both the bipolar and the
unipolar P1-C60 devices have held stable over a test period of
1.0 × 104 to 4.0 × 104 s without loss of integrity; the reading
voltage of −3.0 V was employed. In case of the OFF-states,
both the bipolar and the unipolar P1-C60 devices showed high
Figure 9. Reliability of the polymer devices: (a) bipolar Al/P1-C60(60
nm thick)/ITO and (b) Al/P2-C60(60 nm thick)/ITO devices; (c, d)
unipolar Al/P1-C60(60 nm thick)/Al and (e, f) Al/P2-C60(60 nm
thick)/Al devices. The devices used in the reliability test were stored in
air ambient for 1.5 years. The electrode contact area was 1.0 × 1.0
mm2.
For the P1-C60 and P2-C60 devices fabricated with Al top and
ITO bottom electrodes, which exhibit switching-ON behavior
in the first negative voltage sweeps, the energy barriers (0.37
and 0.51 eV) for electron injection into the LUMO levels from
the Al electrode (Φ(work function) = −4.28 eV) are lower than
those (2.31 and 2.07 eV) for hole injection into the HOMO
levels from the ITO electrode (Φ = −4.80 eV). In the first
negative voltage sweeps the conduction processes are thus
governed by electron injection. On the other hand, in the first
positive voltage sweeps, the energy barriers (2.83 and 2.59 eV)
for hole injection into the HOMO levels from the Al electrode
are higher than those (0.89 and 1.03 eV) for electron injection
into the LUMO levels from the ITO electrode. These facts
indicate that the conduction processes under the first positive
voltage sweeps are also driven by electron injection. However,
the energy barriers to electron injection are higher in the
positive voltage sweeps than in the negative voltage sweeps.
Thus, charge injection is more favorable during the negative
voltage sweeps, which results in the switching of the devices to
the ON-state; in contrast, the higher energy barriers in the
Figure 8. Long-time responses (i.e., retention times) of the ON- and
OFF-states of the polymer devices: (a) Al/P1-C60(60 nm thick)/ITO
and (b) Al/P2-C60(60 nm thick)/ITO devices in bipolar switching
mode; (c) Al/P1-C60(60 nm thick)/Al and (d) Al/P2-C60(60 nm
thick)/Al devices in unipolar switching mode.
8160
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
Figure 10. Switching mechanism proposed for the polymer memory devices. (a) Al/P1-C60(60 nm thick)/ITO or Al/P2-C60(60 nm thick)/ITO
device; (b, c) energy level diagrams of the polymer device before applying electric field; (d, e) energy level diagrams of the polymer device under
applying electric field. The purple dots containing yellow C60 in (a), (c), and (e) represent the fullerene moieties present in the polymer film layer of
the device. The schematic diagram in (c) additionally illustrates some fullerene moieties positioned locally across the film thickness in the highlighted
local area of the polymer film layer in (a), which can act as the charge-trapping sites and also serve as the stepping stones for the flow of the charge
carriers via hopping process; the schematic diagram in (e) shows the fullerene moieties being acted under applying electric field as the electron
trapping sites and being served as the stepping stones (that play together as an filament) for the flow of the charge carriers via hopping process. dtr in
(c) and (e) is the trap depth of the charge-trapping sites, which is a function of the induction and stabilization powers of the charge-trapping site.
positive voltage sweeps prevent the devices switching on. In
case of the P1-C60 and P2-C60 devices with Al top and bottom
electrodes exhibiting switching-ON behavior in negative and
positive voltage sweeps, the energy barriers (0.37 and 0.51 eV)
for electron injection into the LUMO levels are lower than
those (2.83 and 2.59 eV) for hole injection into the HOMO
levels. For the P1-C60 and P2-C60 devices with Al top and Au
bottom electrodes revealing unipolar switching behaviors,
electron injection into the LUMO levels is also more favorable
than hole injection into the HOMO levels. Thus, the
conduction processes in the devices with Al top and Al or Au
bottom electrodes are also dominated by electron injection.
These considerations collectively show that the P1-C60 and P2-
C60 devices with an Al top electrode and ITO, Al, or Au bottom
electrode can be favorably operated by electron injection and
are thus the first n-type nonvolatile polymer memory devices.
Additional analysis using various conduction models19 was
carried out to characterize the conduction mechanisms and
electrical switching behaviors of the polymer films. The OFF-
state I−V data were found to follow trap-limited space charge
limited conduction (SCLC) model (Figure S8). In comparison,
the ON-state data were characterized to follow the ohmic
conduction model (Figure S8). Furthermore, the switched ON
devices showed current levels which were independent of the
size of the device cell (Figures S9 and S10). These results
suggest that heterogeneous localized filaments (i.e., hopping
routes) might be formed under applied electric field. As a result,
trap-limited SCLC and hopping process govern the outstanding
switching behaviors of the P1-C60 and P2-C60 devices.
Taking into account the results described above, the
switching-ON processes of the P1-C60 and P2-C60 layers in
the devices can be understood by considering their charge-
trapping and morphological characteristics as follows. The
charge-trapping sites could originate from the chemical
compositions of the P1-C60 or P2-C60 chains, whose chemical
repeat units are composed of a fullerene moiety, a phenyl unit,
two triazole units, three or five ester linkers, and three ether
linkers, as shown in Figure 1a,b. The fullerene, phenyl, and
ester carbon units can play as electrophilic sites because of their
electron-deficient nature. In contrast, the triazole, ester oxygen,
and ether oxygen groups can function as nucleophilic sites
because of their electron-rich nature. For example, consider a
case where a voltage of current compliance of 0.01 A is applied
to the active polymer film layer. The fullerene moieties and the
other electrophilic units get enriched with electrons, thereby
becoming electron-trapping sites. At that instance, the
nucleophilic groups become enriched with holes, acting as
charge-trapping sites. Taking into consideration induction and
resonance effects, in the polymers the fullerene moiety may
have the highest charge trap and stabilization abilities than any
other groups. Hence, the fullerene moieties can act as major
charge-trapping sites under applied bias; they enable to flow
charge carriers at Vc,ON or higher. The resulting current,
however, is not a simple flow of charge carriers through the
entire polymer layer, but it rather occurs through hopping
process via the fullerene sites (which were dispersed with the
average interdistance of 1.04−1.93 nm in the film) as stepping
stones. Based on these considerations and results, a switching
8161
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
mechanism of the P1-C60 and P2-C60 devices is proposed, as
shown in Figure 10.
ACKNOWLEDGMENTS
■
This study was supported by the National Research Foundation
(NRF) of Korea (Doyak Program 2011-0028678 and Center
for Electro-Photo Behaviors in Advanced Molecular Systems
(2010-0001784)) and the Ministry of Science, ICT & Future
Planning (MSIP) and the Ministry of Education (BK21 Plus
Program and Global Excel Program). This work was also
supported by the Ministry of Education, Culture, Sports,
Science and Technology (MEXT), Japan (Challenging
Exploratory Research, No. 22655030), and by Tokyo Institute
of Technology (SATR program−Supports for Tokyo Tech
Advanced Researchers). The synchrotron X-ray scattering
measurements at the Pohang Accelerator Laboratory were
supported by MSIP, POSTECH Foundation, and POSCO
Company.
The switching-OFF processes of the P1-C60 and P2-C60
devices in their bipolar and unipolar modes involve two
different mechanisms. In the bipolar switching case, applying a
negative voltage with the same current compliance as the
switching-ON process destabilizes the fullerene groups’ ability
to maintain the charged state. Charge destabilization renders
fullerene moieties neutral, thereby decommissioning the
conducting pathways and switching off the device. On the
contrary, the excess current from a positive voltage bias with a
higher current compliance than the switching-ON process
generate heat generation,20 causing repulsive Coulomb
interactions among the trapped charges at the fullerene units
and other groups. Heat generation and repulsive Coulomb
interactions may rupture of the hopping routes made of charge-
trapped fullerene moieties inside the film and switch off the
device.
REFERENCES
■
(1) Kroto, H. W.; Heath, J. R.; O’Brien, S. C.; Curl, R. F.; Smalley, R.
F. Nature 1985, 318, 162−3.
CONCLUSIONS
■
(2) Kratschmer, W.; Lamb, L. D.; Fostiropoulos, K.; Huffman, D. R.
Nature 1990, 347, 354−8.
P1-C60 and P2-C60 were newly designed and synthesized, which
covalently contained fullerene in extremely high levels (61.2−
69.8 wt %). Despite the high fullerene contents, the polymers
demonstrated good solubility in common solvents and easy film
formability, which are generally required in the fabrication of
advance devices. Moreover, they exhibited no aggregation of
fullerene units even in the film state.
(3) (a) Haddon, R. C. Acc. Chem. Res. 1992, 25, 127−133. (b) Prato,
M. Top. Curr. Chem. 1999, 199, 173−84.
(4) Wang, C.; Guo, Z.-X.; Fu, S.; Wu, W.; Zhu, D. Prog. Polym. Sci.
2004, 29, 1079−1141.
(5) (a) Kim, Y.; Cook, S.; Tuladhar, S. M.; Choulis, S. A.; Nelson, J.;
Durrant, J. R.; Bradley, D. D. C.; Giles, M.; McCulloch, I.; Ha, C.-S.;
Ree, M. Nat. Mater. 2006, 5, 197−203. (b) Kim, Y.; Nelson, J.;
Durrant, J. R.; Bradley, D. D. C.; Heo, K.; Park, J.; Kim, H.;
McCulloch, I.; Heeney, M.; Ree, M.; Ha, C.-S. Soft Matter 2007, 3,
117−121. (c) Chung, D. S.; Park, J. W.; Kim, S.-O.; Heo, K.; Park, C.
E.; Ree, M.; Kim, Y.-H.; Kwon, S.-K. Chem. Mater. 2009, 21, 5499−
5507. (d) Nam, S.; Shin, M.; Kim, H.; Ha, C.-S.; Ree, M.; Kim, Y. Adv.
Funct. Mater. 2011, 21, 4527−4534.
The nanoscale thin films of both P1-C60 and P2-C60 in
devices revealed superior digital memory behaviors in unipolar
and bipolar modes under ambient air. The polymer films
additionally demonstrated highly stable permanent memory
behavior. These memory devices were able to operate with high
stability and reliability as well as with high ON/OFF current
ratios in low power consumption. These electrical switching
behaviors were favorably operated by electron injection.
Overall, we succeeded to develop two novel fullerene-based
polymers, P1-C60 and P2-C60, over limitations in synthesis.
Their physicochemical properties were quite significant over
the poor solubilities in solvents and the severe aggregations in
the fullerene units of previously reported fullerene-containing
materials. Moreover, the P1-C60 and P2-C60 devices exhibited
the first reported n-type unipolar and bipolar digital memory
characteristics with high performance and low power
consumption in flash and WORM modes.
(6) (a) Hsu, J.-C.; Liu, C.-L.; Chen, W.-C.; Sugiyama, K.; Hirao, A.
Macromol. Rapid Commun. 2011, 32, 528−533. (b) Kim, H. J.; Jung, J.
H.; Ham, J. H.; Kim, T. W. Jpn. J. Appl. Phys. 2008, 47, 5083−5085.
(c) Paul, S.; Kanwal, A.; Chhowalla, M. Nanotechnology 2006, 17,
145−151. (d) Majumdar, H. S.; Baral, J. K.; Laiho, A.; Ruokolainen, J.;
̈
Ikkala, O.; Osterbacka, R. Adv. Mater. 2006, 18, 2805−2806.
̈
(7) (a) Majumar, H. S.; Baral, J. K.; Osterbacka, R.; Ikkala, O.; Stubb,
H. Org. Electron. 2005, 6, 188−192. (b) Chu, C. W.; Ouyang, J.;
Tseng, J. H.; Yang, Y. Adv. Mater. 2005, 17, 1440−1443. (c) Gao, H.
J.; Sohlberg, K.; Xue, Z. Q.; Chen, H. Y.; Hou, S. M.; Ma, L. P.; Fang,
X. W.; Pang, S. J.; Pennycook, S. J. Phys. Rev. Lett. 2000, 84, 1780−
1783.
(8) (a) Hirsch, A.; Brettreich, M. Fullerenes: Chemistry and Reactions;
Wiley-VCH: Weinheim, Germany, 2005. (b) Michinobu, T.;
Nakanishi, T.; Hill, J. P.; Funahashi, M.; Ariga, K. J. Am. Chem. Soc.
2006, 128, 10384−10385.
(9) Fang, L.; Liu, P.; Sveinbjornsson, B. R.; Atahan-Evrenk, S.;
Vandewal, K.; Osuna, S.; Jimenez-Osees, G.; Shrestha, S.; Giri, G.;
Wei, P.; Salleo, A.; Aspuru-Guzik, A.; Grubbs, R. H.; Houkb, K. N.;
Bao, Z. J. Mater. Chem. C 2013, 1, 5747−5755.
(10) Ling, Q.-D.; Lim, S.-L.; Song, Y.; Zhu, C.-X.; Chan, D. S.-H.;
Kang, E.-T.; Neoh, K.-G. Langmuir 2007, 23, 312−319.
(11) Hahm, S. G.; Kang, N.-G.; Kwon, W.; Kim, K.; Ko, Y.-G.; Ahn,
S.; Kang, B.-G.; Chang, T.; Lee, J.-S.; Ree, M. Adv. Mater. 2012, 24,
1062−1066.
(12) (a) Lai, Y.-S.; Tu, C.-H.; Kwong, D.-L. Appl. Phys. Lett. 2005, 87,
122101. (b) Lin, J.; Ma, D. Appl. Phys. Lett. 2008, 93, 93505. (c) Lim,
S. L.; Ling, Q.; Teo, E. Y. H.; Zhu, C. X.; Chan, D. S. H.; Kang, E.-T.;
Neoh, K. G. Chem. Mater. 2007, 19, 5148−5157.
(13) Iehl, J.; Nierengarten, J.-F. Chem.Eur. J. 2009, 15, 7306−7309.
(14) (a) Zhang, W.; Horoszewski, D.; Decatur, J.; Nuckolls, C. J. Am.
Chem. Soc. 2003, 125, 4870−4873. (b) Michinobu, T.; Okoshi, K.;
ASSOCIATED CONTENT
■
S
* Supporting Information
TGA and DSC data, UV−vis spectra, CV data, AFM images,
and I−V data and analysis results. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail ree@postech.edu, Tel +82-54-279-2120, Fax +82-54-
279-3399 (M.R.).
*E-mail michinobu.t.aa@m.titech.ac.jp, Tel +81-3-5734-3774,
Fax +81-3-5734-3944 (T.M.).
Author Contributions
Y.-G.K. and S.G.H. equally contributed to this work.
Notes
The authors declare no competing financial interest.
8162
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163

Macromolecules
Article
Murakami, Y.; Shigehara, K.; Ariga, K.; Nakanishi, T. Langmuir 2013,
29, 5337−5344.
(15) (a) Lee, B.; Park, Y.-H.; Hwang, Y.; Oh, W.; Yoon, J.; Ree, M.
Nat. Mater. 2005, 4, 147−150. (b) Rho, Y.; Min, J.; Yoon, J.; Ahn, B.;
Jung, S.; Kim, K.; Shah, P.; Lee, J.-S.; Ree, M. NPG Asia Mater. 2012,
4, e29. (c) Rho, Y.; Ahn, B.; Yoon, J.; Ree, M. J. Appl. Crystallogr. 2013,
46, 466−475. (d) Kim, Y. Y.; Ahn, B.; Sa, S.; Jeon, M.; Roth, S. V.;
Kim, S. Y.; Ree, M. Macromolecules 2013, 46, 8235−8244. (e) Rho, Y.;
Kim, C.; Higashihara, T.; Jin, S.; Jung, J.; Shin, T. J.; Hirao, A.; Ree, M.
ACS Macro Lett. 2013, 2, 849−855. (f) Ree, M. Macromol. Rapid
Commun. 2014, 35, 930−959.
(16) (a) Bolze, J.; Ree, M.; Youn, H. S.; Chu, S.-H.; Char, K.
Langmuir 2001, 17, 6683−6691. (b) Kim, K.; Yen, H.-J.; Ko, Y.-G.;
Chang, C.-W.; Kim, D. M.; Kwon, W.; Hahm, S. G.; Liou, G.-S.; Ree,
M. Polymer 2012, 53, 4135−4144. (c) Ko, Y.-G.; Kwon, W.; Yen, H.-J.;
Chang, C.-W.; Kim, D. M.; Kim, K.; Hahm, S. G.; Lee, T. J.; Liou, G.-
S.; Ree, M. Macromolecules 2012, 45, 3749−3758. (d) Kim, K.; Fang,
Y.-K.; Kwon, W.; Pyo, S.; Chen, W.-C.; Ree, M. J. Mater. Chem. C
2013, 1, 4858−4868.
(17) Giacalone, F.; Martín, N. Adv. Mater. 2010, 22, 4220−4248.
(18) Parratt, L. G. Phys. Rev. 1954, 95, 359−369.
(19) (a) Campbell, A. J.; Bradley, D. D. C.; Lidzey, D. G. J. Appl.
Phys. 1997, 82, 6326−6342. (b) Jensen, K. L. J. Vac. Sci. Technol., B
2003, 21, 1528−1544. (c) Mark, P.; Helfrich, W. J. Appl. Phys. 1962,
33, 205−215. (d) Frenkel, J. Phys. Rev. 1938, 54, 647−648.
(e) Laurent, C.; Kay, E.; Souag, N. J. Appl. Phys. 1998, 64, 336−343.
(20) Dearnaley, G.; Stoneham, A. M.; Morgan, D. Rep. Prog. Phys.
1970, 33, 1129−1191.
8163
dx.doi.org/10.1021/ma5021402 | Macromolecules 2014, 47, 8154−8163