The fullerene-based derivative phenyl-C61-butyric acid methyl ester (PCBM) has attracted attention as an n-type transistor channel and an organic photovoltaic material due to its solution processability, stability, and high electron mobility. However, disorder in solution-processed PCBM thin films makes understanding of charge carrier transport mechanisms difficult. We have fabricated PCBM-based field effect transistors and improved device performance by reducing electron trapping at the dielectric–semiconductor interface and improving charge carrier injection. Single-wall carbon nanotubes were successfully integrated into conventional gold electrodes to improve injection efficiency. Meyer–Neldel behavior has been investigated in a study of charge carrier transport in PCBM thin films. Crystallization of PCBM was observed during the measurement. We find that charge carriers hop between molecules in mostly amorphous regions from 145 to 240 K. The transport is facilitated by the formation of more crystallized regions when samples are heated as high as 240 K.

Organic semiconductors have been intensively investigated for applications such as organic light-emitting diodes, organic thin-film transistors (OTFTs), and organic photovoltaic cells (OPVs).1 Fullerene-based derivatives such as phenyl-C61-butyric acid methyl ester (PCBM) have attracted particular attention.2 They exhibit high electron mobility of 10−2 to 10−5 cm2/Vs,3,4 approaching hole mobilities in organic semiconductors. They are solution processable and display desirable morphology when blended with other common donors.5,6 PCBM is soluble in the same organic solvents as some p-type semiconductors, e.g., poly(3-hexylthiophene-2,5-diyl). This simplifies the preparation of blends for the fabrication of OPVs and OTFTs. Compared to fullerene (C60), PCBM has a lower lowest unoccupied molecular orbital (LUMO) level, facilitating electron injection from metal electrodes and p-type semiconductors. Although it is a prototype n-type small molecule, widely used as the electron acceptor in OPVs, the pristine material has been relatively little studied. Solution-processed thin films of PCBM are usually structurally and energetically disordered due to the high concentrations of defects (owing to the inherent properties of organic materials, inclusion of impurities, and solvent molecules).7 This leads to difficulty in understanding the charge carrier transport process. A correct description of these processes is essential for developing thermally stable OPVs and OTFTs.1,8,9

It is well established that various physical processes, including charge transport in disordered organic electronic devices,10,11 are temperature activated and follow the Arrhenius exponential law:12 
X = X 0 e Δ E / k B T ,
(1)
where X is the activated measured or calculated property, X 0 is the activation prefactor, Δ E is the activation energy, k B is the Boltzmann constant, and T is the temperature.
Since the 1920s, it has been found experimentally, even though the underlying mechanisms were not always clear, that for a related group of processes that are temperature activated, the prefactor X 0 obeys the relationship10,13
ln X 0 = ln X 00 + Δ E k B T i s o .
(2)
Combining Eq. (2) with Eq. (1), we obtain
ln X = ln X 00 + Δ E k B T i s o Δ E k B T ,
(3)
where X 00 is the activation prefactor and T i s o is the isokinetic temperature. Equation (3) is known as the Meyer–Neldel rule (MNR) since it was reported by Meyer and Neldel who investigated the conductivity of disordered materials.12 It is also known as the isokinetic law since X becomes independent of Δ E when T = T i s o. The MNR can be represented by a plot of ln X versus inverse temperature, which is called the Arrhenius plot. Lines corresponding to different Δ E intersect at T i s o. MNR has been observed in various systems, such as porous and amorphous silicon,14,15 microcrystalline silicon films,16 glassy materials,17 ionic conductors,18 and organic semiconductors.19,20 It can be used to understand the transport properties of organic electronic materials. In particular, it has been proposed10 that the Meyer–Neldel energy is
E M N = k B T i s o = υ l n κ ,
(4)
where υ is the energy of the excitations, normally phonons in inorganic solids, and infrared or Raman vibrations in organic solids, and κ is a coupling constant between electrons and phonons. Equation (4) may provide important insight into the study of organic semiconductors. For example, it has recently been shown that the proton conductivity in perovskite-type metal oxides can be enhanced by tuning a lattice vibration frequency toward a desirable T i s o.21 Furthermore, the correlation between X 00 and T i s o can be used to identify the conduction mechanism, i.e., polaronic conductors exhibit a positive correlation ( d X 00 / d T i s o > 0) while trap-limited conductors exhibit a negative correlation ( d X 00 / d T i s o < 0).17 

MNR is typically studied by preparing many samples in slightly different ways or closely related compositions. However, it is advantageous, if possible, to use an external parameter, such as pressure or electric field, to vary Δ E for the same sample.22 The organic field effect transistor (OFET) geometry can be used for the study of charge carrier transport (e.g., charge carrier mobility) properties of a semiconducting material. The energy distribution of carriers near the dielectric can then be changed by varying the gate voltage, Vg, thus changing Δ E.22 Here, we report such a study of PCBM.

The observation of MNR in PCBM has previously been reported by Fishchuk et al.23 and von Hauff.24 Fishchuk et al.23 studied the dependence of charge carrier mobility on T at a constant gate voltage, while von Hauff24 investigated charge injection with various metal contacts. Mg exhibited the best charge injection due to the doping of Mg into fullerene derivatives and to MgO bilayer formation at the interface of the metal contact and the semiconductor. The field effect carrier mobility of PCBM was studied by modulating Δ E by varying the gate voltage. However, measurements were reported for only two gate voltages, with extrapolation over different temperatures. Here, we report a more thorough study, covering a wide range of temperatures and voltages.

As reported in the literature,25 device performance in OFETs is limited by low charge injection efficiency, low charge carrier mobility, and electrode/semiconductor interface dynamics. Source–drain currents large enough to yield meaningful measurements require that these be maximized. In particular, charge injection from the metal to the semiconductor plays a major role in determining device performance.24 To gain insight into the fundamental aspects of charge carrier transport, we optimized the performance of OFETs taking care of two aspects: reducing electron trapping at the dielectric–semiconductor interface and increasing charge carrier injection through embedding 1D single-wall carbon nanotubes (SWCNTs) into the conventional gold electrode.4 Schottky barrier formation between a metal electrode and the LUMO or highest occupied molecular orbital (HOMO) level of the semiconductor (depending upon the sign of the charge carriers in the semiconductor) limits the injection efficiency of the transistors. Structural disorder and interfacial traps also affect injection efficiency. A strong electrostatic field at the electrode–semiconductor interface improves the injection efficiency. SWCNTs, with one end connected to a metal contact and the other to the semiconductor, provide such an electrostatic effect. SWCNT arrays improve the charge injection and increase the current (thus giving the appearance of increasing the bulk charge carrier mobility as compared to conventional Au electrodes).4 We have reduced trapping at the PCBM–dielectric interface by coating the dielectric with hexamethyldisilazane (HMDS).26 Both of these procedures led to the improvement of experimentally measured carrier mobility, with the use of SWCNTs exhibiting a greater effect. The linear mobilities of SWCNT-based OTFTs and Au-based OTFTs were 2 × 10−2 and 1 × 10−3 cm2/Vs.4 We studied MNR using optimized transistors in order to investigate charge carrier transport mechanisms in PCBM.

Bottom-contact, bottom-gate transistors were fabricated on highly doped silicon (gate) substrates coated with a silicon dioxide (SiO2) layer (200 nm) acting as the dielectric. Circular concentric source and drain, 5 nm titanium and 40 nm gold, were deposited by e-beam evaporation. Thin films of PCBM were spin-coated (1500 rpm, 100 s) from a 10 mg/ml solution in chlorobenzene under a N2 atmosphere. Prior to spin-coating, the solution was stirred overnight. W = 1500 μ m is the channel width, and L = 20 μ m is the channel length. Some of the devices were thermally treated on a hotplate at 50 °C for 2 h.

SWCNT powder (purchased from RAYMOR NanoIntegris) was sequentially purified using 3 M HNO3, 3 M NaOH, and HCl. SWCNTs were rinsed six times with de-ionized (DI) water to achieve a neutral final solution (pH ∼ 7.8, as indicated by the pH/conductivity meter) and collected by a polytetrafluoroethylene membrane filter. A SWCNT solution was prepared using the method previously reported by Cicoira et al.4 Briefly, 0.5% w/v sodium cholate and 5 mg SWCNT were dissolved in 400 ml of DI water and tip sonicated for 20 min. Subsequently, 4 ml of the solution was put in 80 ml of DI water, giving rise to a concentration of SWCNT ca. 6.25 × 10−4 mg/ml. The final solution was bath sonicated for another 30 min. SWCNTs were then transferred to an amino cellulose membrane filter (0.22 μm Triton-free MCE, purchased from Sigma Millipore) through vacuum filtration. The SWCNTs on the filter were then attached to a precleaned SiO2/Si substrate (i.e., sonication in deionized (DI) water, isopropyl alcohol (IPA), acetone, and UV ozone treatment) to make a SWCNT coating. The samples were immersed in a fresh acetone solution for 54 h to dissolve the membrane. They were then placed inside a tube furnace for thermal annealing at 350 °C for 5 h in air to remove the solution residue and improve the metal adhesion during electrode deposition. The substrate was then transferred to a clean room for further microfabrication. Electrodes were patterned as before. To make a SWCNT-free channel within the interelectrode distance, samples were exposed to O2 reactive ion etching (50 W RF power at 125 m Torr and 8 SCCM etching rate) for 1 min. PCBM films were then deposited and thermally treated as described in Sec. II A.

We have investigated the effect of thermal treatment, exposure to ambient air, and HMDS coating of the SiO2 gate dielectric upon electron trapping at its interface with the PCBM layer. This procedure is well established for many semiconductors.26 HMDS greatly reduces trapping, while thermal treatment and atmospheric exposure have lesser, variable effects. The transistors discussed below were fabricated on HMDS-coated oxide with thermal treatment and without atmospheric exposure.

Measurements of electrical properties were performed using a semiconductor parameter analyzer Agilent B1500A. During measurements, the source electrode was connected to the ground. Samples were measured in a micromanipulator cryogenic vacuum probe station with liquid nitrogen and copper heater to change the temperature. The pressure was maintained below 10−6 Pa. Before each measurement, the samples were allowed to stabilize at temperature for 1 h.

The surface topography of the samples was analyzed by atomic force microscopy under ambient conditions, using a Digital Instruments Dimension 3100 (Santa Barbara, CA) equipped with a Nanoscope V controller (Bruker). The electrodes with SWCNTs were investigated by scanning electron microscopy (SEM); the measurements were carried out at an accelerating voltage of 5 kV in backscattered electron and secondary electron imaging mode using a JEOL FEG-SEM. X-ray diffraction (XRD) spectra of the thin films were scanned by a Bruker D8 diffractometer, using 1.54 Å CuKα radiation.

We fabricated the electrodes with SWCNTs to improve charge carrier injection. The SWCNT electrodes consist of bundles of SWCNTs having one end in the channel and the other connected to the electrode. SEM images of SWCNT electrodes are shown in Fig. S4 in the supplementary material.35, Figure 1 shows the output curves of a PCBM thin film fabricated on a SWCNT-based device at room temperature. Compared to the device fabricated with gold electrodes (see Fig. S5 in the supplementary material35), the current of the SWCNT-based device is improved approximately three orders of magnitude, demonstrating the expected improvement in electron injection.

FIG. 1.

Drain current (Id) as a function of drain voltage (Vd) at several different gate voltages (Vg) for an optimized FET.

FIG. 1.

Drain current (Id) as a function of drain voltage (Vd) at several different gate voltages (Vg) for an optimized FET.

Close modal
By modifying the semiconductor–dielectric interface and improving the charge carrier injection, we optimized the devices for the study of charge carrier transport mechanisms in PCBM. Assuming ohmic contacts, the mobility, μ, is obtained from27 
I d = W C i L μ ( V g V t h V d 2 ) V d .
(5)

In Eq. (5), C i is the insulator capacitance per unit area, I d is the drain current, V d is the drain voltage with the source electrode connected to the ground, V g is the gate voltage, V t h is the threshold voltage, and V d V g V t h. In Fig. 2, we show ln μ versus 1/T. The activation energies, shown in Table I, are calculated from the slopes in Fig. 2, using Eq. (3). We plotted ln μ versus 1/T.

FIG. 2.

Arrhenius plot for PCBM-based OFET, with Vd = 10 V. T varies from 145 to 320 K.

FIG. 2.

Arrhenius plot for PCBM-based OFET, with Vd = 10 V. T varies from 145 to 320 K.

Close modal
TABLE I.

Activation energy of PCBM-based FETs as a function of Vg.

Ea (meV) T = 145–240 KEa (meV) T = 260–310 KEa (meV) T = 312–320 K
Vg = 30 V 168 377 383 
Vg = 32 V 243 264 303 
Vg = 34 V 156 201 254 
Vg = 36 V 149 167 202 
Vg = 38 V 147 150 166 
Vg = 40 V 145 142 175 
Ea (meV) T = 145–240 KEa (meV) T = 260–310 KEa (meV) T = 312–320 K
Vg = 30 V 168 377 383 
Vg = 32 V 243 264 303 
Vg = 34 V 156 201 254 
Vg = 36 V 149 167 202 
Vg = 38 V 147 150 166 
Vg = 40 V 145 142 175 

According to Eq. (3), the slopes of the experimental results are equal to −ΔE/kB. The activation energies obtained from the figure are presented in Table I. At T < 240 K, μ increases with Vg and depends exponentially upon T at a given Vg.

As we may see, the activation energies for PCBM thin films are considerably higher than excitation and thermal energies. This is in agreement with the multiexcitation entropy model and supports it as the explanation of the origin of MNR.10,28 In the low temperature (145–240 K) region, the charge carrier hops between PCBM molecules in a mostly amorphous thin film. Near 240 K, the low values of μ at low gate voltages increase more rapidly as a function of 1/T than at lower temperatures. This increase becomes more rapid at T = 310 K, with μ becoming less dependent upon Vg. We attribute this rapid increase of μ to an annealing effect: the partial crystallization of PCBM at an elevated temperature.29,30 In the high temperature (240–320 K) region, PCBM molecules can diffuse, increasing crystallinity.30 The increased crystallinity of the film facilitates electron hopping. This is confirmed by the measurements described below.

We have performed grazing XRD measurements before and after high T measurement. The results are shown in Fig. 3. The three diffraction peaks in the figure are at 2θ = 3.8° (corresponding to an interplanar distance of 2.3 nm), 5.2° (1.7 nm), and 6.5° (1.4 nm).31 According to the Scherrer relation,32 the mean size of the crystallites of these three peaks are 25.4, 2.8, and 1.4 nm, respectively. Compared with the sample before high T measurement, the peak at 3.8° is sharper, which gives ca. 1.6 nm larger crystallized regions in the thin film. The new peak located at 5.2° also indicates the crystallization process.

FIG. 3.

XRD patterns of a PCBM film deposited on a SWCNT/Au electrode-patterned, HMDS-treated SiO2/Si substrate.

FIG. 3.

XRD patterns of a PCBM film deposited on a SWCNT/Au electrode-patterned, HMDS-treated SiO2/Si substrate.

Close modal

We have measured μ vs T increased during three heating and cooling scans. The 2nd scan was carried out immediately after the 1st, and the 3rd, 12 h after the 2nd. As shown in Fig. 4, μ increased in both scans. The increase in the 2nd scan is attributed to the crystallization process. The further improvement in the 3rd scan is probably due to the full relaxation of PCBM molecules. In an earlier study,33 the device performance decreased slightly. This was attributed to the degradation of the material. However, as shown in Fig. 3, it increased during the entire measurement.

FIG. 4.

Effect of sequential scans in PCBM-based FET, Vd = 10 V, T = 295–380 K. Samples were kept in a vacuum probe station.

FIG. 4.

Effect of sequential scans in PCBM-based FET, Vd = 10 V, T = 295–380 K. Samples were kept in a vacuum probe station.

Close modal

The study of high-performance n-type organic semiconductors is essential to complementary logic circuit fabrication. There are three principal obstacles to usable organic n-type organic transistors. First, the noble metals, Au and Ag, frequently used as electrodes, have work functions that are more suitable for the injection of holes into the HOMO level, rather than electrons into the LUMO level, for typical organic semiconductors. While lower work function metals such as Mg and Ca have lower Schottky barriers, which can ensure better electron injection from the electrodes, they are more reactive with the organic semiconductors and are easily oxidized. Second, oligoacenes and oligothiophenes have relatively small electron affinities, and although the molecules can be chemically tuned with electron withdrawing side groups to transport electrons more efficiently, the final compounds are more sensitive to the atmosphere.34 Third, electrons are trapped at the semiconductor–dielectric interface by hydroxyl groups, present in the form of silanols in the case of the commonly used SiO2 dielectric. These three effects have rendered the n-type transistor elusive. Here, following the literature, we have coated HMDS on the SiO2 gate dielectric to decrease the electron trapping at the dielectric interface with the PCBM film. SWCNTs were successfully embedded in the conventional gold electrode to form a strong electrostatic field at the electrode–semiconductor interface to improve the electron injection efficiency.

The study of Meyer–Neldel behavior in the PCBM thin film shows that the charge carrier transport behavior changed at high temperature, and x-ray diffraction results indicate that the crystallization of PCBM small molecules occurred during the measurement. At low temperature, the charge carriers hop between the molecules in amorphous regions. At high temperature, the amorphous material is annealed, resulting in the formation of more crystallized regions. The crystallization is irreversible so that low temperature conduction is also improved.

A detailed study of the relaxation of PCBM small molecules and an investigation of the morphology and crystallinity of the thin film during the measurement may improve our understanding of its structure and electronic properties. It would be of interest to investigate the relaxation times or frequencies in the PCBM thin film, as they may contribute to the detailed charge carrier transport mechanisms in the PCBM film and the recombination behavior of the electron–hole pairs in OPVs. The detailed study of the structure and the packing of PCBM small molecules under different temperatures might also help to determine whether there is a stacking favorable to electron transport. Such understanding would contribute to the material synthesis of thermally stable OPVs and OTFTs.

In perspective, it has been shown4 that the PCBM thin films can be converted to ambipolar by vacuum annealing the samples at 400 K for several hours. The charge carrier transport properties of electrons and holes in PCBM and the comparison of charge transport mechanisms before and after the conversion of polarity of the thin film might be of interest. The aim of measuring the mobility of PCBM at higher temperatures without being affected by strong electron scattering might be achieved. It could help us to gain insight into a complete picture of charge carrier transport mechanisms in this small molecular organic semiconductor.

C.S. acknowledges financial support from NSERC (DG). We acknowledge the technical support of Yves Drolet.

The authors have no conflicts to disclose.

Zhaojing Gao: Data curation (lead); Investigation (lead); Writing – original draft (lead). Dieudonné Niyonkuru: Methodology (equal); Resources (lead); Writing – review & editing (equal). Arthur Yelon: Methodology (lead); Project administration (lead); Supervision (lead); Writing – review & editing (lead). Clara Santato: Funding acquisition (lead); Project administration (lead); Supervision (lead); Writing – review & editing (lead).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
C.
Groves
,
Rep. Prog. Phys.
80
,
026502
(
2016
).
2.
S. R.
Forrest
,
Organic Electronics: Foundations to Applications
(
Oxford University
,
Oxford
,
2020
).
3.
V. D.
Mihailetchi
,
J. K.
van Duren
,
P. W.
Blom
,
J. C.
Hummelen
,
R. A.
Janssen
,
J. M.
Kroon
,
M. T.
Rispens
,
W. J. H.
Verhees
, and
M. M.
Wienk
,
Adv. Funct. Mater.
13
,
43
(
2003
).
4.
F.
Cicoira
,
C. M.
Aguirre
, and
R.
Martel
,
ACS Nano
5
,
283
(
2011
).
5.
J. C.
Hummelen
,
B. W.
Knight
,
F.
LePeq
,
F.
Wudl
,
J.
Yao
, and
C. L.
Wilkins
,
J. Org. Chem.
60
,
532
(
1995
).
6.
D. L.
Cheung
and
A.
Troisi
,
J. Phys. Chem. C
114
,
20479
(
2010
).
7.
G.
Paterno
,
A. J.
Warren
,
J.
Spencer
,
G.
Evans
,
V. G.
Sakai
,
J.
Blumberger
, and
F.
Cacialli
,
J. Mater. Chem. C
1
,
5619
(
2013
).
8.
H.
Bässler
and
A.
Köhler
,
Unimolecular and Supramolecular Electronics I
(
Springer
,
New York
,
2011
), pp.
1
65
.
9.
V.
Coropceanu
,
J.
Cornil
,
D. A.
da Silva Filho
,
Y.
Olivier
,
R.
Silbey
, and
J.-L.
Brédas
,
Chem. Rev.
107
,
926
(
2007
).
10.
A.
Yelon
,
B.
Movaghar
, and
R.
Crandall
,
Rep. Prog. Phys.
69
,
1145
(
2006
).
11.
P.
Borsenberger
,
L.
Pautmeier
, and
H.
Bässler
,
J. Chem. Phys.
94
,
5447
(
1991
).
12.
W.
Meyer
and
H.
Neldel
,
Phys. Z.
38
,
1014
(
1937
).
13.
W.
Linert
and
R.
Jameson
,
Chem. Soc. Rev.
18
,
477
(
1989
).
14.
Y.
Lubianiker
and
I.
Balberg
,
J. Non-Cryst. Solids
227
,
180
(
1998
).
15.
M.
Kondo
,
Y.
Chida
, and
A.
Matsuda
,
J. Non-Cryst. Solids
198
,
178
(
1996
).
16.
S. K.
Ram
,
S.
Kumar
,
R.
Vanderhaghen
, and
P. Roca
i Cabarrocas
,
J. Non-Cryst. Solids
299
,
411
(
2002
).
17.
K.
Shimakawa
and
F.
Abdel-Wahab
,
Appl. Phys. Lett.
70
,
652
(
1997
).
18.
K.
Ngai
,
Solid State Ion.
105
,
231
(
1998
).
19.
E.
Meijer
,
M.
Matters
,
P.
Herwig
,
D.
De Leeuw
, and
T.
Klapwijk
,
Appl. Phys. Lett.
76
,
3433
(
2000
).
20.
A.
Pivrikas
,
M.
Ullah
,
T. B.
Singh
,
C.
Simbrunner
,
G.
Matt
,
H.
Sitter
, and
N.
Sariciftci
,
Org. Electron.
12
,
161
(
2011
).
21.
P.
Du
,
N.
Li
,
X.
Ling
,
Z.
Fan
,
A.
Braun
,
W.
Yang
,
Q.
Chen
, and
A.
Yelon
,
Adv. Energy Mater.
12
,
2102939
(
2022
).
22.
Z.
Gao
,
M.
Reali
,
A.
Yelon
, and
C.
Santato
,
Mater. Adv.
3
,
7815
(
2022
).
23.
I.
Fishchuk
,
A.
Kadashchuk
,
J.
Genoe
,
V.
Poroshin
, and
H.
Bässler
,
Mol. Cryst. Liq. Cryst.
535
,
1
(
2011
).
24.
E.
von Hauff
, “
Field effect investigations of charge carrier transport in organic semiconductors
,”
Ph.D. thesis
(Universität Oldenburg,
2006
).
25.
F.
Cicoira
and
C.
Santato
,
Organic Electronics: Emerging Concepts and Technologies
(
Wiley
,
Weinheim
,
2013
).
26.
L.-L.
Chua
,
J.
Zaumseil
,
J.-F.
Chang
,
E. C.-W.
Ou
,
P. K.-H.
Ho
,
H.
Sirringhaus
, and
R. H.
Friend
,
Nature
434
,
194
(
2005
).
27.
C. D.
Dimitrakopoulos
and
D. J.
Mascaro
,
IBM J. Res. Dev
45
,
11
(
2001
).
28.
A.
Yelon
,
Monatsh. Chem.
144
,
91
(
2013
).
29.
Y.
Zhang
,
H.-L.
Yip
,
O.
Acton
,
S. K.
Hau
,
F.
Huang
, and
A. K.-Y.
Jen
,
Chem. Mater.
21
,
2598
(
2009
).
30.
X.
Yang
,
J. K.
van Duren
,
M. T.
Rispens
,
J. C.
Hummelen
,
R. A.
Janssen
,
M. A.
Michels
, and
J.
Loos
,
Adv. Mater.
16
,
802
(
2004
).
31.
T.
Lan
,
F.
Soavi
,
M.
Marcaccio
,
P.-L.
Brunner
,
J.
Sayago
, and
C.
Santato
,
Chem. Commun.
54
,
5490
(
2018
).
32.
T.
Erb
,
U.
Zhokhavets
,
G.
Gobsch
,
S.
Raleva
,
B.
Stühn
,
P.
Schilinsky
,
C.
Waldauf
, and
C. J.
Brabec
,
Adv. Funct. Mater.
15
,
1193
(
2005
).
33.
H.
Sirringhaus
,
Adv. Mater.
21
,
3859
(
2009
).
34.
C. R.
Newman
,
C. D.
Frisbie
,
D. A.
da Silva Filho
,
J.-L.
Brédas
,
P. C.
Ewbank
, and
K. R.
Mann
,
Chem. Mater.
16
,
4436
(
2004
).
35.
See the supplementary material online for details of the study, S1–S3, and the increase of current after HMDS treatment.

Supplementary Material