On the trade-off between processability and opto-electronic properties of single wall carbon nanotube derivatives in thin film heterojunctions

Article abstract of DOI:10.1039/c4tc01350f

A flow functionalization route has been employed to derivatize single wall carbon nanotubes (SWCNTs) by thienylphenyl groups. The SWCNT derivatives in the most soluble fraction have been characterized by thermogravimetric analysis, DLS analysis, DFT calculations, and UV-vis-NIR, microRaman and IR spectroscopies to study the degree of functionalization, the concentration of SWCNTs in solution, the dimension of the aggregates in solutions, the density of defects, and the presence of the thienylphenyl groups. Thin-film heterojunctions made of SWCNT derivatives and poly(3-hexylthiophene) (P3HT) have been prepared by various methods employing the Langmuir-Schaefer technique, spin-coating and thermal annealing processes. By comparing the batch and the flow functionalizations, a trade-off between solubility, processability and the thin-film opto-electronic properties has been found as a result of the degree of functionalization.

Full text of DOI:10.1039/c4tc01350f

Journal of
Materials Chemistry C
View Article Online
View Journal
PAPER
On the trade-off between processability and
opto-electronic properties of single wall carbon
Cite this: DOI: 10.1039/c4tc01350f
nanotube derivatives in thin film heterojunctions†
Patrizio Salice,a Camillo Sartorio,b Alessandro Burlini,a Roberto Improta,c
b
a
*
*
Bruno Pignataro and Enzo Menna
A flow functionalization route has been employed to derivatize single wall carbon nanotubes (SWCNTs) by
thienylphenyl groups. The SWCNT derivatives in the most soluble fraction have been characterized by
thermogravimetric analysis, DLS analysis, DFT calculations, and UV-vis-NIR, microRaman and IR
spectroscopies to study the degree of functionalization, the concentration of SWCNTs in solution, the
dimension of the aggregates in solutions, the density of defects, and the presence of the thienylphenyl
groups. Thin-film heterojunctions made of SWCNT derivatives and poly(3-hexylthiophene) (P3HT) have
been prepared by various methods employing the Langmuir–Schaefer technique, spin-coating and
thermal annealing processes. By comparing the batch and the flow functionalizations, a trade-off
between solubility, processability and the thin-film opto-electronic properties has been found as a result
of the degree of functionalization.
Received 24th June 2014
Accepted 27th October 2014
DOI: 10.1039/c4tc01350f
www.rsc.org/MaterialsC
of scalable techniques aimed at controlling these characteris-
tics has made it possible to obtain semiconducting CNTs with
high ambipolar mobilities and high dielectric constants.11,12
Introduction
The use of carbon nanotube (CNT) derivatives brings forth new
challenges and opportunities to the materials science commu-
nity1,2 including elds like optoelectronics, photovoltaics, nano-
lithography, organic electronics, surface coatings, membranes,
sensors, biosensors, and so on.3,4
For instance, in the case of bulk heterojunction (BHJ) solar
cells, the use of CNTs instead of [60]- and [70]-fullerene deriv-
atives may lead to improved performance due to the geometric
and electronic characteristics of CNTs that are well suited to
improve light harvesting, and photo-generation, splitting and
diffusion of the excitonic species.5–9 Indeed, the tubular shape
of CNTs provides a large interfacial area and their size is
compatible with both the exciton diffusion length (6–20 nm)
and the percolation paths (hundreds of nanometers) across the
photoactive layer.
As a consequence, the excitons in CNT-based materials have
low binding energies and thus a low barrier for exciton disso-
ciation making them promising alternatives to [60]- and [70]-
fullerene derivatives for developing heterojunction devices.
Nevertheless, while chemical processes for the preparation
of photoactive layers based on conventional small organic
molecules and polymers have been largely exploited, the use of
carbon nanostructures (CNSs) requires additional develop-
ments. Indeed, the strong tendency of CNTs to aggregate
hampers their compatibilization in truly nanoscaled compos-
ites and may result in phase separation and, ultimately, in the
loss of the foresighted properties associated to their nanometric
dimensions.13 From this scenario it emerges that the rst
requisite in the design of CNT-based BHJs is to improve the
processability of CNTs to tune their interaction with the semi-
conducting polymer within the photoactive blend, where the
key processes of exciton splitting and charge carrier recombi-
nation occur.14 Despite a plethora of CNT-conjugated polymer
blends15,16 reported since the initial ndings of Amaratunga17
and Kymakis,18–20 BHJ solar cells based on these blends oen
fail to afford satisfying photo-conversion efficiencies.21,22
Indeed, suitable processing techniques and affordable chemical
strategies need to be further developed to optimize the blend
morphology and to improve the electronic interaction between
the polymer and the CNTs.23 The Nicholas group has been
amongst the rst to focus on the development of functionali-
zation processes to optimize the interface between conductive
Also, as to the organic thin lm transistor (OTFT) eld, the
geometric parameters of CNTs (i.e., chirality index and length)
determine their electronic properties, in particular the band gap
and the position of their conduction band.10 The development
a
`
Universita di Padova, Dipartimento di Scienze Chimiche, Via Marzolo, 1-35131,
Padova, Italy. E-mail: enzo.menna@unipd.it; Tel: +39-0498275660
b
`
Universita di Palermo, Dipartimento di Fisica e Chimica, V.le delle Scienze, Parco
d'Orleans II, 90128, Palermo, Italy. E-mail: bruno.pignataro@unipa.it; Tel: +39-
09123897983
cConsiglio Nazionale delle Ricerche, Istituto di Biostrutture e Bioimmagini (IBB-CNR),
Via Mezzocannone 16, I-80136, Napoli, Italy
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c4tc01350f
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C

View Article Online
Journal of Materials Chemistry C
Paper
polymers and CNTs.24 The authors reported the non-covalent temperature, between 280 and 1400 nm, data interval ¼ 0.5 nm,
functionalization of (6,5) single-walled carbon nanotubes scan rate ¼ 300 nm minꢁ1, SBW ¼ 2 nm. Dispersions of
(SWCNTs) with a monolayer of poly-(3-hexylthiophene) (P3HT) SWCNTs were achieved using the Sonicator 3000 (Misonix) with
to obtain nanometric heterojunctions with both excellent the following pulse parameters: time on ¼ 3 s, time off ¼ 3 s,
alignment and ultrafast charge-transfer rates. While this power level ¼ 2 (4–6 Watts) for 30 minutes. Dynamic light
strategy is effective in lab-scale quantities, covalent functional- scattering (DLS) measurements of nanotube samples dispersed
ization processes which form stable bonds at the surface of in air-equilibrated chlorobenzene were peꢀrformed with a Zeta-
CNTs are more advisable to prepare functionalized carbon sizer Nano S (Malvern Instruments) at 20 C setting 20 runs of
nanostructures with enhanced solubility and processability on a 10 seconds for each measurement. Raman spectra of CNSs,
larger scale.25,26 Indeed, the chemistry of CNTs is now rich with a drop cast on pre-cleaned glass micro slides (Corning) and
variety of reactions, including oxidation with strong annealed at 110 ꢀC, were recorded with an Invia Renishaw
oxidants,27,28 addition of carbenes,29 addition of organo- Raman microspectrometer (50ꢂ objective) using the 633 nm
lithium,30 1,3-dipolar cycloaddition31 and addition of diazo- line of a He–Ne laser at room temperature with a low laser
nium salts.32,33 The last class of reactions is particularly power. IR spectra of the samples were collected on a Thermo
indicated to modify surface properties (such as solubility, Scientic Nicolet 5700 FT-IR spectrophotometer equipped with
adhesion and biocompatibility) of CNSs.34,35 In the case of an attenuated total reection module (Smart Omni sampler, Ge
CNTs, we have demonstrated that the combined use of suitable crystal). Spectra were measured at a resolution of 4 cmꢁ1
.
dispersing solvents (e.g., 1-cyclohexyl-2-pyrrolidone, CHP)36,37 Thermogravimetric analyses (TGAs) of CNS samples, precipi-
and of ow reactors26,38 enables process intensication while tated by adding methanol and dried at 80 ꢀC at 0.2 mbar for 4 h,
preserving a tight control on the reaction conditions, and thus were carried out with a Q5000IR TGA (TA Instruments) under
ꢀ
on the properties of functionalized CNTs, including solubility, nitrogen by an isotherm at 100 C for 10 minutes followed by
ꢁ1
ꢀ
ꢀ
aggregation size and functionalization degree (FD).
heating at 10 C min rate till 900 C.
Since the interaction of functionalized CNTs with the
The ow reactor used for the functionalization of CNTs has
surrounding medium is determined by the amount and by the been previously described38 and it consisted of a 400 cm segment
distribution of functional groups on their surface,39,40 we of polytetrauoroethylene (PTFE) tubing (i.d. ¼ 800 mm) for a
devised an affordable synthetic strategy in ow conditions that total volume of 2.0 mL. The reagent stream was injected into the
provides a high degree of control on the addition of the reactive reactor heated at 80 ꢀC by an HPLC pump (Model KP-12-01S,
aryl diazonium salts to CNTs and, ultimately, on the properties Flom, Tokyo, Japan) providing a ow rate of 8.0 mL hꢁ1, cor-
of the resulting products.41 These results nd application in the responding to a residence time of 15 minutes. The inlet of the
present work, where we report on the preparation of photoactive reactor was properly tted with a three-way valve to control the
heterojunction devices, as in the case of a blend of P3HT as a loading of the CNT suspension into the PTFE injection loop
p-type electron donor semiconducting polymer and derivatized (O.D. ¼ 2.0 mm, volume ¼ 11 mL) and the outlet was tted with
SWCNTs as n-type electron acceptor semiconducting nano- a mechanical back-pressure regulator (S series Metering Valve,
structures. In particular, SWCNTs were functionalized with Swagelok).
4-(thien-2-yl)phenyl (PhTh) moieties, and the interaction of the
SWCNT-PhTh derivatives with P3HT in thin-lm blends has
been investigated.
Synthesis
Tert-butyl (4-(thiophen-2-yl)phenyl)carbamate (1). The reac-
tion was carried out under nitrogen atmosphere. Pd(PPh3)2Cl2
(160 mg, 0.23 mmol) and an aqueous solution of Na2CO3 (1 M, 9
Experimental
Materials and methods
All the reagents and solvents were purchased from Sigma mmol) were added to a solution of dihydroxy-thien-2-ylborane
Aldrich and were used as received if not otherwise specied. (600 mg, 4.7 mmol) and tert-butyl(4-iodophenyl)carbamate (0.77
SWCNTs were purchased from CNI (lot # P2150) and were used g, 2.9 mmol) in THF (40 mL). The resulting mixture was heated
as received. CHP was distilled at 104 ꢀC and 0.05 mbar prior to at reux. The reaction was monitored by thin layer chroma-
use. Dihydroxy-thien-2-ylborane42 and tert-butyl(4-iodophenyl)- tography (TLC) on silica gel (eluent: ETP/Et2O 4 : 1, Rf ¼ 0.35).
carbamate43 were synthesized according to literature proce- Aer 180 minutes, the reaction was quenched with a saturated
dures. Melting points of compounds were measured with a Q20 aqueous solution of NH4Cl (50 mL) and washed with AcOEt (3 ꢂ
1
DSC (TA Instruments). H and 13C-NMR spectra were collected 50 mL). The organic phase was dried over Na2SO4 and the
using an Avance 300 MHz NMR spectrometer (Bruker) and solvent was removed at a reduced pressure, obtaining a yellow
CDCl3 as a solvent. Multiplicity is given as follows: s ¼ singlet, d solid which was puried by column chromatography on silica
¼ doublet, dd ¼ doublet of doublets, m ¼ multiplet, br ¼ broad gel (eluent: ETP/Et2O 4 : 1, Rf ¼ 0.35). The desired product was
peak. Centrifugations were performed on an MR23i Jouan obtained as a yellow powder (360 mg). 56% yield (0.36 g). m.p. ¼
1
ꢀ
ultracentrifuge equipped with a SWM 180.5 swinging bucket 192–194 C; H-NMR (200 MHz, DMSO-d6): d (ppm) 10.01 (br,
rotor (Thermo Electron Corporation) at 3000 rpm for 10 1H), 7.60 (m, 4H), 7.47 (dd, 1H), 7.41 (dd, 1H), 7.10 (dd, 1H),
minutes. Absorption spectra in air-equilibrated solvents were 2.02 (s, 3H). 13C-NMR (50 MHz, DMSO-d6): d (ppm) 168.3, 143.3,
registered with a Varian Cary 5000 spectrophotometer, at room 138.8, 129.2, 129.0, 126.5, 125.5, 123.5, 120.0, 24.0. Anal. calcd
J. Mater. Chem. C
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
for C15H17NO2S: C, 65.43; H, 6.22; N, 5.09; S, 11.64. Found: C, 15 000–45 000, Aldrich) and SWCNTs by a Langmuir–Schaefer
65.27; H, 6.45; N, 4.89; S, 11.33. technique (LS). In particular, Langmuir lms were deposited
4-(Thien-2-yl)aniline (2). A solution of 1 (170 mg, 0.62 mmol) using a KSV Minitrough apparatus. Ultrapure Millipore ltered
in CH2Cl2 (5 mL) was treated with triuoroacetic acid (TFA, water with resistivity greater than 18.2 MU cm was used as the
57 mL, 0.74 mmol). Aer 45 minutes, the reaction was quenched subphase at a temperature of 25 ꢀC. A solution of 0.1 mg mLꢁ1
with a 1 M aqueous solution of NaOH (100 mL) and washed with of P3HT in chloroform (Aldrich, 99.9%) was randomly spread
AcOEt (3 ꢂ 50 mL). The organic phase was dried over Na2SO4 over the aqueous surface by a microsyringe. Aer solvent
and the solvent was removed at a reduced pressure obtaining evaporation (about 10 min) the oating lms were linearly
the desired product as a yellow powder (107 mg). 99% yield. compressed by means of two mobile barriers at a rate of 5 mm
m.p. ¼ 72–74 ꢀC; 1H-NMR (200 MHz, DMSO-d6): d (ppm) minꢁ1. The ultrathin lms, at a 20 mN mꢁ1 surface pressure,
7.25 (m, 3H), 7.13 (dd, 1H), 6.96 (dd, 1H), 5.21 (br, 1H). 13C-NMR were transferred layer by layer on the indium tin oxide coated
(50 MHz, DMSO-d6): d (ppm) 149.2, 145.6, 128.8, 127.1, 123.3, poly(ethylene terephthalate) (ITO/PET) square substrates (about
122.3, 121.1, 114.7.
1 cm2). To obtain a BHJ, the MHJ architectures were heated at a
Batch functionalization of CNTs (SWCNT-PhTh). As a refer- temperature close to the glass transition temperature of P3HT.
ence, SWCNTs were functionalized in a batch process.32,44–46 This thermal annealing allows increasing the molecular inter-
Briey, freshly prepared 4-(thien-2-yl)aniline (2) (130 mg, diffusion within the bilayer system and induces the formation
0.74 mmol) and isopentylnitrite (100 mL, 0.74 mmol) were added of an interpenetrating system typical of a BHJ. To realize spin-
to a dispersion of SWCNTs (9.0 mg, 0.75 mmol) in CHP (5 mL) coated BHJ, P3HT was dissolved in the SWCNTs solutions. The
pre-heated at 80 ꢀC. Aer 4 h, the reaction mixture was resulting P3HT:SWCNT mixtures were spin-coated onto ITO/
quenched in methanol (100 mL) and then centrifuged at 3000 PET substrates at 1500 rpm for 30 seconds.
rpm for 3 minutes, the supernatant was removed and the black
residue was washed with methanol (5 ꢂ 7 mL). The residual
carbon material was dried under vacuum (0.2 mbar) at 80 ꢀC for
4 h, weighted and characterized.
Characterization of thin lms
Fluorescence spectra were recorded using a Fluoromax-4
(HORIBA Jobin Ivon, Edison, USA) spectrouorometer equipped
with a 150 W xenon arc lamp as the excitation source. Spec-
troscopic analyses were performed directly on freshly prepared
thin lms of both pure P3HT and P3HT/SWCNT hetero-
junctions. Heterojunction quenching data were normalized to
the uorescence of P3HT thin lms.
Flow functionalization of CNTs (SWCNT-PhTh)
A solution of SWCNTs (8.9 mg, 0.74 mmol C) in CHP with
concentration 2 mg mLꢁ1 was sonicated for 1 h. Then the
aniline 2 (130 mg, 0.74 mmol) and isopentylnitrite (100 mL,
0.74 mmol) were added to the dispersion of SWCNTs and
immediately loaded in the injection loop of the ow reactor.
The reaction mixture collected from the reactor was quenched
in methanol (100 mL) and then centrifuged at 3000 rpm for 3
Computational details
minutes. The supernatant was removed and the black residue
was washed with methanol (5 ꢂ 7 mL). The residual carbon Our computations are based on density functional theory (DFT)
ꢀ
material was dried under vacuum (0.2 mbar) at 80 C for 4 h, and its time-dependent extension (TD-DFT) for the study of
weighted and characterized.
excited states. We used the CAM-B3LYP functional (that has
already been successfully applied to the study of dithiophene),47
while solvent effects were included by the Polarizable
Continuum Model (PCM).48 All the calculations were performed
by a Gaussian09 program.49
Characterization of functionalized CNTs
Aliquots of the soluble SWCNT-PhTh derivative were extracted
by means of sonication/centrifugation cycles. For each washing
cycle, 7.0 mL of chlorobenzene were added to the CNT material
obtained aer the functionalization processes. The concentra-
tions of each solution of SWCNT-PhTh were calculated contex- Fluorescence quenching experiments
tually to the TGA measurements. Briey, 500 mL of each solution
In a typical experiment, the intensity of the uorescence emis-
was drop cast and annealed at 120 ꢀC on a platinum crucible for
TGA analysis. The weight measured at the beginning of the TGA
analysis was used to calculate the concentration of SWCNT-
PhTh in solution. The rst extract for each reaction was sub-
jected to UV-vis-NIR, Raman and ATR-IR spectroscopies, DLS
and TGA.
sion of a uorophore solution is measured aer each addition
of a stable dispersion of CNTs. In our case, we added consec-
utive aliquots of a solution of SWCNT-PhTh, both ow- and
batch-functionalized, to a solution of P3HT in chlorobenzene
and we measured the emission at 575 nm aer each addition.
The intensities were then corrected both for the increasing
dilution and for the inner lter effect due to the absorption of
SWCNT-PhTh. Indeed, the absorption of CNTs throughout the
Preparation of thin-lm heterojunctions
Multilayer planar heterojunctions (MHJs) were realized by visible spectrum results in the reabsorption of the light emitted
depositing alternated thin layers of poly(3-hexylthiophene-2,5- at 575 nm by P3HT and could signicantly affect these
diyl) (P3HT regioregular, electronic grade, 99.995%, average Mn measurements (see Fig. 2).
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C

View Article Online
Journal of Materials Chemistry C
Paper
products of the functionalization in batch and in ow condi-
tions.38 Briey, the functionalized carbonaceous materials
obtained at the end of the addition reaction were quenched in
methanol, collected by centrifugation and washed with
methanol to remove the solvent (CHP), byproducts and excess
reactants. At this stage of the derivatization process, the func-
tionalized material consists of a mixture of SWCNTs with
different FDs and solubilities. Being interested in the prepara-
tion of solution-processable SWCNT/P3HT blends, we isolated
soluble fractions of the SWCNT-PhTh derivative by extraction
with chlorobenzene (in which P3HT is soluble). Practically, the
soluble derivatives were collected through dispersion of the
CNT material by pulsed sonication in chlorobenzene, precipi-
tation of the less-soluble aggregates by centrifugation and
collection of the supernatant. These steps were repeated ve
times on each sample, thus obtaining ve fractions with
decreasing concentrations55 for both batch- and ow-processed
SWCNT-PhTh (see Fig. 1).
Results and discussion
In Scheme 1, we report the synthesis of the 4-(thien-2-yl)aniline
(2) used for the preparation of SWCNT-PhTh.
The thienyl group was previously proposed as a motif of
interaction between CNTs and polythiophenes in the photo-
active layer of BHJs.50,51 The derivatization of an aniline with a
thienyl group was achieved by Suzuki–Miyaura coupling
between Boc-protected 4-iodoaniline and 2-thiophenboronic
acid to yield the carbamate 1 followed by deprotection of the
Boc group with triuoroacetic acid. The resulting 4-(thien-2-yl)-
aniline 2 was reacted with isopentylnitrite in CHP to generate in
situ the diazonium salt used for the derivatization of CNTs. We
chose to derivatize SWNCTs produced by the HiPco process due
to their prevailing semiconducting behavior which makes them
suitable electron acceptors in BHJs.52,53 For the functionaliza-
tion of SWCNTs, we initially followed reported approaches
based on reactions in a ask at 80 C for 4 h.54 In these condi-
ꢀ
The comparison between the solubility proles shows that
both functionalization processes led to a signicant increase in
the maximum solubility (0.47 mg mLꢁ1 and 0.37 mg mLꢁ1 for
SWCNT-PhTh synthesized in batch and in ow, respectively) in
chlorobenzene compared to pristine HiPco SWCNTs (0.017 mg
mLꢁ1). Nevertheless, the two solubility proles are remarkably
different. The sample functionalized in batch has a high mean
solubility and is completely extracted within ve fractions. On
the other hand, the rapidly decreasing concentration prole in
the case of ow synthesis indicates that only a fraction of the
starting material has gained enough solubility and thus is
extracted with chlorobenzene within the ve fractions (4.7 mg
over 8.9 mg of starting material).
tions, we expected to obtain highly soluble and extensively
functionalized CNTs. In fact, we have recently observed that an
extensive functionalization by the addition of diazonium salts
to MWCNTs in batch conditions results in the growth of aryl
layers on the surface of CNTs which further enhance their
solubility in common organic solvents (e.g., ethanol).41 Never-
theless, a covalent functionalization inducing a large increase
in the density of C sp3 defects in the C sp2 backbone leads to the
deterioration of the electronic properties of the CNTs.29 As a
consequence, a tight control on the reaction conditions is
necessary to reach a trade-off between solubility of the func-
tionalized CNTs and the resulting density of sp3 defects.
Considering that ow reactors offer higher productivity,
increased safety and tighter control on the reaction parameters
with respect to traditional batch processes, we have introduced
the functionalization of CNTs in ow as a valuable alternative to
traditional batch processes.26,38 Under this perspective, we
functionalized HiPco SWCNTs by the addition of the diazonium
salt of the aniline 2 also in a ow reactor, by setting a ow rate of
The most-soluble fractions for SWCNT-PhTh functionalized
in batch and in ow conditions were further characterized by
UV-vis-NIR, Raman, and ATR-IR spectroscopies, TGA, and DLS.
Using DLS we estimated the dimension of the soluble
aggregates of CNTs in chlorobenzene. Both samples present a
8.0 mL hꢁ1 and a temperature of 80 C. Then, we relied on a
ꢀ
robust protocol to purify, characterize and compare the
Fig. 1 Solubility profiles in chlorobenzene for SWCNT-PhTh derivat-
ized in flow (dark gray) and in batch (light gray).
Scheme 1 Functionalization of SWCNTs via the addition of the dia-
zonium salt of 4-(thien-2-yl)aniline (2).
J. Mater. Chem. C
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Table 1 Computed vertical excitation energies (in eV, values in nm
reported in italics) of 2-(p-tolyl)thiophene and system II in chloroform,
computed at the PCM/TD-CAMB3LYP/6-311+G(2d,2p)//CAMB3LYP/
6-31G(d) level. Oscillator strengths in parentheses
2-(p-tolyl)thiophene
II
Planar saddle point
Non-planar
4.29 (0.60) 288.7
4.52 (0.58) 274.5
4.27 (0.73) 290.3
4.48 (0.73) 277.0
minimum
(see Fig. 3), where the methyl group bears three ethylene
substituents (mimicking the interaction with the nanotube).
Our calculations predict that the interaction with the p system
of the ethylene groups leads to a weak red-shi of nA supporting
our assignment of the feature at 294 nm in SWCNT-PhTh
absorption spectra to the PhTh moiety. The lowest energy
absorption band is due to a strong p–p* transition, that can be
described as deriving from a HOMO / LUMO excitation.
Inspection of the frontier orbitals of II shows that the p
molecular orbitals (MO) of the three ethylene substituents on
the terminal methyl have a small but not negligible contribu-
tion both to the HOMO and to the LUMO. The interaction,
antibonding, i.e. leading to a destabilization of the MO, is larger
for the LUMO, explaining the small red-shi of the HOMO /
LUMO transition.
The same trend is found for both the non-planar absolute
minima and for the planar saddle point that has been suggested
to be the best reference structure for computing the oligothio-
phene absorption spectra.47 Our computed nA are in very good
quantitative agreement with the experiments, especially when
considering that, in a strongly related compound as dithio-
phene, vibrational effects lead to a 0.2–0.3 eV red-shi of the
computed band maximum with respect to nA.47
These ndings suggest that the surface of SWCNTs func-
tionalized in batch is heavily covered with 4-(thien-2-yl)phenyl
functionalities. On the other hand, the spectrum of ow-
synthesized SWCNT-PhTh shows both the plasmon
Fig. 2 Comparison between the UV-vis-NIR spectra in chloroben-
zene of (a) SWCNT-PhTh synthesized in batch (red) and in flow (blue)
conditions. The inset highlights the spectral regions of Van Hove
singularities (S2 and S1). (b) SWCNT-PhTh obtained in batch (red) and
2-(p-tolyl)thiophene (black). The inset highlights the absorbance
bands due to thienylphenyl residues.
distribution of particles with an average solvodynamic diameter
up to 100 nm, consistent with their proposed use as functional
nanoller in BHJs. In particular, SWCNT-PhTh obtained in
batch and that obtained in ow form bundles with an average
diameter of 50 and 100 nm, respectively (see ESI, page S8,
Fig. S7†).
The UV-vis-NIR absorption spectra of the most soluble
extracts of SWCNT-PhTh in chlorobenzene are reported in
Fig. 2a. The inset highlights the lack of the absorbance bands
due to electronic transitions between Van Hove singularities (S2
and S1) in the batch sample. Indeed, the nanotubes function-
alized in batch are characterized by the typical SWCNT plas-
monic absorption centered below 260 nm together with a band
around 294 nm. Such a band is not far from the maximum
absorption wavelength of the model compound 2-(p-tolyl)thio-
phene, as reported in the inset in Fig. 2b, and thus it might be
ascribed to the functional moiety graed to the nanotube
backbone. We resorted to rst principle TD-DFT calculations in
solution to conrm that the functionalization of SWCNTs may
result in a bathochromic shi of 0.10 eV (7 nm) of the absorp-
tion band of the PhTh moiety. In Table 1, the vertical excitation
energies (nA) of 2-(p-tolyl)thiophene computed in an apolar
environment are compared with those obtained for system II,
Fig. 3 Schematic description of the HOMO (left) and the LUMO (right)
the minimal model for a PhTh bonded to the SWCNT surface of the model system II.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C

View Article Online
Journal of Materials Chemistry C
Paper
contribution and the absorption features due to the Van Hove of SWCNTs is maintained. Complementary to Raman spec-
singularities of semiconducting SWCNTs between 550 and 800 troscopy, which is particularly sensitive to the CNT backbone,
nm (S2) and between 1000 and 1400 nm (S1), but lacks the band IR spectroscopy gives information on the functional groups
at 294 nm ascribed to the PhTh functionality. These observa- attached to their surface. The comparison between the IR
tions are consistent with a lower FD compared to the sample spectra of SWCNT-PhTh derivatized in batch and in ow is
processed in batch conditions, allowing it to retain, at least in reported in Fig. 4b. In both cases the C–H stretching modes of
part, the semiconducting characteristics of the pristine HiPco aryl groups (at ꢃ3000 and ꢃ3100 cmꢁ1) are recognizable but the
nanotubes upon ow functionalization.
intensity of the peaks is low, as expected for sparse functional
Further evidence comes from the comparison of the Raman groups graed to the CNT surface. The qualitative evaluation
spectra of the samples from batch and ow functionalization provided so far arguably implies that the sample prepared in
reported in Fig. 4a.
Both functionalization processes resulted in an increase of respect to the sample derivatized in ow conditions.
the ratio between the intensity of the D band at about 1300 cmꢁ1 Since pristine SWCNTs are thermally stable up to 700 ꢀC
batch is more soluble and has a higher defect density with
and that of the G band centered at about 1650 cmꢁ1 (D/G ratio) (ESI, page S6, Fig. S5†), it is normally assumed that organic
with respect to the starting material (see ESI, page S9, Fig. S8†). groups attached to the CNT surface decompose at a lower
This behavior is in agreement with the expected reduction of temperature than the carbon backbone. This allows us to obtain
the size of the sp2 domains, due to the conversion of sp2 a quantitative estimate of the FD by thermogravimetric
carbons to sp3 carbons upon reaction with aryl diazonium salts. analysis.28
Noticeably, the increase in the D/G ratio is more pronounced for
Nevertheless, the thermal analysis of the two samples
the material processed in batch conditions, indicating that it reported in the thermograms and in the corresponding weight-
has endured a more extensive modication of the sp2 backbone loss derivatives in Fig. 5 suggest a more complex behavior. The
than the material processed in ow conditions. Moreover, the sample prepared in batch is characterized by two peaks in the
dramatic reduction of the intensity of the 2D band at 2650 cmꢁ1 weight-loss derivative graph corresponding to
a thermal
in the sample processed in batch reveals a strong perturbation decomposition below 400 ꢀC followed by a second one centered
in the electronic and phononic structure of CNTs.56,57 Finally, at 550 ꢀC. As we previously reported,41 a fast and premature
the radial breathing modes are not subjected to particular
variations for both samples suggesting that the 1D-form factor
Fig.
5 Comparison between (a) thermograms of SWCNT-PhTh
Fig. 4 Comparison between SWCNT-PhTh derivatized in batch (red) derivatized in batch (red) and in flow (blue) (carrier gas: nitrogen;
and in flow (blue) conditions: (a) Raman (the inset highlights the radial heating rate: 10 ꢀC minꢁ1) and (b) between the corresponding weight-
breathing modes) and (b) IR spectra.
loss derivatives.
J. Mater. Chem. C
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
pyrolysis of CNTs derivatized via addition of aryl diazonium same time, the domains must ensure extended percolation
salts is observed when the functionalization is very intensive, so channels for the charge carriers. Among the possible processes
that nanotubes are almost completely degraded at 700 ꢀC. In to prepare thin lms with high control on morphology and
this case the weight loss ascribed to the organic fraction and structure, we exploited the Langmuir–Schaefer technique for
that of the carbon nanostructure are hardly discriminated, the realization of a multi-planar heterojunction (MHJ) made of
thus hampering a reliable calculation of the degree of alternating P3HT (donor) and SWCNT-PhTh (acceptor) thin
functionalization. Interestingly, the weight-loss derivative for layers. In this way, we were able to obtain a high control of the
SWCNT-PhTh prepared in ow conditions shows an additional organization and of the interface between the different
peak centered at 700 C. This may suggest that functionaliza- components.60,61
ꢀ
tion in batch conditions has reduced the thermal stability of
CNTs more markedly than ow-functionalization.
In particular, we prepared a (3 + 3) ꢂ 2 MHJ structure. This
structure consists of three consecutive layers of SWCNT-PhTh
So far we have provided evidence that ow-functionalization derivatized in batch and deposited onto three layers of P3HT,
affords products with good solubility and lower FD with respect repeated twice, for a total of twelve layers. In this way, we
to batch processing. Then, we used uorescence spectroscopy to ensured homogeneous domains for each component, although
prove that the differences in the functionalization result in the donor–acceptor interfacial area is limited. In order to
different interactions with P3HT. We started from the consid- explore a wider donor–acceptor interaction range, in addition to
eration that a charge-transfer interaction between a uorophore the as-prepared MHJ samples (Fig. 7a) the uorescence
and a CNT should provide a non-radiative deactivation pathway quenching measurements were performed on annealed MHJs
for the rst excited state of the uorophore, thus quenching its (Fig. 7b) and in spin-coated thin lms that are both BHJs with a
uorescence. Indeed, evidence of the interactions between larger interface area (Fig. 7c).
porphyrins, phthalocyanines and polycyclic aromatic hydro-
Spin-coating was performed by a solution of P3HT and
carbons with CNTs was obtained with a series of steady-state SWCNT-PhTh in chlorobenzene.
and time-resolved spectroscopic methods.58 The reported
results conrm that charge separation evolves from the excited
state of the uorophore to the SWCNT upon photoexcitation.
This phenomenon is quantiable by the Stern–Volmer equa-
tions from a uorescence quenching experiment.
As reported in Fig. 6, the addition of SWCNT-PhTh func-
tionalized in ow to a solution of P3HT resulted, indeed, in the
reduction of the uorescence emission intensity. Once reported
in the Stern–Volmer plot as a function of SWCNT-PhTh
concentration (i.e., the quencher), we obtained an upward curve
for both the sample functionalized in ow and in batch
conditions. This behavior indicates that two concurrent
mechanisms are possible: a dynamic quenching due to colli-
sions between excited P3HT and SWCNT-PhTh and a static
quenching resulting from the formation of a ground-state
complex between P3HT and SWCNT-PhTh. These results are
consistent with the reported CNT-induced uorescence
quenching of Rhodamine B.59 In that case a consistent ground-
state deactivation was found. Interestingly, uorescence
quenching of P3HT by SWCNT-PhTh functionalized in ow
conditions is more pronounced than for those functionalized in
batch conditions.
On the basis of the results obtained in solution we replicated
the uorescence quenching experiments in a solid-state photo-
active blend composed of P3HT and SWNT-PhTh. Indeed, the
performance of electronic devices is very sensitive to the
nanoscale architecture of the active layer, since key properties
like, for instance, the photovoltaic behavior or the eld-effect
mobility in ambipolar devices, are strongly related to the
structural order. In general, ensuring a good mixing of the two
components (P3HT and SWNT-PhTh) is not a sufficient condi-
tion to obtain a performing active layer. For instance, in organic
Fig. 6 (a) Emission spectra of P3HT in chlorobenzene with increasing
concentrations of SWCNT-PhTh functionalized in flow. (b) Stern–
Volmer plot of the quenching of P3HT by SWCNT-PhTh functionalized
photovoltaics, it is crucial that P3HT and SWCNTs create a
bicontinuous system with domains that have a size comparable
to the length scale of exciton diffusion (i.e., ꢃ5–20 nm). At the in flow (blue squares) and in batch (red triangles).
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C

View Article Online
Journal of Materials Chemistry C
Paper
Fig. 7 Schematic representation of three different P3HT:SWCNT
heterojunctions: (a) multi-planar heterojunction, (b) bulk heterojunc-
tion obtained by thermal annealing of the multi-planar heterojunction,
(c) bulk heterojunction obtained by spin-coating.
Fig. 9 Fluorescence spectra of pristine P3HT (red line), MHJ (blue
line), annealed MHJ for 1500 (green line), annealed MHJ for 50 (purple
line) and BHJ by spin-coating (yellow line) with SWCNT-PhTh deriv-
atized in flow.
Thanks to the approach in Fig. 7b, a modulation of the
characteristics of the heterojunction architectures (i.e., interfa-
cial area and percolation pathways) may be achieved simply by
changing the dynamic parameters (i.e., time and thermal
energy).60 Arguably, the thermal annealing of an MHJ may lead
to an intermediate architecture providing a high interfacial
surface, and thus a high probability for exciton dissociation,
and extended percolation channels for charge transport.60
Fig. 8 shows uorescence spectra of the pristine P3HT and of
three different structures: MHJ, MHJ annealed at 130 ꢀC for
5 minutes and BHJ obtained by spin-coating the soluble extract
of SWCNT-PhTh derivatized in batch. In this case, it is evident
that the presence of nanotubes does not induce any changes in
the luminescence of P3HT, irrespective of the heterostructure
considered. Arguably, this is evidence of the lack of electronic
interaction between P3HT and the CNTs functionalized in batch
conditions.
On the other hand, by performing the experiment with
SWCNT-PhTh derivatized in ow we observed an appreciable
P3HT uorescence quenching in all the architectures (MHJ,
spin-coated BHJ and annealed MHJ) (Fig. 9), spanning from
47% to 70% for the MHJ and spin-coated BHJ, respectively. This
shows that there is a signicant interaction between P3HT and
SWCNT-PhTh functionalized in ow conditions with all the
heterojunction architectures. Nevertheless, the increased
interface area of the BHJ architectures results in a higher uo-
rescence quenching. Moreover, when we annealed the MHJ
ꢀ
(130 C for 5 minutes) we obtained a quenching value (ꢃ70%)
similar to that of thꢀe spin-coated BHJ. Also by changing the
annealing time (130 C for 15 s) we were able to modulate the
uorescence quenching more likely because of a change of the
interfacial area indicating that P3HT and SWCNT-PhTh func-
tionalized in ow are able to undergo a controlled inter-
diffusion.
Conclusions
We have reported an insight into the use of functionalized
SWCNTs for optoelectronic applications by forming hetero-
junction devices with P3HT. In particular, we have provided
evidence that an excessive modication of the SWCNT surface,
due to the high number of carbon atoms converted from sp2 to
sp3 aer the functionalization, results in the loss of the elec-
tronic and thermal properties of pristine nanotubes. In agree-
ment with our previous study, uorescence quenching
measurements of SWCNT-PhTh/P3HT blends show that an
uncontrolled functionalization shields the surface of SWCNTs
thus preventing it from interacting with P3HT. On the other
hand, the use of non-conventional methods for the controlled
functionalization of SWCNTs has allowed us to prepare a
SWCNT-PhTh derivative with a good trade-off between solu-
bility, and thus processability, and preservation of the SWCNT
properties required for an electronic interaction with P3HT.
Fig. 8 Fluorescence spectra of pristine P3HT (red line), MHJ (blue
line), annealed MHJ (green line) and BHJ by spin-coating (purple line)
with SWCNT-PhTh derivatized in batch.
J. Mater. Chem. C
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
24 S. D. Stranks, S. N. Habisreutinger, B. Dirks and
R. J. Nicholas, Adv. Mater., 2013, 25, 4365–4371.
25 S. Niyogi, M. A. Hamon, H. Hu, B. Zhao, P. Bhowmik, R. Sen,
M. E. Itkis and R. C. Haddon, Acc. Chem. Res., 2002, 35, 1105–
1113.
26 P. Salice, D. Fenaroli, C. C. De Filippo, E. Menna,
G. Gasparini and M. Maggini, Chim. Oggi-Chem. Today,
2012, 30, 37–39.
Acknowledgements
This work was supported by MIUR (grants FIRB-Futuro in
Ricerca RBFR08DUX6, FIRB RBAP11C58Y, PRIN 20104XET32)
and Regione del Veneto (SMUPR n. 4148, Polo di Ricerca nel
settore fotovoltaico).
27 E. Menna, F. Della Negra, M. Dalla Fontana and
M. Meneghetti, Phys. Rev. B: Condens. Matter Mater. Phys.,
2003, 68, 193412.
28 M. D'Este, M. D. Nardi and E. Menna, Eur. J. Org. Chem.,
2006, 2517–2522.
29 J. Chen, M. A. Hamon, H. Hu, Y. Chen, A. M. Rao,
P. C. Eklund and R. C. Haddon, Science, 1998, 282, 95–98.
30 B. Gebhardt, R. Graupner, F. Hauke and A. Hirsch, Eur. J.
Org. Chem., 2010, 2010, 1494–1501.
31 P. Salice, E. Rossi, A. Pace, P. Maity, T. Caroglio, E. Menna
and M. Maggini, J. Flow Chem., 2014, 4, 79–85.
32 J. L. Bahr and J. M. Tour, Chem. Mater., 2001, 13, 3823–3824.
Notes and references
1 S. Cataldo, P. Salice, E. Menna and B. Pignataro, Energy
Environ. Sci., 2012, 5, 5919–5940.
2 L. Wang, H. Liu, R. M. Konik, J. A. Misewich and S. S. Wong,
Chem. Soc. Rev., 2013, 42, 8134–8156.
3 D. Liming, H. Pingang and L. Sinan, Nanotechnol., 2003, 14,
1081–1097.
4 D. Jariwala, V. K. Sangwan, L. J. Lauhon, T. J. Marks and
M. C. Hersam, Chem. Soc. Rev., 2013, 42, 2824–2860.
5 V. C. Tung, J.-H. Huang, J. Kim, A. J. Smith, C.-W. Chu and
J. Huang, Energy Environ. Sci., 2012, 5, 7810–7818.
´
´
´
´
´
6 M.-Y. Wu, R. M. Jacobberger and M. S. Arnold, J. Appl. Phys., 33 L. Rodrıguez-Perez, R. Garcıa, M. A. Herranz and N. Martın,
2013, 113, 204504–204515. Chem.–Eur. J., 2014, 20, 7177.
7 D. D. Tune and J. G. Shapter, Energy Environ. Sci., 2013, 6, 34 C. A. Dyke and J. M. Tour, J. Phys. Chem. A, 2004, 108, 11151–
2572–2577. 11159.
8 M.-H. Ham, G. L. C. Paulus, C. Y. Lee, C. Song, K. Kalantar- 35 J. L. Hudson, M. J. Casavant and J. M. Tour, J. Am. Chem. Soc.,
zadeh, W. Choi, J.-H. Han and M. S. Strano, ACS Nano,
2010, 4, 6251–6259.
2004, 126, 11158–11159.
36 J. N. Coleman, Adv. Funct. Mater., 2009, 19, 3680–3695.
9 D. J. Bindl, N. S. Safron and M. S. Arnold, ACS Nano, 2010, 4, 37 S. D. Bergin, Z. Sun, D. Rickard, P. V. Streich, J. P. Hamilton
5657–5664.
and J. N. Coleman, ACS Nano, 2009, 3, 2340–2350.
10 H. Kataura, Y. Kumazawa, Y. Maniwa, I. Umezu, S. Suzuki, 38 P. Salice, P. Maity, E. Rossi, T. Caroglio, E. Menna and
Y. Ohtsuka and Y. Achiba, Synth. Met., 1999, 103, 2555–2558. M. Maggini, Chem. Commun., 2011, 47, 9092–9094.
11 H. Liu, T. Tanaka, Y. Urabe and H. Kataura, Nano Lett., 2013, 39 Z. Salmi, S. Gam-Derouich, S. Mahouche-Chergui,
13, 1996–2003. M. Turmine and M. Chehimi, Chem. Pap., 2012, 66, 369–391.
12 T. Kobayashi, Z. Nie, J. Du, K. Okamura, H. Kataura, 40 M. M. Chehimi, Aryl Diazonium Salts, Wiley-VCH, 2012.
Y. Sakakibara and Y. Miyata, Phys. Rev. B: Condens. Matter 41 P. Salice, E. Fabris, C. Sartorio, D. Fenaroli, V. Figa,
`
Mater. Phys., 2013, 88, 035424.
13 D. W. Schaefer and R. S. Justice, Macromolecules, 2007, 40,
8501–8517.
14 C. Bounioux, E. A. Katz and R. Yerushalmi-Rozen, Polym.
Adv. Technol., 2012, 23, 1129–1140.
15 J. Gao, R. Annema and M. A. Loi, Eur. Phys. J. B, 2012, 85, 1–5.
16 W. Gomulya, J. Gao and M. A. Loi, Eur. Phys. J. B, 2013, 86, 1–
13.
M. P. Casaletto, S. Cataldo, B. Pignataro and E. Menna,
Carbon, 2014, 74, 73–82.
42 G. E. Collis, A. K. Burrell, S. M. Scott and D. L. Officer, J. Org.
Chem., 2003, 68, 8974–8983.
´
´
´
´
´
43 P. Hrobarik, V. Hrobarikova, I. Sigmundova, P. Zahradnık,
M. Fakis, I. Polyzos and P. Persephonis, J. Org. Chem.,
2011, 76, 8726–8736.
`
44 D. Bouilly, J. Cabana, F. o. Meunier, M. Desjardins-Carriere,
17 I. Musa, M. Baxendale, G. A. J. Amaratunga and
W. Eccleston, Synth. Met., 1999, 102, 1250.
F. o. Lapointe, P. Gagnon, F. L. Larouche, E. Adam, M. Paillet
and R. Martel, ACS Nano, 2011, 5, 4927–4934.
18 E. Kymakis, I. Alexandou and G. A. J. Amaratunga, Synth. 45 J. Bahr, E. Mickelson, M. Bronikowski, R. Smalley and
Met., 2002, 127, 59–62. J. Tour, Chem. Commun., 2001, 2001, 193–194.
19 E. Kymakis and G. Amaratunga, Appl. Phys. Lett., 2002, 80, 46 J. Bahr and J. Tour, J. Mater. Chem., 2002, 12, 1952–1958.
112–114.
47 E. Stendardo, F. Avila Ferrer, F. Santoro and R. Improta, J.
Chem. Theory Comput., 2012, 8, 4483–4493.
48 J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 2005,
105, 2999–3094.
49 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota,
20 E. Kymakis, I. Alexandrou and G. A. J. Amaratunga, J. Appl.
Phys., 2003, 93, 1764–1768.
21 E. Kymakis and G. A. J. Amaratunga, Sol. Energy Mater. Sol.
Cells, 2003, 80, 465–472.
22 E. Kymakis and G. Amaratunga, Rev. Adv. Mater. Sci., 2005,
10, 300–305.
23 S. Berson, R. de Bettignies, S. Bailly, S. Guillerez and
B. Jousselme, Adv. Funct. Mater., 2007, 17, 3363–3370.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C

View Article Online
Journal of Materials Chemistry C
Paper
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr,
W. H. Noon, C. Kittrell, J. Ma, R. H. Hauge, R. B. Weisman
and R. E. Smalley, Science, 2002, 297, 593–596.
J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, 54 C. A. Dyke and J. M. Tour, Nano Lett., 2003, 3, 1215–1218.
E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, 55 The concentrations of the solution of SWCNT-PhTh were
J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam,
M. Klene, J. E. Knox, 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,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,
calculated contextually to the TGA measurements. Briey,
500 mL of each solution was drop casted and annealed at
120 ꢀC on a platinum crubicle for TGA analysis. The
weight measured at the beginning of the TGA analysis was
used to calculate the concentration of SWCNT-PhTh in
solution.
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, 56 M. S. Dresselhaus, A. Jorio, M. Hofmann, G. Dresselhaus and
¨
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and
R. Saito, Nano Lett., 2010, 10, 751–758.
D. J. Fox, Gaussian Inc., Wallingford, CT, 2009.
50 M. M. Stylianakis, J. A. Mikroyannidis and E. Kymakis, Sol.
Energy Mater. Sol. Cells, 2010, 94, 267–274.
57 M. S. Dresselhaus, A. Jorio, A. G. Souza Filho and R. Saito,
Philos. Trans. R. Soc. London, A, 2010, 368, 5355–5377.
58 G. Bottari, G. de la Torre, D. M. Guldi and T. Torres, Chem.
Rev., 2010, 110, 6768–6816.
51 A. F. Nogueira, B. S. Lomba, M. A. Soto-Oviedo,
C. R. D. Correia, P. Corio, C. A. Furtado and 59 A. Ahmad, T. Kurkina, K. Kern and K. Balasubramanian,
¨
I. A. Hummelgen, J. Phys. Chem. C, 2007, 111, 18431–18438.
52 S. M. Bachilo, M. S. Strano, C. Kittrell, R. H. Hauge, 60 S. Cataldo, C. Sartorio, F. Giannazzo, A. Scandurra and
R. E. Smalley and R. B. Weisman, Science, 2002, 298, 2361–2366. B. Pignataro, Nanoscale, 2014, 6, 3566–3575.
53 M. J. O'Connell, S. M. Bachilo, C. B. Huffman, V. C. Moore, 61 S. Cataldo and B. Pignataro, Materials, 2013, 6, 1159–
ChemPhysChem, 2009, 10, 2251–2255.
M. S. Strano, E. H. Haroz, K. L. Rialon, P. J. Boul,
1190.
J. Mater. Chem. C
This journal is © The Royal Society of Chemistry 2014