Article
pubs.acs.org/Macromolecules
Electronic Properties of Semiconducting Polymer-Functionalized
Single Wall Carbon Nanotubes
Souzana N. Kourkouli,†,‡ Angeliki Siokou,†,* Andreas A. Stefopoulos,†,‡ Fotini Ravani,† Thomas Plocke,§
Matthias Müller,§ Janina Maultzsch,§ Christian Thomsen,§ Kostas Papagelis,†,∥ and Joannis K. Kallitsis†,‡
†
Foundation for Research and Technology Hellas, Institute of Chemical Engineering Sciences (FORTH-ICE-HT), Stadiou Str.,
Platani Achaias, 26504, Patras, Greece
‡
Department of Chemistry, University of Patras, 26504, Patras, Greece
∥
Department of Materials Science, University of Patras, 26504, Patras, Greece
§
Institut für Festkörperphysik, Technische Universität Berlin, Hardenbergstrasse 36, 10623 Berlin, Germany
ABSTRACT: New hybrid materials were synthesized by functionalizing
single wall carbon nanotubes (SWCNTs) with different semiconducting
species. Vinyl quinoline monomers bearing electron accepting groups or
even its ruthenium complex were used as organic semiconductors. UV−
vis−NIR absorption spectroscopy, X-ray photoelectron spectroscopy and
thermogravimetric analysis were used in order to qualify the resulted
materials. Optical characteristics were studied by photoluminescence
spectroscopy in neutral and acidic media. The electronic structure of the
valence band and the surface electronic properties of the materials in the
form of thin films developed on Si substrates were studied by ultraviolet
photoelectron spectroscopy. Raman spectroscopy revealed the charge
transfer mechanism between the polymer and the carbon nanotube while
the recorded Raman resonance profiles showed significant changes in the
transition energies and the widths of the resonance windows due to chemical modification of SWCNTs.
■
INTRODUCTION
Single wall carbon nanotubes (SWCNTs) have attracted much
attention for the past decade as a material with interesting onedimensional physics and potential applications in various
fields.1 One of the promising applications of SWCNTs is in
optoelectronics due to their excellent electron affinity, thermal
conductivity, and mechanical strength and, of course, their
chemical inertness2 due to the strong carbon−carbon bonds of
the cylindrical graphitic lattice. The disadvantage of insolubility,
mostly driven from the creation of bundles, has been overcome
by functionalization of SWCNT surface using either covalent or
noncovalent modifications. The chemical modification of
SWCNTs has been extensively reviewed3,4 and provides a
means for the combination of the dispersibility improvement
with the introduction of an additional function that is
dependent on the type of the groups used. “Grafting from”
and “grafting to” constitute the two basic methods for the
covalent attachment of polymeric chains to carbon nanotube’s
surface.4 In the current work, the functionalization of SWCNTs
with different semiconducting polymeric quinoline chains was
occurred using the “grafting from” method. SWCNTs were
properly modified to act as initiators for the atom transfer
radical polymerization (ATRP) of quinoline monomers,
affording the new hybrid materials.
In cases that the targeted application of the modified
SWCNTs is their use as optoelectronic materials, attachment of
© 2013 American Chemical Society
semiconducting units is of utmost importance. Recent advances
on the enrichment of the semiconducting character in
SWCNTs5−8 expand the perspective for the use of such
materials in photovoltaic applications.9−12 In these cases, the
control over the electronic properties in the semiconducting
SWCNTs is combined with the solubility improvement by
simply using semiconducting organic species as covalent
SWCNTs modifiers.13−17
Therefore, our approach mainly targets in the formation of
new electron accepting materials and their combination with
classical electron-donor functionalities producing polymer
electron donor−acceptors directly applicable to plastic solar
cells as the polymeric layer or as compatibilizers and stabilizers
of the net polymeric counterparts.18,19 In this case and in order
to increase and ensure higher charge mobilities in the overall
bulk phase, the incorporation of carbon based nanostructures,
such as SWCNTs are highly desired.
The novel composite materials combine the unique properties of this type of semiconducting polymers20,21 with those of
the SWCNTs22,23 giving rise to multifunctional systems with
great potential. The main purpose of the current work is the
detailed characterization of the obtained hybrid materials and
Received: December 17, 2012
Revised: March 1, 2013
Published: March 22, 2013
2590
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
flask BrPyQ (2g, 5.54 mmol) was added 4-vinylphenylboronic (1g,
6.65 mmol), K2CO3 2 M solution (8.3 mL), Pd(PPh3)4 (0.2 g, 0.166
mmol), and toluene (50 mL). The flask was degassed and filled with
argon. The reaction mixture was refluxed for 4 days. Then the mixture
was filtered from paper filter followed by the extraction of the organic
layer with ethyl acetate (3x 40 mL) and distilled water (3x 40 mL).
The organic part was dried with MgSO4 and after filtration the
solution was evaporated at reduced pressure in rotary evaporator
resulting yellow solid. This was dried and then dispersed in methanol
for further purification. The solid was then filtered from Por.3 and the
light yellow solid was vacuum-dried overnight resulting in 1.61g
(76%).
Ruthenium Complex with 6-(4-Vinylphenyl)-2-(2-pyridine)-4phenylquinoline (MRuPyQ). A 50 mL round-bottom flask with 6-(4vinylphenyl)-2-(2-pyridine)-4-phenylquinoline (0.15 g, 0.3916 mmol)
and cis-(by)2-RuCl2 in absolute ethanol (40 mL), was degassed and
filled with argon. The reaction mixture was heated for 3 days at reflux.
The solvent was evaporated resulted in a dark red solid which was
treated with water and filtered for removing the insoluble part.
Addition of ammonium hexafluorophosphate (NH4PF6) (15 g) in 30
mL of water in the soluble part, resulted in the formation of red
crystals. Then the mixture was placed in the refrigerator for 2 h and
then by slow filtration the red solid was collected. Finally, the solid was
washed with distilled cold water and diethyl ether. The solid was
vacuum-dried overnight, resulting in 0.28 g (76%) yield.
Synthesis of Hydroxy-Functionalized SWCNTs (SWCNT−OH). A
20 mL round-bottom flask with SWCNTs (100 mg, 8.33 mequiv C),
p-aminophenol (1 g, 9.17 mmol) and isopentyl nitrite (3 mL), was
degassed and filled with argon reacting at 65 °C for 36 h. Filtration of
the black solid, from Nylon-200 nm membrane, and continuous wash
with acetone was made until a colorless filtrate was obtained. Then the
solid was dispersed in dimethylformamide (DMF) (100 mL) and
stirred at 50 °C for additional purification. Again, filtration from same
type membrane resulted in a black solid, which was vacuum-dried
overnight at 75 °C.
Synthesis of ATRP Initiator Groups onto SWCNTs (SWCNT-Init). A
100 mL round-bottom flask with SWCNT−OH (50 mg),
chloropropionyl chloride (CPC) (3 mL, 30 mmol), dry triethylamine
(4 mL, 31.2 mmol), and distilled DMF (60 mL) was degassed, filled
with argon, and placed in the ultrasonic bath for 20 min. The reaction
mixture was stirred at 100 °C for 2 days. Filtration of the black solid
from Nylon-200 nm membrane and continuous wash with acetone and
dichloromethane until colorless filtrate was made. Then the solid was
dispersed in DMF (100 mL) and stirred at 50 °C for additional
purification. Again, filtration from same type membrane resulted in a
black solid, which was vacuum-dried overnight at 75 °C.
Polymerization of Quinoline Monomers onto SWCNTs (SWCNTCNQ, SWCNT-PyQ, SWCNT-RuPyQ). A 5 mL round-bottom flask with
SWCNT-Init (20 mg), CuBr (10 mg, 0.07 mmol), and PMDETA (15
μL, 0.07 mmol) was degassed and filled with argon. Distilled DMF and
the respective monomer (MCNQ, MPyQ, MRuPyQ) (1 mmol) were
added with a syringe, and then the flask was placed in the ultrasonic
bath for 20 min. The reaction mixture was stirred at 90 °C for 1 day.
Then the mixture was dissolved in chloroform and the suspension was
filtered from Nylon-200 nm membrane for purification from any
unreacted monomers. Then the solid was dispersed in DMF (100 mL)
and stirred at 50 °C for additional purification. Again, filtration from
same type membrane resulted black solid which was vacuum-dried
overnight at 75 °C
Instruments and Measurements. The structures of the
synthesized materials were clarified by 1H NMR spectroscopy with
a Bruker Avance DPX 400 MHz spectrometer.
Centrifugation was made using a Hettich Zentrifugen, Universal
320, and sonication was done on a Bransonic (Branson), ultrasonic
cleaner 2510 model.
TGA experiments were performed on a TA Instruments Q50 series.
The functionalization degree of functionalized SWCNTs was
estimated according to: (% carbon/atomic weight of carbon)/(%
functional group/molecular weight of the function group).
the interactions between the SWCNTs and the semiconducting
polymers.18 Working in this direction, different techniques like
ultraviolet photoelectron spectroscopy (UPS) and X-ray
photoelectron spectroscopy (XPS) were used for the study of
the surface chemistry and electronic properties and UV−vis and
Photoluminescence spectroscopy for the optical characterization. The successful modification of SWCNTs with the
semiconducting polymers was confirmed by thermogravimetric
analysis (TGA) and Raman spectroscopy. The combination of
SWCNTs with semiconducting polymers offers an attractive
route to reinforce the macromolecular compounds as well as to
introduce novel electronic properties based on electronic
interactions and/or morphological alterations between the
two constituents.
■
EXPERIMENTAL SECTION
Materials. SWCNTs were purchased from Carbon Nanotechnologies Inc., Houston, TX (HiPCo form Unidym (lot no. P0261)). (2Amino-5-bromophenyl) (phenyl) methanone and 4-vinylphenylboronic acid were synthesized according to previous reports.24−26 1Bromo-4-nitrobenzene (Aldrich), 99%, p-cyanoacetophenone (Aldrich), methyl 2-pyridylketone (Aldrich), copper bromide 99,999%
(Aldrich), N,N,N′,N′,N″-pentamethyldiethylenetriamine 99% (Aldrich), 2-chloropropionyl chloride (Merck), ruthenium(III) chloride
(Aldrich), 2,2′-bipyridyl 99%(Aldrich), lithium chloride (Aldrich),
ammonium hexafluorophosphate >98% (Aldrich), and isopentyl nitrite
(Flucka) were also used. All solvents were used without further
purification unless otherwise stated. The filtration of SWCNTs was
performed in a typical apparatus using filtration papers purchased from
Sterlitech Corp., (Nylon-200 nm, 47 mm).
Synthetic Procedures. The synthetic routes27−29 of vinylic
monomers 6-(4-vinylphenyl)-2-cyanophenyl-4-phenylquinoline
(MCNQ), 6-(4-vinylphenyl)-2-(2-pyridine)-4-phenylquinoline
(MPyQ), ruthenium complex (MRuPyQ), and hybrid materials
(SWCNT-CNQ, SWCNT-PyQ, SWCNT-RuPyQ) are described
below.
Vinylic Monomer 6-(4-Vinylphenyl)-2-cyanophenyl-4-phenylquinoline (MCNQ). A 50 mL round-bottom flask with 2-amino-5bromophenyl (phenyl) methanone (3g, 10.86 mmol), p-cyanoacetophenone (3.15g, 21.72 mmol), acetic acid (35 mL) and sulfuric acid
95% (0.1 mL) was degassed and filled with argon. The reaction
mixture was refluxed for 3 days. Cooling the reaction at room
temperature resulting in the appearance of a light yellow solid that was
filtrated and washed with pure methanol. The obtained solid (6bromo-2-cyanophenyl-4-phenylquinoline (BrCNQ)) was vacuumdried overnight (3.37g) (80%). Into a 100 mL degassed roundbottom flask BrCNQ (3.3 g, 8.57 mmol) was added 4-vinylphenylboronic (1.9 g, 12.8 mmol), K2CO3 2 M solution (18.1 mL),
Pd(PPh3)4 (0.2 g, 0.166 mmol), and toluene (50 mL). The flask was
degassed and filled with argon and the reaction mixture was refluxed
for 4 days. The mixture was filtered from paper filter followed by the
extraction of the organic layer with ethyl acetate (3 × 40 mL) and
distilled water (3 × 40 mL). The organic part was dried with MgSO4,
and after filtration, the solution was evaporated at reduced pressure in
rotary evaporator resulting in a yellow solid. This was dried and then
dispersed in methanol for further purification. The solid was then
filtered, and the light yellow solid was vacuum-dried overnight
resulting in 2.7 g (80%)
Vinylic Monomer 6-(4-Vinylphenyl)-2-(2-pyridine)-4-phenylquinoline (MPyQ). A 50 mL round-bottom flask with 2-amino-5bromophenyl (phenyl) methanone (3 g, 10.86 mmol), acetylpyridine
(2.631 g, 21.72 mmol), acetic acid (35 mL), and sulfuric acid 95% (0.1
mL) was degassed and filled with argon. The reaction mixture was
refluxed for 3 days. After cooling the solution, a light brown cotton like
solid appeared. Purification was achieved after filtration in a Por.3 filter
and the solid was washed with pure methanol. The white solid (6bromo-2-(2-pyridine)-4-phenylquinoline (BrPyQ)) was vacuum-dried
overnight resulting in 3.37 g (86%) yield. Into a 100 mL round-bottom
2591
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
Chart 1. Chemical Structures of Monomers (MCNQ, MPyQ, MRuPyQ) and Hybrid Materials (SWCNT-CNQ, SWCNT-PyQ,
SWCNT-RuPyQ)
UV−vis spectra were recorded using a Hewlett-Packard 8452A
Diode Array. Continuous wave photoluminescence was measured on a
Perkin-Elmer LS45 spectrofluorometer. Photoluminescence measurements were performed after excitation of the samples at 340 nm. UV−
vis−NIR was conducted on a Perkin-Elmer L 900 spectrometer.
The Raman measurements were carried out using a micro Raman
setup in backscattering geometry, at ambient conditions. The
backscattered light was analyzed by a triple monochromator (Dilor
XY) and detected by a cryogenic charge coupled device (CCD). The
spectrometer was calibrated using a neon lamp. The resonance profiles
were recorded using a dye laser with Rhodamin 6G. The same
experimental setup was used to obtain Raman measurements with the
514.5 nm (2.41 eV) excitation of an Ar+ laser. To avoid problems
related to the inhomogeneities in the samples, spectra were taken at
different positions for each sample. The phonon characteristics were
obtained by fitting Lorentzian lineshapes to the experimental peaks,
after background subtraction Moreover, the Raman spectra at 785 nm
(1.58 eV) excitation energy were recorded using a micro-Raman
(InVia Reflex) spectrometer. The laser power was kept below 0.5 mW
to avoid laser heating effects.
The photoemission experiments were carried out in an ultrahighvacuum system, which consists of a fast entry specimen assembly, a
sample preparation and an analysis chamber with a base pressure <5 ×
10−10 mbar. The system is equipped with a hemispherical electron
energy analyzer (SPECS LH-10) and a UV lamp for UPS
measurements and a dual anode which produced the AlKα and
MgKα X-ray lines. UPS spectra were obtained using the HeI resonance
line (hν = 21.23 eV). The work function (eΦ) was determined for all
the surfaces from the UPS spectra by subtracting their width (i.e., the
energy difference between the analyzer Fermi level and the high
binding energy cutoff), from the HeI excitation energy. For these
measurements a bias of −12.29 V was applied to the sample in order
to avoid interference of the spectrometer threshold in the UPS spectra.
The high and low binding energy and highest occupied molecular
orbital (HOMO) cutoff positions were assigned by fitting straight lines
on the high and low energy cutoffs of the spectra and determining
their intersections with the binding energy axis. The analyzer
broadening was defined via UPS, from the width of the Fermi edge
of a sputtered clean Au foil and it was found to be 0.16 eV. Since this
broadening applies to the entire UV spectrum, the cutoff positions
were corrected only by half this value by adding 0.08 eV at the high
binding energy cutoff and subtracting the same amount from the low
binding energy cutoff. Regarding measurement errors it should be
noted that an error of ±0.1 eV is assigned to the absolute values for
ionization energies, work function and other UP-spectra cutoff features
where the error margin is significant, due to the process of fitting
straight lines. The ionization potential (IP) of the material under
investigation, i.e., the distance between the vacuum level and the
HOMO cutoff, is calculated by adding the absolute values of the
measured work function and the HOMO cutoff. The position of the
lowest unoccupied molecular orbital (LUMO) can only be measured
directly by using inverse photoemission, or it can be deduced from
HOMO−LUMO energy gap measurements.
The unmonochromatized MgKα (1253.6 eV) line was used for the
XPS measurements. The analyzer was working at constant pass energy
(Ep = 36 eV) which giving a full width at half-maximum of the main C
1s XPS peak of 1.4 eV. The XPS core level spectra were analyzed using
a fitting routine, which can decompose each spectrum into individual
mixed Gaussian−Lorentzian peaks after a Shirley background
subtraction. Regarding the measurement errors, for the XPS core
level peaks we estimate that for a good signal-to-noise ratio, errors in
peak positions are of about ±0.05 eV.
■
RESULTS AND DISCUSSION
The synthesis of monomers was performed according to our
previously described methodology,27 and characterized accordingly. The vinyl quinoline pyridine ruthenium complex was
obtained by complexation of the vinylpyridine quinoline using
cis-tripyridineruthenium chloride in ether.28 SWCNTs were
properly modified by the subsequent attachment of phenolic
and isopropyl chloride groups that can act as initiator for the
2592
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
Figure 1. Thermogravimetric analysis (TGA) of pristine SWCNT,
SWCNT-initiator, polymer-functionalized SWCNT, and homopolymers (PPyQ, PCNQ).
atom transfer radical polymerization29 of the vinyl quinoline
monomers. The obtained monomers and hybrid materials are
presented in Chart 1.
The final hybrid materials were characterized first with TGA
and the results proved, in all cases, the attachment of the
quinolines on the nanotubes surface (Figure 1). For SWCNTCNQ and SWCNT-PyQ almost the same amount of quinolines
was immobilized on the SWCNTs, about 11%, while for the
SWCNT-RuPyQ an enhancement of the weight loss (∼27%) is
observed. The homopolymers PCNQ and PPyQ indicate a
single degradation state about 400 °C which is in accordance
with that of the hybrid materials.
The optical characterization of the homopolymers (PPyQ
and PCNQ) and the obtained hybrid materials (SWCNTCNQ, SWCNT-PyQ, SWCNT-RuPyQ) dispersed in DMF and
formic acid (F.A.) was performed with UV−vis, UV−vis-NIR
absorption spectroscopy (Figures 2, parts a and b, and 4) and
photoluminescence spectroscopy (Figures 3, 4). From the
UV−vis spectra and based on the well-known relation between
the band gap energy (E) and the cut off wavelength (λ), E =
h*c/λ (h is Planck’s constant and c is the speed of light), we
were able to estimate the band gap energy of the
homopolymers and the derived hybrid materials. The calculated
values for the SWCNT-CNQ, SWCNT-PyQ and SWCNTRuPyQ materials were 2.8(±0.1) eV, 2.8(±0.1) eV and
2.9(±0.1) eV, respectively. Comparing with those of the
corresponding homopolymers PCNQ, 3.2(±0.1) eV and PPyQ,
3.3(±0.1) eV, the band gap energies of the obtained hybrid
materials are much lower mainly due to the new energy states
introduced by the SWCNTs.
Polyquinolines are pH-responsive due to the protonation of
the quinoline nitrogen atom which red shifts the emission
spectrum. This is shown in Figure 3 where the emission spectra
of hybrid materials (SWCNT-CNQ, SWCNT-PyQ) dispersed
in formic acid are compared to the ones obtained from the non
protonated species. Furthermore, in previous work on
SWCNTs functionalized with quinoline units,18 the optical
behavior of homopolymers and hybrid materials was extensively
studied. Particularly, by increasing the concentration of
quinoline units, photoluminescence intensity in the emission
Figure 2. UV−vis absorption spectroscopy: (a) polymer-functionalized SWCNTs (SWCNT-CNQ and SWCNT-PyQ) dispersed in
DMF and formic acid (FA). (The inset shows the UV−vis−NIR
absorption spectra of pristine SWCNTs and SWCNT−OH in the
region 400 −1300 nm); (b) homopolymers PCNQ and PPyQ in THF
and formic acid (FA).
spectra of homopolymers increases, while in the emission
spectra of hybrid materials the photoluminescence intensity
decreases showing an electron transfer from the quinoline units
to the SWCNTs. In addition, the emission spectra of the
protonated hybrid materials do not show the quenching
normally observed on the polymeric quinolines after protonation due to their immobilization on the SWNT surface, and
the absence of excimer formation.20,21
From Figure 5, it is evident that the relative integrated
intensities of the D to G+ band (ID/IG+) increases dramatically
upon functionalization. The extracted ID/IG+ values for pristine
and functionalized materials are presented in Table 1.
Alternatively, Maultzsch et al.32 argue that the G-band itself
can be defect induced and suggest that the intensity of the D
band should be normalized to the intensity of the second order
mode 2D as a measure for defect concentration in SWCNTs.
The corresponding ratios are also included in Table 1.
The inset of Figure 5 shows the Raman spectra of the studied
materials in the radial breathing mode (RBM) frequency
region, at 785 nm excitation. The spectra were normalized to
the intensity of the G-band. At this laser wavelength the
excitation is primary resonant with v2 → c2 transitions of
2593
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
Figure 5. Raman spectra of pristine and chemically modified CNTs
excited with the 785 nm (1.58 eV) excitation.
Figure 3. Photoluminescence spectroscopy of hybrid materials
(SWCNT-CNQ, SWCNT-PyQ) and homopolymers (PCNQ,
PPyQ) dispersed in DMF and FA.
Table 1. ID/IG+ Ratios for the Studied Samples (λ = 785 nm)
sample
ID/IG+
ID/I2D
pristine
SWCNT-initiator
SWCNT-CNQ
SWCNT-PyQ
SWCNT-RuPyQ
0.07
1.15
1.00
1.28
1.07
0.19
5.80
6.08
6.48
9.37
Table 2. Peak Positions of the G Mode for Pristine and
Functionalized Materials
pristine
SWCNT-CNQ
SWCNT-PyQ
SWCNT-RuPyQ
Figure 4. UV−vis absorption and photoluminescence spectroscopy of
SWCNT-RuPyQ dispersed in DMF.Raman spectroscopy is a powerful
and reliable tool for the characterization of modified SWCNTs at the
molecular level. Figure 5 shows the Raman spectra of pristine and
chemically modified SWCNTs, excited with 785 nm (1.58 eV), in the
D- and G-band region. The D-band is a disorder-induced feature
arising from double resonance Raman scattering process from a
nonzero-center phonon mode.30 It is well-documented that sidewall
functionalization breaks the translational symmetry along the tube axis
causing this mode to become Raman active. Therefore, an increase of
the D-band intensity comprises a fingerprint for successful sidewall
functionalization. The high energy region in the Raman spectrum of
single-wall carbon nanotubes contains two main first-order components, resulting from the in-plane C−C carbon displacements parallel
and perpendicular to the tube axis (tangential G-band), usually labeled
as G+ (1590 cm-1) and G− (1570 cm-1). In metallic carbon nanotubes,
the G- component exhibits a broad asymmetric Breit−Wigner−Fano
(BWF) line shape, resulting from the phonon coupling to an electronic
continuum.31
(1590.2
(1593.7
(1594.9
(1587.7
±
±
±
±
0.8)
2.7)
3.1)
2.6)
cm‑1
cm‑1
cm‑1
cm‑1
Figure 6. Raman resonance profiles of the decorated and nondecorated reference sample for the semiconducting (6,4) tubes.
nanotubes of the (11,0) branch. Interestingly, the decorated
compounds exhibit a significant reduction of the Raman
scattering intensity for the (12,1) branch, located at the region
between 210 and 240 cm−1.
The energy shift of the G-Mode of the samples SWCNTCNQ, SWCNT-PyQ, and SWCNT-RuPyQ compared to the
starting material (black) at an excitation wavelength of 514 nm
semiconducting nanotubes. The members of two branches,
namely the (12,1) and the (11,0) contribute to the spectra.33
The highest intensity RBM feature at 267 cm−1 corresponds to
a diameter of 0.862 nm and can be assigned to (10,2) or (11,0)
2594
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
found that the metal complex leads to a G mode shift in the
opposite direction than the other samples. Because of the
preparation pathway the degree of functionalization of all
samples is the same; the polymerizations were started from the
same modified nanotubes, prepared with the ATRP initiator
groups. This ensures that the influence on the phonon energies
of the high energy mode is not caused by structural changes. As
a consequence, we consider a charge transfer between the
nanotubes and the polymers as responsible for the shifts. As
stated just below the observed opposite behavior for the metal
complex derivative is connected to an enhanced electron−
phonon coupling in this system.
Whereas in graphite the two related phonons are degenerate,
in nanotubes the curvature leads to two peaks, with the G−
peak frequency decreasing with decreasing tube diameter.34,35
Sasaki et al.36 recently reported density functional theory
(DFT) calculations, showing a possible shift for the G+ peak in
zigzag tubes. In case of the metallic zigzag tubes they show an
influence on both modes, because both longitudinal optical
(LO) and transverse optical (TO) modes may couple with
electron−hole pairs. Therefore, the modes show a dependence
of the Fermi energy. Surprisingly, in our measurements only the
G+ peak is affected, whereas typically the G− peak is supposed
to be much more sensitive to any changes of the tube: Kohn
anomalies in doped nanotubes have been predicted by Caudal
et al.,37 however, they limit the phenomena to the G peak. It is
most likely that this influence is also the reason for the
observed effects in the nanotube derivatives. The G mode data
in Table 1 were taken at 514 nm excitation wavelength, where
predominantly metallic tubes are in resonance. The DFT
calculations for each tube show a symmetric behavior around 0
eV.36 Assuming that the pristine sample is slightly doped,
because of impurities, remains from catalysts etc. the starting
material would be in a state of the rising or trailing edge of the
high energy mode calculated frequency shifts. This would cause
the observed shifts with opposite sign for the different moieties.
For the charging of the tubes due to the ruthenium ion, two
concepts are competing: the first is a charging via the whole
polymer chain, as it were “by wire”. The second and more likely
concept takes into account the small distances between each
ion and the tube surface because of the typical polymer
wrapping. This means that lots of ions come very close to the
nanotube surface, the architecture of the metal complex may
even allow a direct contact between them. Because the
polymers are semiconducting their conductivity is supposed
to be low.
Additionally full resonance profiles of the different samples
and tubes were recorded (Figure 6). A CaF2 crystal was used
for normalization of the intensities of the RBM signal for the
different excitation wavelengths. The transition energies of all
our samples are shifted about 30−60 meV to lower energies as
compared to the starting material. Only the SWCNT-PyQ
sample shows an increased full half-width. The shift of the
transition energy originates from the charge transfer, which is
causing a doping effect in the SWCNTs. Because of the doping
effect new acceptor or donor levels develop in the SWCNTs,
decreasing the effective band gap.
The XPS and UPS techniques were also used for a more
thorough characterization of the SWCNTs attached groups as
well as the influence of the various groups attached on the
electronic levels of the hybrid materials. Apart from the hybrid
materials, the pristine SWCNTs and the monomer MCNQ
were studied for comparison purposes.
Figure 7. C 1s (left) and N 1s (right) XPS spectra of SWCNT-CNQ,
SWCNT-CNQ[H] (protonated), the monomer MCNQ, and the
pristine SWCNTs.
Figure 8. UPS spectra of the hybrid materials, the CNQ monomer and
the pristine SWCNTs. On the left side, a detailed view of the area
around the Fermi level.
Table 3. HOMO Positions (Ionization Potentials) of
Pristine SWCNT, Quinoline Monomer (MCNQ), and the
Hybrid Materials in the Form of Film on Si Foil Measured
by UPS and Energy Gaps Calculated from the UV−Vis−NIR
Spectra and the Corresponding LUMO Positions
SWCNT
MCNQ
(monomer)
SWCNT-CNQ
SWCNT-PyQ
SWCNT- RuPyQ
HOMO (eV) ±
0.1
Eg (eV) ±
0.1
LUMO (eV)) ±
0.14
4.3
5.8
−
3.2
−
2.6
5.0
5.3
5.5
2.8
2.8
2.9
2.2
2.5
2.6
is presented in Table 2. To determine the positions of the G
mode, several points on the sample surface were measured and
the average shift was determined. SWCNT-CNQ and SWCNTPyQ showed a blue shift of 4.5 ± 1.8 and 3.4 ± 2.0 cm−1,
respectively, compared to the starting material while SWCNTRuPyQ had a red shift of the high energy mode 2.5 ± 1.7 cm−1.
Comparing the G mode positions of the tubes decorated with
the ruthenium complex with those of the other samples, we
2595
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
Figure 9. Schematic diagram of the pristine SWCNTs and the hybrid materials energy bands.
not discernible probably due to the poor quality of the
spectrum. The N1s spectrum of the quinoline modified
SWCNTs has the same features.
Next, the valence band structures of the materials were
investigated by UPS (Figure 8). The prominent peak of the
SWCNTs valence band, which extends up to the Fermi edge
cutoff, is observed at a binding energy of 3 eV and it is
attributed to C2pπ states, as reported in previous studies.39 The
low-binding-energy region of the monomer CNQ film appears
to have a strong feature at about 5 eV caused by the N lone pair
electrons40 which contribute to the π-system. This feature
becomes less sharp at the hybrid material while the SWCNTPyQ and SWCNT-RuPyQ hybrids have their density of states
close to the Fermi level significantly suppressed. This is better
shown on the left side of Figure 8 where a more detailed view
near the low-binding-energy region is presented. The HOMO
cutoff (valence band edge) positions for the modified
nanotubes are measured at 1.3 ± 0.05 eV for SWCNT-CNQ,
at 1.8 ± 0.05 eV for SWCNT-PyQ and 2.2 ± 0.05 eV for
SWCNT-RuPyQ.
The work function (eΦ) for each material is derived from the
high binding energy cutoff of the UPS spectra and its value for
the SWCNTs is 4.3 ± 0.1 eV in agreement with reported
values.39 The eΦ values for the modified nanotubes are reduced
to 3.7 ± 0.1 eV, 3.5 ± 0.1 eV and 3.4 ± 0.1 eV for SWCNTCNQ, SWCNT-PyQ, and SWCNT-RuPyQ respectively
indicating that electron transfer take place in all three cases
toward the carbon nanotubes leading to a work function
decrease.
The ionization potential, i.e., the HOMO position from the
vacuum level, can be found by adding the HOMO cutoff energy
(distance between the valence band edge and the Fermi level)
to the work function (distance between the Fermi level and the
The C1s peaks originating from the two starting materials i.e.
the pristine SWCNTs sample and the quinoline monomer
(MCNQ) are shown at the bottom and the top of Figure 7
respectively. The spectrum of the pristine nanotubes can be
analyzed into four components. The main one at BE= 284.2 eV
corresponds to carbon atoms in the sp2 bonds of the graphitic
structure while the one shifted by +1 eV from the main peak
corresponds to C−C sp3 bonds. This peak is always apparent in
the C1s spectra of SWCNTs and it is attributed to the presence
of carbon in defect sites such as rings along with defects on the
nanotube walls, vacancies, heptagon-pentagon pairs, kinks as
well as carbon atoms at the edges of the graphitic planes. The
peaks at 286.8 and 288.8 eV are attributed to carbon atoms
attached to oxygen containing functionalities in the form of C−
O(H) and O−CO respectively. Finally the peak close to 291
eV is a characteristic loss feature of the C1s spectrum of
graphitic carbon and represents the energy loss of C1s electrons
due to π−π* transitions that occur during to photoionization
procedure. The main C1s component of the quinoline
monomer appears at 285 eV. The peak at 286.6 eV is
attributed to carbon atoms bonded to nitrogen in pyridinic
(CN−C) and cyano (−CN) bonds, while the one at and
at 287.8 eV is attributed to carbon atoms in −CO (carbonyl)
or O−C−O or OC−N bonds (dihydroxy, dialkoxy, amide).
All these components appear also at the C 1s spectrum of the
quinoline modified SWCNTs sample (second from the bottom
in Figure 7) where the contribution from the graphitic body of
the SWCNTs has been considerably suppressed. The N1s
spectrum of the quinoline monomer is quite noisy and appears
to have a wide peak (fwhm = 3 eV) at binding energy 400.3 eV.
This peak corresponds to N atoms in cyano (−CN) units38
while the peak originating from nitrogen atoms in the pyridine
bonds, which should appear at slightly higher binding energy is
2596
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
through the European Social Fund. The Technische Universität
Berlin authors acknowledge support by the Cluster of
Excellence “Unifying Concepts in Catalysis” coordinated by
the Technische Universität Berlin and funded by the DFG. J.M.
acknowledges support from the European Research Council,
ERC Grant No. 259286.
vacuum level). The LUMO positions can be found by
subtracting the corresponding energy gap values calculated
from the UV−vis−NIR spectra from the ionization potential
values. The HOMO position of all the materials under study
together with the Eg values and calculated LUMO positions are
summarized in Table 3.
In Figure 9 the band diagrams of the pristine and modified
nanotubes are drawn. It is observed that the strongest charge
transfer occurs in the case of the metal complex derivative in
agreement with the results of the Raman studies while the same
complex has the highest LUMO energy which makes it a
promising candidate as electron accepting material.
■
■
CONCLUSIONS
The synthesis and detailed characterization of novel hybrid
materials with electron acceptor properties has been presented
in this work. The hybrid materials were synthesized by
functionalizing SWCNTs with semiconducting polymeric
quinoline chains using the “grafting from” method. TGA
measurements proved the success of the carbon nanotubes
functionalization while UV−vis−NIR allowed the band gap
estimation for the obtained materials. Raman spectroscopy, in
combination with Raman resonance measurements clearly
revealed the charge transfer mechanism between the polymer
and the SWCNTs suggesting that, due to the preparation
pathway, the degree of functionalization of all samples was the
same. Furthermore, it was shown that significant electron−
phonon effects such as Kohn anomalies take place in the metal
complex derivative due to the strong coupling between G+
mode with the electron−hole pairs in carbon nanotubes.The
modification in the HOMO position and the electronic
structure in the valence band of the materials were revealed
by UPS. The attachment of the quinoline molecules in the
SWCNTs body lead to a hybrid material with a HOMO
position which lays between the HOMO positions of the two
starting materials. The herein synthesized hybrid materials are
employed for the development of new composites that
combine the properties of the corresponding semiconducting
polymers with those of the SWCNTs and can be applied in
organic solar cells.
■
REFERENCES
(1) Kanungo, M.; Lu, H.; Malliaras, G. G.; Blanchet, G. B. Science
2009, 323, 234−237.
(2) Avouris, P.; Chen, J.; Freitag, M.; Perebeinos, V.; Tsang, J. C.
Phys. Status Solidi B 2006, 243, 3197−3203.
(3) Karousis, N.; Tagmatarchis, N.; Tasis, D. Chem. Rev. 2010, 110,
5366−5397.
(4) Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M. Chem. Rev.
2006, 106, 1105−1136.
(5) Sandanayaka, A. S. D.; Maligaspe, E.; Hasobe, T.; Ito, O.;
D’Souza, F. Chem. Commun. 2010, 46, 8749−8751.
(6) Oelsner, C.; Schmidt, C.; Hauke, F.; Prato, M.; Hirsch, A.; Guldi,
D. M. J. Am. Chem. Soc. 2011, 133, 4580−4586.
(7) Ehli, C.; Oelsner, C.; Guldi, D. M.; Mateo-Alonso, A.; Prato, M.;
Schmidt, C.; Backes, C.; Hauk, F.; Hirsch, A. Nat. Chem. 2009, 1,
243−249.
(8) Nakayama, K.-I.; Asakura, Y.; Yokoyama, M. Mol. Cryst. Liq. Cryst.
2004, 424, 217−224.
(9) Collison, C. J.; Spencer, S.; Preske, A.; Palumbo, C.; Helenic, A.;
Bailey, R.; Pellizzeri, S. J. Phys. Chem. B 2010, 114, 11002−11009.
(10) Stefopoulos, A. A.; Chochos, C. L.; Prato, M.; Pistolis, G.;
Papagelis, K.; Petraki, F.; Kennou, S.; Kallitsis, J. K. Chem.−Eur. J.
2008, 14, 8715−8724.
(11) Kymakis, E.; Stratakis, E.; Koudoumas, E. Thin Solid Films 2007,
515, 8598−8600.
(12) Li, C.; Chen, Y.; Wang, Y.; Iqbal, Z.; Chhowalla, M.; Mitra, S. J.
Mater. Chem. 2007, 17, 2406−2411.
(13) Stefopoulos, A. A.; Pefkianakis, E. K.; Papagelis, K.;
Andreopoulou, A. K.; Kallitsis, J. K. J. Polym. Sci. A Polym Chem
2009, 47, 2551−2559.
(14) Fischer, J. E. Acc. Chem. Res. 2002, 35, 1079−1086.
(15) Chun, K.-Y.; Lee, C. J. J. Phys. Chem. C 2008, 112, 4492−4497.
(16) Michel, T.; Alvarez, L.; Sauvajol, J. L.; Almairac, R.; Aznar, R.;
Mathon, O.; Bantignies, J. L.; Flahaut, E. J. Phys. Chem. Solids 2006, 67,
1190−1192.
(17) Rao, C. N. R.; Voggu, R. Mater. Today 2010, 13, 34−40.
(18) Chochos, C. L.; Stefopoulos, A. A.; Campidelli, S.; Prato, M.;
Gregoriou, V. G.; Kallitsis, J. K. Macromolecules 2008, 41, 1825−1830.
(19) Stefopoulos, A. A.; Kourkouli, S. N.; Economopoulos, S.;
Ravani, F.; Andreopoulou, A.; Papagelis, K.; Siokou, A.; Kallitsis, J. K.
Macromolecules 2010, 43, 4827−4828.
(20) Jenekhe, S. A.; Chen, X. L. Science 1998, 279, 1903−1907.
(21) Jenekhe, S. A.; Chen, X. L. Science 1999, 283, 372−375.
(22) Tans, S. J.; Verschueren, A. R. M.; Dekker, C. Nature 1998, 393,
49−52.
(23) Dai, H.; Wong, E. W.; Lieber, C. M. Science 1996, 272, 523−
526.
(24) Sybert, P. D.; Beever, W. H.; Stille, J. K. Macromolecules 1981,
14, 493−502.
(25) Davis, R. B.; Pizzini, L. C. J. Org. Chem. 1960, 25, 1884−1888.
(26) Dondoni, A.; Ghiglione, C.; Marra, A.; Scoponi, M. J. Org. Chem.
1998, 63, 9535−9539.
(27) Economopoulos, S. P.; Andreopoulou, A. K.; Gregoriou, V. G.;
Kallitsis, J. K. Chem. Mater. 2005, 17, 1063−1071.
(28) Andreopoulou, A. K.; Kallitsis, J. K. Eur. J. Org. Chem. 2005, 20,
4448−4458.
(29) Coessens, V.; Pintauer, T.; Matyjaszewski, K. Prog. Polym. Sci.
2001, 26, 337−377.
(30) Dresselhaus, M. S.; Dresselhaus, G.; Saito, R.; Jorio, A. Phys. Rep.
2005, 409, 47−99.
AUTHOR INFORMATION
Corresponding Author
*Telephone/Fax: +30 2610 965263/+30 2610 965223. E-mail:
[email protected].
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
This research has been cofinanced by the European Union
(European Social Fund − ESF) and Greek national funds
through the Operational Program “Education and Lifelong
Learning” of the National Strategic Reference Framework
(NSRF), Research Funding Program: Heracleitus II.
(D.276.001.049). Part of the research leading to these results
has received funding from the European Union’s Seventh
Framework Programme (FP7/2007−2013) under the Grant
Agreement No. 246039. Investing in knowledge society
2597
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598
Macromolecules
Article
(31) Brown, S. D. M.; Jorio, A.; Corio, P.; Dresselhaus, M. S.;
Dresselhaus, G.; Saito, R.; Kneipp, K. Phys. Rev. B 2001, 63, 155414.
(32) Maultzsch, J.; Reich, S.; Thomsen, C.; Webster, S.; Czerw, R.;
Carroll, D. L.; Vieira, S. M. C.; Birkett, P. R.; Rego, C. A. Appl. Phys.
Lett. 2002, 81, 2647−2649.
(33) Maultzsch, J.; Telg, H.; Reich, S.; Thomsen, C. Phys. Rev. B
2005, 72, 205438.
(34) Jorio, A.; Souza Filho, A. G.; Dresselhaus, G.; Dresselhaus, M.
S.; Swan, A. K.; Ü nlü, M. S.; Goldberg, B. B.; Pimenta, M. A.; Hafner,
J. H.; Lieber, C. M.; Saito, R. Phys. Rev. B 2002, 65, 155412.
(35) Piscanec, S.; Lazzeri, M.; Robertson, J.; Ferrari, A. C.; Mauri, F.
Phys. Rev. B 2007, 75, 035427.
(36) Sasaki, K.-i.; Farhat, H.; Saito, R.; Dresselhaus, M. S. Phys. E:
Low Dimens. Syst. Nanostruct. 2010, 42, 2005−2015.
(37) Caudal, N.; Saitta, A. M.; Lazzeri, M.; Mauri, F. Phys. Rev. B
2007, 75, 115423.
(38) Pels, J. R.; Kapteijn, F.; Moulijn, J. A.; Zhu, Q.; Thomas, K. M.
Carbon 1995, 33, 1641−1653.
(39) Wang, Y.-Q.; Sherwood, P. M. A. Chem. Mater. 2004, 16, 5427−
5436.
(40) Hammer, P.; Lacerda, R. G.; Droppa, R., Jr; Alvarez, F. Diam.
Relat Mater 2000, 9, 577−581.
2598
dx.doi.org/10.1021/ma3025464 | Macromolecules 2013, 46, 2590−2598