Structure-property relationships in multifunctional thieno(bis)imide-based semiconductors with different sized and shaped N-alkyl ends

Article abstract of DOI:10.1039/c3tc32538e

The relationships between the molecular structure, packing modalities, charge mobility and light emission in organic thin films is a highly debated and controversial issue, with both fundamental and technological implications in the field of organic optoe

Full text of DOI:10.1039/c3tc32538e

Journal of
Materials Chemistry C
View Article Online
View Journal | View Issue
PAPER
Structure–property relationships in multifunctional
thieno(bis)imide-based semiconductors with
different sized and shaped N-alkyl ends†
a
Cite this: J. Mater. Chem. C, 2014, 2,
3448
Manuela Melucci, Margherita Durso,a Cristian Bettini,ab Massimo Gazzano,a
*
Lucia Maini,c Stefano Toffanin,d Susanna Cavallini,d Massimiliano Cavallini,d
Denis Gentili,d Viviana Biondo,e Gianluca Generali,e Federico Gallino,f
Raffaella Capellid and Michele Muccinide
The relationships between the molecular structure, packing modalities, charge mobility and light emission
in organic thin films is a highly debated and controversial issue, with both fundamental and technological
implications in the field of organic optoelectronics. Thieno(bis)imide (TBI) based molecular
semiconductors provide an interesting combination of good processability, tunable self-assembly,
ambipolar charge transport and electroluminescence, and are therefore an ideal test base for
fundamental studies on the structure–property correlation in multifunctional molecular systems. Herein,
we introduce a new class of thieno(bis)imide quaterthiophenes having alkyl side chains of different
shapes (linear, cyclic, branched) and lengths (C1–C8). We found that contrarily to what is generally
observed in most molecular semiconductors, the length of the alkyl substituent does not affect the
optical, self-assembly and charge transport properties of TBI materials. However, different
electroluminescence powers are observed by increasing the alkyl side, this suggesting a potential tool for
the selective modulation of TBI functionalities. A deep experimental and theoretical investigation on this
new family of TBI materials is provided.
Received 23rd December 2013
Accepted 23rd February 2014
DOI: 10.1039/c3tc32538e
www.rsc.org/MaterialsC
In the last few years, the development of a very large library of
small molecules and polymeric p-conjugated semiconductors has
Introduction
provided useful design principles for the predened tailoring of
a number of functional properties.2 P-type semiconductors,
both molecular and polymeric,3 based on different building
blocks,4 have been reported5 with mobility larger than 1 cm2 Vꢀ1 sꢀ1
(ref. 6) up to 17 cm2 Vꢀ1 sꢀ1.7 On the other hand, because of both
synthetic and technological limits, n-type8 and ambipolar mate-
rials9 that are crucial for complementary circuits, sensors, and light-
emitting transistors (OLETs)10 have been less documented.11 Even
fewer examples as well as related deep structure–property investi-
gation have been provided for molecular semiconductors showing
simultaneous efficient electroluminescence and high mobility
ambipolarity.12 In this context, we recently introduced the class of
2,3-thienoimide (TBI) based oligothiophenes (NTxN) as new
multifunctional materials combining good processability, high
crystallinity together with ambipolar or n-type charge transport
combined to electroluminescence in single layer OLETs. Starting
from the rst case, namely C4-NT4N (Chart 1),13 we investigated the
effect of the molecular dissymmetry,14 aromatic p-core type15 and
molecular size16 on the functional properties of the resulting
oligomers.17
Organic semiconductors are subject of intense research for
their crucial role as key components of the next generation of
electronic devices such as displays, thin-lm transistors, solar
cells, sensors, and logic circuits.1 In particular, molecular
tailoring and structure–property relationship investigation have
allowed signicant advances in the device performances,
attracting increasing interest from both industry and academia.
a
`
Consiglio Nazionale delle Ricerche, Istituto per la Sintesi Organica e la Fotoreattivita,
(CNR-ISOF), via P. Gobetti 101, 40129 Bologna, Italy. E-mail: manuela.melucci@isof.
cnr.it
bMIST-ER Laboratory, via P. Gobetti 101, 40129 Bologna, Italy
c
`
Dipartimento di Chimica “G. Ciamician”, Universita di Bologna, Via F. Selmi 2, 40126
Bologna, Italy
dConsiglio Nazionale delle Ricerche, Istituto per lo Studio dei Materiali
Nanostrutturati (CNR-ISMN), via P. Gobetti 101, 40129 Bologna, Italy
eE.T.C. s.r.l., via P. Gobetti 101, 40129 Bologna, Italy
fSAES-GETTERS S. p. A., Viale Italia 77, 20020 Lainate, MI, Italy
† Electronic supplementary information (ESI) available: Synthetic details, optical
properties in solution, differential scanning calorimetry, thin deposit morphology
and thermal behavior, X-ray diffraction, full electrical characterization,
computational details, brightness measurements for C6-NT4N. CCDC 977806.
For crystallographic data in CIF or other electronic format see DOI:
10.1039/c3tc32538e
Alkyl and heteroalkyl end substitution of p-conjugated mate-
rials is a widely exploited design strategy to change the solid state
organization and packing in thin lms, generally resulting in a
3448 | J. Mater. Chem. C, 2014, 2, 3448–3456
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Chart 1 Molecular structure of the compounds herein investigated (C1–C8 linear), C6 branched and cyclic NT4N.
+
great impact on the charge transport.4,18 On this basis, aiming at M.p. >350 C; EI-MS m/z 496 (Mc ); UV-vis (CH2Cl2) lmax ¼ 451
ꢁ
TBI materials with improved electrical performance and at nm, PL (CH2Cl2) lem ¼ 572 nm (lexc ¼ 451 nm). 1H NMR
dened property-specic design strategies, here we report a new (CD2Cl2, TMS/ppm) d 7.37 (s, 2H), 7.33 (d, 3J ¼ 3.6 Hz, 2H), 7.25
3
class of TBI materials having a xed p-conjugated backbone, (d, J ¼ 3.6 Hz, 2H), 3.10 (s, 6H). Anal. calcd for C22H12N2O4S4
length and core type, but differing in the alkyl end substituents. (496.60): C, 53.21; H, 2.44. Found: C, 53.28; H, 2.97%.
In particular, TBI molecules, quaterthiophene sized (T4),
2,20-(2,20-Bithiophene-5,50-diyl)bis(5-propyl-5H-thieno[3,2-c]-
having linear (C1–C8), branched (C6) or cyclic (C6) alkyl end pyrrole-4,6-dione), C3-NT4N, 1b. Red orange powder, 80% yield.
substituents (Chart 1) have been considered. The optical, M.p. 284 ꢁC, MS (70 eV, EI): m/z 552 (Mc+1), UV-vis (CH2Cl2) lmax
thermal, optoelectronic and electrical properties of the newly ¼ 451 nm, PL (CH2Cl2) lem ¼ 572 nm (lexc ¼ 451 nm). 1H NMR
synthesized materials have been investigated by a combined (CDCl3, TMS/ppm): d. 7.33 (s, 2H), 7.26 (d, 3J ¼ 3.6 Hz, 2H), 7.18
3
experimental and theoretical approach and related to the (d, J ¼ 4.0 Hz, 2H), 3.60 (t, 4H), 1.68 (m, 4H), 0.95 (t, 6H). 13C
peculiar molecular end motif.
NMR (CDCl3, TMS/ppm): d 163.9, 162.7, 149.3, 145.2, 137.8,
137.6, 134.7, 126.8, 125.4, 116.7, 40.2, 22.1, 11.3. Anal. calcd for
C
26H20N2O4S4 C, 56.50; H, 3.65, found: C, 56.70; H, 3.91%.
2,20-(2,20-Bithiophene-5,50-diyl)bis(5-hexyl-5H-thieno[3,2-c]-
Experimental
Synthesis
pyrrole-4,6-dione), C6-NT4N, 1d. Orange powder, 78% yield.
M.p. 283 ꢁC, MS (70 eV, EI): m/z 636 (Mc+1), UV-vis (CH2Cl2) lmax
¼ 451 nm, PL (CH2Cl2) lem ¼ 571 nm (lexc ¼ 451 nm).1H NMR
(CDCl3, TMS/ppm): d 7.32 (s, 2H), 7.26 (d, 3J ¼ 3.6 Hz, 2H), 7.18
Compounds 2 and 6 were prepared according to already
reported procedures.13 Compounds 3a–g are commercially
available. The synthesis of the intermediates is reported in the
ESI.†
3
(d, J ¼ 4 Hz), 3.60 (t, 4H), 1.64 (m, 4H), 1.31 (m, 12H), 0.95 (t,
6H). 13C NMR (CDCl3, TMS/ppm): d 163.8, 162.7, 149.3, 145.2,
137.8, 137.6, 134.7, 126.8, 125.4, 116.7, 38.6, 31.4, 28.8, 26.5,
22.5, 14.0. Anal. calcd for C32H32N2O4S4: C, 60.35; H, 5.06,
found: C, 60.39; H, 5.10%.
2,20-(2,20-Bithiophene-5,50-diyl)bis(5-cyclohexyl-5H-thieno-
[3,2-c]pyrrole-4,6-dione), C6cyc-NT4N, 1e. Orange powder, 40%
yield. M.p. >300 ꢁC, MS (70 eV, EI): m/z 632 (Mc+1), UV-vis
(CH2Cl2) lmax ¼ 444 nm, PL (CH2Cl2) lem ¼ 569 nm (lexc ¼ 444
nm). 1H NMR (CDCl3, TMS/ppm): d 7.31 (s, 2H), 7.25 (d, 3J ¼ 4.0
Hz, 2H), 7.18 (d, 3J ¼ 4.0 Hz), 4.00 (m, 2H), 2.14 (m, 4H), 1.79 (m,
All 1H, 13C, NMR spectra were recorded with a Varian Mercury
400 spectrometer operating at 400 MHz (1H). Chemical shis
were calibrated using the internal CDCl3, acetone-d6 or CD2Cl2
resonance, which were referenced to as TMS. Mass spectra were
collected on an ion trap Thermo scientic ISQ spectrometer
operating in electron impact (EI) ionization mode. Melting points
were determined by hot stage optical microscopy.
General procedure for Stille cross-coupling reaction
To a reuxing toluene solution of Pd[(AsPh3)]4 [in situ, 6 mol%, 12H), 1.30 (m, 4H). 13C NMR (CDCl3, TMS/ppm): d 163.9, 162.7,
Pd2(dba)3 (0.011 mmol), AsPh3 (0.088 mmol), 5 ml of tol.] and 149.1, 145.1, 137.7, 126.7, 125.5, 125.3, 116.7, 116.5, 53.4, 51.5,
bromo-thieno[3,2-c]pyrrole-4,6-dione 5 (0.36 mmol, in 3 ml 30.1, 29.7, 26.1, 25.1. Anal. calcd for C32H28N2O4S4: C, 60.73; H,
of toluene), under N2 atmosphere, commercially available 4.46, found: C, 60.79; H, 4.50%.
5,50-bis(tributylstannyl)-2,20-bithiophene 6 (0.18 mmol in 2 ml
2,20-(2,20-Bithiophene-5,50-diyl)bis(5-(2-ethylhexyl)-5H-thieno-
of toluene) was added dropwise. The mixture was reuxed for 48 [3,2-c]pyrrole-4,6-dione), C6br-NT4N, 1f. Red-orange solid, M.p.
+1
ꢁ
h then pentane was added at room temperature and the so 234 C, (70 eV, EI): m/z 632 (Mc ), UV-vis (CH2Cl2) lmax ¼ 451
obtained solid precipitate was puried by ash chromatography nm, PL (CH2Cl2) lem ¼ 571 nm (lexc ¼ 451 nm). 1H NMR (CDCl3,
on silica gel (CH2Cl2–petrol. ether ¼ 60 : 40 / CH2Cl2) and TMS/ppm): d 7.33 (s, 2H), 7.27 (d, 3J ¼ 4.0 Hz, 2H), 7.19 (d, 3J ¼
nally by crystallization from toluene.
4.0 Hz, 2H), 3.55 (t, 4H), 1.92 (m, 2H), 1.35 (m, 16H), 0.95 (m,
2,20-(2,20-Bithiophene-5,50-diyl)bis(5-methyl-5H-thieno[3,2-c]- 12H). 13C NMR (CDCl3, TMS/ppm): d 164.1, 162.9, 149.3, 145.2,
pyrrole-4,6-dione), C1-NT4N, 1a. Dark red powder, 47% yield, 137.8, 137.6, 134.7, 126.8, 125.4, 116.7, 42.5, 38.4, 30.5, 28.5,
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 3448–3456 | 3449

View Article Online
Journal of Materials Chemistry C
Paper
23.8, 23.0, 14.1, 10.4. Anal. calcd for C36H40N2O4S4 C, 62.40; H, radiation and a fast X'Celerator detector. The data of the devices
5.82, found: C, 62.45; H, 5.89%.
(powders) have been collected in the 2q range 2.4–30ꢁ (2.4–80ꢁ)
2,20-([2,20-Bithiophene]-5,50-diyl)bis(5-octyl-4H-thieno[2,3-c]- counting 800 s (200 s) at each step of 0.05ꢁ (0.033ꢁ).
pyrrole-4,6(5H)-dione), C8-NT4N, 1g. Orange powder, 76%
yield. M.p. 242 ꢁC, MS (70 eV, EI): m/z 692 (Mc+1), UV-vis
OFET/OLET fabrication
(CH2Cl2) lmax ¼ 450 nm, PL (CH2Cl2) lem ¼ 568 nm (lexc ¼ 450
The device platform consists of a 420 nm thick PMMA dielectric
nm). 1H NMR (CDCl3, TMS/ppm): d 7.31 (s, 2H), 7.24 (d, 3J ¼ 4.0
lm, on top of a glass/ITO gate electrode, and gold made drain
Hz, 2H), 7.16 (d, 3J ¼ 3.6 Hz), 3.59 (t, 4H), 1.64 (m, 4H), 1.29 (m,
and source top contact. The PMMA lm was spin coated, while
20H), 0.87 (t, 6H). 13C NMR (CDCl3, TMS/ppm): d 163.8, 162.6,
the gold electrodes have been evaporated by high vacuum
149.3, 145.2, 137.8, 137.6, 134.7, 126.8, 125.4, 116.6, 38.6, 31.8,
sublimation through a shadow mask process. OLET devices are
29.2, 29.1, 28.8, 26.8, 22.6, 14.1. Anal. calcd for C36H40N2O4S4: C,
based on 100 nm thick active organic lms, grown by high
vacuum sublimation. The channel length (L) is equal to 70 mm,
62.40; H, 5.82, found: C, 62.43; H, 5.89%.
while the channel width (W) is equal to 12 mm. All opto-elec-
tronic measurements were carried out in an MBraun nitrogen
Optical spectroscopy
Thin-lm UV-visible absorption spectra were recorded with a
JASCO V-550 spectrophotometer. Steady-state photo-
luminescence (PL) was excited using a CW He–Cd laser at 325
nm and collected under transmission conditions (no colour
lter was necessary for truncating laser excitation).
PL emission was collected with a calibrated optical multi-
channel analyzer (PMA-11, Hamamatsu). Quantum yield
measurements of thin-lms were carried out by a stand-alone
and automatic absolute PL Quantum Yield measurements
System (C9920-02, Hamamatsu). The measurements were per-
formed in air and exciting the PL emission at the thin-lm
absorption maximum wavelength.
glove box using a standard SUSS Probe Station equipped with a
Hamamatsu photodiode for light detection.
Computational details
Molecular and periodic calculations have been performed
within spin-polarized DFT, using the hybrid B3LYP functional,19
as implemented in the CRYSTAL09 code.20 The all-electron
Gaussian-type basis sets adopted were 6-31(d1) for oxygen,21
nitrogen22 and carbon,23 31(p1) for hydrogen26 and 8-6311(d1)
for sulphur.24 A post-DFT dispersive contribution, suggested by
Grimme,25 has been added to the computed ab initio total
energy and gradients in order to account long-range dispersion
interactions.
Polarised optical microscopy
Bulk charge transport has been investigated in the small
polaron hopping framework. The intramolecular reorganiza-
tion energy li has been computed both within the adiabatic
potential (AP) approach and through calculations of Huang–
Rhys (HR) parameters.26 The charge transfer integral s is eval-
uated for the nearest neighbouring molecules through the
application of Koopmans' theorem in the energy-splitting in the
Micrographs were recorded with a Nikon i-80 microscope
equipped with epi-illuminator, and cross-polars. The thermal
treatments of the samples were done using a commercial
heating stage (Linkam THMS 600) in air.
Atomic force microscopy (AFM)
AFM images were recorded with a commercial AFM (Multimode dimer (ESID) model.27
8, Bruker) operating in semi-contact mode under ambient
conditions. Si3N4 cantilevers, with a typical curvature radius of a
tip 10 nm, were used. Image analysis was done using the open
Results and discussion
Synthesis
source SPM soware Gwyddion-http://www.gwyddion.net.
The target compounds in Chart 1 were prepared by conven-
tional Stille coupling reaction by bistannylbithiophene 6 with
the proper bromo-TBI (5, Scheme 1). The TBI starting
compounds were prepared by ring opening of the thienoanhy-
dride 2 with the alkylamine (3) followed by intramolecular
dehydration of the intermediates. The isolated yields ranged
from 40% to 80% mainly depending on the nal material
solubilities. Indeed, methyl and cyclohexyl substituted C1-
NT4N and C6cyc-NT4N showed a poor solubility in organic
solvents. The solubility increased on increasing the chain size
or by using branched substituents.
Single crystal X-ray diffraction (SC XRD)
Crystal data for C3-NT4N were collected on an Oxford Xcalibur S
˚
with MoKa radiation, l ¼ 0.71073 A, monochromator graphite
at room temperature; C26H20N2O4S4, M ¼ 552.68, monoclinic, a
ꢁ
ꢀ1
˚
˚
˚
¼ 7.4707 (8) A, b ¼ 7.7479 (6) A, c ¼ 43.819 (6) A, b ¼ 91.54(1) , V
3
˚
¼ 2535.4(5) A , space group C2/c, Z ¼ 4 m (MoKa) ¼ 0.412 mm
,
¼
5239 reections measured, 2925 independent reections (Rint
0.0332). The nal R1 values were 0.0787 (I > 2s(I)). The nal
wR(F2) values were 0.1623 (I > 2s(I)). The nal R1 values were
0.1124 (all data). The nal wR(F2) values were 0.1785 (all data).
The goodness of the t on F2 was 1.167.
Electronic properties
Thin lm X-ray diffraction (XRD)
Insights into the electronic properties of the newly synthesized
The analysis was carried out by means of a PANalytical X'Pert TBI derivatives were obtained by performing optical spectro-
PRO powder diffractometer equipped with a Ni-ltered Cu Ka scopic characterization in diluted solution and thin lms.
3450 | J. Mater. Chem. C, 2014, 2, 3448–3456
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Scheme 1 Synthetic route to the target compounds. (i) Toluene, reflux, (ii) SOCl2 reflux, (iii) TFA, H2SO4, NBS, and (iv) Pd(AsPh3)4, toluene, reflux.
Details in ESI.†
The UV-vis absorption spectra in diluted CH2Cl2 solutions (see
The PL quantum yields (F, Table 1) measured for the alkyl
Fig. S1 and Table S1, ESI†) were almost comparable for all the substituted molecules (from C1 to C8) are almost invariant
compounds with the maximum absorption wavelength located at within the experimental error ranging between 3.5% (C8-NT4N)
about 350 and 450 nm. The emission spectra are featureless with and 6.5% (C6br-NT4N).
the emission maximum wavelength at around 570 nm.
The absorption spectra of vacuum-sublimed thin-lms,
reported in Fig. 1, are well structured with absorption maxima
located at around 345 and 430 nm for compounds having linear
alkyl substituents with C > 1. C1-NT4N was the only compound
with the absorption maximum red-shied to 450 nm. By
increasing the alkyl chain length from C1 to C8 units no relevant
variations in the photoluminescence spectra were observed.
Thermal properties
The thermal behaviour of compounds C1–C8-NT4N was inves-
tigated by combined hot-stage polarized microscopy of thin
deposits (POM) and by differential scanning calorimetry (DSC)
analyses.
DSC traces show multiple intense temperature transitions
beyond the melting one for all compounds with exception for
methyl and cyclohexyl ended derivatives C1-NT4N and C6cyc-
NT4N respectively, revealing liquid crystalline properties for
these compounds (Table 2).
The thermograms of compound C8-NT4N taken as model
systems are depicted in Fig. 2, while those of the other
compounds are reported as ESI (Fig. S2†). Accordingly to DSC,
hot-stage POM revealed LC properties for compounds C3–C8-
NT4N while only thermal decomposition was observed for C1-
NT4N. The transition temperatures and the melting enthalpies
are listed in Table 2.
Thin deposits
The thin deposits of compounds C1–C8-NT4N for POM analysis
were prepared by drop casting a chloroform solution (30 ml of
Table 1 Spectroscopic parameters of thin films of 100 nm thick NT4N
derivatives
Item
labs (nm)
lema (nm)
Fb (%)
C1-NT4N
C3-NT4N
C4-NT4N
C6-NT4N
C6cyc-NT4N
C6br-NT4N
C8-NT4N
350, 450
345, 429
345, 430
346, 426
365, 432
368, 446
345, 427
617
605
603
602
592
604
604
4
6
5
5
4
6.5
3.5
Fig. 1 (a–f) UV-vis and PL spectra vacuum-sublimed thin-films of all
compounds (the spectra of C4-NT4N are reported in Fig. S1, ESI†).
a
b
lexc ¼ 325 nm. Excitation at absorption maximum.
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 3448–3456 | 3451

View Article Online
Journal of Materials Chemistry C
Paper
Table 2 Thermal properties of compounds C1–C8-NT4N. For compound C4-NT4N see ref. 13
Transition temperature [ꢁC] (melting enthalpies [J gꢀ1])
Compound
Heating (20ꢁC minꢀ1
)
Cooling (20ꢁC minꢀ1
)
1a, C1-NT4Na
1b, C3-NT4N
1d, C6-NT4N
1e, C6cyc-NT4N
1f, C6br-NT4N
1g, C8-NT4N
Nd
Nd
274 (3.0), 289 (73.4), 301 (1.1)
244 (1.5), 265 (21.3), 287 (5.9)
344 (1.8), 367 (73.9)
218 (1.1), 235 (26.1)
121 (2.7), 248 (40.5), 279 (12.1)
262 (55.2), 281 (1.0)
219 (0.8), 243 (21.4), 269 (4.9)
329 (15.1)
213 (29.4), 201 (0.6)
231 (41.9), 266 (11.9)
a
No transitions were observed up to 400 ꢁC.
1 g lꢀ1 or saturated solution) on quartz slides in a saturated became a liquid crystal and de-wets forming a homogeneous
atmosphere of solvent vapours. distribution of droplets which exhibit the typical behaviour of a
All thin deposits exhibit similar morphologies,28 the material smectic liquid crystal (Fig. 3c). Accordingly, atomic force
tends to accumulate at the boundaries of the drop because of microscopy (AFM) investigation revealed crystals with well
the inhomogeneous evaporation of the solvent during the dened borders and angles and very slight roughness terraces
shrinking.29 Larger crystals (size >100 mm) whose shape (r.m.s. roughness <1 nm).
depended on the molecular structure, accumulate at the
boundaries while towards the centre of the droplet their sizes
XRD of single crystal, powder, thin lms
decrease till a few mm; furthermore, in the case of compounds
Compound C3-NT4N afforded single crystals useful for crystal
C3-, C6- and C8-NT4N a continuous lm in between the crystals
structure solution and molecular packing determination. The
was observed by atomic force microscopy (AFM) and POM.
molecule is centrosymmetric and the backbone is almost planar
being the two inner thiophene rings strictly coplanar and the
dihedral angle between the thiophene and the thiophene-imide
As a representative case, compound C8-NT4N is herein dis-
cussed while details on POM analysis of all the other
compounds are reported in the ESI (Fig. S3†).
units 1.1(4)ꢁ (Fig. 4a). Differently from what observed for
When deposited on glass C8-NT4N forms a continuous lm
C4-NT4N in previously reported study,13 the thiophene-imide
units are in uncommon syn conformations.30 The p-conjugated
backbones of compound C3-NT4N of adjacent molecules run
parallel to one another with an interlayer distance of ca. 3.54 A
as observed also in the recently reported structure of C4-NT4N.13
on the surface whose thickness increases from the centre to the
boundaries. Under POM, the lm exhibits birefringence
without extinguishing at any direction (Fig. 3a). Furthermore in
the centre of the deposit birefrangent ber-like crystals were
oen observed (Fig. 3b). The ber-like crystals appear as long
The molecules are involved in a bifurcated C–H/O hydrogen
crystallites (>50 mm) randomly distributed on the surface, with
˚
width ranging from 200 nm to 10 mm resulting in an aspect ratio
(length–width) between 1 : 5 and >1 : 10. The bers exhibit a
strong birefringence, however they do not extinguish at any
directions, indicating a polycrystalline nature. In good accor-
dance with the DSC trace (Fig. 2), C8-NT4N thin deposits show a
transition to the LC phase at 255 ꢁC followed by the transition to
the isotropic phase at 280 ꢁC. By heating above 255 ꢁC, C8-NT4N
Fig. 3 Optical micrographs taken by POM (cross-polars are oriented
parallel and perpendicular to the axis of the images) and the AFM
image of C8-NT4N thin deposits. (a) Birefrangent continuous film at
the boundary of the deposit. Bar length 100 mm. (b) Fiber like crystals at
Fig. 2 DSC traces (second heating and cooling steps) at 20ꢁC minꢀ1 of the centre of deposit. Bar length 10 mm. (c) LC phase at 260 ꢁC. Bar is
compound C8-NT4N.
100 mm. (d) AFM image of (b).
3452 | J. Mater. Chem. C, 2014, 2, 3448–3456
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Fig. 5 (a) XRD patterns of compounds C1, C3, C4, C6, C8-NT4N
devices directly exposed to the X-ray beam. In the inset the small angle
region is enlarged and the corresponding interlayer distances are
reported in nm. (b) Schematic representation of the C4-NT4N mole-
cule arrangement on the substrate.
Fig. 4 (a) ORTEP drawing of C3-NT4N (displacement ellipsoids drawn
at the 50% probability level). Unlabeled atoms are related to the
labelled atoms by symmetry operation 1/2 ꢀ x, 1/2 ꢀ y, ꢀz. (b) Arbitrary
units of the crystal packing showing the C–H–O interactions (red
lines). (c) View along the b axis showing the p-stack interactions and
the assembly of the layers.
can reasonably assume that all the investigated C1–C8-NT4N
molecules adopt a similar upright molecular arrangement.
Electrical characterization
Electrical characterization of compounds in Chart 1 was per-
formed in a top contact device conguration (details in the
Experimental section). The representative transfer and multiple
output (n-type) electrical curves of compounds C1 and C8-NT4N
are shown in Fig. 6 and Table 3 summarizes the charge trans-
port and electroluminescence data expressed as device emis-
sion power. Considering the electrical properties, it can be
noted that contrarily to conventional oligothiophenes, for
which cyclohexyl end substitution promotes charge transport,23
bond to form a 1D network (Fig. 4b). The p-stack and hydrogen
bond interactions pack efficiently the molecules to form layers
having in the inner part the conjugated backbone and the alkyl
chains on the surfaces, these layers pile up along the long axis
(Fig. 4c). The XRD pattern calculated from the structural data
ts very well with that experimentally collected for the C3-NT4N
powder sample conrming that the crystal phase is the same
(Fig. S4, ESI†). No signicant differences are observed for the
other compounds showing XRD patterns of the powder remi-
niscent of that of C3-NT4N with an intense peak at a low angle
which suggests the presence of a long axis.31
The XRD plots of vacuum evaporated lms on the PMMA
substrate for compounds C4–C8-NT4N are shown in Fig. 5a. The
high intensity in the 0 0 l reections of the C3-NT3N device
reveals the presence of a strong preferential orientation of the
molecules with respect to the substrate in the lm. The proles
are very similar to one another differing in the position of the
main reections. Indeed each prole shows two or more intense
and sharp reections that are further orders of the interlayer
distance reported in the inset of Fig. 5a.
The value of the distance progressively increases on
increasing the length of the alkyl chain. Based on the prefer-
ential orientation parameter obtained from C3-NT3N, for which
the single crystal structure is known, the sketch of the molec- Fig. 6 Left, n-type locus curves, right, n-type multiple output curves
for compounds C1, and C8-NT4N (a and b, respectively). In each
graph, the blue dotted curves represent the device drain–source
current (referred to the left scale), while the purple dotted curves
represent the emitted optical power (referred to the right scale). The
ular organization in thin lms is depicted in Fig. 5b.
The molecules are arranged in the upright position with respect
to the substrate with a minimal interdigitation of the alkyl ends.
Due to the strict similarity of the XRD patterns and to the
p(n)-type locus curves, multiple output and saturation transfer curves
trend of the interlayer distance with the alkyl chain length, we for all compounds are reported in the ESI.†
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 3448–3456 | 3453

View Article Online
Journal of Materials Chemistry C
Paper
Table 3 Electrical parameters for compounds in Chart 1 measured
under the same experimental conditions. Film thickness 100 nm
me
[cm2 Vꢀ1 sꢀ1
mh
[cm2 Vꢀ1 sꢀ1
EP
[nW]
Compound
]
]
VNT/V
VPT/V
C1-NT4N
C3-NT4N
C4-NT4N
C6-NT4N
C8-NT4N
C6cyc-NT4N
C6br-NT4N
3.0 ꢂ 10ꢀ1
2.4 ꢂ 10ꢀ1
2.9 ꢂ 10ꢀ1
1.7 ꢂ 10ꢀ1
3.2 ꢂ 10ꢀ1
2.9 ꢂ 10ꢀ2
Nd
2.8 ꢂ 10ꢀ3
1.5 ꢂ 10ꢀ4
8.0 ꢂ 10ꢀ5
1.3 ꢂ 10ꢀ4
3.0 ꢂ 10ꢀ3
Nd
27
40
33
20
28
40
Nd
ꢀ61
ꢀ45
ꢀ70
ꢀ66
ꢀ62
Nd
nd
30
47
145
150
Nd
Nd
Nd
Nd
C6cyc-NT4N shows lower electrical performances (Fig. S6, ESI†).
Similarly, the branched C6-tail does not promote either charge
transport or EL. On the other hand, linear alkyl tail compounds
(C1–C8-NT4N) show ambipolar behaviour with a major n-type
contribution to charge transport. The hole mobility, mh, covers a
wide range of values, oscillating between 2 and 3.5 orders of
magnitude lower than electron mobility me that ranges between
1.7 and 3.2 ꢂ 10ꢀ1 cm2 Vꢀ1 sꢀ1. The observed major electron
charge transport was investigated by theoretical calculations in
the framework of the small polaron hopping model.23 A thor-
ough understanding of the correlation between the molecular
structure and charge mobility is provided by the investigation of
the inner reorganization energy li and the charge transfer
integral s. The former is an intrinsic molecular property,
packing independent, while the latter represents the electronic
coupling element between neighbouring molecules and there-
fore critically depends on the order in the solid state. Only
minor differences in li are observed by varying the alkyl-
substituent type and/or length, with all the values included in a
30–40 meV width range (see Table S3†). Moreover, the slight
promotion of p-type transport in terms of li (i.e. lih < lie) is in
strong disagreement with the measured FET mobilities
(Table 3).
Therefore, the predominance of n-type behaviour cannot be
understood at a molecular structure level but requires also the
analysis of the molecular packing. Consequently, the transfer
integrals of C3-NT4N, for which the crystal structure was
determined (and the XRD show comparable thin lm crystal
packing), were calculated to gain an insight into the packing-
charge transport relationships. The possible charge hopping
pathways are shown in Fig. 7.
As shown from data in Table 4, the highest electronic
couplings are computed for the dimers involving rst next-
neighbouring molecule pairs, namely pathways 1, 2 and 3. The
latter shows not negligible electron transfer integral values, se
while only one pathway, namely pathway 1, has a signicant
hole transfer integral sh. All the directions are, therefore,
favourable for electron transport (i.e. me) while only one allows
the hole transport (i.e. mh).
Fig. 7 Charge hopping pathway schemes for molecule C3-NT4N in
the bulk crystal. y and z axis perpendicular and parallel to the molecular
axis, respectively.
Table 4 Computed charge transfer integrals s for all the pathways
having distances lower than 11 A˚ for molecule C3-NT4N. Displace-
ments between molecules reported by r
˚
r [A]
Pathway
sh [meV]
se [meV]
1
2
3
4
5
6
5.38
5.38
7.75
10.76
7.47
10.76
51.56
7.48
3.13
0
0.14
0.41
46.72
74.86
26.26
0.39
0.97
0.15
FET one of about three order of magnitudes. This discrepancy
may be ascribed to the different types of computed and exper-
imental mobility. At the theory level, the charge mobility is
described by Brownian motion of charge carriers with no spatial
constraints, while the experimental FET mobility strictly
depends on the source–drain potential tensor that means the
device conguration (e.g. lm orientation). In this sense, while
electrons can move along any crystal directions thanks to non-
negligible transfer integrals s, hole transport is signicantly
high only along one pathway (see Table 4). If the latter is
forbidden by device conguration, even a damping of some
orders of magnitude may become possible for mh. Thus, by
ignoring the pathway 1, we get m*calc,e ¼ 2.13 ꢂ 10ꢀ1 cm2 Vꢀ1 sꢀ1
and mc*alc,h ¼ 2.20 ꢂ 10ꢀ3 cm2 Vꢀ1 sꢀ1, in much better agreement
with the experimental FET mobility.
Table 5 Charge mobilities for molecule C3-NT4N: experimental FET
mobility mFET and computed bulk mobilities mcalc and m*calc obtained by
considering all the pathways reported in Fig. 7 and excluding pathway
1, respectively
In Table 5 bulk charge mobilities, computed in the non-adia-
batic hopping model mcalc (see ESI† for details), have been
mFET [cm2 Vꢀ1 sꢀ1
1.5 ꢂ 10ꢀ4
]
mcalc [cm2 Vꢀ1 sꢀ1
]
m*calc [cm2 Vꢀ1 sꢀ1
]
reported for a comparison with the experimental FET ones, mFET
.
Hole
1.15 ꢂ 10ꢀ1
1.82 ꢂ 10ꢀ1
2.20 ꢂ 10ꢀ3
2.13 ꢂ 10ꢀ1
The electron mobility shows a good agreement with experi-
ments, while the hole mobility dramatically overestimates the
Electron 2.4 ꢂ 10ꢀ1
3454 | J. Mater. Chem. C, 2014, 2, 3448–3456
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
Journal of Materials Chemistry C
Press, 2010; (b) H. Klauk, Organic Electronics: Materials,
Manufacturing and Applications, Wiley-VCH, Weinheim,
Germany, 2006; (c) Z. Bao and J. Locklin, Organic Field-
Effect Transistors, CRC Press, 2007; (d) C.-A. Di, F. Zhang
and D. Zhu, Adv. Mater., 2013, 25, 313–330.
2 (a) P. M. Beaujuge, H.-N. Tsao, M. R. Hansen, C. M. Amb,
C. Risko, J. Subbiah, R. K. Choudhury, A. Mavrinskiy,
´
¨
W. Pisula, J.-L. Bredas, F. So, K. Mullen and J. R. Reynolds,
J. Am. Chem. Soc., 2012, 134, 8944–8957; (b) J. Mei, Y. Diao,
A. L. Appleton, L. Fang and Z. Bao, J. Am. Chem. Soc., 2013,
135, 6724–6742.
Fig. 8 (a) Images of a device in the OFF state and in the ON state
showing light emission visible by eyes. (b) Electroluminescence
spectra of compounds C3, C6, and C8-NT4N. (c) Optical imaging of
the emissive stripe motion in the channel achieved by tuning the gate
voltage for compound C6-NT4N.
3 (a) J. Mei, D. H. Kim, A. L. Ayzner, M. F. Toney and Z. Bao,
J. Am. Chem. Soc., 2011, 133, 20130–20133; (b) K. Takimiya,
H. Ebata, K. Sakamoto, T. Izawa, T. Otsubo and Y. Kunugi,
J. Am. Chem. Soc., 2006, 128, 12604–12605; (c) J. Li, Y. Zhao,
H. S. Tan, Y. L. Guo, C.-A. Di, G. Yu, Y. Q. Liu, M. Lin,
S. H. Lim, Y. H. Zhou, H. B. Su and B. S. Ong, Sci. Rep.,
2012, 2, 754.
4 C. Wang, H. Dong, W. Hu, Y. Liu and D. Zhu, Chem. Rev.,
2012, 112, 2208–2267.
5 S. Holliday, J. E. Donaghey and I. McCulloch, Chem. Mater.,
2014, 26, 647–663.
6 (a) H. Minemawari, T. Yamada, H. Matsui, J. Tsutsumi,
S. Haas, R. Chiba, R. Kumai and T. Hasegawa, Nature,
2011, 475, 364–367; (b) H. Ebata, T. Izawa, E. Miyazaki,
K. Takimiya, M. Ikeda, H. Kuwabara and T. Yui, J. Am.
Chem. Soc., 2007, 129, 15732–15733.
7 A. Y. Amin, A. Khassanov, K. Reuter, T. Meyer-Friedrichsen
and M. Halik, J. Am. Chem. Soc., 2012, 134, 16548–16550.
8 J. E. Anthony, A. Facchetti, M. Heeney, S. R. Marder and
X. W. Zhan, Adv. Mater., 2010, 22, 3876–3892.
As shown by the curves in Fig. 6 (Fig. S8, ESI†) and data in
Table 3, the emission power signal of the device (taken as
comparative parameter) slightly increases on increasing the tail
length from C3 to C8. Being the thin lms quantum yields
almost comparable for all compounds, further investigation to
understand the exciton generation and decay processes within
the organic electroluminescent transistor device will be carried
out. The EL spectra of compounds C3, C6, C8-NT4N are shown
in Fig. 8b and are similar to the PL ones (Fig. 1b, c and f)
demonstrating that the EL emission truly comes from the
organic active layer. Fig. 8c shows the movement of the emissive
stripe inside the device channel by changing the applied bias, a
typical feature of ambipolar OLETs also previously observed for
other TBI materials including C4-NT4N.14
Conclusions
9 (a) K. Oniwa, T. Kanagasekaran, T. Jin, Md. Akhtaruzzaman,
Y. Yamamoto, H. Tamura, I. Hamada, H. Shimotani, N. Asao,
S. Ikeda and K. Tanigaki, J. Mater. Chem. C, 2013, 1, 4163–
4170; (b) A. R. Mohebbi, J. Yuen, J. Fan, C. Munoz,
M. F. Wang, R. S. Shirazi, J. Seier and F. Wudl, Adv.
Mater., 2011, 23, 4644–4648; (c) X. G. Guo, F. S. Kim,
M. J. Seger, S. A. Jenekhe and M. D. Watson, Chem. Mater.,
2012, 24, 1434–1442; (d) D. T. Chase, A. G. Fix, S. J. Kang,
B. D. Rose, C. D. Weber, Y. Zhong, L. N. Zakharov,
M. C. Lonergan, C. Nuckolls and M. M. Haley, J. Am. Chem.
Soc., 2012, 134, 10349–10352; (e) R. P. Ortiz, H. Herrera,
C. Seoane, J. L. Segura, A. Facchetti and T. J. Marks,
Chem.–Eur. J., 2012, 18, 532–543.
10 (a) R. Capelli, S. Toffanin, G. Generali, H. Usta, A. Facchetti
and M. Muccini, Nat. Mater., 2010, 9, 496–503; (b)
M. Muccini, Nat. Mater., 2006, 5, 605–613; (c) J. Zaumseil
and H. Sirringhaus, Chem. Rev., 2007, 107, 1296–1323; (d)
J. A. Misewich, R. Martel, P. Avouris, J. C. Tsang, S. Heinze
and J. Tersoff, Science, 2003, 300, 783–786.
In summary, we investigated the effect of the N-alkyl end shape
and size on the optical, self-assembly and charge transport
properties of a new rational family of thieno(bis)imide based
materials. Contrarily to what generally observed for conven-
tional oligothiophenes, no signicant effects on the optical
properties, thin lms morphology and charge mobility resulted
from the different structural end-motifs. Major electron charge
transport with mobility up to 0.3 cm2 Vꢀ1 sꢀ1 was measured in
OFET conguration, and rationalized by theoretical calcula-
tions. Tail dependent electroluminescence was also observed
suggesting the potential key role of the N-alkyl end of the TBI
moiety for the selective modulation of the electroluminescence
of these materials for ambipolar single layer OLET device
applications. Further studies in this direction are currently
under way.
Acknowledgements
F.G. is grateful to Dr M. Urbano, Prof. C. Castiglioni and Dr D.
Fazzi for useful discussions.
11 Y. Zhao, Y. Guo and Y. Liu, Adv. Mater., 2013, 25, 5372–5391.
12 (a) H. Tamura, I. Hamada, H. Shang, K. Oniwa,
Md. Akhtaruzzaman, T. Jin, N. Asao, Y. Yamamoto,
T. Kanagasekaran, H. Shimotani, S. Ikeda and K. Tanigaki,
J. Phys. Chem. C, 2013, 117, 8072–8078; (b) T. Sakanoue,
M. Yahiro, C. Adachi, H. Uchiuzou, T. Takahashi and
A. Toshimitsu, Appl. Phys. Lett., 2007, 90, 171118; (c)
Notes and references
1 (a) A. J. Heeger, N. S. Saricici and E. B. Namdas,
Semiconducting and Metallic Polymers, Oxford University
This journal is © The Royal Society of Chemistry 2014
J. Mater. Chem. C, 2014, 2, 3448–3456 | 3455

View Article Online
Journal of Materials Chemistry C
Paper
R. Capelli, F. Dinelli, S. Toffanin, F. Todescato, M. Murgia,
M. Muccini, A. Facchetti and T. J. Marks, J. Phys. Chem. C,
2008, 112, 12993–12999; (d) T. Yamao, Y. Shimizu,
K. Terasaki and S. Hotta, Adv. Mater., 2008, 20, 4109–4112;
(e) K. Yamane, H. Yanagi, A. Sawamoto and S. Hotta, Appl.
2128–2136; (i) H. B. Akkerman, S. C. B. Mannsfeld,
A. P. Kaushik, E. Verploegen, L. Burnier, A. P. Zoombelt,
J. D. Saathoff, S. Hong, S. Atahan-Evrenk, X. Liu,
A. Aspuru-Guzik, M. F. Toney, P. Clancy and Z. Bao, J. Am.
Chem. Soc., 2013, 135, 11006–11014.
Phys. Lett., 2007, 90, 162108; (f) T. Oyamada, H. Sasabe, 19 (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652; (b)
C. Adachi, S. Okuyama, N. Shimoji and K. Matsushige,
Appl. Phys. Lett., 2005, 86, 093505; (g) A. Dadvand,
C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens.
Matter Mater. Phys., 1988, 37, 785–789.
A. G. Moiseev, K. Sawabe, W. H. Sun, B. Djukic, I. Chung, 20 (a) R. Dovesi, R. Orlando, B. Civalleri, C. Roetti,
T. Takenobu, F. Rosei and D. F. Perepichka, Angew. Chem.,
Int. Ed., 2012, 51, 3837–3841.
V. R. Saunders and C. M. Z. Zicovich-Wilson,
Kristallograya, 2005, 220, 571–573; (b) R. Dovesi,
V. R. Saunders, C. Roetti, R. Orlando, C. M. Zicovich-
Wilson, F. Pascale, B. Civalleri, K. Doll, N. M. Harrison,
I. J. Bush, P. D'Arco and M. Llunell, CRYSTAL09
(CRYSTAL09 User's Manual), University of Torino, Torino,
2009.
13 M. Melucci, M. Zambianchi, L. Favaretto, M. Gazzano,
A. Zanelli, M. Monari, R. Capelli, S. Troisi, S. Toffanin and
M. Muccini, Chem. Commun., 2011, 47, 11840–11841.
14 M. Melucci, L. Favaretto, M. Zambianchi, M. Durso,
M. Gazzano, A. Zanelli, M. Monari, M. G. Lobello, F. De
Angelis, V. Biondo, G. Generali, S. Troisi, W. Koopman, 21 M. Corno, C. Busco, B. Civalleri and P. Ugliengo, Phys. Chem.
S. Toffanin, R. Capelli and M. Muccini, Chem. Mater., 2013,
25, 668–676.
15 M. Durso, D. Gentili, C. Bettini, A. Zanelli, M. Cavallini, F. De
Chem. Phys., 2006, 8, 2464–2472.
22 C. Gatti, V. R. Saunders and C. Roetti, J. Chem. Phys., 1994,
101, 10686–10696.
Angelis, M. G. Lobello, V. Biondo, M. Muccini, R. Capelli and 23 M. A. Spackman and A. S. Mitchell, Phys. Chem. Chem. Phys.,
M. Melucci, Chem. Commun., 2013, 49, 4298–4300. 2001, 3, 1518–1523.
16 M. Durso, C. Bettini, A. Zanelli, M. Gazzano, M. G. Lobello, 24 V. Maslyuk, C. Tegenkamp, H. Pfnur and T. Bredow,
F. De Angelis, V. Biondo, D. Gentili, R. Capelli, ChemPhysChem, 2006, 7, 1055–1061.
M. Cavallini, S. Toffanin, M. Muccini and M. Melucci, Org. 25 S. J. Grimme, J. Comput. Chem., 2006, 27, 1787–1799.
Electron., 2013, 14, 3089–3097.
26 (a) A. Troisi, Chem. Soc. Rev., 2011, 40, 2347–2358; (b)
J. L. Bredas, D. Beljonne, V. Coropceanu and J. Cornil,
Chem. Rev., 2004, 104, 4971–5004; (c) S. Di Motta, E. Di
Donato, F. Negri, G. Orlandi, D. Fazzi and C. Castiglioni, J.
Am. Chem. Soc., 2009, 131, 6591.
17 D. Gentili, M. Durso, C. Bettini, I. Manet, M. Gazzano,
R. Capelli, M. Muccini, M. Melucci and M. Cavallini, Sci.
Rep., 2013, 3, 2581.
18 (a) J. Liu, Y. Zhang, H. Phan, A. Sharenko, P. Moonsin,
B. Walker, V. Promarak and T.-Q. Nguyen, Adv. Mater., 27 M. Pope, C. E. Swenberg, Electronic Processes in Organic
2013, 25, 3645–3650; (b) T. Takahashi, S. Li, W. Y. Huang,
Crystals and Polymers, Oxford University Press, New York,
F. Z. Kong, K. Nakajima, B. J. Shen, T. Ohe and K. Kanno,
2nd edn, 1999.
J. Org. Chem., 2006, 71, 7967–7977; (c) Q. Miao, X. L. Chi, 28 (a) M. Cavallini, J. Mater. Chem., 2009, 19, 6085–6092; (b)
S. X. Xiao, R. Zeis, M. Lefenfeld, T. Siegrist,
M. L. Steigerwald and C. Nuckolls, J. Am. Chem. Soc., 2006,
M. Cavallini, M. Facchini, M. Massi and F. Biscarini, Synth.
Met., 2004, 146, 283–286.
128, 1340–1345; (d) J. E. Anthony, D. L. Eaton and 29 R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel
S. R. Parkin, Org. Lett., 2002, 4, 15–18; (e) M. D. Curtis, and T. A. Witten, Nature, 1997, 389, 827–829.
J. Cao and J. W. Kampf, J. Am. Chem. Soc., 2004, 126, 4318– 30 DFT results, however, show that the anti-conformer is
4328; (f) M. Halik, H. Klauk, U. Zschieschang, G. Schimd,
W. Radlik, S. Ponomarenko, S. Kirchmeyer and W. Weber,
J. Appl. Phys., 2003, 93, 2977–2981; (g) A. J. Lovinger,
H. E. Katz and A. Dodabalapur, Chem. Mater., 1998, 10,
energetically favored by 1.1 kcal molꢀ1 versus the syn-
conformer. This energy difference may be overcome by
crystal packing effects (cohesion energy for crystal
structure Ecoh ¼ 19.92 kJ molꢀ1).
3275–3277; (h) A. Kreyes, A. Mourran, Z. H. Hong, 31 The powder pattern of C4-NT4N does not correspond to that
¨
J. B. Wang, M. Moller, F. Gholamrezaie, W. S. C. Roelofs,
reported in ref. 13, hence powder and lm corresponds to a
D. M. de Leeuw and U. Ziener, Chem. Mater., 2013, 25,
new polymorph.
3456 | J. Mater. Chem. C, 2014, 2, 3448–3456
This journal is © The Royal Society of Chemistry 2014