Solution processed ter-anthrylene-ethynylenes for annealing-activated organic field-effect transistors: A structure-performance correlation study

Article abstract of DOI:10.1039/c1jm11957e

OFETs based on new solution-processed ester functionalized 9,10-ter-anthrylene-ethynylenes show a mobility increase of four orders of magnitude, leading to mobilities as high as 4.9 × 10^-2 cm ² V^-1 s^-1 if the deposited film is annealed before contact deposition. The behavior is ascribed to an increase in film order at the dielectric/semiconductor interface as revealed by X-ray studies.

Full text of DOI:10.1039/c1jm11957e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2011, 21, 15186
www.rsc.org/materials
COMMUNICATION
Solution processed ter-anthrylene-ethynylenes for annealing-activated organic
field-effect transistors: a structure–performance correlation study†
Giuseppe Romanazzi,a Antonio Dell’Aquila,a Gian Paolo Suranna,a Francesco Marinelli,b Serafina Cotrone,b
c
Davide Altamura, Cinzia Giannini, Luisa Torsi and Piero Mastrorilli
c
b
a
*
*
*
Received 4th May 2011, Accepted 27th July 2011
DOI: 10.1039/c1jm11957e
detection.10,11 Subsequently, we have investigated other alkyl func-
tionalized ter-anthrylene-ethynylenes confirming the good semi-
conducting performances of the chromophore.12
OFETs based on new solution-processed ester functionalized 9,10-
ter-anthrylene-ethynylenes show a mobility increase of four orders
of magnitude, leading to mobilities as high as 4.9 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1
if the deposited film is annealed before contact deposition. The
behavior is ascribed to an increase in film order at the dielectric/
semiconductor interface as revealed by X-ray studies.
Aiming at further exploring the semiconducting potentials of our
molecular architecture and at obtaining polar materials for potential
applications in chem-OFET, we embarked on the preparation of ter-
anthrylene-ethynylenes functionalised in their terminal position with
ester groups bearing ethyl (E1), n-butyl (E2), and n-octyl (E3) chains
(Fig. 1). These materials show a dramatic improvement in their
OFET performances upon a straightforward thermal annealing
process, a behavior that we deemed worthwhile to further investigate.
The desired target molecules E1–E3 are red powders and were
obtained in 67%, 55%, and 57% yield, respectively, by optimizing our
synthetic approach,10,12 described in Scheme S1 (ESI†). The oligomer
E1 was markedly less soluble with respect to E2 and E3 in common
organic solvents. The structure and purity of E1–E3 were fully
Focused efforts are still required for deepening our knowledge on the
structure–property relations in materials for organic field-effect
transistors (OFETs)1 and especially for solution processable semi-
conductors, which are particularly interesting in this respect.2 Such
studies can be beneficial also for very promising applications
employing OFETs as biological and environmental sensors, (chem-
or bio-OFETs).3 In this framework, the presence of ad hoc chosen
polar functional groups covalently bound to the semiconductor
structure might allow for an improved degree of selectivity towards
chemically homologous species.4 Moreover, the functionalization
of an organic semiconductor with polar groups can increase its
solubility and processability.5 Concerning acene-based organic
semiconductors,6,7 much interest has been focused on molecular
materials based on anthracene due to its high single-crystal mobility.8
Among the strategies aimed at extending the anthracene p-system,7
conjugation extension through the 9,10-positions has recently
emerged.9 In this framework we have described the synthesis and
properties of the ter-anthrylene-ethynylene chromophore, constituted
by a monodispersed architecture in which three neighbouring
anthracenes are linked by two ethynylene groups. The 9,10-bis[(10-
decylanthracen-9-yl)ethynyl]anthracene showed an average field-
effect mobility (mFET) of 1.2 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1 with on/off ratios
higher than 104, and was also used in chem-OFETs for NOx
1
confirmed by H, 13C-APT NMR, FT-IR spectroscopy, HRMS
spectrometry, as well as elemental analysis (ESI†). A very high
thermal stability was revealed by the TGA analysis of E1–E3 that
indicates (Fig. S1†) 5% weight loss temperatures higher than 360 ꢂC.
The spectroscopic properties of E1–E3 were investigated by UV-
Vis and photoluminescence spectroscopy (Fig. S2 and Table S1†)
both in chloroform solution and in the solid state. In solution, the
absorption and emission spectra of the three compounds are almost
superimposable and show a structureless absorption band with
a maximum at 500 nm, ascribed to the p–p* transition of the
aDipartimento di Ingegneria delle Acque e di Chimica (DIAC), Politecnico
di Bari, via Orabona 4, 70125 Bari, Italy. E-mail: p.mastrorilli@poliba.it;
Fax: +39 080 5963611; Tel: +39 080 5963605
bDipartimento di Chimica, Universitaꢀ degli Studi di Bari ‘‘A. Moro’’, via
Orabona 4, 70125 Bari, Italy. E-mail: torsi@chimica.uniba.it; Fax: +39
080 5442026; Tel: +39 080 5442092
cIstituto di Cristallografia - Consiglio Nazionale delle Ricerche (IC-CNR),
via Amendola 122/O, 70126 Bari, Italy. E-mail: cinzia.giannini@ic.cnr.it;
Fax: +39 080 592 9170; Tel: +39 080 592 9167
† Electronic supplementary information (ESI) available: Materials and
methods, synthesis of E1–E3, TGA traces, UV-Vis and PL spectra, CV
scans, output characteristics of OFET. See DOI: 10.1039/c1jm11957e
Fig. 1 Structure of the ester functionalised ter-anthrylene-ethynylenes
E1, E2 and E3.
15186 | J. Mater. Chem., 2011, 21, 15186–15189
This journal is ª The Royal Society of Chemistry 2011

View Article Online
conjugated backbone. The emission spectra in solution show a lem of
561 nm along with a shoulder at longer wavelengths. The Stokes shift
of E1–E3 is of 61 nm, comparable to that observed for the alkyl
functionalised ter-anthrylene-ethynylenes.10,12 The solid state optical
properties of E1 could not be recorded due to the low quality of the
obtained film. The solid state UV-Vis spectra of E2 and E3 show that
the main absorption peak at 512 nm (E2) and 492 nm (E3) is
accompanied by a shoulder at ꢃ550 nm. The solid state photo-
were obtained for the octyl functionalised E3: the average mFET was
4.4 ꢄ 0.1 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1 with on/off ratios of ꢃ104 and Vth
between ꢁ2 V and ꢁ10 V. The best mFET was 4.9 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1.
To the best of our knowledge, this is the first time that such
a dramatic performance improvement is seen on OFET upon
annealing.15 This remarkable behavior prompted us to study the
structural features of the semiconducting thin films before and after
annealing by complementing X-ray specular reflectivity (XSR) and
X-ray diffraction (XRD) techniques. To this purpose, thin-films of
E1–E3 were deposited on SiO2/Si substrates using the conditions
previously described for OFET device fabrication.
luminescence spectra are characterised by a broad emission at lem
¼
650 nm for E2 and at 653 nm for E3. These optical features are
symptomatic of strong intermolecular interactions between the
chromophores in the film. The very similar optical behavior observed
for the three molecules suggests analogous electronic structures for
E1–E3. This was further confirmed by the electrochemical investi-
gations carried out by cyclic voltammetry. In the oxidation scans, the
voltammograms of thin films of E1–E3 (Fig. S3†) are characterized
by the presence of an irreversible event; after calibration with Fc/
Fc+,13 the HOMO levels were derived by the onset of oxidation as
ꢁ5.6 eV for all three compounds.14 This value warrants a good
environmental stability of the structures. On the other hand the onset
of the reduction scans (Fig. S4†) allowed the estimate of the LUMO
of E1–E3 as ꢁ3.2 eV, leading to an electrochemical gap of ꢃ2.4 eV,
slightly higher than the optical band gap value (2.1 eV) extracted
from the absorption band edge (Table S1†).
In Fig. 3 the XSR experimental patterns, together with the corre-
sponding simulatꢂions for thin E1–E3 films both not-annealed (A) and
annealed at 100 C (B) for 30 min in vacuum, as a function of the
incidence angle q are reported.
XSR is a widely used technique to probe the variation of the
electron density in thin films in the direction perpendicular to
substrate plane.17 Interference fringes in reflectivity curves arise from
beam reflection at the top and at the bottom film interfaces (Kiessig
fringes) as well as at any inner interface (parallel to substrate plane)
between regions with different electron density. The d-spacing
between interfaces defining a periodically repeated structure in the
film thickness can be calculated as d ¼ 1/DQ, where Q ¼ 2sin q/l is
the length of the scattering vector, l is the X-ray wavelength, and DQ
is the distance between the Q values of two consecutive maxima in the
reflectivity curve, for a given modulation frequency. If molecules are
oriented and ordered along a given direction, they can form alter-
nating high and low electron density layers (bi-layers) giving rise to
interference fringes that can be detected by XSR. The higher the
interface sharpness, the electron density difference at interfaces, and
sample in-plane homogeneity, the larger the amplitude of reflectivity
fringes. From the first inspection of XSR curves in Fig. 3A and B, the
presence of flatter interfaces and/or higher lateral homogeneity in the
E2–E3 films, with respect to the E1 ones can be deduced. Moreover,
the reflectivity curves of the annealed E2–E3 films show two distinct
frequency modulations, associated to the film thickness and to bi-
layers stacked throughout the film, respectively. The two frequency
modulations are also evident in the not-annealed E3 film, while they
can barely be recognized in the analogous E2 sample, suggesting
a higher degree of molecular ordering in the as-deposited E3 film with
respect to the E2 one. The two modulation frequencies become
markedly visible for E2 after the annealing. Also for E3 the annealing
leads to more pronounced reflectivity fringes, indicating an
improvement in the degree of molecular layers ordering.
Next, E1–E3 were used as active layers in OFET devices fabricated
in top-contact bottom-gate configuration. The active layer was a thin
film of E1–E3 deposited onto a SiO2(300 nm)/n-doped Si substrate by
spin coating from a 0.05/0.1% w/w chloroform solution. The study of
the device features revealed a very intriguing behavior of these
materials.
Two sets of devices were tested: in the first, the gold ohmic contacts
were deposited onto the semiconductor film after the spin coating
steps, without any thermal pretreatment. For the second set of
devices, a thermal treatment at 100 ꢂC for 30 min in vacuum preceded
the gold contact evaporation. This treatment is usually employed to
eliminate traces of volatile impurities and induce an increase of the
grain dimension in the active layer. In some instances, it leads to an
increase in the structural order of the organic films, eventually
improving electronic performances. Usually, the thermal annealing
treatment is carried out on materials already showing quite good
OFET performances and it leads, in the best cases, to non-dramatic
improvements of their field-effect mobility and on/off ratios.15
In our case, the current-voltage characteristics of the E1–E3 based
OFET that were not subjected to any thermal treatment were all of
extremely poor quality, and did not show any appreciable current
modulation (Fig. S5†). On the other hand, while the device based on
E1 showed a very low mFET (ꢃ10ꢁ6 cm2 Vꢁ1 sꢁ1) even after the
annealing, a completely different behavior was observed for the
devices bearing the thermally annealed films of E2 (Fig. S6†) and E3
(Fig. 2), which showed well-shaped transistor characteristics exhib-
iting clear linear and saturation regions along with a dramatic
increase of mobility (mFET z 10ꢁ2 cm2 Vꢁ1 sꢁ1, Table 1) extracted
following standard data treatment.16
In contrast, in the case of the E1 samples the annealing process is
not sufficient to induce a significant improvement of molecular
ordering. The XSR scan of the not-annealed E1 film shows a single
modulation frequency, corresponding to a bi-layer thickness of 2.9
nm. Based on the simulation of the reflectivity curve, only one bi-
layer can be recognized close to the SiO2 surface. The annealed E1
sample shows similar features, with even less visible fringes, which
indicates lower film homogeneity and/or molecular ordering after the
annealing.
Similar to what was observed for alkyl functionalised ter-anthry-
lene-ethynylenes,12 a marked effect of the alkyl chain length was also
observed on the OFET figures of merit. For E2, bearing n-butyl
chains, an average mFET of 2.8 ꢄ 0.3 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1 with an on/
off ratio of 104 was obtained. The devices exhibited very low
threshold voltages (Vth) between ꢁ5 V and ꢁ8 V. Better mobilities
Table 2 reports the values of the overall layer thickness, the mean
values of bi-layer thicknesses obtained from simulations of the XSR
curves, and the derived number of stacked bi-layers for E1–E3 films.
From these values, it is apparent that all samples are characterized
by a similar thickness of the overall film and of the bi-layers. In the E3
thin films the number of stacked bi-layers is the highest, both before
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15186–15189 | 15187

View Article Online
Fig. 2 Output characteristics (left) and transfer characteristics (right) at Vds ¼ ꢁ100 V for the annealed top contact OFET embodying E3 as active
layer, resulting in a mFET of 4.9 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1, an on/off of 104 and a threshold value of ꢁ7.6 V.
Table 2 Multilayer structure of pristine and annealed thin films of E1–
E3 on SiO2, derived by simulation of the XSR curves
Table 1 Figures of merit for the OFET based on E1–E3 (active layer
annealed at 100 ꢂC for 30 min under vacuum)
a
Molecule
mFET/cm2 Vꢁ1 sꢁ1
Ion/Ioff
Vth/V (min/max)
Film
thickness/nm
Average bi-layer
thickness/nm
Number of
stacked bi-layers
Sample
E1
E2
E3
ꢃ10ꢁ6
4.4 ꢄ 0.1 ꢀ 10ꢁ2
102
104
104
ꢁ10
2.8 ꢄ 0.3 ꢀ 10ꢁ2
ꢁ5/ꢁ8
ꢁ2/ꢁ10
E1
11.8
10.6
8.1
9.2
9.9
2.9
—
2.9
3.1
3.3
3.4
1
1
2
3
3
3
E1a
E2
a
E2a
E3
Statistics carried out on 5 individual devices.
E3a
10.4
a
After annealing at 100 ꢂC for 30 min under vacuum.
and after the annealing; moreover, the bi-layers are repeated
throughout the whole film thickness, indicating that the E3 films are
endowed with the highest degree of order in the lamellar stacking. On
the other hand, an improvement in the degree of order is observed in
the annealed E2 sample with respect to the not-annealed E2 one, as
testified by the increase in the number of stacked bi-layers. In
contrast, no significant changes are detected for E1, for which a single
ordered bi-layer is evaluated at most.
In order to gain insight into the nature of the low frequency fringes
observed in reflectivity curves we performed XRD experiments on
much thicker E2–E3 spin coated films obtained from 0.2/0.4% w/w
solutions.
Fig. 4 shows the XRD and XSR patterns as a function of the
scattering angle 2q, for the thicker E3 and E2 films. XSR curves of the
corresponding thin films are also reported for comparison.
Inspection of the XSR part of Fig. 4A reveals that thicker films
show fewer reflectivity features, however the low frequency fringes,
associated with the bi-layer thickness, are still observable (and high-
lighted by square symbols) for the E3 sample. On the other hand
equally spaced diffraction peaks are observed in the XRD patterns of
all thicker samples, regardless of the thermal treatment, which indi-
cates a lamellar ordering of molecules in the film bulk. The peaks
related to the stacking periodicity (d-spacing) of the lamellae are
indicated in Fig. 4 by dotted or continuous line bars, together with the
calculated d-spacing d ¼ 1/DQ, where DQ is the difference in scat-
tering vector length between two consecutive diffraction peaks.17
In Fig. 4A dotted line bars are relevant to the main stacking period
in the E3 film, which is 3.0 nm; a slightly smaller periodicity (2.9 nm)
can also be recognized and marked by continuous line bars.
Fig. 3 XSR (experimental and simulated) for not-annealed (A) and
annealed (B) E1–E3 films. The low frequency modulation fringes are
clearly visible and marked by square symbols for the E3 films.
15188 | J. Mater. Chem., 2011, 21, 15186–15189
This journal is ª The Royal Society of Chemistry 2011

View Article Online
In conclusion, we have reported on the intriguing OFET behavior
of new solution processable 9,10-ter-anthrylene-ethynylenes func-
tionalised with carboethoxy (E1), carbobutoxy (E2), and carboctoxy
(E3) groups. The new compounds were investigated as organic
semiconductors in solution processed OFET devices, which showed
an intriguing activation by thermal annealing. In fact, very low field-
effect mobility was recorded for the as-spun films, while a very good
modulation and a 4 orders of magnitude higher field-effect mobility
crops up for E2 and E3-based devices after a mild annealing of the
active layer. The devices based on E3 showed the best figures of merit
(average mFET: 4.4 ꢀ 10ꢁ2 cm2 Vꢁ1 sꢁ1 and on/off ratios of 104). The
observed behavior was attributed to an increase in molecular order at
the interface between the annealed films and the gate dielectric, as
revealed by XSR characterization.
Part of this work has been supported by the SEED project ‘‘X-ray
synchrotron class rotating anode microsource for the structural
micro-imaging of nanomaterials and engineered biotissues (XMI-
LAB)’’- IIT protocol no. 21537 of 23/12/2009.
Notes and references
ꢁ
1 A. R. Murphy and J. M. J. Frechet, Chem. Rev., 2007, 107, 1066;
Y. Yamashita, Sci. Technol. Adv. Mater., 2009, 10, 024313.
2 A. C. Arias, J. D. MacKenzie, I. McCulloch, J. Rivnay and A. Salleo,
Chem. Rev., 2010, 110, 3; S. Allard, M. Forster, B. Souharce,
H. Thiem and U. Scherf, Angew. Chem., Int. Ed., 2008, 47, 4070;
M. Mas-Torrent and C. Rovira, Chem. Soc. Rev., 2008, 37, 827.
3 T. Someya, A. Dodabalapur, J. Huang, K. C. See and H. E. Katz,
Adv. Mater., 2010, 22, 3799; M. E. Roberts, A. N. Sokolov and
Z. Bao, J. Mater. Chem., 2009, 19, 3351.
4 L. Torsi, G. M. Farinola, F. Marinelli, M. C. Tanese, O. H. Omar,
L. Valli, F. Babudri, F. Palmisano, P. G. Zambonin and F. Naso,
Nat. Mater., 2008, 7, 412.
Fig. 4 XSR (log scale) experimental curves for thin and thick E3 (A) and
E2 (B) films, and XRD (linear scale) experimental curves for E3 (A) and
E2 (B) films before and after annealing.
5 G. Romanazzi, F. Marinelli, P. Mastrorilli, L. Torsi, A. Sibaouih,
€ €
M. Raisanen, T. Repo, P. Cosma, G. P. Suranna and C. F. Nobile,
Tetrahedron, 2009, 65, 9833, and references therein.
6 J. E. Anthony, Angew. Chem., Int. Ed., 2008, 47, 452.
7 J. E. Anthony, Chem. Rev., 2006, 106, 5028.
8 A. N. Aleshin, J. Y. Lee, S. W. Chu, J. S. Kim and Y. W. Park, Appl.
Phys. Lett., 2004, 84, 5382.
9 F. Silvestri, A. Marrocchi, M. Seri, C. Kim, T. J. Marks, A. Facchetti
and A. Taticchi, J. Am. Chem. Soc., 2010, 132, 6108.
10 A. Dell’Aquila, F. Marinelli, J. Tey, P. Keg, Y. M. Lam,
O. L. Kapitanchuk, P. Mastrorilli, C. F. Nobile, P. Cosma,
A. Marchenko, D. Fichou, S. G. Mhaisalkar, G. P. Suranna and
L. Torsi, J. Mater. Chem., 2008, 18, 786.
11 F. Marinelli, A. Dell’Aquila, L. Torsi, J. Tey, G. P. Suranna,
P. Mastrorilli, G. Romanazzi, C. F. Nobile, S. G. Mhaisalkar,
N. Cioffi and F. Palmisano, Sens. Actuators, B, 2009, 140, 445.
12 P. Keg, A. Dell’Aquila, F. Marinelli, O. L. Kapitanchuk, D. Fichou,
P. Mastrorilli, G. Romanazzi, G. P. Suranna, L. Torsi, Y. M. Lam
and S. G. Mhaisalkar, J. Mater. Chem., 2010, 20, 2448.
13 J. Pommerehne, H. Vestweber, W. Guss, R. F. Mark, H. Bassler,
M. Porsch and J. Daub, Adv. Mater., 1995, 7, 551.
14 D. M. De Leeuw, M. M. J. Simenon, A. R. Brown and
R. E. F. Einerhand, Synth. Met., 1997, 87, 53.
15 L. Zhang, C.-a. Di, G. Yu and Y. Liu, J. Mater. Chem., 2010, 20,
7059, and references therein.
16 D. Braga and G. Horowitz, Adv. Mater., 2009, 21, 1473.
Concerning E2 (Fig. 4B), the stacking period derived from the XRD
data is 2.5 nm. The average bi-layer thicknesses estimated from the
simulation of XSR curves are therefore consistent with the d-spacings
derived from XRD analyses. Interestingly, similar trends in stacking
periodicity of the E2–E3 samples are observed for the d-spacings
derived by XSR and XRD analysis, which are in agreement with the
difference in molecular length of E2 with respect to E3. Moreover, the
simultaneous presence of both sharp diffraction peaks and reflectivity
fringes in the XSR curve of the not-annealed E3 thick sample (hardly
visible in the analogous E2 sample) suggests that thick crystalline
domains are present in the E3 film bulk, while a thin laterally
homogeneous bi-layered structure is formed close to the interface
with SiO2. The thickness of such a bi-layered structure seems to be
only slightly affected by the thickness of the deposited film, as shown
by the barely visible fringes in the reflectivity curves of the thicker
films.
The fact that this interface thin layer becomes more ordered after
thermal annealing in the case of the E3 and E2 samples, as testified by
the more intense observed reflectivity fringes (Fig. 3B), can be taken
as the direct cause of the dramatic mobility enhancement recorded for
OFET devices and provides further evidence of the fact that in OFET
two dimensional transport occurs at the interface between the organic
layer and the gate dielectric.18
ꢁ
17 U. Pietsch, V. Holy and T. Baumbach, High Resolution X-Ray
Scattering: From Thin Films to Lateral Nanostructures, Springer-
Verlag, New York, 2nd edn, 2004.
18 A. Dodabalapur, L. Torsi and H. E. Katz, Science, 1995, 268, 270;
F. Dinelli, M. Murgia, P. Levy, M. Cavallini, F. Biscarini and
D. M. de Leeuw, Phys. Rev. Lett., 2004, 92, 116802, and references
therein.
This journal is ª The Royal Society of Chemistry 2011
J. Mater. Chem., 2011, 21, 15186–15189 | 15189