Aligning the band gap of graphene nanoribbons by monomer doping

Article abstract of DOI:10.1002/anie.201209735

A matter of doping: Graphene nanoribbons (GNRs) were generated through an on-surface bottom-up synthesis and selectively doped at their edges by introducing nitrogen atoms in the precursor monomers. While the size of the band gap of 2.8 eV remains almost unchanged upon N substitution, a linear shift of the band structure is observed and corresponds to n-type doping (see picture; CB=conduction band and VB=valence band). Copyright

Full text of DOI:10.1002/anie.201209735

Angewandte
Chemie
DOI: 10.1002/anie.201209735
Carbon Materials
Aligning the Band Gap of Graphene Nanoribbons by Monomer
Doping**
Christopher Bronner,* Stephan Stremlau, Marie Gille, Felix Brauße, Anton Haase,
Stefan Hecht,* and Petra Tegeder*
Silicon-based field-effect transistors (FETs) are the building
blocks of modern digital logic circuitry and therefore part of
virtually every electronic device available today. Over the
past decades, continuous downscaling of existing designs has
met the rising performance requirements, but as the size of
FETs approaches the regime of atomic structures, new
concepts are required to maintain the current pace at which
microelectronics is developing.[1] At small gate channel
lengths, the applicability of quantum mechanical principles
results in several so-called short-channel effects (e.g. reduced
carrier mobility).[2] Since its experimental realization in
2004,[3] graphene has been discussed intensively as a substitute
for doped silicon in FETs because of its high charge carrier
mobility and its unsurpassably low thickness.[1,4] Graphene
transistors have even been realized but cannot be put into the
“off” state because of the lack of a band gap.[5a,b] However,
there are concepts available for opening a band gap, for
example, applying strain[6a,b] along the sheet or biasing
bilayers of graphene.[7a,b] Also, lateral confinement in quasi-
one-dimensional graphene nanoribbons (GNRs) leads to
a band gap, which furthermore is highly sensitive to the width
and edge shape of the GNR, thus opening possibilities to
tailor the electronic properties of a device.[8a,b,c] Indeed, FETs
built from nanoribbons show much higher on/off-ratios than
graphene transistors, which makes them more suitable for
integration into logic devices.[9,10] However, the ability to
control the electronic properties is essential: while the size of
the gap can be engineered by varying the nanoribbon
widths,[8c] the alignment of the GNR band structure with
respect to the Fermi level of a metal electrode is equally
important. Such a shifting of the entire band structure is
observed both in two-dimensional graphene[11] as well as in
chemically synthesized or lithographically patterned GNRs[12]
upon doping, particularly with nitrogen atoms.[13] Using
present doping techniques, the distribution of dopant atoms
will not be well-defined on the nanoscale and the band gap
shift upon nitrogen doping depends on the site of the N atom,
that is, the bonding configuration to neighboring carbon
atoms.[14] Generally, for doped and pristine GNRs, fabrication
remains a challenge as well-established top-down approaches
using lithography[8c] or unzipping of carbon nanotubes[15] yield
relatively wide ribbons with an undetermined edge structure.
Particularly for small widths on the order of a few nanometers
(where the band gap reaches a technologically relevant size)
atomically precise edges are necessary and can be realized
using Br-substituted precursor molecules, which are thermally
activated on a surface and—in a bottom-up synthesis—
covalently assemble to a specific nanostructure.[16a,b] In this
study, we employed the latter approach to prepare GNRs with
an atomically precise edge structure and doping pattern
through polymerization of specific monomers directly on the
Au(111) surface and studied the position and size of the band
gap of these GNRs with surface-sensitive electron spectros-
copies.
Besides straight armchair edge GNRs, another type of
chevron-shaped nanoribbons has previously been fabricated
using an on-surface reaction.[16b] In this process adsorption of
several layers of 6,11-dibromo-1,2,3,4-tetraphenyl-tripheny-
lene (monomer 1 in Scheme 1) on Au(111) and heating at
250 8C leads to desorption of the second and higher layers as
well as halogen dissociation and coupling of the resulting
activated biradical monomers, yielding a sterically crowded
and hence twisted polyphenylene. In a second heating step at
4408C, this polymer undergoes a subsequent cyclodehydro-
genation reaction providing access to the desired chevron-
shaped GNR with armchair edges. Selective substitution of
the parent monomer 1 with either one or two N atoms
provided monomers 2 and 3, respectively, which were used
to generate GNRs with different doping levels (exemplarily
shown for the doubly N doped GNR 5 in Scheme 1) and
accordingly with potentially different electronic structure
properties. Synthesis of the new monomers 2 and 3 was
accomplished by Diels–Alder reactions of an appropriate
cyclopentadienone with either mixed phenylpyridyl-acety-
lene or bispyridylacetylene, followed by immediate chele-
tropic CO extrusion (for details see the Supporting Informa-
tion).
[*] C. Bronner, S. Stremlau, A. Haase, Prof. Dr. P. Tegeder
Fachbereich Physik, Freie Universitꢀt Berlin
Arnimallee 14, 14195 Berlin (Germany)
E-mail: bronner@zedat.fu-berlin.de
petra.tegeder@physik.fu-berlin.de
Dr. M. Gille, F. Brauße, Prof. Dr. S. Hecht
Department of Chemistry, Humboldt-Universitꢀt zu Berlin
Brook-Taylor-Straße 2, 12489 Berlin (Germany)
E-mail: sh@chemie.hu-berlin.de
Prof. Dr. P. Tegeder
Physikalisch-Chemisches Institut
Ruprecht-Karls-Universitꢀt Heidelberg
Im Neuenheimer Feld 253, 69120 Heidelberg (Germany)
[**] We gratefully acknowledge funding by the Focus Area Nanoscale at
the Freie Universitꢀt Berlin, the German Research Foundation
(DFG) through collaborative research center Sfb658, the European
Union via the AtMol project, and the European Science Foundation
via P2M. M.G. is indebted to the Fonds der chemischen Industrie
for providing a Kekulꢁ doctoral fellowship.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201209735.
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Scheme 1. Monomers for the generation of graphene nanoribbons with
defined edge-doping via on-surface polymerization followed by cyclo-
dehydrogenation. Shown are monomers 1–3 with either none, one or
two doping N-atoms, while the synthesis is exemplarily shown for the
case of the doubly N-doped GNR.
Figure 1. On-surface synthesis of the undoped GNR analyzed by vibra-
tional HREELS in specular and off-specular scattering geometry for the
three reaction steps: a) adsorbed monomers, b) polymer and c) gra-
phene nanoribbon. E0 is the primary energy of the incident electrons.
After deposition of the monomers on the surface, we
followed the two heating steps towards GNR formation using
vibrational spectroscopy (Figure 1), namely angle-resolved
high-resolution electron energy loss spectroscopy (HREELS)
in analogy to a similar system.[17] In the first heating step, the
higher monomer layers are desorbed leading to an overall
increase of the elastic peak (Figure 1b). Zero-order desorp-
tion, representative of multilayer desorption, is observed
using a quadrupole mass spectrometer monitoring the C6H5Br
vibrational modes having dipole moments within the phenyl
À
ring plane (including n(C H)) is a consequence of all phenyl
rings lying flat as part of the GNR. To estimate the GNR
coverage after such a preparation temperature programmed
desorption (TPD) experiments on co-adsorbed Xe were used
(see the Supporting Information). From these studies, a GNR
coverage of approximately 2/3 of a monolayer was deter-
mined. The preparation method as well as the character-
ization by means of HREELS and TPD were employed for all
three monomers (1–3), giving similar results. Furthermore,
the electronic HREELS (yielding electronic transition ener-
gies) signature changes drastically during both heating steps
(see the Supporting Information).
fragment ion (m/z = 158 amu). Furthermore, the loss peak at
À1
À
À
257 cm , which we assign to the C Br bending mode g(C
Br), vanishes because the Br atoms are detached from the
À1
À
molecule. The phenyl ring torsion mode t(C C) at 696 cm
and the out-of-plane bending mode g(C H) at 798 cmÀ1 show
À
a higher dipole activity mostly because in the first layer, the
triphenylene moiety lies flat on the surface. A reduced dipole
À1
À
À
activity of the C H stretching vibration n(C H) at 3063 cm
À
in the polymer phase results from a large portion of C H
bonds now pointing parallel to the surface. In the final GNR
Two complementary surface-sensitive spectroscopies
were employed to study the electronic structure, that is,
band gaps and energetic positions of the valence and
conduction bands, of the differently doped GNRs. Electronic
HREELS, which employs higher primary electron energies
compared to the above shown vibrational HREELS, gives
access to the width of the band gap by exciting electronic
À
phase (Figure 1c), the torsion mode t(C C) is shifted to
higher energies because of the changed environment within
the more rigid graphene structure and the increased inter-
action with the substrate. The dipole activity of the out-of-
À
À
plane bending mode g(C H) is increased since all C H bonds
are lying parallel to the surface. The quenching of the various
2
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Figure 3. Summary of band gap size and alignment relative to the
Fermi level of the gold substrate for the three investigated systems.
While the band gap remains unaffected, the entire band structure is
shifted toward lower energies.
marizes the energetic positions of the VB and CB of the
studied GNRs.
Figure 2. a) Electronic HREELS spectra for the three GNRs with differ-
ent doping concentrations together with fits to the valence band (VB)–
conduction band (CB) transition peak. The primary electron energy is
15.5 eV and the spectra are normalized with respect to the VB-CB
transition peak. b) UPS spectra recorded with a photon energy of 6 eV
showing the gold surface state and the valence band (VB) peak for all
three GNRs. UPS spectra were normalized to the VB peak.
In summary, we have successfully doped graphene nano-
ribbons in a well-defined manner by selective nitrogen
substitution of the precursor monomers which does not
interfere with the on-surface synthesis reaction. Using
HREELS in combination with photoelectron spectroscopy,
we could show that the band gap is linearly shifted relative to
the electronic structure of the environment of the GNR but
remains almost unchanged in magnitude, as expected for
pyridine-like nitrogen at the edges of armchair GNRs.[13] The
positions of the bands and thus the size of the band gap agree
well with other experimental data on the pristine GNR.[18]
The independence of the size of the band gap on one hand and
its alignment relative to the Fermi level on the other hand
could prove useful in tailoring the electronic properties of
GNR devices. Doping the molecular precursors in a bottom-
up fabrication technique furthermore allows a well-defined
site selection and dosage of the dopant atoms within defect-
free GNRs. This should add another powerful item to the
toolbox of band gap engineering of graphene nanostructures
and, together with width variation and edge shaping, thus
contribute to controlling their properties towards technolog-
ical applications.
valence band (VB)–conduction band (CB) transitions. While
electronic HREELS gives the size of the respective band gap,
the alignment of the GNR band gap relative to the electronic
structure of the underlying substrate is measured with ultra-
violet photoelectron spectroscopy (UPS). Figure 2a shows
electronic HREEL spectra for the pristine, undoped GNR
produced from monomer 1 and the two doped monomers 2
and 3 in which the VB-CB transition peak has been fitted with
Gaussian fit functions. The peak positions of the pristine
GNR and the singly doped are identical, 2.80 Æ 0.03 eV. The
band gap of the doubly N-doped GNR (5) is slightly lower,
2.71 Æ 0.01 eV, which corresponds to a reduction by 3%. UPS
spectra recorded from the three different GNRs show two
peaks: one at higher energies (i.e. closer to the Fermi level)
that increases in intensity as more nitrogen atoms are present
in the GNR, originates from the gold Shockley surface state.
Its higher intensity relative to the other peak is indicative of
decreasing coverages as we proceed to a higher doping level.
Since the preparation was identical for all three systems, the
varying coverage could be due to repulsive interactions
among the electro-negative N atoms, favoring monomer
desorption in the formation process compared to polymeri-
zation. The second peak originates from the GNR and
exhibits an energy dependence on the degree of doping.
Therefore, we assign the corresponding state to the valence
band. Its binding energy in the pristine GNR is À0.87 Æ
0.03 eV and decreases in the doped GNRs to À1.01 Æ
0.03 eV (one N atom per molecule) and À1.10 Æ 0.05 eV
(two N atoms), respectively. Note that the downward shift in
energy is in agreement with calculations[14] and experiments
on N-doped graphene[11] as well as top-down fabricated
graphene nanoribbons, the edge structure of which is not
atomically precise.[12] We thus observe a continuous down-
shifting of approximately 0.1 eV per N atom. Figure 3 sum-
Received: December 5, 2012
Revised: February 18, 2013
Published online: && &&, &&&&
Keywords: doping · electronic structure · graphene ·
.
molecular electronics · surface chemistry
[1] F. Schwierz, Nat. Nanotechnol. 2010, 5, 487 – 496.
[2] R. Chau, B. Doyle, S. Datta, J. Kavalieros, K. Zhang, Nat. Mater.
2007, 6, 810 – 812.
[3] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang,
S. V. Dubonos, I. V. Grigorieva, A. A. Firsov, Science 2004, 306,
666 – 669.
[4] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I.
Katsnelson, I. V. Grigorieva, S. V. Dubonos, A. A. Firsov, Nature
2005, 438, 197 – 200.
[5] a) I. Meric, M. Y. Han, A. F. Young, B. Ozyilmaz, P. Kim, K. L.
Shepard, Nat. Nanotechnol. 2008, 3, 654 – 659; b) F. Xia, D. B.
Farmer, Y.-M. Lin, P. Avouris, Nano Lett. 2010, 10, 715 – 718.
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[6] a) Z. H. Ni, T. Yu, Y. H. Lu, Y. Y. Wang, Y. P. Feng, Z. X. Shen,
ACS Nano 2008, 2, 2301 – 2305; b) T. M. G. Mohiuddin, A.
Lombardo, R. R. Nair, et al., Phys. Rev. B 2009, 79, 205433.
[7] a) Y. Zhang, T.-T. Tang, C. Girit, Z. Hao, M. C. Martin, A. Zettl,
M. F. Crommie, Y. R. Shen, F. Wang, Nature 2009, 459, 820 – 823;
b) T. Ohta, A. Bostwick, T. Seyller, K. Horn, E. Rotenberg,
Science 2006, 313, 951 – 954.
[8] a) K. Nakada, M. Fujita, G. Dresselhaus, M. S. Dresselhaus,
Phys. Rev. B 1996, 54, 17954 – 17961; b) Y.-W. Son, M. L. Cohen,
S. G. Louie, Phys. Rev. Lett. 2006, 97, 216803; c) M. Y. Han, B.
ꢀzyilmaz, Y. Zhang, P. Kim, Phys. Rev. Lett. 2007, 98, 206805.
[9] X. Li, X. Wang, L. Zhang, S. Lee, H. Dai, Science 2008, 319,
1229 – 1232.
[12] X. Wang, X. Li, L. Zhang, Y. Yoon, P. K. Weber, H. Wang, J.
Guo, H. Dai, Science 2009, 324, 768 – 771.
[13] H. Terrones, R. Lv, M. Terrones, M. S. Dresselhaus, Rep. Prog.
Phys. 2012, 75, 062501.
[14] F. Cervantes-Sodi, G. Csꢁnyi, S. Piscanec, A. C. Ferrari, Phys.
Rev. B 2008, 77, 165427.
[15] D. V. Kosynkin, A. L. Higginbotham, A. Sinitskii, J. R. Lomeda,
A. Dimiev, B. K. Price, J. M. Tour, Nature 2009, 458, 872 – 876.
[16] a) L. Grill, M. Dyer, L. Lafferentz, M. Persson, M. V. Peters, S.
Hecht, Nat. Nanotechnol. 2007, 2, 687 – 691; b) J. Cai, P.
Ruffieux, R. Jaafar, M. Bieri, T. Braun, S. Blankenburg, M.
Muoth, A. P. Seitsonen, M. Saleh, X. Feng, K. Mꢂllen, R. Fasel,
Nature 2010, 466, 470 – 473.
[10] X. Wang, Y. Ouyang, X. Li, H. Wang, J. Guo, H. Dai, Phys. Rev.
Lett. 2008, 100, 206803.
[11] a) D. Wei, Y. Liu, Y. Wang, H. Zhang, L. Huang, G. Yu, Nano
Lett. 2009, 9, 1752 – 1758; b) H. Wang, T. Maiyalagan, X. Wang,
ACS Catal. 2012, 2, 781 – 794.
[17] C. Bronner, F. Leyssner, S. Stremlau, M. Utecht, P. Saalfrank, T.
Klamroth, P. Tegeder, Phys. Rev. B 2012, 86, 085444.
[18] S. Linden, D. Zhong, A. Timmer, N. Aghdassi, J. H. Franke, H.
Zhang, X. Feng, K. Mꢂllen, H. Fuchs, L. Chi, H. Zacharias, Phys.
Rev. Lett. 2012, 108, 216801.
4
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Carbon Materials
C. Bronner,* S. Stremlau, M. Gille,
F. Brauße, A. Haase, S. Hecht,*
P. Tegeder*
&&&& — &&&&
Aligning the Band Gap of Graphene
Nanoribbons by Monomer Doping
A matter of doping: Graphene nanorib-
bons (GNRs) were generated through an
on-surface bottom-up synthesis and
selectively doped at their edges by intro-
ducing nitrogen atoms in the precursor
monomers. While the size of the band
gap of 2.8 eV remains almost unchanged
upon N substitution, a linear shift of the
band structure is observed and corre-
sponds to n-type doping (see picture;
CB=conduction band and VB=valence
band).
Angew. Chem. Int. Ed. 2013, 52, 1 – 5
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!