Aryl Radical Geometry Determines Nanographene Formation on Au(111)

Article abstract of DOI:10.1002/anie.201606440

The Ullmann coupling has been used extensively as a synthetic tool for the formation of C?C bonds on surfaces. Thus far, most syntheses made use of aryl bromides or aryl iodides. We investigated the applicability of an aryl chloride in the bottom-up assembly of graphene nanoribbons. Specifically, the reactions of 10,10′-dichloro-9,9′-bianthryl (DCBA) on Au(111) were studied. Using atomic resolution non-contact AFM, the structure of various coupling products and intermediates were resolved, allowing us to reveal the important role of the geometry of the intermediate aryl radicals in the formation mechanism. For the aryl chloride, cyclodehydrogenation occurs before dehalogenation and polymerization. Due to their geometry, the planar bisanthene radicals display a different coupling behavior compared to the staggered bianthryl radicals formed when aryl bromides are used. This results in oligo- and polybisanthenes with predominantly fluoranthene-type connections.

Full text of DOI:10.1002/anie.201606440

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201606440
German Edition: DOI: 10.1002/ange.201606440
Nanographene
Aryl Radical Geometry Determines Nanographene Formation on
Au(111)
Peter H. Jacobse, Adri van den Hoogenband, Marc-Etienne Moret,*
Robertus J. M. Klein Gebbink, and Ingmar Swart*
Abstract: The Ullmann coupling has been used extensively as
precursor molecule 10,10’-dibromo-9,9’-bianthryl (DBBA).[6]
The coupling is a two-step process, in which first the carbon–
halogen bonds are dissociated thermally and the resulting
(surface-bound) bianthryl radicals polymerize to give poly-
anthrylene.[7] By heating the polyanthrylene chains to over
2508C, cyclodehydrogenation (CDH) is induced, transform-
ing the staggered polyanthrylenes into flat, conjugated nano-
ribbons.[8] CDH releases hydrogen atoms onto the surface,
which has been shown to passivate leftover radicals.[9]
Nanoribbons of various widths, edge structures, and
heteroatom-functionalized GNRs have been synthesized.[10]
Interestingly, all of the precursor molecules were synthesized
by means of aryl halide coupling chemistry. The subsequent
on-surface synthesis has so far also always employed the aryl
bromide Ullmann coupling. Investigating the feasibility of
aryl chloride Ullmann coupling in the on-surface synthesis of
GNR is highly desirable as it may increase the flexibility and
selectivity of the solution-phase synthesis of the precursor
molecules.
À
a synthetic tool for the formation of C C bonds on surfaces.
Thus far, most syntheses made use of aryl bromides or aryl
iodides. We investigated the applicability of an aryl chloride in
the bottom-up assembly of graphene nanoribbons. Specifically,
the reactions of 10,10’-dichloro-9,9’-bianthryl (DCBA) on
Au(111) were studied. Using atomic resolution non-contact
AFM, the structure of various coupling products and inter-
mediates were resolved, allowing us to reveal the important
role of the geometry of the intermediate aryl radicals in the
formation mechanism. For the aryl chloride, cyclodehydroge-
nation occurs before dehalogenation and polymerization. Due
to their geometry, the planar bisanthene radicals display
a different coupling behavior compared to the staggered
bianthryl radicals formed when aryl bromides are used. This
results in oligo- and polybisanthenes with predominantly
fluoranthene-type connections.
T
he Ullmann coupling has proven to be an indispensable
tool for the bottom-up assembly of graphene nanostructures
and covalent organic frameworks on surfaces.[1] Numerous
structures can be accessed with aryl chlorides,[2] aryl bro-
mides[3] and aryl iodides.[4] The combination of different
halogen substituents in a single molecule has been used to
provide an extra degree of freedom as individual Ullmann
couplings can be carried out in sequence.[5] This hierarchical
approach is also useful in the preparation of precursor
molecules, which are typically synthesized by means of
solution-based coupling methods. The presence of two differ-
ent halogens then allows for solution-phase coupling on the
heaviest halogen, followed by surface-mediated Ullmann
coupling on the lightest one, providing increased synthetic
freedom.
Here, we report on our investigation of the applicability of
the aryl chloride Ullmann coupling in the thermal assembly of
graphene nanostructures. Specifically, the precursor 10,10’-
dichloro-9,9’-bianthryl (DCBA) was used, which is the chloro
analogue of the DBBA used by Cai to synthesize 7-armchair
graphene nanoribbons (7-acGNR), as shown in Scheme 1.
Scheme 1. Synthesis of 7-acGNR from bianthryl halides through sub-
sequent Ullmann coupling and cyclodehydrogenation. Numbering is
shown for a single anthryl unit.
Among the most fascinating structures produced with the
Ullmann coupling are graphene nanoribbons (GNR). The
archetypical synthesis, pioneered by Cai et al., uses the
We find that changing the halogen strongly impacts the
course of the reaction: instead of the expected graphene
nanoribbons, we observe the formation of fused oligo-
and polybisanthenes (bisanthene being the cyclodehydrogen-
ized form of bianthryl) with varying degrees of atomic order.
[*] P. H. Jacobse, Dr. I. Swart
Condensed Matter and Interfaces, Debye Institute for Nanomaterials
Science, Utrecht University
PO Box 80000, 3508 TA Utrecht (The Netherlands)
E-mail: i.swart@uu.nl
P. H. Jacobse, A. van den Hoogenband, Dr. M.-E. Moret,
Prof. R. J. M. Klein Gebbink
Organic Chemistry and Catalysis, Debye Institute for Nanomaterials
Science, Utrecht University
À
This difference is proposed to stem from the stronger C Cl
bond allowing for cyclodehydrogenation prior to radical
formation.
First, DCBA molecules were evaporated onto an Au(111)
crystal held at room temperature. As is evident from the STM
image shown in Figure 1a, a well-ordered self-assembled
layer was observed. The apparent height (4.0 Æ 0.5 ꢀ) and
Universiteitsweg 99, 3584 CG Utrecht (The Netherlands)
E-mail: m.moret@uu.nl
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201606440.
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
By having performed the assembly of DCBA in a slow-
heating experiment, we rule out a possible temperature
quenching effect: the hypothetical case that molecules are
heated too quickly to a temperature at which they may
undergo intramolecular reactions before having the chance to
couple.[12] Hence, the rationale behind slow heating is that it
maximizes the probability that all steps in the mechanism are
traversed in the order of increasing activation energy while
giving every single one of them enough time to occur before
possibly being quenched by reactions that happen at higher
temperature. Since both coupling and cyclodehydrogenation
reactions must have occurred for the polybisanthene chains to
form, it was decided to carry out the next experiments at
lower temperatures, closer to the onset of various reactions.
When evaporating molecules onto Au(111) maintained at
2008C, smaller, planar structures are observed. In AFM these
were found to correspond to mostly monomers, dimers and
oligomers of bisanthene (Figure 1e). In the STM images, the
vast majority of dimers appear as a structure where the two
monomers are attached at a 34 Æ 5 degree angle (based on 11
measurements). A few dimers feature a different angle and
appear to be fused even more intricately. AFM (Figure 1 f)
reveals both of these structures to be fused bisanthene dimers,
both featuring the non-alternant fluoranthene-type motif,
containing a single pentagon sandwiched between the two
moieties. Similar attachments are seen in oligomers (Fig-
ure 1d).
Evaporating molecules onto Au(111) heated to 1208C
results mostly in agglomerates of staggered molecules. In
addition, some flat molecules were observed, particularly
near the kink sites of the Au(111) herringbone reconstruction
(Figure 1g). Most of these flat species were identified as
bisanthene, with a smaller fraction of monochlorobisanthene
(Figure 1h). At even lower temperatures, no flat molecules
were observed, indicating that cyclodehydrogenation of
bianthryl species does not take place.
Figure 1. STM and AFM images of DCBA molecules on the Au(111)
surface treated at different temperatures. a) Room temperature.
b) Model of the DCBA. c) Heated to 3608C. d) AFM of structures in
(c). e) Heated to 2008C. f) AFM on structures in e. g) Heated to
1208C. h) AFM on structures in g. All scale bars are 5 nm.
The observation of monochlorobisanthene implies that
the DBCA molecules first undergo cyclodehydrogenation
and only then dehalogenation and polymerization. This order
is reversed with respect to the order found for the bromo-
analog (DBBA)[6,7a] (see Figure 2). This can be rationalized
from the fact that the carbon–chlorine bond is stronger than
the carbon–bromine bond.
When dehalogenation precedes cyclodehydrogenation,
that is, the case of DBBA, bianthryl radicals are formed.
The staggered geometry of bianthryl radicals is essential to
alleviate the steric repulsion between the peri hydrogen
atoms and allow for carbon–carbon coupling at the 10 and 10’
positions.[13] Only this coupling will result in well-defined
graphene nanoribbons. When the order of the reaction is
reversed, radical polymerization can only occur between
planar species that are lying flat on the surface. Here the
radical carbon atom is flanked by two peri-hydrogens, which
pose a significant sterical difficulty for the radical to approach
another molecule—particularly another radical. For the
planar radical, the most accessible position is now the
periodicity (1.2 Æ 0.2 nm) are consistent with a close packing
of DCBA molecules. Interestingly, multiple crystal structures
were identified within the monolayer, which we ascribe to
variations in the three-dimensional orientation of DCBA
molecules relative to each other. As shown in the Supporting
Information, the structures could be interconverted by
applying voltage pulses.
Subsequent slow heating of the sample from room
temperature to 3608C over the course of 1 h results in
a large decrease of adsorbate concentration. The remaining
molecules formed extensive disordered polymer-like chains,
near the Au(111) step edges, as seen in Figure 1c. These
chains have a different conjugation than graphene nano-
ribbons.[11] The formation of disordered chains implies that
the orientation and coupling of subsequent bisanthenes is
random and that branching of the chains is prevalent. This is
confirmed using sub-molecular resolution AFM images (Fig-
ure 1d). This is in obvious contrast with the linear structures
that result from radical–radical coupling of staggered
bianthryl radical intermediates, as is the case for DBBA.
À
corner, as shown in Scheme 2. C C bond formation at this
position gives an intermediate bisanthenylbisanthene, which,
when followed by cyclization, gives the dimeric structures that
2
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
during the bisanthenylbisanthene formation in a single step.
Alternatively, the hydrogen may be removed from the corner
by an initial abstraction through a first-generation radical,
followed by radical–radical coupling to another bisanthene
radical. Both hypothetical couplings will give rise to the same
type of dimers. The dimer radicals may be passivated by H
atoms or participate in further coupling steps to give oligo-
and polybisanthenes. Since coupling can occur at every
corner, ill-defined polybisanthene chains are formed. Upon
increasing the temperature, also more indistinct radical
coupling reactions can take place. Nevertheless, a significant
amount of fluoranthene-type connections can be distin-
guished in the polybisanthenes. We conclude that the three-
dimensional geometry of the intermediate radicals deter-
mines the regioselectivity in aryl–aryl bond formation.
Further evidence of the necessity of radical species in the
coupling reactions was obtained from experiments with the
non-halogenated 9,9’-bianthryl, as described in the Support-
ing Information (S1). Since this molecule cannot dehalogen-
ate to form a radical, it is not expected to participate in
coupling reactions. Indeed, no coupling products could be
observed in experiments with bianthryl.
In summary, we investigated the Ullmann coupling of
DCBA, an aryl chloride, on Au(111). In contrast to the
coupling of the analogous bromide, highly branched struc-
tures were formed, which is due to an inversion of dehaloge-
nation and cyclodehydrogenation steps. This changes the
intermediate radicals from staggered bianthryl-type to planar
bisanthene-type. Because of steric effects these undergo an
alternative coupling reaction, giving rise predominantly to
fluoranthene-type connections. We conclude that the geom-
etry of the aryl radical, in combination with surface confine-
Figure 2. Mechanisms for the nanographene formation from dibromo-
bianthryl (left) and dichlorobianthryl (right).
À
ment determines the regioselectivity of C C bond formation
on surfaces. The notion that the reaction sequence in on-
surface synthesis can be tuned by proper choice of halogen,
combined with the importance of the geometry of reaction
intermediates establishes new design criteria for the prepa-
ration of nanographenes and covalent organic frameworks.
Specifically, the strong aryl chloride bond may be used to
allow for intramolecular reactions before intermolecular
couplings in cases where the analogous bromide gives the
reverse order. Proper design of monomers may furthermore
exploit the steric effects to generate the desired products in
future syntheses.
Experimental Section
The compound 10,10’-dichloro-9,9’-bianthryl was synthesized via
Suzuki coupling of 9-anthrylboronic acid[14] with 9-bromoanthracene,
followed by chlorination of the intermediate 9,9’-bianthryl. See the
Supporting Information for details.
The scanning probe experiments were carried out with a Scienta-
Omicron LT STM/AFM with a commercially available qPlus sensor
(quality factor of Q ꢀ 30000, a resonance frequency of f0 = 19500 Hz
and a peak-to-peak oscillation amplitude of approximately 2 ꢀ),
operating at 4.6 K. The molecules were thermally evaporated onto an
Au(111) surface cleaned by several sputter and annealing cycles. All
AFM images were acquired in constant height mode with CO
terminated tips prepared using standard procedures.[15]
Scheme 2. Coupling mechanism for bisanthene radicals. Only hydro-
gens that play a role in the steric effect are shown explicitly. Radical-to-
corner attack may create a bisanthenylbisanthene intermediate, which
then cyclizes into either fused dimeric structure after pivoting to the
left or to the right around this bond (as denoted by the curved
arrows).
were resolved with AFM (Figure 1 f). The present data do
not, however, allow us to resolve the exact mechanism of
coupling. It may be that the hydrogen at the corner is lost
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
Eichhorn, D. Nieckarz, O. Ochs, D. Samanta, M. Schmittel, P. J.
Szabelski, M. Lackinger, ACS Nano 2014, 8, 7880 – 7889; c) J.
Eichhorn, T. Strunskus, A. Rastgoo-Lahrood, D. Samanta, M.
Schmittel, M. Lackinger, Chem. Commun. 2014, 50, 7680 – 7682;
d) J. Bjçrk, F. Hanke, S. Stafstrom, J. Am. Chem. Soc. 2013, 135,
5768 – 5775; e) K. J. Shi, X. Zhang, C. H. Shu, D. Y. Li, X.-Y. Wu,
P. N. Liu, Chem. Commun. 2016, 52, 8726 – 8729.
Acknowledgements
We gratefully acknowledge funding by a NWO Graduate
Program and the Sector Plan Chemistry and Physics.
Keywords: atomic force microscopy · graphene nanoribbons ·
scanning tunnelling microscopy · surface synthesis ·
Ullmann coupling
[6] J. Cai, P. Ruffieux, R. Jaafar, M. Bieri, T. Braun, S. Blankenburg,
M. Muoth, A. P. Seitsonen, M. Saleh, X. Feng et al., Nature 2010,
466, 470 – 473.
[7] a) A. Batra, D. Cvetko, G. Kladnik, O. Adak, C. Cardoso, A.
Ferretti, D. Prezzi, E. Molinari, A. Morgante, L. Venkataraman,
Chem. Sci. 2014, 5, 4419 – 4423; b) L. Massimi, O. Ourdjini, L.
Lafferentz, M. Koch, L. Grill, E. Cavaliere, L. Gavioli, C.
Cardoso, D. Prezzi, E. Molinari et al., J. Phys. Chem. C 2015, 119,
2427 – 2437.
[8] a) C. Bronner, J. Bjçrk, P. Tegeder, J. Phys. Chem. C 2015, 119,
486 – 493; b) S. Blankenburg, J. Cai, P. Ruffieux, R. Jaafar, D.
Passerone, X. Feng, K. Mꢄllen, R. Fasel, C. A. Pignedoli, ACS
Nano 2012, 6, 2020 – 2025; c) J. Bjçrk, S. Stafstrçm, F. Hanke, J.
Am. Chem. Soc. 2011, 133, 14884.
[1] a) Q. Fan, J. M. Gottfried, J. Zhu, Acc. Chem. Res. 2015, 48,
2484 – 2494; b) M. El Garah, J. M. Macleod, F. Rosei, Surf. Sci.
2013, 613, 6 – 14; c) X. Zhang, Q. Zeng, C. Wang, Nanoscale
2013, 5, 8269 – 8287; d) L. Dong, P. N. Liu, N. Lin, Acc. Chem.
Res. 2015, 48, 2765 – 2774; e) G. Franc, A. Gourdon, Phys. Chem.
Chem. Phys. 2011, 13, 14283; f) J. Mꢁndez, M. F. Lꢂpez, J. A.
Martꢃn-Gago, Chem. Soc. Rev. 2011, 40, 4578 – 4590; g) J. Bjçrk,
J. Phys. Condens. Matter 2016, 28, 083002; h) M. Lackinger,
W. M. Heckl, J. Phys. D 2011, 44, 464011.
[2] a) K. J. Shi, D. W. Yuan, C. X. Wang, C. H. Shu, D. Y. Li, Z. L.
Shi, X. Y. Wu, P. N. Liu, Org. Lett. 2016, 18, 1282 – 1285; b) M.
Kittelmann, M. Nimmrich, R. Lindner, A. Gourdon, A. Kꢄhnle,
ACS Nano 2013, 7, 5614 – 5620; c) H. Zhang, J. H. Franke, D.
Zhong, Y. Li, A. Timmer, O. D. Arado, H. Mçnig, H. Wang, L.
Chi, Z. Wang et al., Small 2014, 10, 1361 – 1368; d) M. Kittel-
mann, P. Rahe, M. Nimmrich, C. M. Hauke, A. Gourdon, A.
Kꢄhnle, ACS Nano 2011, 5, 8420 – 8425.
[3] a) C. Zhang, Q. Sun, H. Chen, Q. Tan, W. Xu, Chem. Commun.
2015, 51, 495 – 498; b) L. Grill, M. Dyer, L. Lafferentz, M.
Persson, M. V. Peters, S. Hecht, Nat. Nanotechnol. 2007, 2, 687 –
691; c) Q. Fan, C. Wang, Y. Han, J. Zhu, J. Kuttner, G. Hilt, J. M.
Gottfried, ACS Nano 2014, 8, 709 – 718; d) C. Sꢅnchez-Sꢅnchez,
S. Brꢄller, H. Sachdev, K. Mꢄllen, M. Krieg, H. F. Bettinger, A.
Nicolaꢆ, V. Meunier, L. Talirz, R. Fasel et al., ACS Nano 2015, 9,
9228 – 9235; e) L. Lafferentz, F. Ample, H. B. Yu, S. Hecht, C.
Joachim, L. Grill, Science 2009, 323, 1193 – 1197; f) R. Gutzler, L.
Cardenas, J. Lipton-Duffin, M. El Garah, L. E. Dinca, C. E.
[9] L. Talirz, H. Sçde, J. Cai, P. Ruffieux, S. Blankenburg, R. Jafaar,
R. Berger, X. Feng, K. Mꢄllen, D. Passerone et al., J. Am. Chem.
Soc. 2013, 135, 2060 – 2063.
[10] a) Y.-C. C. Chen, D. G. de Oteyza, Z. Pedramrazi, C. Chen, F. R.
Fischer, M. F. Crommie, ACS Nano 2013, 7, 6123 – 6128; b) R. R.
Cloke, T. Marangoni, G. D. Nguyen, T. Joshi, D. J. Rizzo, C.
Bronner, T. Cao, S. G. Louie, M. F. Crommie, F. R. Fischer, J.
Am. Chem. Soc. 2015, 137, 8 – 11; c) S. Kawai, S. Saito, S. Osumi,
S. Yamaguchi, A. S. Foster, P. Spijker, E. Meyer, Nat. Commun.
2015, 6, 8098; d) A. Kimouche, M. M. Ervasti, R. Drost, S.
Halonen, A. Harju, P. M. Joensuu, J. Sainio, P. Liljeroth, Nat.
Commun. 2015, 6, 10177; e) P. Ruffieux, S. Wang, B. Yang, C.
Sanchez, J. Liu, T. Dienel, L. Talirz, P. Shinde, C. A. Pignedoli, D.
Passerone et al., Nature 2015, 531, 15.
[11] M. Koch, F. Ample, C. Joachim, L. Grill, Nat. Nanotechnol. 2012,
7, 713.
[12] B. Cirera, N. Gimenez-Agullo, J. Bjork, F. Martinez-Pena, A.
Martin-Jimenez, J. Rodriguez-Fernandez, A. M. Pizarro, R.
Otero, J. M. Gallego, P. Ballester et al., Nat. Commun. 2016, 7,
11002.
ˇ
Szakacs, C. Fu, M. Gallagher, M. Vondrꢅcek, M. Rybachuk
[13] T. Kaneko, N. Tajima, T. Ohno, Jpn. J. Appl. Phys. 2016, 55,
06GF05.
[14] a) E. Clar, Ber. Dtsch. Chem. Ges. 1939, 72, 1645 – 1649; b) S. K.
Talapatra, S. Chakrabarti, A. K. Mallik, B. Talapatra, Tetrahe-
dron 1990, 46, 6047 – 6052.
et al., Nanoscale 2014, 6, 2660 – 2668.
[4] a) J. A. Lipton-Duffin, O. Ivasenko, D. F. Perepichka, F. Rosei,
Small 2009, 5, 592 – 597; b) M. Bieri, M.-T. Nguyen, O. Grçning,
J. Cai, M. Treier, K. Aꢆt-Mansour, P. Ruffieux, C. A. Pignedoli,
D. Passerone, M. Kastler et al., J. Am. Chem. Soc. 2010, 132,
16669 – 16676; c) G. Eder, E. F. Smith, I. Cebula, W. M. Heckl,
P. H. Beton, M. Lackinger, ACS Nano 2013, 7, 3014 – 3021; d) S.
Schlçgl, W. M. Heckl, M. Lackinger, Surf. Sci. 2012, 606, 999 –
1004.
[15] L. Gross, F. Mohn, N. Moll, P. Liljeroth, G. Meyer, Science 2009,
325, 1110 – 1114.
[5] a) L. Lafferentz, V. Eberhardt, C. Dri, C. Africh, G. Comelli, F.
Esch, S. Hecht, L. Grill, Nat. Chem. 2012, 4, 215 – 220; b) J.
Received: July 3, 2016
Published online: && &&, &&&&
4
www.angewandte.org
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Nanographene
Radical geometry: The geometric struc-
ture of the aryl radical involved in the on-
surface synthesis of graphene nanorib-
bons determines the geometric structure
of the ribbon. Using atomically resolved
non-contact atomic force microscopy the
structure of various coupling products
and intermediates were resolved.
P. H. Jacobse, A. van den Hoogenband,
M.-E. Moret,* R. J. M. Klein Gebbink,
I. Swart*
&&&& — &&&&
Aryl Radical Geometry Determines
Nanographene Formation on Au(111)
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!