Synthesis of Graphene Nanoribbons by Ambient-Pressure Chemical Vapor Deposition and Device Integration

Article abstract of DOI:10.1021/jacs.6b10374

Graphene nanoribbons (GNRs), quasi-one-dimensional graphene strips, have shown great potential for nanoscale electronics, optoelectronics, and photonics. Atomically precise GNRs can be bottom-up synthesized by surface-assisted assembly of molecular building blocks under ultra-high-vacuum conditions. However, large-scale and efficient synthesis of such GNRs at low cost remains a significant challenge. Here we report an efficient bottom-up chemical vapor deposition (CVD) process for inexpensive and high-throughput growth of structurally defined GNRs with varying structures under ambient-pressure conditions. The high quality of our CVD-grown GNRs is validated by a combination of different spectroscopic and microscopic characterizations. Facile, large-area transfer of GNRs onto insulating substrates and subsequent device fabrication demonstrate their promising potential as semiconducting materials, exhibiting high current on/off ratios up to 6000 in field-effect transistor devices. This value is 3 orders of magnitude higher than values reported so far for other thin-film transistors of structurally defined GNRs. Notably, on-surface mass spectrometry analyses of polymer precursors provide unprecedented evidence for the chemical structures of the resulting GNRs, especially the heteroatom doping and heterojunctions. These results pave the way toward the scalable and controllable growth of GNRs for future applications.

Full text of DOI:10.1021/jacs.6b10374

Subscriber access provided by University of Rochester | River Campus & Miner Libraries
Article
Synthesis of graphene nanoribbons by ambient-pressure
chemical vapor deposition and device integration
Zongping Chen, Wen Zhang, Carlos-Andres Palma, Alberto Lodi Rizzini, Bilu Liu, Ahmad
N. Abbas, Nils Richter, Leonardo Martini, Xiao-Ye Wang, Nicola Cavani, Hao Lu, Neeraj
Mishra, Camilla Coletti, Reinhard Berger, Florian Klappenberger, Mathias Kläui, Andrea
Candini, Marco Affronte, Chongwu Zhou, Valentina De Renzi, Umberto del Pennino,
Johannes V. Barth, Hans Joachim Räder, Akimitsu Narita, Xinliang Feng, and Klaus Müllen
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b10374 • Publication Date (Web): 04 Nov 2016
Downloaded from http://pubs.acs.org on November 5, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 24
1
2
3
4
5
6
7
8
9
Synthesis of graphene nanoribbons by ambient-
pressure chemical vapor deposition and device
integration
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Zongping Chen1, Wen Zhang1, Carlos-Andres Palma3, Alberto Lodi Rizzini4,5, Bilu
Liu6, Ahmad Abbas6,7, Nils Richter8,9, Leonardo Martini4,5, Xiao-Ye Wang1, Nicola
Cavani4,5, Hao Lu1, Neeraj Mishra10, Camilla Coletti10, Reinhard Berger2, Florian
Klappenberger3, Mathias Kläui8,9, Andrea Candini5, Marco Affronte4,5, Chongwu
Zhou6, Valentina De Renzi4,5, Umberto del Pennino4,5, Johannes V. Barth3, Hans
Joachim Räder1, Akimitsu Narita1, Xinliang Feng2* and Klaus Müllen1*
1Max Planck Institute for Polymer Research, Ackermannweg 10, D-55128 Mainz,
Germany
2Center for Advancing Electronics Dresden (cfaed) & Department of Chemistry and
Food Chemistry, Technische Universität Dresden, Mommsenstrasse 4, 01062 Dresden,
Germany
3Physik-Department, Technische Universität München, James-Franck-Str. 1, D-85748
Garching, Germany
4Dipartimento di Scienze Fisiche, Informatiche e Matematiche, Università di Modena
e Reggio Emilia, I-41125 Modena, Italy
5CNR-NANO, Istituto Nanoscienze, Centro S3, I-41125 Modena, Italy
6Department of Electrical Engineering and Department of Chemistry, University of
Southern California, Los Angeles, California, 90089, United States
7Department of Electrical Engineering, King Abdulaziz University, Abdullah
Sulayman Street, Jeddah 22254, Saudi Arabia
8Johannes Gutenberg Universität-Mainz, Institut für Physik, Staudingerweg 7, 55128
Mainz, Germany
9Graduate School of Excellence Materials Science in Mainz, Staudingerweg 9, 55128
Mainz, Germany
10Center for Nanotechnology Innovation @ NEST, Istituto Italiano di Tecnologia,
Piazza San Silvestro 12, 56127 Pisa, Italy
*e-mail: xinliang.feng@tu-dresden.de; muellen@mpip-mainz.mpg.de
1
ACS Paragon Plus Environment

Page 3 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
ABSTRACT: Graphene nanoribbons (GNRs), quasi-one-dimensional graphene
strips, have shown great potential for nano-scale electronics, optoelectronics and
photonics. Atomically precise GNRs can be “bottom-up” synthesized by surface-
assisted assembly of molecular building blocks under ultrahigh vacuum
conditions. However, a large-scale and efficient synthesis of such GNRs at low
cost remains a significant challenge. Here we report an efficient “bottom-up”
chemical vapor deposition (CVD) process for inexpensive and high-throughput
growth of structurally defined GNRs with varying structures under ambient-
pressure conditions. The high quality of our CVD-grown GNRs is validated by a
combination of different spectroscopic and microscopic characterizations. Facile,
large-area transfer of GNRs onto insulating substrates and subsequent device
fabrication demonstrate their promising potential as semiconducting material,
exhibiting high current on/off ratio up to 6,000 in field-effect transistor devices.
This value is three orders of magnitude higher than values reported so far for
other thin film transistors of structurally defined GNRs. Notably, on-surface
mass-spectrometry analyses of polymer precursors provide unprecedented
evidences for the chemical structures of the resulting GNRs, especially the
heteroatom-doping and heterojunctions. These results pave the way toward the
scalable and controllable growth of GNRs for future applications.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
INTRODUCTION
Graphene is a two-dimensional semi-metallic crystal with zero bandgap,1-3 which
hinders its use in many electronic and optoelectronic devices, where a suitable
bandgap is essential.4-5 It has been predicted by theoretical studies that narrow (<10
nm) graphene nanoribbons (GNRs) with armchair-type edges exhibit semiconducting
behavior, due to the intense quantum confinement and edge effects.6-8 Whereas the
predominant “top-down” approaches, such as lithographical patterning of graphene9-10
and unzipping of carbon nanotubes11-12 can hardly control the width and edge
structure of GNRs, surface-assisted13-15 and solution-mediated16-20 “bottom-up”
methods have been developed to synthesize atomically precise GNRs. Such “bottom-
up” synthesized ultranarrow (~1-2 nm) GNRs have demonstrated large bandgaps of
~1 to 2 eV with visible to near-infrared absorption,16-23 rendering them highly
2
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 24
1
2
3
4
5
6
7
8
9
interesting for a broad range of applications in next-generation transistors, as well as
optoelectronic and photonic devices.24-26 However, the device applications of the
“bottom-up” synthesized GNRs are still elusive, because on one hand, processing of
solution-synthesized GNRs is a major obstacle due to their limited solubility and
tendency to aggregate in dispersions.17, 19-20,27 On the other hand, the amount of GNRs
accessible by surface-assisted method is limited, and the elaborate and costly
ultrahigh vacuum (UHV) equipment restricts the large-scale fabrication and further
use of the GNRs. Sakaguchi and Nakae et al. have previously proposed a use of CVD
method for the on-surface synthesis of GNRs, but low pressure was required for the
process and unambiguous elucidation of the quality and precise structures of the
resulting GNRs has remained elusive.28
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The precise structures of the GNRs “bottom-up” synthesized under UHV have been
demonstrated by high resolution scanning probe microscopy (SPM) together with
various spectroscopic characterizations, including Raman spectroscopy and X-ray
photoelectron spectroscopy (XPS).13-14, 29-32 Nevertheless, there is still often a lack of
direct evidence for the precise chemical structures of the GNRs. For example, C and
N atoms cannot be directly distinguished in GNRs because of their nearly identical
topographic morphology.14, 31 To date, there is thus no direct evidence for the precise
chemical composition of heteroatom-doped GNRs and their heterojunctions.14, 30-31, 33
Here, we demonstrate an efficient ambient-pressure CVD process for inexpensive and
high-throughput growth of structurally defined GNRs over large areas, employing
hydrogen gas as oxygen scavenger. The GNRs by this CVD method exhibit structural
quality and properties comparable with those of the same GNRs synthesized under
UHV conditions, as supported by Raman spectroscopy, XPS, high-resolution electron
energy loss spectroscopy (HREELS), and scanning tunneling microscopy (STM)
characterizations. Homogenous GNR films over areas of square centimeters can be
obtained and successfully transferred to non-conducting wafers, allowing for the
facile device integration. Thus fabricated field-effect transistor (FET) devices exhibit
a large current on/off ratio of up to 6,000, which is significantly higher than values
reported so far for thin film transistors made with structurally defined GNRs.27-28, 34
Moreover, by using graphene electrodes, the contact resistance can be drastically
reduced while preserving the high current on/off ratio. This “bottom-up” CVD
method further allows the growth of N-doped and N, S-co-doped GNRs as well as
3
ACS Paragon Plus Environment

Page 5 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
their heterojunctions, demonstrating the versatility and scalability of this process,
which provides access to a broad class of GNRs with engineered structures and
properties based on molecular-scale design. Importantly, high-resolution mass
spectrometry analyses of their early-stage polyphenylene precursors provided strong
evidences for the chemical composition at the termini as well as the doping and
heterostructures of the CVD-grown GNRs.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION
CVD growth of GNRs
For the growth of chevron-type GNRs, the halogen-functionalized monomer 6,11-
dibromo-1,2,3,4-tetraphenyltriphenylene (DBTT) was used as molecular precursor
(Figs. 1 and S1). In a CVD setup, the DBTT molecules were sublimed by a heating
belt and deposited on an Au/mica surface that was kept in the central zone of a
horizontal tube furnace at a temperature of 250 °C (Figs. 1a and S2). The halogen
groups of the monomers were removed by thermal activation on the Au surface to
form biradical species, which subsequently underwent radical coupling reactions to
form covalent C–C bonds between each monomer, yielding a chevron-type
polyphenylene precursor. Finally, the polyphenylene precursor was annealed to a
higher temperature of 450 °C to establish an extended fully aromatic system, where
the surface-assisted intramolecular cyclodehydrogenation afforded the structurally
defined chevron-type GNRs (Fig. 1a). In contrast to the UHV13 and the three-zone
low-pressure CVD28 processes reported before, a two-zone system and ambient
pressure were used in our synthesis, for the first time demonstrating that the GNRs
can be fabricated efficiently even under such readily available conditions. The results
also proved that the surface-assisted polymerization and cyclodehydrogenation
reactions for the GNR synthesis are pressure-independent, from UHV to ambient
pressure. The CVD growth was carried out under a mixture of Ar and H2, which is
essential to suppress the oxidation of the resulting GNRs as revealed by HREELS and
XPS (vide infra). This CVD method can be further applied to different monomers,
enabling ambient pressure synthesis of heteroatom-doped GNRs as well as GNR
heterojunctions (vide infra).
4
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 24
1
2
3
4
5
6
Mass spectrometry characterization of the intermediate polyphenylene
precursors and GNRs
7
8
9
The polymerization reaction was first studied by high-resolution matrix-assisted laser
desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), which is a
powerful method to assess the atomic composition of complex organic compounds.35
For MALDI-TOF MS measurements, the {(2E)-2-methyl-3-[4-(2-methyl-2-propanyl)
phenyl]-2-propen-1-ylidene}malononitrile (DCTB) matrix was sublimed on the as-
deposited polyphenylene precursors on the Au/mica substrate. The MS spectrum
exhibits regular patterns of polyphenylene precursors from dimer (m/z = ~1,062 Da)
to pentadecamer (m/z = ~7,966 Da), in agreement with the expected mass values,
which confirms the successful polymerization reaction (Fig. 1e). The observed
isotopic distribution is in perfect agreement with the simulated pattern, proving the
precise chemical composition of the polyphenylene precursors (Fig. 1e, inset). The
intensities of these MS peaks decrease for longer polyphenylene precursors, and
precursors larger than fifteen repeating units, i.e., ~12 nm in length were not observed.
We have also successfully investigated the polymerization reaction in CVD growth of
other GNRs, such as N = 7 armchair GNRs (7-AGNRs), using the high-resolution MS
analysis method (Fig. S3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Understanding the nature of the GNR termini is important to reveal the factors
limiting the length of GNRs.29 There are mainly three plausible GNR termini
considering the steps of the CVD growth: radical, Br and H passivation. By
combining high-resolution SPM experiments with large-scale density functional
theory simulations, Talirz et al. have demonstrated that the termini of GNRs grown
under UHV conditions are passivated by hydrogen.29 It was suggested that the H
atoms come from premature cyclodehydrogenation in the polymerization step, which
can directly compete with the polymerization reaction by forming a C−H bond at the
radical termini. However, even with those atomically resolved STM analyses, the
presence of the hydrogen atoms cannot be directly confirmed. Besides, no SPM
investigation has been performed directly on the polyphenylene precursor to reveal its
terminal structure in the polymerization reaction. Here, for the first time, we obtained
stoichiometric information on the polyphenylene termini by the MALDI-TOF MS
analysis as shown in Figs. 1e and S4. No Br terminated polyphenylene precursors can
be detected in the samples polymerized at a temperature higher than 250 °C. Besides,
5
ACS Paragon Plus Environment

Page 7 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
the exact isotopic distribution of each polyphenylene does not match the radical
termini but is in agreement with a H-passivated structure (Fig. S4). These results
indeed agree well with the SPM study on GNR termini.29 However, no signal that can
be attributed to premature cyclodehydrogenation of the precursor can be detected. The
amount of premature cyclodehydrogenation might be too limited to be detected by
MALDI-TOF MS even if it exists. The reported SPM analysis revealed that 15% of
the GNR termini are methylene groups (H2 passivation).29 Nevertheless, MS analysis
does not support such a termination for the polyphenylene precursors.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Whereas the polyphenylene precursors could be well characterized, the MALDI-TOF
MS analysis of the resulting chevron-type GNRs on Au/mica substrates shows no
meaningful signals. We suppose that the planar GNRs strongly interacting with the
gold surface could not be effectively desorbed by MALDI.17, 36 When the GNRs were
transferred onto a silicon wafer, some carbon fragments could be detected (Fig. S5).
Interestingly, all the observed fragments are even-numbered carbon clusters formed
through destructive ablation of the GNRs by the laser irradiation in the MS system,
which is remarkably similar to the LDI MS spectrum observed for graphene.37
UHV characterizations of GNRs
Although the investigation of ambient-pressure samples under UHV conditions is
generally challenging due to the atmospheric contaminants, UHV STM, XPS, and
HREELS are key characterization methods for GNRs. Topographical STM surveys of
the sample after the first polymerization step show chevron-type polyphenylene
precursors (Fig. 1b). High-resolution data reveals zigzag shaped polymer chains in
good agreement with that of previous UHV-STM studies,13 where the tetraphenyl
substitutions marked in green are tilted with respect to the surface, resulting in bright
features in the STM images (Fig. 1c, d). After the final cyclodehydrogenation step, a
planar chevron-type structure of the resulting GNRs is identified in both ambient and
UHV STM investigations (Fig. 2a-c). Higher resolution UHV image reveals
additional imperfections at the submolecular level. The white and cyan arrows in Fig.
2b indicate common defects due to broken phenyl rings and randomly coupled GNRs,
respectively. To be noted is that even the "atomically precise" GNR made under UHV
are not always 100% atomically precise, in terms of the curving and branching.13
Recently, the coupling of several GNRs into complex junction structures has also
6
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 24
1
2
3
4
5
6
7
8
9
been found in the UHV-grown GNRs and revealed by CO-functionalized atomic force
microscopy in atomic resolution.38-39 These coupling junctions between GNRs may be
important tool in the development of graphene-based circuitry and facilitate the
electronic device investigation on GNRs.38 Large scale survey was difficult to obtain
due to adsorbed impurities from ambient pressure condition (bright white protrusions
in Fig. 1b-d and Fig. 2a-c), hindering quantitative statistical analysis of the GNR
quality. Nevertheless, we would like to highlight that we can synthesize GNRs much
faster and facilely under the CVD conditions with a significantly simpler setup, which
can be easily scaled up for applications. We suggest that further preparation of the
samples in clean room conditions and vacuum packaging may improve the quality of
GNRs and also facilitate the UHV STM characterization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Instead, the quality of the CVD-grown chevron-type GNRs was further examined by
comparing the XPS and HREELS spectra with those of a reference GNR sample,
which was grown in-situ under UHV conditions on an Au(111) single crystal13 (Figs.
2e and f). The overlayer coverage, as estimated from the intensity of the C1s peak
relative to the Au 4f peak in the XPS spectrum, is very similar for both samples,
corresponding to 0.5 ML for the UHV-grown GNRs and 0.6 ML for the CVD-grown
GNRs (1 ML is defined as a single GNR layer with surface number density of
2.7×1015 at./cm2, see the Supplementary Information for details). The XPS survey
spectra exhibit only the core level peaks belonging to the GNRs and gold substrate.
The C1s core level spectrum of the CVD-grown GNRs consists of a single sharp
component located at ~284 eV binding energy, corresponding to the sp2 bonded
carbon, which well resembles that of the UHV-grown GNRs (Fig. 2e). No Br
components can be detected, indicating that Br atoms are completely desorbed from
the gold surface. HREELS can precisely characterize the vibrational properties of the
overlayer, serving as a powerful characterization method for GNR thin films
especially when IR absorption is too weak for a reasonable analysis.30, 40 The
HREELS spectra of the two samples are almost identical: both are indeed dominated
by the out-of-plane CH mode manifold, located in the 700-850 cm-1 region, while the
intensity of the in-plane CH stretching mode at 3,063 cm-1 is almost negligible (Fig.
2f). The manifold is built up by four main peaks locates at 754, 782, 806, and 862 cm-
1, which are assignable to out-of-plane modes (C-H), characteristic of different
molecular edge topologies (i.e., TRIO, DUO and SOLO structures in Refs. 41-42). It is
7
ACS Paragon Plus Environment

Page 9 of 24
Journal of the American Chemical Society
1
2
important to notice that in both spectra the feature at ~700 cm-1, which is attributed to
the out-of-plane (C-H) mode of the mono-substituted benzene rings and is therefore
present only for the monomer and polyphenylene precursors, is indeed barely
visible.30 This finding, together with the dominant intensity of out-of-plane modes
relative to in-plane ones, proves that the reaction from the polyphenylene precursors
to flat GNRs is completed.30 The quality of the CVD-grown GNRs is also
documented by the huge intensity of the elastic peak, and the narrow width of the
vibrational modes, which are comparable with those of the UHV-grown GNRs. The
slight differences seen in the spectra might be caused by the atmospheric
contaminants on the Au(111) surface under the ambient-pressure conditions. Besides,
compared with UHV conditions, we have found that hydrogen is important for CVD
growth of high quality GNRs. Normally, the oxygen content in CVD is much larger
than the UHV system and is inevitable, leading to the considerable oxidation of the
resulting GNRs, as evidenced by the HREELS analysis: The HREELS spectra of the
CVD-GNRs grown without H2 displays two extra broad features located at 440 and
1,100 cm-1, which may attribute to C-O-C deformation and stretching modes (Fig. 2f).
The oxidation of the GNRs grown without H2 was further corroborated by XPS where
much higher level of oxygen content was detected (see Fig. S6 and the Supplementary
Information for details). High quality of the CVD-grown GNRs could be achieved
only under the presence of H2 in the growth, which can suppress the oxidation of the
GNRs during the high temperature annealing (Fig. 2f). Therefore, combination of the
high-resolution STM, HREELS, and XPS analyses proved that structurally-defined
GNRs can be synthesized by our CVD method under the presence of H2, with high
quality comparable to that of the UHV-grown GNRs.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Transfer of the GNRs and their Raman and UV-Vis-NIR spectroscopy
characterizations
The Raman spectra of the CVD-grown chevron-type GNRs excited by a 532 nm laser
is presented in Fig. 2d and Fig. S7, where the most intense peaks are at ~1,330 and
1,608 cm-1, typically assigned to D and G bands, respectively.43 Remarkably, sharp
peaks of low-frequency modes can be observed at ~180, ~284, and ~403 cm-1, which
may be related to the specific GNR structure, like the width-specific radial breathing-
like mode (RBLM),13, 19 indicating high uniformity of the chevron-type GNRs. Raman
spectrum of the CVD-grown 7-AGNRs demonstrates perfect agreement with that of
8
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 24
1
2
the same GNR prepared under the UHV condition, including the RBLM at ~400 cm–1
and other smaller features, which further corroborates the highly defined structures of
the CVD-grown GNRs (Fig. S3).
3
4
5
6
7
8
9
Moreover, the quantity of GNRs and the size of the resulting films can be easily
scaled up by using large Au/mica substrates, e.g., with an area of 25×75 mm2 as
shown in Fig. 3a (essentially limited only by the furnace and tube dimensions), which
is much larger than the reported small piece of GNRs by CVD (limited to a size of
~10×10 mm2) using the similar size of quartz tube.28 After the CVD growth, the
GNRs can be transferred onto other substrates with or without the assistance of
poly(methyl methacrylate) (PMMA) (Scheme S2 and S3). The transferred GNR films
possess a high uniformity over an area of square centimeters without any cracks based
on the optical images and Raman mapping (Figs. 3b and S7-9). Atomic force
microscopy of the GNRs transferred on silicon wafer reveals a thickness of ~0.7 nm
(Fig. 3d). The chevron-type GNR films have also been transferred layer by layer on
fused silica for UV-Vis-NIR absorption measurements (Fig. S10). The spectrum
exhibits absorption in the UV and visible region with an absorption maximum at ~450
nm. The optical bandgap of the GNR film was roughly estimated from the absorption
onset at ~700 nm to be approximately 1.8 eV,19-20 suggesting their potential for
electronic device applications.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Electronic properties of GNR-based three-terminal FET devices
Based on the successful transfer of chevron-type GNR films on any desired substrates,
we have fabricated three-terminal FET devices using GNRs as the channel material,
to study their charge transport properties. The first set of devices was fabricated from
chevron-type GNR films transferred on a silicon wafer, where Pd/Ti (50 nm/1 nm)
electrodes were patterned as the contact metals using electron-beam lithography (EBL)
(Fig. 3c). The channel length was ~1-2 µm (see Supplementary Information for more
details). As shown in Figs. 3e-f and S11a, a large gate-modulation to the source-drain
current can be observed (current on/off ratio >1,000). The GNR transistors exhibit a
unipolar p-type conduction, as the source-drain current increases with increasing
negative gate voltages. The p-type conduction might be caused by doping via adatoms
(e.g. oxygen from air), since the device measurements were carried out in ambient
condition.44 Best performing devices exhibit current on/off ratio of >6,000, which is
9
ACS Paragon Plus Environment

Page 11 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
significantly higher (more than three orders of magnitude) than the values reported so
far for other “bottom-up” synthesized GNR thin film network transistors (with typical
on/off ratio of ~2-5),27-28, 34 and even better than the devices on UHV-GNRs with an
ultranarrow channel.44 Fig. 3f shows Ids-Vds curves of a typical GNR transistor. The
non-linear nature of the curves at small bias regimes suggests Schottky-contact
behavior. This observation is understandable because our GNRs are very narrow and
possess a large bandgap. Since the GNR devices reveal p-type conduction in air, the
use of high work function contact materials, such as MoO3-x,45 may offer better
contacts to the valence band of GNRs and deliver better device performance.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Alternatively, reducing the channel length may also improve the device performance,
as the additional resistance between adjacent ribbons is reduced or can be completely
avoided. This has been tested in an additional set of devices where the Pd electrode
was in direct contact with the GNR films and the spacing between the electrodes was
smaller than 50 nm (Fig. 4a). Because of the drastically reduced active channel length,
the same level of current flow can be achieved at one order of magnitude lower
source-drain voltages. However, in that case the influence of the gate voltage is
drastically reduced, which may be attributed to the screening effect of the metal
electrodes on the electric field from the gate. Therefore, either a thinner insulating
barrier layer or a better source-drain electrode material would be needed to retain the
FET-functionality in the short-channel devices. For this purpose, and by taking
advantage of the facile transfer process of our GNRs, we realized another type of
three-terminal FET device by direct transfer of the chevron-type GNRs on pre-
fabricated graphene electrodes, i.e. a graphene device where the channel has been
opened with a gap of approximately 100 nm by EBL and oxygen plasma etching (see
Fig. 4b and the Supplementary Information for details). Relatively thick graphene
(average 10 layers) on C-face of SiC is chosen as the electrode material, which
displays almost no gate dependence and is expected to show limited screening of the
gate potential compared to metal electrodes. The electrical characteristics of the
resulting devices are presented in Fig. 4c-d (see Fig. S11-13 and the Supplementary
Information for more details). The contact resistance is greatly reduced with respect to
the long-channel metal electrodes, due to the shorter channel distance and a superior
GNR/graphene interface. An evident transistor effect (current vs side-gate voltage)
can be observed with a current on/off ratio of >100, limited only by the maximum
10
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 24
1
2
3
4
5
6
7
8
9
applicable gate voltage. Improvement of the channel-gate coupling, for instance by
employing a top gate with a thin high-k dielectric, is expected to further enhance the
device performances. Interestingly, contrary to the p-type conduction observed in the
long-channel devices, the devices based on graphene electrodes show an n-type
conduction, which was previously found also for the UHV-GNR device.44 The Ids-Vds
curves in Fig. 4d remain non-linear, indicating the presence of a Schottky-type contact
also in these devices. We suggest that the performance of these GNR devices can be
further improved by using hBN as its support. Moreover, development of wider GNRs
with smaller bandgaps and growth of longer GNRs through optimization of the CVD
process may further reduce the Schottky barriers and contact resistance for a better
electrical connection to electrodes.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Heteroatom-doped GNRs and GNR heterojunctions
The “bottom-up” CVD procedure can be a general strategy for fabricating a broad
class of GNRs and their hybrids with engineered structures and properties based on
the selection of suitable monomers. As an example, we have demonstrated the CVD
growth of heteroatom-doped chevron-type GNRs, which is a promising way to
modify the electronic structures of GNRs.14, 30 By using the nitrogen and sulphur
containing
monomers
4,4'-(6,11-dibromo-1,4-diphenyltriphenylene-2,3-
diyl)dipyridine (DBNN) and 4-(6,11-dibromo-1,4-diphenyl-3-(thien-3-yl)triphenylen-
2-yl)pyridine (DBNS, Fig. 5 and Scheme S1) as precursors, N-doped and N, S-co-
doped chevron-type GNRs can be grown in the CVD system, respectively (Scheme
S4). High-resolution MALDI-TOF MS analysis clearly validates the polymerization
of the corresponding dehalogenated biradical intermediates, forming N-substituted
and N, S-co-substituted polyphenylene precursors (Figs. 5b, e and S14-15). Contrary
to symmetric monomer DBNN, monomer DBNS has an asymmetric structure with
both pyridyl and thienyl groups. Therefore, it must be noted that the resulting N, S-co-
substituted polyphenylene precursors always contain regioisomeric segments,
depending on the head-to-head or head-to-tail coupling, which can only be
differentiated by atomic-resolution SPM. The successful cyclodehydrogenation of
these precursors into the N-doped and N, S-co-doped GNRs is corroborated by
Raman spectroscopy (Fig. S16). The observation of intense Raman spectra with the
sharp G peak located at ~1,608 cm-1 and D peak located at ~1,330 cm-1 validated the
successful “graphitization” of the polymer precursors into GNR structures. We have
11
ACS Paragon Plus Environment

Page 13 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
also performed XPS and HREELS analyses of the N- and N, S-co-doped GNRs
grown by our CVD method (see Figs. S17-18 and the Supplementary Information for
details).
9
We further fabricated GNR heterojunctions in the CVD system by co-sublimation of
different kinds of structurally complementary monomers. Figure 5a illustrates the
strategy for fabricating a p-N-GNR heterojunctions with seamlessly connected
segments of pristine (p) and N-doped (N) GNRs.14 The monomer DBTT and its N-
substituted equivalent DBNN were co-deposited on an Au/mica substrate kept at
200 °C, resulting in polyphenylene heterojunctions, which could be directly identified
by the MS analysis (Figs. 5a, c, S19 and Scheme S4). The deposited radical
intermediates of the two monomers distribute randomly and uniformly on the gold
surface.15 Therefore, the building blocks in the hybrid polyphenylene precursors are
mixed and fused in a random sequence during the polymerization, exhibiting a
statistical distribution from pure segments of each monomer unit to their
combinations with different ratio. For example, the isotopic distributions of the
resulted trimeric polyphenylene precursors exhibit four kinds of trimer structures,
with the ratio of the two monomer units varying from 3/0, 2/1, 1/2 to 0/3, where the
middle two structures are corresponding to the polyphenylene heterojunctions (Fig.
5c). Moreover, the CVD method allows for the synthesis of polyphenylene
heterojunctions with various chemical compositions simply by using different
monomers. For example, heterojunctions of N-substituted and N, S-co-substituted
polyphenylenes can be fabricated by combining the N- and N, S-co-substituted
monomers (DBNN and DBNS), which have been unambiguously identified by high-
resolution MS analysis (Figs. 5f and S20). These results mark the high versatility of
the CVD process and the great effectiveness of this direct MS analysis method for
investigating the GNR growth process. All these polyphenylene heterojunctions can
be converted to GNR heterojunctions similarly by cyclodehydrogenation upon further
thermal annealing at 400-420 °C, as verified by Raman spectroscopic analyses (Fig.
S16).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 24
1
2
3
4
CONCLUSION
5
6
7
8
9
In summary, the low-cost CVD method has enabled the large-scale “bottom-up”
growth of structurally-defined pristine GNRs and selectively N-doped and N, S-co-
doped GNRs as well as their heterojunctions at the ambient pressure, which can be an
excellent complement to the more elaborate UHV method. Raman, HREELS, XPS,
and STM analyses manifested the high quality of the CVD-grown GNRs well
comparable to those grown under UHV conditions. The surface-mass-spectrometry
analysis can serve as a quick and powerful method to investigate the precise chemical
structures of various GNR precursors as well as other surface-bound molecular
materials. Further studies might also provide an insight into the reaction mechanisms
of the surface-based polymerization and cyclodehydrogenation. Large-area GNR
films can be readily transferred on to arbitrary substrates, where preliminary micro- as
well as nano-scale transistor devices could be fabricated, demonstrating
unprecedentedly strong gate modulation for the “bottom-up” GNRs. Furthermore, our
large-scale synthesis process can be readily applied to other designed monomers for
preparing GNRs with engineered topographic and chemical structures, which is
expected to provide practical solutions to the current challenges in graphene-based
nanoelectronic, optoelectronic and photonic devices.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
ACS Paragon Plus Environment

Page 15 of 24
Journal of the American Chemical Society
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
9
CVD growth of GNRs. Home-made and commercially available (Phasis, Switzerland
and Georg Albert, Germany) 200 nm Au thin films epitaxially grown on mica were
used as substrates for GNR growth. They were cut into pieces with sizes from 10×10
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
to 25×75 mm2 and placed in a quartz tube of outer diameter 35 mm and inner
diameter 33 mm in a horizontal tube furnace (Nabertherm, RT 80-250/11S). Precursor
monomers (for details on the synthesis, see the Supplementary Information) were
loaded upstream of the Au/mica substrates and sublimed at 250-325 °C onto the
Au/mica surface using a heating belt (Thermocoax Isopad S20). The Au/mica
substrates were heated and maintained at 200-250 °C under gas flow of Ar (500
s.c.c.m.) and H2 (100 s.c.c.m.) during monomer sublimation for 5-30 min to induce
dehalogenation and radical addition reaction. After deposition, the samples were post-
annealed at 400-450 °C for 15 min to convert the polyphenylene precursors into
GNRs by cyclodehydrogenation. For fabrication of the heterojunctions, the two
precursor species were evaporated together onto an Au/mica substrate held at 200 °C,
which was subsequently annealed to 400-420 °C. To study the oxidation of the GNRs,
only 500 s.c.c.m. of Ar was used in the growth without H2, the pressure was ~0.7
mbar by a mechanical rotary pump (capable of low pressure down to <10-4 mbar).
Transfer of GNRs. As-grown GNRs/Au/mica substrates were placed on a surface of
hydrofluoric acid (HF) (40 wt. %) solution, which etched away the mica and made the
Au film with GNRs float on the surface of the HF solution. After washing with
ultrapure water, the GNRs/Au films were transferred to a new insulated substrate and
dried at 100 °C for 10 minutes. Finally, the Au films were etched away by gold
etchant (Sigma-Aldrich), leaving only the GNR films on the new substrate. When
transferring GNR films with large areas, spin-coated PMMA thin films were used as a
mechanical support before etching with HF to avoid breaking the GNR films. The
PMMA was washed away by hot acetone in the last step (Scheme S2 and S3).
14
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 24
1
2
3
4
ASSOCIATED CONTENT
5
6
7
Supporting Information
8
9
Experimental details, MALDI-TOF MS, Raman spectra, UV−vis spectra, more
analysis on XPS and HREELS spectra of the GNRs and their FET devices. The
Supporting Information is available free of charge on the ACS Publications website.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Authors
* xinliang.feng@tu-dresden.de
* muellen@mpip-mainz.mpg.de
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was supported by the European Research Council grant on NANOGRAPH
and MolArt, DFG Priority Program SPP 1459, MIUR FIR grant RBFR13YKWX and
project PRIN-GRAF 20105ZZTSE_008, Graphene Flagship (No. CNECT-ICT-
604391), European Union Projects UPGRADE and MoQuaS (contract N.610449),
and the Office of Naval Research BRC Program. The authors thank Dr. Roberto Biagi
and Valdis Corradini (CNR-NANO, Istituto Nanoscienze) for useful discussions, and
Dr. Dieter Schollmeyer (Institute for Organic Chemistry, Johannes Gutenberg
University Mainz) for single-crystal X-ray structural analysis of the monomer DBNS.
X.Y.W. is grateful for the fellowship from Alexander von Humboldt Foundation.
15
ACS Paragon Plus Environment

Page 17 of 24
Journal of the American Chemical Society
1
2
3
4
REFERENCES
5
6
7
8
9
(1)
Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.;
Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Science 2004, 306, 666-669.
(2)
(3)
(4)
Geim, A. K. Science 2009, 324, 1530-1534.
Geim, A. K.; Novoselov, K. S. Nature Mater. 2007, 6, 183-191.
Kotov, V. N.; Uchoa, B.; Pereira, V. M.; Guinea, F.; Castro Neto, A. H. Rev.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Mod. Phys. 2012, 84, 1067-1125.
(5) Castro Neto, A. H.; Guinea, F.; Peres, N. M. R.; Novoselov, K. S.; Geim, A.
K. Rev. Mod. Phys. 2009, 81, 109-162.
(6) Yang, L.; Park, C. H.; Son, Y. W.; Cohen, M. L.; Louie, S. G. Phys. Rev. Lett.
2007, 99, 186801.
(7)
(8)
Barone, V.; Hod, O.; Scuseria, G. E. Nano Lett. 2006, 6, 2748-2754.
Palma, C.-A.; Awasthi, M.; Hernandez, Y.; Feng, X.; Müllen, K.; Niehaus, T.
A.; Barth, J. V. J. Phys. Chem. Lett. 2015, 6, 3228-3235.
(9) Chen, Z.; Lin, Y.-M.; Rooks, M. J.; Avouris, P. Physica E 2007, 40, 228-232.
(10) Han, M. Y.; Ozyilmaz, B.; Zhang, Y.; Kim, P. Phys. Rev. Lett. 2007, 98,
206805.
(11) Jiao, L. Y.; Zhang, L.; Wang, X. R.; Diankov, G.; Dai, H. J. Nature 2009, 458,
877-880.
(12) Kosynkin, D. V.; Higginbotham, A. L.; Sinitskii, A.; Lomeda, J. R.; Dimiev,
A.; Price, B. K.; Tour, J. M. Nature 2009, 458, 872-876.
(13) Cai, J. M.; Ruffieux, P.; Jaafar, R.; Bieri, M.; Braun, T.; Blankenburg, S.;
Muoth, M.; Seitsonen, A. P.; Saleh, M.; Feng, X. L.; Müllen, K.; Fasel, R. Nature
2010, 466, 470-473.
(14) Cai, J. M.; Pignedoli, C. A.; Talirz, L.; Ruffieux, P.; Söde, H.; Liang, L. B.;
Meunier, V.; Berger, R.; Li, R. J.; Feng, X. L.; Müllen, K.; Fasel, R. Nature
Nanotech. 2014, 9, 896-900.
(15) Chen, Y. C.; Cao, T.; Chen, C.; Pedramrazi, Z.; Haberer, D.; de Oteyza, D. G.;
Fischer, F. R.; Louie, S. G.; Crommie, M. F. Nature Nanotech. 2015, 10, 156-160.
(16) Dössel, L.; Gherghel, L.; Feng, X.; Müllen, K. Angew. Chem. Int. Ed. 2011,
50, 2540-2543.
(17) Schwab, M. G.; Narita, A.; Hernandez, Y.; Balandina, T.; Mali, K. S.; De
Feyter, S.; Feng, X.; Müllen, K. J. Am. Chem. Soc. 2012, 134, 18169-18172.
(18) Yang, X.; Dou, X.; Rouhanipour, A.; Zhi, L.; Rader, H. J.; Müllen, K. J. Am.
Chem. Soc. 2008, 130, 4216-4217.
(19) Narita, A.; Feng, X.; Hernandez, Y.; Jensen, S. A.; Bonn, M.; Yang, H.;
Verzhbitskiy, I. A.; Casiraghi, C.; Hansen, M. R.; Koch, A. H.; Fytas, G.; Ivasenko,
O.; Li, B.; Mali, K. S.; Balandina, T.; Mahesh, S.; De Feyter, S.; Müllen, K. Nature
Chem. 2014, 6, 126-132.
(20) Vo, T. H.; Shekhirev, M.; Kunkel, D. A.; Morton, M. D.; Berglund, E.; Kong,
L.; Wilson, P. M.; Dowben, P. A.; Enders, A.; Sinitskii, A. Nature Commun. 2014, 5,
3189.
(21) Narita, A.; Verzhbitskiy, I. A.; Frederickx, W.; Mali, K. S.; Jensen, S. A.;
Hansen, M. R.; Bonn, M.; De Feyter, S.; Casiraghi, C.; Feng, X.; Müllen, K. ACS
Nano 2014, 8, 11622-11630.
(22) Chen, L.; Hernandez, Y.; Feng, X.; Müllen, K. Angew. Chem. Int. Ed. 2012,
51, 7640-7654.
16
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 24
1
2
3
4
5
(23) Narita, A.; Wang, X. Y.; Feng, X. L.; Müllen, K. Chem. Soc. Rev. 2015, 44,
6616-6643.
(24) Lee, S. J.; Kim, J.-Y.; Kim, H. P.; Kim, D.; da Silva, W. J.; Schneider, F. K.;
Mohd Yusoff, A. R. b.; Jang, J. Chem. Commun. 2015, 51, 9185-9188.
(25) Osella, S.; Narita, A.; Schwab, M. G.; Hernandez, Y.; Feng, X.; Müllen, K.;
Beljonne, D. ACS Nano 2012, 6, 5539-5548.
(26) Bonaccorso, F.; Sun, Z.; Hasan, T.; Ferrari, A. C. Nature Photonics 2010, 4,
611-622.
(27) Abbas, A. N.; Liu, G.; Narita, A.; Orosco, M.; Feng, X. L.; Müllen, K.; Zhou,
C. W. J. Am. Chem. Soc. 2014, 136, 7555-7558.
(28) Sakaguchi, H.; Kawagoe, Y.; Hirano, Y.; Iruka, T.; Yano, M.; Nakae, T. Adv.
Mater. 2014, 26, 4134-4138.
(29) Talirz, L.; Söde, H.; Cai, J.; Ruffieux, P.; Blankenburg, S.; Jafaar, R.; Berger,
R.; Feng, X.; Müllen, K.; Passerone, D.; Fasel, R.; Pignedoli, C. A. J. Am. Chem. Soc.
2013, 135, 2060-2063.
(30) Bronner, C.; Stremlau, S.; Gille, M.; Brausse, F.; Haase, A.; Hecht, S.;
Tegeder, P. Angew. Chem. Int. Ed. 2013, 52, 4422-4425.
(31) Zhang, Y.; Zhang, Y. F.; Li, G.; Lu, J. C.; Lin, X.; Du, S. X.; Berger, R.; Feng,
X. L.; Müllen, K.; Gao, H. J. Appl. Phys. Lett. 2014, 105, 023101.
(32) Batra, A.; Cvetko, D.; Kladnik, G.; Adak, O.; Cardoso, C.; Ferretti, A.; Prezzi,
D.; Molinari, E.; Morgante, A.; Venkataraman, L. Chem. Sci. 2014, 5, 4419-4423.
(33) Vo, T. H.; Shekhirev, M.; Kunkel, D. A.; Orange, F.; Guinel, M. J. F.; Enders,
A.; Sinitskii, A. Chem. Commun. 2014, 50, 4172-4174.
(34) Zschieschang, U.; Klauk, H.; Mueller, I. B.; Strudwick, A. J.; Hintermann, T.;
Schwab, M. G.; Narita, A.; Feng, X. L.; Müllen, K.; Weitz, R. T. Adv. Electron.
Mater. 2015, 1, 1400010.
(35) Hillenkamp, F.; Peter-Katalinic, J., MALDI MS: a practical guide to
instrumentation, methods and applications. John Wiley & Sons: 2013.
(36) Palma, C. A.; Diller, K.; Berger, R.; Welle, A.; Björk, J.; Cabellos, J. L.;
Mowbray, D. J.; Papageorgiou, A. C.; Ivleva, N. P.; Matich, S.; Margapoti, E.;
Niessner, R.; Menges, B.; Reichert, J.; Feng, X.; Rader, H. J.; Klappenberger, F.;
Rubio, A.; Müllen, K.; Barth, J. V. J. Am. Chem. Soc. 2014, 136, 4651-4658.
(37) Kong, X.; Huang, Y.; Chen, Y. J. Mass. Spectrom. 2012, 47, 523-528.
(38) Dienel, T.; Kawai, S.; Söde, H.; Feng, X.; Müllen, K.; Ruffieux, P.; Fasel, R.;
Gröning, O. Nano Lett. 2015, 15, 5185-5190.
(39) Kawai, S.; Saito, S.; Osumi, S.; Yamaguchi, S.; Foster, A. S.; Spijker, P.;
Meyer, E. Nature Commun. 2015, 6, 8098.
(40) Bronner, C.; Leyssner, F.; Stremlau, S.; Utecht, M.; Saalfrank, P.; Klamroth,
T.; Tegeder, P. Phys. Rev. B 2012, 86, 085444.
(41) Centrone, A.; Brambilla, L.; Renouard, T.; Gherghel, L.; Mathis, C.; Müllen,
K.; Zerbi, G. Carbon 2005, 43, 1593-1609.
(42) Tommasini, M.; Lucotti, A.; Alfè, M.; Ciajolo, A.; Zerbi, G. Spectrochimica
Acta A 2016, 152, 134-148.
(43) Ferrari, A. C.; Meyer, J. C.; Scardaci, V.; Casiraghi, C.; Lazzeri, M.; Mauri,
F.; Piscanec, S.; Jiang, D.; Novoselov, K. S.; Roth, S.; Geim, A. K. Phys. Rev. Lett.
2006, 97, 187401.
(44) Bennett, P. B.; Pedramrazi, Z.; Madani, A.; Chen, Y. C.; de Oteyza, D. G.;
Chen, C.; Fischer, F. R.; Crommie, M. F.; Bokor, J. Appl. Phys. Lett. 2013, 103,
253114.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17
ACS Paragon Plus Environment

Page 19 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
(45) Chuang, S.; Battaglia, C.; Azcatl, A.; McDonnell, S.; Kang, J. S.; Yin, X. T.;
Tosun, M.; Kapadia, R.; Fang, H.; Wallace, R. M.; Javey, A. Nano Lett. 2014, 14,
1337-1342.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 24
1
2
3
4
Figures
5
6
7
8
9
a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
DBTT
Polyphenylene precursors
GNRs
b
c
d
e
100
x946
2655.1208
2654.1313
2656.1331
2657.1228
2653.1194
2658.1130
0
1000
1500
2000
2500
3000
3500
4000
4500
5000
5500
6000
6500
7000
7500
8000
m/z(Da)
Figure 1. CVD growth of chevron-type GNRs and characterization of the
intermediate polyphenylene precursors. a, CVD reaction scheme to chevron-type
GNRs. b, 5 K UHV STM topography of the GNR polyphenylene precursors on
Au/mica. c, d Zoom-in highlights undecameric and dodecameric polymer sections,
where the tetraphenyl substitutions are marked in green (also the structure model in
the inset of c and a). Contamination adsorbed during ambient-pressure CVD (white
protrusions) make high-resolution, long-range surveys challenging. Vt = 1 V, It = 100
pA. e, Mass spectrum of chevron-type polyphenylene precursors before
cyclodehydrogenation, with a regular sequence of peaks from dimer (m/z = ~1,062 Da)
to pentadecamer (m/z = ~7,966 Da). Inset shows the isotopic distributions of the
pentamer precursor, which agree well with the simulated results given in dotted lines.
19
ACS Paragon Plus Environment

Page 21 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
2 nm
a
b
c
e
C1s
Au 4f
281 282 283 284 285 286 287 288
Au 4d
9
UHV-GNRs
Au 4p
C1s
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CVD-GNRs
1 nm
0
100
200
300
400
500
600
700
800
d
284
Binding energy (eV)
G
180
f
(C-H)
400
400
UHV-GNRs
D
100
200
300
500
Raman shift (cm-1
)
D+D'
CVD-GNRs grown with H
2
2D
2D'
(C-O-C)
64
CVD-GNRs grown without H
2
0
500
1000
1500
2000
2500
3000
400
800
1200 1600 2000 2400 2800 3200
Raman shift (cm-1)
Energy loss (cm-1
)
Figure 2. Microscopic and spectroscopic characterizations of chevron-type GNRs.
a, Ambient STM topography of the CVD-grown GNRs on Au/mica. A GNR model is
also shown in the inset to scale. Vt = 450 mV, It = 15 pA. b-c, 5 K UHV STM data
depicts the chevron-type structure. Arrows highlight common GNR imperfections. Vt
= 1 V, It = 50 pA. d, Raman spectra of GNRs transferred on fused silica, measured at
532 nm, the inset is the magnified low-frequency mode (black oblong, bottom left). e,
f, XPS and HREELS spectra of the UHV-grown and CVD-grown GNRs. The
HREELS spectra of CVD-GNRs grown without H2 is also showed in f for comparison.
In the inset of e, the C1s core level spectra are presented.
20
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 24
1
2
3
4
5
6
7
8
a
d
b
c
SiO2/Si
GNRs
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
10 µm
e
f
14
25
20
15
10
5
V
V
gs
ds
12
10
8
-100 V
-80 V
-60 V
-40 V
-20 V
0 V
0 V
10 V
20 V
30 V
40 V
50 V
6
GNRs
Ti/Pd
4
SiO2
Si
0
2
1 µm
0
-5
-100
-50
0
50
100
-45
-30
-15
0
15
30
45
3
0
-3
V
(V)
gate
V
(V)
ds
0.5
1.0
1.5
2.0
2.5
Distance (m)
Figure 3. Transfer of CVD-grown chevron-type GNRs and micro-channel FET
device fabrication. a, Photograph of a 25×75 mm2 GNRs/Au/mica plate. b, Optical
microscope image of the GNR film transferred on SiO2/Si substrate, showing the
uniformity inside the film. The folding area can be found at the edge of the film. c,
Optical microscope image of the devices built on GNR film. d, AFM topographical
image of GNR film transferred on silicon wafer. The z-scale ranges from –16.5 to
16.5 nm. The cross sectional high analysis is indicated along the black line. e,
Transfer curves of a typical GNR thin film transistor measured at different Vds. The
results show that GNR transistors exhibit unipolar p-type behavior with on/off current
ratio of >1,000. The highest on/off ratio we measured is 6,000. f, Ids-Vds curves of the
GNR transistors. The curves display highly non-linear relationship, indicating a
Schottky contact between electrodes and GNR channel.
21
ACS Paragon Plus Environment

Page 23 of 24
Journal of the American Chemical Society
1
2
3
4
5
6
b
a 15
SiC
substrate
V
Metallic
side-gate
gs
10
-10 V
-5 V
+5 V
0 V
5
7
0
8
9
-5
Cr/Au
GNRs
-10
-15
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Graphene
Pd
-3
-2
V
-1
0
1
2
3
V
(V)
ds
c
d
5
V
ds
gs
4
5 V
0 V
-10 V
-20 V
-30 V
-35 V
1.2 V
4
3
2
1
0
1.0 V
0.8 V
0.6 V
0.4 V
2
0
-2
-40
-30
-20
-10
0
-1.0
-0.5
0.0
(V)
0.5
1.0
V
(V)
V
gate
ds
Figure 4. Electrical characteristics of nano-channel FET devices fabricated with
CVD-grown chevron-type GNRs. a, Ids-Vds curves of the short-channel devices at
various gate voltages. The inset shows the scanning electron micrograph (SEM) of a
typical short-channel device with an electrode spacing of ~46 nm. Scale bar, 100 nm.
b-d, Short-channel side-gate devices with graphene contacts. b, SEM image of the
device on SiC substrate. Scale bar, 300 nm. c, Transfer curves of a typical GNR
device with graphene electrodes, exhibiting an n-type conduction with high current
on/off ratios. d, Ids-Vds curves at different gate voltages. The non-linear behaviour of
the Ids-Vds curves indicates Schottky contacts with slightly asymmetric barriers.
22
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 24 of 24
1
2
3
4
a
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
polyphenylene heterojunctions
GNR heterojunctions
b
1600.7871
1599.7742
100
N-doped
1601.7828
1602.7963
1603.7926
0
1590
1595
1600
1605 m/z(Da)
1605 m/z(Da)
1605 m/z(Da)
1595.5887
1596.6176
c100
p–N junctions
1594.6300
1593.6018
1597.6293
1592.6088
1598.6238
1599.6362
0
1590
1595
1600
d
1593.6791
100
pristine
1592.6686
1594.6724
1595.6836
1596.6776
0
1590
1595
1600
e
f
1615.6326
100
1614.6150
N, S-co-doped
1616.6328
1613.6152
1612.6158
1617.6337
1618.6345
0
1600
1610
1620
m/z(Da)
m/z(Da)
m/z(Da)
100
1610.5157
1614.4771
Heterojunctions
1609.5173
1615.4769
1616.4946
1617.4775
1605.5447
1604.5479
1599.5685 1600.5986
1600
0
1610
1620
g100
N-doped
1600.6741
1599.6787
1601.6697
1602.6829
0
1600
1610
1620
Figure 5. Heteroatom-doping and heterojunctions of polyphenylene precursors
and GNRs. a, CVD reaction scheme to GNR heterojunctions by co-sublimation of
DBTT/DBNN and DBNS/DBNN. b-d, Isotopic analysis of the N-doped (b) and
pristine (d) polyphenylene trimers and their co-deposited p-N polyphenylene
heterojunctions (c) by high-resolution MALDI-TOF MS. e-g, Isotopic analysis of the
N, S-co-doped (e) and N-doped polyphenylene trimers (g) and their co-deposited
polyphenylene heterojunctions (f) by high-resolution MALDI-TOF MS.
23
ACS Paragon Plus Environment

Page 25 of 24
Journal of the American Chemical Society
1
2
3
4
TOC
5
6
14
V
ds
12
0 V
10 V
20 V
7
10
8
9
30 V
40 V
50 V
8
6
GNRs
Ti/Pd
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
SiO2
2
Si
0
-100
-50
0
50
100
V
(V)
gate
For Table of Contents Only
24
ACS Paragon Plus Environment