The synthesis and STM/AFM imaging of 'Olympicene' benzo[cd]pyrenes

Article abstract of DOI:10.1002/chem.201404877

H-Benzo[cd ]pyrene ('Olympicene') is a polyaromatic hydrocarbon and non-Kekul fragment of graphene. A new synthetic method has been developed for the formation of 6H-benzo[cd]pyrene and related ketones including the first time isolation of the unstable alcohol 6H-benzo[cd]pyren-6-ol. Molecular imaging of the reaction products with scanning tunnelling microscopy (STM) and non-contact atomic force microscopy (NC-AFM) characterised the 6H-benzo[cd]pyrene as well as the previously intangible and significantly less stable 5 H-benzo[cd ]pyrene, the fully conjugated benzo[cd]pyrenyl radical and the ketones as oxidation products.

Full text of DOI:10.1002/chem.201404877

DOI: 10.1002/chem.201404877
Full Paper
&
Synthetic Methods
The Synthesis and STM/AFM Imaging of ‘Olympicene’
Benzo[cd]pyrenes
Anish Mistry,[a] Ben Moreton,[a] Bruno Schuler,[b] Fabian Mohn,[b] Gerhard Meyer,[b]
Leo Gross,[b] Antony Williams,[c] Peter Scott,[a] Giovanni Costantini,[a] and David J. Fox*[a]
Abstract: H-Benzo[cd]pyrene (‘Olympicene’) is a polyaromatic
hydrocarbon and non-Kekulꢀ fragment of graphene. A new
synthetic method has been developed for the formation
of 6H-benzo[cd]pyrene and related ketones including
the first time isolation of the unstable alcohol
6H-benzo[cd]pyren-6-ol. Molecular imaging of the reaction
products with scanning tunnelling microscopy (STM) and
non-contact atomic force microscopy (NC-AFM) character-
ised the 6H-benzo[cd]pyrene as well as the previously intan-
gible and significantly less stable 5H-benzo[cd]pyrene, the
fully conjugated benzo[cd]pyrenyl radical and the ketones as
oxidation products.
Introduction
omatic molecules can be difficult to identify due to the high
number of possible isomers and overlapping NMR signals. STM
and AFM can however distinguish single molecules of such iso-
mers[13,23] and are therefore set to become extremely powerful
analytical tools in synthetic and materials chemistry. These
techniques are already being used to study graphene[26–27] and
smaller related polyaromatic hydrocarbons (PAHs).[12–16,28–29]
These latter types of molecules have been revealed as ex-
tremely versatile molecular building blocks for a number of re-
cently proposed applications in organic photovoltaics[30] and
electronics such as organic field effect transistors.[31–33] Interest
in PAHs was further motivated by their analogy with graphene,
of which they effectively represent finite sections with elec-
tronic properties determined by quantum confinement ef-
fects.[34–35] Moreover, PAHs are products of incomplete combus-
tion,[36–39] and can be found in the interstellar medium of galax-
ies.[40]
One of the most studied small PAHs is pentacene[12,33,41–44]
which has five benzene rings fused in a line. If the position of
two of these rings is altered, the result is the non- Kekulꢀ ben-
zo[cd]pyrene skeleton which was chosen for this study. Ben-
zo[cd]pyrene does not satisfy simple valence rules for a typical
Kekulꢀ aromatic hydrocarbon but benzo[cd]pyrenyl radical,[45–46]
anion[47–48] and cation[47,49] have been observed. Here we devel-
op the synthesis of 6H-benzo[cd]pyrene (’Olympicene’) and
characterise it and its derivatives on a metallic substrate by
using a combination of STM and NC-AFM.
Traditionally the primary methods of compound identification
for organic synthesis are melting point analysis, elemental
analysis, single crystal X-ray crystallography, refractive index
measurement and UV/Vis and infra-red spectroscopy. More re-
cently, the invention of nuclear magnetic resonance spectrom-
etry and high resolution mass spectrometry have been the
most significant developments in organic structure determina-
tion and remain the corner stones of synthetic chemistry re-
search. All of the techniques mentioned above are, of course,
based on the analysis of relatively large numbers of molecules,
either in the condensed or gas phase. Advances in atomic-
scale surface characterization, such as scanning tunnelling mi-
croscopy (STM) and atomic force microscopy (AFM), have al-
lowed a revolution in structure analysis by allowing for the
imaging[1–4] and stereochemical determination[5–11] of single
molecules. The development of high resolution non-contact
AFM (NC-AFM) methods have resolved molecular structure at
the atomic level.[12–20] Indeed it is now even possible to record
“before” and “after” atomic images of reactions induced on
single molecules with atomic resolution.[21–25] Unknown polyar-
[a] A. Mistry, B. Moreton, Prof. P. Scott, Dr. G. Costantini, Dr. D. J. Fox
Department of Chemistry
University of Warwick
Gibbet Hill Road, Coventry
CV4 7AL (UK)
Although the benzo[cd]pyrene skeleton was reported by
Clar[50] in 1932, the synthesis of H-benzo[cd]pyrene was first re-
ported in 1936 by Scholl and Meyer[51] involving the condensa-
tion of pyrene 1 with glycerol to give a ketone (thought to be
3-oxo-3H-benzo[cd]pyrene 2 or less probably the 5H-isomer
5) followed by a reduction with phosphorus to give 4,5-hydro-
3H-benzo[cd]pyrene 3 (Scheme 1). This trimethylene com-
pound was then selectively dehydrogenated by distillation
over hot copper to produce a hydrocarbon drawn as 3H-ben-
E-mail: d.j.fox@warwick.ac.uk
[b] Dr. B. Schuler, Dr. F. Mohn, Dr. G. Meyer, Dr. L. Gross
IBM Research - Zurich
Sꢀumerstrasse 4, 8803 Rꢁschlikon (Switzerland)
[c] Dr. A. Williams
Royal Society of Chemistry
US office, 904 Tamara’s Circle
Wake Forest, NC 27587 (USA)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201404877.
Chem. Eur. J. 2015, 21, 2011 – 2018
2011
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 2. i) SOCl2; ii) AlCl3, CH2Cl2; iii) LiAlH4, THF, 30% (3 steps); iv) Amber-
lyst-15, 54%.
Scheme 1. i) glycerol, H2SO4; ii) phosphorus; iii) Cu, heat.
Scheme 3. Additional products formed in the reaction of carboxylic acid 9 in
polyphosphoric acid at 808C under a nitrogen atmosphere.
zo[cd]pyrene 4. A year later Vollmann et al. published an exten-
sive study[52] in this area including the syntheses of both oxo-
H-benzo[cd]pyrene and H-benzo[cd]pyrene and concluded that
the structures made by Scholl and Meyer were 6-oxo-6H-ben-
zo[cd]pyrene 7 and 6H-benzo[cd]pyrene 8 rather than their re-
spective 3H or 5H-isomers 2,4–6. The syntheses of the 6H-
ketone 7[53–54] and the 6H-hydride 8[55] were subsequently inde-
pendently repeated by Clar and Johnson[56] as part of ultravio-
let spectroscopy studies of polyaromatic compounds.
pyrene 5, identified by comparison with the data published by
Arnbjerg et al.,[59] as well as trimethylene hydrocarbons 3[54]
and 13.[60] In addition, a trace quantity of pyrene 1 was also
isolated. A seventh rather unstable compound was also isolat-
ed. Analytical data suggest that this unstable compound is the
previously unisolated 6-hydroxy-6H-benzo[cd]pyrene 14. Over-
1
all the ratio of crude products by H NMR spectrometry for this
Results and Discussion
PPA mediated reaction is 5, 31%; 7, 30%; 8, 25.5%; 14, 6%; 3,
4%; 13, 3.5%, along with a trace quantity of pyrene 1. Reason-
able quantities of compounds 5 (30%), 7 (29%) and 8 (26%)
could be isolated by chromatography.
Synthesis
Reid and Bonthrone’s synthesis of a number of benzo[cd]pyr-
ene compounds involved a more mechanistically defined ring
closure of 3-(pyren-1’-yl)propionic acid 9 in liquid HF.[49] The re-
sulting ketone 11 is reduced and the alcohol 12 dehydrated to
form H-benzo[cd]pyrene as a mixture of isomers. The work de-
scribed herein started with a repeat of Reid and Bonthrone’s
synthesis using modified ring-closure conditions (Scheme 2).
The NMR spectra of the H-benzo[cd]pyrene are consistent with
the formation of the 6H-isomer 8 as the major product.[57] The
synthesis of H-benzo[cd]pyrene via acid chloride 10 (Scheme 2)
was however rather long and an alternative route was sought.
Polyphosphoric acid (PPA) was chosen to mediate the ring
closure of acid 9. Heating the carboxylic acid at 808C in PPA in
a nitrogen atmosphere for 24 h did not result in the formation
We propose that PPA mediates the ring closure of carboxylic
acid 9 to give ketone 11 that reacts further in these conditions.
A mechanism is proposed in Scheme 4 in which the enol form
of the ketone (15) donates a hydride to another molecule of
ketone 11 producing 5-hydroxy-4,5-hydro-3H-benzo[cd]pyrene
12 and 5-oxo-5H-benzo[cd]pyrene 5. The alcohol then dehy-
drates in the acidic reaction conditions to produce 3H-ben-
zo[cd]pyrene 4 which isomerises by proton transfer to the 6H-
isomer 8. Mechanisms for the formation of the minor compo-
nents 1, 3 and 13 are discussed in the supplementary informa-
tion.
This reaction was repeated a number of times and it was
shown that, if heated for long enough in PPA, the 6H-ben-
zo[cd]pyrene 8 and 5-oxo-5H-benzo[cd]pyrene 5 in the reaction
mixture would be converted into 6-oxo-6H-benzo[cd]pyrene 7.
It was therefore concluded that on long heating in PPA: 1) 6H-
benzo[cd]pyrene 8 is slowly oxidised by trace levels of air to 6-
oxo-6H-benzo[cd]pyrene 7, and 2) 5-oxo-5H-benzo[cd]pyrene 5
1
of ketone 11. Rather, analysis of the H NMR spectrum of the
crude reaction product indicated the formation of, among
other compounds, 6H-benzo[cd]pyrene[47] 8 and 6-oxo-6H-ben-
zo[cd]pyrene[58] 7 (Scheme 3). Purification by column chroma-
tography also resulted in the isolation of 5-oxo-5H-benzo[cd]-
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2012
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 5. 6H-Benzo[cd]pyrene 8 catalysed isomerisation of ketone 5.
Scheme 4. Formal disproportionation of ketone 11.
isomerises to 6-oxo-6H-benzo[cd]pyrene 7. Though it is
known[58] that 6H-benzo[cd]pyrene 8 oxidises in air over time
due to the relatively weak methylene CÀH bonds,[61] it has not
been reported that 5-oxo-5H-benzo[cd]pyrene 5 isomerises to
the 6H-ketone 7.[49,59]
increases its conjugation in creating ketone 5. 6H-Benzo[cd]-
pyrene 8 could also act as a hydride donor in acidic conditions,
also producing a conjugated molecule, that is, the benzo[cd]-
pyrenium cation 18 (Scheme 5). Isomerisation of the product
alcohol 5-hydroxy-5H-benzo[cd]pyrene 17 via the same aro-
matic cation to 6-hydroxy-6H-benzo[cd]pyrene 14 could then
occur. The 6H-alcohol 14 would then transfer a hydride back
to the remaining benzo[cd]pyrenium cation 18 regenerating 6-
oxo-6H-benzo[cd]pyrene 7 and 6H-benzo[cd]pyrene 8. To con-
firm this, ketone 5 was heated in PPA with added 6H-ben-
zo[cd]pyrene 8, resulting in its isomerisation to 6-oxo-6H-ben-
zo[cd]pyrene 7.
The redox disproportion of the benzo[cd]pyrenium cation 18
to the compounds 7 and 8 has been previously reported by
Reid and Banthone.[49] While they proposed alcohol 14 as an
intermediate, they could not isolate it. The isolated alcohol 14
degrades in base-washed and dried CDCl3 and peaks corre-
sponding to 6-oxo-6H-benzo[cd]pyrene 7 and 6H-benzo[cd]pyr-
ene 8 (Scheme 6) appeared during 13C NMR spectroscopic anal-
ysis (see the Supporting Information).
To support the proposed mechanism, 5-oxo-4,5-hydro-3H-
benzo[cd]pyrene 11 and 5-oxo-5H-benzo[cd]pyrene 5 were in-
dependently heated at 808C in PPA in a nitrogen atmosphere,
as in the earlier experiments. 5-Oxo-4,5-hydro-3H-benzo[cd]pyr-
ene 11 reacted to give a similar range of products as from the
reaction of carboxylic acid 9, with the exception that no 7,8-di-
hydro-9H-cyclopenta[a]pyrene 13 was produced. Extended
heating of 5-oxo-5H-benzo[cd]pyrene 5 in PPA at 808C under
either a nitrogen or air atmosphere resulted in no reaction and
5-oxo-5H-benzo[cd]pyrene 5 was returned unchanged. Similar
treatment of 6H-benzo[cd]pyrene 8 resulted in complete oxida-
tion to 6-oxo-6H-benzo[cd]pyrene 7. These reactions with com-
pounds 11 and 5 support the proposed disproportionation
shown in Scheme 4, but do not support the simple acid-cata-
lysed isomerisation of 5-oxo-5H-benzo[cd]pyrene 5 to its 6H-
isomer 7.
While the acid-catalysed or, more specifically, the PPA-cata-
lysed isomerisation of ketone 5 is not readily rationalised, it
was evident from our initial reactions that the isomerisation
takes place if acid 9 or ketone 11 is the starting material. This
result therefore suggested that another reaction component
was mediating the isomerisation. In Scheme 4 it is proposed
that 5-hydroxy-4,5-hydro-3H-benzo[cd]pyrene 12 is formed by
donation of a hydride from the enol of 5-oxo-4,5-hydro-3H-
benzo[cd]pyrene (15). The donation of a hydride from the enol
According the above experiments, 6-oxo-6H-benzo[cd]pyr-
ene 7 and 6H-benzo[cd]pyrene 8 seem to be thermodynami-
cally more stable than their respective 5H isomers. This obser-
vation prompted a set of DFT calculations. For six H-benzo[cd]-
pyrene isomers (2H, 3H, 5H, 6H, 5bH, 11aH) and four oxo-H-
benzo[cd]pyrene isomers (2H, 3H, 5H, 6H), structures were
energy minimised and zero-point energies calculated using the
B3LYP functional[62–63] and the 6-31G(d)[64] basis using Firefly
software.[65] Single-point energy calculations were then per-
formed on the ten final geometries using the B3LYP-D3(BJ)
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2013
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
8.[69] Recently, however, Valentine et al. have studied all of the
possible isomers of H-benzo[cd]pyrene and shown that the 6H-
isomer 8 is the most stable.[70]
STM characterisation
As a further means of characterisation, the benzo[cd]pyrene
molecules were deposited in ultra-high vacuum (UHV) onto an
atomically flat and clean copper crystal and imaged by means
of scanning tunnelling microscopy. Initially, STM was used with
the aim of identifying the different species based upon their
appearance. Three benzo[cd]pyrenes were chosen for this
study, 6H-benzo[cd]pyrene 8 and ketones 5 and 7. Clean
Cu(111) substrates were prepared by cycles of Ar+ sputtering
(1 keV, 12 mA/cm2) and annealing (823 K) in UHV (base pressure
3ꢂ10À10 mbar). For the STM measurements, the molecules
were deposited from a Knudsen cell at 383 K onto room tem-
perature Cu(111) before cooling to 77 K for imaging. All three
molecules are imaged as bright trapezoidal features with very
similar dimensions in STM, so that it is not possible to directly
differentiate between them based on their individual appear-
ance. However, the way in which they assemble is markedly
different, as shown in Figure 2. 6H-Benzo[cd]pyrene 8 mole-
cules (Figure 2a) assemble into islands with a close-packed ar-
rangement. The island size ranges from two to several tens of
molecules, depending on the local molecular coverage. Oxo-H-
benzo[cd]pyrenes 7 and 5 (Figure 2b and c, respectively) form
only smaller aggregates which do not increase in size with in-
creasing coverage. The formation of relatively small molecular
aggregates instead of large extended islands could be due
either to a restricted surface diffusivity of the molecules on
Cu(111) or to long-range repulsive interactions caused by mo-
lecular dipoles.[71] Both effects are however not expected to
significantly alter the (short range) intermolecular bonding
configurations.
Scheme 6. Disproportionation of alcohol 14.
Figure 1. Benzo[cd]pyrene position numbering, and additional possible H-
benzo[cd]pyrene and oxo-H-benzo[cd]pyrene isomers.
Table 1. DFT energies calculated at B3LYP-D3(BJ)/def2-QZVP (including
zero-point energy (ZPE) at B3LYP/6-31G(d)) of H-benzo[cd]pyrene and
oxo-H-benzo[cd]pyrene isomers (see Figure 1).
Substitution
position
Relative H-isomer energy
Relative oxo-H isomer energy
[a]
[a]
[kJmolÀ1
]
[kJmolÀ1
]
2H
3H
5H
6H
82.2 (19)
24.5 (4)
29.7 (6)
0 (8)
69.5 (20)
22.9 (2)
27.5 (5)
0 (7)
Not possible
Not possible
Dimers and trimers constitute the great majority of these is-
lands, which can be classified into seven main types (Figure 3).
There are two different trimer units, open and closed trimers
(OT and CT, respectively). In the OT the molecules arrange with
their longer edge perpendicular to a line pointing radially from
the centre of the cluster, whereas in the CT the packing is
closer and the longer edge faces towards a neighbouring mol-
ecule. Most of the dimers can be divided into two large
5bH
11aH
164.4 (21)
127.9 (22)
[a] Compound number is given in parentheses (see text and schemes).
functional[66] and the def2-QZVP
basis set[67] as recommended by
Grimme
and
co-workers.[68]
These calculations are consistent
with the 6H-isomers being the
most stable for both H-ben-
zo[cd]pyrene and oxo-H-ben-
zo[cd]pyrene (Figure 1, Table 1).
The relative stabilities of H-ben-
zo[cd]pyrenes were calculated
with DFT by Bonfanti et al. and
the 5H-isomer 6 was found to
be the most stable but without
Figure 2. STM images of a) 8 (5ꢂ1014 molecules/cm2), b) 7 (2ꢂ1014 molecules/cm2) and c) 5 (1ꢂ1015 molecules/
consideration of the 6H-isomer cm2) deposited at room temperature on Cu(111) and imaged at 77 K.
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2014
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
zo[cd]pyrene 7 has the oxygen of one molecule positioned
symmetrically between the two hydrogens of the other mole-
cule, whereas for 5-oxo-5H-benzo[cd]pyrene 5 the oxygen is in
an asymmetric position, effectively interacting with only one
hydrogen. Likewise for the OHH dimer, symmetrical ketone 7
can form two H-bonds whereas non-symmetrical ketone 5 can
only form a single H-bond. Ultimately, this difference in H-
bonding arrangement might be the cause of the LHT and OHH
configurations being favoured by ketone 7. A similar reasoning
can be applied to the OHT and LHH dimers, with ketone 5
being expected to have the more favourable H-bond configu-
rations. A preference of non-symmetrical 5 over symmetrical
ketone 7 to form the LHH dimer is indeed observed (7 1.6%
and 5 5.3%), but this is not the case for the OHT dimer that
occurs with a similar frequency for the two ketones (7 6.4%
and 5 5.9%). This shows that simple arguments based on the
number of intermolecular hydrogen bonding-type interactions
allowed by the molecular structures and their relative orienta-
tion do not fully capture the experimental observations (see
further discussion in the Supporting Information).
Gas-phase semi-empirical AM1 calculations were performed
to attain a more quantitative evaluation of the formation
energy for each of the aggregates proposed in Figure 3, but
these were also not able to rationalise the measured distribu-
tion of island types (see the Supporting Information).
This hints at an integral role played by the Cu(111) surface in
determining the formation of these structures, which can be
expected since the substrate coupling for 6-oxo-6H-benzo[cd]-
pyrene 7 is substantially higher compared to the physisorbed
6H-benzo[cd]pyrene 8.[72] Moreover, for some of the more open
structures shown in Figure 3, we cannot exclude the presence
of substrate Cu adatoms involved in coordination bonds with
molecules 5 and 7. Full-scale DFT calculations of the clusters
and the substrate would be very resource expensive and go
well beyond the scope of this work.
Figure 3. Left: representative STM images and model configurations as ob-
tained from the direct overlay of the molecular structure onto the STM
images. Right: relative population frequencies for the seven main different
aggregates of the 6H-ketone (sym, 7), and 5H-ketone (asym, 5) (see Support-
ing Information for discussion).
classes: head to head (HH), where the long edges of the mole-
cules are oriented towards one another, and head to tail (HT),
where the long edge of one molecule faces the short edge of
the other. Two further classifications can be identified within
these two main families, depending on whether the molecules
are in line (LHH and LHT) or have an offset (OHH and OHT).
The last type of dimer aggregate, denoted as angled dimer
(AD), appears similar to an open trimer with one of the mole-
cules removed. Each of these structures is expected to be sta-
bilised mainly by hydrogen bonding-type interactions between
the oxygen and the aromatic hydrogens. Although all seven
types of assemblies are observed in both compounds 7 and 5,
their relative frequencies vary for the two isomers, as shown in
Figure 3. A statistical analysis of the data (250 clusters for 5
and 170 clusters for 6) exhibits a very similar population of
trimers but significant differences with respect to the distribu-
tion of the dimers. The largest difference in relative frequencies
is observed for the LHT dimer (7 12.4% and 5 5.9%) and the
OHH dimer (7 16.5% and 5 6.5%). This can be rationalised by
considering the different positions of the oxygen and the pos-
sible hydrogen bonding-type interactions that can be corre-
spondingly formed: for the LHT configuration, 6-oxo-6H-ben-
NC-AFM
The results of the previous section demonstrate that STM can
partially characterise the different molecular species by their
distinct aggregation state. To unambiguously identify the three
benzo[cd]pyrenes by their individual appearance, we used NC-
AFM, which has recently emerged as a powerful analytical tool
to characterise single adsorbed molecules with atomic resolu-
tion.[12] By using functionalised tips, it has been shown that the
structure of unknown molecules can be elucidated[13,15] and
bond-order differences between individual aromatic CÀC
bonds discriminated.[14]
To investigate the individual benzo[cd]pyrenes we deposited
the molecules at a sample temperature of 10 K. The corre-
sponding measurements, shown in Figure 4 and Figure 5, were
acquired at 5 K. Figure 4 shows constant-height AFM measure-
ments of the observed species recorded with a CO tip. As dis-
cussed above, both our synthetic studies and those reported
by others[58] have shown that 6H-benzo[cd]pyrene 8 is oxidized
in air to 6-oxo-6H-benzo[cd]pyrene 7. NC-AFM analysis con-
firmed this, showing, in addition to 6H-benzo[cd]pyrene 8 (Fig-
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2015
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Unexpectedly, NC-AFM also identified trace amounts (0.5–
5%) of 5H-benzo[cd]pyrene 6 (Figure 4d) along with the 6H-
isomer 8. This NC-AFM image is the first evidence of the exis-
tence of compound 6, which has previously not been isolated
or even detected in mixtures. It is potentially easily isomerised
by acid catalysis and DFT calculations (see earlier) predict it to
be significantly less stable than 6H-isomer 8. Trace amounts of
5-oxo-5H-benzo[cd]pyrene 5 (Figure 4e) were also present in
the partially oxidized sample. Although previous reports of the
oxidation of 6H-benzo[cd]pyrene 8 characterised the most
stable 6-oxo ketone-isomer 7 as the major product, the imag-
ing of 5-oxo-5H-benzo[cd]pyrene 5 shows that reaction can
occur at less thermodynamically favourable positions. While
Reid and Bonthrone[49] have demonstrated that benzo[cd]pyr-
ene cation 18 reacts preferentially at the 6-position, oxidation
by air of 6H-benzo[cd]pyrene 8 may very well occur via the
fully conjugated benzo[cd]pyrene radical 23. Molecular orbital
calculations[49] predict that the 6-position would be the most
reactive for benzo[cd]pyrene cations, radicals or anions as it
has the largest orbital coefficient of the LUMO, SOMO or
HOMO, respectively. Identification of 5-oxo-5H-benzo[cd]pyrene
5 in the product mixture shows that, as for the cation 18,[49]
the 5-position is the next most favourable reaction site. It is in-
teresting to note that no further isomers of H-benzo[cd]pyrene
(e.g., 19, 4, 21, 22) or oxo-H-benzo[cd]pyrene (e.g., 2 or 20)
could be identified in the mixture by NC-AFM, presumably be-
cause they are too unstable and, if formed, may have isomer-
ised to more stable compounds.
Figure 4. AFM images with CO tip functionalisation of a) 6H-benzo[cd]pyrene
8; b) 6-oxo-6H-benzo[cd]pyrene 7; c) benzo[cd]pyrene radical 23; d) 5H-ben-
zo[cd]pyrene 6; e) 5-oxo-5H-benzo[cd]pyrene 5; in images a) and d) CH2
groups can be seen as bright regions; in images b) and e) C=O can been
seen as dark regions; f) The structure of benzo[cd]pyrene radical 23. The
black scale bar equals 0.5 nm in all the images.
The benzo[cd]pyrene radical 23 (Figure 4c,f) has not been
previously isolated but can be generated on the surface from
6H-benzo[cd]pyrene 8 by removing one hydrogen atom from
the 6-position by positioning the STM tip above it and apply-
ing a voltage pulse (1.6 V at 10 pA) (Figure 5). Note that, in
images subsequent to the removal of the hydrogen from 6H-
benzo[cd]pyrene 8, we often observed dark stripes in the fast-
scanning direction (see Figure 5c). We relate these stripes to
the removed hydrogen that gets trapped for short periods of
time in the tunnelling junction, causing a decrease of the tip
height. After a few images, the hydrogen is usually moved out
of the scanned region (Figure 5d). The radical 23, which is left
after the dehydrogenation, is strongly chemisorbed on
Cu(111)[72] and has a tilted planar geometry.[73] This tilt causes
a gradual decrease of the AFM contrast of the carbon back-
bone, where the brightest contrast (most repulsive region) is
seen on atom and bond positions at the molecule short edge,
which is adsorbed farthest from the surface.
Figure 5. STM images at 5 K with a Cu terminated tip. a) STM image of two
6H-benzo[cd]pyrenes 8 (white) and two CO molecules (black). The area
within the dashed box is shown in b) and c); b) STM image of 6H-benzo[cd]-
pyrene 8 with ring showing position of the STM tip above the 6-carbon of 8
where the voltage pulse is applied (1.6 V at 10 pA, see text); c) STM image
of the same area as in b) but after the voltage pulse is applied showing the
benzo[cd]pyrene radical 23 (white) and features (dark stripes) that we assign
to the dissociated hydrogen atom; d) STM image of a wider area showing
6H-benzo[cd]pyrene 8 (‘Olympicene’), 6-oxo-6H-benzo[cd]pyrene 7 (‘Ketone’),
benzo[cd]pyrene radical 23 (‘Radical’) and a carbon monoxide molecule
(‘CO’). The black scale bar equals 2 nm in all the images.
Conclusion
ure 4a), a substantial amount of 6-oxo-6H-benzo[cd]pyrene 7
(Figure 4b) (ꢀ20-40%). The two species can be readily distin-
guished by their characteristic contrast at the position of the
attached ketone or CH2 group: the CH2 group appears brighter
(repulsive interaction) and the ketone darker (attractive interac-
tion). This is partially related to chemical contrast, but also
a consequence of adsorption-induced local relaxations out of
the molecular plane. The detailed adsorption characteristics
are discussed elsewhere.[73]
The synthetic and imaging study presented here combines
new discoveries in the long-standing area of polyaromatic
chemistry with recent cutting-edge advances in analytical sur-
face science. This innovative methodology makes it possible
for problems in compound identification, long-encountered in
organic synthesis, to be solved at a single-molecule level. The
identification of the previously unseen, and relatively unstable
molecule, 5H-benzo[cd]pyrene 6, exemplifies the potential of
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2016
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[15] K. Ø. Hanssen, B. Schuler, A. J. Williams, T. B. Demissie, E. Hansen, J. H.
Andersen, J. Svenson, K. Blinov, M. Repisky, F. Mohn, G. Meyer, J.-S.
Svendsen, K. Ruud, M. Elyashberg, L. Gross, M. Jaspars, J. Isaksson,
Angew. Chem. Int. Ed. 2012, 51, 12238–12241; Angew. Chem. 2012, 124,
12404–12407.
[16] D. G. de Oteyza, P. Gorman, Y.-C. Chen, S. Wickenburg, A. Riss, D. J.
Mowbray, G. Etkin, Z. Pedramrazi, H.-Z. Tsai, A. Rubio, M. F. Crommie,
F. R. Fischer, Science 2013, 340, 1434–1437.
this combined approach. The single-atom molecular manipula-
tion used in the formation of the ‘olympicene’ radical and the
detailed characterisation of aggregated complexes will lead to
the design and on-surface use of new polyaromatic ‘mono-
mers’ for the assembly of complex molecular architectures. We
expect the methodology presented in this publication to
quickly develop into a new technique for the analysis of un-
known compounds or complex product mixtures in synthetic
chemistry.
[17] A. M. Sweetman, S. P. Jarvis, H. Sang, I. Lekkas, P. Rahe, Y. Wang, J.
Wang, N. R. Champness, L. Kantorovich, P. Moriarty, Nat. Commun. 2014,
5, 3931.
ˇ
[18] N. Pavlicek, B. Fleury, M. Neu, J. Niedenfꢃhr, C. Herranz-Lancho, M.
Ruben, J. Repp, Phys. Rev. Lett. 2012, 108, 086101.
[19] J. Zhang, P. Chen, B. Yuan, W. Ji, Z. Cheng, X. Qiu, Science 2013, 342,
611–614.
Experimental Section
[20] R. Pawlak, S. Kawai, S. Fremy, T. Glatzel, E. Meyer, ACS Nano 2011, 5,
6349–6354.
[21] S.-W. Hla, L. Bartels, G. Meyer, K.-H. Rieder, Phys. Rev. Lett. 2000, 85,
2777–2780.
[22] J. R. Hahn, W. Ho, Phys. Rev. Lett. 2001, 87, 166102.
[23] J. Henzl, M. Mehlhorn, H. Gawronski, K.-H. Rieder, K. Morgenstern,
Angew. Chem. Int. Ed. 2006, 45, 603–606; Angew. Chem. 2006, 118,
617–621.
All experimental details for isolated compounds, including proce-
dures, characterisation and copies of spectra, and computationally
calculated structures and energies and semi-empirical aggregation
calculations are available in the Supporting Information.
Acknowledgements
[24] J. Repp, G. Meyer, S. Paavilainen, F. E. Olsson, M. Persson, Science 2006,
312, 1196–1199.
[25] P. Liljeroth, J. Repp, G. Meyer, Science 2007, 317, 1203–1206.
[26] M. P. Boneschanscher, J. van der Lit, Z. Sun, I. Swart, P. Liljeroth, D. Van-
maekelbergh, ACS Nano 2012, 6, 10216–10221.
[27] J. van der Lit, M. P. Boneschanscher, D. Vanmaekelbergh, M. Ijꢅs, A.
Uppstu, M. Ervasti, A. Harju, P. Liljeroth, I. Swart, Nat. Commun. 2013, 4,
2023.
[28] B. Schuler, S. Collazos, L. Gross, G. Meyer, D. Pꢀrez, E. Guitiꢄn, D. PeÇa,
Angew. Chem. Int. Ed. 2014, 53, 9004–9006; Angew. Chem. 2014, 126,
9150–9152.
[29] A. Riss, S. Wickenburg, P. Gorman, L. Z. Tan, H.-Z. Tsai, D. G. de Oteyza,
Y.-C. Chen, A. J. Bradley, M. M. Ugeda, G. Etkin, S. G. Louie, F. R. Fischer,
M. F. Crommie, Nano Lett. 2014, 14, 2251–2255.
[30] Y.-J. Cheng, S.-H. Yang, C.-S. Hsu, Chem. Rev. 2009, 109, 5868–5923.
[31] C. D. Dimitrakopoulos, P. R. L. Malenfant, Adv. Mater. 2002, 14, 99–117.
[32] M. Halik, H. Klauk, U. Zschieschang, T. Kriem, G. Schmid, W. Radlik, K.
Wussow, Appl. Phys. Lett. 2002, 81, 289–291.
A.M. and D.J.F. acknowledges financial support from the Uni-
versity of Warwick. G.C. and B.M. acknowledge financial sup-
port from the EPSRC (EP/D000165/1), from the EU through the
ERC Starting Grant “VISUAL-MS”, from the Royal Society
through Grant n. RG100917 and from the Warwick-Santander
Fund. Part of the equipment used in this research was ob-
tained through Birmingham Science City: “Innovative Uses for
Advanced Materials in the Modern World” with support from
Advantage West Midlands and part funded by the European
Regional Development Fund. B.S., F.M., G.M. and L.G. acknowl-
edge financial support from the European project ARTIST (con-
tract no. 243421) and the European Research Council advanced
grant CEMAS.
[33] R. Schroeder, L. A. Majewski, M. Grell, Appl. Phys. Lett. 2003, 83, 3201–
3203.
Keywords: isomerization
·
polyaromatic hydrocarbon
·
[34] J. Wu, W. Pisula, K. Mꢃllen, Chem. Rev. 2007, 107, 718–747.
[35] K. Mꢃllen, J. P. Rabe, Acc. Chem. Res. 2008, 41, 511–520.
[36] J. L. Durant, W. F. Busby, A. L. Lafleur, B. W. Penman, C. L. Crespi, Mutat.
Res.-Genet. Tox. 1996, 371, 123–157.
[37] J. L. Durant, A. L. Lafleur, E. F. Plummer, K. Taghizadeh, W. F. Busby, W. G.
Thilly, Environ. Sci. Technol. 1998, 32, 1894–1906.
[38] M. P. Fraser, G. R. Cass, B. R. T. Simoneit, Environ. Sci. Technol. 1998, 32,
2051–2060.
[39] M. P. Hannigan, G. R. Cass, B. W. Penman, C. L. Crespi, A. L. Lafleur, W. F.
Busby, W. G. Thilly, B. R. T. Simoneit, Environ. Sci. Technol. 1998, 32,
3502–3514.
radical · scanning probe microscopy · synthesis
[1] J. Frommer, Angew. Chem. Int. Ed. Engl. 1992, 31, 1298–1328; Angew.
Chem. 1992, 104, 1325–1357.
[2] T. A. Jung, R. R. Schlittler, J. K. Gimzewski, Nature 1997, 386, 696–698.
[3] J. K. Gimzewski, C. Joachim, Science 1999, 283, 1683–1688.
[4] W. Ho, J. Chem. Phys. 2002, 117, 11033–11061.
[5] G. P. Lopinski, D. J. Moffatt, D. D. M. Wayner, R. A. Wolkow, Nature 1998,
392, 909–911.
[6] A. Kꢃhnle, T. R. Linderoth, B. Hammer, F. Besenbacher, Nature 2002, 415,
891–893.
[40] A. G. G. M. Tielens, Annu. Rev. Astron. Astrophys. 2008, 46, 289–337.
[41] N. Koch, I. Salzmann, R. L. Johnson, J. Pflaum, R. Friedlein, J. P. Rabe,
Org. Electron. 2006, 7, 537–545.
[7] M. Alemani, M. V. Peters, S. Hecht, K.-H. Rieder, F. Moresco, L. Grill, J. Am.
Chem. Soc. 2006, 128, 14446–14447.
[8] M. Lingenfelder, G. Tomba, G. Costantini, L. Colombi Ciacchi, A. De Vita,
K. Kern, Angew. Chem. Int. Ed. 2007, 46, 4492–4495; Angew. Chem.
2007, 119, 4576–4579.
[42] J. Lagoute, K. Kanisawa, S. Fçlsch, Phys. Rev. B 2004, 70, 245415.
´
[43] J. Repp, G. Meyer, S. M. Stojkovic, A. Gourdon, C. Joachim, Phys. Rev.
Lett. 2005, 94, 026803.
[44] W. H. Soe, C. Manzano, A. De Sarkar, N. Chandrasekhar, C. Joachim, Phys.
Rev. Lett. 2009, 102, 176102.
[9] B. Moreton, Z. Fang, M. Wills, G. Costantini, Chem. Commun. 2013, 49,
6477–6479.
[10] R. Fasel, M. Parschau, K.-H. Ernst, Nature 2006, 439, 449–452.
[45] A. H. Reddoch, D. H. Paskovich, Chem. Phys. Lett. 1969, 3, 351–352.
[46] J. Nowakowski, Chem. Phys. Lett. 1972, 13, 589–591.
[47] I. Murata, K. Yamamoto, O. Hara, Tetrahedron Lett. 1974, 15, 2047–2048.
[48] A. Streitwieser, Jr., J. M. Word, F. Guibe, J. S. Wright, J. Org. Chem. 1981,
46, 2588–2589.
ˇ
[11] N. Pavlicek, C. Herranz-Lancho, B. Fleury, M. Neu, J. Niedenfꢃhr, M.
Ruben, J. Repp, Phys. Status Solidi B 2013, 250, 2424–2430.
[12] L. Gross, F. Mohn, N. Moll, P. Liljeroth, G. Meyer, Science 2009, 325,
1110–1114.
[13] L. Gross, F. Mohn, N. Moll, G. Meyer, R. Ebel, W. M. Abdel-Mageed, M.
Jaspars, Nat. Chem. 2010, 2, 821–825.
[49] D. H. Reid, J. Chem. Soc. 1965, 5920–5926.
[50] E. Clar, Ber. Dtsch. Chem. Ges. B 1932, 65, 1425–1428.
[51] R. Scholl, H. K. Meyer, Ber. Dtsch. Chem. Ges. B 1936, 69, 152–158.
[14] L. Gross, F. Mohn, N. Moll, B. Schuler, A. Criado, E. Guitiꢄn, D. PeÇa, A.
Gourdon, G. Meyer, Science 2012, 337, 1326–1329.
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2017
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[52] H. Vollmann, H. Becker, M. Corell, H. Streeck, G. Langbein, Justus Liebigs
Ann. Chem. 1937, 531, 1–159.
[66] S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456–
1465.
[53] E. Clar, Chem. Ber. 1948, 81, 520–527.
[67] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–3305.
[68] H. Kruse, L. Goerigk, S. Grimme, J. Org. Chem. 2012, 77, 10824–10834.
[69] M. Bonfanti, S. Casolo, G. F. Tantardini, A. Ponti, R. Martinazzo, J. Chem.
Phys. 2011, 135, 164701–164710.
[70] A. J. S. Valentine, D. A. Mazziotti, J. Chem. Phys. A 2013, 117, 9746–9752.
[71] S. Kervyn, N. Kalashnyk, M. Riello, B. Moreton, J. Tasseroul, J. Wouters,
T. S. Jones, A. De Vita, G. Costantini, D. Bonifazi, Angew. Chem. Int. Ed.
2013, 52, 7410–7414; Angew. Chem. 2013, 125, 7558–7562.
[72] W. Liu, B. Schuler, Y. Xu, N. Moll, G. Meyer, L. Gross, A. Tkatchenko, to be
published.
[54] E. Clar, D. G. Stewart, J. Am. Chem. Soc. 1953, 75, 2667–2672.
[55] E. Clar, D. G. Stewart, J. Am. Chem. Soc. 1952, 74, 6235–6238.
[56] A. W. Johnson, J. Org. Chem. 1959, 24, 833–836.
[57] J. Morgenthaler, C. Ruchardt, Eur. J. Org. Chem. 1999, 2219–2230.
[58] C. Tintel, M. Vanderbrugge, J. Lugtenburg, J. Cornelisse, Recl. Trav. Chim.
Pay. B. 2010, 102, 220–223.
[59] J. Arnbjerg, M. J. Paterson, C. B. Nielsen, M. Jørgensen, O. Christiansen,
P. R. Ogilby, J. Chem. Phys. A 2007, 111, 5756–5767.
[60] H. Lee, R. G. Harvey, J. Org. Chem. 1982, 47, 4364–4367.
[61] S. E. Stein, R. L. Brown, J. Am. Chem. Soc. 1991, 113, 787–793.
[62] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[63] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys. Chem.
1994, 98, 11623–11627.
[73] B. Schuler, W. Liu, A. Tkatchenko, N. Moll, G. Meyer, A. Mistry, D. Fox, L.
Gross, Phys. Rev. Lett. 2013, 111, 106103.
[64] W. J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys. 1972, 56, 2257–
2261.
Received: August 15, 2014
[65] A. A. Granovsky, Firefly version 7.1.G, www http://classic.chem.msu.su/
gran/firefly/index.html.
Published online on December 2, 2014
Chem. Eur. J. 2015, 21, 2011 – 2018
www.chemeurj.org
2018
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim