This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

Author's personal copy

Atmospheric contaminants on graphitic surfaces

David Martinez-Martin a,b, Raphael Longuinhos c, Jesus G. Izquierdo d,Antonela Marele a, Simone S. Alexandre c, Miriam Jaafar a, Jose M. Gomez-Rodrıguez a,Luis Banares d, Jose M. Soler a, Julio Gomez-Herrero a,*

a Dep. Fısica de la Materia Condensada and Condensed Matter Physics Center (IFIMAC), Univ. Autonoma de Madrid, Campus de Cantoblanco,

28049 Madrid, Spainb ETH Zurich, Department of Biosystems Science and Engineering, CH-4058 Basel, Switzerlandc Univ. Federal de Minas Gerais, Dep. Fisica, BR-30123970 Belo Horizonte, MG, Brazild Univ. Complutense de Madrid, Fac. Ciencias Quimicas, Dep. Quimica Fisica 1, E-28040 Madrid, Spain

A R T I C L E I N F O

Article history:

Received 10 October 2012

Accepted 17 April 2013

Available online 24 April 2013

A B S T R A C T

Kelvin probe force microscopy images show that the surface potential of graphite changes

with time as the contamination covers its surface. Using mass spectrometry we identify

the molecular mass of the contaminants to be compatible with that of tetracene, a polycy-

clic aromatic hydrocarbon (PAH), and its isomers. A combination of desorption and Kelvin

probe force microscopy experiments plus theoretical calculations confirms that these mol-

ecules are the main contaminant for graphitic surfaces in air ambient conditions. Interest-

ingly, when the sample temperature is increased above �50 �C the molecules are desorbed

and the surface potential becomes fairly homogeneous, suggesting that graphitic surfaces

should be almost atomically clean above this temperature. PAHs are potent atmospheric

pollutants, potentially carcinogenic, that consist of fused aromatic rings. Incomplete com-

bustion of organic materials can increase the concentration of PAHs in the atmosphere,

which in urban regions is enough to totally cover the surface of graphite in a time period

that varies from minutes to a few hours. One of the consequences of the adsorption of mol-

ecules on graphene is the doping of its surface and the variation of the charge neutrality

point originated by the charge transfer between graphene and the contamination layer.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Carbon-based materials and in particular those presenting

sp2 bonding [1–4] (graphite, graphene and carbon nanotubes)

are usually classified as chemically inert. Nevertheless, some

kind of surface contamination is expected when the material

is exposed to air ambient conditions. This matter is not a sim-

ple academic question, on the contrary it has relevant impli-

cations in technology. For instance, the lubricant properties of

graphite are known to depend critically on the presence of ad-

sorbed layers [5] and hence graphite is an excellent lubricant

in air ambient conditions but is not in vacuum.

A thorough revision of the literature shows a variety of

proposals for the possible contaminants of graphite surfaces,

including generic hydrocarbons (CxHy), metallic atoms, oxy-

gen, and sulfur [6,7]. Taking into account the highly inert nat-

ure of the graphite surface, van der Waals forces are expected

to play the most important role. Thus, although chemisorp-

tion can occur at point defects, steps, and grain boundaries,

physisorption would be the predominant adsorption mecha-

nism. Water has been also reported as a possible contaminant

for graphite in ambient air [8,9]. However, graphite is known

to be highly hydrophobic and hence it is difficult to foresee

water as a major contaminant for graphite.

0008-6223/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.carbon.2013.04.056

* Corresponding author.E-mail address: julio.gomez@uam.es (J. Gomez-Herrero).

C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9

Avai lab le a t www.sc ienced i rec t .com

journal homepage: www.elsevier .com/ locate /carbon

Author's personal copy

Since graphite is a very common substrate in scanning

probe microscopy, a technique with extreme surface sensitiv-

ity, contamination on graphite has been a topic of intense de-

bate since the very first years of scanning tunneling

microscopy (STM) [10,11]. More recently, a work by Lu et al.

[12] reported electrostatic force microscopy images showing

striking surface potential contrast on highly oriented pyro-

lytic graphite (HOPG). The authors attributed this effect to a

high electrical contact resistance between different regions

of graphite separated by defects but a later comment to this

work by Sadewasser and Glatzel [13] noticed that there were

not enough experimental proofs supporting that idea and

suggested surface contamination as a likely origin of the con-

trast. Contamination was also vaguely referenced as the ori-

gin of different electrostatics effects in graphene.

In this work, we present Kelvin probe force microscopy

(KPFM) images [14] at variable temperature, a technique closely

related to the aforementioned electrostatic force microscopy,

which show the adsorption and desorption of molecules on

graphite surfaces in real time. Using laser desorption ioniza-

tion (LDI) with reflectron geometry time-of-flight mass spec-

trometry (RTOFMS), we infer the mass of the contaminants,

which is compatible with the polycyclic aromatic hydrocarbon

(PAH) tetracene isomers. Density functional theory (DFT) calcu-

lations of tetracene physisorbed on a graphite surface confirms

the desorption energies observed in the experiments. Our re-

sults are relevant for many studies in graphitic surfaces carried

out in ambient air conditions where the effects of molecular

adsorption on sample surfaces should be taken into account

for a correct interpretation of the results.

2. Experimental

The KPFM experiments have been carried out in air, dry nitro-

gen and in a high vacuum chamber (base pressure �10�7 -

mbar) equipped with an atomic force microscope (AFM)

from Nanotec Electronica. The instrument is controlled by

the WSxM software [15]. All the data presented in this work

have been acquired in the frequency modulation and drive

amplitude modulation modes [16,17], using metalized cantile-

vers with a nominal force constant of �2.5 N/m and a reso-

nance frequency of �70 kHz for the fundamental mode. The

KPFM images were recorded by modulating the tip bias volt-

age at the resonance frequency of the second cantilever bend-

ing mode (�420 kHz). The bias voltage amplitude was kept as

low as possible (50–100 mV) to minimize possible

perturbations.

Samples of both high (ZYA) and low (ZYH) quality HOPG

(from NT-MDT and Advanced Ceramic) and natural graphite

(from Naturgraphit GmbH) with large terraces have been used

in the experiments, yielding similar results.

The mass spectrometry experiments were carried out on a

second high vacuum system (base pressure �10�8 mbar)

equipped with the LDI RTOFMS technique. LDI mass spectra

were acquired from HOPG samples using the fourth harmonic

of a Nd:YAG laser (266 nm), with a pulse width of 8 ns and a

repetition rate of 10 Hz (additional details in SI).

Additionally to the experimental results we have per-

formed density functional theory (DFT) calculations, using

the efficient implementation in the SIESTA code [18,19] of

the van der Waals density functional (vdW-DF) of Dion et al.

[20], which has already been used successfully for aromatic

molecules on graphite [21,22] (see supplementary informa-

tion for technical details).

3. Results and discussion

3.1. KPFM measurements

In the experiments the topography information is acquired

simultaneously with the KPFM images. KPFM images are

quantitative maps of the contact potential difference (CPD)

defined as CPD = (utip � usample)/e, where ux are the tip and

sample work functions respectively. Thus, sample regions

with lower surface potential are assigned to higher work

functions.

Fig. 1a–d shows four topography images acquired at room

temperature (RT) (298 K) and high vacuum (P = �10�6 mbar)

along with the simultaneously acquired KPFM maps

(Fig. 1e–h), extracted from a movie of 239 images (see video

in the supplementary material), corresponding to a total

lapsed time of 17 h. The CPD evolves rapidly during the first

3 h (the HOPG surface was cleaved at time t = 0 in ambient

air); after this time just 10% of the surface remains at the

highest potential that initially dominates the scanned area.

The last snapshot shows a surface close to the final potential

with a few regions remaining at the initial high electric poten-

tial. Fig. 1(i) shows the CPD histograms corresponding to the

images depicted in Fig. 1(e–h). Two peaks, separated by

�190 mV, are clearly visible corresponding to the two CPDs

observed in these images. It is possible to obtain an indepen-

dent confirmation of this magnitude by measuring the fre-

quency shift (proportional to the force gradient) as a

function of the tip-sample bias on both regions. Since the

force gradient between tip and sample scales with V2, the re-

sult of the experiment is a parabola. The position of the max-

imum gives the minimum electrostatic force between tip and

sample that is indeed the information obtained from KPFM

images. The voltage difference between the minima, for

parabolas taken on the bright and dark regions (see Fig. 1j),

is �200 mV, in good agreement with the value obtained from

KPFM. Importantly, the CPD variations and their geometrical

distribution are compatible to those reported in [12] where

the images where acquired in a dry nitrogen atmosphere.

A second set of experiments is presented in Fig. 2, where a

natural graphite sample, exhibiting very large terraces, is im-

aged in the same vacuum chamber (1 · 10�5 mbar) but now as

a function of temperature. In this case as well, the sample

was first cleaved in air and immediately inserted in the vac-

uum chamber. Fig. 2a shows the topography image taken at

RT, Fig. 2f is the corresponding KPFM image. The intensity

of CPD variations is similar to those reported in Fig. 1. Up to

eight different areas, enclosing islands both in the topography

and KPFM images can be observed. The height of these is-

lands in the topography is approximately 0.8 nm. As temper-

ature increases the islands tend to disappear and the regions

with higher CPD begin to grow. Finally, at 337 K CPD is almost

uniform corresponding to that of the brightest areas in Fig. 1.

34 C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9

Author's personal copy

As the sample is cooled down to RT, the regions with lower

CPD grow again.

We have confirmed the reversibility of this effect in many

different experiments with different graphite samples. If a

large surface area cold finger at liquid nitrogen temperature

is introduced into the chamber, reducing the base pressure

and increasing the pumping rate for large molecules, then

the surface remains at high CPD for much longer after cleaving

the graphite sample in high vacuum. Therefore, we conclude

that the CPD distribution observed in the HOPG samples is re-

lated to molecular adsorption on the graphite surface, as sug-

gested in [13]. Importantly, if the sample is exfoliated, exposed

to and measured in ambient air conditions for hours, we also

observe CPD distributions, but much less defined. This is prob-

ably a consequence of the adsorption of water on the cantile-

ver tip (metalized cantilever tips are usually hydrophilic) that

tends to reduce the electrostatic contrast [23]. However, if the

humidity is reduced by blowing dry nitrogen, the KPFM con-

trast evolves to that observed in vacuum, what is compatible

with that reported by Lu et al. [12].

3.2. Time of flight experiments

In order to identify the molecular species adsorbed on the

HOPG surface, we have carried out LDI RTOFMS experiments.

Fig. 3a shows a typical mass spectrum where in addition to

the intense peaks assigned to alkaline reactive metals, a clear

but less intense peak, corresponding to a molecular mass of

228 Da, is observed after introducing an air leak into the vac-

uum chamber at 1 · 10�5 mbar for about 1 h. The spectrum

was very reproducible from day to day and on several acqui-

sitions during a single day, using different HOPG samples.

This mass can be assigned to a hydrocarbon with empiric

formula C18H12. There are five PAHs compatible with this

Fig. 1 – (a–d) AFM topographies depicting the characteristic terraces of a HOPG surface. In order to enhance the terraces the

derivative of the topography is shown. (e–h) Kelvin probe microscopy maps extracted from a movie simultaneously acquired

with the topography one. The first frame is taken about 20 min after cleaving the HOPG surface with the chamber pressure at

1 · 10�5 mbar. The corresponding times for the images are 0, 4, 40 min and 15 h. The changes in the surface potential are

correlated with variations on the topographies images (see green arrows). (i) Histograms of the surface potential distribution

for figure (e–h). The peaks represent the two color regions clearly seen in the KPM images. For the case of the histogram

corresponding to figure (e) the voltage difference between peaks is 194 mV. (j) Frequency shift (force gradient) vs. tip bias

voltage taken on two different surface spots belonging to the two voltage regions observed in (e). (k) Percentage of high

voltage regions area as a function of time. (For interpretation of the references to colour in this figure legend, the reader is

referred to the web version of this article.)

C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9 35

Author's personal copy

molecular formula: tetracene, benz[a]anthracene, benzo[c]-

phenanthrene, chrysene and triphenylene (see inset in Fig. 3).

The smaller peak observed at 129 ± 1 Da is compatible with

the mass of naphthalene isomers, but the desorption tempera-

ture of physisorbed molecules with this mass is well below

room temperature (see Fig. 3b). Hence, we attribute this peak

to a fragment of tetracene isomers. In particular C10H6 has a

mass of 126 Da. Notice that a higher peak height in the mass

spectrum does not necessarily imply a higher abundance in

the sample, since it is related to a number of factors, including

the ability of the chemical species to be desorbed/ionized by the

laser pulse (see for instance Getty et al. [24]) We consider that if

alkaline metal ions are chemisorbed they should be at the sur-

face defects (steps, grain boundaries, vacancies, etc.) and that

they represent a tiny amount of the contaminants, or otherwise

graphite would not be so hydrophobic.

According to Fig. 3b, the desorption temperature of PAHs

is, to a good approximation, proportional to their molecular

weight. Small molecules such as N2, O2 or CO2, highly abun-

dant in the atmosphere, are too light to be physisorbed on

graphite at RT. In fact, the minimum molecular weight is

about that observed in the present mass spectra. While larger

molecules can also be adsorbed, their atmospheric concentra-

tion is lower. In fact, the presence of tetracene isomers in the

atmosphere is indeed very rare (few parts in billions). None-

theless, since they are the most abundant, among those com-

patible with physisorption, they finally prevail as the main

contaminant of HOPG at RT.

In an attempt to visualize directly the adsorbed molecules

on the HOPG surfaces, we have also carried out several STM

experiments, at temperatures as low as 50 K, but always with

negative results (see Supplementary information for details).

The absence of molecular resolution in the STM images sug-

gests that the contamination layers are too thick for stable

tunnelling, resulting in mechanical contact between the tip

and the sample surface.

3.3. Water adsorption

We now come back to the issue of water adsorption on graph-

ite. We have also observed, via dissipation images [28] (Fig. 4),

that the adsorption of molecules on graphite influences the

mechanical response of the surface. More specifically, there

is a one to one correspondence between low CPD and low dis-

sipation (Fig. 4b–c). Using bimodal AFM in ambient condi-

tions, Proksch [29] reported images where the mechanical

properties of a HOPG surface were also resolved at the nano-

meter scale. The geometric distribution of regions with differ-

ent mechanical contrast was very similar to the KPFM maps

shown in our Figs. 1, 2 and 4. Proksch attributed this observa-

tion to the possible adsorption of water on the sample sur-

face. However, the close correspondence between dissipation

and CPD maps, along with our LDI and DFT results, suggests

that these variations can be better explained by adsorption of

PAHs. To further confirm this hypothesis, we have also ob-

tained bimodal AFM images (Fig. S2) of the contaminated sur-

face in vacuum observing the same contrast as in Ref. [29]. In

order to further discard water as a major contaminant of

graphite, we have carried out contact angle measurements of

water droplets immediately after cleaving the graphite surface

and after 19 h in the vacuum chamber with the surface already

covered by molecules. The result is that the contact angle

(Fig. S3) is even higher for the contaminated surface than for

the clean graphite. This experimental evidence indicates that

the contamination layer is hydrophobic, as expected for PAHs,

and not hydrophilic, as expect for a water layer.

3.4. Density functional theory calculations

To crosscheck the experimental evidences for tetracene iso-

mers as the most likely contaminants of graphite, we have

carried out DFT calculations of several aromatic molecules

on graphite. We have considered different geometrical

Fig. 2 – Natural Graphite. The images show the evolution of the topography (a–e) and the surface potential (f–j) as a function of

the temperatures (298, 305, 318, 323 and 337 K, respectively). As the temperature increases the CPD becomes uniform. The

color scale for each figure has been equalized. Figure (a–e) are high pass filtered to enhance the edges.

36 C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9

Author's personal copy

arrangements with the molecules lying either horizontally

(h), i.e., parallel to the surface; vertically (v), i.e., with the long-

er axis perpendicular to the surface, in a configuration similar

to that of the tetracene solid [30]; or sideways (s), i.e., perpen-

dicular to the surface but with the longer axis parallel to it.

Obviously, we cannot rule out the possibility of other arrange-

ments with lower energy, but we can reach several safe con-

clusions from the calculations. The distance between the

carbon atoms of the adsorbed molecule and those of the

graphite layer, is �0.344 nm in the parallel geometries, some-

what smaller than in the perpendicular structures

(�0.354 nm). The calculated variation in work function dW,

relative to that obtained for clean graphite (4.3 eV) is largest

in the horizontal geometries, especially for the long tetracene

molecules (+0.13 eV (h) vs �0.05 eV (v), and �0.24 eV (s)). The

experimental variation (0.22 eV) is therefore consistent with

that of the horizontal geometry. The calculated adsorption

energy Eads, however, favors the sideways structure, but the

differences are not so large as to rule out the other geome-

tries, or a coexistence of the different orientations, depending

on the coverage. Consistently with the experimental behavior

of the desorption temperature, we find that Eads is approxi-

mately proportional to the number of carbon atoms in the

molecule (�70 meV/atom). Also, the interaction between ad-

sorbed molecules, at optimal separations (Eint ��0.1 eV, in-

cluded in Eads), suggests that the adsorbed molecules tend

to condense into ordered islands, rather than to form a homo-

geneous two dimensional phase. The theoretical results sug-

gest a complex situation where several molecules can be

coadsorbed for disordered layers.

Although the absolute values of work functions are typi-

cally underestimated somewhat in DFT calculations, this

underestimation is rather small in the case of graphite. Thus,

we obtain a value of 4.26 eV, in perfect agreement with, [31]

versus an experimental value of �4.5 eV. Furthermore, we fo-

cus on work function changes upon adsorption, which de-

pend mostly on changes of the surface dipoles, which are

generally well described in these calculations.

3.5. PAHs as atmospheric pollutants

PAHs are potent atmospheric pollutants that consist of fused

aromatic rings, which can be originated by incomplete

combustion of organic materials. Some of them are highly

carcinogenic, like benz[a]anthracene, the first carcinogenic

chemical to be discovered. The concentration of PAHs in the

atmosphere depends on a number of factors. Low molecular

weight PAHs (two and three rings) occur in the atmosphere

in the vapour phase whereas multi-ringed PAHs (five-rings)

are bound to particles. Intermediate molecular weight PAHs

(four-rings) are partitioned between the vapour and particu-

late phases, depending on atmospheric temperature [32,33].

Moreover, the atmospheric concentration of PAHs changes

Fig. 3 – (a) Time flight spectrum showing a clear peak at 228

the mass expected for tetracene isomers. The inset shows

the different isomers of tetracene. (b) Shows the variation of

the desorption temperature on graphite as a function of the

molecular mass of. The data for desorption temperatures for

Bencene, naphthalene, coronene and ovalene have been

obtained from Ref. [25], antracene Ref. [26] and pentacene

Ref. [27]. The desorption temperature for C18H12 has been

interpolated.

Fig. 4 – (a) AFM topography of a HOPG surface taken in high vacuum. (c and d) Dissipation and KPFM images of the same area

shown in (a), there is a clear correspondence between regions with low CPD and regions with low dissipation.

C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9 37

Author's personal copy

up to four orders of magnitude between rural and urban areas

with intense traffic [34]. We have carried out the experiments

in the Autonomous University of Madrid, placed �5 km away

from Madrid’s urban limit. Barrado et al. [35] reported concen-

tration of four rings PAHs ranging between 278 pg/m3 (winter)

and 132 pg/m3 (summer) in a similar area of Madrid. For these

concentrations, using the kinetic theory of gases, one can

estimate that the time required to cover a sample of graphite

with a monolayer of four-ring PAHs is in the order of 1 h, com-

patible with our observations. For the case of the experiments

performed in vacuum, the molecules that are adsorbed on the

graphite surface are in equilibrium with those present on the

vacuum chamber wall (we do no bake out the chamber) that

are in turn adsorbed from the atmosphere when the chamber

is opened. We have also observed that the time required to

cover a graphite layer is not always the same. Nevertheless,

the desorption temperature never changes.

4. Conclusions

Contamination on graphite is an issue that can easily drive to

data misinterpretation [12,13,29,36,37] due to the lack of exper-

iments. Whereas this work reports experimental and theoreti-

cal data on the molecular adsorption on graphite, we are quite

confident that other sp2 based carbon surfaces should be also

contaminated with this class of molecules [18]. Importantly,

the surface can be kept free of molecular contamination just

by increasing the temperature above 323 K. This work also sug-

gests the importance of graphitic materials as potential traps

for PAHs. In particular, carbon materials with high surface

areas such as activated carbon or, even better, carbon nanotube

forests [38], can efficiently adsorb four-ring PAHs at room tem-

perature, reducing the concentration of potential chemical car-

cinogens in the atmosphere.

Finally, the findings of this work indicate that a correct

interpretation of experimental data related with the elec-

tronic structure of graphene and nanotubes should take into

account the adsorption of molecules when the experiments

are carried out in ambient air conditions.

Acknowledgements

The authors would like to acknowledge funding by the Span-

ish MICINN Projects MAT2010-20843-C02-02 CTQ2008-02578,

MAT2010-14902 and FIS2009-12721-C04-02, Consolider Pro-

jects CSD2010-00024, CSD2007-00013, and CAM Project

S2009/MAT-1467. The EMBO Postdoctoral Fellowship ALTF

506-2012. The facilities provided by the Centro de Asistencia

a la Investigacion de Espectroscopia Multifotonica y de Fem-

tosegundos (UCM) are gratefully acknowledged.

R E F E R E N C E S

[1] Dai HJ, Hafner JH, Rinzler AG, Colbert DT, Smalley RE.Nanotubes as nanoprobes in scanning probe microscopy.Nature 1996;384:147–50.

[2] Baughman RH, Zakhidov AA, de Heer WA. Carbon nanotubes– the route toward applications. Science 2002;297:787–92.

[3] Stankovich S, Dikin DA, Dommett GHB, Kohlhaas KM,Zimney EJ, Stach EA, et al. Graphene-based compositematerials. Nature 2006;442:282–6.

[4] Geim AK. Graphene: status and prospects. Science2009;324:1530–4.

[5] Savage RH. Graphite lubrication. J Appl Phys 1948;19:1–10.[6] Metois JJ, Heyraud JC, Takeda Y. Experimental conditions to

obtain clean graphite surfaces. Thin Solid Films1978;51:105–17.

[7] Sen A, Bercaw JE. Thermal-desorption studies on hydrogen-terminated and oxygen-terminated graphite surfaces. J PhysChem 1980;84:465–6.

[8] Freund J, Halbritter J, Horber JKH. How dry are dried samples?Water adsorption measured by stm. Microsc Res Tech1999;44:327–38.

[9] Luna M, Colchero J, Gil A, Gomez-Herrero J, Baro AM.Application of non-contact scanning force microscopy to thestudy of water adsorption on graphite, gold and mica. ApplSurf Sci 2000;157:393–7.

[10] Soler JM, Baro AM, Garcia N, Rohrer H. Interatomic forces inscanning tunneling microscopy – giant corrugations of thegraphite surface. Phys Rev Lett 1986;57:444–7.

[11] Mamin HJ, Ganz E, Abraham DW, Thomson RE, Clarke J.Contamination-mediated deformation of graphite by thescanning tunneling microscope. Phys Rev B 1986;34:9015–8.

[12] Lu YH, Munoz M, Steplecaru CS, Hao C, Bai M, Garcia N, et al.Electrostatic force microscopy on oriented graphite surfaces:coexistence of insulating and conducting behaviors. Phys RevLett 2006;97:076805–9.

[13] Sadewasser S, Glatzel T. Comment on ‘‘electrostatic forcemicroscopy on oriented graphite surfaces: coexistence ofinsulating and conducting behaviors’’. Phys Rev Lett2007;98:269701.

[14] Nonnenmacher M, Oboyle MP, Wickramasinghe HK. Kelvinprobe force microscopy. Appl Phys Lett 1991;58:2921–3.

[15] Horcas I, Fernandez R, Gomez-Rodriguez JM, Colchero J,Gomez-Herrero J, Baro AM. WSxM: a software for scanningprobe microscopy and a tool for nanotechnology. Rev SciInstrum 2007;78:013705–13.

[16] Albrecht TR, Grutter P, Horne D, Rugar D. Frequency-modulation detection using high-q cantilevers for enhancedforce microscope sensitivity. J Appl Phys 1991;69:668–73.

[17] Jaafar M, Martinez-Martin D, Cuenca M, Melcher J, Raman A,Gomez-Herrero J. Drive-amplitude-modulation atomic forcemicroscopy: from vacuum to liquids. Beilstein J Nanotechnol2012;3:336–44.

[18] Ordejon P, Artacho E, Soler JM. Self-consistent order-ndensity-functional calculations for very large systems. PhysRev B 1996;53:R10441–4.

[19] Soler JM, Artacho E, Gale JD, Garcia A, Junquera J, Ordejon P,et al. The siesta method for ab initio order-n materialssimulation. J Phys Condens Matter 2002;14:2745–79.

[20] Roman-Perez G, Soler JM. Efficient implementation of avan der waals density functional: application todouble-wall carbon nanotubes. Phys Rev Lett2009;103:096102–96106.

[21] Chakarova-Kack SD, Schroder E, Lundqvist BI, Langreth DC.Application of van der waals density functional to anextended system: adsorption of benzene and naphthalene ongraphite. Phys Rev Lett 2006;96:013017–13033.

[22] Chakarova-Kack SD, Vojvodic A, Kleis J, Hyldgaard P, SchroderE. Binding of polycyclic aromatic hydrocarbons and graphenedimers in density functional theory. New J Phys2010;12:013017–13033.

[23] Sugimura H, Ishida Y, Hayashi K, Takai O, Nakagiri N.Potential shielding by the surface water layer in kelvin probeforce microscopy. Appl Phys Lett 2002;80:1459–61.

[24] Getty SA, Brinckerhoff WB, Cornish T, Ecelberger S, Floyd M.Compact two-step laser time-of-flight mass spectrometer for

38 C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9

Author's personal copy

in situ analyses of aromatic organics on planetary missions.Rapid Commun Mass Spectrom 2012;26:2786–90.

[25] Zacharia R, Ulbricht H, Hertel T. Interlayer cohesive energy ofgraphite from thermal desorption of polyaromatichydrocarbons. Phys Rev B 2004;69:155406–13.

[26] Bussolotti F, Han SW, Honda Y, Friedlein R. Phase-dependentelectronic properties of monolayer and multilayeranthracene films on graphite [0001] surfaces. Phys Rev B2009;79:245410–7.

[27] Goetzen J, Kaefer D, Woell C, Witte G. Growth and structure ofpentacene films on graphite: Weak adhesion as a key forepitaxial film growth. Phys Rev B 2010;81:085440–52.

[28] Luthi R, Meyer E, Bammerlin M, Baratoff A, Howald L, GerberC, et al. Ultrahigh vacuum atomic force microscopy: Trueatomic resolution. Surf Rev Lett 1997;4:1025–9.

[29] Proksch R. Multifrequency, repulsive-mode amplitude-modulated atomic force microscopy. Appl Phys Lett2006;89:015501–12.

[30] Burns DM, Iball J. Refinement of the structure of chrysene.Proc R Soc London Ser A 1960;257:491–514.

[31] Chan KT, Neaton JB, Cohen ML. First-principles study ofmetal adatom adsorption on graphene. Phys Rev B2008;77:235430–42.

[32] Howsam M, Jones KC, Ineson P. Pahs associated with theleaves of three deciduous tree species. I – concentrations andprofiles. Environ Pollut 2000;108:413–24.

[33] Howsam M, Jones KC, Ineson P. Pahs associated with theleaves of three deciduous tree species. II: Uptake during agrowing season. Chemosphere 2001;44:155–64.

[34] Srogi K. Monitoring of environmental exposure to polycyclicaromatic hydrocarbons: A review. Environ Chem Lett2007;5:169–95.

[35] Barrado AI, Garcia S, Barrado E, Perez RM. Pm2.5-bound pahsand hydroxy-pahs in atmospheric aerosol samples:Correlations with season and with physical and chemicalfactors. Atmos Environ 2012;49:224–32.

[36] Dan YP, Lu Y, Kybert NJ, Luo ZT, Johnson ATC. Intrinsicresponse of graphene vapor sensors. Nano Lett2009;9:1472–5.

[37] Schedin F, Geim AK, Morozov SV, Hill EW, Blake P, KatsnelsonMI, et al. Detection of individual gas molecules adsorbed ongraphene. Nat Mater 2007;6:652–5.

[38] Hata K, Futaba DN, Mizuno K, Namai T, Yumura M, Iijima S.Water-assisted highly efficient synthesis of impurity-freesingle-waited carbon nanotubes. Science2004;306:1362–4.

C A R B O N 6 1 ( 2 0 1 3 ) 3 3 – 3 9 39