Oxygen etching mechanism in carbon-nitrogen (CNx) domelikenanostructuresJ. J. Acuña, C. A. Figueroa, D. Biggemann, M. U. Kleinke, and F. Alvarez Citation: J. Appl. Phys. 103, 124907 (2008); doi: 10.1063/1.2948941 View online: http://dx.doi.org/10.1063/1.2948941 View Table of Contents: http://jap.aip.org/resource/1/JAPIAU/v103/i12 Published by the AIP Publishing LLC. Additional information on J. Appl. Phys.Journal Homepage: http://jap.aip.org/ Journal Information: http://jap.aip.org/about/about_the_journal Top downloads: http://jap.aip.org/features/most_downloaded Information for Authors: http://jap.aip.org/authors

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions

Oxygen etching mechanism in carbon-nitrogen „CNx… domelikenanostructures

J. J. S. Acuña,1 C. A. Figueroa,1 D. Biggemann,2 M. U. Kleinke,1 and F. Alvarez1,a�

1Instituto de Física “Gleb Wataghin,” Unicamp, C.P. 6165, 13083-970 Campinas, SP, Brazil2Laboratório Nacional de Luz Síncrotron, C.P. 6192, 13084-971 Campinas, SP, Brazil

�Received 3 March 2008; accepted 18 April 2008; published online 24 June 2008�

We report a comprehensive study involving the ion beam oxygen etching purification mechanism ofdomelike carbon nanostructures containing nitrogen. The CNx nanodomes were prepared on Sisubstrate containing nanometric nickel islands catalyzed by ion beam sputtering of a carbon targetand assisting the deposition by a second nitrogen ion gun. After preparation, the samples wereirradiated in situ by a low energy ion beam oxygen source and its effects on the nanostructures werestudied by x-ray photoelectron spectroscopy in an attached ultrahigh vacuum chamber, i.e., withoutatmospheric contamination. The influence of the etching process on the morphology of the samplesand structures was studied by atomic force microscopy and field emission gun–secondary electronmicroscopy, respectively. Also, the nanodomes were observed by high resolution transmissionelectron microscopy. The oxygen atoms preferentially bond to carbon atoms by forming terminalcarbonyl groups in the most reactive parts of the nanostructures. After the irradiation, the remainingnanostructures are grouped around two well-defined size distributions. Subsequent annealingeliminates volatile oxygen compounds retained at the surface. The oxygen ions mainly react withnitrogen atoms located in pyridinelike structures. © 2008 American Institute of Physics.�DOI: 10.1063/1.2948941�

I. INTRODUCTION

The discovery of different allotropic forms of carbon hasopened new applications of this versatile element.1–3 Amongthe variety of studied carbon nanostructures, nanotubes�NTs�, nano-onions, and nanodomes �NDs� are of interest forapplications such as field emitting devices.4,5 The electronicproperties of these structures can be doped to control photo-emission properties. Indeed, nitrogen incorporation is ex-pected to change the electronic properties of the material byacting as a donor impurity when incorporated into a graphi-ticlike structure.4,6 Nevertheless, the undesirable amorphouscompounds and defects introduced by nitrogen in the synthe-sis process degrade the morphology behavior of thenanostructures.7 Therefore, in order to improve their proper-ties, a later treatment is normally employed, such as oxygenpurification. In standard procedures applied to NTs, amor-phous carbon phases �a-C� present in the raw material areeliminated by oxidation in air or pure oxygen at 750 °C.8–10

The purifications take place because, thermodynamically, theorganized structures such as fullerenes, NDs, NTs, nano-onions, and nanohorns containing graphitic planes are morestable than a-C and defective structures.11,12 In spite of thebroadly used oxygen etching purification process, the mecha-nisms leading to a purer material are not fully understood.Consequently, the study of the microscopic mechanisms in-volved in oxygen ion beam irradiation of CNx nanostructurescould lead to improving the understanding of the process andcontrolling the final morphology and electronic properties of

the material.13,14 This is so because the ion energy, beamcomposition, and oxygen doses are accurately controlled inbroad ion beam sources �Kaufman cells�.15

In this paper we report a comprehensive study of the ionoxygen chemical etching mechanisms occurring in CNX NDsobtained by ion beam assisted deposition �IBAD�. The evo-lution of the physical-chemical properties of the nanostruc-tures after in situ oxygen ion beam irradiation was studied byx-ray photoelectron spectroscopy �XPS� in a ultra highvacuum �UHV� chamber attached to the depositionchamber.16 Afterward, ex situ atomic force microscopy�AFM�, field emission gun–scanning electron microscopy�FEG-SEM�, and high resolution transmission electron mi-croscopy �HRTEM� were performed in order to studychanges in the morphological and structural characteristics ofthe material.

II. EXPERIMENTAL

The CNx NDs were prepared by IBAD on nickel islandspreviously deposited on Si �100� p-doped ��15 � /cm2� mir-ror polished substrates �10�10 mm2�, 1 mm thick. Detailsof the deposition chamber are reported elsewhere.15 A SiO2

layer of �600 nm was grown before the nickel deposition bywet oxidation at 1050 °C on the Si wafers to act as a barrierto prevent the formation of nickel silicide. The nickel islands��1–5 nm size� were grown on the substrates by in situsputtering of a Ni �99.9995%� target and subsequent 3 minannealing. The annealing treatment induces the nickel par-ticles to coalesce into tiny dots �islands�. Details of the Niisland preparation were reported elsewhere.17 The CNx nano-structures were deposited by sputtering a high purity graphite�99.9995%� target �Ar+, 1450 eV, and 13 mA /cm2�. Simul-a�Electronic mail: alvarez@ifi.unicamp.br.

JOURNAL OF APPLIED PHYSICS 103, 124907 �2008�

0021-8979/2008/103�12�/124907/6/$23.00 © 2008 American Institute of Physics103, 124907-1

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions

taneously, a nitrogen beam assists the growth �8 eV,0.2 mA /cm2�. The substrate temperature during the nano-structure growth was maintained at �650 °C for the wholeprocess, i.e., from the nickel deposition to CNx formation.The chamber base pressure was �8�10−7 Pa, and the depo-sition pressure maintained at �3�10−4 Pa for the nickel andthe CNx films deposition.

An oxygen ion beam from a direct current Kaufmansource, 3 cm diameter, was employed to perform in situ theroom temperature purification of the CNx nanostructures,employing different oxidation times. The nominal oxygenion energy and current-density beam were fixed at 10 eV and4.2 mA /cm2, respectively. In order to eliminate volatilecompounds, samples were heated at �400 °C for 10 min.

First, the samples were analyzed by XPS before and af-ter the oxygen etching, without any annealing treatment. TheXPS spectra were obtained using the Al K� line �1486.6 eV�and a VG-CLAMP-2 electron cylindrical analyzer.15 The to-tal apparatus resolution was mainly due to x-ray linewidth��0.85 eV�. The relative atomic composition on the samplesurfaces was determined by the standard procedure of inte-grating the corresponding core level peak elements, properlyweighted by the photoemission cross section.18 The inelasticscattering spectra backgrounds were subtracted using Shirl-ey’s method.19 The spectra were deconvoluted by a standardmultiples 50%–50% Gaussian–Lorentzian peak fitting proce-dure. For further characterization of the nanostructures afterannealing, AFM �Topometrix TMX 2000 SPM�, SEM-FEG�JEOL JSM6330F�, and cross section HRTEM �JEOL JEM-3010 URP� measurements with a LaB6 emission gun operat-ing at 300 keV were performed.

III. RESULTS

A. Morphological and structural results

In previous works we have reported the deposition ofcarbon nanostructures coating nickel particles�“domes”�.16,17,20 Figure 1 shows a typical image obtained byHRTEM of similar structures to those that were studied inthis paper. Figure 1 shows the micrograph of a typical irra-diated sample. Figure 2 shows the FEG-SEM images of atypical CNx deposited sample obtained in the secondary elec-tron mode, before �top� and after oxygen etching �bottom�.The two pictures were obtained in similar conditions. FromFig. 2 �top�, it was possible to determine that the particlessizes span from �40 to 150 nm. The poor contrast in thispicture indicates low peak-valley ratio and quite uniform ma-terial. In comparison, 90 s oxidation irradiation notably in-creases contrast, suggesting larger peak-valley ratio and dif-ferent material covering the particles and the valleys �Fig. 2,bottom�.

The oxygen irradiation effect on the sample topology iswell observed by AFM measurements �Fig. 3�. The picturesclearly display the peak-valley ratio enhancement on oxygentime irradiation. Due to the presence of amorphous carbon,nonpurified samples present smoother surfaces. On the con-trary, after oxygen etching the roughness of the samples in-

creases from �3.9 to 5.2 nm rms since the amorphous ma-terial occupying the sites between Ni islands is in parteliminated.

B. XPS results

For the x-ray photon energy used in the experimentsreported in this paper, an �50 Å material depth is probed bythe photoemitted electrons.18 Therefore, XPS is a good toolto quantify oxygen etching effect on the material surface. Asremarked above, the annealing effects eliminate the volatileoxygen compounds obtained after the irradiation procedure.Therefore, we shall proceed to discuss the results obtained inthe irradiated samples before and after annealing.

Figure 4 shows the atomic content of C, N, and O as afunction of the oxidation time, before and after annealingprocess. The decreasing �increasing� carbon �oxygen� as afunction of the oxidation time indicates the formation of sur-face oxide compounds. On the other hand, the nitrogen con-tent maintains a quite constant value of �3.1 at. %, suggest-ing that oxygen is weakly interacting with nitrogen. Figures5�a�–5�c� show the binding energy �BE� bands associatedwith the C 1s, N 1s, and O 1s core level electrons before andafter the thermal annealing of a typical sample, oxygen irra-diated for 90 s. Before the annealing process the effects ofthe oxygen on the structures are the following: �1� the C 1sand N 1s bands shift to higher binding energies; �2� the O 1sband increases its intensity with a remarkable contribution at531.6 eV on higher oxidation doses.

After the annealing, the linewidth of the band associatedwith the electron C 1s is narrower �Fig. 5�a��. On the otherhand, the N 1s and O 1s bands show two well-defined peaks,indicating that the annealing process enhances two preferen-tial environments for the N atoms �see Figs. 5�b� and 5�c��.

FIG. 1. Cross section HRTEM image of a typical nickel particle covered bygraphene planes after oxygen radiation �52 s of ion oxygen etching�.

124907-2 Acuña et al. J. Appl. Phys. 103, 124907 �2008�

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions

IV. DISCUSSION

A. Morphological and structural characteristics

As remarked in Sec. III A, the AFM pictures shows thatthe peak-valley ratio increases on oxygen irradiation, indicat-ing that defective �amorphous� carbon phases are being re-moved. The aspect of the domes is also increased, as ob-served in Fig. 3 for increasing oxidation times.

Figure 6 shows the top view size distribution of fourrepresentative samples series obtained from FEG-SEM im-ages at different oxidation times. The histogram analysisshows that the left side of the distribution diminishes onoxygen irradiation. Indeed, after 52 s oxidation a quite well-defined two-size distribution increases centered at �80 and�130 nm, respectively. Further irradiation increases thepeak-valley ratio helping to resolve quite well the dome dis-tribution. The sharper structures observed by AFM imagesreinforce these results since after even short irradiationtimes, a much better definition of the NDs is observed �Fig.3�.

In order to quantify the effect of the oxidation-time irra-diation on the ND sizes, the experimental histograms weredeconvoluted. The procedure confirms that two componentsare enough to obtain a very good fitting in all the studied

samples. As an example, Fig. 7 �inset� shows the fitting forthe pristine sample. Figure 7 shows the evolution of the twocomponents’ separation distance as a function of the irradia-tion time.

Therefore, one can conclude that the oxidation processspecifically attacks the most reactive parts of the dome struc-tures creating two different types of domes. Indeed, a non-specific oxidation process �the same chemical reactivity�should yield a unique size dome distribution.

FIG. 2. FEG-SEM images of the CNx nanostructures �domelike�. Top: Pris-tine sample. Bottom: 90 s of oxygen plasma etching.

FIG. 3. AFM images of the NDs distribution: �a� pristine sample, �b� 16 s,�c� 52 s, and �d� 90 s oxidation time.

124907-3 Acuña et al. J. Appl. Phys. 103, 124907 �2008�

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions

B. Chemical reactivity

Table I summarizes the most probable chemical constitu-ents compatible with the XPS spectra. PC1 is the band asso-ciated with photoelectrons involved in pure graphitic struc-

ture, i.e., aromatic CuC bonds.21 The CuN bonds aregenerally associated with three principal complex groups:first, aromatic structures containing C atoms bonded to oneN �PC2�, second, nonaromatic structures containing C atomsbonded to one N and sp3 C bonded to one N �PC3�, andthird, aromatic or nonaromatic C atoms bonded to two N�PC4�.15,22 The PC4 band is controversial, and sometimes it isascribed as related to simple and double carbon-oxygenbonds uCuOu and uCOOu �located at 286.6 and288.2 eV, respectively�.23,24 Finally, a contribution associ-ated with the energy loss photoelectrons due to collectiveexcitation of valence electrons, i.e., �-plasmon �PB1�, is alsoindicated. However, this structure is sometimes associatedwith shake up transitions ��→�*� characteristic in aromaticcompounds.25

The band associated with the N 1s core level electronswas deconvoluted by considering two main contributions.One contribution is associated with N atoms bonded to C innonaromatic environment, terminal N, and N bridges �la-beled PN1�. The second contribution is associated with N

FIG. 7. Separation distance between the two components obtained by de-convoluting the histogram from Fig. 6 as a function of the oxidation irra-diation time. The dash line is a guide for the eyes. Inset: Deconvolution ofthe pristine sample.

FIG. 4. Atomic content of the topmost sample layer as a function of theoxidation time studied by XPS. Filled symbols: Before annealing. OpenSymbols: After annealing. The analyzed elements are indicated in the plot.Symbols corresponding to N are overlapping. The dash lines are guides forthe eyes.

FIG. 5. ��a�–�c�� Spectral band associated with the C 1s, N 1s, and O 1selectron core levels before and after oxygen irradiation �90 s� in the oxygen-irradiated experiments. Solid line: Pristine sample; dash line and dash-dot-dot lines for the oxidized sample; before and after annealing process,respectively.

FIG. 6. Distribution �in nm� obtained from the analysis of FEG-SEM im-ages. �a�–�d� are the pristine, 8 s, 52 s, and 90 s oxygen irradiated samples.

124907-4 Acuña et al. J. Appl. Phys. 103, 124907 �2008�

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions

atoms in aromatic ring sites �labeled PN2�.21,26–30 Finally, inthe case of the spectra associated with O 1s core level elec-trons, the band was deconvoluted in two contributions: Oatoms associated with C in double and single bonds, labeledPO1 and PO2, respectively. Due to the low N concentration��3 at. % � in the domes, NuO contributions were ex-cluded.

The relative evolution of the different ratio band contri-butions to the XPS spectra is particularly enlightening inunderstanding the oxygen etching mechanism. This is so be-cause the energy shifting of the BEs associated with differentcore level electrons’ contributions to the spectra are inti-mately associated with the components’ chemical environ-ments. Figures 8–10 show the relative contribution evolutionassociated with the bands from Table I as a function of theirradiation time, before and after annealing the samples.

The observed PO2 /PO1 ratio decreasing on oxygen irra-diation before sample annealing indicates that the O atomson the surfaces are preferentially forming C double bonds�Fig. 8, top�. This is understood since, from a kinetic point ofview, it is easier to form a CvO group terminal group�PO1� than two single bonds in an arrangement of threeatomic centers �PO2�.31 After the annealing process, the satu-ration limit for O content is reduced from �16 to 6.5 at. %.

Moreover, the PO2 /PO1 ratio remains constant, suggest-ing that terminal CvO is preferentially eliminated from thesample surface during the annealing treatment. Moreover,annealing induces reagreements of atoms by eliminatingvolatile species such as CO and CO2. Figure 9 shows theevolution of the PC4 /PC1 ratios as a function of oxidationirradiation time. We point out that the observed BE shift tohigher energies �see Fig. 5�a�� of the C 1s band is related toCuO bonds, i.e., the contribution to the spectra from PC4

species. In agreement with O 1s band evolution, the aug-menting of the PC4 /PC1 ratio is related to C atoms doublybonded to O atoms �CvO group�. The observed constancyof the PC4 /PC1 ratio after annealing suggests that the com-ponents associated with OuC bonds were eliminated, i.e.,the contributions such as pure aromatic CuC bonds and Cbonded to N do not present appreciable changes due to oxy-gen etching. In summary, these results suggest that C atomsare essentially oxidized as terminal carbonyl.

The analysis of the BE associated with the N 1s electronis very important in the interpretation of the oxygen etchingmechanism. Figure 10 shows the evolution of the PN2 /PN1

TABLE I. Summary of the different contributions involved in the XPSspectra deconvolution.

Band BE �eV� Chemical bonding

PC1 284.3�0.1 Aromatic CuCPC2 285.2�0.1 Aromatic C3N structuresPC3 286.1�0.1 Nonaromatic C3N phasePC4 287.4�0.1 Nonaromatic C2N2, uCvO, uCuOu, and

uCOOu

PB1 290.0�0.1 Plasmons and/or shake upPN1 398.3�0.5 Nonaromatic N–sp2CPN2 400.4�0.5 N aromatic clusters and NwNPO1 531.6�0.5 uCvOPO2 533.1�0.5 uCuOu

FIG. 8. Evolution of the PO2 /PO1 ration as a function of the oxidation time.The dash line is a guide for the eyes.

FIG. 9. Evolution of the PC4 /PC1 ration as a function of the oxidation time.The dash line is a guide for the eyes.

FIG. 10. Evolution of the PN2 /PN1 ratio as a function of the oxidation time.The dash line is a guide for the eyes

124907-5 Acuña et al. J. Appl. Phys. 103, 124907 �2008�

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions

ratios as a function of oxidation irradiation time. After an-nealing, the content of N atoms bonded to aromatic clusterand NwN groups decreases due to oxygen etching. We notethat the PN2 band is located at higher BE than the PN1 bandcontribution. This chemical shift indicates that N atoms con-tributing to PN2 have indeed lost electron density, i.e., theirlocal environment is more positive.18 Therefore, this situa-tion increases the N chemical reactivity with O, a strongelectronegative atom, i.e., N bonded in aromatic and py-ridinelike structures �PN2� is preferentially etched by O. Wesuggest, however, that O mainly destroys pyridinelike struc-tures since N in aromatic rings is linked �besides relativelyweak �-bonds� by stronger � bonds.

From the above analysis, the following mechanism oc-curring during the oxygen etching is proposed. First, the O2

+

ions landing on the domes have enough energy as to disso-ciate ��10 eV� in atomic oxygen. Therefore, the O atomsform bonds with the more reactive part of the ND structure.The existence of two types of dome sizes after this step sug-gests a preferential site for O attachment, such as terminalvolatile carbonyl groups �CO and CO2� subsequently elimi-nated by the annealing treatment. Also, oxygen atoms pref-erentially react with N bonded in pyridinelike structures.

V. CONCLUSIONS

The oxygen etching mechanism of grapheme nanostruc-tures �domes� containing nitrogen was studied by ion beam�O2

+� irradiation. The chemical evolution of the species in-volved in the process was analyzed by photoemission elec-tron spectroscopy �XPS�. The morphological and structuralanalysis of the nanostructures was studied by HRTEM, AFM,and FEG-SEM.

The AFM images show the evolution of the nanostruc-tures on oxygen etching. Two types of ND distribution sizesare obtained following the original two-size nickel originaldistribution of catalyzer metal used to grow the nanostruc-tures. HRTEM images show that the graphene layers coatingthe nickel particles maintain the original structure after oxi-dation. This suggests that oxygen preferentially etches car-bon atoms forming terminal carbonyl groups on the mostreactive parts of the material �CO and CO2�. Compoundscontaining nitrogen atoms bonded to aromatic and pyridinestructures are etched by oxygen. We suggest, however, thatthe pyridine groups are more reactive and are preferentiallyeliminated than stable aromatic benzene rings. This is alsoshown by the AFM images where the amorphous intervalleyamorphous material is eliminated by oxygen irradiation andposterior annealing.

ACKNOWLEDGMENTS

This work is part of the Ph.D. thesis of J.J.S.A. and waspartially sponsored by Fapesp, Project No. 97/12069-0.

J.J.S.A. and F.A. are CNPq fellows. C.A.F. is a FAPESPfellow �Project No. 04/01977-9�. At present, C.A.F. is at theCCET, UCS, 95070-560, Caxias do Sul-RS, Brazil. The elec-tron microscopy work has been performed with the JEOLJSM6330F �SEM-FEG� and JEOL JEM-3010 URP �HR-TEM� microscopes of the LME/LNLS, Campinas, SP, Brazil.

1H. W. Kroto, J. R. Heath, S. C. O’Brien, R. F. Curl, and R. E. Smalley,Nature �London� 318, 162 �1985�.

2S. Iijima, Nature �London� 354, 56 �1991�.3D. Ugarte, Nature �London� 359, 707 �1992�.4P. Paredez, C. A. Figueroa, L. F. Zagonel, F. R. Reichert, C. R. M. Ribeiro,S. Point, C. Godon, T. M. Minea, and F. Alvarez, J. Nanosci. Nanotechnol.2, 188 �2005�.

5W. A. de Hexer, A. Châtelain, and D. Ugarte, Science 270, 1179 �1995�.6J. J. Li, W. T. Zheng, Z. S. Jin, X. Wang, H. J. Bian, G. R. Gu, Y. N. Zhao,S. H. Meng, X. D. He, and J. C. Han, J. Vac. Sci. Technol. B 21, 2382�2003�.

7B. S. Satyanarayana, A. Hart, W. I. Milne, and J. Robertson, Appl. Phys.Lett. 71, 1430 �1997�.

8T. W. Ebbesen, P. M. Ajayan, H. Hiura, and K. Tanigaki, Nature �London�367, 519 �1994�.

9K. Tohji, T. Goto, H. Takahashi, Y. Shinoda, N. Shimizu, B. Jeyadevan,and I. Matsuoka, Nature �London� 383, 679 �1996�.

10R. Droppa, Jr., C. T. M. Ribeiro, R. A. Zanatta, M. C. dos Santos, and F.Alvarez, Phys. Rev. B 69, 045405 �2004�.

11X. Y. Zhu, S. L. Lee, Y. H. Lee, and Thomas Frauenheim, Phys. Rev. Lett.85, 2757 �2000�.

12M. G. Fyta, I. N. Remediakis, P. C. Kelires, Phys. Rev. B 67, 035423�2003�.

13H. T. Fang, C. G. Liu, L. Chang, L. Feng, L. Min, H. M. Cheng, Chem.Mater. 16, 5744 �2004�.

14Y. Lui, L. Lui, P. Lui, L. Sheng, and S. Fan, Diamond Relat. Mater. 13,1609 �2004�.

15P. Hammer, N. M. Victoria, and F. Alvarez, J. Vac. Sci. Technol. A 16,2491 �1998�.

16J. J. S. Acuña, C. A. Figueroa, M. E. H. Maia da Costa, P. Paredez, C. T.M. Ribeiro, and F. Alvarez, J. Non-Cryst. Solids 352, 1314 �2006�.

17P. Paredez, M. C. Marchi, M. E. H. Maia da Costa, C. A. Figueroa, M. U.Klienke, C. T. M. Ribeiro, J. C. Sánchez-López, T. C. Rojas, and F. Al-varez, J. Non-Cryst. Solids 352, 1303 �2006�.

18J. F. Moulder, W. F. Stickle, P. E. Sobol, and K. D. Bomben, in Handbookof X-ray Photoelectron Spectroscopy, edited by J. Chastain, �Perkin-Elmer,Minnesota, USA, 1992�, p. 25.

19D. A. Shirley, Phys. Rev. B 5, 4709 �1972�.20P. Paredez, M. E. H. Maia da Costa, L. F. Zagonel, C. T. M. Ribeiro, and

F. Alvarez, Carbon 45, 2678 �2007�.21P. Hammer, R. G. Lacerda, R. Droppa, Jr., and F. Alvarez, Diamond Relat.

Mater. 9, 577 �2000�.22D. Marton, K. J. Boyd, and J. W. Rabalais, Int. J. Mod. Phys. B 9, 3527

�1995�.23Y. Xing, L. Li, C. C. Chusuei, and R. V. Hull, Langmuir 21, 4185 �2005�.24W. H. Lee, J. G. Lee, and P. J. Reucroft, Appl. Surf. Sci. 171, 136 �2001�.25D. Briggs and M. P. Seah, Practical Surface Analysis, 2nd ed. �Willey,

New York, 1993�, Vol. 1.26J. C. Sánchez-López, C. Donnet, F. Lefèbvre, C. Fernández-Ramos, and

A. Fernández, J. Appl. Phys. 90, 675 �2001�.27H. Sjöström, S. Stafström, M. Boman, and J. E. Sundgren, Phys. Rev. Lett.

75, 1336 �1995�.28S. Souto and F. Alvarez, Appl. Phys. Lett. 70, 1539 �1997�.29C. Speath, M. Kühn, F. Riehter, U. Fa1ke, M. Mietsehold, R. Kilper, and

U. Kreissig, Diamond Relat. Mater. 7, 1727 �1998�.30A. Mansour and D. Ugolini, Phys. Rev. B 47, 10 201 �1993�.31Y.-R. Luo, Handbook of Bond Dissociation Energies in Organic Com-

pounds �CRC, Boca Raton, FL, 2003�, pp. 223–224.

124907-6 Acuña et al. J. Appl. Phys. 103, 124907 �2008�

Downloaded 17 Jul 2013 to 143.106.108.110. This article is copyrighted as indicated in the abstract. Reuse of AIP content is subject to the terms at: http://jap.aip.org/about/rights_and_permissions