Letter pubs.acs.org/JPCL
Resonance Charges to Encode Selection Rules in Inelastic Electron Tunneling Spectroscopy Shiri R. Burema and Marie-Laure Bocquet* Ecole Normale Supérieure de Lyon, Laboratoire de Chimie, CNRS UMR 5182, 46 Allée d’Italie, 69364 CEDEX07 Lyon, France S Supporting Information *
ABSTRACT: From extensive simulations of a set of covalently grafted phenyl derivatives onto Cu(111), we derive a simplistic rule that selectively predicts the onset of stretching vibrations in inelastic electron tunneling spectroscopy (IETS) with the scanning tunneling microscope. Indeed the rise (extinction) of the highest-frequency modes is found to correlate to the accumulation (depletion) of π electron density at the metal−organic contact point. This π electron density can be fine-tuned by the usage of (de) activating aromatic substituent at different ring positions. This finding provides a simple analysis tool that can be used to reveal structural characteristics on the atomic scale by IETS.
SECTION: Surfaces, Interfaces, Porous Materials, and Catalysis
O
Some of the detected IETS signatures involve only out-ofplane (OOP) wagging modes in the low-frequency region.20 Although these may be sufficient to characterize the adsorbed molecule (especially in conjunction with STM imaging), they do not qualify as an unique molecular footprint as highfrequency stretching vibrations do. It is therefore important to favor the detection of such modes by IETS. However, this is not straightforward because the current selection rules contain many intricate factors at play, like the frontier orbital control and the symmetry matching between electronic states and nuclear movements. In this letter we propose a charge descriptor to account for an efficient activation of the high-frequency stretching modes in IETS for a family of phenyl moieties grafted onto a metal surface. For this type of metal−organic junction, we show that the prevailing ingredient is a negative net charge of π electron density (even slightly) at the point, where the organic molecule binds covalently the metal surface. Additionally, we demonstrate that the contrary also holds true: a (small) positive polarization of the π system at the anchoring site totally quenches all stretching vibrations in the IETS spectrum. The above correlation is based on exhaustive ab initio investigations (see Computational Section for details) of a family of monosubstituted phenyl-compounds adsorbed in an upright position on the hollow sites of a Cu(111) surface. In this adsorption configuration (called head-up), the phenyl radical binds after H abstraction from benzene (or I abstraction if iodobenzene is taken as a precursor).24 As has been shown
rganic molecules adsorbed on metal surfaces are of great interest for a multitude of applications that impact all domains of chemistry. Examples range from heterogeneous catalysts1 for synthesis, molecular electronics,2 and nanodevices3 in physical chemistry to bases for biochemical molecular recognition.4 It is essential to characterize these systems finely to ensure their proper functioning. Traditionally, the scanning tunneling microscope (STM) has been used to obtain an image on the atomic scale of such chemisorbed species.5 Nevertheless, this technique often lacks the necessary resolution for an definite identification. The use of the STM probe to perform inelastic electron tunneling spectroscopy (IETS)6,7 is now emerging as a complementary approach that holds great promise for its discriminative potential.8−10 Because in IETS the tunneling electrons from the STM tip stimulate vibrational excitations in the adsorbed system when the tunneling bias voltage surpasses the ℏω threshold energy, it yields knowledge of the active modes and their energies. These are unique fingerprints of the adsorbate’s adsorption site, orientation, and chemical nature, forming all together an accurate basis for a detailed structural identification,11 especially in conjunction with dedicated theory approaches.12−17 Indeed, examples exist in the literature where IETS successfully differentiated similar molecules being adsorbed with distinct denticity (e.g., phenyl and benzyl13) or hybridization (e.g., ethylene and acetylene18 or propene, propadiene, and propyne19), with opposite orientations (e.g., nitrobenzene aligned vertically and pointing up- and downward relative to the surface20 or aligned horizontally and parallel to the surface21) or varying configurations (e.g., fullereneterminated molecules22), and being subjected to heteroatom substitution (e.g., N versus S containing five-membered heterocycles,23 or nitrobenzene versus benzoate20). © XXXX American Chemical Society
Received: August 28, 2012 Accepted: October 3, 2012
3007
dx.doi.org/10.1021/jz3012832 | J. Phys. Chem. Lett. 2012, 3, 3007−3011
The Journal of Physical Chemistry Letters
Letter
elsewhere,20 head-up adducts of functionalized phenyl are equally stable, regardless of the substituent, with adsorption energies around −2.5 eV, preventing any diffusion or desorption under the application of a voltage ramp. Whereas the direct attachment of the function (called head-down) can sometimes be energetically preferred over these head-up forms, it generally leads to versatile bound species with various adsorption energies, which complicates their spectroscopic comparison. Hence, we only consider the isoenergetic series of grafted fragments shown in Figure 1.
Figure 1. Investigated monosubstituted phenyl compounds/Cu(111) in this study. The π electronic density at the metal−organic contact ranges from excess (left) to depletion (right), as guessed by mesomery.
Using conventional electrophilic aromatic substitution rules for the (de)activating effect of substituents, we can easily sketch four possible cases to tune the π electron density at the anchoring carbon atom. These are schematically depicted in decreasing order in Figure 1. For an electron-donating group (L) as −NH2 placed in para position to the surface, resonance effects yield the highest negative π charge at the metal−organic contact. An electronwithdrawing group (Z), as −NO2 and −COOH in the same ring position, gives the highest positive π charge at this interface point via conjugation. Intermediate situations are obtained when L or Z groups are placed in a meta position with respect to the surface. For meta L groups, the net π charge at the metal−organic contact becomes slightly positive to compensate for the local negative charge accumulations on neighboring atoms. On the contrary, for meta Z groups this anchoring carbon atom becomes slightly anionic as it responds to adjacent local positive polarizations. Because many authors have suggested that IET spectra can be traced back to the electronic structure around Fermilevel,13,20,25−32 we first discuss this property. In Figure 2, we show PDOS plots for all phenyl-based adsorbates studied here. These PDOS plots indicate how gas-phase molecular orbitals (MOs) of the radicals hybridize with the metal surface upon adsorption. Only resonances having contributions in the probed regime close to Fermi-energy are shown, along with a schematic illustration of the MO they represent. As we can see from Figure 2, all para and meta substituents have a broad σ singly occupied molecular orbital (SOMO) and a peaked π* lowest unoccupied molecular orbital (LUMO) almost symmetrically disposed in energy below and above the Fermi level. In general, we can thus state that the character of the MOs and SOMO-LUMO gap of gas-phase para- and metasubstituted fragments with the same functional group is similar, except for the LUMO coefficient on the contact carbon atom. Also their overall PDOS characteristics (Figure 2) display almost no difference. This was confirmed by the fact that resembling constant-current STM images at +0.2 V were obtained for para and meta isomers (partially published elsewhere,20,25 not shown for brevity). Following the state-of-
Figure 2. PDOS plots onto frontier MOs with contributions around Fermi-level (EF) and their corresponding amplitudes (black = σ MO, red = π* MO).
the-art literature on IETS, this practical identicalness in electronic structure around Fermi-level should imply that similar IETS spectra are expected, regardless of the ring position of the functional group. The presence of a strongly hybridized molecular state with σ character at the Fermi level (broad SOMO band) should ensure a large electron vibration coupling with stretch modes, and it could be anticipated that some stretching modes will be active for all selected fragments. However, a strikingly different result arises from our IETS simulations. The IETS relative intensities per vibration calculated at the center of the adsorbates are shown in Figure 3. As we can see in Figure 3, for the COOH substituent, in para position, almost no IETS-signal was detected. By moving it to meta, however, suddenly more vibrations become visible in its IETS spectrum. For NO2, the transition from para to meta substitution on the phenyl ring also leads to the appearance of extra signals, although the effect is less drastic. We recall Figure 1 in which we show that para-COOH and -NO2 cause a strong positive polarization of the π system at the point where they are anchored to the surface, whereas their meta analogues yield a small negative charge at this position. This sheds light onto a possible charge descriptor, the π electron density at the metal− organic contact, for which enhancement can be linked to the IETS activity of a surface-adsorbed compound. We confirm this postulation by comparing the IETS-spectra of meta- and paraNH2. As Figure 3 shows, when NH2 is substituted at the para position, the molecule shows a larger IETS activity than in the case for the meta position. NH2 indeed causes an increase in π electron density at the metal−organic contact when situated at para, whereas at meta it yields a small positive charge (Figure 1). 3008
dx.doi.org/10.1021/jz3012832 | J. Phys. Chem. Lett. 2012, 3, 3007−3011
The Journal of Physical Chemistry Letters
Letter
retain its gas phase value (of one electron). For the latter two compounds the binding is solely dative through the lone pair nitrogen, although this does not modify the perpendicular π electron density. As such, all molecules have a zero net π charge at the metal−organic contact. Following our hypothesis, this has a neutral effect on the manifestation of stretching modes in IETS: it implies that these are not inhibited, as only a (slight) positive π polarization is detrimental to their IETS activity. This justifies that it was possible to observe them previously.13,32 In the general picture, by using the aforementioned differences in IETS spectra of meta and para substituted phenyl-compounds, it is also possible to distinguish between various ring positions of the same functional group using IETS as a resolution technique. To finalize, we propose a qualitative picture of the complex IETS phenomenon. For this, we take the perspective that the frontier orbitals design two tunneling channels of different types: σ that lies in the molecular plane and π that has a perpendicular orientation. Inelastic electron tunneling occurring through the σ channel should lead to the visibility of stretching vibrations because these match the σ symmetry by being in-plane motions. When transport occurs through the perpendicular π channel, however, stretching modes do not match the symmetry and thus are forbidden. IET activity of OOP modes with the same π symmetry is then favored. When the net π charge at the metal−organic contact is zero, the σ channel is intrinsically prominent over the π channel because it has a larger hybridization with metallic states. As such, stretching modes are visible in IETS (as previously discussed for pristine phenyl, benzene, pyridine, and α-pyridil). When a positive π polarization is induced at the metal−organic contact (the molecular entrance) by chemical substitution, the negatively charged tunneling electron is attracted toward it and hence dragged into the π channel present at that binding point: the stretching modes turn silent and OOP modes are seen. On the contrary, a negative π polarization at the metal− organic contact will repel the tunneling electron from entering this π channel and favor the stretching mode activity through the remaining σ channel. In conclusion, we have performed DFT simulations of IETS for a series of functionalized phenyl compounds substituted in meta and para positions relative to the surface. We find that high-frequency stretching modes become IETS-active when electron-withdrawing groups as COOH and NO 2 are substituted meta or when electron-donating groups as NH2 are linked para. This sensitizing effect is attributed to an increase in π electron density at the metal−organic contact point, which can be evaluated by mesomeric forms at no computational cost. Contrarily to previous assumptions in literature, we show that the electronic structure around Fermi level does not play an all-determining role. These findings further emphasize the discriminative power of IETS, as it is capable to resolve different ring positions in addition to the already known capability to distinguish between chemical species and adsorption configurations. Because much benefit can be made of this accurate single-molecule resolution, we hope that these chemical insights will aid the development of IETS to become a well-understood and standard characterization tool in surface science.
Figure 3. IETS relative intensities per vibration calculated on the center (above the phenyl-ring) of the monosubstituted phenyl fragments/Cu(111). A red line indicates the experimental detection threshold of 1%. Green = active stretching vibrations; blue = active OOP wagging vibrations; uncolored = inactive. Indicated values are vibration energies in millielectronvolts. Because of the lack of symmetry of the meta fragments, off-center positions (above the substituent) cannot be the same for all cases and thus are not relevant for comparison.
Let us now look closely at the type of vibrations that become IETS-active due to an increased π electron density at the metal−organic contact. In Figure 3 we see that for meta-NO2 (meta-COOH) compared with para-NO2 (para-COOH), additional signals are measured for modes at 391 (390 meV), 87 (83 meV), and 49 meV (49 meV); for para-NH2 relative to meta-NH2, two extra vibrations become active at 449 and 436 meV. Because the modes at 87 (83 meV) and 49 meV (49 meV) of meta-NO2 (meta-COOH) are “shoulder peaks” of the already known OOP wagging modes of B1 symmetry, we will focus more on the rest. From the vibrational analyses including symmetry assignments (see the Supporting Information), we see that all of these extra modes are stretching modes at high frequencies. For meta-NO2 (meta-COOH) at 391 meV (390 meV) a C−H stretch of phenyl is detected; for para-NH2, we find a NH2 antisymmetric stretch at 449 meV and NH2 symmetric stretch at 436 meV. Hence, an increased π electron density at the metal−organic contact may be viewed as an activator for high-frequency stretching modes, or, taking the inverse perspective, a positive polarization at that point may be viewed as a quencher of such vibrations. We now further generalize this hypothesis by confirming its applicability to published examples of IETS-active C−H stretches from other aromatic compounds. We find that these have been reported for pristine phenyl, benzyne, pyridine and α-pyridil.13,32 The former two systems are covalently bound via axial overlap and the perpendicular π electron density should
■
COMPUTATIONAL SECTION We modeled all monosubstituted phenyl molecules on Cu(111) (described in Figure 1) by applying periodic boundary 3009
dx.doi.org/10.1021/jz3012832 | J. Phys. Chem. Lett. 2012, 3, 3007−3011
The Journal of Physical Chemistry Letters
Letter
conditions to p(5 × 5) slabs containing three monolayers of copper and 25 Å vacuum. We performed periodic density functional theory (DFT) calculations at the generalized gradient approximation (GGA) level, using the Perdew− Burke−Ernzerhof (PBE) functional,33,34 as implemented in the Vienna ab initio simulation package (VASP).35,36 The projected augmented wave (PAW) method37,38 was used with a plane-wave basis set being cut off at 400 eV and k-point sampling of 3 × 3 × 1. Geometries were optimized by relaxing the molecular degrees of freedom as well as those of the first metal layer with a convergence criterion of 0.02 eV/Å. The lowest two metal layers were kept rigid in their standard (111)slab configuration, defined by the theoretical lattice parameter of bulk copper (3.64 Å). Vibrations were evaluated by diagonalizing the dynamical matrix obtained from VASP while only taking the degrees of freedom of the adsorbate into account. The electronic structure in the tunneling energy regime has been analyzed elsewhere,32 first by the computation of the projected density of states (PDOS).39 Additionally, constant current STM images at +0.2 V were calculated using Tersoff−Hamann theory40,41 using a specific improvement of the wave function decay in the vacuum region.42 Finally, IETS simulations were carried out for each vibration using a manybody extension43,44 of the Tersoff−Hamann theory, described before.25 IETS intensities were captured at a tip position above the center of the molecule, which is the experimentally most relevant position.
■
(5) Gimzewski, J. K.; Joachim, C. Nanoscale Science of Single Molecules Using Local Probes. Science 1999, 283, 1683−1688. (6) Stipe, B. C.; Rezaei, M. A.; Ho, W. Single-Molecule Vibrational Spectroscopy and Microscopy. Science 1998, 280, 1732−1735. (7) Ho, W. Single-Molecule Chemistry. J. Chem. Phys. 2002, 117, 11033−11061. (8) Ueba, H. Motions and Reactions of Single Adsorbed Molecules Induced by Vibrational Excitation with STM. Surf. Rev. Lett. 2003, 10, 771−796. (9) Lorente, N.; Rurali, R.; Tang, H. Single-Molecule Manipulation and Chemistry with the STM. J. Phys.: Condens. Matter 2005, 17, S1049−S1074. (10) Komeda, T. Chemical Identification and Manipulation of Molecules by Vibrational Excitation via Inelastic Tunneling Process with Scanning Tunneling Microscopy. Prog. Surf. Sci. 2005, 78, 41−85. (11) Lauhon, L. J.; Ho, W. Single-Molecule Chemistry and Vibrational Spectroscopy: Pyridine and Benzene on Cu(001). J. Phys. Chem. A 1999, 104, 2463−2467. (12) Lin, L.-L.; Wang, C.-K.; Luo, Y. Inelastic Electron Tunneling Spectroscopy of Gold−Benzenedithiol−Gold Junctions: Accurate Determination of Molecular Conformation. ACS Nano 2011, 5, 2257−2263. (13) Bocquet, M.-L.; Lesnard, H.; Lorente, N. Inelastic Spectroscopy Identification of STM-Induced Benzene Dehydrogenation. Phys. Rev. Lett. 2006, 96, 096101−4. (14) Ren, H.; Yang, J.; Luo, Y. Simulation of Inelastic Electronic Tunneling Spectra of Adsorbates from First Principles. J. Chem. Phys. 2009, 130, 134707−5. (15) Okabayashi, N.; Paulsson, M.; Ueba, H.; Konda, Y.; Komeda, T. Inelastic Tunneling Spectroscopy of Alkanethiol Molecules: HighResolution Spectroscopy and Theoretical Simulations. Phys. Rev. Lett. 2010, 104, 077801−4. (16) Kim, Y.; Hellmuth, T. J.; Bürkle, M.; Pauly, F.; Scheer, E. Characteristics of Amine-Ended and Thiol-Ended Alkane SingleMolecule Junctions Revealed by Inelastic Electron Tunneling Spectroscopy. ACS Nano 2011, 5, 4104−4111. (17) Long, D. P.; Troisi, A. Inelastic Electron Tunneling Spectroscopy of Alkane Monolayers with Dissimilar Attachment Chemistry to Gold. J. Am. Chem. Soc. 2007, 129, 15303−15310. (18) Gaudioso, J.; Lee, H. J.; Ho, W. Vibrational Analysis of Single Molecule Chemistry: Ethylene Dehydrogenation on Ni(110). J. Am. Chem. Soc. 1999, 121, 8479−8485. (19) Parschau, M.; Rieder, K. H.; Hug, H. J.; Ernst, K. H. SingleMolecule Chemistry and Analysis: Mode-Specific Dehydrogenation of Adsorbed Propene by Inelastic Electron Tunneling. J. Am. Chem. Soc. 2011, 133, 5689−5691. (20) Burema, S. R.; Bocquet, M. L. Sensitizers in Inelastic Electron Tunneling Spectroscopy: A First Principles Study of Functional Aromatics on Cu(111). Nanotechnology 2012, 23, 315702−10. (21) Morgenstern, K. On the Interpretation of IETS Spectra of a Small Organic Molecule. J. Phys.: Condens. Matter 2011, 23, 484007−8. (22) Fock, J.; Sørensen, J. K.; Lörtscher, E.; Vosch, T.; Martin, C. A.; Riel, H.; Kilså, K.; Bjørnholm, T.; van der Zant, H. A Statistical Approach to Inelastic Electron Tunneling Spectroscopy on FullereneTerminated Molecules. Phys. Chem. Chem. Phys. 2011, 13, 14325− 14332. (23) Gaudioso, J.; Ho, W. Single-Molecule Vibrations, Conformational Changes, and Electronic Conductivity of Five-Membered Heterocycles. J. Am. Chem. Soc. 2001, 123, 10095−10098. (24) Xi, M.; Bent, B. E. Iodobenzene on Cu(111): Formation and Coupling of Adsorbed Phenyl Groups. Surf. Sci. 1992, 278, 19−32. (25) Burema, S. R.; Lorente, N.; Bocquet, M. L. A Theoretical Rationalization of a Total Inelastic Electron Tunneling Spectrum: The Comparative Cases of Formate and Benzoate on Cu(111). J. Chem. Phys. 2012, 136, 244507−7. (26) Lorente, N.; Persson, M.; Lauhon, L. J.; Ho, W. Symmetry Selection Rules for Vibrationally Inelastic Tunneling. Phys. Rev. Lett. 2001, 86, 2593−2596.
ASSOCIATED CONTENT
S Supporting Information *
Detailed vibrational analyses of all investigated monosubstituted phenyl molecules on Cu(111). This material is available free of charge via the Internet http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are grateful for support from the FP7Marie Curie Actions of the European Commission, via the Initial Training Network SMALL (MCITN-238804). We thank the Pôle Scientifique de Modélisation Numérique (PSMN), part of the Fédération Lyonnaise de Calcul Haute Performance (FLCHP) of the Centre Blaise Pascal (CBP), Lyon, France for computational resources.
■
REFERENCES
(1) Ertl, G.; Knoezinger, H.; Schueth, F.; Weitkamp, J. Handbook of Heterogeneous Catalysis; Wiley-VCH: Weinheim, Germany, 2008. (2) Joachim, C.; Ratner, M. A. Molecular Electronics: Some Views on Transport Junctions and Beyond. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 8801−8808. (3) Moresco, F.; Meyer, G.; Rieder, K.-H.; Tang, H.; Gourdon, A.; Joachim, C. Conformational Changes of Single Molecules Induced by Scanning Tunneling Microscopy Manipulation: A Route to Molecular Switching. Phys. Rev. Lett. 2001, 86, 672−675. (4) Kuehnle, A.; Linderoth, T. R.; Hammer, B.; Besenbacher, F. Chiral Recognition in Dimerization of Adsorbed Cysteine Observed by Scanning Tunnelling Microscopy. Nature 2002, 415, 891−893. 3010
dx.doi.org/10.1021/jz3012832 | J. Phys. Chem. Lett. 2012, 3, 3007−3011
The Journal of Physical Chemistry Letters
Letter
(27) Bocquet, M.-L.; Lorente, N. Probing the Proton Location in a Water Bilayer on Pd (111) by Inelastic Spectroscopy Simulations. J. Chem. Phys. 2009, 130, 124702−6. (28) Alducin, M.; Sanchez-Portal, D.; Arnau, A.; Lorente, N. MixedValency Signature in Vibrational Inelastic Electron Tunneling Spectroscopy. Phys. Rev. Lett. 2010, 104, 136101−4. (29) Maddox, J. B.; Harbola, U.; Liu, N.; Silien, C.; Ho, W.; Bazan, G. C.; Mukamel, S. Simulation of Single Molecule Inelastic Electron Tunneling Signals in Paraphenylene−Vinylene Oligomers and Distyrylbenzene[2.2]paracyclophanes. J. Phys. Chem. A. 2006, 110, 6329−6338. (30) Franke, K. J.; Schulze, G.; Pascual, J. I. Excitation of Jahn−Teller Active Modes during Electron Transport through Single C60 Molecules on Metal Surfaces. J. Phys. Chem. Lett. 2009, 1, 500−504. (31) Persson, B. N. J.; Baratoff, A. Inelastic Electron Tunneling from a Metal Tip: The Contribution from Resonant Processes. Phys. Rev. Lett. 1987, 59, 339−342. (32) Lesnard, H.; Bocquet, M.-L.; Lorente, N. Dehydrogenation of Aromatic Molecules under a Scanning Tunneling Microscope: Pathways and Inelastic Spectroscopy Simulations. J. Am. Chem. Soc. 2007, 129, 4298−4305. (33) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77, 3865−3868. (34) Perdew, J. P.; Burke, K.; Ernzerhof, M. Generalized Gradient Approximation Made Simple [Phys. Rev. Lett. 77, 3865 (1996)]. Phys. Rev. Lett. 1997, 78, 1396−1396. (35) Kresse, G.; Furthmueller, J. Efficiency of Ab-Initio Total Energy Calculations for Metals and Semiconductors using a Plane-Wave Basis Set. Comput. Mater. Sci. 1996, 6, 15−50. (36) Kresse, G.; Furthmueller, J. Efficient Iterative Schemes for Ab Initio Total-Energy Calculations using a Plane-Wave Basis Set. Phys. Rev. B 1996, 54, 11169−11186. (37) Kresse, G.; Joubert, D. From Ultrasoft Pseudopotentials to the Projector Augmented-Wave Method. Phys. Rev. B 1999, 59, 1758− 1775. (38) Bloechl, P. E. Projector Augmented-Wave Method. Phys. Rev. B 1994, 50, 17953−17979. (39) Hoffmann, R. A Chemical and Theoretical Way to Look at Bonding on Surfaces. Rev. Mod. Phys. 1988, 60, 601−628. (40) Tersoff, J.; Hamann, D. R. Theory and Application for the Scanning Tunneling Microscope. Phys. Rev. Lett. 1983, 50, 1998− 2001. (41) Tersoff, J.; Hamann, D. R. Theory of the Scanning Tunneling Microscope. Phys. Rev. B 1985, 31, 805−813. (42) Lorente, N. Theory of Tunneling Currents and Induced Dynamics at Surfaces. In Handbook of Surface Science; Hasselbrink, E., Lundqvist, B. I., Eds.; North-Holland: Amsterdam, 2008; Chapter 12 (43) Lorente, N.; Persson, M. Theory of Single Molecule Vibrational Spectroscopy and Microscopy. Phys. Rev. Lett. 2000, 85, 2997−3000. (44) Lorente, N. Mode Excitation Induced by the Scanning Tunneling Microscope. Appl. Phys. A: Mater. Sci. Process. 2004, 78, 799−806.
3011
dx.doi.org/10.1021/jz3012832 | J. Phys. Chem. Lett. 2012, 3, 3007−3011
