12202
J. Phys. Chem. C 2010, 114, 12202–12206
Scanning Tunneling Microscopy Investigation of Tris(phthalocyaninato)yttrium Triple-Decker Molecules Deposited on Au(111) Hironari Isshiki,‡,† Jie Liu,‡,† Keiichi Katoh,† Masahiro Yamashita,† Hitoshi Miyasaka,† Brian K. Breedlove,† Shinya Takaishi,† and Tadahiro Komeda*,§,‡ Department of Chemistry, Graduate School of Science, Tohoku UniVersity, Aramaki-Aza-Aoba, Aoba-Ku, Sendai 980-8578, Japan, Institute of Multidisciplinary Research for AdVanced Materials, Tohoku UniVersity, 2-1-1, Katahira, Aoba-Ku, Sendai, 980-0877, Japan, and CREST, JST, 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan ReceiVed: January 31, 2010; ReVised Manuscript ReceiVed: May 28, 2010
We studied triple-decker tris(phthalocyaninato)yttrium (Y2Pc3) molecules deposited on a Au(111) surface with a low-temperature scanning tunneling microscope (STM). It is shown that the triple-decker molecule can be successfully transferred to the Au(111) surface by a sublimation method in the ultrahigh vacuum condition. A monolayer film of Y2Pc3 is observed with a height of ∼0.55 nm from the bare Au(111) surface. The molecules are adsorbed with a flat-lying configuration and a pseudosquare lattice is formed. Inside each molecule, eight bright protrusions were observed in the occupied-state images, which correspond to the high density-of-state (DOS) area on both sides of four phenyl rings in the top phthalocyanine (Pc). Scanning tunneling spectroscopy (STS) data show distinct features both in occupied and in unoccupied states. The variation of STS spectra when the tip moved from the ligand position to the center of the molecule was small, showing a limited contribution from the center metal. Introduction Phthalocyanine (Pc) molecules have been investigated intensively for various applications,1 which is now extended to a new target such as field-effect transistors. In addition, a successful synthesis of a variety of sandwich-type double- and triple-decker Pc complexes has produced new functions of Pc. For example, double- and triple-decker Pc molecules with lanthanoid metal atoms as linkers (MPc2, M2Pc3) show the properties of single molecule magnets (SMMs).2-5 Such properties attract attention for the applications to the quantum computing and the spintronics devices.6,7 It is a crucial step to control the assembly of molecular components on surfaces in the development of molecular-scale devices. The development of scanning tunneling microscopy (STM) has enabled imaging and positional control of individual molecules.8 The STM observation of the sandwich complex is more challenging because of the nonplanar characteristics; unlike planar monophthalocyanine, the double- and triple-decker sandwich complexes are cylindrical rather than disklike. Takami et al. has successfully observed STM images of double- and triple-decker complexes of Pcs at the liquid-solid interface between 1-phenyloctane and highly oriented pyrolytic graphite (HOPG);9 however, a triple-decker molecule of (Pc)Y(Pc*)Y(Pc*) [Pc* ) 2,3,9,10,16,17,23,24-octakis(octyloxy)phthalocyaninato] was not ordered at the liquid and solid interface. More recently, an STM observation of a film of a triple-decker Pc complex formed at a liquid/solid interface was reported.10 It is known that the film grown by a molecular sublimation in ultrahigh vacuum (UHV) condition can produce a high quality * To whom correspondence should be sent. E-mail: komeda@ tagen.tohoku.ac.jp. † Graduate School of Science, Tohoku University. ‡ Institute of Multidisciplinary Research for Advanced Materials, Tohoku University. § JST.
film and enables a precise understanding of its structural and electronic configurations. We have reported that a film of double-decker Pc molecules can be formed by a thermal deposition method.11-13 In this paper, we show high resolution STM images and scanning tunneling spectroscopy (STS) of triple-decker Pc (Y2Pc3) molecules on the Au (111) surface at liquid He temperature, in which the molecules were deposited with sublimation of the molecules from a heated Ta boat. It was observed that the triple-decker molecules can be transferred to the metal surface without a dissociation of molecules. A pseudosquare lattice was observed in the film. In addition, internal structures of the molecules were successfully observed and we propose a tentative model of the lattice and the molecule configuration of the Y2Pc3 film. STS data show distinct features with differences from that obtained on YPc2 films. Experiment All experiments were done in the UHV condition. The Au(111) single crystal was used for the substrate, which was prepared with a standard Ar+ sputtering and annealing process. Y2Pc3 molecules were synthesized by ourselves, and the details are described elsewhere.13 The degassing of the molecules was performed carefully by heating a container of the Ta boat for several hours prior to evaporation. The YPc2 and Y2Pc3 molecule deposition was executed under a flux rate of ∼0.1 monolayer (ML) per minute with the Au substrate kept at room temperature. No subsequent annealing was executed after the deposition. The substrate cleaning, molecule deposition, and low-temperature STM observation were done in ultrahigh vacuum (UHV) chambers that are connected by metal valves. The sample can be transferred between them without being exposed to a nonUHV condition. The STM head is placed in a tubelike stainless chamber, which has been inserted into a He dewar. The He
10.1021/jp101349v 2010 American Chemical Society Published on Web 06/24/2010
STM of (Y2Pc3) on Au(111) dewar is attached to an air-suspended table and located below the floor level. The assembly of the tube scanner, STM tip, sample holder, and inertia slider for the coarse motion of the tip are suspended by springs for a vibration isolation. The sample temperature was kept at ∼4.5 K for the STM/STS experiments described in this article. STS spectra were obtained using a lockin amplifier in which a modulation voltage of 10 mV was superimposed to the tunneling bias voltage. The first principles calculations were performed by the VASP code, employing a plane wave basis set and the PAW potentials to describe the behavior of valence electrons.14-18 The exchange-correlation energies are described by the PerdewWang (PW91) functions with the generalized gradient approximations (GGA).19 Each Y2Pc3 molecule consists of 170 atoms (96 C atoms, 24 N atoms, 48 H atoms, and 2 Y atoms). To reduce the computational burden, we have tried modeling the Au(111) surface by using a three-layer Au atomic sheet with (111) normal orientation in which Au atoms are fixed. We find this is reasonable because the adsorption of Y2Pc3 on Au is considerably weak. In the structural optimization process, an energy cutoff of 300 eV is used for the plane waves, and a plane wave cutoff of 400 eV is used for the calculation of intermolecular interactions. The simulations of the STM images have been performed on the basis of the Tersoff-Hamann model20,21 using the program STRender.22 Results and Discussion First we discuss the stucutre of a Y2Pc3 molecule. Each Y2Pc3 has two Y atoms sandwitched by three Pc planes. The structure analysis using an X-ray diffraction (XRD) with a bulk singlecrystal of the molecules can give detailed information about the conformation of the stacking of Pc planes. Unfortunately, due to a limited amount of synthesized samples, XRD data of Y2Pc3 molecules neither of a single crystal nor a thin film was obtained. However, the XRD data, such as for Bi2Pc3 formed byasublimed-crystallization,23 Cd2(OAc)3 (Oac)1,4,8,11,15,18,22,25octakis(hexyl)phthalocyanine) formed by a recrystallization from a solvant,24 and Tb2(obPc)3 (obPc ) dianion of 2,3,9,10,16,17,23, 24-octabutoxyphthalocyanine) in the shape of a single-crystal,25 were reported, in which we can see a common Pc stacking configuration. For the Bi2Pc3 case, the middle Pc ligand is rotated 38.3° with respect to the top and bottom Pcs. The top Pc is just slightly (∼0.5°) rotated with respect to the bottom Pc. The configuration of the three Pc can be seen for the other two molecules, and the twist angles were reported to be ∼34.2° and ∼32° for the Cd2(OAc)3 and Tb2(obPc)3 cases, respectively.24,25 Thus we considered a model for the Y2Pc3 molecule with the twist angle of 30° that was further examined by using the VASP calculation. The result after the structure optimization is shown in Figure 1; a top view in (a) and a tilted side view in (b). We found that a local minimum structure can be obtained at the twist angle of 30° after calculation of the structural optimization starting from the same twist angle. The top view illustrated in Figure 1a shows that the top Pc eclipses the bottom Pc and the central Pc is staggered. However, the energy differences between the local minima of other twist angles are too small to determine the precise twist angle with the computational resource we have. As additional information, we found that the outer planes have a bent configuration rather than a flat plane (Figure 1b). Similar bent configurations of the outer Pc planes can be seen in the report of Cd2(OAc)3.24 Figure 1c shows an STM image of a Y2Pc3 film on Au(111). The Au bare substrate (dark area) can be seen at the upper-left corner of the image, and the rest of the image is occupied by a
J. Phys. Chem. C, Vol. 114, No. 28, 2010 12203
Figure 1. (a) Top view and (b) tilted side view of a Y2Pc3 tripledecker molecule. (c) STM image of the Y2Pc3 film on the Au(111) surface. Both the film and Au bare surface are seen (45 × 27 nm2, IT ) 0.4 nA, V ) -0.8 V, T ) 4.7 K). (d) Cross section along the white line in (c).
film of Y2Pc3. The domain size of this film is roughly 70 × 70 nm2. It should be noted that we observed very few isolated molecules on the substrate, and almost all molecules were taken into some islands of the film. Inside the film, we can identify a pseudo-square lattice. One of the perimeters of the lattice is parallel to the [11j0] direction of the substrate. Figure 1d shows a cross section of the image of Figure 1c along the marked white line, which reads that the height of the film is ∼0.55 nm from the bare Au substrate. We consider that this is an appropriate height of a monolayer film of the tripledecker Pc with the following reasons. The height of a single Pc layer has been reported as ∼0.14 nm on Au surface.26 For double-decker molecules, we have reported the height of YPc2 (double-decker) as ∼0.42 nm on the Au(111) surface,8,12 and Vitali et al. reported that the TbPc2 molecule deposited on the Cu(111) surface shows a height of ∼0.3 nm.27 We can also notice that a film of [(Pc′)2TbIII] (Pc′ ) octabutoxy-substituted phthalocyaninato ligand) molecules deposited on HOPG shows a height of ∼0.3 nm.28 Comparing with these results, we consider that the observed height of ∼0.55 nm is a reasonable one for the monolayer film of the triple-decker Y2Pc3 molecules. We stress that triple-decker Pc molecules were successfully transferred to the substrate by a thermal evaporation method. Considering a high quality surface obtained by the vacuum sublimation technique, this is a very important premise for further research on the electronic structure or the physical properties of triple-decker molecules. Magnified occupied-state STM images of the film are shown in Figure 2a,b, which were obtained with a tunneling current of 0.4 nA and a sample voltage of -0.8 V. The center of the molecule corresponds to a dark circle of the image. We marked four molecules A-D in Figure 2a, which make a unit of the pseudosquare lattice of the Y2Pc3 film. We show a simulated STM image for the occupied state near the Fermi level in Figure 2c using the VASP calculation and the STRender visualization software. Basically left and righthand sides of each phenyl ring are protruded, which give a total of eight bright spots, marked 1-8 in the simulation image. With the knowledge of an expected STM image of a molecule, we can analyze the STM image in more detail. For each Y2Pc3 molecule, we can identify eight bright spots surrounding the center of the molecule. They are marked by 1-8 in Figure
12204
J. Phys. Chem. C, Vol. 114, No. 28, 2010
Isshiki et al.
Figure 3. Schematic model of Y2Pc3 on Au(111). Yellow circles are Au atoms. Top and bottom Pcs are shown in green and blue, respectively, the latter of which are eclipsed by the former in A and C. s and t are unit vectors of the gold substrates.
Figure 2. Magnified images of Y2Pc3 film on Au(111). (a) Occupied state image (5.2 × 5.2 nm2, IT ) 0.4 nA, V ) -0.8 V). Models of the top Pc of the Y2Pc3 molecule are superimposed for molecules A-D. Numbers 1-8 correspond to bright spots originated from molecule A. The numbers are corresponding to (c). (b) Same as (a) but observed at a different position of the surface. Crosses connecting the spots 2-6 and 4-8 are superimposed. (c) Simulated STM image of the occupied state. Eight bright spots are located on both sides of the phenyl rings, which are numbered 1-8. (d) Unoccupied state image (5.2 × 5.2 nm2, IT ) 0.4 nA, V ) 0.8 V). Note (a) and (d) were taken at exactly the same area. (e) Simulated STM image of the unoccupied state.
2a,b. At the positions of bright spots 1, 3, 5, and 7, we also expect the bright spots of the neighboring molecules. For example, spot 3 looks like a convolution of the spots of molecules A and B. The other spots (2, 4, 6, and 8) originate from a single molecule. Here we examine an ordering of rotational angles of Pc planes. It was executed by checking a direction of a symmetry line of Pc planes. Spots 1, 3, 5, and 7 are broad since they are
shared by neighboring molecules. Thus we examined the symmetric lines connecting spots 2 and 6 (also 4 and 8). Interestingly, they show a slight rotation in an alternate manner and the nearest neighbor molecules are rotated with each other. They are illustrated in Figure 2b by white and blue crosses. All white (blue) orthogonal crosses are directing to the same direction. The blue crosses are rotated ∼8° clockwise from the white crosses. It should be mentioned that the symmetry line connecting the diagonal phenyl rings is aligned with the [011j] direction within an accuracy of 0.5° for the molecules with the white crosses. In an unoccupied-state image, we see a different shape of the molecule. The image of Figure 2d was obtained with a tunneling current of 0.4 nA and a sample voltage of 0.8 V. The image was obtained in the same area as that of Figure 2a, and we superimposed the identical models of the top Pc. Comparing with the images of Figure 2a, we notice that the four spots shared by neighboring molecules are more highlighted. The simulation image of the unoccupied state is shown in Figure 2e. The change from Figure 2c is that the spots around the phenyl ring are broadened. With the VASP simulation, we could not reproduce the four highlighted spots, but spread electronic states probably make an overlap between the neighboring molecules stronger and make the four distinguished spots. We consider the lattice vectors of the film with respect to the Au(111) lattice. The tentative models are illustrated in Figure 3, in which the marks of the molecules A-D correspond to the ones of Figure 2a. The lattice can be defined with two vectors a (D f B) and b (D f C). With two unit vectors of the lattice (shown by s and t in Figure 3), the two sets are described like following;
() (
)( )
0 5 s a ) 6 -3 t b
in which a and b are parallel to the [11j0] and [112j] directions, respectively. a has the length of 5a where a corresponds to the nearest-neighbor distance of the Au(111) surface (∼0.288 nm).
STM of (Y2Pc3) on Au(111)
J. Phys. Chem. C, Vol. 114, No. 28, 2010 12205 TABLE 1: Energy of the Peak Positions Observed in Figure 4 peak
energy (mV)
peak
energy (mV)
A1 A2 A3 A4 A5 B1 B2 B3 B4 B5
-1229 -629 570 792 1200 -1345 -706 670 864 1318
C1 C2 C3 C4 C5 D1 D2 D3 D4 D5
-904 -595 150 839 1060 -974 -471 190 873 1068
TABLE 2: Comparison of the Enegy Positions of Corresponding Peaks in A and B of Figure 4
Figure 4. dI/dV spectra of Y2Pc3 film (labeled A, B) and YPc2 film (labeled C, D). A and C were measured on the bright lobes of the molecules, while B and D were at the center of the molecules.
The length of b is 3 · 31/2a, which is longer than 5a by ∼4%. The size and the direction of the lattice agree with the observed periodicity of the images of Figure 2. We consider the azimuthal angle of the top Pc as follows. First, we assume the bottom Pc follows the film configuration of the molecules of H2Pc,29 FePc,30 and CoPc,31 on the Au(111) surface, in which the film lattice is formed by the vectors a and b proposed in Figure 3 and the azimuthal rotation is aligned so the line connecting the centers of the diagonal phenyl rings is parallel to the [011j] direction, which is rotated 60° from the [11j0] direction. Second, the top and bottom Pc planes have the same azimuthal angle, which was discussed above with examples of other triple-decker Pc molecules. As noted above, the azimuthal angles of the top Pc of molecules B and C are rotated ∼8° in the clockwise direction, and the bottom Pc appears in the top view due to this rotation, which are sketched by a blue skeleton model in Figure 3. The driving force of the rotation of the molecules is not clear, but a part of it should be a steric repulsion between the neighboring Pc ligands. These models are superimposed on the images of Figure 2a,d and can reproduce the observed STM images well. Figure 4 shows dI/dV spectra obtained on the Y2Pc3 film (labeled A and B) and the YPc2 film (C and D). Spectra A and C were measured at bright lobes corresponding to both sides of the phenyl rings, while B and D are measured at the center of molecules (STM image of YPc2 film were shown in elsewhere12). The spectrum of the bare Au(111) surface showed no feature except the surface state at -0.5 V. The energies of the marked features of the spectra are summarized in Table 1. First we examine the difference of the spectra when the tip was positioned on the ligand and the molecule-center position of the Y2Pc3 molecule. Overall, the peaks in the spectra A have corresponding features in B, which are similar in shape and relative intensity. It was reported that the STS spectra obtained at the lobe and the metal positions have large differences for the CoPc molecule adsorbed on the Au(111) surface,32 in which the d-orbital induced state was observed only when the tip was positioned at the center of the molecule and the spectrum taken at the position showed no ligand features. This indicates that, for the Y2Pc3 molecule, the contribution from the metal to the local DOS is limited.
peak
peak
energy B/A
A1 A2 A3 A4 A5
B1 B2 B3 B4 B5
1.09 1.12 1.18 1.09 1.10
It is intriguing to notice that the peak pisitions of B are shifted away from the Fermi level both in occupied and in unoccupied states if compared with the corresponding peaks in A. In addition, the shift is larger for the peaks with more energy separation from the Fermi level. This might be more clearly seen in Table 2 in which the peak energies are compared between A and B as their ratio. We consider that the shift is systematic rather than an accidental shift due to some tip effect. The shift of peak positions in STS spectra obtained with molecules was reported previously.33,34 Gopakumar et al. reported a shift of the highest occupied molecular orbital (HOMO) level away from the Fermi level when the tip-sample distance was decreased, while the lowest unoccupied molecular orbital (LUMO) level was pinned for the system of metal-Pc molecules adsorbed on a graphite surface.34 For the mechanism, they considered the preferential polarization of the center metal together with the shift of the energy levels of the metal orbitals with the presence of the electric field formed by the tip and the substrate. As expected from the reversed direction of the polarization for the occupied and the unoccupied state, the behavior of the peak shifts are unsymmetric around the Fermi level, which cannot explain the symmetric peak shift observed in our experiment. Deng and Hipps examined the actual bias voltage between the tip and the molecule, when the resistance between the molecule and the substrate is high.33 The tip-substrate bias voltage (Vb) is the sum of voltage drops between tip-molecule (Vt-m) and molecule-substrate (Vm-s); Vb ) Vt-m + Vm-s. In the STS measurement, we change Vb and show the current variation with Vb in the dI/dV spectra. However, the horizontal axis should be Vt-m for a precise argument of the electronic state of the molecule. They discussed that Vm-s should be small due to the expected small molecule-substrate resistance (Rm-s: Rm-s ∼ 0.1 MΩ for Xe-metal case35). However, when we measure larger molecules, the conductance through the molecule itself should be included in Rm-s. The Y2Pc3 molecule has a height of ∼8 Å from the substrate, and the resistance of the molecule should not be negligible. Also, when we measure the center of the molecule, the density of states at the position is very small, as revealed by the caluculation. This indicates that not only the tip has to be very close to the molecule but also the current path to the substrate should have a high resistance
12206
J. Phys. Chem. C, Vol. 114, No. 28, 2010
for the tunneling electrons to reach to the substrate. The current (It) is roughly proportional to the Vb, It ) cVb where c is a constant. The voltage drop of Vm-s is then expressed by Vm-s ) Rm-s × cVb, which is followed by the next formula Vt-m ) Vb - Vm-s ) (1 - Rm-s × c)Vb. To obtain a certain Vt-m, the required Vb is Vt-m/(1 - Rm-s × c), which is larger for a large Rm-s. Since Rm-s is high at the center of the molecule, the required Vb can be shifted away from the Fermi level that accounts for the observed shift of the dI/dV peaks. As pointed out by Deng and Hipps,33 the arugument should employ theoretical simulations for the conductance through molecules. For inelastic tunneling spectroscopy (IETS), the experimental and theoreteical simulations are showing excellent agrement,36,37 which should be extended for the issues of the energy shift in STS. Next we compare the dI/dV spectra of Y2Pc3 and YPc2. A peak at ∼150 mV can be seen in the spectrum C for the doubledecker molecule, which is completely missing in the tripledecker molecule. We speculate that this feature is derived from an interaction between the top Pc and the metal substrate, which is weaker for the triple-decker case. We need a theoretical calculation to discuss further. Summary We showed an STM/STS observation of triple-decker Y2Pc3 molecules deposited on Au(111) surface combined with a theoretical simulation. It is shown that the triple-decker phthalocyanine molecule, Y2Pc3, can be successfully transferred to the Au(111) surface by a sublimation method in UHV. A monolayer film of Y2Pc3 was observed with a height of ∼0.55 nm from the bare Au(111) surface. The molecules were adsorbed with a flat-lying configuration and a pseudosquare lattice was formed, which can be defined with two unit vectors of a and b; a parallel to [11j0] direction with a length of 5a (a is the nearest neighbor distance of Au(111)) and b parallel to [112j] direction with a length of 3 · 31/2a. We consider a model of the azimuthal rotation angle of each molecule, in which the bottom Pc aligns one of two diagonal directions of the molecules to the [011j] direction. This model can overall reproduce the observed STM images with an assumption that the top and bottom Pc has a same azimuthal angle, which is deduced from previous triple-decker Pc studies. However, in more detail, a slight rotation of the upper Pc of ∼8° in an alternate manner was found. STS spectra of Y2Pc3 were obtained both at the Pc ligand position and at the center of the molecule. The spectra obtained at the two sites showed similar peaks, indicating a small contribution from the metal atom for the STS spectra. However, the peak energies showed a systematic shift, which might be related to a high resistivity of the molecule. The spectra were further compared with the ones of YPc2 molecules, which shows a disappearance of electronic states near the Fermi level in Y2Pc2 that were observed in the YPc2 case.
Isshiki et al. References and Notes (1) Phthalocyanines: Properties and Applications; VCH: New York, 1993; Vol. 3. (2) Barlow, D. E.; Hipps, K. W. J. Phys. Chem. B 2000, 104, 5993. (3) Ishikawa, N.; Sugita, M.; Ishikawa, T.; Koshihara, S.; Kaizu, Y. J. Am. Chem. Soc. 2003, 125, 8694. (4) Ishikawa, N.; Sugita, M.; Tanaka, N.; Ishikawa, T.; Koshihara, S. Y.; Kaizu, Y. Inorg. Chem. 2004, 43, 5498. (5) Ishikawa, N. Polyhedron 2007, 26, 2147. (6) Gatteschi, D.; Sessoli, R. Angew. Chem., Int. Ed. 2003, 42, 268. (7) Bogani, L.; Wernsdorfer, W. Nat. Mater. 2008, 7, 179. (8) Gopakumar, T. G.; Lackinger, M.; Hackert, M.; Muller, F.; Hietschold, M. J. Phys. Chem. B 2004, 108, 7839. (9) Takami, T.; Arnold, D. P.; Fuchs, A. V.; Will, G. D.; Goh, R.; Waclawik, E. R.; Bell, J. M.; Weiss, P. S.; Sugiura, K.-i.; Liu, W.; Jiang, J. J. Phys. Chem. B 2006, 110, 1661. (10) Lei, S.-B.; Deng, K.; Yang, Y.-L.; Zeng, Q.-D.; Wang, C.; Jiang, J.-Z. Nano Lett. 2008, 8, 1836. (11) Zhang, Y.-F.; Isshiki, H.; Katoh, K.; Yoshida, Y.; Yamashita, M.; Miyasaka, H.; Breedlove, B. K.; Kajiwara, T.; Takaishi, S.; Komeda, T. J. Phys. Chem. C 2009, 113, 14407. (12) Zhang, Y. F.; Isshiki, H.; Katoh, K.; Yoshida, Y.; Yamashita, M.; Miyasaka, H.; Breedlove, B. K.; Kajiwara, T.; Takaishi, S.; Komeda, T. J. Phys. Chem. C 2009, 113, 9826. (13) Katoh, K.; Yoshida, Y.; Yamashita, M.; Miyasaka, H.; Breedlove, B. K.; Kajiwara, T.; Takaishi, S.; Ishikawa, N.; Isshiki, H.; Zhang, Y. F.; Komeda, T.; Yamagishi, M.; Takeya, J. J. Am. Chem. Soc. 2009, 131, 9967. (14) Kresse, G.; Hafner, J. Phys. ReV. B 1994, 49, 14251. (15) Kresse, G.; Hafner, J. Phys. ReV. B 1993, 47, 558. (16) Kresse, G.; Furthmuller, J. Phys. ReV. B 1996, 54, 11169. (17) Kresse, G.; Furthmuller, J. Comput. Mater. Sci. 1996, 6, 15. (18) Blochl, P. E. Phys. ReV. B 1994, 50, 17953. (19) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Phys. ReV. B 1992, 46, 6671. (20) Tersoff, J.; Hamann, D. R. Phys. ReV. Lett. 1983, 50, 1998. (21) Tersoff, J.; Hamann, D. R. Phys. ReV. B 1985, 31, 805. (22) Spisak, D. STRender, computer program; Institute of Materials Physics, University of Vienna: Vienna, Austria, 2007; http://homepage. univie.ac.at/daniel.spisak/. (23) Benihya, K.; Mossoyan-De´neux, M.; Hahn, F.; Boucharat, N.; Terzian, G. Eur. J. Inorg. Chem. 2000, 2000, 1771. (24) Chambrier, I.; Hughes, D. L.; Swarts, J. C.; Isare, B.; Cook, M. J. Chem. Commun. 2006, 3504. (25) Katoh, K.; Kajiwara, T.; Nakano, M.; Nakazawa, Y.; Wernsdorfer, W.; Ishikawa, N.; Breedlove, B. K.; Yamashita, M. J. Am. Chem. Soc., submitted for publication. (26) Kro¨ger, J.; Jensen, H.; Ne´el, N.; Berndt, R. Surf. Sci. 2007, 601, 4180. (27) Vitali, L.; Fabris, S.; Conte, A. M.; Brink, S.; Ruben, M.; Baroni, S.; Kern, K. Nano Lett. 2008, 8, 3364. (28) Gomez-Segura, J.; Diez-Perez, I.; Ishikawa, N.; Nakano, M.; Veciana, J.; Ruiz-Molina, D. Chem. Commun. 2006, 2866. (29) Isshiki, H.; Liu, J. Komeda, T. Manuscript to be published. (30) Gao, L.; Ji, W.; Hu, Y. B.; Cheng, Z. H.; Deng, Z. T.; Liu, Q.; Jiang, N.; Lin, X.; Guo, W.; Du, S. X.; Hofer, W. A.; Xie, X. C.; Gao, H. J. Phys. ReV. Lett. 2007, 99. (31) Takada, M.; Tada, H. Chem. Phys. Lett. 2004, 392, 265. (32) Takada, M.; Tada, H. Jpn. J. Appl. Phys. Part 1 2005, 44, 5332. (33) Deng, W.; Hipps, K. W. J. Phys. Chem. B 2003, 107, 10736. (34) Gopakumar, T. G.; Meiss, J.; Pouladsaz, D.; Hietschold, M. J. Phys. Chem. C 2008, 112, 2529. (35) Yazdani, A.; Eigler, D. M.; Lang, N. D. Science 1996, 272, 1921. (36) Okabayashi, N.; Konda, Y.; Komeda, T. Phys. ReV. Lett. 2008, 100, 217801. (37) Okabayashi, N.; Paulsson, M.; Ueba, H.; Konda, Y.; Komeda, T. Phys. ReV. Lett. 2010, 104, 077801.
JP101349V