Geometric, Electronic, and Optical Properties of a Superatomic

Jun 22, 2012 - Journal of Chemical Education · Journal of Chemical Information and Modeling .... The heterodimer consisting of halogen-like Sc@Si16 an...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/JPCC

Geometric, Electronic, and Optical Properties of a Superatomic Heterodimer and Trimer: Sc@Si16−V@Si16 and Sc@Si16−Ti@Si16−V@ Si16 Takeshi Iwasa and Atsushi Nakajima* JST-ERATO, Nakajima Designer Nanocluster Assembly Project, 3-2-1 Sakado, Takatsu-ku, Kawasaki 213-0012, Japan, and Department of Chemistry, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223-8522, Japan S Supporting Information *

ABSTRACT: The geometric, electronic, and optical properties of a heterodimer and trimer consisting of metal-encapsulating silicon cage clusters, M@Si16 (M = Sc, Ti, V) with D4d symmetry, are studied using density functional computations to explore the possibility of using these clusters as building blocks for a nanometer scale heteroassembly. In this study, among the possible low-lying geometries, the linearform of the hetero-oligomers is adopted as a model system, where the D4d monomers are covalently bonded by facing their squares in an eclipsed fashion. The heterodimer consisting of halogen-like Sc@Si16 and alkaline-like V@Si16 has a dipole moment of 7.63 D, and its occupied and virtual frontier orbitals are localized to V@Si16 and to Sc@Si16, respectively. Some of the inner molecular orbitals exhibit superatomic bonding and antibonding character. The electronic excitations involve charge-transfer states mainly from V@Si16 to Sc@Si16 in the optical energy region. The linear heterotrimer of Sc@Si16−Ti@Si16−V@Si16, formed by inserting the rare-gas-like Ti@Si16, has a larger dipole moment of 15.6 D and one or more localized frontier orbitals compared to the dimer. We propose possible formation routes to realize the present hetero-oligomers using photoexcitation or energy-selective electron injection into several LUMOs of the monomers that are suitable for linear-oligomerization. phase characterized.20−27 Using the density functional computations, the geometric and electronic properties for homogeneous aggregations of M@Si16 (M = Sc, Ti, V) have been extensively investigated by Torres et al.18,19 As for further advanced functionalities, heterogeneously integrated aggregations of these M@Si16 clusters are intriguing. With the aim of investigating the physicochemical properties of such heteroassemblies, we here report density functional computations for simplified model oligomers consisting of different species of M@Si16. In particular, we focus on a heterodimer consisting of halogen-like Sc@Si16 and alkaline-like V@Si16 because a dimer consisting of these species is neutral in the ground state, thereby not requiring additional counterions. In addition to the dimer, we also study a heterotrimer of Sc@ Si16−Ti@Si16−V@Si16, in which a rare-gas-like Ti@Si16 is inserted between Sc@Si16 and V@Si16 aiming at improving, for instance, the charge separations. For studying the electronic and optical properties of these high-symmetry oligomers, we adopt the choice of a linear form for the hetero-oligomers, such a structure has been found by Kumar et al. for the Zr@Si16 dimer12 and by Torres et al.18,19 for homogeneous aggregations of M@Si16. In the latter case, the authors showed that this

1. INTRODUCTION Assembling or combining different atoms, molecules, or materials having different functionalities into one structure is one of the most fundamental approaches in chemistry and material science to obtaining desired and novel properties. For example, a heteroassembly at the molecular level has been proposed and demonstrated to show rectification1,2 or a charge separation active layer for photovoltaic cells.3−6 In the semiconductor industry, heterojunctions, especially p−n junctions, serve as the foundation of modern electronic applications, including transistors, solar cells, and emitting diodes.7 For assembly at the nanometer scale, the use of multielement clusters with magic-number stability as a building block has attracted much attention because of their superatomlike spherical geometries, and tunable electronic properties obtained by varying the constituent atoms.8 To date, a number of mono- and heteroatomic clusters have been made and are known to show superatomic behavior, including C60,9 aluminum10 or noble-metal cluster compounds,11 and so on. Among them, we focus here on metal encapsulating silicon cage clusters, M@Si16 (M = Sc−, Ti, V+), as a building block for heteroassembly because of their isoelectronic magic-number behavior, and the fact that their sizes and high-symmetry geometries are similar.12−19 Indeed, a series of M@Sin clusters incorporating various endohedral metal atoms, ranging from transition metal to lanthanide atoms, have been synthesized and their electronic properties in the gas © 2012 American Chemical Society

Received: March 22, 2012 Revised: June 1, 2012 Published: June 22, 2012 14071

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077

The Journal of Physical Chemistry C

Article

linear form is not at the global energy minimum but is indeed one of the low-lying species, both for a dimer and a trimer.

2. COMPUTATIONAL DETAILS The geometry optimizations employing Kohn−Sham (KS) density functional computations are performed with TURBOMOLE 6.2 and 6.3,28,29 adopting the B3LYP30,31 functional and def-SV(P)32 basis sets having double-ζ valence-quality plus polarization, assuming the singlet states. The natural bond orbital (NBO) analysis33,34 and electronic absorption spectra for the heterodimer and trimer at the above-mentioned optimized geometries are computed using Gaussian0935 within the time-dependent (TD) KS linear response theory,36−39 employing 6-31G (d,p) basis sets and the CAM-B3LYP functional.40 In fact, it has been known that the calculations give a better description of charge-transfer excitations,41 which are also found for the dimer and trimer in this study. The line spectra of absorption peaks are convoluted by a Lorentz function whose width is taken to be 0.5 eV. To analyze absorption spectra with special emphasis on their charge transfer characters, the electronic transitions are separated in terms of KS orbitals to visualize the charge transfers between Sc@Si16 and V@Si16 monomers. Let us define that each excitation peak consists of several transitions between KS orbitals, for instance from orbital i to a with the coefficient xai , whose square is normalized to 1 when the sum runs over all the contributions to an excitation. At first, we perform the Mulliken population analysis and calculate the population of V@Si16 in orbital i, denoted as p(i). Using these values together with the kth absorption intensity I(k), the increase and decrease of the charge of V@Si16 can be expressed as pincrease = ∑a p(a)(xia ) 2 and pdecrease = ∑i p(i)(xia) 2 , respectively, where i, j,... and a, b,... denote occupied and virtual orbitals, respectively. The total increase in charge of V@ Si16 is calculated as pincrease − pdecrease, where a negative value means a gain of charge for Sc@Si16. Finally, these values are multiplied by I(k) and plotted over simulated UV−vis spectra.

Figure 1. Geometric structure of the heterodimer Sc@Si16−V@Si16. Each Si4 plane is alphabetically labeled.

approximately C4v symmetry with slight deviations around the open-square face of V@Si16, lowering the dimer symmetry to C2v within 0.01 au; as mentioned above, this form of the oligomers is taken in this article as the simplest model system for a heteroassembly of the M@Si16 clusters. We have confirmed that this structure is actually one of the low-lying isomers at our computational level. Details are provided in the Supporting Information, where geometric and energetic data are given for all the dimeric structures obtained in our computations. The interatomic distances and the Wiberg bond indices obtained from the NBO analysis are summarized in Table 1, Table 1. Interatomic Distances (Å) and Wiberg Bond Indices (In Parentheses) of the Heterodimer; See Figure 2 for the Atomic Labels Sia−Sia 2.31 (1.02) Sid−Sid 2.50 (0.72) Sif−Sig 2.36 (0.92) Sc−Sia 3.10 (0.22) V−Sie 3.36 (0.09)

3. RESULTS AND DISCUSSION The results of the NBO analysis and the electronic and optical properties presented here are obtained with single-point and response calculations at the level of CAM-B3LYP/6-31G(d,p) using Gaussian09 at the optimized geometries obtained at the B3LYP/def-SV(P) level using TURBOMOLE 6.3.1. While the Wiberg bond indices42 obtained by the NBO analysis computed with these two types of functionals show negligible differences, the differences in the electronic properties are notable. The dipole moments of the heterodimer, for example, obtained with CAM-B3LYP and B3LYP are 7.6 and 5.7 D, respectively. Furthermore, the occupied KS orbital energies are about 1−2 eV lower and HOMO−LUMO gap energies are about 1 eV larger with CAM-B3LYP than those with B3LYP. Comparisons of the density of states and absorption spectra for the dimer and trimer with these functionals are given in the Supporting Information. 3.1. Geometric Structure of the Heterodimer: ScSi16− VSi16. Figure 1 shows the optimized structure of the heterodimer Sc@Si16−V@Si16, where all the Si4 planes are alphabetically labeled for later use. The constituent D4d monomers have two square faces facing each other and eight pentagon faces, and in the dimer, the two monomers bond to each other by facing their squares. The dimer possesses

Sia−Sib 2.30 (1.08) Sid−Sie (interface) 2.45 (0.80) Sig−Sih 2.30 (1.01)/ 2.32 (0.98) Sc−Sib 3.03 (0.34) V−Sif 3.10 (0.38)

Sib−Sic 2.41 (0.87) Sie−Sie 2.39/2.40 (0.86) Sih−Sih 2.32 (1.00)

Sic−Sid 2.34 (0.97) Sie−Sif 2.35 (0.96)

Sc−Sic 2.86 (0.40) V−Sig 2.58 (0.52)/ 2.64 (0.46)

Sc−Sid 2.87 (0.32) V−Sih 2.59 (0.59)/ 2.78 (0.52)

where Sia−Sib, for example, indicates the bond between the nearest neighboring Si atoms belonging to two Si4 planes labeled as a and b in Figure 1. The Si−Si bonds are of mixed sp2−sp3 character, as reported for Zr@Si16 by Kumar et al.,43 but all the Si−Si bonds are single according to the Wiberg bond index, while the M−Si bonds are half for the dimer, as well as for Ti@Si16. While the Sc and V atoms in the D4d monomers are located almost at the center of the cage and the distance between two metal-to-square faces are equal, in the dimer, these atoms shift toward the positive z direction, resulting in these two metal-to-square face distances being different: the Sc−Sia and V−Sie bond lengths are longer than Sc−Sic,d and V−Sig,h by more than 0.2 Å. Except for the h-Si4 plane (see Figure 1), the four Si atoms in the Si4 planes are bonded almost equivalently with negligible deviations; for example, the bond length deviations of Sie−Sie and Sig−Sih are about 0.02 Å at maximum. The h-Si4 plane bends from the square due to the interaction with V−Sih, where two of these bond lengths are 2.59 Å and the other two are 2.78 Å, lowering the symmetry of the dimer from 14072

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077

The Journal of Physical Chemistry C

Article

C4v to C2v. The vanadium atom interacts rather strongly with four of the eight Si atoms in the g and h planes with bond lengths of 2.58−2.59 Å, and these four Si atoms are arranged in a distorted tetrahedron. The rest of the Si atoms interact rather weakly with the V atom, with bond lengths of 2.64−2.78 Å. The bond lengths for Sc−Sia and V−Sie are more than 3 Å, suggesting weak, or even negligible, interaction between M and Si atoms. 3.2. Electronic and Optical Properties of the Heterodimer of ScSi16−VSi16. Figure 2a shows the inner shell

with different orbital energies. This can be ascribed to the weak spatial overlaps of Px and Py compared to the Pz, which has a sigma character and thus a large spatial overlap. The frontier orbitals from HOMO − 2 to LUMO + 2 are shown in Figure 2b. The HOMO−LUMO gap of the heterodimer is 2.80 eV, implying a high stability of the dimer. In contrast to the inner orbitals, these KS orbitals are localized on either monomer. This indicates that the frontier orbitals of the heterodimer can be described well by those of the corresponding monomers even when they are bonded. The charge transfer from Sc@Si16 to V@Si16 calculated by summing the natural charges is 0.15, and the natural charges of the Sc and V atoms are −0.29 and 0.10, respectively. Due to the charge separation, Sc@Si16 and V@Si16 are positively and negatively charged, respectively, and thus, the heterodimer has a large dipole moment of 7.63 D. It should be noted that the dimer is dissimilar to an ionically bonded diatomic counterpart, such as NaCl, in which these atoms are approximately described as Na+Cl−. In sharp contrast, as seen in Figure 2a, the bond between Sc@Si16 and V@Si16 is covalent, which is also suggested by the Wiberg indices shown in Table 1. As reported previously, however, ionic forms of Sc@Si16− and V@ Si16+ complete the electronic closing, forming isoelectronic superatoms similar to the magic-numbered Ti@Si16 cluster. Irrespective of the preference for electronically ionic species, the heterodimer consists of covalent bonds because the monomers have similar orbital energies. According to the calculation for the neutral species of Sc@Si16/V@Si16, with fully relaxed geometries (Cs/C1 symmetries) starting from the D4d geometries, the orbital energies for the HOMO and LUMO obtained by B3LYP/def-SVP are −5.68/−5.65 and −4.45/− 4.01 eV, respectively. Owing to the large dipole moment, it is expected that the heterojunction composed of Sc@Si16 and V@Si16 can provide a strong dipole layer. As for the further application of the M@Si16 clusters as building blocks for optoelectronic devices, let us consider the heterodimer as a p−n junction where Sc@Si16 and V@Si16 are regarded as being p- and n-type materials. At conventional p−n interfaces, an electron usually transfers from the positive layer to the negative layer to satisfy the electronic shell closing in both the p and n layers, creating what is called the depletion region. Interestingly, this p−n junction of the C4v dimer, Sc@Si16 −V@Si 16 , can work without the usual intermediate insulator layer. In the present geometry shown in Figure 1, the heterodimer of Sc@Si16 and V@Si16 are almost in their neutral states and the frontier orbitals of the dimer are not delocalized in the heterodimer. These results suggest a promising possibility of these monomers being used to produce a cluster-based p−n junction. Figure 2c shows the density of states (DOS) of the heterodimer. In order to focus on the cluster assembly, the DOS is shown separately for the Sc@Si 16 and V@Si 16 monomers. As shown in Figure 2b, the HOMO − 2 to LUMO + 2 region consists almost exclusively of the molecular orbitals of each monomer. Owing to the localization in the KS orbitals of the heterodimer, photoexcitation can induce a clear charge transfer from V@Si16 to Sc@Si16 over a wide range of the absorption spectrum, as shown in Figure 2d. Here, the colored areas indicate the amount of charge transfer. The red area indicating the charge transfer from V@Si16 to Sc@Si16 is large, while the opposite direction of charge transfer, indicated in green, can hardly be seen. When the charge transfer from V@Si16 to Sc@Si16 is induced by photoexcitation, each

Figure 2. (a) Inner KS orbitals that has superatomic S-, Pz-, and Dz2like spatial distributions; (b) the occupied and virtual frontier orbitals from HOMO − 2 to LUMO + 2 that are, respectively, localized on V@Si16 and Sc@Si16; (c) the density of states in which black arrows indicate HOMO and LUMO; and (d) the absorption spectrum of the heterodimer. The charge-transfer rates from V@Si16 to Sc@Si16 (red) and in the opposite direction (green) are superimposed on the absorption spectrum.

superatomic-like KS orbitals of the heterodimer having bonding and antibonding characters of S, Pz, and Dz2, where the z-axis is defined by the Sc and V atoms. The S and P orbitals are about 13 eV below the HOMO, while the D orbitals are about 10 eV below; the bonding and antibonding orbital energies are S (−20.36 and −19.96 eV), Pz (−19.41 and −18.38 eV), and Dz2 (−17.32 and −16.16 eV). The makeup of these orbitals indicates the superatomic nature of M@Si16 clusters. The π superatomic bonding or antibonding orbitals such as Px−Px are not found, whereas Px and Py are localized on either monomer 14073

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077

The Journal of Physical Chemistry C

Article

indices. Interestingly, the results to the heterodimer similarly hold for the heterodimer. In the optimized structure, the titanium atom also shifts toward the h-Si4 plane, and the Ti− Sig,h bond distances are 2.75 and 2.66 Å, respectively. These values are intermediate between those of Sc−Si and V−Si bonds. Intercluster bond lengths seem to be insensitive to the difference in the endohedral atom, i.e., Sid−Sie (2.44 Å) and Sih−Sii (2.43 Å). The four atoms belonging to each Si4 plane are also equivalent for Ti@Si16, and the C4v symmetry is broken only at the edge of the V@Si16, i.e., Sik and Sil planes. Figure 4a,b shows the frontier orbitals from HOMO − 5 to LUMO + 5 and the DOS, respectively. As in the case for the

constituent monomer can satisfy the electronic shell closings, implying that these charge-transfer excited states can work as a good charge separator. Namely, when the adjacent layers of V@ Si16 and Sc@Si16 are formed, the cluster assembled layers have the vital functionality of a photoswitching device. It should be noted that the photoinduced charge separation has been observed over a wide energy range only for one dimer unit. In other words, the heterodimer itself exhibits the ability of efficient charge separation without the need for doping atoms or molecules. 3.3. Geometric, Electronic, and Optical Properties of the Heterotrimer: Sc@Si16−Ti@Si16−V@Si16. As mentioned above, the calculations suggest that the heterodimer of V@ Si16−Sc@Si16 can work as a p−n junction having a large dipole moment. Thus, a natural question arises whether the electronic properties can be improved or not by inserting an insulating layer, similar to those used in silicon-based devices. Among the M@Si16, the rare-gas like Ti@Si16 is stable in its neutral state, and thus, it can be regarded as an insulating unit in the heterodimer-assembly. To examine this possibility via a model calculation, we have optimized the geometric structure for the Sc@Si16−Ti@Si16−V@Si16 trimer, starting from D4d monomers arranged one-dimensionally with the same bonding motif as the heterodimer. The optimized geometric structure is shown in Figure 3, and Table 2 summarizes the bond lengths and the Wiberg bond

Figure 3. Geometric structure of the heterotrimer Sc@Si16−Ti@Si16− V@Si16. See Figure 1 for the alphabetical labels.

Table 2. Interatomic Distances (Å) and Wiberg Bond Indices (In Parentheses) of the heterotrimer; See Figure 4 for the Atomic Labels Sia−Sia 2.31 (1.02) Sid−Sie (interface) 2.44 (0.82) Sih−Sih 2.51 (0.69) Sik−Sil 2.30 (1.01)/ 2.32(0.98) Sc−Sia 3.11 (0.22) Ti−Sie 3.57 (0.05) V−Sii 3.40 (0.08)

Sia−Sib 2.30 (1.08) Sie−Sie

Sib−Sic 2.40 (0.87) Sie−Sif

Sic−Sid 2.33 (0.97) Sif−Sig

Sid−Sid 2.51 (0.72) Sig−Sih

2.38 (0.88) Sih−Sii (interface) 2.43 (0.82)

2.35 (0.95) Sii−Sii

2.35 (0.95) Sii−Sij

2.34 (0.94) Sij−Sik

2.39 (0.86/0.87)

2.35 (0.95)

2.36 (0.92/0.93)

Sc−Sic 2.87 (0.40) Ti−Sig 2.75 (0.48) V−Sik 2.57 (0.52)/ 2.63 (0.46)

Sc−Sid 2.86 (0.32) Ti−Sih 2.66 (0.46) V−Sil 2.59 (0.60)/ 2.76(0.54)

Figure 4. (a) Occupied and virtual frontier orbitals from HOMO − 5 to LUMO + 5 for the Sc@Si16−Ti@Si16−V@Si16 trimer; (b) the density of states; and (c) the absorption spectrum of the heterotrimer. The charge-transfer rates from V@Si16 to Sc@Si16 (red) and in the opposite direction (green) are superimposed on the absorption spectrum.

dimer, the DOS is divided into the contributions from the monomers, using red (Sc@Si16), green (Ti@Si16), and blue (V@Si16) colors. Comparing with the heterodimer, one or two frontier orbitals localized on either Sc@Si16 or V@Si16 are found, as well as the orbitals localized on Ti@Si16. The HOMO−LUMO gap is 2.44 eV, which is ca. 0.4 eV lower than that of the heterodimer but still a large value. The natural charges of Sc, Ti, and V are 0.29, 0.09, and 0.10, respectively. The summed charges for Sc@Si16, Ti@Si16, and V@Si16 are 0.22, −0.01, and −0.20, respectively, and the dipole moment is 15.6 D. By inserting Ti@Si16 between Sc@Si16 and V@Si16, therefore, considerable improvements are found in the localizations in the frontier orbitals, the summed charges for

Sil−Sil 2.32 (1.00) Sc−Sib 3.02 (0.34) Ti−Sif 3.07 (0.38) V−Sij 3.10 (0.38)

14074

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077

The Journal of Physical Chemistry C

Article

Sc@Si16 and V@Si16, and the dipole moment, suggesting that Ti@Si16 can play a role as an insulating layer. From these results, a stable and strong dipole layer can be expected by inserting Ti@Si16 in between the Sc@Si16 and V@Si16 layers, just as an insulating layer is inserted between p- and n-type materials. Figure 4c shows the absorption spectrum of the heterotrimer, where the charge-transfer rates from V@Si16 to Sc@Si16 (red) and the opposite (green) are also shown. Comparing to the dimer, in this energy range below 3 eV, the charge transfer is a one-way process from V@Si16 to Sc@Si16 because there are no opposite processes indicated by green in Figure 4c. The electronic excitation at 1.91 eV consists of two components of charge transfers from V@Si16 to Sc@Si16 and excitations localized on Sc@Si16, with a ratio of approximately 1:1. Similarly, the electronic excitation at 2.37 eV consists of electronic transitions localized on Sc@Si16 (∼12%) and Ti@ Si16 (∼24%), although the localized character is less prominent than that of the 1.91 eV excitation. Localized excitations within the insulating layer (Ti@Si16) might possibly be utilized as photoswitching conductance mechanism through a p−i−n junction consisting of the M@Si16 clusters. 3.4. On the Oligomerization. Finally, let us mention possible routes to experimentally realize the heterodimer and trimer modeled here with a one-dimensional geometry where the constituent D4d monomers are covalently connected through their Si4 square faces in an eclipsed fashion. Since the M@Si16 clusters have been synthesized in the gas-phase with a magic-numbered behavior,20 heterojunctions can plausibly be fabricated by sequential vapor depositions of different clusters onto a substrate, which would result in the coexistence of different cluster islands or as thin layers. The heterodimer would probably be found at a domain boundary of islands or at an interface of layered structures that could be fabricated with the aid of well-controlled sequential depositions of Sc@Si16 and V@Si16. Similarly, the heterotrimer might also be found at an interface when an insulating monolayer of Ti@ Si16 is inserted between layers of Sc@Si16 and V@Si16. Since the sizes of the constituent superatomic clusters are similar to each other, the sequential depositions seem to be plausible routes to provide the correct local structure of linear oligomers. As a matter of fact, however, the formation of a covalent bond between the constituent monomers is required to achieve the present oligomers, which generally needs a certain amount of energy in the form of heat or light. Here, we propose that the covalent bonds for the linear-oligomerization at the interface of the heteroassembly can possibly be made by laser irradiation or energy-selective electron injection to suitable unoccupied orbitals. To investigate this possibility, we have first confirmed that the three M@Si16 monomers actually have the frontier orbitals that are suitable for linear-oligomerization. As an example, the HOMO − 5 to LUMO + 5 region of Ti@Si16 are shown in Figure 5a. LUMO + 1, 2, 4, and 5 seem to be suitable for linearoligomerization because these orbitals elongate along the longaxis of the oligomers, i.e., normal to the Si4 planes, not the Si5 planes. Second, Figure 5b shows absorption spectra for the different monomers, computed using TURBOMOLE at the B3LYP/def-SV(P) level. The line spectra in red are calculated by summing up the contributions of each electronic transition whose final states have the suitable spatial distributions for linear-oligomerization, as shown in Figure 5a, and then, the summed contribution is multiplied by the original line spectral

Figure 5. (a) Frontier orbitals from HOMO − 5 to LUMO + 5 of Ti@Si16. (b) Lorentzian convoluted absorption spectra (black) and the original line spectra (green) of the M@Si16 monomers (M = Ti, Sc−, V+, Sc, and V), together with the top and side views for the monomers. The red line spectra show the electronic transitions whose final states seem to be suitable for the oligomerization in a linear-form. These final states are indicated by the black arrows in panel a.

intensities, shown in green in Figure 5b. The red spectra suggest that the linear-oligomerization is achievable in the optical energy region from 1 to 3 eV. Some absorption peaks consist mainly of these transitions. Thus, we expect that, after sequential vapor depositions, the desired oligomerizations could be triggered either by optical excitations at appropriate wavelengths that can induce dimerizations or by electron injections to suitable unoccupied states using scanning tunneling microscopy, as reported in the case of C 60 oligomerization.44 14075

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077

The Journal of Physical Chemistry C

Article

(4) Nakamura, E.; Isobe, H. Acc. Chem. Res. 2003, 36, 807−815. (5) Imahori, H.; Fukuzumi, S. Adv. Funct. Mater. 2004, 14, 525−536. (6) Deibel, C.; Dyakonov, V. Rep. Prog. Phys. 2010, 73, 096401−1− 096401−39. (7) Sze, S. M.; Ng, K. K. Physics of Semiconductor Devices, 3rd ed.; Wiley-Interscience: Hoboken, NJ, 2007. (8) Claridge, S. A.; Castleman, A. W.; Khanna, S. N.; Murray, C. B.; Sen, A.; Weiss, P. S. ACS Nano 2009, 3, 244−255. (9) Feng, M.; Zhao, J.; Petek, H. Science 2008, 320, 359−362. (10) Bergeron, D. E.; Castleman, A. W.; Morisato, T.; Khanna, S. N. Science 2004, 304, 84−87. (11) Walter, M.; Akola, J.; Lopez-Acevedo, O.; Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Whetten, R. L.; Grönbeck, H.; Häkkinen, H. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 9157−9162. (12) Kumar, V.; Kawazoe, Y. Phys. Rev. Lett. 2001, 87, 045503−1− 045503−4. (13) Reveles, J. U.; Khanna, S. N. Phys. Rev. B 2006, 74, 035435−1− 035435−6. (14) Koyasu, K.; Atobe, J.; Akutsu, M.; Mitsui, M.; Nakajima, A. J. Phys. Chem. A 2007, 111, 42−49. (15) Torres, M. B.; Fernández, E. M.; Balbás, L. C. Phys. Rev. B 2007, 75, 205425−1−205425−12. (16) Furuse, S.; Koyasu, K.; Atobe, J.; Nakajima, A. J. Chem. Phys. 2008, 129, 064311−1−064311−6. (17) Lau, J. T.; Hirsch, K.; Klar, P.; Langenberg, A.; Lofink, F.; Richter, R.; Rittmann, J.; Vogel, M.; Zamudio-Bayer, V.; Möller, T.; et al. Phys. Rev. A 2009, 79, 053201−1−053201−5. (18) Torres, M. B.; Fernández, E. M.; Balbás, L. C. Int. J. Quantum Chem. 2011, 111, 444−462. (19) Torres, M. B.; Fernández, E. M.; Balbás, L. C. J. Phys. Chem. C 2011, 115, 335−350. (20) Koyasu, K.; Akutsu, M.; Mitsui, M.; Nakajima, A. J. Am. Chem. Soc. 2005, 127, 4998−4999. (21) Koyasu, K.; Atobe, J.; Furuse, S.; Nakajima, A. J. Chem. Phys. 2008, 129, 214301−1−214301−7. (22) Beck, S. M. J. Chem. Phys. 1987, 87, 4233−4234. (23) Beck, S. M. J. Chem. Phys. 1989, 90, 6306−6312. (24) Zheng, W.; Nilles, J. M.; Radisic, D.; Bowen, K. H. J. Chem. Phys. 2005, 122, 071101−1−071101−4. (25) Jaeger, J. B.; Jaeger, T. D.; Duncan, M. A. J. Phys. Chem. A 2006, 110, 9310−9314. (26) Janssens, E.; Gruene, P.; Meijer, G.; Wöste, L.; Lievens, P.; Fielicke, A. Phys. Rev. Lett. 2007, 99, 063401−1−063401−4. (27) Cantera-López, H.; Balbás, L. C.; Borstel, G. Phys. Rev. B 2011, 83, 075434−1−075434−9. (28) TURBOMOLE, V6.2 2010 and V6.3 2011; University of Karlsruhe: Karlsruhe, Germany, 2009; available from http://www. turbomole.com. (29) Ahlrichs, R.; Bär, M.; Häser, M.; Horn, H.; Kölmel, C. Chem. Phys. Lett. 1989, 162, 165−169. (30) Lee, C.; Yang, W. T.; Parr, R. G. Phys. Rev. B 1988, 37, 785− 789. (31) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (32) Schäfer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571−2577. (33) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735−746. (34) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88, 899−926. (35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;

4. CONCLUDING REMARKS In summary, the geometric, electronic, and optical properties of a heterodimer and trimer consisting of the D4d monomers M@ Si16 (M = Sc, Ti, V), covalently bonded by facing their squares in an eclipsed fashion, have been studied using density functional computations to explore the possibility of heteroassembled clusters as a building block at the nanometer scale. Both the heterodimer of Sc@Si16−V@Si16 and the heterotrimer of Sc@Si16−Ti@Si16−V@Si16 show large HOMO−LUMO gaps (2.8 and 2.4 eV), superatomic-like inner orbitals, and frontier orbitals localized on Sc@Si16 and V@Si16 for the virtual and occupied states. Together with the large dipole moments (7.6 and 15.6 D), the charge transfer states in electronic excitation spectra from V@Si16 to Sc@Si16 suggest practical applications for the construction of a strong dipole layer, p−n or p−i−n junctions, and photovoltaic cells. Possible routes exist for realizing the linear-oligomers using photoexcitations or energy-selective electron injections to proper LUMOs whose spatial distribution are suitable for linear-oligomerization of the D4d monomers that possibly exist as one of the energetically low-lying isomers. The electronic states of both the dimer and trimer feature spatially separated frontier orbitals in the ground states and charge-transfer states in the electronic excited states, which result in the electronic spectra of the hetero-oligomers lying in the visible region. Since the constituents of M@Si16 strongly resemble each other with high-symmetry geometries, the close-packed assembly of these superatomic monomers might open new avenues in nanocluster-assembled device engineering for electronics and optics. In the future, the electronic and optical properties of hetero-oligomers in other forms, or consisting of other monomers with different endohedral atoms, will be studied, as well as their bulk phases and the dynamical aspects of their charge transfer states.



ASSOCIATED CONTENT

S Supporting Information *

Geometric structures, total energies, and HOMO−LUMO gaps of the low-lying isomers for the monomers and the dimer along with considered initial geometries for the dimer, as well as the density of states and absorption spectra for the dimer and trimer computed with B3LYP and CAM-B3LYP. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work is partly supported by the Science Research Promotion Fund from the Promotion and Mutual Aid Corporation for Private Schools of Japan. The computations were partly performed at the Research Center for Computational Science, Okazaki, Japan.



REFERENCES

(1) Aviram, A.; Ratner, M. A. Chem. Phys. Lett. 1974, 29, 277−283. (2) Reed, M. A.; Zhou, C.; Muller, C. J.; Burgin, T. P.; Tour, J. M. Science 1997, 278, 252−254. (3) Hiramoto, M.; Fujiwara, H.; Yokoyama, M. Appl. Phys. Lett. 1991, 58, 1062−1064. 14076

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077

The Journal of Physical Chemistry C

Article

Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision B.01; Gaussian, Inc.: Wallingford, CT, 2009. (36) Casida, M. E. In Recent Advances in Density Functional Methods Part I; Chong, C. P., Ed.; World Scientific: Singapore, 1995; p 155. (37) Bauernschmitt, R.; Ahlrichs, R. Chem. Phys. Lett. 1996, 256, 454−464. (38) Bauernschmitt, R.; Häser, M.; Treutler, O.; Ahlrichs, R. Chem. Phys. Lett. 1997, 264, 573−578. (39) Furche, F. J. Chem. Phys. 2001, 114, 5982−5992. (40) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393, 51−57. (41) Jacquemin, D.; Perpéte, E. A.; Scuseria, G. E.; Ciofini, I.; Adamo, C. J. Chem. Theory Comput. 2008, 4, 123−135. (42) Wiberg, K. B. Tetrahedron 1968, 24, 1083−1096. (43) Kumar, V.; Majumder, C.; Kawazoe, Y. Chem. Phys. Lett. 2002, 363, 319−322. (44) Nakaya, M.; Aono, M.; Nakayama, T. ACS Nano 2011, 5, 7830− 7837.

14077

dx.doi.org/10.1021/jp302752g | J. Phys. Chem. C 2012, 116, 14071−14077