ARTICLE pubs.acs.org/JPCC
First-Principle Study on Structural and Electronic Properties of Pristine and Adsorbed LiF Nanotubes Su-Fang Wang,† Li-Yong Chen,‡ Yan Zhang,§ Jian-Min Zhang,*,† Vincent Ji,§ and Ke-Wei Xu|| †
College of Physics and Information Technology, Shaanxi Normal University, Xian 710062, Shaanxi, P. R. China Shaanxi Key Laboratory of Condensed Matter Structures and Properties, Northwestern Polytechnical University, Xian 710129, Shaanxi, P. R. China § ICMMO/LEMHE UMR CNRS 8182, Universite Paris-Sud 11, 91405 Orsay Cedex, France State Key Laboratory for Mechanical Behavior of Materials, Xian Jiaotong University, Xian 710049, Shaanxi, P. R. China
)
‡
ABSTRACT: A systematic density functional theoretical study of the single-walled LiF nanotubes (NTs) in armchair (n,n) and zigzag (n,0) (2 < n < 11) configurations is presented. Full geometry and spin optimizations with unrestricted symmetry have been performed. Both zigzag and armchair LiF NTs can exist stably because of their large binding energies that are slightly smaller than the bulk’s. Furthermore, the insulating character is observed for them. Given single transition-metal (TM) atom adsorbed stable (8,0) LiF NT, it is found that TM atom almost locates on the top site of the nearest F atom, regardless of the initial site. Dramatically, the half-metallic character is obtained for V-, Cr-, and Mn-adsorbed (8,0) LiF NTs. The results suggest that these systems could be used in the field of quasi-1D spintronics devices for producing nearly 100% spin polarized currents.
1. INTRODUCTION Recent years have seen an explosive interest in the studies of 1D and quasi-1D nanostructures, such as nanotubes (NTs) and nanowires (NWs), since the discovery of carbon nanotubes (CNTs) by Iijima and coworkers.1,2 Owing to its unique quasi-1D atomic structure and superb mechanical and electronic properties, the single-walled CNT has been playing a significant role in emerging nanotechnology3 and becoming the important building blocks for nanoelectronic applications,46 chemical and biological sensing,7,8 and nanocomposites.9 Besides CNTs, a considerable number of different composite nanoscale tubular structures has either already been fabricated, based on crystals such as BN and SiC,1013 or suggested, as in the case of BeO14 and MgO15,16 NTs. Carbon, SiC, BN, and BeO NTs differ in their increasing ionicity. As a consequence, the properties of SiC, BN, and BeO NTs are different from those of the covalently bonded, homopolar CNTs.17 Most notably, the first row elements of the periodic table also include lithium (Li) and fluorine (F), except the above-mentioned elements. Lithium fluoride (LiF) in bulk form can be widely used in honing glass, copper, and aluminum soldering process and saltmelted chemical process as a menstruum, recommended as a space technology storage solar radiation heat agent, used in aluminum electrolysis and metallurgical industry. High-pure LiF could be used in making fluoride glass and producing spectrometer and X-ray monochromator prisms. Unfortunately, to the best of our knowledge, there have been no reports on whether the singlewalled LiF NTs exist and no reports on the stability and application of LiF NTs and the metal decoration of single-walled LiF NTs. r 2011 American Chemical Society
Theoretically, density functional theories (DFTs) are very successful techniques for understanding the electronic, structural, and vibrational properties of large molecular systems. The VASP is an efficient DFT code developed recently1821 for studying 3D bulk systems with periodic boundary condition. DFT within local density approximation (LDA) and generalized gradient approximation (GGA) have been used for CNTs and 3d TMs chain or nanowires encapsulated inside CNTs during the past decade,2224 for BN NTs and 3d TMs chain or nanowires encapsulated in and adsorbed on BN NTs during the past decade,2529 for SiC NTs and defects in SiC NTs in the last couple of years,3035 and even for BeO NTs.14 CNTs have been found to be either metallic or semiconducting depending on their helicity;3638 both BN and SiC NTs are semiconducting, and their structural as well as electronic properties depended in characteristic ways on the chirality and the diameter of the tubes.2535 In contrast, the band gap progression in zigzag and armchair BeO NTs shows a very peculiar behavior for small diameters. No band gap breakdown occurs, and the gap goes through a minimum for zigzag BeO NTs.17 Most notably, the first known example is that of BN NTs,10 which were synthesized after theoretical prediction.26,39 We hasten to point out that to the best of our knowledge no ab initio study has been reported in the literature on LiF NTs. It is expected that the theoretical research could Received: May 14, 2011 Revised: November 27, 2011 Published: December 12, 2011 1650
dx.doi.org/10.1021/jp2045146 | J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C
ARTICLE
Table 1. Calculated Lattice Parameters (Å), Elastic Constants (GPa), and Bulk Modulus (GPa) of the Bulk LiF Compared with the Experimental Results elastic constants (GPa) bulk modulus
lattice parameters (Å) this work
4.025
C11
C12
C44
114.251 45.813 33.046
experiment 4.017,43,44 4.027945 11245
45.645
3245
(GPa) 62.051 62.0345
also play an important role in the development of the experimental study. In this Article, the structural and electronic properties of the single-walled pristine and TM atoms (TM = V, Cr, Mn, and Fe) adsorbed LiF NTs have been investigated by using the first-principles projector-augmented wave (PAW) potential within the DFT framework under the GGA. The rest of the Article is organized as follows. In Section 2, the calculation method and models of both the zigzag (3,0) and armchair (3,3) LiF NTs are given. In Section 3, the band structures, the density of states (DOS), and the charge density of single-walled pristine and TM atoms (TM = V, Cr, Mn, and Fe) adsorbed LiF NTs with spin-polarized calculations are analyzed and discussed. Finally, the conclusions of the work are presented in Section 4.
2. COMPUTATIONAL METHOD AND MODELS The calculations are implemented with the VASP1821 within the framework of DFT using the projector-augmented-wave (PAW)40 potential and a plane-wave basis set. To treat electron exchange and correlation, we chose the PerdewBurkeErnzerhof41 formulation of the GGA, which yields the correct ground-state structure of the combined systems. The cutoff energy for the plane waves is chosen to be 400 eV, which is much higher than the cutoff energies provided for the elements considered here in the PAW potentials. We set the periodicity of the LiF NTs along the tube axis direction and the 16 Å vacuum spaces perpendicular to the tube axis direction to eliminate the interaction between periodic images. The Brillouin zone integration is performed by using the gamma-centered Monkhorst-Pack scheme.42 The force convergence criterion is set to be 0.02 eV/Å, and all the geometric structures of LiF NTs are fully relaxed to minimize the total energy of the system until a precision of 104eV is reached. The 2s1 and 2s2p5 electrons are taken as the valence electrons for Li and F atoms, respectively. For 3d TM atoms, the 4s23d3, 4s13d5, 4s23d6, and 4s23d7 electrons are taken as the valence electrons for V, Cr, Mn, and Fe atoms, respectively. To test the quality of the above parameters, we have calculated the lattice parameters, elastic constants, and bulk modulus of the bulk LiF, and the calculated results are listed in Table 1 together with the experimental results for comparison. It can be seen that the calculated results are in good agreement with the experimental results. The starting geometries of these tubular structures have been obtained by simply rolling of graphene like sheet of Li and F atoms placed at different nodes of the honeycomb lattice to form a tube. The structure of an NT may be described entirely in terms of the length and chirality. The chirality and diameter are uniquely defined in terms of the magnitude of the components of the chiral vector C Bh = na B1 + ma B2 (n,m), where n,m are integers and B a 1,a B2 are the unit vectors of a hexagonal,
Figure 1. Left (left panel) and side (right panel) views of both (a) the zigzag (3,0) and (b) armchair (3,3) LiF NTs, in which the green and gray balls denote Li and F atoms, respectively.
Bh connects graphene-like sheet.46 Therefore, the chiral vector C two crystallographically equivalent sites on the sheet. The particular chiralities known as “zigzag” and “armchair” are named for the cases of m = 0 and m = n, respectively. The term “chiral” is used for all other values of n and m. Here we have undertaken a systematic study of LiF NTs in the zigzag and armchair structures with n from 3 to 10. Zigzag and armchair LiF NTs have been chosen because of their structural simplicity and suitability as limiting cases in a broader study of chiral tubes. In the zigzag (n,0) tubes, the two sides of each hexagons are parallel to the tubular axis, whereas in armchair (n,n), the two sides of each hexagons are perpendicular to the tube axis.47 Figure 1a,b shows the left (left panel) and side (right panel) views of the initial geometric structures of both the zigzag (3,0) and armchair (3,3) LiF NTs as examples, in which the green and gray balls denote Li and F atoms, respectively.
3. RESULTS AND DISCUSSION 3.1. Pristine LiF NTs. The smallest zigzag and armchair tubes studied here are (3,0) and (3,3) tubes, and the largest structures are (10,0) and (10,10), respectively. All LiF NTs studied here have been fully relaxed to minimize their energies. To determine the stability of the single-walled LiF NTs with zigzag and armchair configurations, we have calculated the atomic binding energy for every system. The atomic binding energy, Eb, is defined as
Eb ¼
aEðLiÞ þ bEðFÞ-EðLia Fb Þ ða þ bÞ
ð1Þ
where a and b are the numbers of Li and F atoms in the per unit cell, respectively. E(Li), E(F), and E(LiaFb) are the spinoptimized ground-state total energies of the isolated Li atom, F atom, and LiF NTs, respectively. For all tubes, the ground state is a singlet, indicating the absence of both any unpaired electron and any magnetic behavior. Figure 2a displays the atomic binding energy Eb as a function of the tube diameter for the zigzag and armchair LiF NTs. Dramatically, for both zigzag and armchair tubes, the atomic binding energies tend to increase monotonically in Figure 2a. This general monotonous increased trend in binding energy is also noticed in single-walled SiGe NTs48 and armchair GeC NTs.49 It is well known that distances between the atoms play a major role in stabilization, and as the tube diameter increases, both short- and long-range atomic interactions possibly help stabilize the infinite NTs. Also, the larger diameter tube surfaces are more planar in comparison with highly curved smaller diameter tubes; this flatness of surface reduces the σπ hybridizations, increasing the stability of larger diameter NTs.48,49 Another important feature of Figure 2a is that the energy relationship is nonlinear. This indicates that the binding 1651
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C
ARTICLE
Figure 2. (a) Binding energy, (b) bond length, (c) radial buckling, and (d) band gap for the zigzag (n,0) (black space) and armchair (n,n) (red circle) LiF NTs. The solid lines are drawn to guide the eye.
energy is dependent not only on the tube diameter but also on aspects affected by the diameter such as the curvature of the NT.46 For the zigzag (n,0) LiF NTs that are chosen in our study, the smallest value of binding energy is found to be 4.032 eV/ atom for the (3,0) tube, and the largest is 4.108 eV/atom for the (10,0), and thus the average value of the atomic binding energy is estimated to be 4.070 eV/atom. In a similar way, for the armchair (n,n) LiF NTs, the average value of the binding energy is estimated to be 4.096 eV/atom. It is noticed that the difference in the atomic binding energy between the zigzag structure and the armchair structure is just ∼0.026 eV/atom. Therefore, we think that the zigzag (n,0) and the armchair (n,n) LiF NTs possess uniform stabilization for a given value of n. Also, the LiF flat film (sheet) shown in Figure 3a having the same geometry as the LiF NT walls has been considered here. The optimized length of the LiF bond is 1.82 Å. Figure 3b shows the calculated energy band with majority spin and minority spin of the LiF flat film (sheet). One readily identifies the symmetry in the proximity of the Fermi level between majority spin and minority spin for the energy bands. This indicates that the LiF flat film (sheet) has no spin splitting in the proximity of the Fermi level and thus no magnetic moment. A large band gap up to ∼7 eV is observed, indicating the insulating character for the LiF flat film (sheet). Furthermore, the obtained atomic binding energy of ∼4.082 eV is similar to the average value for the zigzag and armchair LiF NTs. The calculated atomic binding energies reveal that the formation processes of both LiF NTs and flat film (sheet) are exothermic, even if they are slightly smaller than the value of the bulk LiF (4.231 eV), implying the feasibility of the LiF NTs.
Figure 3. (a) Optimized structure and (b) band structure with majority spin and minority spin of the LiF flat film (sheet). The Fermi level is set to zero energy and indicated by the red solid line.
When a graphene-like sheet is rolled into a tubular cylinder, the corresponding bonds between the atoms are strained and distorted. The degree of the strain and distortion depends on the spatial positions of the atoms, which is clearly distinct in the zigzag and armchair NTs. All relaxed tubes reported here were initially built with all atoms equally spaced and at same radial distance from the tube axis. After relaxation, the average LiF bond lengths versus diameter for the zigzag and armchair LiF NTs are shown in Figure 2b. Obviously, the average LiF bond length decreases with increasing the tube diameter. Of course, there is also an interesting parameter describing the fact that anions and cations relax from their ideal atomic positions when a graphene-like sheet is rolled into a single cylindrical tube. The anions move slightly outward toward the tube axis, whereas 1652
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C
ARTICLE
the cations move inward with respect to their initial positions. This may lead to a decrease in the total energy of the system due to the lowered electronelectron repulsion. Therefore, the tube surface becomes buckled. After relaxation, the radial geometry of the tubular structure is characterized by two concentric cylindrical tubes, with an outer anionic and an inner cationic cylinders. The radial buckling, which indicates the strength of the buckling, is defined by β ¼ r̅ F r̅ Li
ð2Þ
where r F and r Li are the mean radii of the anion and cation cylinders, respectively. The values for the radial buckling of both zigzag (n,0) and armchair (n,n) LiF NTs as a function of the tube diameter are plotted in Figure 2c. For both zigzag (n,0) and armchair (n,n) LiF NTs, the cations (Li+) shift inward toward the NT axis, whereas the anions (F) tend to move outward with respect to their initial positions. So, both zigzag (n,0) and armchair (n,n) LiF NTs may be considered as F-coated tubes with the reconstructions of the tube surfaces being quite prominent. Obviously, the amount of radial buckling β decreases with increasing diameter for both zigzag (n,0) and armchair (n,n) LiF NTs and vanishes in the limit of very large tubes diameters. This result is similar to the calculation result for the zigzag SiC, BN, and BeO NTs.17 Surprisingly, the zigzag (n,0) and armchair (n,n) LiF NTs have almost negligible radial buckling β, which is smaller than the corresponding values obtained for SiC, BN, and BeO NTs, owing to the fact that the LiF bond length is larger than that of the corresponding NTs. Second, the radial bucklings β of armchair (n,n) LiF NTs are smaller than those of zigzag (n,0) ones for a given n, which can basically be explained by the smaller curvature surface. The band gap as a function of the tube diameter is plotted in Figure 2d. The fundamental band gaps of zigzag (n,0) and armchair (n,n) LiF NTs are denoted by the black square and red circular symbols, respectively. Several interesting features are to be noticed. First, all LiF NTs are insulating in nature, with a wide spectrum of band gap ranging from 6.648 to 6.828 eV for zigzag tubes, and 6.257 to 6.807 eV for armchair ones. Second, for small diameter NTs (4 e n e 7), the band gaps of zigzag NTs are larger than those of armchair ones, but for larger diameter NTs (8 e n e 10), the conclusion is the other way round. In ref 14, the authors conclude that the band gap is independent of the tube diameter amounting to ∼5 eV for all BeO NTs considered.14 Our results clearly reveal that this conclusion does not apply here. Instead, the band gap depends on NT diameter for d e 8.5 Å, and their values range from 6.200 to 6.800 eV. More importantly, the zigzag and armchair LiF NTs with wide band gaps can potentially be used in some nano-optical-based applications and in the low voltage-based nanoelectronics circuits as insulators where the excitation energy is not enough to overcome the gap barrier. Better insight into the distribution of the electrons with energy can be gained from the analyses of electronic band structures and density of states (DOS). Figure 4 shows band structure (insets) along the high symmetry line Γ-Z of the 1D NT Brillouin zone and DOS for the zigzag (n,0) and armchair (n,n) LiF NTs. The results of Figure 4 show that (i) both zigzag and armchair LiF NTs have large band gap, meaning both of them are insulators. (ii) A direct (indirect) band gap is exhibited for the zigzag (armchair) LiF NTs. (iii) The DOS values of both occupied states below the Fermi level and unoccupied states above
Figure 4. Total density of states (DOS) and band structure (insets) along the high symmetry line Γ-Z of the 1D nanotube Brillouin zone for (a) the zigzag (n,0) and (b) the armchair (n,n) LiF NTs. The Fermi level is set to zero energy and indicated by the vertical dashed lines. 1653
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C
ARTICLE
Table 2. Calculated Binding Energies (Eb) for Four TM Atoms Adsorbed at Five Different Sites on Outer Surface of the Zigzag (8,0) LiF NT, and the Net Magnetic Moment (μa) of the Adsorption Atoms Compared with That (μi) of the Isolated Atoms TM V
Figure 5. Left (left panel) and side views (right panel) of the band decomposed charge distribution of the CBM and VBM states of the 1D Brillouin zone at the Γ point in armchair (6,6) LiF NT. Li and F atoms are denoted by the green and gray balls, respectively.
Cr
Mn
Fe
Figure 6. Side view of five possible adsorption sites for a single TM atom adsorbed on the zigzag (8,0) LiF NT. Green and gray balls denote Li and F atoms, respectively.
the Fermi level increase with increasing n due to the atom number, and thus the electron number per unit cell increases with increasing n. It is known that its curvature generally decreases the distance between neighboring sites, especially along the circumference to a certain extent when rolling the sheet up into tubular form. Much more importantly, the respective charge densities start to overlap inside the cylinder. This is illustrated in Figure 5 by the band decomposed charge distribution of the conduction band minimum (CBM) and valence band maximum (VBM) states of the 1D Brillouin zone at the Γ point in armchair (6,6) LiF NT. We concentrate in the following on the CBM and VBM states because they are mainly responsible for the band gap, as shown in Figure 4 with red and blue lines. The Figure 5 reveals that in the armchair (6,6) LiF NT the CBM state is localized at more electronegativity F atom. In fact, the CBM state largely derives from the pz-s hybridization orbital electrons because the CBM state is of an ionic pz character, which gives rise to a fairly small dispersion of the respective uppermost valence band. In contrast, the charge distribution of the VBM state very clearly demonstrates that a major redistribution of the charge density takes place inside the tube building up a ringlike superposition. 3.2. Transition Metal (TM = V, Cr, Mn, Fe) Atom Adsorbed on the (8,0) LiF NT. Taking the zigzag (8,0) LiF NT with moderate diameter as one example, we chose a tetragonal supercell containing 32 Li and 32 F atoms as well as a single TM atom; the length of c (in the axial direction) is twice of the periodicity of the (8,0) LiF NT. The distance (10.74 Å) between two neighboring adsorbed TM atoms is large enough to neglect the interaction between two neighboring TM atoms. Figure 6 shows the side
initial sites
Eb (eV)
μa(μB)
μi(μB) 4.240
F
1.494
3.942
L A
1.500 1.497
3.943 3.941
Z
1.497
3.942
H
1.496
3.943
F
3.569
5.053
L
3.459
5.050
A
3.419
4.982
Z
3.559
5.055
H F
3.567 0.512
5.056 4.549
L
0.519
4.663
A
0.513
4.782
Z
0.508
4.744
H
0.511
4.747
F
0.935
3.003
L
0.942
3.328
A Z
0.942 0.943
3.322 3.489
H
0.944
3.478
5.185
4.949
3.547
view of different adsorption sites of a single TM atom adsorbed on the zigzag (8,0) LiF NT. The five different sites are (i) the top site of the Li atom (L), (ii) the top site of the F atom (F), (iii) the hollow site of the Li3F3 hexagon ring (H), (iv) the bridge site over an axial LiF bond (A), and (v) the bridge site over a zigzag LiF bond (Z). Four 3d TM atoms (V, Cr, Mn, and Fe) have been considered to be adsorbed on the outer surface of the pristine zigzag (8,0) LiF NT. For each species of the adsorption, five different initial adsorption sites (as shown in Figure 6) are selected to examine the interaction between the (8,0) LiF NT and single TM atom. To investigate the stability of different adsorption configurations, we have calculated the binding energy (Eb) of all the cases. Here the binding energy is given by Eb ¼ EðLiFNTÞ þ EðTMÞ EðLiFNT þ TMÞ
ð3Þ
where E(LiFNT), E(TM), and E(LiFNT+TM) are the spinpolarized total energies (per supercell) of the pristine LiF NT, a single TM atom, and TM atom adsorbed on LiF NT, respectively. All calculated values of the binding energies are listed in the third column of Table 2 together with the net magnetic moment (μa) of the binding atoms compared with that (μi) of the isolated atoms. Dramatically, the positive binding energy is obtained for each case, indicating an attractive interaction between the (8,0) LiF NT and the single TM atom, and thus the adsorption process is exothermic. In addition, for each TM atom adsorbed on (8,0) LiF NT at five different sites initially, the final binding energies are nearly identical because of the similar optimized configurations. Actually, to a certain extent, charge transfer can validate the trend in binding energy. Figure 7 displays the charge density difference on the plane containing the TM, F, and Li atoms between the more stable TM-adsorbed (8,0) LiF NTs and the sum of the 1654
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C
Figure 7. Charge density difference on the plane containing the TM (square), F (rotundity), and Li (triangle) atoms between the more stable TM-adsorbed (8,0) LiF NTs and the sum of the isolated TM atoms and the zigzag (8,0) LiF NT. The red solid line is for charge accumulation, whereas the blue dashed line is for charge depletion. Contour starts from 0.200 and ends in 0.200 (elections/Å3) and is spaced by a factor of 0.002 (elections/Å3).
Figure 8. Stable adsorption configurations for a single Cr atom adsorbed on the zigzag (8,0) LiF NT initially at F, L, A, Z, and H sites, where the green, gray, and wine balls denote the Li, F, and Cr atoms, respectively.
isolated TM atoms and the zigzag (8,0) LiF NT. We can see that there is an accumulation of electronic density in the region between the TM and F atoms in these four cases, serves as an evidence for covalent bonding between them. Because F is a highly electronegative atom, it is expected to attract electrons from the adsorbed TM atoms. A purely ionic interaction would result in an increase in the electronic density of F in a symmetric fashion around the F atom. Figure 7, in contrast, shows that electron density is removed from the region around the F atom and is added in the region between the F and TM atoms. The degree of electron density accumulation between the F and V as well as Cr atoms is more pronounced than that between the F and Mn as well as Fe atoms and hence is weaker in binding energy for the Mnand Fe-adsorbed (8,0) LiF NTs. Using Cr case as an example, the stable adsorption configurations of Cr atom on the zigzag (8,0) LiF NT at the different F, L, A, Z, and H sites initially are shown in Figure 8. It can be
ARTICLE
Figure 9. Total densities of states (DOS) with the majority (red) and minority (black) spins for the more stable adsorption configurations. The Fermi level is set to zero energy and indicated by the vertical dashed lines.
clearly seen that after full structural optimization the Cr atom almost locates on the top site of the nearest F atom, regardless of the initial location. In fact, the optimized adsorption configurations at five different initial sites are the same adsorption configuration shown in Figure 8f by circumvolving either clockwise or anticlockwise the same angle, and, the adsorption configuration at the F site is of more favorable energy (Eb = 3.569 eV). These results coincide with the calculation result of the binding energy (Table 2) and are relevant to the ion-binding character between F and TM atoms. In detail, the valence charge is strongly accumulated around the F atom and thus there is large asymmetry in charge distribution, resulting in the adsorption atom moving toward the F atom and stronger binding between them. To obtain better insight into the distribution of the electrons with energy, we have done the analyses of the density of states (DOS) for the adsorption configurations. The total DOS with the majority and minority spins (red and black lines) for the more stable adsorption configurations is presented in Figure 9. Remarkably, for V-, Cr-, and Mn-adsorbed (8,0) LiF NT, comparing the total DOS at the Fermi level of majority spin (red) with that of minority spin (black), only one type of spins (either majority spin or minority spin) passes through the Fermi level. This implies that these systems are half-metallic and could be used in the field of quasi-1D spintronics devices for producing 100% spin polarized currents, which is similar to beryllium-deficient BeO NTs.50 For Fe-adsorbed (8,0) LiF NT, at the Fermi level, there is not only the DOS of majority electrons but also the DOS of minority electrons, resulting in the decrease in the spin polarization at the Fermi level. Therefore, the magnetic moments of the adsorption Fe atom in Fe-adsorbed (8,0) LiF NT are smaller than those of the V, Cr, or Mn atoms in V-, Cr-, or Mn-adsorbed (8,0) LiF NT due to the smaller spin splitting of the Fe case.
4. CONCLUSIONS In summary, under GGA, the structural and electronic properties have been investigated for the pristine zigzag (n,0) and armchair (n,n) LiF NTs (3 e n e 10), and TM atoms 1655
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C (TM = V, Cr, Mn and Fe) adsorbed (8,0) LiF NT by using the first-principles PAW potential within DFT framework. The following conclusions are obtained: (1) For pristine single-walled armchair (n,n) and zigzag (n,0) (3 e n e 10) LiF NTs, the average value of the binding energy for the zigzag (n,0) LiF NTs is estimated to be 4.0704 eV/atom, whereas that for the armchair (n,n) LiF NTs is estimated to be 4.0961 eV/atom. Therefore, we believe that the zigzag (n,0) and the armchair (n,n) LiF NTs possess uniform stabilization for a given value of n. We point out that the calculated average atomic binding energy for LiF NTs is slightly smaller than that for bulk LiF (4.2312 eV) and thus could exist stably. Both zigzag and armchair LiF NTs have the insulating characters, and at the Γ point of the 1D Brillouin zone a direct (indirect) band gap is exhibited for the zigzag (armchair) LiF NTs. Therefore, the zigzag and armchair LiF NTs with wide band gaps can potentially be used in some nano-opticalbased applications and in the low voltage-based nanoelectronics circuits as insulators, where the excitation energy is not enough to overcome the gap barrier. (2) The positive binding energies have been obtained for the TM atoms (TM = V, Cr, Mn, and Fe) at five different sites adsorbed on (8,0) LiF NT, and thus the adsorption processes are exothermic, and after they are fully optimized, the TM atoms almost locate on the top site of the nearest F atom despite the different initial locations. Comparing the total DOS at the Fermi level of majority spin with that of minority spin, one readily identifies the adsorbed systems as having a high spin polarization and magnetic moment. In particular, for V-, Cr-, and Mn-adsorbed (8,0) LiF NT, only one type of spins (either majority spin or minority spin) passes through the Fermi level, implying that these systems are half-metallic and could be used in the field of quasi-1D spintronics devices for producing 100% spin polarized currents.
’ AUTHOR INFORMATION Corresponding Author
*Tel: +862985308456. E-mail:
[email protected].
’ ACKNOWLEDGMENT We would like to acknowledge the National Natural Science Foundation of China (grant no. 51071098) and the State Key Development for Basic Research of China (grant no. 2010CB631002) for providing financial support for this research. ’ REFERENCES (1) Iijima, S. Nature 1991, 354, 56–58. (2) Ichihashi, T. Nature 1993, 363, 603–605. (3) Robertson, J. Mater. Today 2004, 7, 46–52. (4) Tans, S. J.; Devoret, M. H.; Dai, H.; Thess, A.; Smalley, R. E.; Geerligs, L. J.; Dekker, C. Nature 1997, 386, 474–477. (5) Rueckes, T.; Kim, K.; Joselevich, E.; Tseng, G. Y.; Cheung, C. L.; Lieber, C. M. Science 2000, 289, 94–97. (6) Bachtold, A.; Hadley, P.; Nakanishi, T.; Dekker, C. Science 2001, 294, 1317–1320. (7) Wong, S. S.; Joselevich, E.; Woolley, A. T.; Cheung, C.-L.; Lieber, C. M. Nature 1998, 394, 52–55.
ARTICLE
(8) Chen, R. J.; Bangsaruntip, S.; Drouvalakisdagger, K. A.; Kam, N. W. S.; Shim, M.; Li, Y.; Kim, W.; Utzdagger, P. J.; Dai, H. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 4984–4989. (9) Zhan, G. D.; Kuntz, J. D.; Wan, J.; Mukherjee, A. K. Nat. Mater. 2003, 2, 38–42. (10) Chopra, N. G.; Luyken, R. J.; Cherry, K.; Crespi, V. H.; Cohen, M. L.; Louie, S. G.; Zettl, A. Science 1995, 269, 966–967. (11) Sun, X. H.; Li, C. P.; Wong, W. K.; Wong, N. B.; Lee, C. S.; Lee, S. T.; Teo, B. K. J. Am. Chem. Soc. 2002, 124, 14464–14471. (12) Tenne, R.; Margulis, L.; Genut, M.; Hodes, G. Nature 1992, 360, 444–446. (13) Tenne, R.; Homyonfer, M.; Feldman, Y. Chem. Mater. 1998, 10, 3225–3238. (14) Sorokin, P. B.; Fedorov, A. S.; Chernozatonskii, L. A. Phys. Solid State 2006, 48, 398–401. (15) Enyashin, A. N.; Shein, I. R.; Ivanovskii, A. L. Phys. Rev. B 2007, 75, 193408. (16) Shein, I. R.; Enyashin, A. N.; Ivanovskii, A. L. Phys. Rev. B 2007, 75, 245404. (17) Baumeier, B.; Kr€uger, P.; Pollmann, J. Phys. Rev. B 2007, 76, 085407. (18) Kresse, G.; Hafner, J. Phys. Rev. B 1993, 47, 558–561. (19) Kresse, G.; Hafner, J. Phys. Rev. B 1994, 49, 14251–14269. (20) Kresse, G.; Furthm€uller, J. Comput. Mater. Sci. 1996, 6, 15–50. (21) Kresse, G.; Furthm€uller, J. Phys. Rev. B 1996, 54, 11169–11186. (22) Sorescu, D. C.; Jordan, K. D.; Avouris, P. J. Phys. Chem. B 2001, 105, 11227–11232. (23) Yagi, Y.; Briere, T. M.; Sluiter, M. H. F.; Kumar, V.; Farajian, A. A.; Kawazoe, Y. Phys. Rev. B 2004, 69, 075414. (24) Sun, Y.; Yang, X.; Ni, J. Phys. Rev. B 2007, 76, 035407. (25) Vaccarini, L.; Goze, C.; Henrard, L.; Hernandez, E.; Bernier, P.; Rubio, A. Carbon 2000, 38, 1681–1690. (26) Blase, X.; Rubio, A.; Louie, S. G.; Cohen, M. L. Europhys. Lett. 1994, 28, 335–340. (27) Kim, Y. H.; Chang, K. J.; Louie, S. G. Phys. Rev. B 2001, 63, 205408. (28) Okada, S.; Saito, S.; Oshiyama, A. Phys. Rev. B 2002, 65, 165410. (29) Jhi, S. H.; Roundy, D. J.; Louie, S. G.; Cohen, M. L. Solid State Commun. 2005, 134, 397–402. (30) Menon, M.; Richter, E.; Mavrandonakis, A.; Froudakis, G.; Andriotis, A. N. Phys. Rev. B 2004, 69, 115322. (31) Xiang, H. J.; Yang, J.; Hou, J. G.; Zhu, Q. Phys. Rev. B 2003, 68, 035427. (32) Guo, G. Y.; Lin, J. C. Phys. Rev. B 2005, 71, 165402. (33) Zhao, M.; Xia, Y.; Li, F.; Zhang, R. Q.; Lee, S. T. Phys. Rev. B 2005, 71, 085312. (34) Gali, A. Phys. Rev. B 2006, 73, 245415. (35) Baierle, R. J.; Piquini, P.; Neves, L. P.; Miwa, R. H. Phys. Rev. B 2006, 74, 155425. (36) Saito, R.; Fujita, M.; Dresselhaus, G.; Dresselhaus, M. S. Appl. Phys. Lett. 1992, 60, 2204–2206. (37) Hamada, N.; Sawada, S. I.; Oshiyama, A. Phys. Rev. Lett. 1992, 68, 1579–1581. (38) Blase, X.; Benedict, L. X.; Shirley, E. L.; Louie, S. G. Phys. Rev. Lett. 1994, 72, 1878–1881. (39) Rubio, A.; Corkill, J. L.; Cohen, M. L. Phys. Rev. B 1994, 49, 5081–5084. (40) Kresse, G.; Joubert, D. Phys. Rev. B 1999, 59, 1758–1775. (41) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865–3868. (42) Monkhorst, H. J.; Pack, J. D. Phys. Rev. B 1976, 13, 5188–5192. (43) Wyckhoff, R. W. G. Crystal Structures, 2nd ed.; Interscience Publishers: New York, 1963; Vol. 1. (44) Vogt, J.; Weiss, H. Surf. Sci. 2002, 501, 203–213. (45) Brightcrystals Technology. http://www.brightcrystals.com/readnew.asp?NewsID=4945, 10, 8, 2011. (46) Barnard, A. S.; Russo, S. P. J. Phys. Chem. B 2003, 107, 7577– 7581. 1656
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
The Journal of Physical Chemistry C
ARTICLE
(47) Dresselhaus, M. S.; Dresselhaus, G.; Saito, R. Phys. Rev. B 1992, 45, 6234–6242. (48) Rathi, S. J.; Ray, A. K. Chem. Phys. Lett. 2008, 466, 79–83. (49) Rathi, S. J.; Ray, A. K. Nanotechnology 2008, 19, 335706. (50) Gorbunova, M. A.; Shein, I. R.; Makurin, Yu. N.; Ivanovskaya, V. V.; Kijko, V. S.; Ivanovskii, A. L. Physica E 2008, 41, 164–168.
1657
dx.doi.org/10.1021/jp2045146 |J. Phys. Chem. C 2012, 116, 1650–1657
