Journal of Nanoparticle Research

, Volume 11, Issue 6, pp 1405–1420

Interactions of Fe atom with single wall armchair SiC nanotubes: an ab initio study

Authors

  • Kazi M. Alam
    • Department of PhysicsUniversity of Texas at Arlington
    • Department of PhysicsUniversity of Texas at Arlington
Research Paper

DOI: 10.1007/s11051-008-9529-2

Cite this article as:
Alam, K.M. & Ray, A.K. J Nanopart Res (2009) 11: 1405. doi:10.1007/s11051-008-9529-2

Abstract

A systematic study of Fe atom encapsulation and adsorption in armchair SiC nanotubes (SiCNT) with diameters in the range of 5.313 to 10.582 Å has been performed using hybrid density functional theory and a finite cluster approximation. A detailed comparison of the binding energies, equilibrium positions, Mulliken charges, and spin magnetic moments of Fe atoms has been performed for three types of nanotubes. The electronic states, HOMO–LUMO gaps, and changes in gaps with respect to the bare nanotube gaps have been investigated as well. Our results show that the properties of SiCNT can be modified by Fe atom encapsulation and adsorption. Binding energies of the encapsulated and adsorbed systems indicate that these structures are stable and show site dependence. For both cases a significant band gap decrease is observed for type 1 nanotubes enabling band gap tailoring. This decrease is not observed for the other two types with a larger diameter. All structures are found to have magnetic ground states with high magnetic moments indicating the possibility of them being used in spintronics applications.

Keywords

Fe atomSiC nanotubeInteractionsHybrid density functional theoryElectronic and geometric structuresNanoscale modeling and simulation

Introduction

Nanoscience and nanotechnology continue to be very exciting fields of research due to their far-reaching scientific, technological, and societal applications and implications (Roco et al. 2000). Among the various nanostructures, carbon nanotubes have been investigated most extensively due to their many fascinating electronic, mechanical, and optical properties (Dresselhaus et al. 2001; Iijima 1991; Iijima and Ichihashi 1993). They provide ideal model systems to study electrical transport properties of one-dimensional nanostructures and molecules (Wind et al. 2003; Javey et al. 2004). This one-dimensional behavior arises from their extremely small diameters comparable to the electron de Broglie wave length. Theoretical (Andriotis et al. 2000; Lee et al. 1997; Kong et al. 1999; Fagan et al. 2003; Durgun et al. 2003; Durgun et al. 2004; Dag et al. 2004) and experimental (Zhang et al. 2000; Zhang and Dai 2000) research on interactions of transition metal atoms with single wall carbon nanotubes (SWCNTs) have reported dramatic changes in the structural, electronic, and magnetic properties of the nanotubes and provide valuable information about possible uses of them as nanowires, nanomagnets, interconnects in molecular electronic network, spintronics, recording media, and magnetic links, among others. Doping metal atoms inside silicon clusters and nanotubes has also acquired significant attention since this stabilizes the structures (Singh et al. 2003; Hiura et al. 2001). In addition, it was found that electrical, magnetic, and optical properties of the silicon nanostructures can be tailored by changing the metal atoms.

Silicon carbide (SiC) in bulk form is one of the hardest materials and is very suitable for electronic devices designed for operations in extreme environments. SiC, with its wide band gap, high thermal conductivity, and radiation resistance, is particularly important for use in high temperature and radiation environments. SiC nanotubes (SiCNT) have been successfully synthesized by different groups (Sun et al. 2002; Keller et al. 2003; Borowiak-Palen et al. 2005; Taguchi et al. 2005a, b; Hu et al. 2004; Nhut et al. 2002; Pham-Huu et al. 2001; Huczko et al. 2005), and we comment only on a few of them for the sake of brevity. Sun et al. have reported the synthesis of SiCNT trough a substitutional reaction with Si atoms replacing half of the C atoms from a multi-walled carbon nanotube. The observed SiCNT were also multi-walled but with larger inter-planar spacing than those of multi-walled carbon nanotubes. This indicated weak coupling between inner and outer tubes and the possibility of separating them with ease. Borowiak-Palen et al. produced SiCNT based on high-temperature reactions between silicon powders and multi-walled carbon nanotubes. Hu et al. formed SiCNT by reacting CH4 with SiO. SiC nanotubes are expected to have some advantages over carbon nanotubes. Theoretical studies (Zhao et al. 2005, Mavrandonakis et al. 2003; Miyamoto and Yu 2002; Menon et al. 2004) performed on the structure and stability of SiCNTs have shown that the most stable structures contain Si and C atoms with the ratio of 1:1, and nanotubes with only Si–C bonds have higher binding energies/atom than those containing Si–Si and C–C bonds in addition to Si–C bonds. We have carried out detailed studies on the stability of Si60 fullerene-like cages by substitutional and endohedral placements of carbon atoms. Stability is higher when the Si and the C atoms are in separate sub-units on the cage (Huda and Ray 2004, 2008; Srinivasan et al. 2005, Ray and Huda 2006; Huda et al. 2007). We have also studied the evolution of geometric and electronic properties with tube diameters of three different types of single wall armchair and zigzag SiCNT (Alam and Ray 2007, 2008). Though type 1 nanotubes, containing only Si–C bonds, were found to be more stable, other two types, containing Si–C, Si–Si, and C–C bonds, were predicted to play possible important roles in molecular electronic networks for having lower band gaps with higher diameters. Among all three types of SiCNT we studied, type 1 nanotubes have valence charges strongly accumulated around C atoms, making the nanotube walls highly reactive to the external atom or group of atoms. This asymmetry has been exploited to band structure modification by side wall decoration with H, CH3, SiH3, N, NH, NH2 and Si or C substitution by N atom (Zhao et al. 2005a, b, Li et al. 2005, He et al. 2006). In our previous study of armchair nanotubes, we demonstrated that within the same helicity and under the constraint of Si to C ratio as 1:1, the dominant factor in deciding the stability, tube morphology, and electronic behavior is the relative position of Si and C atoms, and consequently the nature of chemical bonds. The cohesive energy of the nanotubes increased and approached saturation as the number of atoms increased, with hybridization of Si and C atoms on the tube surface causing radial buckling and all three types of bare armchair SiCNT were semiconductors. Type 1 tubes have the largest band gaps. Strong ionic type bonding localized the electronic states which results in wide band gap for the type 1 nanotubes. Types 2 and 3 nanotubes were found to have significantly lower gaps. Type 2 nanotubes exhibit a zigzag type trend in gap and diameter relationship, while band gaps for type 3 nanotubes decrease monotonically with the increasing diameters. It was predicted that type 2 and 3 tubes might exhibit metallic behavior at higher diameters (Alam and Ray 2007, 2008).

Unlike carbon and silicon nanotubes, interaction of transition metal atoms with SiCNT has not been explored extensively. Recently, first principles method has been used to investigate the adsorption of Ti atom on type 1 single wall SiCNT with the corresponding adsorption of hydrogen molecules (Meng et al. 2007). Using periodic boundary conditions and the Dmol3 suite of software, Zhao and Ding have very recently studied the silicon carbide nanotubes functionalized by transition metal atoms. They primarily studied interactions with a (8, 0) zigzag SiC nanotube and four of them, not including Fe, with a (6, 6) armchair type 1 nanotube (Zhao and Ding 2008). They concluded that transition metal-SiC nanotube materials could be used in various interesting applications, such as nanomagnets and hydrogen storage (Mukherjee and Ray 2008). To the best of our knowledge, no other study has been reported regarding the interaction of transition metal atoms with SiCNT. As is known, the most common experimental methods for fabricating carbon nanotubes are arc discharge, laser ablation, and chemical vapor deposition. Transition metals such as Mn, Fe, Co, and Ni or metal alloys are required as catalyst in those methods. As mentioned before SiCNTs were synthesized using carbon nanotubes as template, so it is particularly important to analyze the interaction of these metal atoms with both CNT and SiCNT. In this present work, we have systematically studied the encapsulation and adsorption of an Fe atom in all three types of single wall armchair SiCNT.

Computational method

As in our previous works (Alam and Ray 2007, 2008), we have used hybrid density functional theory to investigate the interaction of SiCNT with Fe atoms. In particular, we have used the B3LYP (Becke 1993, 1998; Lee et al. 1988) hybrid functional and the Los Alamos National Laboratory double-ζ basis set (Dunning and Hay 1976) (Lanl2DZ) as implemented in the Gaussian 03 suite of programs (Frisch 2003). For iron and silicon atoms, the Hay–Wadt pseudo potential (Hay and Wadt 1985) and the associated basis set are used for the 10 core electrons and the valence electrons, respectively. For carbon and hydrogen atoms, an all electron basis set, specifically the Dunning–Huzinaga double-zeta basis set has been employed. Hybrid density functional theory incorporating Hartree (HF) exchange (Hehre et al. 1986; Young 2001) with density functional theory (DFT) exchange-correlation (Perdew et al. 1982) has proved to be an efficient method for many systems. It has been recently verified that hybrid functionals can reproduce the band gaps of semiconductors and insulators quite well (Muscat et al. 2001; Heyd and Scuseria 2004). The SiCNTs have been represented by finite clusters with dangling bonds terminated by hydrogen atoms to simulate the effect of infinite nanotubes. All the computations reported here have been performed at the supercomputing center of the University of Texas at Arlington.

Results and discussions

In this study, an Fe atom is placed at the center of each type of nanotubes from (3, 3) to (6, 6) with the aim of exploring comparative evolution and modification of geometric and electronic properties from bare nanotubes along with the relative binding energies of Fe atoms inside those nanotubes. Placing an Fe atom inside nanotubes with higher diameters beyond (6, 6) would not provide significant information due to negligible interaction with the tube wall. In our previous work (Alam and Ray 2007, 2008), we have noted that as the size of the nanotubes increased from (3, 3) to (11, 11), binding energies tended to saturate for all three types of nanotubes. From (3, 3) to (6, 6), the increase in binding energies was more prominent and then a slow variation continues up to (11, 11). For this reason and for obvious severe demand on computational resources, we opted to concentrate on Fe adsorption at different sites of (3, 3) and (6, 6) for all three types of nanotubes. Figure 1 shows the relative positions of Si and C atoms in different types of nanotubes. In the type 1 arrangement, with only Si–C bonds, silicon and carbon atoms are placed alternatively without any adjacent Si or C atoms. In types 2 and 3 arrangements with Si–C, Si–Si, and C–C bonds, the nearest neighbors of each Si atom consist of two C atoms and another Si atom and vice versa. The difference between type 2 and type 3 lies in the relative spatial positions of the Si and C atoms. If we consider one layer perpendicular to the tube axis, in type 3, Si and C atoms are alternating, while in type 2 each layer contains either Si or C atoms. Figure 1 also shows the different adsorption sites, five for type 1 and seven for types 2 and 3 nanotubes.
https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig1_HTML.gif
Fig. 1

Atomic arrangements and different adsorption sites for (a) type 1, (b) type 2, and (c) type 3 nanotubes. The carbon atoms are yellow and silicon atoms are green. The dashed lines represent the orientation of tube axis

There are three major adsorption sites: top, bridge, and hollow. C top and Si top refer to the positions directly perpendicular to the tube wall along the C and Si atoms, respectively. There are two major bridge sites named as normal and zigzag bridge. The relative orientation of those two sites with respect to the tube axis is clearly visible in the figures. Those two bridge sites are named according to their terminal atoms. Hollow sites are at the middle of the hexagons. In types 1 and 3, only one type of hexagon is present and there is only one hollow site. In type 2, two different hexagons are present and this gives rise to two different hollow sites. The first hollow site is associated with the hexagon containing four C and two Si atoms, whereas the second hollow site is positioned along the center perpendicular to hexagon containing two C and four Si atoms.

The binding energy for the encapsulated or adsorbed Fe atom is obtained from the expression below.
$$ E_{\text{b}} = E\left( {\text{SiCNT}} \right) + E\left( {\text{Fe}} \right) - E\left( {{\text{Fe}} + {\text{SiCNT}}} \right) $$
(1)
where E (SiCNT) and E (Fe) are the ground state energies of the bare nanotube and Fe atom, respectively. E (Fe + SiCNT) represent the ground state energy of the Fe encapsulated or adsorbed SiCNT. Table 1 shows the stoichiometry and the binding energies of the Fe atom inside the nanotube from (3, 3) to (6, 6) for all types. Though type 1 bare nanotubes were found to be most stable in our previous work (Alam and Ray 2007, 2008), we note from Table 1 that type 2 tubes have the strongest interactions with Fe atoms followed by type 3 and then type 1. As the diameter or the tube size increases, interactions between Fe atom and tube wall decrease for all types. Figure 2 shows this variation of binding energy/atom with respect to tube diameter. After optimization Fe atoms were found to retain their initial positions (center of the tube) with negligible deviation along the tube axis. The diameters did not change upon Fe atom encapsulation, from that of bare nanotubes. The Mulliken charges on the Fe atom are shown in Table 2. For the smallest nanotubes, charge transfers occur from the tube to the Fe atom. This direction of charge transfer continues for all type 1 larger nanotubes but for the other two types the direction is reversed as we increase the tube diameter. From (4, 4) to (6, 6) tubes, Fe atom transfers charge to the tube in case of type 2 and type 3 tubes. We believe that this is due to the presence of two other types of bonds, namely, Si–C and C–C in types 2 and 3 nanotubes and since both Si and C atoms are more electronegative compared to Fe. Figure 3 demonstrates how the highest occupied molecular orbital (HOMO) is localized to the Fe atom in type 1 (3, 3). The undoped tube HOMO was uniformly distributed. Similar behavior was observed for the other two types. The total charge density for the smallest tube (3, 3) is shown in Fig. 4. We note that for type 2, there is significant overlap between the charge of the Fe atom with the SiCNT, followed by that for type 3 and type 1. The more lateral expansion of the blue region for types 2 and 3 relative to type 1 is reminiscent of our previous observation of buckling effect, where buckling was almost negligible in type 1 nanotubes (Alam and Ray 2007, 2008).
Table 1

Cohesive energies/atom (in eV) for Fe atom encapsulated in armchair SiCNT

Nanotube

Stoichiometry

Total number of atoms

Binding energy of Fe atom (eV) inside the nanotube

Type 1

Type 2

Type 3

SiC (3, 3)

Si30C30H12Fe1

73

1.712

3.149

2.634

SiC (4, 4)

Si40C40H16Fe1

97

0.557

1.104

0.833

SiC (5, 5)

Si50C50H20Fe1

121

0.535

0.632

0.634

SiC (6, 6)

Si60C60H24Fe1

145

0.491

0.561

0.553

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig2_HTML.gif
Fig. 2

Cohesive energy/atom (eV) versus tube diameter (Å) for all three types of nanotubes

Table 2

Tube diameters (in Å) and Mulliken charges of encapsulated Fe atoms

Nanotube

Tube diameter (Å)

Mulliken charge of Fe atom

Type 1

Type 2

Type 3

Type 1

Type 2

Type 3

SiC (3, 3)

5.313

5.412

5.375

−0.296

−0.294

−0.172

SiC (4, 4)

7.022

7.162

7.116

−0.140

0.529

0.411

SiC (5, 5)

8.760

8.936

8.841

−0.071

0.217

0.152

SiC (6, 6)

10.451

10.694

10.582

−0.049

0.081

0.058

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig3_HTML.gif
Fig. 3

The localization of highest occupied level for type 1 (3, 3) bare (left) and Fe atom doped (right) nanotubes. The green, yellow, and black atoms are Si, C, and Fe, respectively. Red transparent lobes show the positive and blue transparent lobes show the negative values of the wave function

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig4_HTML.gif
Fig. 4

Top and side views of the calculated total charge density (blue color) of Fe encapsulated (3, 3) for type 1 (left), type 2 (middle), and type 3 (right) nanotubes. The green, yellow, and black atoms are C, Si, and Fe, respectively

Tables 3, 4, 5, 6, 7, and 8 show Fe atom adsorption energies, distances of Fe atom from the tube wall, nearest C–Fe and Si–Fe distances, and the Mulliken charges of Fe atom for (3, 3) and (6, 6) nanotubes of three types. Only (3, 3) and (6, 6) tubes were considered here since no significantly new results were expected by studying the (4, 4) and (5, 5) tubes and also because we are interested in the curvature effects of the tubes. There are five possible sites for type 1 and seven sites for other two types described earlier. For type 1 (3, 3), Si–C normal bridge site is the most stable site and Si top is the least stable site, with a difference in adsorption energy of 0.67 eV. The corresponding sites for type 2 (3, 3) are second hollow and C–C normal bridge, with a difference in adsorption energy of 1.09 eV. In case of type 3 (3, 3), they are hollow and C–C zigzag bridge, with the corresponding difference being 1.42 eV. The Si–C zigzag bridge for type 1 (6, 6) and C top for types 2 and 3 (6, 6) are the most stable Fe adsorption sites. The difference in adsorption energies between the most and least favored sites for type 1 (6, 6) is about 0.62 eV. These values are 0.90 and 0.41 eV for type 2 (6, 6) and type 3 (6, 6), respectively. Another noticeable observation is that Fe prefers to bond with C than with Si, a common feature for all types for both (3, 3) and (6, 6) consistent with our Fe–C and Fe–Si dimer calculations, where the binding energies/atom are 1.81 and 1.26 eV, respectively. Using the DMol3 suite of software, for Ti adsorption on type 1 (5, 5) SiCNT, Meng et al. (2007) found the most favorable site to be the C top site. We also note that Fe adsorption energy decreases with increasing radius (or decreasing curvature) of the tube. Therefore by creating regions of different curvature on a single SiCNT by radial deformation or be designing different types of tubes, it is possible to get different values of adsorption energies. The average adsorption energy on a type 1 (3, 3) is 1.515 eV and that on a (6, 6) is 1.100 eV. For type 2, the corresponding numbers are 2.658 and 1.350 eV and for type 3, 2.564 and 1.581 eV. So type 2 and type 3 nanotubes have better Fe adsorption capacity as a whole; in other words, complete Fe coating is predicted to be more facile for these two types of SiCNT. Despite possessing highest binding energy/atom for the bare tube, type 1 tubes have the least possibility of being coated with Fe atoms. This implies Fe–Fe interaction can dominate over Fe–nanotube interaction which might result in forming isolated Fe clusters than complete Fe coating.
Table 3

Adsorption energies of Fe atoms (eV), distances of Fe atoms from tube walls (Å), nearest C–Fe distances (Å), nearest Si–Fe distances (Å), and Mulliken charges of Fe atoms adsorbed on type 1 (3, 3) armchair SiC nanotube

Site

Adsorption energy of Fe atom (eV)

Distance of Fe atom from tube wall (Å)

Nearest C–Fe distance (Å)

Nearest Si–Fe distance (Å)

Mulliken charge of Fe atom

C top

1.527

1.738

1.930

2.492

0.195

Si top

1.201

2.134

2.826

2.415

0.236

Hollow

1.567

1.350

2.219

2.552

0.281

Si–C normal Bridge

1.871

1.763

2.068

2.380

0.248

Si–C zigzag bridge

1.410

1.997

2.101

2.425

0.252

Table 4

Adsorption energies of Fe atoms (eV), distances of Fe atoms from tube walls (Å), nearest C–Fe distances (Å), nearest Si–Fe distances(Å), and Mulliken charges of Fe atoms adsorbed on type 1 (6, 6) armchair SiC nanotube

Site

Adsorption energy of Fe atom (eV)

Distance of Fe atom from tube wall (Å)

Nearest C–Fe distance (Å)

Nearest Si–Fe distance (Å)

Mulliken charge of Fe atom

C top

1.237

2.234

2.330

2.675

0.022

Si top

0.928

2.476

3.189

2.518

0.139

Hollow

0.714

1.544

2.505

2.420

0.126

Si–C normal Bridge

1.289

2.117

2.168

2.441

0.113

Si–C zigzag bridge

1.333

2.167

2.191

2.493

0.094

Table 5

Adsorption energies of Fe atoms (eV), distances of Fe atom from tube walls (Å), nearest C–Fe distances (Å), nearest Si–Fe distances (Å), and Mulliken charges of Fe atoms adsorbed on type 2 (3, 3) armchair SiC nanotube

Site

Adsorption energy of Fe atom (eV)

Distance of Fe atom from tube wall (Å)

Nearest C–Fe distance (Å)

Nearest Si–Fe distance (Å)

Mulliken charge of Fe atom

C top

2.983

2.275

2.963

2.456

0.134

Si top

2.836

2.244

2.914

2.451

0.160

First hollow

2.028

1.005

2.155

2.638

0.465

Second hollow

3.116

1.153

2.020

2.415

−0.652

C–C normal bridge

2.022

1.933

2.250

2.332

0.151

Si–Si normal bridge

2.745

2.124

3.261

2.349

−0.135

Si–C zigzag bridge

2.875

2.265

2.874

2.461

0.155

Table 6

Adsorption energies of Fe atoms (eV), distances of Fe atom from tube walls (Å), nearest C–Fe distances (Å), nearest Si–Fe distances (Å), and Mulliken charges of Fe atoms adsorbed on type 2 (6, 6) armchair SiC nanotube

Site

Adsorption energy of Fe atom (eV)

Distance of Fe atom from tube wall (Å)

Nearest C–Fe distance (Å)

Nearest Si–Fe distance (Å)

Mulliken charge of Fe atom

C top

1.765

2.137

2.255

2.376

0.182

Si top

1.489

2.447

3.517

2.618

0.159

First hollow

0.866

1.641

2.502

2.600

0.254

Second hollow

0.932

1.749

2.960

2.391

−0.275

C–C normal bridge

1.237

1.618

2.145

2.582

0.200

Si–Si normal bridge

1.446

2.021

3.514

2.428

0.047

Si–C zigzag bridge

1.717

2.337

2.722

2.469

0.133

Table 7

Adsorption energies of Fe atoms (eV), distances of Fe atom from tube walls (Å), nearest C–Fe distances (Å), nearest Si–Fe distances (Å), and Mulliken charges of Fe atoms adsorbed on type 3 (3, 3) armchair SiC nanotube

Site

Adsorption energy of Fe atom (eV)

Distance of Fe atom from tube wall (Å)

Average C–Fe distance (Å)

Average Si–Fe distance (Å)

Mulliken charge of Fe atom

C top

2.862

1.542

2.071

2.413

0.182

Si top

2.157

2.618

3.604

2.372

0.202

Hollow

3.241

1.638

2.158

2.448

0.232

Si–C normal bridge

2.867

1.625

2.167

2.419

0.212

Si–C zigzag bridge

2.860

1.604

2.132

2.434

0.193

C–C zigzag bridge

1.823

1.602

2.078

2.425

0.187

Si–Si zigzag bridge

2.139

2.098

3.496

2.412

0.082

Table 8

Adsorption energies of Fe atoms (eV), distances of Fe atom from tube walls (Å), nearest C–Fe distances (Å), nearest Si–Fe distances (Å), and Mulliken charges of Fe atoms adsorbed on type 3 (6, 6) armchair SiC nanotube

Site

Adsorption energy of Fe atom (eV)

Distance of Fe atom from tube wall (Å)

Average C–Fe distance (Å)

Average Si–Fe distance (Å)

Mulliken charge of Fe atom

C top

1.745

1.969

2.392

2.551

0.233

Si top

1.434

2.430

3.577

2.583

0.132

Hollow

1.664

2.435

3.227

2.679

0.171

Si–C normal bridge

1.332

2.032

2.139

2.443

0.219

Si–C zigzag bridge

1.684

1.954

2.552

2.618

0.226

C–C zigzag bridge

1.647

1.785

2.101

2.434

0.225

Si–Si zigzag bridge

1.562

2.317

3.492

2.417

0.066

Mulliken charge analysis shows electron transfer occurs from Fe atom to the nanotube for all sites of all types except for second hollow site of type 2 (6, 6) and (3, 3) and Si–Si normal bridge of type 2 (3, 3) where Fe atom gains 0.275e, 0.652e, and 0.135e from the tube, respectively. No trend between amount of charge transfer and adsorption energy of the site was found here, suggesting the local structure as a whole determines the amount and direction of charge transfer as well as the binding strength of the adsorbent. From adsorption energies, it can be inferred that the most favorable sites for the three types of (6, 6) nanotubes C top site of type 2 has the maximum adsorption energy, followed by C top of type 3 and Si–C zigzag bridge site of type 1. The corresponding order of the sites for (3, 3) is hollow of type 3, second hollow of type 2, and Si–C normal bridge site of type 1. When Fe atom was placed on C top and Si top sites of (6, 6) SiCNT, the atom slightly migrated toward the neighboring normal or zigzag bridge sites but did not deviate completely from the initial positions. This slight migration was observed towards the hollow sites in case of (3, 3) nanotubes. This is the reason we see slight differences between the distances of Fe atom from the tube wall and the Fe–C and Fe–Si distances for the C top and Si top sites. Figures 5, 6, and 7 show the HOMO localizations for the most and the least preferred sites for all three types (6, 6) of nanotubes with Fe atom adsorbed on them. The more localization of the “negative” and “positive” values of the wave function is found for the most preferred sites. The difference in localization is not much perspicuous for the type 3 sites. The reason is attributed to the fact that the difference in adsorption energies between the most and least preferred sites for type 3 configuration is small compared to those for the other two types.
https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig5_HTML.gif
Fig. 5

The localization of highest occupied level for Fe atom adsorbed on type 1 (6, 6) nanotube. The most preferred Si–C zigzag bridge site (top) and least preferred hollow site (bottom). The green, yellow, and black atoms are Si, C, and Fe, respectively. Red transparent lobes show the positive and blue transparent lobes show the negative values of the wave function

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig6_HTML.gif
Fig. 6

The localization of highest occupied level for Fe atom adsorbed on type 2 (6, 6) nanotube. The most preferred C top site (top) and least preferred first hollow site (bottom). The green, yellow, and black atoms are Si, C, and Fe, respectively. Red transparent lobes show the positive and blue transparent lobes show the negative values of the wave function

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig7_HTML.gif
Fig. 7

The localization of highest occupied level for Fe atom adsorbed on type 3 (6, 6) nanotube. The most preferred C top site (top) and least preferred Si–C normal bridge site (bottom). The green, yellow, and black atoms are Si, C, and Fe, respectively. Red transparent lobes show the positive and blue transparent lobes show the negative values of the wave function

In nanotube and fullerene-based technology, molecular electronic structures are required with a wide variety of mechanical, electronic, optical and magnetic properties. Band gap tailoring is an important issue for different purposes especially when coupling of nanostructures is concerned in a complex network. Tables 9, 10, and 11 show the HOMO–LUMO gaps of the (Fe + SiCNT) system and the change of the gaps with respect to the bare nanotube gaps (Alam and Ray 2007, 2008). In our previous calculation for bare nanotubes, band gaps for type 1 nanotubes were found to be larger than bulk 3C–SiC gap, while type 2 and type 3 nanotubes had significantly lower band gaps. Unlike the other two types, band gap for type 3 nanotubes showed a monotonous decreasing trend with increasing tube diameter. Table 9 and Fig. 8 show the variation of these gaps with tube diameter for the Fe atom encapsulated in type 1 nanotube. There is a continuous decrease in gap with increasing tube diameter for the Fe encapsulated tube. This trend is reversed from the bare tube gap variation. There is a significant decrease in gap upon Fe atom encapsulation for type 1 nanotubes. In case of types 2 and 3 only small tubes show this reduction, while it remains almost unchanged for the tubes with smaller curvature or higher diameter. These results indicate that band gaps can be tuned by Fe doping inside the type 1 tubes only. Table 10 and Fig. 9 show the variation of gaps with tube diameter for the Fe atom encapsulation in type 2 nanotubes. This variation for type 3 is shown in Table 11 and Fig. 10. The HOMO–LUMO gaps and the corresponding changes in gaps from the bare nanotubes for different adsorption sites of three types of SiCNTs have been presented in Tables 12, 13, 14, 15, 16, and 17. Our previous results showed the bare nanotube (6, 6) gaps to be 2.932, 0.876, and 0.835 eV for types 1, 2, and 3, respectively. These gaps were 2.776, 1.487, and 1.216 eV for the (3, 3) tubes in three configurations. For (6, 6) we observe the same trends as in the encapsulation case, as gap change is significant for type 1 only. For all the five sites the reduction in band gap is close to 1 eV with the maximum decrease occurred for C top site. Adsorbing Fe atom can slightly decrease the gap when it is at either the C top or the first hollow site of type 2 (6, 6) nanotube. All other sites do not show any significant change. It is interesting to notice that type 3 sites do not show any remarkable gap modification upon Fe adsorption on any of the seven sites. Unlike all other sites Si top and Si–Si zigzag bridge sites show a very small gap increase upon Fe atom adsorption. In case of (3, 3) type 1 sites also showed the maximum reduction of gaps upon Fe adsorption. However, unlike (6, 6), type 2 and type 3 (3, 3) tubes show noticeable band gap modification. All these results suggest that band gap tuning upon single Fe atom interaction with armchair SiCNT is mostly feasible for type 1 structures. Figure 11 demonstrates energy density of states (DOS) for pristine type 1 (6, 6) nanotube and the corresponding Fe atom adsorbed nanotube that has the maximum adsorption energy (Si–C zigzag bridge site). The DOS is built by fitting a Gaussian function in each eigenvalue and then summing them up. The Gaussian width used to broaden the eigenvalues is 0.05 eV and E = 0 refers to the HOMO. It is worth noting that there are very few isolated extra created states near the HOMO within the pseudogap in case of the adsorbed nanotube. Similar observation was found for the other sites of type 1 nanotubes. This significant decrease in band gap for type 1 nanotubes upon Fe atom interaction might play an important role at the metal nanotube junction in molecular electronic networks or other potential applied areas of band gap engineering.
Table 9

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms encapsulated in type 1 armchair SiCNT

Nanotube

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

(3, 3)

1.764

1.012

5A

3.054

(4, 4)

1.728

1.095

5A

3.989

(5, 5)

1.521

1.368

5A

3.995

(6, 6)

1.409

1.523

5A

4.010

Table 10

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms encapsulated in type 2 armchair SiCNT

Nanotube

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

(3, 3)

1.012

0.475

5A

2.917

(4, 4)

0.773

0.033

5A

3.405

(5, 5)

0.821

0.018

5A

4.023

(6, 6)

0.813

0.063

5B

4.047

Table 11

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms encapsulated in type 3 armchair SiCNT

Nanotube

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

(3, 3)

0.573

0.643

5A

3.040

(4, 4)

0.473

0.611

5A

3.654

(5, 5)

0.865

0.040

5A

3.995

(6, 6)

0.825

0.010

5B

4.003

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig8_HTML.gif
Fig. 8

Bare nanotube HOMO–LUMO gap (eV), gap for the Fe-doped tube and change (decrease) in gap of the doped tube with respect to the bare tube gap versus tube diameter (Å) for type 1 nanotubes

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig9_HTML.gif
Fig. 9

Bare nanotube HOMO–LUMO gap (eV), gap for the Fe doped tube and change (decrease) in gap of the doped tube with respect to the bare tube gap versus tube diameter (Å) for type 2 nanotubes

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig10_HTML.gif
Fig. 10

Bare nanotube HOMO–LUMO gap (eV), gap for the Fe-doped tube and change (decrease) in gap of the doped tube with respect to the bare tube gap versus tube diameter (Å) for type 3 nanotubes

Table 12

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms adsorbed on type 1 (3, 3) armchair SiC nanotube

Site

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic State

Spin magnetic moment (μB) of Fe atom

C top

2.129

0.647

3A

2.602

Si top

2.212

0.564

5A

3.851

Hollow

1.972

0.804

3A

2.894

Si–C Normal bridge

1.942

0.834

5A

3.731

Si–C Zigzag bridge

1.754

1.022

5A

3.737

Table 13

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms adsorbed on type 1 (6, 6) armchair SiC nanotube

Site

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

C top

1.807

1.125

5A

3.968

Si top

2.158

0.774

5A

3.926

Hollow

1.972

0.960

3A

2.659

Si–C Normal bridge

1.998

0.934

5A

3.919

Si–C Zigzag bridge

1.853

1.079

5A

3.865

Table 14

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms adsorbed on type 2 (3, 3) armchair SiC nanotube

Site

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

C top

1.110

0.377

5A

3.071

Si top

1.092

0.395

5A

3.074

First hollow

1.055

0.432

5A

3.046

Second hollow

1.070

0.417

5A

2.639

C–C normal bridge

0.762

0.725

5A

2.918

Si–Si normal bridge

1.061

0.426

5A

3.209

Si–C zigzag bridge

1.066

0.421

3A

2.986

Table 15

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms adsorbed on type 2 (6, 6) armchair SiC nanotube

Site

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

C top

0.752

0.124

5A

3.337

Si top

0.860

0.016

5A

3.266

First hollow

0.549

0.327

5A

3.256

Second hollow

0.838

0.038

3A

2.554

C–C normal bridge

0.824

0.052

3A

2.830

Si–Si normal bridge

0.841

0.035

5A

3.480

Si–C zigzag bridge

0.852

0.024

3A

2.905

Table 16

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms adsorbed on type 3 (3, 3) armchair SiC nanotube

Site

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

C top

0.858

0.358

5A

3.148

Si top

0.843

0.373

5A

3.155

Hollow

1.450

−0.234

5A

3.373

Si–C Normal bridge

1.005

0.211

5A

3.468

Si–C zigzag bridge

0.870

0.346

5A

3.211

C–C zigzag bridge

0.614

0.602

5A

2.868

Si–Si zigzag Bridge

0.987

0.229

3A

2.853

Table 17

HOMO–LUMO gaps (in eV), change in HOMO–LUMO gaps (in eV) from bare nanotubes, electronic states, and spin magnetic moments (μB) of Fe atoms adsorbed on type 3 (6, 6) armchair SiC nanotube

Site

HOMO–LUMO gap (eV)

Change in HOMO–LUMO gap (eV)

Electronic state

Spin magnetic moment (μB) of Fe atom

C top

0.786

0.049

5A

3.680

Si top

0.885

−0.050

5A

3.217

Hollow

0.773

0.062

5A

3.203

Si–C Normal bridge

0.779

0.056

3A

2.659

Si–C zigzag bridge

0.780

0.055

5A

3.744

C–C zigzag bridge

0.774

0.061

3A

2.762

Si–Si zigzag Bridge

0.868

−0.033

3A

2.783

https://static-content.springer.com/image/art%3A10.1007%2Fs11051-008-9529-2/MediaObjects/11051_2008_9529_Fig11_HTML.gif
Fig. 11

Density of states for type 1 (6, 6) bare (top) and the corresponding Fe adsorbed (for the most stable Si–C zigzag bridge site) (bottom) nanotubes

All the (Fe + SiCNT) systems we studied here have magnetic ground states. Because a bare SiCNT has a nonmagnetic ground state (Alam and Ray 2007, 2008), the net spin of (Fe + SiCNT) system originates from the magnetic moment of the encapsulated and adsorbed Fe atom. Tables 12, 13, 14, 15, 16, and 17 show the electronic states and the spin magnetic moments of Fe atom inside and outside the tube. It is evident that encapsulated Fe atom loses its magnetic moment more inside the smaller tubes where interaction is stronger. Mulliken population analysis shows that the more 4 s electron migration to 4p and 3d orbitals of Fe atom is responsible for this quenching of magnetic moment in case of smaller diameter nanotubes. Magnetic moment of Fe atom placed at the center inside (8, 0) carbon nanotube (Fagan et al. 2003) was found to be 2.36 μB. These values were 2.46 for (3, 3) CNT (Mao et al. 2005) and 2.3 for (4, 4) CNT (Yagi et al. 2004). From Tables 9, 10, and 11 we see these values are higher for all tubes of all types which indicate that Fe atom retains its magnetic moment inside SiCNT more than it does inside CNT. From the values of the magnetic moments of Fe atom for different adsorption sites of three types of tubes, it is obvious that Fe atom magnetic moment are more quenched on type 2 and 3 sites than type 1 sites for (3, 3) and (6, 6), consistent with higher adsorption energy of Fe atom for those sites discussed earlier. The high magnetic moments generated upon encapsulation and adsorption of Fe atom in single wall SiCNT should have important implications in magnetic device applications.

Conclusions

We have studied the interactions of an Fe atom with single wall armchair SiCNT in three different configurations. Our results show that interesting properties can be obtained by Fe atom doping inside and adsorbing on the surface of the SiCNT. Encapsulated Fe atom at the center of the tube has maximum stability in type 2, and then type 3, followed by type 1 nanotubes. For type 1 (6, 6), Si–C zigzag bridge, type 2 (6, 6) C top and type 3 (6, 6) C top sites are the most preferred sites for Fe atom adsorption. Si–C normal bridge, second hollow and the hollow sites are the most stable sites for type 1 (3, 3), type 2 (3, 3), and type 3 (3, 3) nanotubes when outer wall Fe atom adsorption is concerned. Type 2 and 3 nanotubes are predicted to be better candidates for coating with Fe than the most stable type 1 nanotubes. For all the adsorption sites Fe atom transferred charge to the nanotube with some exceptions. There is an obvious relationship between most and least preferred sites and the associated HOMO localization as HOMO localization is more prominent for the most preferred sites. Type 1 nanotubes are most sensitive to band gap modification upon encapsulation and adsorption of adatoms than the other two types. The gaps remain almost unaffected for the other two types when metal atom is placed inside the tube. As the tube size (diameter) increases or in other words curvature decreases gap remains unchanged from the bare tube gap, in response to Fe atom adsorption for type 2 and 3. All the structures studied here have magnetic ground states with very high magnetic moments compared to nonmagnetic ground states of the corresponding bare nanotubes. When the interaction between the Fe atom and the nanotube is stronger, magnetic moment of the Fe atom gets quenched. These small magnetic structures could be useful in magnetic device and spintronic applications. We hope this analysis will be helpful for further exploration of the interactions of other transition metal atoms with silicon carbide nanotubes.

Acknowledgments

The authors gratefully acknowledge the partial support from the Welch Foundation, Houston, Texas (Grant No. Y-1525).

Copyright information

© Springer Science+Business Media B.V. 2008