APP下载

Electronic Properties of Armchair Graphene Nanoribbons with Oxygenterminated Edges: A Density Functional Study①

2015-08-31GEHongYuWANGGuoDepartmentofChemistryCapitalNormalUniversityBeijing100048China

结构化学 2015年5期

GE Hong-Yu WANG Guo(Department of Chemistry, Capital Normal University, Beijing 100048, China)

Electronic Properties of Armchair Graphene Nanoribbons with Oxygenterminated Edges: A Density Functional Study①

GE Hong-Yu WANG Guo②
(Department of Chemistry, Capital Normal University, Beijing 100048, China)

Armchair graphene nanoribbons with different proportions of edge oxygen atoms are analyzed in this study using the crystal orbital method, which is based on density functional theory. Although buckled edges are present, all the nanoribbons are energetically favorable. Unlike the adjacent edge oxygen atoms, the isolated edge oxygen atoms cause semiconductor-metal transitions by introducing edge states. For graphene nanoribbons with all oxygen atoms on the edges,band gap and carrier mobility vary with ribbon width. Furthermore, this behavior is different from that of hydrogen-passivated graphene nanoribbons because of different effective widths, which are pictorially presented with crystal orbitals. The carrier mobilities are as 18%~65% magnitude as those of hydrogen-passivated nanoribbons and are of the order of 103cm2·V−1·s−1.

graphene nanoribbon, semiconductor-metal transition, crystal orbital, carrier mobility, density functional theory;

1 INTRODUCTION

Graphene has attracted extensive attention because of its excellent mechanical and electronic properties[1]. Graphene is a good candidate for nextgeneration electronic devices because of its superior carrier mobility[2, 3]. The finite width of graphene nanoribbons[4, 5]especially, of armchair graphene nanoribbons (AGNRs)[4]can increase the band gap of graphene[6], thus resulting in a high on/off ratio,which is beneficial for device applications. AGNRs with hydrogen-passivated edges exhibit interesting zigzag patterned variation in band gap with width. More excitingly, this theoretically predicted band gap alternation[4]has been recently verified with experiments in ref.[7].

Foreign atoms can significantly affect the electronic properties of graphene[8, 9]. Graphene nanoribbons can be fabricated using top-down lithographic method via oxygen plasma etching[5]. Oxygen atoms as foreign atoms can passivate their active edges. Moreover, ketone structures are also present at the edges of graphene oxide[10]that is prepared by treating graphite with strong oxidizing acids. When compared with graphene, the carrier mobilities are significantly degraded in graphene nanoribbons[11-14]吧However, the influence of edge oxygen atoms has not yet been investigated. Therefore, in this study,we investigate whether the high carrier mobilities are maintained in graphene nanoribbons with oxygen-terminated edges. In addition, the influence of edge oxygen atoms on the electronic properties of graphene nanoribbons is analyzed.

Zigzag graphene nanoribbons with oxygen-terminated edges have been reported to have nearly zero or zero band gaps through theoretical investigations[16, 17]; therefore, a high on/off ratio cannot be easily achieved during electronic device operation. In this study, the properties of AGNRs with different proportions of oxygen atoms at the edges are investigated using the crystal orbital method, which is based on the density functional theory. The edge oxygen atoms are shown to induce semiconductor-metal transitions in graphene nanoribbons. Furthermore, the carrier mobilities are analyzed on the basis of deformation potential theory.

2 METHOD

Primitive cells and supercells of 9-AGNR are used to investigate the influence of edge oxygen atoms. The primitive cell in Fig. 1(a) with H1replaced by an oxygen atom is denoted as 1cell-1,and 2cell-12 denotes the supercell in Fig. 1(b) with H1and H2replaced by two oxygen atoms. The structures are fully optimized using the crystal orbital method via the CRYSTAL06 program[18]. Hybrid density functional B3LYP (VWN5)[19, 20]and Bloch functions constructed with 6-21 G(d) basis set are used in the calculations. The default values of convergence criteria are used in the program. A Monkhorst-Pack sampling with 41 k-points in the first Brillouin zone is used in the self-consistent iteration procedure and is sufficient for obtaining the converged energy and derived electronic properties. A uniform k-points sampling with 800 points is used in the non-iterative band structure calculation for fitting the parameters.

Fig. 1. Models of (a) primitive and (b) supercell of 9-AGNR,(c) optimized structure of 9-AGNR-O (1cell-1234)

According to the deformation potential theory,carriers are primarily scattered by acoustic phonons[21]. For the one-dimensional (1D) oxygenterminated AGNRs, carrier mobilities are given by where, and a0is a lattice parameter of the 1D structure., whereis the energy at the top of the valence band or bottom of the conduction band, andThe stretching modulus (C) and deformation potential constant (E1) of the scattering process due to longitudinal acoustic phonons along the 1D direction are obtained under deformed geometries (a = 0.99, 0.995, 1.005, and 1.01 a0)[22]. In addition, the carrier effective mass (m*) is determined by fitting the frontier bands. It should be noted that carriers without a very sharp density of state near the edge of frontier bands can contribute to the real conduction process in an energy range wider than the thermal energy kBT. In this study, a value of 10kBT[23]is used to calculate the carrier effective mass and carrier mobility. The deformation potential approach has been successfully applied to 1D carbon nanostructures[24-28]. The carrier mobility of graphene calculated using the deformation potential theory[26]is in good agreement with the experimental results[29].

3 RESULTS AND DISCUSSION

3. 1 Structures and stabilities

All optimized structures have closed-shell configurations, which are further verified by spinpolarized calculations using the VASP program[30]. Planar geometries are maintained at the center of nanoribbon, whereas buckled edges can be found when two oxygen atoms are linked to adjacent carbon atoms (1cell-13, 1cell-123, 1cell-1234,2cell-13, 2cell-123, and 2cell-136). In the supposed planar structures (with all the atoms in a plane), the distance between adjacent edge oxygen atoms is in the range of 1.4~1.5 Å; thus, the edge oxygen atoms are too close to one another to be non-bonded. Moreover, the VASP calculations show that the planar structures with all oxygen atoms on the edges are not stable[31]. The stress is released when the oxygen atoms shift to different directions. The buckled structure of 1cell-1234 is shown in Fig. 1(c). As can be seen, the two oxygen atoms at the edge deviate from the original plane of the graphene nanoribbons, and the carbon atoms at the edge deviate from the plane by approximately 10º.

The formation of graphene nanoribbons with hydrogen-terminated edges is energetically favorable when compared with those formed by parent graphene and hydrogen[32]. For graphene nanoribbons with dangling bonds at the edges, the edge oxygen atoms should passivate and stabilize the active bonds. For example, the following reaction can be used to compare the stability of AGNRs with oxygenterminated AGNRs.

In the above reaction, the oxygen-terminated structure has four oxygen atoms at the two edges,and the energy change per edge oxygen atom (ΔE) is −1.59 eV, thus indicating that edge oxidization is energetically favorable. The ΔE values for the other oxygen-terminated structures can be calculated in a similar manner. From Table 1, we can see that all the edge oxidizations are favorable. For the primitive cell, the formation of 1cell-1234 is the most exothermic. For the primitive cells with two edge oxygen atoms (edge oxygen proportion of 1/2), the structure of 1cell-13 is the most stable with a ΔE value of −1.56 eV. The ΔE values for 1cell-12 and 1cell-14 are −0.94 and −0.98 eV, respectively. It is surprising that the structure 1cell-13 with deformed edges is more stable, which will be discussed later. Some structures corresponding to the supercells in Table 1 are also analyzed to investigate more edge oxygen proportions. The 2cell-13 structure with adjacent oxygen atoms and deformed edge is more stable than the 2cell-12 structure with isolated oxygen atoms.

Table 1. Energy Change per Edge Oxygen Atom (ΔE in eV) from AGNR to Edge-oxidized Graphene Nanoribbons, Density of States per Primitive Cell at Fermi Level (DOS in eV−1) of the Edge-oxidized Graphene Nanoribbons, and Edge Oxygen Proportions

Because it is energetically favorable to replace hydrogen atoms with oxygen atoms, the structure with only oxygen atoms on the edges is the most stable. To investigate the variation in properties withnanoribbon width (N), the AGNR with all the edge hydrogen atoms replaced with oxygen atoms (denoted as N-AGNR-O, N = 10~18) is also analyzed. The optimized N-AGNR-O also has buckled geometries at the two edges. In all the AGNR-Os, the two carbon atoms at each edge deviate by approximately 10° while the planar geometries are maintained at the center, which is similar to the structure of 9-AGNR-O (1cell-1234). The ΔE values for these nanoribbons are in the range of −1.61~−1.58 eV. This narrow ΔE range can be attributed to oxidization and the buckled structures that are localized at the edges, and the values are not affected by width.

3. 2 Electronic properties of 9-AGNR with oxygen-terminated edges

The AGNRs with oxygen-terminated edges can be classified into three categories according to their band structures. For the primitive cell, the first category includes 1cell-13, 1cell-1234, and all the AGNR-Os. These structures have even number of oxygen atoms, and the oxygen atoms are linked to the adjacent edge carbon atoms. For example, the band structure of 1cell-1234 is shown in Fig. 2(a). These structures are all semiconductors, and direct band gaps are present at the center of the first Brillouin zone. The second category includes 1cell-1 and 1cell-123. The 1cell-123 has an isolated oxygen atom along with two stable adjacent oxygen atoms. These structures with odd number of oxygen atoms have metallic band structures. For example, the band structure of 1cell-1 with a half-filled band is shown in Fig. 2(b). When there are odd numbers of edge oxygen atoms, the number of electrons in the cell must be also odd; therefore, the structures have half-filled bands. The last category includes 1cell-12 and 1cell-14. The band structure of 1cell-12 is shown in Fig. 2(c). Although these structures have even number of oxygen atoms, their band structures still exhibit metallic character because of the two partially filled bands.

Fig. 2. Band structures of (a) 1cell-1234, (b) 1cell-1 and (c) 1cell-12

To further study the three types of AGNRs with oxygen-terminated edges, frontier crystal orbitals are constructed and analyzed. In 1cell-1, the unpaired electron should be near the oxygen atom. The extension of this edge state is localized at and near the oxygen atom, as shown in Figs. 3(a) and 3(b). The combination of p orbital of oxygen and the p orbitals of edge carbon has a high energy, and they comprise the frontier crystal orbitals at the Fermi level. The bandwidth is narrow and approximately 0.83 eV owing to the edge nature; therefore, the density of states (DOS) is high (1.32 states per eV, as shown in Table 1). For 1cell-123, the frontier crystal orbitals in Fig. 3(c) and 3(d) are also localized edge states generated by the isolated oxygen atom.

Fig. 3. (a) Real and (b) image part of frontier crystal orbitals of 1cell-1, (c) real and (d) image part of frontier crystal orbitals of 1cell-123, (e) HOCO and (f) LUCO of 1cell-13, (g) real and (h) image part of frontier crystal orbitals of 1cell-12 at α point, (i) real and (j) image part of frontier crystal orbitals of 1cell-12 at β point, (k) HOCO and (l) LUCO of 9-AGNR-O, (m) HOCO and (n) LUCO of 10-AGNR-O

The highest occupied crystal orbital (HOCO) and the lowest unoccupied crystal orbital (LUCO) of 1cell-13 shown in Fig. 3(e) and 3(f) are delocalized orbitals involving π states. The two unpaired electrons generated by the two adjacent oxygen atoms form a pair. The energy of the occupied edge state deceases, and the corresponding band does not appear in the frontier states.

The frontier crystal orbitals of 1cell-12 at α and β points (Fig. 2(c)) are shown in Fig. 3(g) and 3(h) and Fig. 3(i) and 3(j), respectively. The imaginary part at α point in Fig. 3(h) and the real part at β point in Fig. 3(i) exhibit edge characteristic. The real part at α point in Fig. 3(g) and the imaginary part at β point in Fig. 3(j) are combinations of isolated p atomic orbitals. Because the two oxygen atoms are not linked to adjacent edge carbon atoms, two edge states exist. The linear combination of the two states and some isolated p atomic orbitals generate two sets of frontier crystal orbitals. As can be seen, the energies at these two points are similar, and two partially filled frontier bands exist in 1cell-12.

The occupied edge localized state has a higher energy than the delocalized state. Therefore, 1cell-12 and 1cell-14 are less stable than 1cell-13. Although 1cell-13 has buckled edges, the pair of high-energy electrons releases more energy, thus increasing the stability of structures having adjacent edge oxygen atoms. Therefore, the 1cell-13 structure with deformed edges is more stable.

The supercells also have similar properties with primitive cells. The 2cell-13 structure with two adjacent oxygen atoms has a semiconductor band structure. The 2cell-1, 2cell-123, and 2cell-126 structures with odd number of oxygen atoms have half-filled metallic bands, and the 2cell-12 structure with two separated oxygen atoms has two partially filled bands. One or two half-filled metallic bands can be also attributed to one or two edge states. The band structures and crystal orbitals of the supercells are similar to those of the primitive cells, which are not shown for clarity. Moreover, the 2cell-12 structure with isolated oxygen atoms is less stable than the 2cell-13 structure with adjacent oxygen atoms.

Isolated edge oxygen atoms can induce semiconductor-metal transitions in AGNRs. The DOS values listed in Table 1 are in the range of 1.10~1.43 states per eV per primitive cell for 1cell-1, 1cell-123,2cell-1, 2cell-123, and 2cell-136 with a half-filledband. The DOS increases to 2.15~2.70 states per eV per primitive cell for the 1cell-12, 1cell-14, and 2cell-12 structures with two partially filled bands. Although the DOS values at the Fermi level are an order of magnitude greater than those of singlewalled carbon nanotubes[33], they are still an order of magnitude less than that of the doped C60[34]. The electron-phonon superconducting temperature should not be higher than that of doped C60. AGNRs with oxygen-terminated edges are not suitable for high-temperature superconducting materials according to the electron-phonon coupling theory. Furthermore, metallic structures are less stable than semiconducting structures. Because semiconducting structures with only oxygen atoms on the edges are the most stable, the electronic properties of these structures are discussed below.

3. 3 Electronic properties of N-AGNR-O

In AGNR-Os, all the band gaps exist at the center of the first Brillouin zone, thus indicating that they are all semiconductors. The band gaps shown in Fig. 4(a)vary with N. The band gaps of 10-AGNR-O,13-AGNR-O, and 16-AGNR-O are less than 0.3 eV,whereas the band gaps of the other structures are in the range of 0.8~1.4 eV. Similar to hydrogenpassivated AGNRs, AGNR-Os can be classified into three groups: N = 3q, N = 3q + 1 and N = 3q + 2,where q is a positive integer. The band gaps of AGNR-Os are in the order of gap3q> gap3q+2>gap3q+1. These values are in agreement with the results obtained with the VASP program[31]but different from the band gaps of AGNRs: gap3q+1>gap3q> gap3q+2[6, 25]. Oxidation can typically increase the band gap; for example, the band gaps of 14-AGNR and 14-AGNR-O are 0.20 and 1.04 eV,respectively. However, this is not always true for graphene nanoribbons. For instance, the band gaps of 13-AGNR and 13-AGNR-O are 1.22 eV and 0.09 eV, respectively because of the different band gap sequences of AGNR-O and AGNR, which will be discussed later. Nevertheless, band gap decreases with ribbon width and is zero for two-dimensional (2D) graphene.

The stretching modulus of 9-AGNR-O is 270 eV·Å−1, and this value gradually increases to 544 eV·Å−1for 18-AGNR-O. The stretching moduli of N-AGNR-Os are 83%~95% of those of corresponding AGNRs calculated using the same density functional and basis set[25], thus indicating that mechanical performance decreases slightly owing to the buckled edges. Nevertheless, the elastic constant is similar to that of AGNR of 1 TPa.

For N-AGNR-Os (N = 10~18), the hole and electron effective masses within 10 kBT from the frontier band edges are in the ranges of 0.09~0.22 and 0.10~0.35 m0, respectively, where m0is the mass of a free electron. For 9-AGNR-O, these values are much greater (0.34 and 0.98 m0) because of the narrow π conjugated system. The carriers in AGNR-O are heavier than those in AGNR (0.07~0.21 and 0.07~0.22 m0for holes and electrons,respectively)[25], so we can see that the edge oxidation decreases the carrier mobilities by enhancing the carrier mass.

For a geometry deformed along a 1D direction,the deformation potential constant can be calculated from the band edge shift with respect to the proportion of lattice deformation[22]. The deformation potential constant of AGNR-O is unbalanced. For example, the valence (E1v) and conduction band deformation potential constants (E1c) of 9-AGNR-O are 9.3 and 6.3 eV, respectively. As shown in Fig. 4(b),E1vis greater than E1cwhen N = 3q + 2, whereas E1cis greater than E1vwhen N = 3q or 3q+1 except for only the narrow 10-AGNR-O structure. The higher deformation potential constants are in the range of 9.3~12.0 eV, whereas the small values fall in the 5.0~7.1 eV region. This will significantly affect the carrier mobility as it depends on the power of deformation potential constant in Eq. 1.

The differences in deformation potential constants can be explained by crystal orbitals. Because deformation potential depends on band edge shift,the frontier crystal orbitals at the band edges can give pictorial descriptions. In Figs. 3(k) and 3(l), the HOCO and LUCO of 9-AGNR-O are localized anddelocalized with respect to the 1D deformation direction, respectively. Upon deformation, the variation in extended delocalized orbital along the 1D conjugated line with bond length is less than that of the localized orbital. This results in a smaller energy variation in the delocalized orbital, thus leading to a smaller deformation potential constant. For other AGNR-Os except 10-AGNR-O, delocalized orbitals are present at the valence (conduction)band edge when N = 3q or 3q+1 (3q+2). From 2D graphene to 1D AGNR-O, quantum confinement causes the variation in the properties of frontier orbitals, thus resulting in unbalanced deformation potential constants.

It should be noted that the carbon atoms at the two edges of 9-AGNR-O shown in Figs. 3(k) and 3(l)also contribute to the frontier crystal orbitals because of edge oxidation and the buckled structure. The number of carbon atoms involved in the conjugation should not include the edge carbon atoms. Although the p orbitals of the edge oxygen atoms are present in the frontier crystal orbitals, they are isolated from the main conjugated orbitals at the center. The electronic properties of N-AGNR-O are similar to those of (N-2)-AGNR. For AGNR-O when N = 3q+2, E1vis greater than E1c, whereas in the case of AGNR, this is true when N = 3q[25]. Considering the edge situation, the two rules are the same. For the band gap alternation discussed above, the difference between AGNR-O and AGNR is because of the same reason.

In 10-AGNR-O, the frontier orbitals shown in Figs. 3(m) and 3(n) are not distributed over all the central carbon atoms. This narrow nanoribbon with a very small band gap has isolated conjugated structures, and the two carbon lines at the center do not contribute to the frontier crystal orbitals. The total number of carbon atoms in the conjugated structures changes; therefore, the deformation potential constants of 10-AGNR-O do not obey the above rule.

Fig. 4(c) shows the variation in carrier mobility with N. The electron mobility (μe) is greater than the hole mobility (μh) when N = 3q + 2 because E1cis less than E1v. For N = 3q or 3q + 1, μhis greater than μe. The only exception is 10-AGNR-O because of the unusual isolated conjugated lines. The unbalanced deformation potential constants cause unbalanced carrier mobilities. The polarity[24]of the carrier mobilities depends on the polarity of the deformation potential constants, and the different rules for the carrier mobilities of AGNR-O and AGNR[25]can be considered to be the same, as in the case of deformation potential constant.

Fig. 4. (a) Band gaps, (b) deformation potential constants and (c) carrier mobilities of AGNR-O

Carrier mobility typically increases with width for larger stretching moduli and larger conjugated structures. The highest value of 4096 cm2·V−1·s−1corresponds to the electron mobility of 17-AGNR-O,and this value is 40% that of 15-AGNR[25]. A weaker stretching modulus caused by edge oxygen and edge-buckled structures, higher deformation potential constant, and slightly higher carrier effective mass are responsible for the reduction in mobility. However, the reduction is not significant. The major carrier mobilities of N-AGNR-O are as 18%~65% magnitude as those of (N-2)-AGNR. The major carrier mobilities are of the order of 103cm2·V−1·s−1and are still quite high for nanoelectronic devices.

4 CONCLUSION

Armchair graphene nanoribbons with oxygenterminated edges are analyzed on the basis of density functional theory. When compared with hydrogen-terminated nanoribbons, all the nanoribbons with different proportions of edge oxygen atoms are stable even though buckled edges are present. The nanoribbons can be classified into three categories depending on their band structures. The first category has semiconductor band structures with adjacent edge oxygen atoms, whereas the second category has metallic band structures with a half-filled band having an isolated edge oxygen atom. The last category of AGNRs has metallic band structures with two partially filled bands having two isolated edge oxygen atoms. The edge states, which are shown with crystal orbitals, are responsible for the semiconductor-metal transitions.

The band gap varies with width for the most stable AGNR having only oxygen atoms on the edges. The band gaps are in the order of gap3q>gap3q+2> gap3q+1. The AGNR-O and AGNR have different rules because of the buckled edges of AGNR-O, which do not contribute to the frontier crystal orbitals. For graphene nanoribbons, edge oxidation affects the band gap and the polarity of carrier mobility depending on the different quantum confinements because the edge carbon atoms that do not contribute to the frontier crystal orbitals are excluded when determining the ribbon width. The edge oxygen atoms have limited negative impact on carrier mobilities, and the highest mobilities are still of the order of 103cm2·V−1·s−1.

REFERENCES

(1) Geim, A. K. Graphene: status and prospects. Science 2009, 324, 1530-1534.

(2) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Katsnelson, M. I.; Grigorieva, I. V.; Dubonos, S. V. Two-dimensional gas of massless Dirac fermions in graphene. Nature 2005, 438, 197-200.

(3) Zhang, Y.; Tan, Y.; Stormer, H. L.; Kim, P. Experimental observation of the quantum Hall effect and Berry's phase in graphene. Nature 2005, 438,201-204.

(4) Son, Y. W.; Cohen, M. L.; Louie, S. G. Energy gaps in graphene nanoribbons. Phys. Rev. Lett. 2006, 97, 216803-1-216803-4.

(5) Han, M. Y.; Özyilmaz, B.; Zhang, Y.; Kim, P. Energy band-gap engineering of graphene nanoribbons. Phys. Rev. Lett. 2007, 98,206805-1-206805-5.

(6) Wallace, P. R. The band theory of graphite. Phys. Rev. 1947, 71, 622-634.

(7) Magda, G. Z.; Jin, X.; Hagymási, I.; Vancsó, P.; Osváth, Z.; Nemes-Incze, P.; Hwang, C.; Biró, L. P.; Tapasztó, L. Room-temperature magnetic order on zigzag edges of narrow graphene nanoribbons. Nature 2014, 514, 608-611.

(8) Jia, T. T.; Sun, B. Z.; Lin, H. X.; Li, Y.; Chen, W. K. Bonding of hydroxyl and epoxy groups on graphene: insights from density functional calculations. Chin. J. Struct. Chem. 2013, 32, 1475-1484.

(9) Guan, X. M.; Zhang, H. Y.; Zhang, M.; Luo, Y. H. Opening band gap of graphene by chemical doping: a first principles study. Chin. J. Struct. Chem. 2014, 33, 513-518.

(10) Szabó, T.; Berkesi, O.; Dékány, I. DRIFT study of deuterium-exchanged graphite oxide. Carbon 2005, 43, 3186-3189.

(11) Li, X.; Wang, X.; Zhang, L.; Lee, S.; Dai, H. Chemically derived, ultrasmooth graphene nanoribbon semiconductors. Science 2008, 319, 1229-1232.

(12) Wang, X.; Ouyang, Y.; Li, X.; Wang, H.; Guo, J.; Dai, H. Room-temperature all-semiconducting sub-10-nm graphene nanoribbon field-effect transistors. Phys. Rev. Lett. 2008, 100, 206803-1-206803-4.

(13) Jiao, L.; Wang, X.; Diankov, G.; Wang, H.; Dai, H. Facile synthesis of high-quality graphene nanoribbons. Nature Nanotechnol. 2010, 5, 321-325.

(14) Lin, M.; Ling, C.; Agapito, L. A.; Kioussis, N.; Zhang, Y.; Cheng, M. M.; Wang, W. L.; Kaxiras, E.; Zhou, Z. Approaching the intrinsic band gap in suspended high-mobility graphene nanoribbons. Phys. Rev. B 2011, 84, 125411-1-125411-7.

(15) Fang, T.; Konar, A.; Xing, H.; Jena, D. Mobility in semiconducting graphene nanoribbons: phonon, impurity, and edge roughness scattering. Phys. Rev. B 2008, 78, 205403-1-205403-8.

(16) Hod, O.; Barone, V.; Peralta, J. E.; Scuseria, G. E. Enhanced half-metallicity in edge-oxidized zigzag graphene nanoribbons. Nano. Lett. 2007, 7,2295-2299.

(17) Lee, G.; Cho, K. Electronic structures of zigzag graphene nanoribbons with edge hydrogenation and oxidation. Phys. Rev. B 2009, 79,165440-1-165440-12.

(18) Dovesi, R.; Saunders, V. R.; Roetti, C.; Orlando, R.; Zicovich-Wilson, C. M.; Pascale, F.; Civalleri, B.; Doll, K.; Harrison, N. M.; Bush, I. J.; D'Arco,Ph.; Llunell, M. CRYSTAL06 User's Manual. University of Torino, Italy 2007.

(19) Becke, A. D. Density functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 1993, 98, 5648-5652.

(20) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys. Rev. B 1988, 37, 785-789.

(21) Bardeen, J.; Shockley, W. Deformation potentials and mobilities in non-polar crystals. Phys. Rev. 1950, 80, 72-80.

(22) Wang, G.; Huang, Y. Theoretical study on the co-crystal composed of poly(diiododiacetylene) and bis(nitrile) oxalamide. J. Phys. Chem. Solids 2007,68, 2003-2007.

(23) Grimvall, G. The Electron-phonon Interaction in Metals (pp 16-17). North-Holland Publishing Company, Amsterdam 1981.

(24) Long, M.; Tang, L.; Wang, D.; Wang, L.; Shuai, Z. Theoretical predictions of size-dependent carrier mobility and polarity in graphene. J. Am. Chem. Soc. 2009, 131, 17728-17729.

(25) Wang, G. Effect of edge-hydrogen passivation and saturation on the carrier mobility of armchair graphene nanoribbons. Chem. Phys. Lett. 2012, 533,74-77.

(26) Xi, J.; Long, M.; Tang, L.; Wang, D.; Shuai, Z. First-principles prediction of charge mobility in carbon and organic nanomaterials. Nanoscale 2012,4, 4348-4369.

(27) Wang, G.; Huang, Y. Crystal orbital study on the zigzag single-walled carbon nanotubes with multi-dichlorocarbene addition. J. Phys. Chem. Solids 2008, 69, 2531-2534.

(28) Qiao, W.; Bai, H.; Li, R.; Huang, Y. Structures and electronic properties of the one-dimensional C60O polymers. Comput. Mater. Sci. 2012, 55,175-179.

(29) Morozov, S. V.; Novoselov, K. S.; Katsnelson, M. I.; Schedin, F.; Elias, D. C.; Jaszczak, J. A.; Geim, A. K. Giant intrinsic carrier mobilities in graphene and its bilayer. Phys. Rev. Lett. 2008, 100, 016602-1-016602-4.

(30) Kresse, G.; Furthmüller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B 1996, 54,11169-11186.

(31) Simbeck, A. J.; Gu, D.; Kharche, N.; Satyam, P. V.; Avouris, P.; Nayak, S. K. Electronic structure of oxygen-functionalized armchair graphene nanoribbons. Phys. Rev. B 2013, 88, 035413-1-035413-10.

(32) Wassmann, T.; Seitsonen, A. P.; Saitta, A. M.; Lazzeri, M.; Mauri, F. Structure, stability, edge states, and aromaticity of graphene ribbons. Phys. Rev. Lett. 2008, 101, 096402-1-096402-4.

(33) Wang, G.; Huang, Y. A density functional study on transition-metal-coated single-walled carbon nanotubes. J. Phys. Chem. C 2008, 112, 9128-9132.

(34) Gunnarsson, O. Superconductivity in fullerides. Rev. Mod. Phys. 1997, 69, 575-606.

Received 23 December 2014; accepted 20 January 2015

① This work was supported by the National Natural Science Foundation of China (No. 21203127), the Beijing Higher Education Young Elite Teacher Project (YETP1629) and the Scientific Research Base Development Program of the Beijing Municipal Commission of Education

② Corresponding author. E-mail: wangguo@mail.cnu.edu.cn

10.14102/j.cnki.0254-5861.2011-0611


登录APP查看全文