The spin-dependent transport of transition metal encapsulated B₄₀ fullerene

Abstract

The all-boron fullerene, B₄₀, has been successfully experimentally synthesized [Zhai et al., Nat. Chem., 6, 727 (2014)]. Compared with C₆₀, the smaller cage-like structure is more suitable to dope with metal atoms. Based on density functional theory and nonequilibrium Green’s function method, we investigate the spin-dependent transport of transition metal atom-encapsulated B₄₀ fullerene, i.e., X@B₄₀ (X = Fe, Mn, Ni, and Co), which are contacted with Au electrodes. The transmission spectra of Fe- and Mn-doped systems are spin-polarized, and those of Ni-doped ones are spin-unpolarized. Interestingly, in Co-doped systems, the transmission is highly spin-polarized for the hexagonal doping case, but spin-unpolarized for the heptagonal doping case. Further investigation shows that the screening effect of the electrodes on the magnetism of Co is the underlying physical mechanism, which is found to be robust to the electrode material. We believe that these findings are very useful for developing spintronic devices.

---

Introduction

Since the discovery of the first fullerene (C₆₀),¹ extensive research on carbon fullerene has been carried out.^2–4 Until now, various structural and electronic properties have been investigated, and various potential applications have been proposed, e.g., data storage,⁵ switching⁶ and transistors.⁷ Although a great deal of research on carbon fullerene has been performed, fullerene cage-like structures composed of other elements have always been sought including boron, which is one of the nearest neighbours of carbon in the periodic table. For instance, a B₃₈ fullerene analogue has been predicted using first principles swarm structure searching calculations.⁸

Recently, a breakthrough has been achieved in that an all-boron fullerene (B₄₀) has been successfully experimentally synthesized for the first time.⁹ B₄₀ possesses four heptagons and two hexagons with D_2d symmetry, different from C₆₀.¹ In addition, its diameter is smaller than that of C₆₀.¹⁰ As the first all-boron fullerene, it has attracted increasing attention, and some interesting properties have been revealed. For example, Yang et al.¹¹ reported that, when contacted with Au electrodes, B₄₀ could exhibit a remarkable negative differential resistance (NDR) and large rectification ratio.

Besides its intrinsic features, another attraction of fullerenes is their unique cage-like structure, which is suitable to dope with other elements and molecules to modify the electronic structure. For example, it has been found that doping the carbon fullerene structure could allow it to exhibit various electronic and transport properties.^12–14 Especially, doping with magnetic metals could allow interesting magnetic properties and spin-dependent transport features to be exhibited.^15–20 Actually, compared with C₆₀, B₄₀ is more suitable to dope with other atoms or clusters because of its smaller cage diameter.^9,10 However, a study on magnetic atom-encapsulated B₄₀ and its spin-related transport is still lacking.

In this work, we investigate the spin-dependent transport properties of transition metal atom-encapsulated B₄₀ fullerene, i.e., X@B₄₀ (X = Fe, Mn, Ni, and Co), contacted with Au electrodes using density functional theory (DFT) and nonequilibrium Green’s function (NEGF). For B₄₀, its hexagonal and heptagonal holes are suitable candidates for doping sites to accommodate metal atoms inside to form endohedral X@B₄₀,¹⁰ like X@C₆₀.^21–23 In this paper we confine our studies to these two hollow doping site cases, i.e., hollow hexagonal and heptagonal ones. It is found that Fe- and Mn-doped systems exhibit highly spin-polarized transmission and Ni-doped systems exhibit completely spin-unpolarized transmission. However, Co-encapsulated B₄₀ systems present strange behavior, where the transmission spectra for the hexagonal hole-doped case are highly spin-polarized, but those for the heptagonal hole-doped case are spin-unpolarized. Through further investigation, we find that this is caused by the screening effect of the electrodes on the magnetism of the Co atom, which is robust to the electrode material. Our results are quite useful for the manipulation of spin-dependent transport, showing potential application in future spintronic devices.

Methodology

Our investigation is based on a first principles technique. To explore the transport of the system precisely, we carried out calculations based on the combination of DFT and NEGF, which are performed by making use of the Atomistix Toolkit package (ATK).^24,25 We used a mesh cutoff energy of 150 Ry. The k-point mesh in the Monkhorst–Park scheme²⁶ is set to be 1 × 1 × 100. The Perdew–Burke–Ernzerhof (PBE)²⁷ formulation of the generalized gradient approximation (GGA) for the exchange–correlation functional and the double-zeta polarized (DZP) basis set of the local numerical orbitals were applied in our calculations. The tolerance of the iteration is chosen as 4 × 10⁻⁵ hartrees. Before calculating transport properties, we performed geometric optimization for the isolated metal atom doped B₄₀, until all the forces are less than 0.02 eV Å⁻¹. After optimizing the doped molecules, we contact them with electrodes without geometric optimization for the contact region, where the electrode–molecule distance (d₀) is determined by minimizing the total energy of the whole system. Such a contact method is widely used in transport investigations.^12,28 To avoid interactions with adjacent images, a sufficient vacuum space of the supercell (more than 10 Å) was selected. As the transport under low bias is mainly dominated by zero-bias conditions, we focus our research on this situation in this work.

Results and discussion

Previous studies have shown that transition metal encapsulated systems exhibit interesting magnetic features.^29–33 In this paper, we chose Fe, Mn, Ni and Co to dope in B₄₀ fullerene. In both theory and experiment, it is found that these (Fe, Co, Mn, Ni) transition metal encapsulated systems exhibit interesting magnetic properties.^33–37 Thus, it is expected interesting spin-related transport behavior would be found in these transition metal encapsulated B₄₀ systems. Corresponding two-probe systems shown in Fig. 1 were built to calculate the spin-dependent transport properties. The (001) surface of centered cubic (fcc) Au with a pyramidal tip is selected as the left and right electrode. For the sake of clarity, a transition metal atom (X) doped in B₄₀ is denoted by X@B₄₀, and Au–X@B₄₀–Au denotes the two-probe system contacted with Au electrodes. Considering the D_2d symmetry of B₄₀, the doping sites, and the orientation of B₄₀ relative to the electrodes, we choose and define four kinds of two-probe configurations, i.e., R6E7, R6E6, R7E6 and R7E7, see Fig. 1(b)–(e) for the respective central molecular parts. For instance, R6E7 denotes the configuration where the X atom is doped on the hollow hexagonal site of B₄₀ and the heptagonal ring contacts with the Au electrodes, see Fig. 1(a).

Fig. 1 (a) Geometric view of the Au–X@B₄₀–Au system (in R6E7 configuration). R6E7 denotes that the X atom is doped on the hollow hexagonal site of B₄₀ and the heptagonal ring contacts with the Au electrodes. (b)–(e) The central parts of X@R6E7, X@R6E6, X@R7E6, and X@R7E7 respectively. X = Fe, Mn, Ni, and Co. The molecule–electrode distances (d₀) are 1.8 and 1.6 Å for the hexagonal (E6) and heptagonal (E7) doping-site cases, respectively.

For Fe-doped systems, the spin-dependent transmission spectra are shown in Fig. 2. The Fermi level, E_F, is set to be the zero energy. As the electron transport properties mainly depend on transmission around the Fermi level, we pay attention to the transmission in this area. For the Fe@R6E6 system in Fig. 2(a), where the Fe atom is doped on the hollow hexagonal site and the molecular hexagonal ring contacts with the electrodes, there is a wide up-spin transmission peak around the Fermi level. Meanwhile, the down-spin transmission is much smaller in this energy region. We define the spin polarization here as (up − down)/(up + down). Apparently, the spin polarization around E_F is quite large, which would be very useful in spintronic devices. However, when we change the orientation of the molecule–electrodes to Fe@R6E7, where the molecular heptagonal ring contacts with the electrodes, the down-spin transmission does not change a lot but the up-spin transmission peak splits into two peaks around E_F, see Fig. 2(b). Although the up-spin transmission undergoes a large variation, the spin-polarization of the transmission is still quite large. However, when doping the Fe atom on the hollow heptagonal site, it becomes a little complicated. For the Fe@R7E6 configuration in Fig. 2(c), the up-spin transmission decreases and the down-spin transmission increases around E_F. As a result, the spin-polarization becomes smaller compared with that of Fe@R6E6 and of Fe@R6E7. When changing the molecular orientation to Fe@R7E7 in Fig. 2(d), the up-spin and down-spin transmissions are almost the same around the Fermi level. Consequently, the spin-polarization is nearly zero in this energy region. Thus, the spin-polarization of the transmission is sensitive to both the doping site and the molecular orientation relative to the electrodes in Fe-doped systems. From Fig. 2(a)–(d), one finds that the spin-polarization of the transmission decreases, i.e., the transmission changes from a highly spin-polarized state to a spin-unpolarized state. Therefore, the modulation of spin-polarization could be realized by adjusting the configurations, which would be helpful for the design of spintronic devices.

Fig. 2 Transmission spectra of the Au–Fe@B₄₀–Au system. (a)–(d) Denote Fe@R6E6, Fe@R6E7, Fe@R7E6 and Fe@R7E7, respectively. Up-spin: solid, black line. Down-spin: dashed, red line. The Fermi level is set to be the zero energy.

For Mn-doped systems, the spin-dependent transmission spectra are shown in Fig. 3. For the configuration of Mn@R6E6 in Fig. 3(a), both the up- and down-spin transmissions split into two peaks around the Fermi energy. From the figure, one could conclude that the spin-polarization around this energy region is not large. For the configuration of Mn@R6E7 in Fig. 3(b), the up- and down-spin transmissions are almost the same around E_F, which result in almost zero spin-polarization. While, from Fig. 3(b)–(d), the difference between up- and down-spin transmissions becomes larger and larger, and the spin-polarization becomes higher and higher. Thus a similar configuration modulated spin-polarization effect, like the Fe-doped case, is also found in the Mn-doped systems.

Fig. 3 Transmission spectra of the Au–Mn@B₄₀–Au system. (a)–(d) Denote Mn@R6E6, Mn@R6E7, Mn@R7E6 and Mn@R7E7, respectively. Up-spin: solid, black line. Down-spin: dashed, red line. The Fermi level is set to be the zero energy.

However, the Ni-doped case is quite different from the Fe- and Mn-systems. Fig. 4 shows the spin-dependent transmission spectra of this case. For each configuration, in the energy regions both near and far from E_F, the up-spin transmission is as large as the down-spin one. That is to say, the electronic transmission is completely spin-unpolarized for the entire energy region (the spin-polarization is zero). Comparing the transmission spectra of Ni@R6E6 in Fig. 4(a) and Ni@R7E6 in Fig. 4(c), one finds that they are quite similar around the Fermi energy, as are the transmission spectra of Ni@R6E7 in Fig. 4(b) and Ni@R7E7 in Fig. 4(d). Apparently, the doping site has little effect on the transmission of the two-probe systems in the Ni-doped cases. However, going from Ni@R6E6 in Fig. 4(a) to Ni@R6E7 in Fig. 4(b) [or from Ni@R7E6 in Fig. 4(c) to Ni@R7E7 in Fig. 4(d)], the transmission undergoes a large change around E_F. So in Ni-doped systems, transport is more sensitive to the orientation of the molecule relative to the electrodes, and both doping site and orientation have no effect on the spin-polarization of transport, which remains zero.

Fig. 4 Transmission spectra of the Au–Ni@B₄₀–Au system. (a)–(d) Denote Ni@R6E6, Ni@R6E7, Ni@R7E6 and Ni@R7E7, respectively. Up-spin: solid, black line. Down-spin: dashed, red line. The Fermi level is set to be the zero energy.

Interestingly, Co-doped systems exhibit quite strange behavior, different from that of the former three cases, i.e., Fe-, Mn- and Ni-doped ones. When the Co atom is encapsulated in the hollow hexagonal site of B₄₀, the system exhibits spin-polarized transmission around E_F, for both the two kinds of orientations [Co@R6E6 in Fig. 5(a) and Co@R6E7 in Fig. 5(b)]. However, when the Co atom is doped in the hollow heptagonal site, the transmission spectra become completely spin-unpolarized, including the deep energy regions. In other words, when we change the doping site from hollow hexagonal one to heptagonal one, a transition of the transmission from spin-polarized state to spin-unpolarized state occurs. Thus, the doping site is crucial to the spin transport in Co-doped systems. For the orientation of the molecule, it is also important to the transmission, as the transmission spectra change a lot from Co@R6E6 in Fig. 5(a) to Co@R6E7 in Fig. 5(b) [or from Co@R7E6 in Fig. 5(c) to Co@R7E7 in Fig. 5(d)]. But, apparently this is not the key factor that triggers the (spin-) polarized–unpolarized transition.

Fig. 5 Transmission spectra of the Au–Co@B₄₀–Au system. (a)–(d) Denote Co@R6E6, Co@R6E7, Co@R7E6 and Co@R7E7, respectively. Up-spin: solid, black line. Down-spin: dashed, red line. The Fermi level is set to be the zero energy.

In order to figure out the mechanism of spin-dependent transport, we calculated the Mulliken population of the X atom in the isolated X@B₄₀ molecule, which is not contacted with the electrodes. Table 1 shows the Mulliken populations for up-spin, down-spin, and their difference, i.e., the magnetic moment. One finds that the magnetic moments of Fe and Mn atoms are quite large for both of the two kinds of doping site configurations, i.e., 2.030, 2.111, 1.479 and 2.868 μ_B for Fe@R6, Fe@R7, Mn@R6 and Mn@R7, respectively. Obviously, it is these non-zero magnetic moments in X@B₄₀ that result in the spin-polarized transmission of the Fe- and Mn-doped systems. For Ni-doped systems, the magnetic moments are zero. As Au electrodes and isolated Ni@B₄₀ (for both of the two doping cases) both have no magnetic moment, it is reasonable that the combination of them, Au–Ni@B₄₀–Au, exhibits spin-unpolarized transport. Thus, for Fe-, Mn- and Ni-doped systems, the non-zero/zero magnetic moment of the molecule results in spin-polarized/unpolarized transport. However, this rule is not valid in Co-doped systems. For Co-doped systems, the magnetic moment of the Co atom in Co@R6 is 0.868 μ_B. Although it is smaller than that of the Fe and Mn systems, it can still induce spin-polarized transmission in the two-probe systems, as shown in Fig. 5(a) and (b). The magnetic moment of Co in Co@R7 is 0.988 μ_B, which is even larger than that of Co@R6 (0.868 μ_B), but the corresponding two-probe systems exhibit completely spin-unpolarized transmission; see Co@R7E6 and Co@R7E7 in Fig. 5(c) and (d), respectively. It is interesting to find out the physical mechanism underlying this strange behavior.

Table 1 Mulliken populations of the X metal atom in isolated X@B₄₀ which is not contacted with the Au electrodes (X = Fe, Mn, Ni, and Co) for up-spin, down-spin, and their difference (up-down, i.e., the magnetic moment). The unit of magnetic moment is μ_B

Structure	R6			R7
	↑	↓	↑ − ↓	↑	↓	↑ − ↓
Fe@B40	4.791	2.761	2.030	4.825	2.714	2.111
Mn@B40	4.055	2.576	1.479	4.713	1.845	2.868
Ni@B40	4.744	4.744	0	4.742	4.742	0
Co@B40	4.702	3.834	0.868	4.758	3.770	0.988

---

The two-probe system is constructed by contacting the molecule with electrodes. Although the Au electrode is non-magnetic, it may still influence the magnetic properties of the molecule and then affect transport. In the above, we conclude that it is the non-zero and zero magnetic moments of the isolated molecules that result in spin-polarized (Fe- and Mn-doped cases) and unpolarized (Ni-doped case) transport, respectively. To confirm this in the two-probe system and also to explore the mechanism in Co-doped systems, we calculated the Mulliken populations for the Au–X@B₄₀–Au systems. The results are shown in Table 2. As the encapsulated transition metal atom is the source of the magnetic moment in the whole two-probe system, only the magnetic moments of this atom are presented, i.e., the Mulliken population of the X metal atom in Au–X@B₄₀–Au (X = Fe, Mn, Ni, and Co). From Table 2, one finds that the magnetic moment is very large for the Fe- and Mn-doped systems, and for the Ni-doped ones, the magnetic moments are (almost) zero. This is consistent with the results of the isolated doping molecules. For Co-doped systems, the magnetic moments of Co are quite different between the two kinds of doping-site cases. In the R6 doping cases, the magnetic moments are 0.898 and 0.465 μ_B for the configurations of R6E6 and R6E7, respectively. Although compared with the isolated molecule (0.868 μ_B) the magnetic moment of Co in R6E7 becomes smaller, it is still a non-zero value. These results are consistent with the transmission spectra of R6E6 and R6E7, which are both spin-polarized [see Fig. 5(a) and (b) respectively]. However, in the Co@R7 doping cases, the magnetic moments become almost zero, i.e., 0.002 and 0.001 μ_B for R7E6 and R7E7 respectively. These magnetic moments are completely quenched compared with that in the isolated R7 doping molecule (0.988 μ_B). This phenomenon is quite interesting, and it is no doubt induced by contacting with electrodes, which we will discuss in the following section. It is easy to know that the spin-unpolarized transmission spectra of Co@R7E6 and Co@R7E7 result from the (almost) zero magnetic moments of the two-probe systems.

Table 2 The Mulliken populations of the X metal atom in Au–X@B₄₀–Au (X = Fe, Mn, Ni, and Co) for up-spin, down-spin, and their difference (up-down, i.e., the magnetic moment). The unit of magnetic moment is μ_B

Structure	R6E6			R6E7			R7E6			R7E7
	↑	↓	↑ − ↓	↑	↓	↑ − ↓	↑	↓	↑ − ↓	↑	↓	↑ − ↓
Fe@B40	4.686	2.876	1.810	4.598	2.974	1.624	4.707	2.846	1.861	4.612	2.966	1.646
Mn@B40	4.372	2.268	2.104	4.176	2.460	1.716	4.625	1.950	2.675	4.508	2.093	2.415
Ni@B40	4.752	4.750	0.002	4.747	4.747	0	4.744	4.742	0.002	4.749	4.748	0.001
Co@B40	4.723	3.825	0.898	4.510	4.045	0.465	4.283	4.281	0.002	4.288	4.287	0.001

---

For the hollow heptagonal doping site (Co@R7), the magnetic moment of the Co atom goes down to almost zero by contacting the electrodes. Thus, the magnetism of Co could be seen as being screened by the electrodes. Actually, such a screening effect has been found before in transition metal encapsulated Si cage systems.³³ In those systems, the magnetism of Co atom is largely screened by the electrodes, but that of Fe and Mn is not.³³ Consequently, the transport of the Fe- and Mn-doped Si-cage two-probe systems is spin-polarized and of the Co-doped system is spin-unpolarized,³³ consistent with that of our systems. Apparently, the screening effect of the electrodes is crucial to such Co-doped systems.

The screening behavior between the two kinds of doping-site cases (R6 and R7) appears quite different. Next, we will show that the underlying mechanisms actually are the same. The difference is actually induced by the different stable electrode–molecule distances, which are determined by minimizing the total energy of the system. These values vary with configurational change and this results in different strengths of the screening effect.

To explore the screening effect of the electrodes on the magnetic moment of Co, we changed the distance between the molecule and Au electrodes to see the variation of the transmission (it should be noted such configurations are not stable, and this is only used to observe the screening effect). As we know, the screening effect is realized through electrode–molecule coupling. Changing the distance will change the coupling, and result in its weakening and strengthening for longer and shorter distances, respectively. Consequently, the strength of screening effect on the magnetism of Co would change correspondingly. According to this, the magnetic moments of the R6-doping cases might be quenched, and those of the R7-doping cases might possess a non-zero magnetic moment, different from the stable structure situations. Thus, it is expected that spin-unpolarized transmission would be seen in Co@R6E6 and Co@R6E7, and spin-polarized transmission in Co@R7E6 and Co@R7E7.

Fig. 6 and 7 show the transmission spectra under the distances of d = d₀ + Δd (here, Δd = 0.4 and −0.8 Å for Co@R7E6 and Co@R7E7, respectively, and d₀ is the distance in the stable structure for each configuration). They are representative of the two categories (increasing and decreasing the molecule–electrode distances). In Fig. 6, one finds that all the transmission spectra become spin-polarized. Especially, the transmission spectra of Co@R7E6 and Co@R7E7 become spin-polarized [Fig. 6(c) and (d)], although the spin-polarization of Co@R7E7 is not very large. As discussed above, they are completely spin-unpolarized for their stable structures [d = d₀, Fig. 6(c) and (d)]. Obviously, elongating molecule–electrode distance weakens the coupling and screening effect, and then the magnetism of Co is preserved, which results in spin-polarized transmission.

Fig. 6 Transmission spectra of the Au–Co@B₄₀–Au system (here, the distance between the Co@B₄₀ and Au electrodes is increased by 0.4 Å relative to the stable structure). (a)–(d) Denote Co@R6E6, Co@R6E7, Co@R7E6 and Co@R7E7, respectively. Up-spin: solid, black line. Down-spin: dashed, red line. The Fermi level is set to be the zero energy.

Fig. 7 Transmission spectra of the Au–Co@B₄₀–Au system (here, the distance between the Co@B₄₀ and Au electrodes is decreased by 0.8 Å relative to the stable structure). (a)–(d) Denote Co@R6E6, Co@R6E7, Co@R7E6 and Co@R7E7, respectively. Up-spin: solid, black line. Down-spin: dashed, red line. The Fermi level is set to be the zero energy.

Surprisingly, when we decrease the molecule–electrode distance by 0.8 Å relative to the stable configurations, all the transmission spectra tend to be spin-unpolarized, see Fig. 7. Only the transmission spectrum of Co@R6E6 still exhibits minor spin-polarization. No doubt the shorter distance strengthens the coupling and the screening effect, and then quenches the magnetism and transmission.

Thus, in the Co-doped system, changing molecule–electrode distance could modulate the coupling, the screening effect and the corresponding transmission, which even changes from spin-polarized to spin-unpolarized states. To see this more clearly, we take Co@R7E6 and Co@R7E7 as examples and calculate their molecular projected self-consistent Hamiltonian (MPSH) varying with distance. The MPSH is obtained by projecting the Hamiltonian of the two-probe system onto the scattering region. Then it could be seen as an isolated molecule and the energy spectrum calculated, where the electrodes’ influence has been taken into account. For our system, the projection region is shown in Fig. 1(a), denoted by the rectangle. As transport is mainly dominated by the electronic structures near the Fermi level, here we chose the nearest orbital to E_F to plot the distribution in real space. For Co@R7E6, it is the highest occupied molecular orbital (HOMO) in up-spin, and for Co@R7E7, it is the lowest unoccupied molecular orbital (LUMO) in down-spin. They are plotted in the left and right panels of Fig. 8, respectively. Note that for the purpose of comparison the isovalues of all the plotted isosurfaces are the same. For the cases when Δd = 0 and −0.8 Å (the middle and right panels of Fig. 8), the transmission spectra are spin-unpolarized, i.e., Fig. 8(d), (g), (e) and (h), and the corresponding MPSH orbitals distribute throughout the whole two-probe systems. For those, there are large couplings between electrodes and the molecule, while when the distance is increased, the transmission becomes spin-polarized, see Fig. 8(f) and (i), and the corresponding MPSH orbitals become quite different from the former ones. The orbitals are mostly localized in the electrodes and there is only a little distribution in the molecule [Fig. 8(c) and (l)], suggesting a weak coupling between them. Consequently, it weakens the screening effect and results in spin-polarized transport.

Fig. 8 (a)–(c) Molecular projected self-consistent Hamiltonians (MPSH) of the HOMO in up-spin for Co@R7E6 where Δd = −0.8, 0 and 0.4 Å, respectively. (d)–(f) Transmission spectra of Co@R7E6 where Δd = −0.8, 0 and 0.4 Å, respectively. (g)–(i) MPSH of the LUMO in up-spin for Co@R7E7 where Δd = −0.8, 0 and 0.4 Å, respectively. (j)–(l) Transmission spectra of Co@R7E7 where Δd = −0.8, 0, and 0.4 Å, respectively. The isovalue of all the plotted isosurfaces is 0.01 e Å⁻³.

In the above, Au electrodes are used in the two-probe systems. In order to figure out whether the screening effect works for other metal electrodes in the Co-doped systems, we changed Au to Ag, Cu and Al for the electrodes, and calculated the corresponding transmission spectra (not shown). Surprisingly, the screening effect still exists for all of them. When we changed the electrode–molecule distance, the transition of the transmission from spin-polarized to spin-unpolarized states are all observed (for the Ag, Cu and Al cases). That is to say, the screening effect is not sensitive to the electrode material, but robust to it, suggesting great application potential.

For practical application, the stability of the system is crucial. A more detailed investigation on this is necessary, which requires systematic study. In this work, we focused on transport properties. Actually, the robustness of the screening effect to the electrodes material suggests that intrinsic features of the system have been revealed.

Conclusions

In summary, we investigated the transport properties of Co-, Fe-, Mn- and Ni-encapsulated B₄₀ fullerenes contacted with Au electrodes through first principles calculations. The Fe- and Mn-doped systems exhibit spin-polarized transmission, and the Ni-doped ones exhibit completely spin-unpolarized transmission. However, for the Co-encapsulated B₄₀ systems, the transmission of the hexagonal hole-doped case is highly spin-polarized, but that of the heptagonal hole-doped one is spin-unpolarized. It is found that this is caused by the electrodes screening the magnetism of the Co atom. Further analysis shows that the screening effect is robust to the electrode material. Our results would be very useful in the fabrication of future spintronic devices.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (NSFC11504178, NSFC11374162 and NSFC11247033), the Natural Science Foundation of Jiangsu Province (BK20150825 and TJ215009), the Scientific Research Foundation of Nanjing University of Posts and Telecommunications (NY214010), and the Fundamental Research Funds for the Central Universities (NS2014073).

References

1. H. W. Kroto, J. R. Heath, S. C. OBrien, R. F. Curl and R. E. Smalley, Nature, 1985, 318, 162–163.
2. Z. Q. Fan, K. Q. Chen, Q. Wan, B. S. Zou, W. H. Duan and Z. Shuai, Appl. Phys. Lett., 2008, 92, 263304.
3. F. P. OuYang, H. Xu and T. J. Fan, J. Appl. Phys., 2007, 102, 064501.
4. Y. D. Guo, X. H. Yan and Y. Xiao, Appl. Phys. Lett., 2011, 98, 163107.
5. M. Nakaya, S. Tsukamoto, Y. Kuwahara, M. Aono and T. Nakayama, Adv. Mater., 2010, 22, 1622–1625.
6. F. Moresco, G. Meyer, K. H. Rieder, H. Tang, A. Gourdon and C. Joachim, Phys. Rev. Lett., 2001, 86, 672–675.
7. K. Yoshida, I. Hamada, S. Sakata, A. Umeno, M. Tsukada and K. Hirakawa, Nano Lett., 2013, 13, 481–485.
8. J. Lv, Y. C. Wang, L. Zhu and Y. M. Ma, Nanoscale, 2014, 6, 11692.
9. H. J. Zhai, Y. F. Zhao, W. L. Li, Q. Chen, H. Bai, H. S. Hu, Z. A. Piazza, W. J. Tian, H. G. Lu and Y. B. Wu, et al., Nat. Chem., 2014, 6, 727–731.
10. H. Bai, Q. Chen, H. J. Zhai and S. D. Li, Angew. Chem., Int. Ed., 2015, 54, 941–945.
11. Z. Yang, Y. L. Ji, G. Q. Lan, L. C. Xu, X. G. Liu and B. S. Xu, Solid State Commun., 2015, 217, 38–42.
12. X. J. Zhang, M. Q. Long, K. Q. Chen, Z. Shuai, Q. Wan, B. S. Zou and Y. Zhang, Appl. Phys. Lett., 2009, 94, 73503.
13. T. Menke, D. Ray, J. Meiss, K. Leo and M. Riede, Appl. Phys. Lett., 2012, 100, 093304.
14. N. Xin, H. Huang, J. X. Zhang, Z. Dai and L. Gan, Angew. Chem., 2012, 124, 6267–6270.
15. M. B. Javan, N. Tajabor, M. Behdani and M. R. Rokn-Abadi, Phys. Rev. B: Condens. Matter Mater. Phys., 2010, 405, 4937–4942.
16. M. B. Javan and N. Tajabor, J. Magn. Magn. Mater., 2012, 324, 52–59.
17. M. B. Javan, N. Tajabor, M. R. Roknabadi and M. Behdani, Phys. E, 2011, 43, 1351–1359.
18. G. P. Li, R. F. Sabirianov, J. Lu, X. C. Zeng and W. N. Mei, J. Chem. Phys., 2008, 128, 074304.
19. L. Senapati, J. Schrier and K. B. Whaley, Nano Lett., 2004, 4, 2073–2078.
20. A. Alexandrov and N. Mott, J. Supercond., 1994, 7, 599–605.
21. J. R. Heath, S. C. O’brien, Q. Zhang, Y. Liu, R. F. Curl, F. K. Tittel and R. E. Smalley, J. Am. Chem. Soc., 1985, 107, 7779–7780.
22. C. Thilgen, Angew. Chem., Int. Ed., 2012, 51, 587–589.
23. Y. Kubozono, H. Maeda, Y. Takabayashi, K. Hiraoka, T. Nakai, S. Kashino, S. Emura, S. Ukita and T. Sogabe, J. Am. Chem. Soc., 1996, 118, 6998–6999.
24. M. Brandbyge, J. L. Mozos, P. Ordejón, J. Taylor and K. Stokbro, Phys. Rev. B: Condens. Matter Mater. Phys., 2002, 65, 165401.
25. J. Taylor, H. Guo and J. Wang, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 245407.
26. H. J. Monkhorst and J. D. Pack, Phys. Rev. B: Solid State, 1976, 13, 5188.
27. J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett., 1996, 77, 3865–3868.
28. H. Hao, X.-H. Zheng, L.-L. Song, R.-N. Wang and Z. Zeng, Phys. Rev. Lett., 2012, 108, 017202.
29. L. Ma, J. Zhao, J. Wang, B. Wang and G. Wang, Phys. Rev. B: Condens. Matter Mater. Phys., 2007, 75, 045312.
30. C. K. Yang, J. J. Zhao and J. P. Lu, Phys. Rev. Lett., 2003, 90, 257203.
31. A. Singh, T. Briere, V. Kumar and Y. Kawazoe, Phys. Rev. Lett., 2003, 91, 146802.
32. E. Durgun and S. Ciraci, Phys. Rev. B: Condens. Matter Mater. Phys., 2006, 74, 125404.
33. L. Z. Kong and J. R. Chelikowsky, Phys. Rev. B: Condens. Matter Mater. Phys., 2008, 77, 073401.
34. M. Gruber, F. Ibrahim, S. Boukari, H. Isshiki, L. Joly, M. Peter, M. Studniarek, V. Da Costa, H. Jabbar and V. Davesne, et al., Nat. Mater., 2015, 14, 981C984.
35. Q. Wang, Q. Sun, P. Jena and Y. Kawazoe, Phys. Rev. B: Condens. Matter Mater. Phys., 2009, 79, 115407.
36. Y.-D. Guo, X.-H. Yan and Y. Xiao, Appl. Phys. Express, 2014, 7, 075001.
37. Y.-D. Guo, X.-H. Yan and Y. Xiao, Appl. Phys. Lett., 2014, 104, 063103.

---