Appl. Phys.
A (2016) 122:366
DOI 10.1007/s00339-016-9891-3
Electron transport properties and spin modulation effect analysis
for C60, Au@C60 and Ag@C60
Fangyuan Wang1 • Guiqin Li1
Received: 15 September 2015 / Accepted: 19 November 2015 / Published online: 14 March 2016
Ó Springer-Verlag Berlin Heidelberg 2016
Abstract The electron transport properties of C60,
Au@C60 and Ag@C60 junctions between Au electrodes are
investigated with density functional theory and the first-
principle nonequilibrium Greens function. Density of states
and transmission possibility as functions of energy are
examined, and current–voltage characteristics of all these
models are calculated with and without spin correction.
With spin correction, we find a spin-current difference in
Au@C60 as large as an order of magnitude. It may be used
to obtain high-purity arbitrary spin current. By comparing
the calculation results with and without spin correction, we
attribute this spin difference to the spin resonance between
transmission electrons and the valence electron of endo-
hedral Au atom.
1 Introduction
The study of molecular electronics is an important field for
the future of electronic industry and has attracted more and
more interests all over the world in the past decades [1–6].
With their unique characteristics from the classical elec-
tronic components such as molecular rectification, negative
differential resistance and molecular switching, molecular
electronics gradually show more broad prospects for elec-
tronic circuit engineering and electronic manufacturing
technology through many experimental investigations and
theoretical studies [7–11].
& Guiqin Li
ligqin@mail.tsinghua.edu.cn
1
Department of Physics, Tsinghua University, Beijing 100084,
People’s Republic of China
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Organic electronic devices are a focus in the field of
molecular electronics due to their low-dimensional prop-
erties and assembling characteristics. Zero-dimensional
fullerene (C60) [12–16], one-dimensional carbon atomic
wire [17–23], carbon nanotube (CNT) [24, 25] or graphene
nanoribbon (GNR) [26], and two-dimensional graphene
[27] are significant low-dimensional organic electronic
devices and have been systematically studied in theory and
experiment. Also the construction of hybrid structures with
same dimensional or even different dimensional organic
materials has attracted intense interests recently in order to
extend the functionality of carbon nanostructures [28].
Another way to extend the possible performance of
organic devices is combining with atoms, molecules,
chemical groups or other inorganic structures to modulate
the organic devices. As one of the most promising candi-
dates for practical application in all kinds of organic
electronic devices, fullerene C60 molecule has been inten-
sively investigated as a core component of molecular
devices [29–33]. Among all kinds of fullerene-related
materials, endohedral metallofullerenes with unique elec-
trical and magnetic properties which are mainly caused by
the interaction between the metal core and the carbon cage,
have received significant attention [34–37].
The usual endohedral metal atoms chosen in the previ-
ous references are alkali metals, alkaline-earth metals or
rare earth elements. As we choose gold lead to connect our
molecule junction model, we believe Au atom may be a
better choice to modulate transfer current since it contains
the same electron energy level structure with gold leads.
We perform a first-principles theoretical investigation
on the electron transport properties of C60, Au@C60 and
Ag@C60. The calculation method is using the program
smeagol [38–40], which combines density functional the-
ory (DFT) with first-principles nonequilibrium Greens
123
366 Page 2 of 9
function (NEGF). Before transmission calculation, conju-
gate gradients (CG) optimization is performed to obtain
stable structure for each model. We will mainly focus on
the different electronic transport properties of these three
materials and discuss the causes by analysing their electron
distribution and transmission possibility at different biases.
Also the importance of introducing spin ?1/2 to the elec-
tron of endohedral Au atom will be shown in this paper,
and with spin correction we will obtain a controllable spin
electronic filter which may be very important for the future
of quantum computation.
The paper is organized as follows. Section 2 gives a
brief description of the theoretical procedure and compu-
tational models. For the centre endohedral metal atom
model, Sect. 3 presents the normal calculation results and
Sect. 4 gives the results when spin correction is considered.
Section 5 confirms the transmission mechanism by ana-
lysing a Au@C60 model in which the endohedral Au atom
moves to the side of carbon cage. At last, Sect. 6 draws
conclusions and discusses possible future directions.
2 Computational models and methods
To study the transport of endohedral metallofullerenes, we
create an ideal model in quantum electron circuit: A C60
molecule is connected to two quasi-one-dimensional Au
nanowires with S atoms. The nanowire has periodic bulk
gold structure and extends infinitely along (111) direction.
Every gold bulk consists of two 3-atom layers and one
7-atom layer. Two S atoms here are used to raising the
energy of tunnelling electrons, so that the spin state of the
endohedral Au atom can be protected from the spin inter-
action of tunnelling spin electrons. Each S atom is
anchored to the centre of a 3-atom Au layer and connected
to the carbon cage through an aligned C atom.
The molecular junction in our model is divided into
three parts: left electrode, right electrode and central scat-
tering region. The left and right electrodes are calculated as
periodic gold bulks first, and then we take bulks’ Hamil-
tonian and self-energy to calculate the central scattering
part. The scattering region contains one gold bulk at both
side of the C60 molecule as illustrated in Fig. 1. Endohedral
metal atom (Au or Ag) is set in the centre of the carbon
cage. For simplicity, the two S atoms and their nearest C
atoms are located on z axis as well as the central Au atom
in each electrode bulk and the endohedral metal atom in
C60.
During CG optimization, the maximum number of CG
steps is 30 and the maximum atomic displacement for each
step is 0.2 Bohr. Force tolerance is set to 0.04 eV/Å, stress
tolerance is set to 1.0 GPa and unit cell relaxation is
123
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
W. Fangyuan, L. Guiqin
Fig. 1 Scattering region of the molecular junctions. C60 cage is
connected to the left and right Au electrodes with S atoms. Grey balls
in the centre represent C atoms, blue balls and yellow balls on each
side represent Au atoms and S atoms, respectively. The green ball in
C60 cage represents endohedral metal atom and can be just set as
empty
permitted during optimization. Note that here we only
perform CG optimization without coordinates constraints
to C60 model, but fix the relative position between endo-
hedral metal atom and C60 cage for Au@C60 and Ag@C60
in Sects. 3 and 4. This is just because we want to analysis
the modulation effect of endohedral metal atom at first, and
the centre position of carbon cage is the best choice to
eliminate the structure changes of carbon cage. We will
discuss a Au@C60 model with endohedral Au atom relaxed
in Sect. 5 and analyse the effect of orbital hybridization to
spin current.
After CG process, the distance between two Au atoms in
electrode bulk is 2.884 Å which is typically the Au–Au
bond length in gold metal lattice. We also obtain the dis-
tance 2.526 Å between S atom and its nearest three Au
atoms, and distance 1.970 Å between S atom and its
nearest C atom.
Electrical transport calculations are carried out by pro-
gram smeagol with single-f basis plus polarization orbitals
and standard local density approximation (LDA) [41]. In
gold electrode bulk calculation, The k-points were sampled
by a 1  1  100 uniform grid to ensure good periodicity
along z direction, and tolerance of 1  106 is set to get
accurate results for Hamiltonian and self-energy of semi-
infinite Au lead. While during the scattering region cal-
culation, the integral mixing parameter is set to 0.01 and
the iteration tolerance is set to 5  105 .
The total current I along z direction is related to electron
distribution and transmission coefficient, which is expres-
sed as the integral over the injection energy of the tun-
nelling electron E, as follows [42]
Z
e
I¼ dE TðE; VÞðf ðE  lL Þ  f ðE  lR ÞÞ ð1Þ
h
where h is the Planck constant, e is the elementary charge,
lL=R ¼ EF  eV/2 is the electrochemical potentials, and
f(x) is the Fermi function for a given temperature T.
T(E, V) is the transmission coefficient which is usually bias
dependent as well as energy dependent.
Electron transport properties and spin modulation effect analysis for C60, Au@C60 and Ag@C60
3 Computational result without spin correction
In order to study the basic characteristics of these three
models and illustrate the importance of spin correction, we
perform normal transport calculations without spin cor-
rection at first. According to the computational result
without spin correction, the current as a function of the bias
is shown in Fig. 2. Current curves of C60, Au@C60 and
Ag@C60 rise together with a similar trend. The modulation
of endohedral metal atom makes the charging effect show
different current rising point from C60 to Ag@C60. Espe-
cially for Ag@C60, the green curve has an obvious turning
point at about 0.5 V. The current of Ag@C60 increases very
slowly at first when bias is lower than 0.5 V, after that it
increases rapidly and overtakes the other two lines soon. In
the following paragraphs, we will discuss these behaviours
by analysing their electron distribution and transmission
possibility.
Figure 3 shows the density of states (DOS) distributions
of (a) C60, (b) Au@C60 and (c) Ag@C60. The short dot
lines, short dash lines and solid lines represent electron
distribution at bias of 0, 0.36 and 0.9 V, respectively.
When bias increases, there is a significant peak clustering
effect in the energy regions of [-6.9, -6.5] and [-6.4,
-5.9] eV at both sides of the Fermi energy. DOS peaks in
those two regions tend to gather as bias rising and finally
connect to each other to form cluster. At the maximum
voltage value 0.9 V (solid line), we can see from (a) to
(c) the two clustering regions shift to the left on the energy
axes as a whole and always maintain a 0.4 eV band gap
between them. Note that the higher solid line peak
-6.06 eV in Fig. 3c is constituted by two overlapping
Fig. 2 I–V curves of C60, Au@C60 and Ag@C60 in the bias range
from 0.0 to 0.9 V. Current flows from the left electrode to the right
electrode. Respectively, the red line with square symbols represents
the current of C60, blue line with diamond symbols represents
Au@C60, and green line with circle symbols represents Ag@C60
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 3 of 9 366
single peaks, so all the four single peaks in the right
clustering region [-6.4, -5.9] eV are included in the bias
window at maximum bias 0.9 V in Fig. 3c. Since there are
more transmission channels containing in bias window for
Ag@C60, the current of Ag@C60 at final bias 0.9 V is
higher than the other two.
Figure 4 shows the transmission coefficient of these
three systems for different energy as well as different bias.
The purple dash lines indicate the range of bias window
under different bias. As bias rising and bias window
broadening, the transmission peaks in the vicinity of Fermi
energy at 0.0 V bias drift to the left side through the energy
range of left DOS clustering region [-6.9, -6.5] eV .
When bias is higher than 0.6 V, there is also a transmission
plateau rising in the right part of bias window, of which the
energy range just overlaps with the right DOS clustering
region [-6.4, -5.9] eV. From Eq. (1), we know the
behaviour of total current is determined by the integral of
Fermi electron distribution and transmission probability in
these two regions.
In the very beginning when bias is 0.0 V, transmission
peaks at Fermi energy are separated from the left and right
DOS region. It means no electron of these three scattering
systems is distributed near Fermi energy at 0.0 V bias and
the transmission channels are formed by unoccupied
molecule orbitals. Because of the DOS peak clustering
effect and charging effect, the energy of transmission
channels decreases when bias increases. Thus we find the
(a)
(b)
(c)
Fig. 3 DOS of C60, Au@C60 and Ag@C60 in the energy range from
-7.3 to -5.4 eV. The purple dash lines represent the bias window at
0.9 V. The short dot lines, short dash lines and solid lines represent
electron distribution at bias of 0, 0.36 and 0.9 V, respectively. The
colours a red, b blue and c green represent the DOS of C60, Au@C60
and Ag@C60, respectively. The Fermi energy is -6.464 eV for C60,
-6.451 eV for Au@C60 and -6.444 eV for Ag@C60
123
366 Page 4 of 9
(a)
(b)
(c)
Fig. 4 Transmission coefficient of C60, Au@C60 and Ag@C60 in the
energy range from -7.3 to -5.4 eV and bias range from 0.0 to 0.9 V.
The purple dash lines represent the bias window and the colour a red,
b blue and c green represent the transmission coefficient of C60,
Au@C60 and Ag@C60, respectively
transmission peaks near Fermi energy at 0.0 V bias in Fig.
4, drift to the left side when voltage rises. Since there is a
overall energy shift of the DOS peaks from Fig. 3a, c,
current rising caused by charging effect appears a delay
phenomenon from C60 to Ag@C60 in Fig. 2 when the
energy of transmission peaks drift through the left DOS
region. When bias is higher than 0.6 V and the right
transmission plateau rises, those transmission peaks in the
right part of bias window play a more important role in the
total current integral. Because the energy range of trans-
mission plateau in Fig. 4 and the energy range of right
DOS clustering region in Fig. 3 match each other very
well, transmission peaks in the right energy region should
be formed by occupied molecule orbitals. Since Ag@C60
contains a higher plateau in Fig. 4, its total current will
123
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
W. Fangyuan, L. Guiqin
gain a rapid growth when bias increases over 0.6 V. This is
consistent with the previous conclusion that Ag@C60
should contain more transmission channels in bias window
at final bias. Since the plateau of Au@C60 is lower than the
one of C60, the final current of Au@C60 should be lower
than the current of C60 at final bias 0.9 V as we see in
Fig. 2.
Further studies on projected density of states (PDOS)
show that the electron distribution of endohedral Ag atom
in Ag@C60 system forms a peak -5.61 eV at bias 0.0 V
(short dot line) in Fig. 3c and drifts to the right side as bias
rising. So the energy peak of Ag@C60 is always out of bias
window and does not participate in the transmission pro-
cess directly. Similarly in Fig. 3b, the electron distribution
of endohedral Au atom in Au@C60 system forms a peak
-6.23 eV at bias 0.0 V (shot dot line) and a peak
-5.83 eV at bias 0.9 V (solid line). Even this peak of
endohedral Au atom is close to the Fermi energy at 0.0 V
bias, the right drift keeps it always out of bias window
during bias rising from 0.0 V to 0.9 V. The above analysis
of PDOS illustrates that the endohedral atoms in this two
systems do not participate in the transmission process
directly, and the peak drift of their electron distribution
means the potential energy of their valence electron
increases when the electric field caused by the applied
voltage increases.
4 Computational result with spin correction
When we take spin into consideration, not surprisingly, the
spin computational result of C60 exactly coincides with the
previous one since it is a symmetrical structure. For the
other two models, we set the endohedral metal atom as spin
?1/2 and perform spin computation. The computational
result of Ag@C60 is very close to the one without spin
effect that we have obtained in the previous section, and
the relative deviation between spin-up and spin-down
current is only about 103 which is very difficult to observe
in the experiment. This is because the endohedral Ag atom
does not participate in the transmission process and its spin
direction cannot efficiently influence transmission elec-
trons. At last, the computational result of Au@C60 illus-
trates a significant current divergence which is shown in
Fig. 5.
In Fig. 5, spin-up current (orange line with up-triangle
symbols) is greater than spin-down current (cyan line with
down-triangle symbols) and has a significant negative
differential resistance (NDR) effect which we will discuss
later in the end of this section. While the behaviour of spin-
down current is quite close to the current calculated with-
out spin effect (blue dash line). In particular, when voltage
Electron transport properties and spin modulation effect analysis for C60, Au@C60 and Ag@C60 Page 5 of 9 366

(a)

(b)

Fig. 6 DOS of Au@C60 in the energy range from -7.3 to -5.4 eV.
The purple dash lines represent the bias window at 0.9 V. The short
dot lines, short dash lines and solid lines represent electron
distribution at bias of 0, 0.36 and 0.9 V, respectively. The colours
Fig. 5 I–V curves of Au@C60 in the bias range from 0.0 to 0.9 V a orange and b cyan represent the DOS of spin-up and spin-down
when spin effect is considered. Current flows from the left electrode electrons. The Fermi energy is -6.454 eV which is very close to
to the right electrode. Respectively, the dark yellow line with circle -6.451 eV the one we obtained before without spin effect
symbols represents the total spin current (spin-up and spin-down) of
Au@C60, orange line with up-triangle symbols represents the spin-up
part, and cyan line with down-triangle symbols represents the spin-
down part, respectively. The blue dash line with diamond symbols
(a)
represents the result of Au@C60 without spin effect considered, which
we have obtained before and placed here as a reference

is below 0.3 V, spin-up electrons contribute most of the

total current and differ spin-down current as large as an
order of magnitude. It means that the spin ?1/2 of endo-
hedral Au atom increases the spin-up current but affects
little to the spin-down current, and thus the total current
should be spin characterized and larger than the one
without spin correction. More specifically, this spin-current
divergence is caused by the spin interaction including spin- (b)
orbit coupling and exchange correlation. Note that we use
more dots in Fig. 5 to show the current peak of spin-up
current clearly, and current without spin effect (blue dash
line) now shows NDR between [0.1, 0.3] eV which does
not display in Fig. 2. This behaviour is a nature phe-
nomenon of C60 which has been reported by previous work
in C60 [43] as well as its hybrid structures [44] and should
also display for C60 and Ag@C60 if we use more dots in
Fig. 5.
Figure 6 shows the DOS distributions of spin-up and Fig. 7 Transmission coefficient of spin-up and spin-down electrons
spin-down electrons in Au@C60 system. Compared to Fig. in Au@C60 system in the energy range from -7.3 to -5.4 eV and
bias range from 0.0 to 0.9 V. The purple dash lines represent the bias
3b, the electron distributions of spin-up and spin-down are window and the colours a orange and b cyan represent the
quite similar to the one ignoring electron spin effect, and transmission coefficient of spin-up and spin-down electrons,
only the electron of endohedral Au atom that we set its spin respectively
as ?1/2 reduces its energy. PDOS analysis shows the spin
?1/2 electron of endohedral Au atom forms a peak between Fig. 7a and b is the transmission peak on the right
-6.29 eV at bias 0.0 V (shot dot line) and a peak -5.96 eV part of bias window is which formed by the spin ?1/2
at bias 0.9 V (solid line) in Fig. 6a which is the significant electron of endohedral Au atom. Even the right drift keeps
difference between Fig. 6a, and b. this peak always out of bias window, actually it is quite
Figure 7 shows the transmission coefficient of spin-up close to the left electrochemical potentials lL ¼ EF þ
and spin-down electrons. The most significant difference eV=2 (the right purple dash line) that makes it possible to

123

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

366 Page 6 of 9 W. Fangyuan, L. Guiqin

matches a nonzero Fermi distribution. In particular, when

bias is lower than 0.15 V, transmission coefficient in
Fig. 7a has a lager value in bias window while the one in
Fig. 7b reduces to zero quickly in the centre of bias win-
dow. So when we integrate transmission coefficient and
electron distribution to calculate spin current, the spin-up
current increases quite rapidly than spin-down current at
Fig. 8 Au@C60 junctions after optimization. C60 cage is connected to
bias 0.15 V as shown in Fig. 5. the left and right Au electrodes with S atoms. Grey balls in the centre
We find a transmission peak on the right border of bias represent C atoms, blue balls and yellow balls on each side represent
window in Fig. 7a, and its energy matches the DOS peak of Au atoms and S atoms, respectively. The green ball in the bottom side
endohedral Au atom in Fig. 6a at bias 0, 0.36 and 0.9 V. It of C60 cage represents endohedral Au atom
means when lead electrons match the energy as well as the
spin of endohedral Au atom, they will have a great possi-
bility to transport from one electrode to another. So spin
resonance transmission mechanism is the main reason why
the spin-up current of Au@C60 is quite larger than the spin-
down current. Since the energy of Ag atom is always out of
bias window, its spin resonance does not contribute to the
integral of total current. Thus the difference between
computation result of Ag@C60 with and without spin
correction should be quite small.
In Fig. 7a, the spin resonance transmission peak disap-
pears at bias 0.3 V and then appears out of the bias window at
0.45 V. In Fig. 6a, we can see the peak -6.29 eV of endo-
hedral Au atom is on the left of other peaks in right DOS
clustering region at bias 0 V (the dot line). At the final bias
0.9 V, its peak -5.96 eV is on the right of other peaks (the
solid line). It suggests that other electrons of scattering
Fig. 9 I–V curves of optimized Au@C60 in the bias range from 0.0 to
region may interfere with the spin resonance if they are on the 0.9 V. Current flows from the left electrode to the right electrode. the
same energy level. Thus the spin-up current will decrease dark yellow line with circle symbols represents the total spin current
rapidly, and negative difference resistance (NDR) phe- (spin-up and spin-down) of Au@C60, orange line with up-triangle
nomenon should appear in Fig. 5 (the orange line). In par- symbols represents the spin-up part, and cyan line with down-triangle
symbols represents the spin-down part, respectively. The blue dash
ticular, when detuning happens at bias 0.3 V, the spin line with diamond symbols represents the result of Au@C60 without
resonance transmission channel disappears in Fig. 7a and spin effect considered
spin-up current equals to spin-down current in Fig. 5.
Figure 9 shows the current–voltage dependence for each
spin current, we can see the spin divergence is quite small
5 Spin analysis for Au@C60 with endohedral Au
than the one in Fig. 5. The maximum difference between
atom relaxed
spin-up and spin-down current is only 37 % of the former
one at bias 0.18 eV, and this is also the only deviation
To confirm this spin resonance transmission mechanism,
between the total current with and without spin correction.
we perform conjugate gradients (CG) optimization to
From Figs. 10 and 11, we can see the DOS distributions
Au@C60 again, but relax the constraints of endohedral Au
and transmission coefficients of spin-up and spin-down
atom. After CG process, the endohedral Au atom moves to
current as well as current without spin correction. There are
the bottom side of the carbon cage (Fig. 8), the distances
only a few small differences between spin-up and spin-
between it and its nearest two carbon atoms are 1.124 and
down electrons; for example, in Fig. 11a there is an obvi-
1.409 Å, respectively. The orbital hybridization between
ous peak falling at bias 0.9 V and energy -6.8 eV which is
Au atom and carbon cage makes the valence orbital of
different in Fig. 11b. In some extent, the curves of spin-up
endohedral Au atom no longer being occupied by a single
electron are closer to the curves without spin correction.
electron, but by two electrons with opposite spin directions.
This is because the hybridized orbital of spin-up electron is
Thus the spin resonance transmission mechanism will hold
fitted from atom orbitals of Au atom, but the hybridized
for both spin electrons, and the spin divergence will
orbital of spin-down electron is fitted from atom orbitals of
decrease to a rather small level (Fig. 9).

123

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Electron transport properties and spin modulation effect analysis for C60, Au@C60 and Ag@C60
(a)
(b)
(c)
Fig. 10 DOS of optimized Au@C60 in the energy range from -7.3 to
-5.4 eV. The purple dash lines represent the bias window at 0.9 V. The
short dot lines, short dash lines and solid lines represent electron
distribution at bias of 0, 0.36 and 0.9 V, respectively. Colours a orange
and b cyan represent the DOS of spin-up and spin-down electrons.
Colour c blue represents the DOS distribution without spin correction
C atom. So the final calculation results of spin-up electron
will closer to those without spin correction, as both of them
are using the same orbital fitting way.
Since the valence orbital of endohedral Au atom con-
tains both spin electrons, the spin resonance transmission
mechanism now performs for both spin-up and spin-down
currents. Then the total current, of course, is larger than the
one when Au atom is in the centre of carbon cage. It is
worth noting that the spin-up current peak values in Fig. 5
(2.399 lA) and Fig. 9 (2.281 lA) are very close, indicating
that these two current peaks as well as the NDR phenom-
ena after them are formed by the same mechanism.
PDOS analysis shows that the valence electron of
endohedral Au atom does not form a single high peak any
more, but forms several low hybridization peaks. Specifi-
cally, The centre endohedral Au atom forms only a single
PDOS peak -5.95 eV at final bias 0.9 V with the same
height as the one in Fig. 6a, but the bottom endohedral Au
atom forms PDOS peaks -7.18, -6.29, -6.03 and
-5.50 eV, and their heights are quite lower than those in
Fig. 10a. So these four peaks in Fig. 10a are all hybridized
peaks, and they are formed by endohedral Au atom and
carbon cage atoms together. These hybridized peaks do not
raise their energy any more either when electric field
caused by applied voltage increases. So in Fig. 11a, the
transmission peak -6.35 eV at bias 0.3 V remains standing
where it has already disappeared in Fig. 7a. Because of this
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 7 of 9 366
(a)
(b)
(c)
Fig. 11 Transmission coefficient of optimized Au@C60 system in the
energy ranges from -7.3 to -5.4 eV and bias range from 0.0 to
0.9 V. The purple dash lines represent the bias window. Colours
a orange and b cyan represent the transmission coefficient of spin-up
and spin-down electrons. Colour c blue represents the transmission
coefficient without spin correction
peak standing, the NDR phenomenon does not happen until
0.3 V in Fig. 9, while the same phenomenon has already
happened at 0.17 V in Fig. 5.
Note that the peak -6.03 eV in Fig. 10 and the peak
-5.95 eV in Fig. 6 are all formed by valance orbital of
endohedral Au atom. Single peak -5.95 eV in Fig. 6 is
formed by endohedral Au atom only, which means the
valance orbital is occupied by a spin-up electron that we
have specified in the beginning. The PDOS peak -6.03 eV
of endohedral Au atom is lower than the one in Fig. 10,
which indicates there are other electrons participate in the
formation of DOS peak -6.03 eV, so the valance orbital of
optimized model is not occupied only by electron of
endohedral Au atom, but also by electrons of hybridized C
atoms.
123
366 Page 8 of 9
6 Conclusions
We have investigated electronic and transport properties of
C60, Au@C60 and Ag@C60 by combining DFT and NEGF.
DOS analysis illustrates a peak clustering effect, which
means the bond electrons in the system around Fermi
energy will close their energy to each other as bias rising.
Because of this kind of bound electron redistribution, the
transmission peaks will also be redistributed as the mole-
cule orbitals change their energy eigenvalues. This clus-
tering effect together with charging effect determines the
shape of I–V curve, and the energy shift due to the endo-
hedral Au or Ag atom leads to a delay of current rising
from C60 to Ag@C60.
Further PDOS analyses explained the different com-
puting results of Au@C60 in Sects. 3 and 4. Introducing
spin ?1/2 to endohedral Au atom results a lower electron
energy and makes its transmission peak close enough to the
left electrochemical potentials. Thus the transmission peak
of endohedral Au atom which does not transport electrons
in Sect. 3 actually does play an important part in the
integral of total current, especially for low-bias cases which
is shown in Sect. 4.
After conjugate gradients optimization, endohedral Au
atom moves to the side of carbon cage. The hybridization
between endohedral Au atom and carbon cage makes the
spin resonance transmission mechanism perform for both
spin-up and spin-down currents. It will increase the spin-
down current as well as the total current, but the difference
between spin-up and spin-down currents shall decrease to a
lower level either.
The large difference between spin-up and spin-down
currents in Au@C60 at low bias is a promising effect for
application. If we realize the spin direction of transmission
electrons, ‘up’ and ‘down’ , is actually determined by the
spin ?1/2 of the endohedral Au atom that we have set in the
beginning of Sect. 4. This is because different spin direction
of the endohedral Au atom does not change the eigenvalues
of Hamiltonian, and the eigenstates just simply switch their
spin to a corresponding direction. Thus the calculation result
for ‘up’ and ‘down’ will not change at all. Note that even
though different spin settings relate to the same calculation
result, the real spin direction is quite important for the
quantum information carrying in it. By controlling the spin of
endohedral Au atom with laser directly or through electron
spin resonance (ESR), we can easily obtain and modulate
arbitrary spin current with a high purity. By measuring the
spin direction of spin current, we are able to read the spin
information in the endohedral Au atom.
Actually when we use laser to manipulate spin direction,
the endohedral Au atom may oscillate in the carbon cage.
In this case, the spin resonance transmission mechanism
123
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
W. Fangyuan, L. Guiqin
shall still work, but the effect of hybridization will change
depending on time. DFT with NEGF cannot deal with a
moving atom; however, this further study will rely on more
advanced theory.
Acknowledgments This work is supported by the National Basic
Research Program of China (Grants No. 2011CB921602) and the
NSFC (Grants No. 20121318158).
References
1. J. Chen, M.A. Reed, A.M. Rawlett, J.M. Tour, Large on–off
ratios and negative differential resistance in a molecular elec-
tronic device. Science 286, 1550 (1999)
2. J.M. Tour, Molecular electronics. Synthesis and testing of com-
ponents. Acc. Chem. Res. 33, 791 (2000)
3. J. Park, A.N. Pasupathy, J.I. Goldsmith, C. Chang, Y. Yaish, J.R.
Petta, M. Rinkoski, J.P. Sethna, H.D. Abruna, P.L. McEuen, D.C.
Ralph, Coulomb blockade and the Kondo effect in single-atom
transistors. Nature 417, 722 (2002)
4. A. Nitzan, M.A. Ratner, Electron transport in molecular wire
junctions. Science 300, 1384 (2003)
5. A. Salomon, D. Cahen, S. Lindsay, J. Tomfohr, V.B. Engelkes,
C.D. Frisbie, Comparison of electronic transport measurements
on organic molecules. Adv. Mater. 15, 1881 (2003)
6. R.A. Wassel, C.B. Gorman, Establishing the molecular basis for
molecular electronics. Angew. Chem. Int. Ed. 43, 5120 (2004)
7. K. Stokbro, J. Taylor, M. Brandbyge, J. Mozos, P. Ordejón,
Theoretical study of the nonlinear conductance of di-thiol ben-
zene coupled to Au(111) surfaces via thiol and thiolate bonds.
Comput. Mater. Sci. 27, 151 (2003)
8. S. Ke, H.U. Baranger, W. Yang, Molecular conductance: chem-
ical trends of anchoring groups. J. Am. Chem. Soc. 126, 15897
(2004)
9. M. Galperin, M.A. Ratner, A. Nitzan, Hysteresis, switching, and
negative differential resistance in molecular junctions: a polaron
model. Nano Lett. 5, 125 (2005)
10. S. Yeganeh, M. Galperin, M.A. Ratner, Switching in molecular
transport junctions: polarization response. J. Am. Chem. Soc.
129, 13313 (2007)
11. M. Smeu, R.A. Wolkow, H. Guo, Conduction pathway of p-
stacked ethylbenzene molecular wires on Si(100). J. Am. Chem.
Soc. 131, 11019 (2009)
12. C. Kaun, T. Seideman, Current-driven oscillations and time-de-
pendent transport in nanojunctions. Phys. Rev. Lett. 94, 226801
(2005)
13. E.I. Altman, R.J. Colton, The interaction of C60 with noble metal
surfaces. Surf. Sci. 295, 13 (1993)
14. Q. Zeng, X. Chu, M. Yang, D.Y. Wu, Spinorbit coupling effect
on AuC60 interaction: a density functional theory study. Chem.
Phys. 395, 82 (2012)
15. X. Zhang, M. Long, K. Chen, Z. Shuai, Q. Wang, B.S. Zou, Y.
Zhang, Electronic transport properties in doped C60 molecular
devices. Appl. Phys. Lett. 94, 073503 (2009)
16. H. Xu, S. Jia, L. Chen, Negative differential resistance behavior
in doped C82 molecular devices. Appl. Phys. A 19, 299 (2012)
17. C. Jin, H. Lan, L. Peng, K. Suenaga, S. Iijima, Deriving carbon
atomic chains from graphene. Phys. Rev. Lett. 102, 205501
(2009)
18. M. Brandbyge, J. Mozos, P. Ordejón, J. Taylor, K. Stokbro,
Density-functional method for nonequilibrium electron transport.
Phys. Rev. B 65, 165401 (2002)
Electron transport properties and spin modulation effect analysis for C60, Au@C60 and Ag@C60
19. B. Song, S. Sanvito, H. Fang, Anomalous IV curve for mono-
atomic carbon chains. New J. Phys. 12, 103017 (2010)
20. E. Durgun, R.T. Senger, H. Mehrez, H. Sevincli, S. Ciraci, Size-
dependent alternation of magnetoresistive properties in atomic
chains. J. Chem. Phys. 125, 121102 (2006)
21. R. Rivelino, R.B. dos Santos, F. de Brito Mota, Conformational
effects on structure, electron states, and raman scattering prop-
erties of linear carbon chains terminated by graphene-like pieces.
J. Phys. Chem. C 114, 16367 (2010)
22. B. Wang, Y. Yu, L. Zhang, Y. Wei, J. Wang, Oscillation of
dynamic conductance of Al–Cn–Al structures: nonequilibrium
Greens function and density functional theory study. Phys. Rev. B
79, 155117 (2009)
23. K. Walczak, G. Platero, Spin-dependent transport through mag-
netic nanojunctions. Cent. Eur. J. Phys. 4, 30 (2006)
24. Z. Zanolli, J. Charlier, Spin transport in carbon nanotubes with
magnetic vacancy-defects. Phys. Rev. B 81, 165406 (2010)
25. G.B. Abadir, K. Walus, D.L. Pulfrey, Basis-set choice for DFT/
NEGF simulations of carbon nanotubes. Appl. Phys. A 8, 1
(2009)
26. L. Pisani, J.A. Chan, B. Montanari, N.M. Harrison, Electronic
structure and magnetic properties of graphitic ribbons. Phys. Rev.
B 75, 064418 (2007)
27. P. Chang, H. Liu, B.K. Nikolić, First-principles versus semi-
empirical modeling of global and local electronic transport
properties of graphene nanopore-based sensors for DNA
sequencing. Appl. Phys. A 13, 847 (2014)
28. R.N. Wang, X.H. Zheng, L.L. Song, Z. Zeng, First-principles
study on electron transport of carbon dumbbells C60–Cn–C60.
J. Chem. Phys. 135, 044703 (2011)
29. Z. Fan, K. Chen, Q. Wan, B.S. Zou, W. Duan, Z. Shuai, Theo-
retical investigation of the negative differential resistance in
squashed C60 molecular device. Appl. Phys. Lett. 92, 263304
(2008)
30. X. Zhang, M. Long, K. Chen, Z. Shuai, Q. Wan, B.S. Zou, Y.
Zhang, Electronic transport properties in doped C60 molecular
devices. Appl. Phys. Lett. 94, 073503 (2009)
31. X. Zhong, R. Pandey, A.R. Rocha, S.P. Karna, Can Single-atom
change affect electron transport properties of molecular nanos-
tructures such as C60 fullerene? J. Phys. Chem. Lett. 1, 1584
(2010)
32. X.H. Zheng, X.L. Wang, Z.X. Dai, Z. Zeng, Tuning the transport
properties of a (C60)2 bridge with electron and hole dopings.
J. Chem. Phys. 134, 044708 (2011)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 9 of 9 366
33. S. Yuan, S. Wang, Q. Mei, Q. Ling, L. Wang, W. Huang, Effects
of electrodes and nitrogen-atom locations on electron transport in
C59N molecular junctions: a first-principles study. J. Phys.
Chem. C 118, 617 (2014)
34. N. Okamura, K. Yoshida, S. Sakata, K. Hirakawa, Electron
transport in endohedral metallofullerene Ce@C82 single-mole-
cule transistors. Appl. Phys. Lett. 106, 043108 (2015)
35. M. Chattopadhyaya, M.M. Alam, S. Sen, S. Chakrabarti, Elec-
trostatic spin crossover and concomitant electrically operated
spin switch action in a ti-based endohedral metallofullerene
polymer. Phys. Rev. Lett. 109, 257204 (2012)
36. S. Kaneko, L. Wang, G.F. Luo, J. Lu, S. Nagase, S. Sato, M.
Yamada, Z. Slanina, T. Akasaka, M. Kiguchi, Electron transport
through single endohedral Ce@C82 metallofullerenes. Phys. Rev.
B 86, 155406 (2012)
37. L. Senapati, J. Schrier, K.B. Whaley, Electronic transport,
structure, and energetics of endohedral Gd@C82 metallo-
fullerenes. Nano Lett. 4, 2073 (2004)
38. A.R. Rocha, V.M. Garcia-Suarez, S. Bailey, C. Lambert, J. Fer-
rer, S. Sanvito, Spin and molecular electronics in atomically
generated orbital landscapes. Phys. Rev. B 73, 085414 (2006)
39. I. Rungger, S. Sanvito, Algorithm for the construction of self-
energies for electronic transport calculations based on singularity
elimination and singular value decomposition. Phys. Rev. B 78,
035407 (2008)
40. T. Ohto, I. Rungger, K. Yamashita, H. Nakamura, S. Sanvito, Ab
initio theory for current-induced molecular switching: melamine
on Cu(001). Phys. Rev. B 87, 205439 (2013)
41. J.P. Perdew, A. Zunger, Self-interaction correction to density-
functional approximations for many-electron systems. Phys. Rev.
B 23, 5048 (1981)
42. Y. Meir, N.S. Wingreen, Landauer formula for the current
through an interacting electron region. Phys. Rev. Lett. 68, 2512
(1992)
43. A.V. Danilov, S.E. Kubatkin, S.G. Kafanov, T. Bjørnholm,
Strong electronic coupling between single C60 molecules and
gold electrodes prepared by quench condensation at 4 K. A single
molecule three terminal device study. Faraday Discuss. 131, 337
(2006)
44. X. Zheng, W. Lu, T.A. Abtew, V. Meunier, J. Bernholc, Negative
differential resistance in C60-based electronic devices. ACS
Nano 4, 7205 (2010)
123
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:

1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at

onlineservice@springernature.com