Superlattices and Microstructures 98 (2016) 458e472

Contents lists available at ScienceDirect

Superlattices and Microstructures

journal homepage: www.elsevier.com/locate/superlattices

Monte Carlo simulation of magnetic properties of a nano-

graphene bilayer in a longitudinal magnetic field
Wei Wang*, Ruijia Liu, Dan Lv, Xiaohong Luo
School of Science, Shenyang University of Technology, Shenyang 110870, China

a r t i c l e i n f o a b s t r a c t

Article history: Monte Carlo simulation has been used to study the magnetic properties of a nano-
Received 30 July 2016 graphene bilayer which consists of the upper layer A with spin-3/2 and the bottom
Received in revised form 28 August 2016 layer B with spin-5/2. The effects of the single-ion anisotropy, the intralayer exchange
Accepted 28 August 2016
coupling and the longitudinal magnetic field on the magnetization, the susceptibility, the
Available online 30 August 2016
blocking temperature and hysteresis loops of the mixed-spin nano-graphene bilayer sys-
tem have been examined detailedly. In particular, the variations of the blocking temper-
Keywords:
ature with different intralayer exchange couplings, single-ion anisotropies, and the
Nano-graphene bilayer
Magnetization
longitudinal magnetic field are obtained for the present system. Many multiple hysteresis
Susceptibility loop behaviors have also been found, depending on the combinations of both the upper
Blocking temperature and bottom layer magnetizations in the longitudinal magnetic field. Through a compari-
Monte Carlo simulation son, our results obtained are according well with other theoretical researches and
experimental results.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Over one decade, graphene has become the focus of carbon-based materials due to its remarkable physical and chemical
properties. It sparkles with many performances, such as electrical energy storage, high mechanical stability, and low-
intensity. With the development of nanotechnology, it is likely to make graphene become one of the lightest and thinnest
two-dimensional materials, which results from its unique structure of benzene's six-membered rings. Furthermore, the
discovery of nano-graphene's original magnetic behaviors has broadened its potential prospect in many fields, such as
biosensor [1], solar cells [2], linear magneto resistance [3], environmental science [4], chemical engineering [5] ect.
On one hand, the exploration of graphene in experiment has received great interests. Graphene has been manufactured by
diverse methods at present, which advances the process in the preparation of nano-graphene. Zhong et al. applied the
method of high concentration of carbon source under low pressure and cool slowly to prepare double-layer graphene with
homogeneous layers [6]. Graphene has also been synthesized by the process of graphite's ultrasonic treatment in water or
sodium dodecyl benzene sulfonate. The results illustrate that the majority of graphene under five layers possess high elec-
trical conductivity [7]. Moreover, light transmittance of nano-graphene has been improved through the technique of
exfoliation-inserting layers again-inflation [8].
On the other hand, theoretical explanations can lay the foundation for experimental evidences. In recent years, a number
of theoretical models and methods have been devoted to exploring the physical properties of nano-graphene. The density-

* Corresponding author.
E-mail address: ww9803@126.com (W. Wang).

http://dx.doi.org/10.1016/j.spmi.2016.08.045
0749-6036/© 2016 Elsevier Ltd. All rights reserved.
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 459

functional theory is employed to study electronic and magnetic properties of cobalt-vacancy defect complexes in graphene
[9]. The results show that both the vacancycobalt separation and the sub-lattice of the vacancy relative to cobalt can
determine the magnetization induced on the cobalt atom. The mean-field theory has been applied to study the magnetic
properties of graphene nanostructures [10]. It is found that the magnetic field and the external in-plane electric play
important roles in the ground phase diagram. What is noticeable is that the edge-state spins arranged parallel to one another
in a zigzag edge express a very strong ferromagnetic interaction, which makes the nano-graphene become the excellent
ferromagnet with a higher Curie temperature [11]. A Boumali has studied the mean energy, the free energy and the specific
heat of graphene structure [12]. L. Xu et al. have probed into the topological phases in biased ferromagnetic bilayer graphene
with an external ferromagnetic exchange coupling under the magnetic field [13]. Besides, Peres et al. have analyzed the
variations of magnetic susceptibility of graphene with temperature [14]. The effect of nono-magnetic defects on the
magnetization in graphene has also been examined by V.V. Nelayev et al. with ab-initio simulation [15]. By Green's functional
method and density functional theory, X. Zhong et al. have studied transport properties and the stacking dependent elec-
tronic structure of bi-layer graphene nano-ribbons suspended between gold electrodes [16]. Using effective-field theory (EFT)
with correlations, the magnetic properties of a nano-graphene bilayer described by a transverse Ising model have been
investigated in an external magnetic field [17]. The results revealed that the exchange coupling, the single-ion anisotropy, the
transverse field and the external magnetic field have profound impacts on the magnetization, the susceptibility and the
blocking temperature of the nano-graphene bilayer.
It is well known that Monte Carlo simulation is a successful method for mixed-spin Ising models, which can provide proper
evidences for the analysis of the physical origin of magnetism. By modeling the single-layer and double-layer nano-graphene
mixed-spin Ising systems, Monte Carlo simulation has been used to investigate the effects of the defects, the size effect, the
crystal field and the external magnetic field on the magnetization, the blocking temperature and hysteresis loops of the nano-
graphene system with mixed spin s ¼ 3/2 and spin S ¼ 5/2 [18e20]. It has been discovered that the double hysteresis loop
behaviors exist for some certain physical parameters. In our previous work, the thermodynamic properties, such as internal
energy and the specific heat of a nano-graphene bilayer in the external magnetic field have been discussed by Monte Carlo
simulation [21]. Although many efforts have been devoted to understanding the physical properties of the nano-graphene
bilayer, less attention has been paid to the magnetic properties of the nano-graphene bilayer subjected to the external
magnetic field by using Monte Carlo simulation. In particular, the blocking temperature dependences of different physical
parameters have not been further clarified. Thus, we intend to explore whether the nano-graphene bilayer system can ex-
press richer hysteresis loop behaviors. Therefore, our intention of this paper is to study the effects the exchange coupling, the
single-ion anisotropy and the external magnetic field on the magnetization, the susceptibility, the blocking temperature and
hysteresis loops of the nano-graphene bilayer.
The outline of this paper is classified as follows. The model, the Hamiltonian of the system and the calculation procedure of
Monte Carlo simulation have been discussed detailedly in Section 2. In Section 3, we shall show typical simulation results for
the magnetization, the susceptibility, the blocking temperature and the hysteresis loops. The summary and conclusions are
obtained finally in Section 4.

2. Model and Monte Carlo simulation

The structure of the nano-graphene bilayer is depicted in Fig. 1. It consists of different spins with the antiferromagnetic
intralayer exchange coupling in the upper layer ([Y, s ¼ 3/2) and with the ferromagnetic intralayer exchange coupling in the
bottom layer ([[, S ¼ 5/2). Similar three-dimensional structures have also been described in Refs. [17,21]. Then we shall
present the Hamiltonian of the system as follows:
0 1
X X X X  X 2 X X
H ¼ JA sziA szjA  JB SziB SzjB  JAB sziA SzjB  DA sziA 2  DB SzjB  h@ sziA þ SzjB A (1)
i;j i;j i;j i j i j

where the first three sums only operate the nearest-neighbours, sziðjÞA represents the spin operator on the upper layer
ðsziA ¼ ±3=2; ±1=2Þ, and SziðjÞB represents the spin operator on the bottom layer ðSzjB ¼ ±5=2; ±3=2; ±1=2Þ.Both JA (<0) and JB
(>0) denote the exchange couplings in the upper and bottom layer, respectively. JAB stands for the interlayer exchange
coupling between the upper and bottom layer. DA and DB are the single-ion anisotropies in the upper and bottom layer,
respectively. h denotes the external magnetic field.
For our simulation, the method used is the standard single-spin-flip importance method on the basis of the Metropolis
algorithm [22], which is constructed on the N  L three-dimensional honeycomb lattice. We define N and L as the number of
spins in each layer and the thinness of the bilayer (L ¼ 2), respectively. Further simulations have also been carried out for the
choice of the size of system. For the variations of N from 48 to 108, we have not found obvious differences. Therefore, such is
the case with N ¼ 48 for the selection of our simulation, shown in Fig. 1. As a guarantee, not less than 5000 Monte Carlo steps
(MCS) per site were eliminated to balance the system before averaging thermal quantities over the next 15000 MCS. The
physical quantities are calculated as follows:
The upper and bottom layer magnetizations MA, MB are expressed by:
460 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

Fig. 1. The nano-graphene bilayer with the honeycomb structure has been described here. JA and JB represent the intralayer antiferromagnetic exchange cou-
plings of different spins in the upper layer ([Y, s ¼ 3/2) and bottom layer ([[, S ¼ 5/2), respectively. JAB denotes the intralayer antiferromagnetic exchange
coupling between layers.

1 X z
MA ¼ 〈 S 〉
N i iA
(2)
1 X z
MB ¼ 〈 SjB 〉
N
j

and the total magnetization M per site

1
M¼ ðM þ MB Þ (3)
2 A

The upper and bottom layer susceptibility are given by:

 
cA ¼ Nb < MA2 >  < MA > 2
  (4)
cB ¼ Nb < MB2 >  < MB > 2

Moreover, the magnetic susceptibility of per site:

 
c ¼ 2Nb < M2 >  < M > 2 (5)

With b ¼ 1/kBT, where kB (¼1) is defined as the Boltzmann constant. In order to simplify, here kB ¼ 1.
As is well known, super paramagnetism is a main problem related to the thermal stability of high density magnetic
recording materials. Since the blocking temperature TB is the characteristic one, at which the recording bit maintains
maximum, TB can be one of important parameters in examining the thermal stability of the present system. When the
temperature T is higher than TB, the system shows super paramagnetic phase, whereas performs strong ferromagnetic or
ferromagnetic phase for the case of T < TB. In general, experimentally TB may be obtained from the measurements of ZFC (zero-
field-cooled) and FC (field-cooled) magnetization curves [23,24]. In the low external magnetic field (10Oe or 20Oe), there
exists one bifurcation point in the ZFC and FC curves, at which the temperature corresponds to TB and the system shows
irreversible magnetic behaviors when T is lower than TB. Apart from this, theoretically one can determine TB from the peak
susceptibility curve versus the temperature in previous studies [17e21], which has also been applied in the present work.
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 461

3. Numerical results and discussions

In this section, we adopt JAB as the reduced unit of temperature and energy, then fix jJABj ¼ 1. On account of the antifer-
romagnetic intralayer exchange coupling in the upper layer, we suppose the initial directions of spins in the bottom layer
ðSziðjÞB Þ and half of spins in the upper layer ðsziA Þ as well as the longitudinal magnetic field h are along the positive z-direction,
but those of the rest spins szjA in the upper layer are along the negative z-direction. From the theoretical point of view, we
think it necessary and meaningful to compare our results by Monte Carlo simulation with those in Ref. [17] by EFT. Therefore,
in the simulation, we adopt the values of physical parameters JA ¼ 0.6, JB ¼ 0.8, DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25, which are
same to those in Ref. [17].

3.1. Temperature dependence of magnetization and susceptibility

Firstly, typical results are presented to show the effects of the exchange coupling, the single-ion anisotropy and the
longitudinal magnetic field on the magnetization and susceptibility of the nano-graphene bilayer system.
Fig. 2(a)e(d) show the temperature dependences of the total magnetization M, the upper and bottom layer magnetizations
MA, MB, the total magnetic susceptibility c and the upper and the bottom layer magnetic susceptibilities cA, cB for several
values of JA when JB ¼ 0.8, DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25. In Fig. 2(a), it can be seen that M declines from the same
saturation value to the remanence with T increasing for different JA(¼0.2, 0.8, 2.0, 3.0). The saturation value of the total
magnetization is M ¼ [2  5/23/2 þ 3/2]/4 ¼ 1.25 in the external magnetic field. It is interestingly observed that M labeled as
JA ¼ 0.2 firstly increases to the maximum and then decreases with the increase of T. The analysis for the behavior is

Fig. 2. The temperature dependences of the total magnetization M, the upper and the bottom magnetizations MA, MB, the total susceptibility c, and the upper and
the bottom magnetic susceptibilities cА, cВ for the nano-graphene bilayer with JB ¼ 0.8, DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25. The numbers on the curves are the
values of the upper exchange coupling JA.
462 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

attributed to the competition between the temperature and the external magnetic field at low temperatures. The external
magnetic field plays the larger role before the maximum and afterwards works weaker than the temperature. As is plotted in
Fig. 2(b), we can clearly observe the change processes of MA and MB from their saturation values (MA ¼ 0, MB ¼ 2.5) with T
increasing, respectively. The upper magnetization MA firstly increases and then decreases, whereas the bottom magnetization
MB decreases all the way to their remanences. In Fig. 2(c), the total susceptibility first increases to the maximum and sub-
sequently starts to decrease with increasing T in the susceptibility curve, where the maximal value of the curve corresponds to
the temperature is called as the blocking temperature (TB) [17e21]. It can be found that TB increases with increasing jJAj. From
this figure, we can obtain TB ¼ 4.29, 5.36, 7.05, 9.08 for JA ¼ 0.2, 0.8, 2.0, 3.0, respectively. The susceptibility curves in the
upper and bottom layers are extremely consistent with the total susceptibility c from Fig. 2(d). Namely, they at first reach the
maximum and then reduce with increasing T, which are the characteristics of the second-order transition phase for the
magnetic system in the longitudinal magnetic field.
The effect of JB on the magnetization and the magnetic susceptibility with JA ¼ 0.6, DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25 are
shown in Fig. 3(a)e(d). Compared to the effect of JA on the magnetization and the magnetic susceptibility, we can acquire
richer information and note that the effect of JB on the magnetization seems stronger than that of JA. In addition to the same
saturation magnetization value (M ¼ 1.25) in Fig. 2(a), one can notice the extra saturation magnetization value
M ¼ 0.75 at T ¼ 0 from Fig. 3(a). The main difference of the saturation values originates from diverse spin configurations at
T ¼ 0 with two spin states S ¼ 3/2 and S ¼ 5/2 in the bottom layer and only one spin state s ¼ 3/2 in the upper layer. Thus, the
calculation of the extra saturation value is M ¼ [2  3/23/2 þ 3/2]/4 ¼ 0.75. It is clearly seen that the M curves present non-
monotone behaviors in low temperatures when JB  0.6. At the same temperature, M increases with increasing JB. From
Fig. 3(b), we can find that JB makes a strong difference to the saturation values of the bottom layer MB (¼1.5, 2.5). As plotted in

Fig. 3. The temperature dependences of the total magnetization M, the upper and the bottom magnetizations MA, MB, the total susceptibility c, and the upper and
the bottom magnetic susceptibilities cА, cВ for the nano-graphene bilayer when JA ¼ 0.6, DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25. The numbers on the curves are the
values of the bottom exchange coupling JB.
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 463

Fig. 3(c), one can note that the positions of the maximum of the magnetic susceptibilities tend to higher temperatures on
increasing JB. This suggests that the blocking temperatures are increasing with increasing JB. It is worth mentioning that there
also exists a small peak at low temperatures besides the blocking temperature TB. The fact can be explained as an abrupt
change in M curve with regard to the small values of JB (¼0.2, 0.6) in Fig. 3(a). From Fig. 3(d), we can see a similar behavior that
a second peak occurs in the cB curve for the small values of JB (¼0.2, 0.6) at low temperatures, which gives a direct evidence of
the double-peak behavior in c curve in Fig. 3(c). These results are excellently consistent with the previous theoretical works
[17,18] and the study with the method of Green function [25].
Next, we shall express the temperature dependences of the magnetization and the magnetic susceptibility of the system
for different values of DA for JA ¼ 0.6, JB ¼ 0.8, DB ¼ 1.5 and h ¼ 1.25, as shown in Fig. 4(a)e(d). In Fig. 4(a), one can notice
that the system presents two distinct saturation values (M ¼ 1.25, 1.5). The explanations for the saturation magnetization
value M ¼ 1.5 are attained by Fig. 4(b). The saturation magnetization value of the upper layer is MA ¼ 0.5 and that of the
bottom layer is MB ¼ 2.5, then it results in the total magnetization (2.5 þ 0.5)/2 ¼ 1.5. Furthermore, we find that the
magnetization curves remain almost unchangeable for different DA. This indicates that the effect of DA on M is slight, and that
both MA and MB seem insensitive to the effect of DA. One can see from c curves in Fig. 4(c), the blocking temperature increases
with decreasing jDAj. It is clearly noticed that there also exists a small peak before the blocking temperature in the c curve for
DA ¼ 2.0, which results from the sudden drop of the upper magnetization MA at the low temperatures in Fig. 4(b). In Fig. 4(d),
it is clearly seen that a remarkable peak in the cА curve at the low temperature is stressed on the effect of DA (¼2.0).
Fig. 5(a)e(d) show the temperature dependences of the magnetization and the magnetic susceptibility for different values
of DB with fixed JA ¼ 0.6, JB ¼ 0.8, DA ¼ 0.5 and h ¼ 1.25. From Fig. 5(a), one can see that the profiles of the M curves become
more plentiful compared with Fig. 4(a). This illustrates that the effect of DB on M seems stronger than that of DA. It is clear that

Fig. 4. The temperature dependences of the total magnetization M, the upper and the bottom magnetizations MA, MB, the total susceptibility c, and the upper and
the bottom magnetic susceptibilitiescА, cВ for the nano-graphene bilayer with fixed JA ¼ 0.6, JB ¼ 0.8, DB ¼ 1.5 and h ¼ 1.25. The numbers on the curves are the
values of the upper single-ion anistropy DA.
464 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

Fig. 5. The temperature dependences of the total magnetization M, the upper and the bottom magnetizations MA, MB, the total susceptibility c, and the upper and
the bottom magnetic susceptibilities cА, cВ for the nano-graphene bilayer when JA ¼ 0.6, JB ¼ 0.8, DA ¼ 0.5 and h ¼ 1.25. The numbers on the curves are the
values of the bottom single-ion anistropy DB.

M decreases as jDBj increases for the fixed temperature. There are three possibilities for the saturation values M (¼1.25, 0.75,
0.25) at T ¼ 0 in M curves. This result comes from various spin states (S ¼ 5/2, 3/2, 1/2) in the bottom layer at T ¼ 0, whereas
only one spin state (s ¼ 3/2) in the upper layer. One can easily find from Fig. 5(b) that three saturation values of MB (¼2.5, 1.5,
0.5) appear for different values of DB because the spin state (S ¼ 5/2) of sublattice B in the bottom layer accesses to the low
state (S ¼ 3/2 or 1/2) for the stronger anisotropy DB. From Fig. 5(c), we can also see that the double-peak behavior occurs in the
c curve for DB ¼ 2.5. Note that the blocking temperature TB moves opposite to the increase of T. Comparing the effects of DB
and DA on TB, one can draw a conclusion that both increasing jDAj and jDBj make TB decrease. From a point of energy, this
phenomenon can be explained as follows: the stronger DA and DB may lead to the stronger the single-ion anisotropy energy of
the system, which makes spins flip more easily and TB decreases with the increase of values of jDAj, jDBj reverse to the effect of
JA, JB. Such similar results have been achieved in a mixed-spin (2, 1/2) ferromagnetic system [26]. In Remarkable, our results on
magnetizations are in excellent agreement with the recent experimental evidences of nano-graphene since the experimental
observations have revealed the existence of various maximums in the magnetization curves at low temperatures, suggesting
the coexistence of different types of magnetic states [27,28].
Finally, the effect of the longitudinal magnetic field h on the magnetization and the magnetic susceptibility are shown with
JA ¼ 0.6, JB ¼ 0.8, DA ¼ 0.5 and DB ¼ 1.5 in Fig. 6(a) and (b). In Fig. 6(a), M increases as h increases with the fixed tem-
perature. This means that the longitudinal magnetic field can stabilize the longitudinal magnetization of the system.
Therefore, the remanence can be improved by means of enhancing h. From Fig. 6(b), the height of the peaks of the magnetic
susceptibility c curves decreases and these c curves become more smoothly with increasing h. Such similar behavior has been
observed in other magnetic Ising systems [29e31]. In addition, one can notice that the blocking temperature also increases on
increasing h. Thus, it is remarkable that the effect of h on TB is similar to that of JA and JB, but opposite to that of DA and DB.
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 465

Fig. 6. The temperature dependences of the total magnetization M, the upper and the bottom magnetizations MA, MB, the total susceptibility c, and the upper and
the bottom magnetic susceptibilities cА, cВ for the nano-graphene bilayer with fixed JA ¼ 0.6, JB ¼ 0.8, DA ¼ 0.5 and DB ¼ 1.5. The numbers on the curves are
the values of the longitudinal magnetic field h.

Comparison of our results with the study by using EFT [17] for the same physical parameters, we obtain some similar con-
clusions, which prove our results reliably and reasonably. However, there persist differences between us. For example, our
simulation results of TB (¼2.06, 2.47, 3.35, 4.99 and 8.28) for different values of JB (¼0.2, 0.4, 0.6, 0.8 and 1.2) are smaller than
those of EFT [17]. This phenomena can be made clear that Monte Carlo simulation has considered the spin contributions of all
lattice points rather than the nearest-neighbour points as EFT [17].

3.2. The blocking temperature

In this subsection, we will devote to the effects of the exchange couplings, the single-ion anisotropies and the longitudinal
magnetic field on the blocking temperature TB of the nano-graphene bilayer in Figs. 7e12. As mentioned above, the blocking
temperature TB is determined by means of the broad maximum exhibited in the susceptibility curve.
Fig. 7 shows the blocking temperature TB associated dependences of both intralayer exchange couplings JA and JB within
the upper and bottom layer for DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25. With the increase of jJAj, TB firstly increases slowly and then
quickly for the weaker JB, whereas almost remains constant for the stronger JB. And we can also see the approximately linear
increase of TB for the weaker jJAj as well as the initial increase and subsequent decrease for the greater jJAj with increasing JB.
These behaviors are in accordance with those based on Monte Carlo simulation in Ref. [18].

Fig. 7. The blocking temperature TB for the nano-graphene bilayer under the effects of the upper and bottom exchange couplings JA, JB in three dimensional space
(TB, JA, JB) for DA ¼ 0.5, DB ¼ 1.5 and h ¼ 1.25.
466 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

The united effects of both single-ion anisotropies DA, DB in the upper and bottom layer on TB with JA ¼ 0.6, JB ¼ 0.8 and
h ¼ 1.25 are displayed in Fig. 8. From the figure, we can find that DA and DB have the similar effects on TB, but different from
those of JA and JB. Namely, TB may be reduced by enhancing jDAj and jDBj. It is clearly seen that TB decreases to the certain value
and then remains constant with increasing jDAj and jDBj, respectively. Comparing Fig. 7 with Fig. 8, we can conclude that the
effects of the bottom layer in the exchange coupling (JB) and the single-ion anisotropy (DB) on the blocking temperature
express more strongly than those of the upper layer (JA, DA). It is well-known fact that the single-ion anisotropies (DA, DB) can
destroy the configurations of the system in the ground state. Thus, the stronger single-ion anisotropy forces spins to flip more
easily and stay in low spin states, which make the blocking temperature TB become lower.
Fig. 9 shows the united effects of the exchange coupling JA and the longitudinal magnetic field h on TB with fixed JB ¼ 0.8,
DA ¼ 0.5 and DB ¼ 1.5. TB increases lineally as h increases for the same JA. Also, TB remains constant and seems insensitive to
the variation of JA when h is fixed. This suggests that JA affects TB slightly. TB dependences of JB and h is displayed in Fig. 10 for
JA ¼ 0.6, DA ¼ 0.5 and DB ¼ 1.5. One can notice that the effects of JB and h on TB are comparative. Namely, both the
increasing of JB and h may be beneficial to increase of TB.
Fig. 11 shows the united effects of the upper single-ion anisotropy DA and the longitudinal magnetic field h on TB when
fixed JA ¼ 0.6, JB ¼ 0.8 and DB ¼ 1.5. With the increase of jDAj, TB declines and then maintains constant for the lower h
compared with increasing all the way for the higher h. Moreover, TB decreases with the decrease of h for fixed DA. Thus, we can
know that DA produces the opposite effect on TB contrast with h. It is obvious that the weaker DA and h have greater influences
on TB. Fig. 12 shows the united effects of the bottom single-ion anisotropy DB and the external magnetic field h on TB for
JA ¼ 0.6, JB ¼ 0.8 and DA ¼ 0.5. From this figure, one can see that TB increases as h increases with the fixed DB, which is more
obvious in the stronger external magnetic field. Besides, one can see that TB first reduces sharply and then increases slowly on
increasing jDBj when fixed h. For comprehensive analysis of the above results, we can remark that the increasing of jDAj and
jDBj may prompt the blocking temperature to decline contrary to JA, JB and h. Our results are consistent with those of Ref. [17]
within EFT and Refs. [18e20] with Monte Carlo stimulation.

3.3. Hysteresis loops

In this subsection, let us concentrate on the effects of the exchange couplings and the single-ion anisotropies upon the
hysteresis loops of the system, as shown in Figs. 13e16. From these figures, we can find that the hysteresis loops are all
symmetric for the external magnetic field in the range of 10 < h < 10.
Fig. 13 shows the effect of JA on the hysteresis loops for the nano-graphene bilayer with JB ¼ 0.5, DA ¼ 0.5, DB ¼ 5.0 and
T ¼ 0.5. As the value of JA changes from 0.5 to 2.5, the hysteresis loop increases from single loop to triple loops. It can be
noticed that the coercivity increases with increasing jJAj, whereas the remanence of the system remains unchangeable. The
effect of JB on the hysteresis loops are shown in Fig. 14 for JA ¼ 1.5, DA ¼ 0.5, DB ¼ 5.0 and T ¼ 0.5. The remanences and the
coercivities of the hysteresis loops for different JB have the same tendencies as those for different JA when JB increases. Namely,
the coercivity can be improved by increasing JB and the remanence maintains constant. It is interesting that the effect of JB on
the shape of the hysteresis loop is also different from that of JA in some ways. Under the influence of JB, the system displays

Fig. 8. The blocking temperature TB for the nano-graphene bilayer under the effects of the upper and bottom single-ion anisotropies DA, DB in three dimensional
space (TB, DA, DB) for JA ¼ 0.6, JB ¼ 0.8 and h ¼ 1.25.
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 467

Fig. 9. The blocking temperature TB for the nano-graphene bilayer under the effects of the upper exchange coupling JA and the longitudinal magnetic field h in
three dimensional space (TB, JA, h) for JB ¼ 0.8, DA ¼ 0.5 and DB ¼ 1.5.

Fig. 10. The blocking temperature TB for the nano-graphene bilayer under the effects of the bottom exchange coupling JB and the longitudinal magnetic field h in
three dimensional space (TB, JB, h) for JA ¼ 0.6, DA ¼ 0.5 and DB ¼ 1.5.

richer hysteresis behaviors. More multiple hysteresis loops behaviors such as triple and quintuple hysteresis loops have been
observed.
We shall focus on the effect of DA on the hysteresis loops in Fig. 15 with JA ¼ 2.5, JB ¼ 0.5, DB ¼ 5.0 and T ¼ 0.5. With the
change of DA from 1.0 to 2.5, the number of loops changes firstly from three to five. When DA ¼ 4.0, the hysteresis loop
turns into the central small one. Further changing DA, the system again displays the triple loops for DA ¼ 4.5. In addition, on
increasing jDAj, the coercivity and the remanence decrease gradually and the area of every single loop reduces. The multiple
hysteresis loops are found, depending on the combinations of both upper and bottom layer magnetizations in the longitudinal
magnetic field. We should mention that the multiple hysteresis loop behaviors have also been discovered in nanoparticles
[32,33], cubic metal nanotubes [34], nanowires [35], nanoscaled thin films [36] and molecular-based magnetic materials [29].
The double and triple hysteresis loops have been observed in the nano-graphene [18e20]. Experimentally similar interesting
triple hysteresis loop phenomenon has also been discussed in nanoparticles [37] and a series of FeGa/CoFeB, FeGa/Py and
CoNip/Cu, multilayered nanowires [38]. It is worthwhile to note that the triple hysteresis loop behavior may have potential
applications in multi-state memory devices.
The effect of DB on the hysteresis loops is presented in Fig. 16 for JA ¼ 0.5, JB ¼ 1.2, DA ¼ 0.5 and T ¼ 0.5. It is clearly
observed that the shape of the loop remains single and the area of the loop declines with increasing jDBj. At the same time,
468 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

Fig. 11. The blocking temperature TB for the nano-graphene bilayer under the effects of the upper anisotropy DA and the longitudinal magnetic field h in three
dimensional space (TB, DA, h) for JA ¼ 0.6, JB ¼ 0.8 and DB ¼ 1.5.

Fig. 12. The blocking temperature TB for the nano-graphene bilayer under the effects of the bottom anisotropy DB and the longitudinal magnetic field h in three
dimensional space (TB, DB, h) for JA ¼ 0.6, JB ¼ 0.8 and DA ¼ 0.5.

both the coercivity and the remanence also decrease. It is noted that the hysteresis loop expresses the step effect with
increasing jDBj, namely magnetization plateaus appears for some values of h. In order to explain the origin of the magneti-
zation plateaus phenomenon clearly, we have plotted the hysteresis loops of the total system and two layers for DB ¼ 0.5 and
DB ¼ 2.5 with JA ¼ 0.5, JB ¼ 1.2, DA ¼ 0.5, T ¼ 0.5 in Fig. 17(a) and (b), respectively. From Fig. 17(a), for weaker DB ¼ 0.5, we
can observe that the atoms of the bottom layer have only one spin state SzjB ¼ 5=2, but the atoms of the upper layer appear two
states which is firstly at sziA ¼ 3=2 state and then at sziA ¼ 0 state with decreasing external magnetic field h from 10 to 0. As a
result, we can find that there exists one magnetization plateau in the hysteresis loops of the total system for DB ¼ 0.5. A finite
external magnetic field on dotted line is referred to as the critical field hc (hc ¼ 6.75), where transition happens between
magnetization plateaus. On the other hand, for the case of stronger DB ¼ 2.5 in Fig. 17(b), we can notice that the atoms of the
bottom layer have two spin states which is firstly at SzjB ¼ 5=2 state and then at SzjB ¼ 3=2 state with decreasing the external
magnetic field h from 10 to 0 except that the spin states of atoms of the upper layer have the same tendency with Fig. 17(a).
Thus, three magnetization plateaus appear at around critical fields hc1 ¼ 6.5, hc2 ¼ 1.5 and hc3 ¼ 0 in the hysteresis loops of the
total system for DB ¼ 2.5. It is concluded that the existence of these magnetization plateaus reflects the variations of the spin
configurations in the system, which is due to the competitions among the anisotropy, the longitudinal magnetic field and
temperature. Here it should be mentioned that the low-dimensional spin systems may exhibit magnetization plateaus at low
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 469

Fig. 13. The hysteresis loops of the nano-graphene bilayer for different values of JA with fixed JB ¼ 0.5, DA ¼ 0.5, DB ¼ 5.0 and T ¼ 0.5.

Fig. 14. The hysteresis loops of the nano-graphene bilayer for different values of JB when fixed JA ¼ 1.5, DA ¼ 0.5, DB ¼ 5.0 and T ¼ 0.5.
470 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

Fig. 15. The hysteresis loops of the nano-graphene bilayer for various values of DA with JA ¼ 2.5, JB ¼ 0.5, DB ¼ 5.0 and T ¼ 0.5.

Fig. 16. The hysteresis loops of the nano-graphene bilayer for various values of DB when JA ¼ 0.5, JB ¼ 1.2, DA ¼ 0.5 and T ¼ 0.5.
 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472 471

Fig. 17. The hysteresis loops of the nano-graphene bilayer system and two layers with JA ¼ 0.5, JB ¼ 1.2, DA ¼ 0.5 and T ¼ 0.5 for two cases of (a) DB ¼ 0.5 and
(b) DB ¼ 2.5.

temperatures or ground state in the external magnetic field. But we also recognize that the number of the magnetization
plateaus does not satisfy the 2S þ 1 criterion for our shown results [39,40]. The behaviors are similar to that of magnetic
nanoparticle in Ref. [41], but are different from those [39,40,42], possibly due to the difference between our models.

4. Conclusions

In summary, the Monte Carlo simulation has been used to investigate the magnetic properties of the nano-graphene
bilayer in the longitudinal magnetic field. We have found that the exchange coupling, the single-ion anisotropy and the
longitudinal magnetic field act as important roles in the magnetic properties of the nano-graphene system. It is shown that
the bottom layer of nano-graphene in the exchange coupling and the single-ion anisotropy have more important influences
on the magnetic properties than the upper layer. The effects of JA and JB on the blocking temperature TB are parallel to h but
reserve to DA and DB. Namely, TB decreases with increasing jDAj, jDBj but increases as JA, JB and h increase. In a certain parameter
range, the system can display multiple hysteresis loops such as triple and quintuple hysteresis loops. It has also been found
that the coercivity can be improved by increasing the exchange couplings (jJAj, JB) and decreasing the anisotropies (jDAj, jDBj).
We desire that our results for the magnetic properties of the nano-graphene may offer a favorable guidance to the further
exploration of nano-graphene.

Acknowledgement

This project was supported by Ph.D Start-up Fund of Liaoning province (Grant No. 20141067), Youth Training Fund of
Shenyang University of Technology (Grant No. 005609), College Students' Innovative and Entrepreneurial Training Program of
Liaoning province (Grant No. 20160684), Research Fund for the Doctoral Program of Higher Education of China (Grant No.
20122102110007), Natural Sciences Foundation of Liaoning province, China (Grant No. 2013020109) and Program for Liaoning
Innovative Research Team in University (Grant No. LT2014004).

References

[1] E. Roy, S. Patra, D. Kumar, R. Madhuri, P.K. Sharma, Biosens. Bioelectron. 68 (2015) 726.
[2] Y. Song, X.M. Li, C. Mackin, X. Zhang, W.J. Fang, T. Palacios, H.W. Zhu, J. Kong, Nano Lett. 15 (2015) 2104.
[3] F. Kisslinger, C. Ott, C. Heide, E. Kampert, B. Butz, E. Spiecker, S. Shallcross, H.B. Weber, Nat. Phys. 11 (2015) 650.
[4] Y. Chen, C.L. Tan, H. Zhang, L.Z. Wang, Chem. Soc. Rev. 44 (2015) 2681.
[5] L.J. Xu, W. Chu, L. Gan, Chem. Eng. J. 263 (2015) 435.
[6] S. Lee, K. Lee, Z. Zhong, Nano Lett. 10 (2010) 4702.
[7] M. Lotya, Y. Hernandez, J. Coleman, J. Am. Chem. Soc. 131 (2009) 3611.
[8] D. Li, M.B. Mueller, S. Gilje, Nat. Nanotechnol. 2 (2008) 101.
[9] A.T. Raji, E.B. Lombardi, Phys. B 464 (2015) 28.
[10] K. Szalowski, J. Magn. Magn. Mater. 382 (2015) 318.
[11] K. Wakabayashi, M. Sigrist, M. Fujita, J. Phys. Soc. Jpn. 67 (1998) 2089.
[12] A. Boumali, Phys. Scr. 90 (2015) 45702.
[13] L. Xu, Y. Zhou, J. Zhang, Solid. State Commun. 41 (2015) 212.
[14] N. Peres, F. Guinea, A.H. Castro Neto, Phys. Rev. B 73 (2006) 125411.
[15] V.V. Nelayev, A.I. Mironchik, Mater. Phys. Mech. 9 (2010) 26.
472 W. Wang et al. / Superlattices and Microstructures 98 (2016) 458e472

[16] X. Zhong, R. Pandey, S.P. Karna, Carbon 50 (2012) 784.

[17] W. Jiang, Y. Yang, A. Gong, Carbon 95 (2015) 190.
[18] R. Masrour, L. Bahmad, A. Benyoussef, J. Magn. Magn. Mater. 324 (2012) 3991.
[19] R. Masrour, L. Bahmad, A. Benyoussef, M. Hamedoun, E.K. Hlil, J. Supercond. Nov. Magn. 26 (2013) 679.
[20] L. Bahmad, R. Masrour, A. Benyoussef, J. Supercond. Nov. Magn. 25 (2012) 2015.
[21] X.H. Luo, W. Wang, D. Chen, S. Xu, Phys. B 491 (2016) 51.
[22] N. Metropolis, A.W. Rosenbluth, M.N. Rosenbluth, A.H. Teller, E. Teller, J. Chem. Phys. 21 (1953) 1087.
[23] L. Wang, J. Ding, H.Z. Kong, Phys. Rev. B 64 (2001) 214410.
[24] Y.W. Du, Prog. Phys. 13 (1993) 255.
[25] W. Jiang, A.B. Guo, Chin. Phys. 16 (2007) 3832.
[26] G.M. Buendıa, J.A. Liendo, J. Phys. Condens. Matter 9 (1997) 5439.
[27] C.N.R. Rao, K.S. Subrahmanyam, H.S.S.R. Matte, B. Abdulhakeem, A. Govindaraj, B. Das, et al., Sci. Technol. Adv. Mater. 11 (2010) 054502.
[28] H.S.S.R. Matte, K.S. Subrahmanyam, C.N.R. Rao, J. Phys. Chem. C 113 (2009) 9982.
[29] W. Wang, D. Lv, F. Zhang, J. Bi, J. Chen, J. Magn. Magn. Mater. 385 (2015) 16.
[30] W. Jiang, B.D. Bai, V.C. Lo, W. Liu, J. Appl. Phys. 97 (2005) 103077e103081.
[31] W. Wang, W. Jiang, D.D. Chen, S.Y. Xu, IEEE Trans. Magn. 51 (2015) 1300805.
[32] E. Kantar, B. Deviren, M. Keskin, Eur. Phys. J. B 86 (2013) 253.
[33] Y. Yuksel, E. Vatansever, H. Polat, J. Phys. Condens. Matter 24 (2012) 436004.
[34] J.Y. Liang, C.L. Zou, W. Jiang, X.X. Li, Phys. E 69 (2015) 81.
[35] W. Jiang, X.X. Li, L.M. Liu, J.N. Chen, F. Zhang, J. Magn. Magn. Mater. 353 (2014) 90.
[36] H. Magoussi, A. Zaim, M. Kerouad, Superlattices Microstruct. 89 (2016) 188.
[37] H. Zhou, Y. Yiu, M.C. Aronson, S.S. Wong, J. Solid State Chem. 181 (2008) 1539.
[38] N. Lupu, M. Lostun, H. Chiriac, J. Appl. Phys. 107 (2010) 09E315.
[39] M. Oshikawa, M. Yamanaka, I. Affleck, Phys. Rev. Lett. 78 (1997) 1984.
[40] C. Ekiz, H. Yaraneri, J. Magn. Magn. Mater. 318 (2007) 49.
[41] W. Jiang, L.M. Liu, X.X. Li, Q. Deng, H.Y. Guan, F. Zhang, A.B. Guo, Phys. B 407 (2012) 3933.
[42] W. Wang, W. Jiang, D. Lv, Phys. Status. Solidi B 249 (2012) 190.