close

Вход

Забыли?

вход по аккаунту

?

j.intermet.2017.10.014

код для вставкиСкачать
Intermetallics xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Intermetallics
journal homepage: www.elsevier.com/locate/intermet
Atomic structure, stability and electronic properties of S(Al2CuMg)/Al
interface: A first-principles study
Xingzhi Panga,b, Wenchao Yangb, Jianbing Yangc, Mingjun Pangd, Yongzhong Zhana,b,∗
a
School of Materials Science and Engineering, South China University of Technology, Guangzhou, Guangdong 510641, PR China
College of Materials Science and Engineering, Guangxi University, Nanning, Guangxi, 530004, PR China
c
Xingjian College of Science and Liberal Arts, Guangxi University, Nanning, Guangxi, 530004, PR China
d
SAIC-GM-Wuling Automobile Co., Ltd., Liuzhou, Guangxi, 545007, PR China
b
A R T I C L E I N F O
A B S T R A C T
Keywords:
Density functional theory
Al2CuMg/Al interface
Atomic structure
Work of separation
Interface energy
Interfacial models of Al2CuMg/Al were investigated by first-principles calculations based on density functional
theory. Two types of Al2CuMg(001)/Al(021) interface structures were investigated in consideration of two
different terminations for Al2CuMg(001) surface (Al-terminated and CuMg-terminated). The interaction of interfaces was analyzed by the optimized atomic structures. The ideal work of adhesion (Wad) of the Al2CuMg
(001)/Al(021) interfaces was also calculated. The results show that the interface model with CuMg-terminated is
more stable than that of Al-terminated. It is also demonstrated from the values of interfacial energy (γint) that the
CuMg-terminated interface is more thermodynamically stable. The calculated electronic properties, including
charge density distribution and density of states, reveal that there is a significant hybridization among the
interfacial Cu 3d, Mg 3p and Al 3p states. It is the main reason why CuMg-terminated interface is more stable.
1. Introduction
S phase (Al2CuMg) in Al series alloy has been considered as one of
the most important strengthening precipitates, which has been attracted attention and investigated extensively in recent years [1,2].
There are extensive experimental and theoretical investigations on its
crystal structure [3], morphology [4], growth kinetics [5] and
strengthening mechanisms [6]. Interestingly, it is observed from experiments that Al2CuMg phase are formed in two ways: discontinuous
precipitation and continuous precipitation.
Discontinuous precipitation occurs usually on grain boundaries. In
cast AleZneMgeCu alloy, it is observed that the Al2CuMg phase are
directly formed during solidification process and grew along the grain
boundaries at the medium temperature range (from 350 °C to 450 °C)
[7,8]. In order to dissolve the coarse Al2CuMg phase, which is greatly
influenced on the strength and stress corrosion resistance of alloy,
homogenization treatment is needed. During homogenization treatment, the phase transformation of the solid solution Mg(Zn,Cu,Al)2→Al2CuMg is observed at high temperature [9]. Liu et al. [10] proposed that this phase transition was very difficult when Zn content was
higher than 8% (mass fraction). Unfortunately, the crystallographic
orientation relationships between Al2CuMg phase and Al matrix are few
observed in this system.
∗
However, the continuous precipitation sequence in AleCueMg alloys is widely reported to be: supersaturated solid solution (SSS)→GP I
zone→GP II zone→materstable S'→stable S [11]. The Al2CuMg phase
has a laths-shaped morphology as along 〈100〉Al with {012}Al habit
[12]. According to early investigation [13], three kinds of crystallographic orientation relationships between Al2CuMg phase and Al
matrix are proposed as follow: [100]S//[100]Al, [001]S//[021]Al, and
[010]S//[012 ]Al. Later, Radmilovic et al. [14] observed two types of
Al2CuMg/Al interfaces by using quantitative high resolution electron
microscopy: (001)S//(021)Al and (043)S//(021)Al.
It is well known that the interfaces between precipitate phase and
matrix play an important role in the room temperature toughness and
the high temperature strength of alloys [15]. Therefore, it is worthwhile
to explore the atomic structure and chemical bonding of the Al2CuMg/
Al interfaces in order to understand exactly the strength mechanism of
Al alloys. However, it is still difficult to gain systematic information of
Al2CuMg/Al interface from the experiments.
Recently, the first-principles method has been successfully performed to evaluate the interface properties between Al matrix and
ceramic or precipitate [16–18]. It is of great significance to reveal interface behavior (i. e. interface stability, adhesion strength, atomic
bonding) of Al matrix and ceramic or precipitate. To date, there are few
experimental and theoretical methods are available to quantitatively
Corresponding author. School of Materials Science and Engineering, South China University of Technology, Guangzhou, Guangdong 510641, PR China.
E-mail address: zyzmatres@aliyun.com (Y. Zhan).
http://dx.doi.org/10.1016/j.intermet.2017.10.014
Received 19 May 2017; Received in revised form 23 September 2017; Accepted 19 October 2017
0966-9795/ © 2017 Published by Elsevier Ltd.
Please cite this article as: Pang, X., Intermetallics (2017), http://dx.doi.org/10.1016/j.intermet.2017.10.014
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
orthorhombic BRe3-type structure with space group Cmcm (No. 63). For
orthorhombic Al2CuMg phase, the calculated lattice constants are
a = 4.028 Å, b = 9.338 Å and c = 7.131 Å, which are also in good
agreement with the experimental and other theoretical values [28,29].
In addition, our calculated formation energy for Al2CuMg (−16.5 kJ/
mol) is slightly higher than other calculated value (−19.5 kJ/mol)
[30]. These results show that the adopted parameters in our calculations can ensure enough precision to conduct the subsequent calculations.
Fig. 2 shows the total and partial density of state (DOS) of Al2CuMg
phase, where the Fermi level is set as the origin of the energy scale. It
can be seen that the region from −10 eV to Fermi level is mainly
composed of the hybridization of Al 3p and the delocalized Cu 3d states
with some Mg 2p states, constituting the bonding states below the
Fermi level. The finite DOS value at the Fermi level implies metallic
nature, suggesting strong covalent character. Antibonding levels above
the Fermi level are mainly the hybridization of Al 3p and Mg 2p. Our
calculated results are also consistent with other calculated results [29].
investigate Al2CuMg/Al interface behavior from atomic level point of
view.
In the present work, the Al2CuMg(001)/Al(021) interface was
chosen as our research objective. We mainly concentrate on the atomic
structure, stability and electronic properties of the Al2CuMg/Al interface by using the first-principles method for a better understanding of
the mechanisms responsible of Al2CuMg phase in Al matrix.
2. Calculation method
In this paper, we investigated the Al2CuMg/Al interface by using the
first-principles total energy program CASTEP (Cambridge Serial Total
Energy Package) within the framework of density functional theory
[19]. The plane-wave ultra-soft pseudopotential method was used to
describe the interactions between ionic core and valence electrons [20].
Electron exchange and correlation were treated within the generalized
gradient approximation (GGA) using the Perdew–Burke–Ernzerh (PBE)
functional [21]. In the present calculations, the plane wave cutoff energy was chosen to be 400 eV. Integrations in the Brillouin–zone were
performed using special k points generated with Monkhorst–Pack mesh
[22]. The Pulay scheme of density mixing was applied for the evaluation
of
energy
and
stress
[23].
The
Brodyden–Fletcher–Goldfarb–Shanno (BFGS) minimization scheme was employed
for geometry optimization to complete the relaxation of atoms in supercells [24]. The calculation of total energy and electronic structure
were followed by cell optimization with self consistent field (SCF) tolerance of 5.0 × 10−7 eV/atom. The maximum ionic displacement was
set at 5 × 10−4 Å and maximum stress within 0.02 GPa. A convergence
criterion of 0.01 eV/Å was used for the maximum ionic HellmannFeynman force. For calculation of the electronic density of states (DOS),
we used the linear the tetrahedron Blöchl method with corrections [25].
3.2. Surface properties
The Al(021) surface was modeled by a slab of 3–11 atomic layers
separated by a vacuum region of 15 Å. The 2 × 2 supercells with
12 × 6 × 2 Monkhorst–Pack k-points in the Brillouin zone were used to
calculate the Al(021) surface in this work. The unit cell of Al2CuMg is
stacked with atomic layers along the c direction in the following order:
ABAABA (A and B denote Al-terminated and CuMg-terminated layers,
respectively). Therefore, there are two kinds of surfaces for Al2CuMg
(001): Al-terminated (Surface I) as well as CuMg-terminated (Surface
II), as shown in Fig. 1 (c) and (d). Both surfaces were modeled by a slab
of 3–12 atomic layers separated by a vacuum region of 15 Å, which was
found to be sufficient to prevent interactions between periodic images.
In this case, the Brillouin zone was sampled using a 12 × 6 × 8 grid.
Generally, it is important to make sure the two slabs are thick enough to show the bulk-like character interiors. Therefore, the convergence tests on the Al(021) and Al2CuMg(001) surfaces with respect
to slab thickness were performed. Here, we used surface energy to estimate the convergence of surface thickness, which is one of the basic
qualities to describe stabilities of surface. It is converges to a fixed value
when attaining a critical thickness [33]. Surface energy is calculated
according to the formula as below
3. Results and discussion
3.1. Bulk properties
In order to assess the accuracy of our computational method, a
series of calculations on the bulk properties of Al and Al2CuMg are
performed. Table 1 lists the calculated values for each material, together with the available data from experiments and other calculations.
The calculated lattice constant for bulk face centered cubic (fcc) Al is
a = 4.047 Å, which agrees well with the experimental value (4.050 Å)
[26] and other theoretical value [27]. Meanwhile, the plane spacing of
Al(021) is 0.905 Å, which is agreed with those determined by the experiments [14]. The deviations from first-principles calculation are in
the reasonable range of computational errors.
As shown in Fig. 1 (a) and (b), Al2CuMg phase crystallizes in the
Eslab −
Esurf =
Al
A2CuMg
Method
Present
Experiments
Calculations
Present
Experiments
Calculations
Lattice parameters (Å)
Bulk modulus
(GPa)
a
b
c
4.047
4.050a
4.050c
4.028
4.012e
4.050f
–
–
–
9.338
9.265e
9.279f
–
–
–
7.131
7.124e
7.206f
Nslab
Nbulk
2A
bulk
(1)
where Eslab and Ebulk are total energies of the surface slab and the bulk
unit cell, respectively. Nslab and Nbulk are numbers of atoms in the
surface slab and the bulk unit cell, respectively. A is the surface area of
supercell. The factor 2 accounts for the double surface of the supercell.
The equation (1) is generally applied to calculate the surface energy of
stoichiometric surface model.
We have conducted surface energies of Al(021) slabs thickness
ranging from 3 to 11 layers, as shown in Fig. 3. It is found that the
surface energies are converged to about 1.0 J/m2 (dashed line in Fig. 3)
for the thickness of Al(021) being equal or larger than 9 atomic layers.
It means that the slab with more than 9 atomic layers exhibits bulk-like
interiors. Therefore, all of the following calculations are based on the
slab with 9 layers for Al(021) surface.
The surface energies of two types Al2CuMg(001) surfaces were also
calculated by Equation (1). The number of atomic layers (n) was set as
3, 6, 9 and 12, respectively. The results are listed in Table 2. It is found
that the surface energies of the Surface I with more than six layers can
converge to 1.40 J/m2. The surface energies of Surface II nearly remain
unchanged with the increasing of slab thickness and converge to about
1.23 J/m2 when n ≥ 6. Therefore, the Al2CuMg(001) surfaces with 6
atomic layers were adopted in the following calculations to insure the
Table 1
Comparison of lattice parameters and bulk modulus between calculations and experiments for bulk Al and A2CuMg.
Bulk
( )E
78.43
79.00b
77.32d
72.57
–
75.210f
a
Ref. [26].
Ref. [31].
c
Ref. [27].
d
Ref. [32].
e
Ref. [28].
f
Ref. [29].
b
2
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
Fig. 1. The unit cell of bulk Al2CuMg: (a)
Top view, (b) Side view; Unrelaxed Al2CuMg
(001) with Al-terminated (c) and with
CuMg-terminated (d) surface slab unit cells
in our calculations, where the white, red and
blue spheres represent Mg atom, Al atom
and Cu atom, respectively. All the surface
slabs have a 1 × 1 surface geometry. (For
interpretation of the references to colour in
this figure legend, the reader is referred to
the web version of this article.)
Fig. 2. Calculated total and partial density of states for Al2CuMg.
Table 2
Surface energy of the Al2CuMg (001) surface with different thickness (atomic layer).
Surface
Number of layers, n
Surface energy (J/m2)
Model I
(Al-terminated)
3
6
9
12
3
6
9
12
1.42
1.40
1.40
1.39
1.25
1.23
1.23
1.24
Model II
(CuMg-terminated)
bulk-like interior. Furthermore, the surface energy for Surface II is
smaller than that of Surface I, which implies that the CuMg-terminated
surface is more stable than the Al-terminated surface.
Fig. 4 shows the total density of states of the top two layers for the
Al2CuMg(001) surfaces with different terminations, comparing with
that of the corresponding atoms in the bulk Al2CuMg. It can be seen
from Fig. 4(a) that the DOS of bulk Al2CuMg, the Al-terminated and
CuMg-terminated surfaces look similar to each other. However, the
Fig. 3. Convergence of the surface energy with respect to the number of layers for Al(021)
(1 × 2) surfaces.
3
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
Fig. 4. The total density of states of the top two surface layers of
the Al2CuMg (001) surface with different terminations compared
with that of the corresponding atoms in the bulk Al2CuMg (001):
(a), (b), (c) and (d) correspond to the DOS of surface, Al atom, Cu
atom, and Mg atom, respectively.
4
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
when Al2CuMg(001) surface is gradually close to the Al(021) surface.
Therefore, three atomic layers near interface are considered in order to
estimate the interfacial stability.
For Model I, the calculated AleAl distances at interface are 2.575
and 2.786 Å for Al(2)–Al(i) and Al(3)–Al(i), respectively. The AleAl
distance in the bulk Al is 2.863 Å. It can be seen that the distance between the interfacial Al atom of Al2CuMg(001) side and the nearest
neighbor Al atom becomes smaller than that of the bulk Al, suggesting
that the formation of strong interaction. The distances of Al(1)eAl(2)
and Al(1)eAl(3) in Model II are 2.877 and 2.841 Å, respectively. This is
mainly influenced by the Cu(i) atom of Al2CuMg(001) side.
Moreover, the calculated bond lengths of AleCu metallic bond for
Al(i)–Cu(i), Al(ii)–Cu(i) and Al(iii)–Cu(i) in Model I are 2.530, 2.597
and 2.509 Å, respectively. The corresponding AleCu bonds in bulk
Al2CuMg are 2.547, 2.547 and 2.522 Å, respectively. It can be seen that
the change of AleCu bonds in Model I are attributed to the Cu(i) atom
moves towards to the interface along the c direction. Compared with
Model I, the Cu(i) atom has a larger transformation toward to the Al
(021) surface (see Fig. 5(b)). These indicate that a greater interaction
on the atomic configurations for Model II is occurred by bonding.
In addition, the CueMg bond length in bulk Al2CuMg is 2.843 Å.
They are changed into 2.787 and 2.645 Å after structural relaxation for
Model I, respectively. The corresponding CueMg bonds in Model II are
2.792 and 2.630 Å, respectively. It means that the interaction of Cu and
Mg atoms in the interface is stronger than that in the interior structure.
Above all, those results imply that the binding strength of Model II is
larger than that of Model I.
entire peak of DOS for two type surfaces are shifted to the higher energy
level compared with that of bulk Al2CuMg. Note that the shift distance
of DOS for CuMg-terminated surface are the most greatest.
The DOS of Al atoms in bulk Al2CuMg, the Al-terminated and CuMgterminated surfaces are quite different, as shown in Fig. 4(b). It is because the Al atoms in the Al-terminated and CuMg-terminated surfaces
are the first layer and the second layer, respectively. The central layers
can be regarded as those in the bulk. Compared with that of bulk
Al2CuMg, the DOS of Al atoms in the second layer of CuMg-terminated
surface are shifted toward to the higher energy level. Differently, the
main peak of Al atoms for the first layer of Al-terminated surface moves
to conduction band.
For the Cu and Mg atoms in both surfaces, the entire peaks of DOS
are shifted to the higher energy level compared with that of corresponding bulk atoms. It can be deduced that the upward shift of the
valence bands for CuMg-terminated surface is mainly attributed to the
states from the Cu and Mg atoms in the first layer.
Thus, the surface states appearing in both surfaces are caused by the
atoms in the top two layers. This was mainly due to the unsaturated
chemical bonds and the unbalanced atomic interactions, which is
caused by the decrease of neighboring atoms. It is the reason why the
CuMg-terminated surface is more stable than the Al-terminated surface.
3.3. Interfaces
3.3.1. Atomic structures
Based on the experimental observations [14], Al2CuMg(001)/Al
(021) interface models are built in our calculations, as illustrated in
Fig. 5 (a) and (b). According to previous statements on surface convergence, two kinds of Al2CuMg(001)/Al(021) interfaces were modeled
by stacking six-layers 1 × 1 Al2CuMg(001) slab on 1 × 2 nine-layers Al
(021) slab. Both Al-terminated and CuMg-terminated Al2CuMg(001)
surfaces are employed to simulate the Al2CuMg/Al interfaces, which
are named Model I and Model II, respectively. A 15 Å vacuum layer was
included in the supercells in order to avoid the image interactions. The
Brillouin zone was sampled in the k-space using the Monkhorst–Pack
scheme within 12 × 6 × 2 mesh points. The obtained lattice parameters of Al2CuMg(001) surface (a = 4.028 Å, b = 9.338 Å) are close
to that of Al(021) surface (a = 4.050 Å, b = 9.056 Å). The lattice
mismatch between Al(021) and Al2CuMg(001) was calculated by
Bramfitt planar lattice misfit formula [34]. From it, the mismatch between Al(021) surface and Al2CuMg(001) surface are 0.5% and 3.1%
for a and b, respectively. It is indicated that the interface between Al
(021) surface and Al2CuMg(001) surface is quite coherent. It is in good
agreement with the results of HREM observation [14], confirming that
this orientation relationship for Al and Al2CuMg is the most effective
based on Bramfitt misfit theory [34].
All atoms at both interfaces were relaxed during the geometry optimization. The relaxed atomic structures for two interface models are
given in Fig. 5 (a) and (b). It is obviously that the atomic stacking sequences in both interface models are same as that in bulk Al or Al2CuMg
with only the interfacial layer, indicating that the atomic rearrangement degree by structural relaxation is slight. Fig. 5 (a) and (b) show
some selected atoms at or near interface region, which is represented by
the short dash lines. The atoms have been designated accordingly to
their distances from the interface, e.g., atom Al1 is closer to the interface region than atom Al2 and so forth.
It is obvious from Fig. 5 (a) and (b) that the interactions between Al
atoms in Mode I interface are much larger than that of the pure bulk Al.
The Al(1) atom in Al(021) surface slightly moves close to the Al(i) atom
in Al2CuMg(001) surface along the c direction (perpendicular to the
interface). The Al(i) atom slightly moves along the b direction, causing
significant change on coordinates of surrounding atoms. As for Mode II
interface, the Al(2) atom in Al(021) surface moves toward to the Cu(i)
atom in Al2CuMg(001) surface along the c direction. It means that the
interaction between Al(021) and Al2CuMg(001) becomes stronger
3.3.2. The work of interface adhesion and thermal stability
As an important parameter, the ideal work of adhesion of the interface Wad, is the key to predict the mechanical properties of an interface [35]. The value of Wad is the lower limit on the work needed for
a real cleavage experiment [36]. It can be calculated by using the following equation:
Wad =
Al2CuMg(001)
Al(021)
Eslab
+ Eslab
− EAl/Al2CuMg
A
Al(021)
Eslab
Al2CuMg(001)
Eslab
(2)
and
represent the total energies of relaxed Al
where
(021) and Al2CuMg(001) surface models, respectively. EAl/Al2CuMg denotes the total energy of Al2CuMg(001)/Al(021) interface. A is the interface area.
The calculated values of Wad for Model I and Model II, together with
interfacial distance d0, are presented in Table 3. Obviously, it can be
seen that Model II has an interfacial separation of 1.372 Å, which is
nearly 0.345 Å larger than that of Model I (1.027 Å). Furthermore, the
interfacial distance in Model II is shorter than the distance between Al
layer and CuMg layer in bulk Al2CuMg (1.393 Å), which results in the
relatively strong interaction. Compared with the distance of AleAl
layers in the bulk Al2CuMg (0.780 Å), the interfacial distance in Model I
is much larger, meaning that there is a relatively weak interaction in the
Model I interface. What is more, the value for Wad of Model II is
2.412 J/m2, which is larger than that of Model I (1.768 J/m2). That is to
say, the obtained results imply that Model II shows strong localized
hybridization between Al2CuMg(001) surface and Al(021) surface, and
forms more stable interface than Model I.
To date, there are many TEM experiments revealing the interface
structures between Al2CuMg and Al matrix. Winkelman et al. [37] have
observed the orientation relationship (001)S//(021)Al by using highresolution electron microscopy and diffraction. They also found the
atomic arrangement for the interface of Al2CuMg and Al matrix in this
orientation relationship is similar to the CuMg-terminated interface
model in our work as shown in Fig. 5(b). Based on HRTEM observation,
Chen and Wu et al. [30,38] also confirmed that CuMg-terminated interface is easier to form in Al matrix, indicating that our calculations are
consistent with earlier experiment reports.
5
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
Fig. 5. The supercell model of Al2CuMg(001)/Al(021) interface
with two different configurations. (a) Model I (Al-terminated),
and (b) Model II (CuMg-terminated), where the white, red and
blue spheres represent Mg, Al atom and Cu atom, respectively.
(For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Al2CuMg(001), respectively.
Table 3 lists the calculated interfacial energies for Al2CuMg(001)/Al
(021) interface with two models. Generally, the lower the interfacial
energy is, the more stable the interface structure is [40]. It can be seen
that the interfacial energies of Model I and Model II are 0.616 J/m2 and
-0.198 J/m2, respectively. It is indicated that the latter is more thermodynamically stable than the former. That is to say, the interface
between CuMg-terminated Al2CuMg(001) surface and Al(021) surface
is predicted to be preferred equilibrium stacking sequence. It should be
noted that the interface energy of Model II interface is negative, which
is clearly indicated that the attraction between CuMg-terminated
Al2CuMg(001) surface and Al(021) surface is very strong. Such information could be used to improve models of precipitation kinetics.
Table 3
The calculated results of the interfacial separation (d0), the ideal work of adhesion (Wad)
and interfacial energy (γint), corresponding to two models.
Models
d0 (Å)
Wad (J/m2)
γint (J/m2)
Model I
Model II
1.027
1.372
1.768
2.412
0.616
−0.198
3.3.3. Interfacial energy
The interfacial energy is an important parameter governing the
nucleation and growth of precipitates during heat treatment, but is
difficult to measure experimentally. It originates from the change in the
interfacial atomic chemical bonding and the structure strain. Thus, it is
necessary to calculate interfacial energy to evaluate the stability of an
interface. The interfacial energy γ can be given as following [39]:
Al2CuMg(001)
Al(021)
γint = Esurf
+ Esurf
− Wad
3.3.4. Electronic structure
In order to gain insight into the nature of Al2CuMg(001)/Al(021)
interface, we have also explored the electronic structure to examine the
bonding properties, as illustrated in Fig. 6. The red area represents the
depletion of electronic charge while the blue area represents the
(4)
Al(021)
Al2CuMg(001)
and Esurf
are the surface energies of Al(021) and
where Esurf
6
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
Fig. 6. Charge density difference around the interface for both
interface models: (a) Model I, (b) Model II. The dashed-lines
indicate the interface plane.
indicates that the formation of AleCu, MgeCu and AleMg covalent
bonding. It is also observed from Fig. 7(b) that the interior Al layer in
Al2CuMg (001) side is very similar to that of interior Al atoms of Al
(021) side. It is the main reason why Model II is more stable than Model
I.
accumulation of electronic charge. It can be seen from Fig. 6 that the
higher electrons are mainly located around the Cu atoms, indicating
that the electrons of Mg and Al atoms transfer toward to Cu atom in
forming typically ionic characteristic of a metallic bonding, respectively. Liu et al. [30] also confirmed that the bulk Al2CuMg has relatively strong metallic CueMg bonding and CueAl bonding with an
ionic character.
In Fig. 6(a), we can find that a significant charge accumulation
occurs at interfacial Al atoms in the Model I interface, indicating that
AleAl bonds are formed in the interface. Only a relatively weak AleCu
chemical bond is formed in the Model I interface. From Fig. 6(b), the
obvious change transfer from interfacial atoms of each side to interface
is observed in the Model II interface, leading to a localized characteristic and the formation of strong covalent bonding. This means that
interaction in Model II interface is stronger, which is a main reason why
Model II interface yields larger Wad value. These evidently illustrate
that there are more charges transfer or hybridization contributing to the
interfacial bonds for Model II, thus, CuMg-terminated interface is more
stable. However, there is no change in the charge density distribution is
observed in far away from interface.
For a further insight into the nature of bonding and the variations of
the charge distribution for both interfaces, the electronic densities of
state plots are analyzed, as shown in Fig. 7.
For the Model I interface, the PDOS curve of interfacial Al atom of
Al2CuMg (001) side is slightly different from that of the Al atom in bulk
interior, as shown in Fig. 7(a). There are overlap states between the
interfacial Al atoms of Al(021) and Al2CuMg(001) in the range of
−4.0 eV to 2.0 eV, which are the hybridization between the Al 3p
orbits. This suggests that the interface bonding shows certain covalent
feature. Furthermore, the Al 3p orbits of the first layer Al atom in
Al2CuMg(001) surface are overlapped with the Cu 3d and Mg 3p orbits,
especially Al 3p and Cu 3d, forming AleCu covalent bonding near interface. Therefore, the AleAl and AleCu interactions are a combination
of strong covalent characteristic across the Model I interface.
A key feature of the calculated PDOS for the Model II interface is
that the Cu 3d and Mg 3p orbits of the first layer in Al2CuMg(001)
surface are slightly different from that of Cu and Mg in bulk interior
(see Fig. 7(b)). These suggest the strong influence of Cu and Mg atoms
on the interfacial electronic states. Moreover, significant hybridization
can be observed among interfacial Cu 3d, Mg 3p and Al 3p states, which
4. Conclusion
A systematic study of Al2CuMg(001)/Al(021) interfaces have been
conducted by first principles method, which aimed at determining the
stable interface structure, evaluating the adhesion energy, and providing insight into the nature of interfacial bonding. These models
yielded the following results:
(1) Considering two terminations of Al2CuMg(001) surfaces (Al-terminated and CuMg-terminated), the CuMg-terminated surface is more
stable than that of the Al-terminated surface.
(2) The results of interface energy indicate that the interface between
CuMg-terminated Al2CuMg(001) surface and Al(021) surface is
predicted to be preferred equilibrium stacking sequence.
(3) For both interfaces, the CuMg-terminated interface has the larger
Wad (2.412 J/m2), indicating that it is more stable interface than Alterminated interface.
(4) Electronic structure analysis shows that the main reason why
CuMg-terminated interface is more stable is that there is a significant hybridization among the interfacial Cu 3d, Mg 3p and Al 3p
states in CuMg-terminated interface while a relatively weak interactions of AleAl and AleCu in the Al-terminated interface.
Acknowledgements
This research work is jointly supported by the National Key R & D
Program of China (2016YFB0301400), the National Natural Science
Foundation of China (51361002), and the Training Plan of High-Level
Talents of Guangxi University (2015), and Ministry-province jointlyconstructed cultivation base for state key laboratory of processing for
non-ferrous metal and featured materials, Guangxi Zhuang
Autonomous Region (GXKFJ16-02).
7
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
Fig. 7. Total and partial density of states of interfacial atoms
in (a) Model I, (b) Model II interfaces. The vertical dashed
lines indicate the location of Fermi level.
8
Intermetallics xxx (xxxx) xxx–xxx
X. Pang et al.
[20]
[21]
[22]
[23]
[24]
[25]
[26]
References
[1] T. Hashimoto, X. Zhang, X. Zhou, P. Skeldon, S.J. Haigh, G.E. Thompson, Corros.
Sci. 103 (2016) 157.
[2] J. Wang, B. Zhang, B. Wu, X.L. Ma, Corros. Sci. 105 (2016) 183.
[3] M.J. Styles, C.R. Hutchinson, Y. Chen, A. Deschamps, T.J. Bastow, Acta Mater 60
(2012) 6940.
[4] P. Jia, Y. Cao, Y. Geng, L. He, N. Xiao, J. Cui, Mater. Sci. Eng. A 612 (2014) 335.
[5] M.J. Styles, R.K.W. Marceau, T.J. Bastow, H.E.A. Brand, M.A. Gibson,
C.R. Hutchinson, Acta Mater 98 (2015) 64.
[6] Z.Q. Feng, Y.Q. Yang, B. Huang, X. Luo, M.H. Li, M. Han, M.S. Fu, Acta Mater 59
(2011) 2412.
[7] G. Peng, K. Chen, S. Chen, H. Fang, Mater. Sci. Eng. A 641 (2015) 237.
[8] Y. Deng, Z. Yin, F. Cong, Intermetallics 26 (2012) 114.
[9] X. Fan, D. Jiang, Q. Meng, L. Zhong, Mater. Let. 60 (2006) 1475.
[10] Y. Liu, D. Jiang, W. Xie, J. Hu, B. Ma, Mater. Charact. 93 (2014) 173.
[11] Y.C. Lin, Y.-C. Xia, Y.-Q. Jiang, H.-M. Zhou, L.-T. Li, Mater. Sci. Eng. A 565 (2013)
420.
[12] K.S. Kumar, S.A. Brown, J.R. Pickens, Acta Mater 44 (1996) 1899.
[13] J.M. Silcock, J. Inst. Met. 89 (1960–1961) 203.
[14] V. Radmilovic, R. Kilaas, U. Dahmen, G.J. Shiflet, Acta Mater 47 (1999) 3987.
[15] L. Bourgeois, C. Dwyer, M. Weyland, J.-F. Nie, B.C. Muddle, Acta Mater 59 (2011)
7043.
[16] J. Huang, E. Tea, G. Li, C. Hin, Appl. Surf. Sci. 406 (2017) 128.
[17] Y. Xian, R. Qiu, X. Wang, P. Zhang, J. Nucl. Mater. 478 (2016) 227.
[18] J. Li, M. Zhang, Y. Zhou, G. Chen, Appl. Surf. Sci. 307 (2014) 593.
[19] M.D. Segall, P.J.D. Lindan, M.J. Probert, C.J. Pickard, P.J. Hasnip, S.J. Clark, et al.,
J. Phys. Condens. Matter 14 (2002) 2717.
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
9
D. Vanderbilt, Phys. Rev. B 41 (1990) 7892.
J.P. Perdew, Y. Wang, Phys. Rev. B 45 (1992) 13244.
H.J. Monkhorst, J.D. Pack, Phys. Rev. B 13 (1976) 5188.
G. Kresse, J. Furthmuller, Phys. Rev. B 54 (1996) 11169.
B.G. Pfrommer, M. Côté, S.G. Louie, M.L. Cohen, J. Comput. Phys. 131 (1997) 233.
P.E. Blöchl, O. Jepsen, O.K. Andersen, Phys. Rev. B 49 (1994) 16223.
P. Villars, Pearson's Handbook of Crystallographic Data, ASM International,
Materials Park, OH, 1997.
Y.F. Han, Y.B. Dai, J. Wang, D. Shu, B.D. Sun, Appl. Surf. Sci. 257 (2011) 7831.
B. Heying, R.D. Hoffmann, R. Pöttgen, Z. Naturforsch. B 60 (2005) 491.
C.M. Li, S.M. Zeng, Z.Q. Chen, N.P. Cheng, T.X. Chen, Comput. Mater. Sci. 93
(2014) 210.
Z.R. Liu, J.H. Chen, S.B. Wang, D.W. Yuan, M.J. Yin, C.L. Wu, Acta Mater 59 (2011)
7396.
G.S.H. Wang, Single Crystal Elastic Constants and Calculated Aggregate Properties:
a Handbook, MIT Press, Cambridge, 1971.
J. Li, Y.Q. Yang, X. Luo, Comput. Mater. Sci. 62 (2012) 136.
H.W. Hugosson, O. Eriksson, U. Jansson, A.V. Ruban, P. Souvatzis, I.A. Abrikosov,
Surf. Sci. 557 (2004) 243.
B.L. Bramfitt, Metall. Mater. Trans. B (1970) 1987.
M. Lipkin, D.R. Clarke, A.G. Evans, Acta Mater 46 (1998) 4835.
H.S. Abdelkader, H.I. Faraoun, C. Esling, J. Appl. Phys. 110 (2011) 044901.
G.B. Winkelman, K. Raviprasad, B.C. Muddle, Acta Mater 55 (2007) 3213.
S.B. Wang, J.H. Chen, M.J. Yin, Z.R. Liu, D.W. Yuan, J.Z. Liu, C.H. Liu, C.L. Wu,
Acta Mater 60 (2012) 6573.
D. Myers, Surfaces, Interfaces, and Colloids: Principles and Applications, second ed.,
John Wiley & Sons, Inc., New York, 1999.
K. Li, Z.G. Sun, F. Wang, N.G. Zhou, X.W. Hu, Appl. Surf. Sci. 270 (2013) 584.
Документ
Категория
Без категории
Просмотров
3
Размер файла
1 103 Кб
Теги
014, 2017, intermetal
1/--страниц
Пожаловаться на содержимое документа