Phase Stability and Site Preference of Tb-Fe-Co-V Compounds

The effect of cobalt on the structural properties of intermetallic Tb3Fe27.4−xCoxV1.6 with Nd3(Fe,Ti)29 structure has been studied by using interatomic pair potentials obtained through the lattice inversion method. Calculated results show that the preferential occupation site of the V atom is found to be the 4i(Fe3) site, and Fe atoms are substituted for Co atoms with a strong preference for the 8j(Fe8) site. The calculated lattice constants coincide quite well with experimental values. The calculated crystal structure can recover after either an overall wide-range macrodeformation or atomic random motion, demonstrating that this system has the stable structure of Nd3(Fe,Ti)29. All these prove the effectiveness of interatomic pair potentials obtained through the lattice inversion method in the description of rare-earth materials.


Introduction
In 1993, it had been first found that the structure of the compound Nd 2 Fe 19− Ti suggested to be a Nd 3 (Fe,Ti) 29type structure with monoclinic symmetry [1]. After lots of investigation and disputation [2][3][4], it indicates that the Space Group (S.G.) of Nd 3 (Fe,Ti) 29 -type structure belongs to A2/m rather than P2 1 /c.
As well known, the substitution of cobalt for iron in rare-earth-transition metal intermetallics has a remarkable effect on their magnetic properties. Comparing with iron, cobalt has a different electronic structure and therefore lead to a different local anisotropy. Hence, the structure and the magnetic properties of 3 : 29 compounds are strongly affected by the substitution of iron by cobalt. However, when more than 40% of substitution of iron by cobalt atoms occurs, a large number of stabilizing atoms are demanded. If not, a disordered modification of hexagonal Th 2 Ni 17 -type structure will be formed [10,11,13,14].
So far, there are some investigations on the Tb-Fe-Co-V and Tb-Fe-Co-Cr systems [9,[11][12][13][14]. For Tb 3 Fe 29− V compounds, in the temperature range between 5 K and 200 K, a FOMP occurs [15]. The spin reorientation has been observed in the Tb-Fe-Co-V system, and the Curie temperature and saturation magnetisation decrease with the increase of Co content [13].
In this paper, our investigation focuses on the phase stability and site preference of Co in Tb 3 Fe 27.4− Co V 1.6 ( = 0.0, 0.1, 0.2, 0.3 and 0.4) compounds. The interatomic pair potential are obtained by Chen's lattice inversion method, whose theory is exhibited in the second part. The comparison of the calculated results with the experimental data is shown 2 The Scientific World Journal in Section 3. And the fourth part is the conclusion and discussion.

Methodology
Chen et al. proposed a concise inverse method [16][17][18][19][20][21] based on the modified Möbius inversion in number theory, which can be applied to obtain the photon density of states [16], to solve the inverse blackbody radiation problem for remote sensing [16], to unify the Debye and Einstein approximations in a general mathematics system [20], and to extract the interatomic pair potentials from ab initio calculated cohesive energy curves in pure metals and intermetallic compounds [17-19, 22, 23], metal/ceramic interface [24], and metal/oxide interface [25], as well as in carbides with complex structures [26][27][28], with high convergence speed.
Here, we take a single element crystal as an example to explain how to use Chen's lattice inversion theorem to obtain the interatomic pair potential from the first-principle cohesive energy curve. Suppose that the crystal cohesive energy can be expressed as the sum of interatomic pair potentials where is the nearest neighbor distance, ⃗ is the position vector of the th atom in the lattice, 0 ( ) is the th nearest neighbour distance with 0 (1) = 1, and 0 ( ) is the th coordination number. Obviously, { 0 } is a totally ordered set but, in most cases, not a multiplicative semigroup. In this case, we need to extend it to a totally ordered multiplicative semigroup { } with (1) = 0 (1) = 1. Namely, for any ( ), ( ) ∈ { }, we have ( ) ( ) ∈ { }. Now ( ) can be rewritten as where the extended coordination number ( ) satisfies the following rule: It is clear that { ( )} is uniquely determined by crystal geometrical structure. Now we can extract pair potential from (2) as where ( ), = 1, 2, . . . , ∞, and any ( ) satisfying ( ) | ( ), { ( )} can be determined by Pair potential (eV) where is the Kronecker delta. It is easy to prove that (5) has and only has a single solution set { ( )}, which, like {( )}, is uniquely determined by crystal geometrical structure, not related to element category. This means that we can take advantage of number theory to extract the interatomic pair potentials form cohesive energy curve(s). At this stage, modified Möbius inversion can be adopted to extract the pair potential curve data and then fit these data on the basis of Morse function [29] to determine the potential parameters. Consider where indicates the distance between two atoms and 0 is the equilibrium distance, with 0 and being parameters without unit. For the readers' convenience, several important potential parameters are shown in Figure 1.

Calculated Results
In this paper, energy minimization is carried out using a conjugate gradient method. The cut-off radius is 14Å. In order to reduce statistical fluctuation, we take the super-cell containing 1280 atoms, (Tb 3 Fe 27.4− Co V 1.6 ) 40 , for simulation.
There are no reports in the literature on the existence of the binary structure Tb 3 Fe 29 , which can be seen as the prototype of R 3 (Fe,T) 29 . In the calculation procedure, the initial lattice constants of Tb 3 Fe 29 are randomly chosen in a certain range. Under the control of the interatomic pair potentials, the energy minimization is carried out. After repeated relaxation, the space group maintains A2/m, the atomic site occupation is similar to that of Nd 3 (Fe,Ti) 29 , and the structure will be stabilized with lattice constants    Table 1). The randomness of the initial structure in a certain range and the stability of the final structure illustrate that Tb 3 Fe 29 has the topological invariability with respect to the existing Nd 3 (Fe,Ti) 29 structure. Hence, it furnishes convincing evidence that the interatomic pair potentials are reliable for the study of structural material characteristics.
Substitute the atoms of ternary element vanadium for the randomly chosen iron atoms at a certain lattice site, and then make the lattice relaxation. In this calculation procedure, energy minimization is taken as a criterion of the stability. The cohesive energy of Tb 3 Fe 29− V on the content of ternary addition is evaluated and illustrated in Figure 2. The values of energy in the figure are statistically averaged over the calculations of 20 samples. Figure 2 shows that the cohesive energy is lower when V atoms are substituted for Fe atoms at the 4i(Fe3) site than the other sites, so V should prefer the 4i (Fe3)   values [13], which are shown in Table 2. Substitute the V atoms for a randomly selected part of Fe atoms at the 4i(Fe3) sites, thus forming the (Tb 3 Fe 27.4 V 1.6 ) 40 . Then make use of the conjugate gradient method to minimize the system energy, as an approximation of a practical relaxation process. Furthermore, we take the (Tb 3 Fe 27.4 V 1.6 ) 40 structure with each atom random shifted 0.6Å from their equilibrium position to test the structural stability. Each atom in a disturbed cell can recover its equilibrium position under the interaction of interatomic pair potentials. The results show that the lattice constants are in good agreement with the experimental data [13], still retaining = 10.5433Å, = 8.4647Å, = 9.7196Å, and = 97.0681 ∘ (Table 3). Thus, the stability of the lattice and the effectiveness of the interatomic pair potentials are verified. In Tb 3 Fe 27.4− Co V 1.6 compounds, substitute the atoms of quaternary element cobalt for the randomly chosen iron atoms at a certain lattice site, and then make the lattice relaxation. Based on the ternary compounds Tb 3 Fe 27.4 V 1.6 and the site occupation of cobalt in quaternary system Tb 3 Fe 27.4− Co V 1.6 has been investigated. The dependences of the cohesive energy of Tb 3 Fe 27.4− Co V 1.6 on the content of quaternary addition are evaluated and illustrated in Figure 3. It is shown that the calculated cohesive energy increases least significantly while cobalt atoms occupy 8j(Fe8) site. Therefore, Fe atoms are substituted for Co atoms with a strong preference for the 8j(Fe8) site. The calculated lattice constants are compared with the experimental values [13], which are shown in Table 2. It states clearly that the lattice parameters , (but not ) of the compounds are decreased with increasing cobalt concentration due to the fact that the iron atoms are replaced by the smaller cobalt atoms.

4
The Scientific World Journal   The stability of the calculated structure is further checked through molecular dynamics. The cell constants are traced to higher temperatures, as shown in Table 4. Using MD (molecular dynamics) NPT ensemble, with = 1 atm, = 0.001 ps, dynamic simulations for Tb 3 Fe 27.4− Co V 1.6 are carried out at temperatures of 300, 500, and 700 K. The lattice constants change very little with respect to temperature variation, thus the structural stability is again verified. The comparisons of potential energy and kinetic energy at different temperatures are shown in Table 5. It can be seen from the table that both the values of potential energy and kinetic energy increase with the increasing temperature, but the contribution of interatomic potentials to the internal energy of the system is much larger than that from thermal motion energy. Therefore, the crystal structure at different temperature is basically determined by the interatomic pair potentials.

Conclusions
In this paper, we have made an investigation of the phase stability and effect of cobalt atoms element on the structural properties of Tb 3 Fe 27.4− Co V 1.6 crystals. The energies of the Tb-Fe-Co-V systems are calculated from these effective interatomic pair potentials. The calculation results show that vanadium atoms preferentially occupy 4i(Fe3) site, and cobalt prefer the 8j(Fe8) site. Besides, due to the fact that iron atoms are replaced by the smaller cobalt atoms, the lattice parameters , (but not ) of the compounds decreased with increased concentration.
Above all, with the interatomic potentials from lattice inversion, the structural properties have been well reproduced and are in agreement with the experimental data. This suggests that the potentials are successfully used to calculate the structural properties of rare-earth compounds The Scientific World Journal 5