1: %\documentclass[prb,aps,twocolumn,showpacs,superscriptaddress,amssymb]{revtex4}
2: \documentclass[prb,aps,preprint,showpacs,superscriptaddress]{revtex4}
3: \usepackage{graphicx}
4:
5: \begin{document}
6:
7: \title{A hard metallic material: Osmium Diboride}
8:
9: \author{Z. Y. Chen}
10: \affiliation{Hefei National Laboratory for Physical Sciences at
11: Microscale,
12: University of Science and Technology of
13: China, Hefei, Anhui 230026, People's Republic of China}
14:
15: \affiliation{USTC Shanghai Institute for Advanced Studies,
16: University of Science and Technology of China,
17: Shanghai 201315, People's Republic of China}
18:
19: \author{H. J. Xiang}
20: \affiliation{Hefei National Laboratory for Physical Sciences at
21: Microscale,
22: University of Science and Technology of
23: China, Hefei, Anhui 230026, People's Republic of China}
24:
25: \affiliation{USTC Shanghai Institute for Advanced Studies,
26: University of Science and Technology of China,
27: Shanghai 201315, People's Republic of China}
28:
29:
30: \author{Jinlong Yang}
31:
32: \thanks{Corresponding author. E-mail: jlyang@ustc.edu.cn}
33:
34: \affiliation{Hefei National Laboratory for Physical Sciences at
35: Microscale,
36: University of Science and Technology of
37: China, Hefei, Anhui 230026, People's Republic of China}
38:
39: \affiliation{USTC Shanghai Institute for Advanced Studies,
40: University of Science and Technology of China,
41: Shanghai 201315, People's Republic of China}
42:
43: \author{J. G. Hou}
44: \affiliation{Hefei National Laboratory for Physical Sciences at
45: Microscale,
46: University of Science and Technology of
47: China, Hefei, Anhui 230026, People's Republic of China}
48:
49: \author{Qingshi Zhu}
50: \affiliation{Hefei National Laboratory for Physical Sciences at
51: Microscale,
52: University of Science and Technology of
53: China, Hefei, Anhui 230026, People's Republic of China}
54:
55: \affiliation{USTC Shanghai Institute for Advanced Studies,
56: University of Science and Technology of China,
57: Shanghai 201315, People's Republic of China}
58:
59: \date{\today}
60:
61:
62: \begin{abstract}
63: We calculate the structural and electronic properties of OsB$_{2}$
64: using density functional theory with or without taking into account
65: spin-orbit (SO) interaction. Our results show that the bulk modulus
66: with and without SO interaction are 364 and 365 Gpa respectively,
67: both are in good agreement with experiment (365-395 Gpa).
68: The evidence of covalent bonding of Os-B, which plays an important
69: role to form a hard material, is indicated both in charge
70: density, atoms in molecules analysis, and density of states
71: analysis. The good metallicity and
72: hardness of OsB$_{2}$ might suggest its potential application as
73: hard conductors.
74: \end{abstract}
75:
76: \pacs{71.15.Mb, 71.20.-b, 62.20.Dc, 81.05.Bx}
77:
78: \maketitle
79:
80: Hard materials are of great interest due to their
81: superior properties of higher compressional strength, thermal
82: conductivity, refractive index, and chemical stability beside
83: higher hardness.\cite{appl} Besides the two well-known super-hard
84: materials of diamond and cubic boron nitride ($c$-BN),
85: experimental and theoretical efforts are devoted to searching a
86: new class of hard meterials. Covalent materials are much better
87: candidates for high hardness than ionic compounds because
88: electrostatic interactions are omnidirectional and yield low
89: bond-bending force constants, which result in low shear
90: modulus.\cite{cov, sci} Some potential hard materials such as C3N4
91: polymorphs,\cite{c3n4} Si3N4,\cite{si3n4} BC2N,\cite{bc2n} and
92: B6O,\cite{b6o} have been proposed and investigated intensively in
93: recent years.
94:
95:
96: The pure Osmium (Os) has an exceptionally high bulk modulus
97: (395-462 GPa) while its hardness is of only 400
98: kg/mm$^{2}$.\cite{os} The high bulk modulus of Os mainly due to
99: the high valence electron density, while the low hardness is
100: related to the metallic bonds and the hexagonal close-packed (HCP)
101: crystal structure of Os. It has been found that many transition
102: metals, such as W, Ru, and Zr, which are soft in their pure
103: metals, can be converted into hard materials by combining with
104: small covalent bond-forming atoms such as boron, carbon, oxygen,
105: or nitrogen.\cite{wc1, wc2, RuO, wc3}
106: For Os related materials, OsO$_{2}$ has been predicted to be very
107: stiff,\cite{RuO} and some compounds
108: like OsB$_{0.11}$, OsB$_{2}$, and Os$_{2}$B$_{3}$ have already
109: been synthesized.\cite{osb,osb2geom}
110: %% The experiment has confirmed that the
111: %% OsB$_{2}$ has an orthorhombic lattice(space group No. 59) with the
112: %% lattice parameters a=4.6832$\pm$0.0001\AA, b=2.8717$\pm$0.0002\AA,
113: %% and c=4.0761$\pm$0.0001\AA.\cite{osb2geom}
114: Recently, Cumberland {\it et al.} studied experimentally the
115: mechanical properties of the OsB$_{2}$ compound, and showed that
116: it is an ultra-incompressible (365-395 GPa) and hard
117: ($\geq$2000kg/mm$^{2}$) material.\cite{exp,sci}
118:
119: In this brief report, we carry out density functional calculations
120: on the electronic properties and the bulk modulus of OsB$_{2}$,
121: using the highly accurate full potential linearized augmented
122: plane wave plus local orbital (APW + LO) method as implemented in
123: the Wien2k code.\cite{wien2k} The nonoverlapping muffin-tin (MT)
124: sphere radiis of 2.02 and 1.57 bohr are used for the Os and B
125: atoms, respectively. Expansion in spherical harmonics for the
126: radial wave functions are taken up to l=10. The value of
127: R$_{MT}$K$_{max}$ (the smallest muffin tin radius multiplied by
128: the maximum k value in the expansion of plane waves in the basis
129: set) which determines the accuracy of the basis set used is set to
130: $7.0$. The total Brillouin zones are sampled with 6000 k points,
131: i.e., 672 k points in the irreducible wedge of the Brillouin zone.
132: For the exchange-correlation energy functional, we use the local
133: density approximation (LDA) in our calculations. Since Os is a
134: heavy element, a fully relativistic calculation is performed for
135: core states, whereas the valence states are treated in both scalar
136: and fully relativistic fashions. The spin-orbit coupling (SOC) is
137: self-consistently added via the second variational step scheme
138: with relativistic p$_{1/2}$ local orbitals included.\cite{kune} We
139: find there is no spontaneous spin polarization in OsB$_{2}$ in the
140: framework of the local spin density approximation (LSDA) or
141: LSDA+SO methods. Hereafter for the sake of simplicity, we only
142: discussed the results from the spin-restricted calculations.
143:
144:
145: OsB$_2$ has an orthorhombic lattice (space group $Pmmn$, No. 59)
146: with the experimental lattice parameters ${\bf a}=4.6832 \pm
147: 0.0001 $, ${\bf b}=2.8717 \pm 0.0002 $, and ${\bf c}=4.0761 \pm
148: 0.0001 $ \AA. In the orthorhombic structure, two Os atoms occupy
149: the 2a Wyckoff sites and four B atoms occupy the 4f sites, as
150: shown in Fig.~\ref{fig1}. We optimize both lattice and internal
151: coordinates to get a relaxed structure for OsB$_2$.
152: The total energy per unit cell as a
153: function of volume is calculated and the results are fitted using the
154: Murnaghan equation of state to get the bulk
155: modulus.\cite{murna}
156: %% The total energy-volume curves are compared
157: %% in Fig.~\ref{fig1}(b), in the absence and in the presence of SO.
158: %% The total energy in the presence of SO is much lower than that in
159: %% the absence of SO.
160: The corresponding equilibrium lattice parameters, bulk modulus
161: from both LDA and LDA+SO methods are compared with the
162: experimental result, as shown in Table~\ref{table1}. We can see
163: that the equilibrium lattice parameters calculated both in the
164: absence and in the presence of SO are very close with the
165: experimental value. And the bulk modulus calculated in the absence
166: and in the presence of SO are 365 and 364 Gpa, respectively, both
167: are in good agreement with the experimental result (365-395
168: GPa).\cite{exp}
169:
170:
171: As discussed above, the hardness of a material is largely
172: determined by the bonding type in the system. Thus the electronic
173: properties of OsB$_2$ such as the electron density, band
174: structure, and the density of state (DOS) are calculated in the
175: absence and in the presence of SO at the equilibrium geometries.
176: We find that the electron density from the LDA calculation is
177: almost the same as that from the LDA+SO calculation. Here we show
178: the electron density for OsB$_2$ from the LDA+SO calculation in
179: Fig.~\ref{fig2}. Two charge density contour plots (one in the
180: plane including Os1, Os2, B2, and B3, and the other in the plane
181: including Os1, B1, and B2) are shown in Fig.~\ref{fig2}(a) and
182: (b), respectively. There are some electrons between Os atoms and
183: theirs neighbor B atoms, indicating covalent bonding in OsB$_2$.
184: To gain a more detailed insight into the bonding characters of
185: OsB$_2$, we plot the difference density (the difference between
186: total density and superposition of atomic densities) in
187: Fig.~\ref{fig2}(c). We can see that some electrons are transferred
188: from both B and Os atoms to B-B bonds and Os-B bonds. Clearly, two
189: neighbor B atoms form very strong covalent bond. And the bonding
190: between the Os atom and its neighbor B atoms is also covalent. The
191: fact that Os-B bond is covalent rather than ionic is also evident
192: from Bader's atoms in molecules (AIM) theory,\cite{aim} which
193: indicates there is only about 0.05 electron transferred from an Os
194: atom to B atoms. Moreover, there is some region where the charge
195: density is rather low, implying the bonding in this system is
196: directional but not metallic, as can be seen from
197: Fig.~\ref{fig2}(a) and (b). Since there are eight B atoms around
198: an Os atom, forming a three dimensional network instead of a
199: plane, the anisotropy in OsB$_2$ is not expected to be very large.
200: As discussed above, the strong covalent bonding of Os-B plays an
201: important role to form a hard material, since the highly
202: directional bonding is needed to withstand both elastic and
203: plastic deformations. OsB$_{2}$ might be potential superhard
204: material for effectively cutting ferrous metals, including steel,
205: since diamond reacts with steel producing iron carbide.\cite{sci}
206: A recent report showed that the hardness of the Fe-based alloys is
207: less than 1300kg/mm$^{2}$,\cite{Fe} which is far less than that of
208: the OsB$_{2}$($\geq$2000kg/mm$^{2}$).\cite{exp}
209:
210: Although the LDA electron density is almost indiscernible from
211: that from the LDA+SO calculation, we note that there are some
212: differences in the band structures and DOS between two results.
213: Fig.~\ref{fig3}(a) and (b) show the band structures near the Fermi
214: energy in the absence or presence of SO, respectively. Due to the
215: spin-orbit interaction, some bands are split: some bands crossing
216: each other in Fig.~\ref{fig3}(a) no longer cross in
217: Fig.~\ref{fig3}(b), and more noticeably almost all bands in the
218: direction of M-N-K are split in the LDA+SO result. Except the band
219: splitting in the presence of SO, the both band structures have
220: main features in common. Both Fig.~\ref{fig3}(a) and (b) show that
221: the bands are dispersive but not flat indicating not large
222: anisotropy in OsB$_{2}$. We note that there are several bands
223: crossing the Fermi level which are rather dispersive in the whole
224: Brillouin zone, indicating good electronic mobility in OsB$_{2}$.
225: The metallicity is uncommon in superhard materials. Most of the
226: superhard materials are insulators or semicondunctor.\cite{wc3}
227: This special character of OsB$_{2}$ should bring it special
228: application on electron conductivity. Fig.~\ref{fig4} show the
229: total and partial DOS with or without SOC which are obtained using
230: a modified tetrahedron method of Bl\"ochl {\it et al.}.\cite{tetr}
231: We can see some minor differences between the results with or
232: without SOC. Fig.~\ref{fig4}(e) also implies the good metallicity
233: due to the substaintially large total DOS at the Fermi level. From
234: the calculated partial DOS in Fig.~\ref{fig4}, one can see that
235: the electronic structure of OsB$_{2}$ is governed by a strong
236: hybridization between the Os-d and B-p states, while with a rather
237: small contribution from the Os-p and B-s states. This strong
238: hybridization of Os-d and B-p also indicates the strong covalent
239: bonding of Os-B.
240:
241: Besides the bulk modulus, we want to mention that the shear modulus is also
242: relevant for the non-cubic OsB$_{2}$ materials.
243: The shear modulus can be computed from the elastic constants.
244: Unfortunately, the force calculation when including the spin-orbit coupling has
245: not been implemented in the Wien2k package yet. As an
246: approximation, in the calculation of elastic constants, we keep
247: the relative positions of the atoms fixed at small elastic strains of
248: the lattice. Both calculations without SOC or with SOC are carried
249: out. The procedure for computing elastic constants is similar to that
250: used by Zhou {\it et al.}.\cite{OsB2_PAW}
251: For comparison, we also compute the elastic constants using VASP\cite{vasp} with
252: LDA but no SOC included.
253: In these calculations, we perform two series of calculations with
254: atomic internal coordinates fixed or relaxed, respectively.
255: Our results reported in Table~\ref{table2}.
256: We can see that the Voigt shear modulus $G$ calculated
257: using the Wien2k package (235.1
258: and 251.8 GPa when including SOC or not, respectively.) are
259: in consistent with that (253.6 GPa) calculated using VASP with fixed
260: atomic internal coordinates, but are much larger than that (197.1 GPa)
261: calculated using VASP with a relaxation of the atomic internal coordinates.
262: The results produced by VASP are in good agreement with those reported
263: by Zhou {\it et al.}.\cite{OsB2_PAW}
264: Thus we find that is not a very good approximate that
265: computing the shear modulus with atomic internal coordinates fixed.
266: The accurate shear modulus with SOC included might be computed in our future
267: works.
268:
269:
270: In conclusion, the mechanical and electronic properties of
271: OsB$_{2}$ are studied using the highly accurate full potential
272: linearized APW + LO method. The OsB$_{2}$ is found to be metallic,
273: which is uncommon in hard materials. The calculated bulk modulus
274: with and without SOC are 364 and 365 GPa respectively, which are
275: in good agreement with the experiment (365-395GPa).\cite{exp} The
276: evidence of covalent bonding of Os-B, which plays an important
277: role to form a hard material, is manifested both in the difference
278: charge density, AIM analysis, and DOS analysis. Our results
279: indicate that OsB$_{2}$ might serve as hard conductors.
280:
281:
282: % Acknowledgement.
283: This work is partially supported by the National Project for the
284: Development of Key Fundamental Sciences in China (G1999075305,
285: G2001CB3095), by the National Natural Science Foundation of China
286: (50121202, 10474087, 20533030), by the USTC-HP HPC project, and by
287: the Virtual Laboratory for Computational Chemistry of CNIC and
288: Supercomputing Center of CNIC, Chinese Academy of Sciences.
289:
290:
291: \begin{thebibliography}{99}
292: \bibitem{appl} A. G. Thornton and J. Wilks, Nature {\bf 274}, 792 (1978).
293: \bibitem{cov}J. Haines, J. M. L$\acute{e}$ger, and G. Bocquillon,
294: Annu. Rev. Mater. Res. {\bf 31}, 1 (2001).
295: \bibitem{sci} R. B. Kaner, J. J. Gilman, and S. H. Tolbert, Science {\bf
296: 308}, 1268 (2005).
297: \bibitem{c3n4} T. S. Wang, D. L. Yu, Y. J. Tian, F. R. Xiao, J. L. He, D. C. Li,
298: W. K. Wang, and L. Li, Chem. Phys. Lett. {\bf 334}, 7 (2001).
299: \bibitem{si3n4}J. He, L. Guo, D. Yu, R. Liu, Y. Tiana, and H. T. Wang, Appl. Phys. Lett. {\bf 85}, 5571 (2004).
300: \bibitem{bc2n} H. Sun, S.-H. Jhi, D. Roundy, M. L. Cohen, and S. G. Louie,
301: Phys. Rev. B {\bf 64}, 094108 (2001); Y. Zhang, H. Sun, and C.
302: Chen, Phys. Rev. Lett. {\bf 93}, 195504 (2004); Y. Zhang, H. Sun,
303: and C. Chen, Phys. Rev. Lett. {\bf 94}, 145505 (2005).
304: \bibitem{b6o} D. W. He, Y. S. Zhao, L. Daemen, J. Qian, T. D. Shen, and T. W. Zerda, Appl. Phys. Lett. {\bf 82}, 643 (2002).
305: \bibitem{os}M. Hebbache and M. Zemzemi, Phys. Rev. B {\bf 70}, 224107 (2004).
306: \bibitem{wc1} M. Lee and R. S. Gilmore, J. Mater. Sci. {\bf 17}, 2657 (1982).
307: \bibitem{wc2} D. M. Teter, Mater. Res. Soc. Bull. {\bf 23}, 22 (1998).
308: \bibitem{wc3} J. Haines, J. M. L$\acute{e}$ger, and A. Atouf, J. Am.
309: Ceram. Soc. {\bf 78}, 445 (1995).
310: \bibitem{RuO} U. Lundin, L. Fast, L. Nordstrom, B. Johansson, J. M. Wills, and O. Eriksson,
311: Phys. ReV. B {\bf 57}, 4979 (1998).
312: \bibitem{osb} L. Stuparevi$\acute{c}$ and D. $\breve{Z}$ivkovi$\acute{c}$, Journal of Thermal Analysis
313: and Calorimetry, {\bf 76}, 975 (2004).
314: \bibitem{osb2geom} R. B. JR. Roof and C. P. Kempter, J. Chem. Phys. {\bf 34}, 1473 (1962).
315: \bibitem{exp} W. C. Robert, B. W. Michelle, J. G. John, M. C.
316: Simon, H. T. Sarah, and B. K. Richard, J. Am. Chem. Soc. {\bf
317: 127}, 7264 (2005).
318: \bibitem{wien2k} P. Blaha, K. Schwarz, G. K. H. Madsen, D. Kvasnicka, and J.
319: Luitz, in WIEN2K, \emph{An Augmented Plane Wave Plus Local
320: Orbitals Program for Calculating Crystal Properties}, edited by K.
321: Schwarz (Techn. Universit$\ddot{a}$t Wien, Austria, 2001).
322: \bibitem{kune}J. Kunes, P. Novak, R. Schmid, P. Blaha and K. Schwarz, Phys. Rev. B {\bf
323: 64}, 153102 (2001).
324: \bibitem{murna}F. D. Murnaghan, Proc. Natl Acad. Sci. USA {\bf 30},
325: 244 (1944).
326: \bibitem{aim}R. F. W. Bader, \emph{Atoms in Molecules-A Quantum Theory} (Oxford
327: University Press, Oxford, 1990);
328: F. W. Klieger-Konig, R. F. W. Bader, and T. H. Tang,
329: J. Comput. Chem. {\bf 3}, 317 (1982).
330: \bibitem{Fe} M. Stoicaa, J. Eckerta, S. Rotha, Z.F. Zhangc, L. Schultza, and W.H.
331: Wang, Intermetallics {\bf 13}, 764 (2005).
332:
333: \bibitem{tetr} P. E. Bl\"ochl, O. Jepsen, and O. K. Andersen,
334: Phys. Rev. B {\bf 49}, 16223 (1994).
335:
336: \bibitem{OsB2_PAW}Z. F. Hou, cond-mat/0601216 (unpublished).
337:
338: \bibitem {vasp} G. Kresse and J. Hafner, Phys. Rev. B {\bf 47}, 558 (1993).
339: \end{thebibliography}
340: \newpage
341:
342: \begin{table}
343: \caption{Structural parameters and bulk modulus of OsB$_{2}$.}
344: \label{table1}
345: \begin{tabular}{cccc}
346: \hline\hline
347: & LDA & LDA+SO & Experiment \\
348: \hline
349: Lattice parameters(\AA):& & &\\
350: $\bf a$&4.6433&4.6581&4.6832$\pm$0.0001 \cite{osb2geom}\\
351: $\bf b$&2.8467&2.8700&2.8717$\pm$0.0002 \cite{osb2geom}\\
352: $\bf c$&4.0432&4.0560&4.0761$\pm$0.0001 \cite{osb2geom}\\
353: Bulk modulus(GPa)&365&364&365-395 \cite{exp}\\
354: \hline\hline
355: \end{tabular}
356: \end{table}
357:
358:
359: \begin{table}
360: \caption{Calculated elastic constants and Voigt shear modulus $G$ of
361: orthorhombic OsB$_{2}$. Unless otherwise stated, all results are
362: computed with atomic internal coordinates unrelaxed.
363: All values are in units of GPa. }
364: \label{table2}
365: \begin{tabular}{cccccc}
366: \hline\hline
367: & Wien2K LDA & Wien2K LDA+SO & VASP LDA & VASP LDA (relaxed) & Other VASP
368: LDA (relaxed)\cite{OsB2_PAW} \\
369: \hline
370: $C_{11}$ & 628.9& 608.5& 622.4& 597.2& 597.0 \\
371: $C_{12}$ & 194.7& 198.3& 185.8& 188.7& 198.1 \\
372: $C_{13}$ & 235.0& 220.9& 225.3& 217.5& 206.1 \\
373: $C_{22}$ & 627.8& 590.3& 624.9& 584.5& 581.2 \\
374: $C_{23}$ & 126.8& 129.9& 124.1& 164.0& 142.6 \\
375: $C_{33}$ & 923.2& 855.7& 910.4& 833.8& 825.0 \\
376: $C_{44}$ & 185.5& 175.1& 182.0& 80.2& 70.1 \\
377: $C_{55}$ & 313.5& 292.6& 311.4& 214.5& 212.0 \\
378: $C_{66}$ & 218.6& 205.9& 233.9& 209.2& 201.3 \\
379: $G$ & 251.8 &235.1 & 253.6 &197.1 &193.8 \\
380: \hline\hline
381: \end{tabular}
382: \end{table}
383:
384:
385:
386: \begin{figure}[!hbp]
387: \caption{(Color online) The crystal structure for OsB$_{2}$. The lattice vectors are
388: denoted as $\bf{a}$, $\bf{b}$, and $\bf{c}$.}
389: \label{fig1}
390: \includegraphics[width=6.0cm]{fig1.eps}
391: \end{figure}
392:
393: \begin{figure}[!hbp]
394: \caption{(Color online) (a) The valence charge density contour (in
395: e/a.u.$^{3}$) plot in the plane including Os1, Os2, B2, and B3.
396: (b) and (c) show respectively the valence charge density and
397: difference charge density contour (in e/a.u.$^{3}$) plot in the
398: plane including Os1, B1, and B2.
399: The charge density shown here is the result from
400: the LDA+SO calculation. The LDA charge density has similar character.
401: Os1, Os2, B1, B2, and B3 refer to the atoms labeled in Fig.~\ref{fig1}(a).}
402: \label{fig2}
403: \includegraphics[width=6cm]{fig2.eps}
404: \end{figure}
405:
406: \begin{figure}[!hbp]
407: \caption{(a) Band structure of OsB$_{2}$ from the LDA
408: calculation. (b) Band structure of OsB$_{2}$ from the LDA+SO
409: calculation. $\Gamma=(0,0,0)$,
410: $M=(0,0.5,0)$, $N=(0.5,0.5,0)$, $K=(0.5,0,0)$, $A=(0,0,0.5)$.}
411: \label{fig3}
412: \includegraphics[width=6cm]{fig3.eps}
413: \end{figure}
414:
415: \begin{figure}[!hbp]
416: \caption{(Color online) Partial and total DOS of OsB$_{2}$ from the
417: LDA and LDA+SO calculations.}
418: \label{fig4}
419: \includegraphics[width=7cm]{fig4.eps}
420: \end{figure}
421:
422: \end{document}
423: