1: %% LC7763
2: %%Keywords: crystal field, low-spin state, spin-orbit coupling, Co$^{3+}$ ion
3:
4:
5: \documentstyle[aps,12pt]{revtex}
6: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
7: %TCIDATA{TCIstyle=Article/art2.lat,aps,revtex}
8:
9: %TCIDATA{OutputFilter=LATEX.DLL}
10: %TCIDATA{LastRevised=Tuesday, March 11, 2003 11:16:33}
11: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}
12: %TCIDATA{Language=American English}
13:
14: \begin{document}
15: \title{Electronic structure of the V$^{3+}$ ion in V$_{2}$O$_{3}^{\ast }$ }
16: \author{R.J. Radwanski}
17: \address{Center for Solid State Physics, S$^{nt}$ Filip 5,31-150Krakow,Poland%
18: \\
19: Institute of Physics, Pedagogical University, 30-084 Krakow, Poland}
20: \author{Z. Ropka}
21: \address{Center for Solid State Physics, S$^{nt}$ Filip\\
22: 5,31-150Krakow,Poland.\\
23: e-mail: sfradwan@cyf-kr.edu.pl; http://www.css-physics.edu.pl}
24: \maketitle
25:
26: \begin{abstract}
27: We have attributed magnetism and electronic structure of V$_{2}$O$_{3}$ to
28: the V$^{3+}$ ions. We claim that the V$^{3+}$ ion in V$_{2}$O$_{3}$ should
29: be considered as described by the quantum numbers {\it S}=1 and {\it L}=3.
30: Such quantum numbers result from two Hund's rules. The resulting electronic
31: structure is much more complex than considered up to now, but it describes,
32: e.g. the insulating ground state and the lowered magnetic moment, of 1.2 $%
33: \mu _{B}$. It turns out that the intra-atomic spin-orbit coupling is
34: indispensable for the physically adequate description of electronic and
35: magnetic properties of V$_{2}$O$_{3}$.
36:
37: Keywords: highly-correlated electron system, crystal field, V$^{3+}$ ion,
38: spin-orbit coupling, Mott insulators
39:
40: PACS: 71.70.E, 75.10.D, 75.30.Gw
41: \end{abstract}
42:
43: \pacs{71.70.E, 75.10.D, 75.30.Gw;}
44: \date{(11.03.2003)}
45:
46: \section{Introduction}
47:
48: V$_{2}$O$_{3}$ attracts the substantial scientific interest by more than 50
49: years \ \cite{1}. Despite of it there is still strong discussion about the
50: description of its properties and its electronic structure. There is a
51: long-standing controversy between a {\it S}=1 model without an orbital
52: degeneracy \cite{2} and the $S$=1/2 orbitally degenerate model of Castellani
53: et al.\cite{3}. Recently an ''orbitally degenerate spin-1 model for the
54: insulating V$_{2}$O$_{3}$'' has been proposed in Ref. \cite{4,5} whereas
55: spin-1 model with three degenerate orbitals quite recently was worked out by
56: Di Matteo \cite{6}.
57:
58: In all of these considerations it is agreed that V$_{2}$O$_{3}$ is an
59: insulating antiferromagnet with the Neel temperature of 160 K. At that
60: temperature also the low-temperature insulating state transforms to a
61: metallic state without long-range magnetic order. V$_{2}$O$_{3}$ has
62: basically the rhombohedral corundum Al$_{2}$O$_{3}$ structure. It is agreed
63: that V ions in the corundum structure sit predominantly in the oxygen
64: octahedron with some further distortions, Fig. 1 of Ref. \cite{4,5}. It is
65: also agreed that the V$^{3+}$ ion has two {\it d} electrons, i.e. that the
66: charge fluctuations are practically absent \cite{2,3,4}. The basis for all
67: theories is the description of the V$^{3+}$ ion and its electronic structure.
68:
69: The aim of this paper is to point out that the V$^{3+}$ ion in V$_{2}$O$_{3}$
70: should be described by the quantum numbers $S$=1 and $L$=3.
71:
72: \section{Theoretical outline}
73:
74: The V$^{3+}$ ion has two 3$d$ electrons. According to us these two electrons
75: form the highly-correlated atomic-like electronic system 3$d^{n}$ ($n$=2)
76: that is described by the resultant quantum numbers $S$=1 and $L$=3. Its
77: electronic structure contains 21 discrete states. According to the Quantum
78: Atomistic Solid State Theory, QUASST, these states are preserved also when
79: the paramagnetic ion becomes the full part of a solid \cite{7}. The
80: resultant quantum numbers $S$=1, $L$=3 and $^{3}F$ ground term, Fig. 1a,
81: come out from two Hund's rules (two, i.e. 1$^{o}$ the maximal $S$ and 2$^{o}$
82: the maximal $L$ for two $d$ electrons). Hund's rules reflect strong
83: intra-atomic interactions. Thus, the demand for keeping the Hund rules is
84: equivalent to an idea that the paramagnetic atom preserves much of their
85: internal atomic structure being the part of a solid. Such the atomic-like 3$%
86: d^{n}$ system interacts with the charge and spin surrounding in the solid.
87: The interaction with the charge surrounding we approximate by means of the
88: crystal-field interactions. The symmetry of the crystal-field interactions
89: depends on the local symmetry and in the corrundum structure is
90: predominantly octahedral. The calculated discrete electronic structure in
91: the octahedral crystal field, Fig. 1b, and in the presence of the
92: intra-atomic spin-orbit coupling is shown in Fig. 1c. To such electronic
93: structure we superimpose the spin-dependent interactions to account for the
94: magnetic state. The self-consistent calculations are performed similarly to
95: that presented in Ref. \cite{8} for FeBr$_{2}$. Having the electronic
96: structure, both in magnetic and paramagnetic state, we can calculate the
97: free Helmholtz energy and the resulting thermodynamical properties by means
98: of the statistical physics.
99:
100: \section{Results and discussion}
101:
102: The shown energy level scheme of the 3$d^{2}$ system in the octahedral
103: crystal field and in the presence of the spin-orbit coupling, with the
104: octahedral crystal field dominating the intra-atomic spin-orbit coupling,
105: contains 21 states. As seen from Fig. 1c the dominating octahedral crystal
106: field leaves 9 lowest states well separated from others. These 9 states are
107: spread over 60 meV and their eigenfunctions and energies determine the
108: detailed properties at room and lower temperatures. The properties at zero
109: temperature are determined by the properties of the ground state. The
110: detailed splitting depends on value of the spin-orbit coupling and on the
111: distortions.
112:
113: The ground state is an accidental non-magnetic doublet. This doublet
114: structure is easily split by means of lattice distortions and/or
115: spin-dependent interactions. They yield a singlet ground state of the V$%
116: ^{3+} $ ion in the atomic scale. Despite of the single-ion singlet ground
117: state the magnetism can easily develop as the closely lying states work as a
118: pseudo-doublet. The similar moment-induced mechanism is well known in
119: praseodymium compounds \cite{9}. As the result of local distortions the
120: local V moment can be modelled, both its value and the direction. We get a
121: value of 1.2 $\mu _{B}$ for the V$^{3+}$-ion magnetic moment in the ordered
122: state at 0 K - this value well reproduces the experimental value. We use the
123: octahedral CEF parameter $B_{4}^{0}$ = -40 K and the spin-orbit coupling $%
124: \lambda _{s-o}$=+150 K. The sign of $B_{4}$ comes out directly from the
125: calculations of the octupolar potential for the octahedral oxygen-anion
126: surroundings.
127:
128: Our approach yields in the very natural way the insulating ground state. Due
129: to strong correlations all electrons after the charge transfer, from the
130: metal atoms to oxygens, during the formation of the compound, are localized
131: at the oxygens. Within our approach we can understand why the cubic
132: perovskite structure distorts. By the distortion the system gains energy.
133: From 1937, after Jahn and Teller's work we know that system will
134: spontaneously distort to remove the accidental double degeneracy of the
135: ground state. Our understanding of 3$d$-ion compounds is close to an
136: original idea of Van Vleck from 1932, that electronic and magnetic
137: properties are largely determined by the atomic-like electronic structure.
138:
139: \section{Conclusions}
140:
141: We have attributed magnetism and electronic structure of V$_{2}$O$_{3}$ to
142: the V$^{3+}$ ions. We claim that the V$^{3+}$ ion in V$_{2}$O$_{3}$ should
143: be considered as described by the quantum numbers $S$=1 and $L$=3. Such
144: quantum numbers result from two Hund's rules. We are convinced that the
145: orbital degree of freedom, related to the orbital quantum number $L$ and the
146: intra-atomic spin-orbit coupling are indispensable for the physically
147: adequate description of electronic and magnetic properties of V$_{2}$O$_{3}$%
148: . Most up-to-now approaches employ the one-electron $t_{2g}$ (ground state)
149: and excited $e_{g}$ orbitals and next consider a molecular V-V unit. Within
150: the Quantum Atomistic Solid-State theory we have shown that the value and
151: the direction of the local magnetic moment is predominantly governed by the
152: local lattice symmetry, e.g. Ref. \cite{6}. We get a value of 1.2 $\mu _{B}$
153: for the V$^{3+}$ ion magnetic moment - this value is very sensitive to the
154: local distortions and contains the substantial orbital moment. We are
155: convinced that our description is also useful for the description of the V$%
156: ^{3+}$ ions in orthovanades LaVO$_{3}$ and YVO$_{3}$ \cite{10}. This paper
157: has been at first rejected due to ''an unusual theoretical concept'' in the
158: situation where virtually all other workers in the field over the last
159: twenty years have adopted the V-V molecular unit in the unit cell as a basis
160: for their discussion.'' For us, the atomic approach as the start for the
161: analysis of a compound is obvious owing to a presently well-established, but
162: originally very doubtful, an idea of Dalton and Davy that all solids are
163: built from atoms. So different approaches assure the useful discussion and
164: we believe that the publishing of our paper enables the open scientific
165: discussion and allows to solve the problem in the future. As far as
166: correlations are regarding our approach to V$_{2}$O$_{3}$ is in the very
167: strongly-correlated limit, though these correlations are primarily on-site
168: and atomic-like. Developing of very different theories, despite of 20 years
169: of studies and in the very last time \cite{10,11,12}, we take as indication
170: that there is very urging need for novel theoretical approaches even such
171: ''unusual'' as the one presented in this paper. However, we can add that our
172: model provides both the orbital and spin moment \cite{8,13}. We came out
173: with our approach to 3$d$-atom containing compounds in 1996 \cite{14} when
174: the orbital moment was not yet revealed experimentally. Thus, we take the
175: possibility of accounting for the experimentally-revealed orbital moment as
176: the great plus for our theoretical approach. But surely there is still a lot
177: to do in order to describe details of experimental details, but we feel that
178: the atomic-scale properties are essentially important for description of
179: magnetic and electronic properties of 3$d$-atom compounds.
180:
181: * This paper has been submitted 31.05.2002 to Strongly Correlated Electron
182: Conference in Krakow, SCES-02 getting a code MOT026. It has been presented
183: at the Conference but has been rejected with referee's arguments that ''The
184: authors should make a convincing case for their adoption of a strictly
185: atomic model in the discussion of the physical properties of the V$^{3+}$
186: ion in V$_{2}$O$_{3}$ when virtually all other workers in the field over the
187: last twenty years have adopted the V-V molecular unit in the unit cell as a
188: basis for their discussion. .... The statements that the V cation is ``in
189: the almost perfect oxygen octahedron with some further distortions'' and
190: ``The symmetry of the crystal field interaction \ldots\ is predominantly
191: octahedral'' and the caption to Fig. 1 are misleading. The oxygen
192: environment is severely distorted, which renders suspect any scheme based on
193: the disposition of atomic states in a perfect octahedral environment.'' Our
194: answer of 10.09.2002 '' for us your general objection ``why we use the
195: atomic picture'' as the start is curious as from two hundred years after
196: Davy, Dalton and others chemists all we know is that a solid is built from
197: atoms. Thus atomic starting point is simply obvious. It is not our problem
198: why other people in the field start in the very different way. Your
199: objection that we imply ``a strictly atomic scheme'' is not true. Also, that
200: we ``take no account of the milieu in which the V$^{3+}$ ions find
201: themselves in a V$_{2}$O$_{3}$ lattice'' -- is completely misunderstanding
202: of our job. We say just opposite -- the milieu, we use a more physical word
203: surroundings (both the charge and spin) is fundamentally important for the
204: realization in a solid of the given atomic-scale ground state and
205: consequently of the whole compound. It is a role of lattice distortions
206: widely discussed by us. In connection to distortions we fully agree that
207: ``The oxygen environment is severely distorted'' but each consideration in
208: literature starts from the octahedron. Very important is that our electronic
209: structure, shown in Fig. 1c, will be only modified by distortions -- the
210: number of states will be preserved. Modification means the removal of
211: degeneracies shown, the change of the energy separations and shape of
212: eigen-functions. Next, our model provides both orbital and spin moment, Refs
213: 8,13. We came out with our approach to 3$d$-atom containing compounds in
214: 1996 when the orbital moment was not yet derived experimentally. Thus, we
215: take the possibility of accounting for the orbital moment as the great plus
216: for our theoretical approach.'' did not find the understanding of the
217: SCES-02 Committee. Surely, our many-electron CEF\ electronic structure is
218: completely different from the one-electron crystal-field electronic
219: structure with two electrons put on the t$_{2g}$ orbitals generally
220: discussed in the presently-in-fashion theories of 3$d$-atom containing
221: compounds.
222:
223: The paper has been given under the law and scientific protection of the
224: Rector of the Jagellonian University in Krakow, of University of Mining and
225: Metallurgy and of Polish Academy of Sciences.
226:
227: \begin{references}
228: \bibitem{1} J.\ Spalek, A.M.\ Oles and J.M. Honig, Phys.Rev. B {\bf 28},
229: 6802 (1983); J. Spalek and P. Gopalan, Phys.Rev.Lett. {\bf 64}, 2823 (1990).
230:
231: \bibitem{2} S.Yu. Ezhov, V.I. Anisimov, D.I. Khomskii, and G.A.Sawatzky,
232: Phys.Rev.Lett. {\bf 83}, 4136 (1999).
233:
234: \bibitem{3} C. Castellani, C.R. Natoli, and J.Ranninger, Phys.Rev. B {\bf 18}%
235: , 4945 (1978); {\bf 18}, 4967 (1978); {\bf 18}, 5001 (1978).
236:
237: \bibitem{4} F. Mila, R. Shiina, F.-C. Zhang, A Joshi, M. Ma, V. Anisimov,
238: and T.M. Rice, Phys.Rev.Lett. {\bf 85}, 1714 (2000).
239:
240: \bibitem{5} R. Shiina, F. Mila, F.-C. Zhang, and T.M. Rice, Phys. Rev. B
241: {\bf 63}, 144422 (2001).
242:
243: \bibitem{6} S. Di Matteo, N. B. Perkins, and C.R. Natoli, Phys. Rev. B {\bf %
244: 65}, 054413 (2002).
245:
246: \bibitem{7} R.J. Radwanski, R. Michalski, and Z. Ropka, Acta Phys. Pol. B
247: {\bf 31}, 3079 (2000); R.J. Radwanski and Z. Ropka,
248: http://xxx.lanl.gov/abs/cond-mat/0010081.
249:
250: \bibitem{8} Z. Ropka, R. Michalski, and R.J. Radwanski, Phys. Rev. B {\bf 63}%
251: , 172404 (2001).
252:
253: \bibitem{9} R. Michalski, Z. Ropka, and R.J. Radwanski, {\it J. Phys.:
254: Condens. Matter} {\bf 12}, 7609 (2000).
255:
256: \bibitem{10} G.R. Blake, T.T.M. Palstra, Y. Ren, A.A. Nugroho, and A.A.
257: Menovsky, Phys.Rev.Lett. {\bf 87}, 245501 (2001).
258:
259: \bibitem{11} I. Chatterjee, J. Phys.: Condens. Matter {\bf 13}, 109 (2001).
260:
261: \bibitem{12} S. Di Matteo, N. B. Perkins, and C.R. Natoli, J. Phys.:
262: Condens. Matter {\bf 14}, L37 (2001).
263:
264: \bibitem{13} R.J. Radwanski, Z. Ropka, Acta Phys. Pol. A {\bf 97}, 963
265: (2000); http://xxx.lanl.gov/abs/cond-mat/0005471.
266:
267: \bibitem{14} R.J. Radwanski and Z. Ropka, {\it Relativistic effects in the
268: electronic structure for the 3d paramagnetic ions},
269: http://xxx.lanl.gov/abs/cond-mat/9907140.
270: \end{references}
271:
272: Figure caption:
273:
274: Fig. 1. Electronic structure of the highly-correlated 3$d^{2}$ electronic
275: system occurring in the V$^{3+}$ ion calculated with the octahedral CEF
276: parameter $B_{4}^{0}$ = -40 K and the spin-orbit coupling $\lambda _{s-o}$%
277: =+150 K. According to the QUASST theory such the electronic structure is
278: expected to be largely preserved in a solid. In real V$_{2}$O$_{3}$ this
279: structure is slightly modified by off-octahedral distortions and magnetic
280: interactions.
281:
282: \end{document}
283: