1: \documentclass[abbrv,aps,prb,twocolumn,showpacs,preprintnumbers,amsmath,amssymb,
2: superscriptaddress,nofootinbib]{revtex4}
3:
4: \usepackage{graphicx}% Include figure files
5: \usepackage{dcolumn}% Align table columns on decimal point
6: \usepackage{bm}% bold math
7: % You should use BibTeX and apsrev.bst for references
8: % Choosing a journal automatically selects the correct APS
9: % BibTeX style file (bst file), so only uncomment the line
10: % below if necessary.
11: %\bibliographystyle{apsrev}
12:
13: \begin{document}
14: \title{Local correlations and hole doping in NiO}
15: \author{J. Kune\v{s}}
16: \affiliation{Theoretical Physics III, Center for Electronic Correlations and Magnetism, Institute of Physics,
17: University of Augsburg, Augsburg
18: 86135, Germany}
19: \email{jan.kunes@physik.uni-augsburg.de}
20: \affiliation{Institute of Physics,
21: Academy of Sciences of the Czech Republic, Cukrovarnick\'a 10,
22: 162 53 Praha 6, Czech Republic}
23: \author{V.~I.~Anisimov}
24: \affiliation{Institute of Metal Physics, Russian Academy of
25: Sciences-Ural Division, 620041 Yekaterinburg GSP-170, Russia}
26: \author{A.~V.~Lukoyanov}
27: \affiliation{Ural State Technical University-UPI,
28: 620002 Yekaterinburg, Russia}
29: \author{D. Vollhardt}
30: \affiliation{Theoretical Physics III, Center for Electronic Correlations and Magnetism, Institute of Physics,
31: University of Augsburg, Augsburg
32: 86135, Germany}
33: \date{\today}
34:
35: \begin{abstract}
36: Using a combination of {\it ab initio} bandstructure methods and dynamical mean-field theory
37: we study the single-particle spectrum of the prototypical charge-transfer insulator
38: NiO. Good agreement with photoemission and inverse-photoemission spectra
39: is obtained for both stoichiometric and hole-doped systems.
40: In spite of a large Ni-$d$ spectral weight at the top of the valence band the doped holes are found
41: to occupy mainly the ligand $p$ orbitals. Moreover, high hole doping
42: leads to a significant reconstruction of the single-particle spectrum accompanied
43: by a filling of the correlation gap.
44: \end{abstract}
45: \pacs{71.27.+a, 71.10.-w, 79.60.-i}
46: \maketitle
47:
48: \section{Introduction}
49: Already in 1937, at the outset of modern solid state physics, de Boer and Verwey \cite{deboer} drew attention
50: to the surprising properties of materials with incompletely filled 3$d$-bands, such as NiO. This
51: observation prompted Peierls and Mott \cite{mott} to discuss the interaction between the electrons.
52: Ever since transition metal oxides (TMOs) were investigated intensively.
53: This interest further increased when it was discovered that
54: TMOs display an amazing multitude of ordering and electron correlation phenomena, including
55: high temperature superconductivity, colossal magnetoresistance and Mott metal-insulator transitions \cite{imada}.
56: In the late 1950's MnO and NiO were taken as the textbook examples of antiferromagnets.
57: However, when the importance of local Coulomb correlations in the transition metal $d$-shell
58: was realized TMOs were considered candidates for Mott insulators \cite{mott}.
59: In the mid 1980's Zaanen, Sawatzky and Allen (ZSA) introduced their classification of
60: TMOs and related compounds into Mott-Hubbard and charge-transfer (CT) systems \cite{zaa85}.
61: In the early TMOs the ligand $p$-band is located well below the transition metal $d$-band
62: and thus plays a minor role in the low energy dynamics. Such a case, called
63: Mott-Hubbard system in the ZSA scheme, is well described by a multi-band Hubbard model.
64: On the other hand, the late TMOs belong to the CT type where the $p$-band is situated between
65: the interaction split $d$-bands. A more general Hamiltonian where
66: the $p$-states are explicitly included is then needed, which can be viewed as a combination
67: of multi-band Hubbard and Anderson lattice models.
68: A major impulse for detailed investigations of CT systems, and especially of their hole doped regime,
69: came with the discovery of high temperature superconductivity in cuprate perovskites.
70: While the standard three-band Hamiltonian for cuprates \cite{emery} contains only one $d$-orbital
71: per lattice site, the description of cubic transition metal monoxides, the prominent member of which
72: is NiO, requires the full set of $d$-orbitals. The latter are of interest not only
73: for fundamental research, but play an important role also in fields such as geophysics \cite{cohen}.
74: Furthermore, recent progress in high pressure experiments \cite{yoo}
75: made the insulator-to-metal transition in some TMOs accessible in the laboratory,
76: providing yet another stimulus for theoretical investigations.
77:
78: We report a computational study of NiO combining {\it ab initio}
79: band structure calculations in the local density approximation (LDA) with the dynamical
80: mean-field theory (DMFT), an approach known as LDA+DMFT \cite{ldadmft}.
81: By treating the local correlations and the Ni $3d$ - O $2p$ hybridization on the same
82: footing we provide a description of the full valence and conduction band spectra
83: of a CT system with strong hybridization.
84: We will show that a good quantitative agreement with
85: photoemission and inverse-photoemission data can be obtained thereby.
86: This provides a solid foundation for the
87: subsequent investigation of hole doping, a question of broader interest mainly in the context
88: of cuprates. It will be shown that the behavior of the doped holes
89: clearly reveals the CT character of NiO.
90:
91: NiO is a type II anti-ferromagnet ($T_N=523$ K) with
92: a magnetic moment of almost 2 $\mu_B$ and a large gap surviving well above $T_N$.
93: The standard LDA band theory predicts NiO to be a metal \cite{mat72},
94: or an antiferromagnetic insulator \cite{ter84} if spin polarization is allowed.
95: A severe underestimation of the gap and the magnetic moment
96: suggests, however, that the Slater antiferromagnetic state
97: obtained within LDA does not describe the true nature
98: of NiO. On the other hand exact diagonalization studies on small clusters
99: were quite successful in describing the single- and two-particle
100: spectra \cite{fuj84}, showing that the local Coulomb interactions are important.
101: This made it clear that an explicit treatment of Coulomb interactions within the
102: $3d$ shell is needed, and methods such as
103: LDA+U \cite{ani91}, self-interaction correction \cite{sva91}, or GW \cite{ary95} were introduced.
104: The static, orbitally dependent self-energy of LDA+U
105: enforces a separation of the occupied and unoccupied $d$-bands and thus opens a gap comparable
106: to experiment. This in turn leads to a significant improvement of the description of static properties
107: such as the local moment or the lattice dynamics \cite{savrasov}.
108: However, the LDA+U method is limited to an ordered state and does not
109: yield the electronic excitations and the effect of doping correctly.
110:
111: A systematic inclusion of dynamical correlations was made possible by
112: dynamical mean-field theory \cite{met89,rmp,pt}.
113: Since its introduction DMFT proved to be a powerful
114: tool for the investigation of electronic systems with strong local correlations.
115: In connection with band-structure methods, the LDA+DMFT scheme \cite{ldadmft} provides access to material
116: specific single-particle spectra as well as more general correlation functions.
117: Applications of LDA+DMFT so far were mostly limited to Mott-Hubbard systems, where
118: the ligand states are integrated out before the correlation problem is solved.
119: Recently, Ren {\it et al.} \cite{ren06} applied this approach to NiO and were able to
120: obtain a realistic gap and the near-gap spectra. However, this approach takes
121: into account only $d$-electrons, such that
122: the orbital character of the valence and conduction band are bound to be the same,
123: the high frequency incoherent features in the valence band are missing, and
124: the hole doping cannot be described realistically.
125: In this work we go beyond such limitations by working with the ligand $p$ states explicitly.
126:
127: \begin{table}
128: \caption{\label{tab:1} Orbital occupancies and the local moment on the Ni site for
129: different hole dopings.}
130: \begin{ruledtabular}
131: \begin{tabular}{c|cccc}
132: $n_h$ & $n_{e_g}$ & $n_{t_{2g}}$ & $n_p$ & $m_d$ \\
133: \hline
134: 0 & 0.547 & 1.000 & 0.969 & 1.85 \\
135: 0.6 & 0.531 & 0.994 & 0.885 & 1.61 \\
136: 1.2 & 0.530 & 0.980 & 0.800 & 1.45 \\
137: \end{tabular}
138: \end{ruledtabular}
139: \end{table}
140: \section{Computational details}
141: LDA+DMFT proceeds in two steps: (i) construction of the effective Hamiltonian
142: from converged LDA calculation, and (ii) solution of the corresponding DMFT equations.
143: Here we use the projection onto Wannier functions \cite{ani05} to obtain an eight-band $p-d$ Hamiltonian
144: \begin{equation}
145: \label{eq:ham}
146: \begin{split}
147: H=&\sum_{\mathbf{k},\sigma}\bigl(h_{\mathbf{k},\alpha\beta}^{dd}d_{\mathbf{k}\alpha\sigma}^{\dagger}
148: d_{\mathbf{k}\beta\sigma}+h_{\mathbf{k},\gamma\delta}^{pp}p_{\mathbf{k}\gamma\sigma}^{\dagger}
149: p_{\mathbf{k}\delta\sigma}+ \\& h_{\mathbf{k},\alpha\gamma}^{dp}d_{\mathbf{k}\alpha\sigma}^{\dagger}
150: p_{\mathbf{k}\gamma\sigma}+h_{\mathbf{k},\gamma\alpha}^{pd}p_{\mathbf{k}\gamma\sigma}^{\dagger}
151: d_{\mathbf{k}\alpha\sigma}\bigr)+\\&\sum_{i,\sigma,\sigma'}U_{\alpha\beta}^{\sigma\sigma'}n^d_{i\alpha\sigma}n^d_{i\beta\sigma'}.
152: \end{split}
153: \end{equation}
154: Here $d_{\mathbf{k}\alpha\sigma}$ and $p_{\mathbf{k}\gamma\sigma}$
155: are Fourier transforms of $d_{i\alpha\sigma}$ and $p_{i\gamma\sigma}$, which annihilate the $d$ or $p$ electron
156: with orbital and spin indices $\alpha\sigma$ or $\gamma\sigma$ in the $i$th unit cell,
157: and $n^d_{i\alpha\sigma}$ is the corresponding occupation number operator.
158: The elements of $U_{\alpha\beta}^{\sigma\sigma'}$ matrix are parameterized
159: by $U$ and $J$. The constrained LDA calculation yields $U$=8 eV and $J$=1 eV \cite{ani91}.
160: To account for the Coulomb interaction already present in LDA
161: we renormalize the $dd$-diagonal elements of the LDA Hamiltonian by the double counting correction
162: \begin{equation}
163: h^{dd}_{\mathbf{k},\alpha\beta}=\Tilde{h}^{dd}_{\mathbf{k},\alpha\beta}(\mathbf{k})-(N_{\text{orb}}-1)
164: \bar{U}n_{LDA}\delta_{\alpha\beta}
165: \end{equation}
166: where $n_{LDA}$ is the average LDA occupation per orbital and $N_{\text{orb}}$=10 is the total number of
167: orbitals within the shell.
168: \begin{figure}
169: \includegraphics[width=\columnwidth,clip]{total3.ps}
170: \caption{\label{fig:total}(color online) Theoretical Ni-$d$ (solid line) and O-$p$ (shaded) resolved spectral densities compared
171: to photoemission and inverse photoemission data obtained at
172: 120 eV and 66 eV photon frequencies after Ref. \onlinecite{saw84}. Gaussian broadening of 0.6 eV full width at
173: half maximum corresponding to the experimental resolution was applied to the theoretical curves.
174: The circle marks position of the $d^{10}\underline{L}$ excitation.}
175: \end{figure}
176:
177: Next we iteratively solve the DMFT equations on the Matsubara
178: contour, a key part of which is the auxiliary impurity problem solved by quantum Monte-Carlo (QMC) method \cite{qmc}.
179: The results reported here were obtained at T=1160 K.
180: To obtain the single-particle spectral functions analytic continuation to real frequencies
181: is performed using the maximum entropy method \cite{mem}.
182: Applying QMC to a gapped system requires careful assessment of ergodicity and autocorrelation issues \cite{fw}.
183: \begin{figure}
184: \includegraphics[angle=270,width=\columnwidth,clip]{resolved.ps}
185: \caption{\label{fig:resolved}(color online) Orbitally resolved O $2p$ and Ni $3d$ spectral densities
186: (offset for better resolution).
187: Substantial spectral weight transfer relative to the LDA results (see inset)
188: is observed.}
189: \end{figure}
190:
191: \section{Results and discussion}
192: \subsection{Single particle spectra}
193: The orbital occupations, shown in Table \ref{tab:1}, and the local moment of 1.85 $\mu_B$
194: obtained from the paramagnetic DMFT solution correspond to a $d^8$ groundstate of the Ni ion
195: with two ferromagnetically coupled holes of $e_g$ symmetry.
196: In Fig. \ref{fig:total} the calculated spectral densities resolved into Ni $3d$
197: and O $2p$ contributions
198: are compared to photoemission and inverse photoemission data \cite{saw84}.
199: Using the full $p-d$ Hamiltonian we are able to cover the entire
200: valence and conduction bands spectra. Features corresponding to $4s$ and $4p$
201: bands at 10 eV and 13 eV, respectively, are not included in the theoretical spectrum.
202: As shown by Eastman and Freeouf \cite{eastman}
203: the relative intensity of the $2p$ contribution increases with decreasing photon energy.
204: Therefore the 120 eV spectrum is dominated by Ni $3d$ emission, while at 66 eV photon energy
205: the O $2p$ contribution peaked around -4 eV is resolved (for a detailed
206: orbital decomposition see Ref. \cite{eastman}).
207: The theoretical spectrum very well reproduces the experimental features, including
208: the size of the gap, the $d$ character of the conduction band, the broad $d$ peak at -9 eV,
209: the position of the $p$-band, and the strong $d$ contribution at the top
210: of the valence band.
211: While the gap and the Hubbard subbands can be described already with the
212: static theory (LDA+U) \cite{ani91}, a dynamical treatment is apparently needed
213: to capture the substantial redistribution of spectral weight between the incoherent (-9 eV)
214: and resonant (-2 eV) features in the $d$ spectrum.
215: For a more detailed analysis we show in Fig. \ref{fig:resolved}
216: the spectral density resolved into $e_g$ and $t_{2g}$ representations.
217: In agreement with other studies \cite{ren06,fuj84}
218: we find the conduction band to have a pure $e_g$ character.
219: The dominant feature of the valence spectrum is a distribution of spectral
220: weight between the broad peak at high frequency and sharp peak(s) at the gap edge.
221: The origin of these structures was discussed by Fujimori {\it et al.} \cite{fuj84} in terms
222: of eigenstates of a NiO$_6$ cluster. Emitting a $d$ electron from the
223: $d^8$ state the system can end up either directly in a $d^7$ final state (broad peak)
224: or in a $d^8\underline{L}$ final state (sharp peak), with a ligand hole, due to $p-d$ electron transfer.
225: We use this picture to discuss our DMFT results.
226: A sizable $d^8\underline{L}$ resonant peak appears in the filled $t_{2g}$ band,
227: in spite of only a weak $p-t_{2g}$ hybridization. This is interpreted as being due to
228: the emission of a $t_{2g}$ electron followed by an electron transfer from oxygen
229: to the partially filled $e_g$ shell.
230: The $e_g$ spectrum exhibits even richer structure.
231: In the conduction sector we observe, besides the dominant $d^9$ peak, also
232: a tiny high frequency feature due to the $d^{10}\underline{L}$ final state, an accurate resolution of
233: which is difficult using the maximum entropy method.
234: The the resonant $d^8\underline{L}$ peak exhibits a pronounced splitting in the $e_g$ channel.
235: We ascribe this to formation of the Zhang-Rice bound state
236: \cite{zan89} at the gap edge, which was studied in detail in Ref.
237: \onlinecite{bala}.
238: Similar effect was also observed on a simpler two band p-d model
239: \cite{kun,zol99}.
240: In the present work an accurate resolution of the fine structure of
241: the resonant peak is not possible since the spin-flip terms in the
242: interaction are not included.
243: \begin{figure}
244: \includegraphics[angle=270,width=\columnwidth,clip]{dope2.ps}
245: \caption{\label{fig:dope2}(color online) Ni-$d$ and O-$p$ resolved spectral densities for a hole concentration
246: $n_h=0.6$ (offset for better resolution). The inset shows a comparison of $e_g$ spectral densities for hole
247: concentrations $n_h=0.6$ and $1.2$.}
248: \end{figure}
249:
250: \subsection{Hole doping}
251: Next we discuss hole doping of NiO. An experimental realization can be found in
252: Li$_x$Ni$_{1-x}$O studied in the doping
253: range $x$=0.02-0.4 \cite{linio}.
254: Using the Hamiltonian of the undoped system
255: the replacement of $x$ Ni$^{2+}$ ions by
256: Li$^{1+}$ ions introduces on average $n_h=x/(1-x)$ hole per Ni site.
257: As crude as this approximation may be we believe that the essential
258: physics of p-d weight transfer is captured correctly.
259: In Fig. \ref{fig:dope2} we show
260: the single-particle spectral densities for $n_h=0.6$ corresponding to $x=0.38$. There is no significant
261: difference between the $t_{2g}$ spectra in the doped and undoped cases, but the $e_g$ spectral function changes
262: significantly. Most notably the Mott gap is filled, while the Hubbard subbands are preserved as distinct features.
263: This is also observed in experiment, as shown in Fig. \ref{fig:linio2}.
264: A quite different behavior was reported in the single-band Mott insulator \cite{pru93}, where
265: the gap survives doping while the coherent peak merges with one of the Hubbard subbands.
266: Further hole doping of NiO leads to a spectral weight transfer
267: from both upper- and lower-Hubbard-subbands to the quasiparticle part of the $e_g$ spectrum
268: (see inset of Fig. \ref{fig:dope2}), which can be viewed as an enhancement
269: the itinerant character of the system reflected also in the decrease of the local moment (see Table \ref{tab:1}).
270: The $p$ spectral density suggests that much of the doped hole
271: is accommodated on the oxygen sites. That this is indeed the case can be seen from
272: the reduction of the $p$ orbital occupancy shown in Table \ref{tab:1}.
273: In an uncorrelated system doping results in a mere shift of the chemical potential
274: and the variation of the orbital occupancy is given by
275: the spectral density at the chemical potential.
276: This reasoning clearly fails in a multi-orbital correlated system such as NiO.
277: While the Ni-$d$ spectral density
278: at the top of the valence band in the stoichiometric system (or at the chemical potential
279: in the doped system) is comparable to the O-$p$ spectral density, doped holes almost exclusively
280: reside at the oxygen sites.
281: \begin{figure}
282: \includegraphics[angle=270,width=\columnwidth,clip]{LiNiO.ps}
283: \caption{\label{fig:linio2}(color online) Theoretical Ni-$d$ spectral densities for electron addition and
284: electron removal obtained for $n_h=0.6$ hole-doped NiO compared to the photoemission
285: and inverse photoemission spectra \cite{linio} of Li$_{0.4}$Ni$_{0.6}$O (the experimental
286: baseline is offset for better readability).}
287: \end{figure}
288:
289: \section{Conclusions}
290: We used the LDA+DMFT(QMC) approach to compute single-particle spectra of the prototypical CT insulator
291: NiO. By including the ligand $p$ states and the on-site Coulomb interaction within the same framework we were able
292: provide a full description of the valence band spectrum and, in particular, of the distribution
293: of spectral weight between the lower Hubbard band and the resonant peak at the top of the valence band.
294: Good agreement with the available photoemission and inverse-photoemission data was found without need
295: for adjustable parameters. Importantly, the present method allows us to study the hole doped regime
296: where we find that the doped holes are distributed mainly among the ligand sites.
297: A high hole doping leads to the filling of the correlation gap and a significant
298: transfer of the $d$-spectral weight form the incoherent part of the spectrum.
299:
300: \acknowledgements
301: We thank W. E. Pickett and R. T. Scalettar for numerous discussions at the early stage of the code development.
302: J.K. was sponsored by a Research Fellowship of the Alexander von Humboldt Foundation.
303: J.K. and D.V. acknowledge partial support by the SFB 484 of the Deutsche Forschungsgemeinschaft.
304: V.I.A. and A.V.L were supported by the Russian Foundation for Basic Research
305: under the grants RFFI-06-02-81017, RFFI-04-02-16096, and RFFI-03-02-39024,
306: and by the Netherlands Organization for Scientific Research through NWO 047.016.005.
307: A.V.L acknowledges support from the Dynasty Foundation and International
308: Center.
309:
310: \begin{thebibliography}{33}
311: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
312: \expandafter\ifx\csname bibnamefont\endcsname\relax
313: \def\bibnamefont#1{#1}\fi
314: \expandafter\ifx\csname bibfnamefont\endcsname\relax
315: \def\bibfnamefont#1{#1}\fi
316: \expandafter\ifx\csname citenamefont\endcsname\relax
317: \def\citenamefont#1{#1}\fi
318: \expandafter\ifx\csname url\endcsname\relax
319: \def\url#1{\texttt{#1}}\fi
320: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
321: \providecommand{\bibinfo}[2]{#2}
322: \providecommand{\eprint}[2][]{\url{#2}}
323:
324: \bibitem[{\citenamefont{de~Boer and Verwey}(1937)}]{deboer}
325: \bibinfo{author}{\bibfnamefont{J.~H.} \bibnamefont{de~Boer}} \bibnamefont{and}
326: \bibinfo{author}{\bibfnamefont{E.~J.~W.} \bibnamefont{Verwey}},
327: \bibinfo{journal}{Proc. Phys. Soc. }{\bibinfo{volume}{{\bf 49}, No. 4S}},
328: \bibinfo{pages}{59} (\bibinfo{year}{1937}).
329:
330: \bibitem[{\citenamefont{Mott}(1937)}]{mott}
331: \bibinfo{author}{\bibfnamefont{N.~F.} \bibnamefont{Mott}},
332: \bibinfo{journal}{Proc. Phys. Soc.} {\bibinfo{volume}{{\bf 49}, No. 4S}},
333: \bibinfo{pages}{57} (\bibinfo{year}{1937}).
334:
335: \bibitem[{\citenamefont{for a review~see M.~Imada
336: et~al.}(1998)\citenamefont{for a review~see M.~Imada, Fujimori, and
337: Tokura}}]{imada}
338: \bibinfo{author}{\bibnamefont{For a review~see M.~Imada}},
339: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Fujimori}}, \bibnamefont{and}
340: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Tokura}},
341: \bibinfo{journal}{Rev. Mod. Phys.} \textbf{\bibinfo{volume}{70}},
342: \bibinfo{pages}{1039} (\bibinfo{year}{1998}).
343:
344: \bibitem[{\citenamefont{Zaanen et~al.}(1985)\citenamefont{Zaanen, Sawatzky, and
345: Allen}}]{zaa85}
346: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Zaanen}},
347: \bibinfo{author}{\bibfnamefont{G.~A.} \bibnamefont{Sawatzky}},
348: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~W.} \bibnamefont{Allen}},
349: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{55}},
350: \bibinfo{pages}{418} (\bibinfo{year}{1985}).
351:
352: \bibitem[{\citenamefont{Emery}(1987)}]{emery}
353: \bibinfo{author}{\bibfnamefont{V.~J.}~\bibnamefont{Emery}},
354: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{58}},
355: \bibinfo{pages}{2794} (\bibinfo{year}{1987}).
356:
357: \bibitem[{\citenamefont{Cohen et~al.}(1997)\citenamefont{Cohen, Mazin, and
358: Isaak}}]{cohen}
359: \bibinfo{author}{\bibfnamefont{R.~E.} \bibnamefont{Cohen}},
360: \bibinfo{author}{\bibfnamefont{I.~I.} \bibnamefont{Mazin}}, \bibnamefont{and}
361: \bibinfo{author}{\bibfnamefont{D.~G.} \bibnamefont{Isaak}},
362: \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{275}},
363: \bibinfo{pages}{654} (\bibinfo{year}{1997}).
364:
365: \bibitem[{\citenamefont{Yoo et~al.}(2005)\citenamefont{Yoo, Maddox, Klepeis,
366: Iota, Evans, McMahan, Hu, Chow, Somayazulu, H\"ausermann et~al.}}]{yoo}
367: \bibinfo{author}{\bibfnamefont{C.~S.} \bibnamefont{Yoo}},
368: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Maddox}},
369: \bibinfo{author}{\bibfnamefont{J.-H.~P.} \bibnamefont{Klepeis}},
370: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Iota}},
371: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Evans}},
372: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{McMahan}},
373: \bibinfo{author}{\bibfnamefont{M.~Y.} \bibnamefont{Hu}},
374: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Chow}},
375: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Somayazulu}},
376: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{H\"ausermann}},
377: \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rev. Lett.}
378: \textbf{\bibinfo{volume}{94}}, \bibinfo{pages}{115502}
379: (\bibinfo{year}{2005}).
380:
381: \bibitem{ldadmft}
382: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Held}},
383: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Nekrasov}},
384: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Keller}},
385: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Eyert}},
386: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Bl\"umer}},
387: \bibinfo{author}{\bibfnamefont{A.~K.} \bibnamefont{McMahan}},
388: \bibinfo{author}{\bibfnamefont{R.~T.} \bibnamefont{Scalettar}},
389: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Pruschke}},
390: \bibinfo{author}{\bibfnamefont{V.~I.} \bibnamefont{Anisimov}},
391: \bibnamefont{and}
392: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Vollhardt}},
393: \bibinfo{journal}{phys. stat. sol. (b)} \textbf{\bibinfo{volume}{243}},
394: \bibinfo{pages}{2599} (\bibinfo{year}{2006});
395: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Kotliar}},
396: \bibinfo{author}{\bibfnamefont{S.~Y.} \bibnamefont{Savrasov}},
397: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Haule}},
398: \bibinfo{author}{\bibfnamefont{V.~S.} \bibnamefont{Oudovenko}},
399: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Parcollet}},
400: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{C.~A.}
401: \bibnamefont{Marianetti}}, \bibinfo{journal}{Rev. Mod. Phys.}
402: \textbf{\bibinfo{volume}{78}}, \bibinfo{pages}{865} (\bibinfo{year}{2006}).
403:
404: \bibitem[{\citenamefont{Mattheiss}(1972)}]{mat72}
405: \bibinfo{author}{\bibfnamefont{L.~F.} \bibnamefont{Mattheiss}},
406: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{5}},
407: \bibinfo{pages}{290} (\bibinfo{year}{1972}).
408:
409: \bibitem[{\citenamefont{Terakura et~al.}(1984)\citenamefont{Terakura, Williams,
410: Oguchi, and K\"ubler}}]{ter84}
411: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Terakura}}
412: \bibnamefont{{\it et~al.}}, \bibinfo{journal}{Phys. Rev. B}
413: \textbf{\bibinfo{volume}{30}}, \bibinfo{pages}{4734} (\bibinfo{year}{1984}).
414:
415: \bibitem[{\citenamefont{Fujimori et~al.}(1984)\citenamefont{Fujimori, Minami,
416: and Sugano}}]{fuj84}
417: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Fujimori}},
418: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Minami}}, \bibnamefont{and}
419: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Sugano}},
420: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{29}},
421: \bibinfo{pages}{5225} (\bibinfo{year}{1984}).
422:
423: \bibitem[{\citenamefont{Anisimov et~al.}(1991)\citenamefont{Anisimov, Zaanen,
424: and Andersen}}]{ani91}
425: \bibinfo{author}{\bibfnamefont{V.~I.} \bibnamefont{Anisimov}},
426: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Zaanen}}, \bibnamefont{and}
427: \bibinfo{author}{\bibfnamefont{O.~K.} \bibnamefont{Andersen}},
428: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{44}},
429: \bibinfo{pages}{943} (\bibinfo{year}{1991}).
430:
431: \bibitem[{\citenamefont{Svane and Gunnarsson}(1991)}]{sva91}
432: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Svane}} \bibnamefont{and}
433: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Gunnarsson}},
434: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{65}},
435: \bibinfo{pages}{1148} (\bibinfo{year}{1991}).
436:
437: \bibitem[{\citenamefont{Aryasetiawan and Gunnarsson}(1995)}]{ary95}
438: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Aryasetiawan}} \bibnamefont{and}
439: \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Gunnarsson}},
440: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{74}},
441: \bibinfo{pages}{3221} (\bibinfo{year}{1995}).
442:
443: \bibitem[{\citenamefont{Savrasov and Kotliar}(2003)}]{savrasov}
444: \bibinfo{author}{\bibfnamefont{S.~Y.} \bibnamefont{Savrasov}} \bibnamefont{and}
445: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Kotliar}},
446: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{90}},
447: \bibinfo{pages}{056401} (\bibinfo{year}{2003}).
448:
449: \bibitem[{\citenamefont{Metzner and Vollhardt}(1989)}]{met89}
450: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Metzner}} \bibnamefont{and}
451: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Vollhardt}},
452: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{62}},
453: \bibinfo{pages}{324} (\bibinfo{year}{1989}).
454:
455: \bibitem[{\citenamefont{Georges et~al.}(1996)\citenamefont{Georges, Kotliar,
456: Krauth, and Rozenberg}}]{rmp}
457: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Georges}},
458: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Kotliar}},
459: \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Krauth}}, \bibnamefont{and}
460: \bibinfo{author}{\bibfnamefont{M.~J.} \bibnamefont{Rozenberg}},
461: \bibinfo{journal}{Rev. Mod. Phys.} \textbf{\bibinfo{volume}{68}},
462: \bibinfo{pages}{13} (\bibinfo{year}{1996}).
463:
464: \bibitem[{\citenamefont{Kotliar and Vollhardt}(2004)}]{pt}
465: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Kotliar}} \bibnamefont{and}
466: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Vollhardt}},
467: \bibinfo{journal}{Phys. Today} \textbf{\bibinfo{volume}{57(3)}},
468: \bibinfo{pages}{53} (\bibinfo{year}{2004}).
469:
470: \bibitem[{\citenamefont{Ren et~al.}(2006)\citenamefont{Ren, Leonov, Keller,
471: Kollar, Nekrasov, and Vollhardt}}]{ren06}
472: \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Ren}},
473: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Leonov}},
474: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Keller}},
475: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kollar}},
476: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Nekrasov}}, \bibnamefont{and}
477: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Vollhardt}},
478: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{74}},
479: \bibinfo{pages}{195114} (\bibinfo{year}{2006}).
480:
481: \bibitem[{\citenamefont{Anisimov et~al.}(2005)\citenamefont{Anisimov, Kondakov,
482: Kozhevnikov, Nekrasov, Pchelkina, Allen, Mo, Kim, Metcalf, Suga
483: et~al.}}]{ani05}
484: \bibinfo{author}{\bibfnamefont{V.~I.} \bibnamefont{Anisimov}},
485: \bibinfo{author}{\bibfnamefont{D.~E.} \bibnamefont{Kondakov}},
486: \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Kozhevnikov}},
487: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Nekrasov}},
488: \bibinfo{author}{\bibfnamefont{Z.~V.} \bibnamefont{Pchelkina}},
489: \bibinfo{author}{\bibfnamefont{J.~W.} \bibnamefont{Allen}},
490: \bibinfo{author}{\bibfnamefont{S.-K.} \bibnamefont{Mo}},
491: \bibinfo{author}{\bibfnamefont{H.-D.} \bibnamefont{Kim}},
492: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Metcalf}},
493: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Suga}}, \bibnamefont{et~al.},
494: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{71}},
495: \bibinfo{pages}{125119} (\bibinfo{year}{2005}).
496:
497: \bibitem[{\citenamefont{Sawatzky and Allen}(1984)}]{saw84}
498: \bibinfo{author}{\bibfnamefont{G.~A.} \bibnamefont{Sawatzky}} \bibnamefont{and}
499: \bibinfo{author}{\bibfnamefont{J.~W.} \bibnamefont{Allen}},
500: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{53}},
501: \bibinfo{pages}{2339} (\bibinfo{year}{1984}).
502:
503: \bibitem[{\citenamefont{Hirsch and Fye}(1986)}]{qmc}
504: \bibinfo{author}{\bibfnamefont{J.~E.} \bibnamefont{Hirsch}} \bibnamefont{and}
505: \bibinfo{author}{\bibfnamefont{R.~M.} \bibnamefont{Fye}},
506: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{56}},
507: \bibinfo{pages}{2521} (\bibinfo{year}{1986}).
508:
509: \bibitem[{\citenamefont{Jarrell and Gubernatis}(1996)}]{mem}
510: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Jarrell}} \bibnamefont{and}
511: \bibinfo{author}{\bibfnamefont{J.~E.} \bibnamefont{Gubernatis}},
512: \bibinfo{journal}{Physics Reports} \textbf{\bibinfo{volume}{269}},
513: \bibinfo{pages}{133} (\bibinfo{year}{1996}).
514:
515: \bibitem[{\citenamefont{Kune\v{s}}({\natexlab{a}})}]{fw}
516: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kune\v{s}}},
517: \bibinfo{note}{unpublished}.
518:
519: \bibitem[{\citenamefont{Eastman and Freeouf}(1975)}]{eastman}
520: \bibinfo{author}{\bibfnamefont{D.~E.} \bibnamefont{Eastman}} \bibnamefont{and}
521: \bibinfo{author}{\bibfnamefont{J.~L.} \bibnamefont{Freeouf}},
522: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{34}},
523: \bibinfo{pages}{395} (\bibinfo{year}{1975}).
524:
525: \bibitem[{\citenamefont{Zhang and Rice}(1988)}]{zan89}
526: \bibinfo{author}{\bibfnamefont{F.~C.} \bibnamefont{Zhang}} \bibnamefont{and}
527: \bibinfo{author}{\bibfnamefont{T.~M.} \bibnamefont{Rice}},
528: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{37}},
529: \bibinfo{pages}{3759} (\bibinfo{year}{1988}).
530:
531: \bibitem[{\citenamefont{Ba{\l}a et~al.}(1994)\citenamefont{Ba{\l}a, Ole\'s, and
532: Zaanen}}]{bala}
533: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Ba{\l}a}},
534: \bibinfo{author}{\bibfnamefont{A.~M.} \bibnamefont{Ole\'s}},
535: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Zaanen}},
536: \bibinfo{journal}{Phys. Rev. Lett} \textbf{\bibinfo{volume}{72}},
537: \bibinfo{pages}{2600} (\bibinfo{year}{1994}).
538:
539: \bibitem[{\citenamefont{Kune\v{s}}({\natexlab{b}})}]{kun}
540: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kune\v{s}}},
541: \bibinfo{note}{unpublished}.
542:
543: \bibitem[{\citenamefont{Z\"olfl et~al.}(1999)\citenamefont{Z\"olfl, Maier,
544: Prunschke, and Keller}}]{zol99}
545: \bibinfo{author}{\bibfnamefont{M.~B.} \bibnamefont{Z\"olfl}},
546: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Maier}},
547: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Prunschke}},
548: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Keller}},
549: \bibinfo{journal}{Eur. Phys. J. B} \textbf{\bibinfo{volume}{7}},
550: \bibinfo{pages}{377} (\bibinfo{year}{1999}).
551:
552: \bibitem[{\citenamefont{van Elp et~al.}(1992)\citenamefont{van Elp, Eskes,
553: Kuiper, and Sawatzky}}]{linio}
554: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{van Elp}},
555: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Eskes}},
556: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Kuiper}}, \bibnamefont{and}
557: \bibinfo{author}{\bibfnamefont{G.~A.} \bibnamefont{Sawatzky}},
558: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{45}},
559: \bibinfo{pages}{1612} (\bibinfo{year}{1992}).
560:
561: \bibitem[{\citenamefont{Pruschke et~al.}(1993)\citenamefont{Pruschke, Cox, and
562: Jarrell}}]{pru93}
563: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Pruschke}},
564: \bibinfo{author}{\bibfnamefont{D.~L.} \bibnamefont{Cox}}, \bibnamefont{and}
565: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Jarrell}},
566: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{47}},
567: \bibinfo{pages}{3553} (\bibinfo{year}{1993}).
568:
569: \end{thebibliography}
570: \end{document}
571: