1: % ****** Start of file template.aps ****** %
2: %%
3: %%
4: %% This file is part of the APS files in the REVTeX 4 distribution.
5: %% Version 4.0 of REVTeX, August 2001
6: %%
7: %%
8: %% Copyright (c) 2001 The American Physical Society.
9: %%
10: %% See the REVTeX 4 README file for restrictions and more information.
11: %%
12: %
13: % This is a template for producing manuscripts for use with REVTEX 4.0
14: % Copy this file to another name and then work on that file.
15: % That way, you always have this original template file to use.
16: %
17: % Group addresses by affiliation; use superscriptaddress for long
18: % author lists, or if there are many overlapping affiliations.
19: % For Phys. Rev. appearance, change preprint to twocolumn.
20: % Choose pra, prb, prc, prd, pre, prl, prstab, or rmp for journal
21: % Add 'draft' option to mark overfull boxes with black boxes
22: % Add 'showpacs' option to make PACS codes appear
23: % Add 'showkeys' option to make keywords appear
24:
25: %\documentclass[aps,prl,preprint,groupedaddress,showpacs]{revtex4}
26: %\documentclass[aps,prl,preprint,superscriptaddress]{revtex4}
27: \documentclass[aps,prl,twocolumn,groupedaddress,showpacs]{revtex4}
28:
29: \usepackage{graphicx}
30:
31: % You should use BibTeX and apsrev.bst for references
32: % Choosing a journal automatically selects the correct APS
33: % BibTeX style file (bst file), so only uncomment the line
34: % below if necessary.
35: %\bibliographystyle{apsrev}
36:
37: \begin{document}
38:
39: % Use the \preprint command to place your local institutional report
40: % number in the upper righthand corner of the title page in preprint mode.
41: % Multiple \preprint commands are allowed.
42: % Use the 'preprintnumbers' class option to override journal defaults
43: % to display numbers if necessary
44: %\preprint{}
45:
46: %Title of paper
47: \title{Strong Correlations Produce the Curie-Weiss Phase of Na$_{x}$CoO$_2$}
48:
49: % repeat the \author .. \affiliation etc. as needed
50: % \email, \thanks, \homepage, \altaffiliation all apply to the current
51: % author. Explanatory text should go in the []'s, actual e-mail
52: % address or url should go in the {}'s for \email and \homepage.
53: % Please use the appropriate macro foreach each type of information
54:
55: % \affiliation command applies to all authors since the last
56: % \affiliation command. The \affiliation command should follow the
57: % other information
58: % \affiliation can be followed by \email, \homepage, \thanks as well.
59: \author{Jan O. Haerter, Michael R. Peterson, B. Sriram Shastry}
60: %\email[]{Your e-mail address}
61: %\homepage[]{Your web page}
62: %\thanks{}
63: %\altaffiliation{}
64: \affiliation{Physics Department, University of California, Santa Cruz, CA 95064}
65:
66:
67: %Collaboration name if desired (requires use of superscriptaddress
68: %option in \documentclass). \noaffiliation is required (may also be
69: %used with the \author command).
70: %\collaboration can be followed by \email, \homepage, \thanks as well.
71: %\collaboration{}
72: %\noaffiliation
73:
74: \date{\today}
75:
76: \begin{abstract}
77: Within the $t$-$J$ model we study several experimentally accessible properties of the
78: 2D-triangular lattice system Na$_x$CoO$_2$, using a numerically exact
79: canonical ensemble study of 12 to 18 site triangular toroidal clusters as well as the
80: icosahedron. Focusing on the doping regime of $x\sim0.7$, we study
81: %the temperature dependent chemical potential, specific heat,
82: the temperature dependent specific heat,
83: magnetic susceptibility and the dynamic Hall coefficient
84: $R_H(T,\omega)$ as well as the magnetic field dependent thermopower.
85: We find a crossover between two phases near $x \sim 0.75$
86: in susceptibility and field suppression of the thermopower arising from
87: strong correlations.
88: An interesting connection is found between the
89: temperature dependence of the diamagnetic susceptibility and the
90: Hall-coefficient. We predict a large thermopower enhancement, arising
91: from {\em transport corrections} to the Heikes-Mott formula, in a
92: model situation where the sign of hopping is reversed from that
93: applicable to Na$_x$CoO$_2$.
94: \end{abstract}
95:
96: % insert suggested PACS numbers in braces on next line
97: \pacs{72.15.Jf, 65.90.+i, 71.27.+a}
98: % insert suggested keywords - APS authors don't need to do this
99: %\keywords{}
100:
101: %\maketitle must follow title, authors, abstract, \pacs, and \keywords
102: \maketitle
103:
104: The physics of the 2-dimensional triangular lattice system sodium
105: cobaltate Na$_x$CoO$_2$ (NCO) is fascinating\cite{nco_general}, combining strong electron
106: correlations and thermoelectric physics. A Curie-Weiss metallic phase
107: for dopings $x\sim0.7$ has been reported\cite{ong_nature} where
108: %an unusual mix of
109: %behaviors, between good metals and insulators, challenges theory
110: %severely.
111: the observed physical variables display an unusual mix of behaviors
112: that are hybrid between those of good metals and of insulating
113: systems challenging theory severely.
114: The thermopower of NCO is nearly ten times higher than
115: expected from typical metals; generating excitement in the engineering
116: and material science communities, particularly applied to
117: thermoelectric devices. Here we show that strong electron
118: correlations, along with the geometrically frustrated lattice of NCO,
119: hold the key to explaining this mysterious state of matter. We examine
120: several experimentally accessible properties of NCO within the $t$-$J$
121: model\cite{kumar,baskaran,lee}. We find that strong electron correlations capture the essential
122: physics and our results compare well with experiment.
123:
124: In NCO the low spin Cobalt ion valence
125: fluctuates between a Co$^{4+}$ (spin $1/2$) and a Co$^{3+}$
126: (spin $0$) configuration; the number of Co$^{3+}$ states is precisely
127: $x$. The Co ions form a triangular lattice, and
128: photoemission\cite{hasan,yang_1} is consistent with a single, hole-like, band with
129: hopping $t<0$ and $n=1+x$ electrons satisfying the Luttinger volume count.
130:
131:
132: The $t$-$J$ model describes strongly correlated electron systems by forbidding double
133: occupancy of lattice sites. We apply this model to NCO after an electron-hole transformation, requiring $t\rightarrow -t$ and hole doping $|1-n|$. A non-zero $J$ couples nearest neighbor
134: electrons via their spin degree of freedom. For such
135: strongly correlated systems perturbation theory is doomed to failure from the
136: outset and we make progress through numerical exact diagonalization
137: on systems containing 12, 14, and 18 sites on toroidal clusters (periodic
138: boundary conditions (BC)) and on ladder clusters (open BC in one
139: direction)\cite{long_paper}. Thermodynamics is considered within
140: the canonical ensemble.% (see Methods).
141:
142: The Hilbert spaces of these finite systems
143: are very big (up to $\sim$ 80,000 states) and grow exponentially with the number of sites.
144: Therefore, all available symmetries are used to reduce the dimension of
145: the matrices that arise to large but manageable proportions.
146: However, Peierls phase factors \cite{sss} are needed to
147: describe an applied magnetic field, which breaks or reduces the translational
148: invariance, thereby limiting us somewhat. By using a judicious
149: choice of the BC and of phases on
150: bonds\cite{kohmoto}, we achieve a fairly small non-zero flux per
151: plaquette of $\pi/N_f$; where $N_f$ is the total number of
152: triangular faces on the lattice\cite{long_paper}. The ladder
153: systems, however, enable an infinitesimal flux to be chosen.
154:
155: For NCO, photoemission supports a value for the hopping of $t=-100$ K and we adopt it in this work. This value is suggested by the ARPES data\cite{hasan} on the loss
156: of coherence of the quasiparticles as well as the dispersion in the composition range $x\sim .7$. The $T$ dependence of the chemical potential $\mu(T)-\mu(0)$ is another route to estimating $t$ \cite{footnote1}.
157:
158: %{\em
159: %Jan: Pl create a footnote with this below: I dont know what to do with your bibtex:
160: %
161: %Footnote [ At $x\sim0.8$, the data obtained in Ref\cite{fujimori} suggest a somewhat larger value $t\sim -300^0$K, but this value conflicts somewhat with the thermopower data of Ref\cite{terasaki}, taken to measure $\mu(T)$ by the Heikes Mott formula $S_{HM}$.]
162: %}
163:
164:
165: Fig. 1(a) shows the electronic specific heat $C_v(T)$, and
166: is compared with that for non-interacting electrons with the same
167: hopping. We find that the effect of correlations is a shift in the
168: peak to a smaller temperature and suppression of its overall
169: weight. This is expected since the Gutzwiller projection in the $t$-$J$
170: model reduces the number of available states and hence the
171: entropy. Due to a finite system size induced gap in the spectrum, we expect an
172: exponential behavior of $C_v(T)$ for $T\leq 20$ K, the typical gap
173: value. Taking this into account, we are able to extract the linear
174: electronic contribution $\gamma T$. The value of $\gamma$ is enhanced
175: by $\sim 1.5$ over the non-interacting value; this enhancement
176: depends only slightly on $J$ (neglecting the exponential increase at $T<20K$ due to the finite-system-induced gap) and varies with system size by $0.2$. $J=40$ K (i.e., $0.4|t|$), which is fixed by the experimental system through a comparison of the Curie-Weiss temperature with computations\cite{ong_nature, long_paper}.
177:
178: \setlength{\unitlength}{1cm}
179: \begin{figure}
180: \begin{picture}(0,0)(0,0)
181: \put(-.6,8.2){{\large (a)}}
182: \put(-.6,6.5){{\large (b)}}
183: \put(-.6,5.3){{\large (c)}}
184: \put(-.6,4.1){{\large (d)}}
185: \put(-.6,2.8){{\large (e)}}
186: \put(-.6,1.5){{\large (f)}}
187: \end{picture}
188: \includegraphics[width=7cm]{fig1.ps}
189: \caption{(color online) Specific heat and susceptibility.
190: (a), specific heat $C_v(T)$ for $x=0.72$, computed on the 18-site cluster,
191: comparison of $J=0$ (bottom) with $J/|t|=0.4$ (middle) and bare (Hubbard $U=0$)
192: specific heat (top), dotted straight lines show linear fits and
193: $\gamma$ values for $J=0$ and $U=0$ in units of $mJ/(mol\cdot$K$^2$). (b)-(f),
194: susceptibility $\chi(T)$ for dopings around $x \sim 0.7$. The dotted curves
195: indicate the bare susceptibility, and arrows indicate the evolution of
196: $J/|t|$ from $0$ to $0.5$ in steps of $0.1$ (red to yellow). Note the change of scale
197: in different panels. These results combine two different clusters, a
198: 12-site torus ($x=0.58$, $0.67$, $0.75$) and a 14-site torus
199: ($x=0.71$, $0.79$). The difference in $x=0.71$ and $x=0.75$
200: shows that $\chi(T)$ transitions from a Curie-Weiss to Pauli
201: paramagnetic behavior in this range.}
202: \end{figure}
203:
204: In Fig. 1(b)-(f)
205: we present the spin susceptibility $\chi(T)$ for several dopings
206: around $x\sim0.7$. In the
207: band limit $x\rightarrow 1$, as in the upper two panels (as well as
208: results not shown), we find the
209: expected weakly T dependent but $J$ insensitive Pauli paramagnetic
210: behavior. When $x$ is lowered below $x=0.75$ (bottom three panels),
211: $\chi(T)$ shows strong Curie-Weiss-like $T$ and $J$ dependence, and is
212: significantly renormalised from the non-interacting value at low
213: $T$. This indicates a crossover to the strong-correlation induced
214: local moment behavior for $x < 0.75$ which closely parallels
215: experimental findings\cite{ong_nature}. In this Curie-Weiss
216: phase, the behavior at high $T$ is described by the
217: Curie-Weiss form
218: $\chi(T)=\frac{1}{3}\frac{1}{v}\frac{\mu_B^2p_{eff}^2}{k_B(T-\theta)}$
219: with a negative Weiss temperature $\theta$ and effective magnetic
220: moment $p_{eff}$, $v=V/N$ is the unit cell volume. When continuing
221: the analysis to $x\rightarrow 0$, antiferromagnetic (AFM) correlations increase and we find
222: that $\theta(x,J)=-c J_{eff}(x)$ where $J_{eff}(x)= J(1+c'' x
223: |t|) + c' x |t|$, with $c= 4.0$, $c'=0.01425$, and $c''=-0.9175$. The $c'$
224: term originates in the kinetic antiferromagnetism of the frustrated
225: lattice\cite{counter_nagaoka}, and signifies that even in the absence
226: of $J$, there is a tendency for AFM order, i.e., in a direction
227: opposite to the usual Nagaoka mechanism for the square
228: lattice\cite{anderson_nagaoka}.
229:
230: Experimentally, the Hall
231: coefficient of NCO is remarkable in many respects. Most striking is
232: the unbounded linear increase with temperature of the Hall coefficient
233: $R_H$. To understand this we perform the brute force exact summations
234: of Kubo's formulae for various conductivities\cite{sss} by
235: introducing a level width, i.e., a broadening $\omega \rightarrow
236: \omega + i \eta$ with $\eta$ equal to the mean energy level
237: spacing. In addition, we evaluate the high frequency limit\cite{kumar} of
238: $R_H$ (called $R^*_H$) for all $T$. Recall that the high $T$
239: estimates of $R^*_H$ led to a prediction\cite{kumar,sss} of the
240: linear T dependence of the Hall constant for NCO, which was
241: successfully verified\cite{wang_hall}. We are thus able to provide a
242: purely theoretical benchmarking of this idea as well, subject of
243: course to the limitations of the finite size clusters.
244:
245: Focusing on the region of doping around $x\sim 0.7$ Fig. 2(a)
246: shows the Hall coefficient as a function of
247: temperature and frequency. We find that the Hall coefficient is
248: relatively insensitive to frequency, in keeping with the original
249: expectations\cite{shastry_1,kumar,sss}. All curves show a minimum
250: near $T=100$ K, and an unbounded linear increase for $T>200$ K. The slope of
251: $R_H^*$ as found in the clusters is in agreement with results from
252: high-temperature expansions\cite{kumar}. The experimental curve,
253: unlike theory, has a change of sign and also a pronounced minimum at
254: $T\sim 100$ K. We cannot reproduce the change of sign. While we could
255: fit the high $T$ slope more accurately, it requires a smaller value
256: of hopping $t \sim 35$ K, as already noted\cite{kumar}. Such a
257: choice would make most other variable fits less sensible. Hence we
258: conclude that while the data\cite{wang_hall} is largely as
259: expected from theory, in detail it is still not possible to reconcile
260: the change in sign as well as the magnitude of the slope with theory.
261: Therefore, the Hall coefficient still offers a considerable challenge
262: to both theorists and experimentalists.
263:
264: In an alternate effort to understand further the data\cite{wang_hall}
265: at lower $T\sim 100$ K, we note that the curvature
266: of the (Landau) diamagnetic susceptibility $\chi_{d}$ (obtained by
267: inserting Peierls phases) and the Hall constant are curiously
268: related in our computation via
269: \begin{equation}\label{chiL_RH_eq}
270: T\partial^2\chi_d/\partial T^2=f(x)\partial^2 R_H/\partial T^2\;,
271: \end{equation}
272: where $f(x)$ is a
273: function depending on doping $x$. We integrate equation (\ref{chiL_RH_eq})
274: to arrive at the Hall
275: constant from our $\chi_{d}$, yielding a good overall fit of
276: data. The problem with the slope and the negative
277: intercept at low $T$ are forgiven in this approach since one fits the
278: two constants, but it does capture the ubiquitous minimum at $T \sim
279: 100$ K, with a slight $J$ dependence\cite{long_paper}. While the
280: Landau diamagnetism is very small compared to the paramagnetic
281: susceptibility for narrow band systems, it might yet be accessible to
282: experiments. This is so since it is anisotropic in space (being
283: planar), in contrast to the isotropic Pauli susceptibility, and hence
284: torque magnetometry\cite{torque} could help disentangle these
285: terms.
286:
287: \begin{figure}
288: \begin{picture}(0,0)(0,0)
289: \put(-.7,7.){{\large (a)}}
290: \put(-.7,3.6){{\large (b)}}
291: \put(-.7,1.5){{\large (c)}}
292: \end{picture}
293: \includegraphics[width=6.7cm]{fig2.ps}
294: \caption{(color online) Hall coefficient and thermopower.
295: (a), Comparison of several results for the Hall coefficient $R_H(T)$ at $x=0.75$ with experiment in ref. \cite{wang_hall} at $x=0.71$
296: (red squares): $R_H^*$ (blue dot-dash), $R_H({\omega}=0)$ (blue dashed), $R_H$
297: (orange solid) derived from $\chi_d$ (at $x=0.83$, ladder), the dc-limit
298: required a broadening of the frequency $\omega\rightarrow
299: \omega+i\eta$ with $\eta\approx 3|t|$ to eliminate finite-size
300: artifacts. All results are for $12$-site clusters and $J/|t|=0$.
301: (b), Infinite frequency thermopower $S^*$ vs. $T$ for a $12$-site torus at $x=0.75$ and $x=0.67$.
302: The solid black and
303: dashed blue lines correspond to $J/|t|=0$, and $0.4$ at $x=0.75$, respectively, while
304: the solid orange and dashed-dotted red lines correspond to $J/|t|=0$, and $0.4$ at $x=0.67$,
305: respectively. $S^*(T)$ for $t=-100$ K relevant for NCO. The diamonds
306: and stars represent measured thermopower for NCO at $x=0.68$ from ref.~\cite{ong_nature}
307: and ref.~\cite{terasaki}. (c), Our prediction for $S^*(T)$ for the case
308: when the sign of the hopping is reversed ($t=100$ K.)}
309: \end{figure}
310:
311: The thermopower of NCO is striking in its large
312: magnitude\cite{terasaki} $\sim 100\mu V$/K and also in its surprising sensitivity to an
313: applied magnetic field\cite{ong_nature}. A formulation for the
314: thermopower has been recently given\cite{shastry_1} in the high
315: frequency limit in the same spirit as the Hall constant. We note
316: that the thermopower can be written as a sum of a transport and the
317: Heikes-Mott term as $S(\omega,T)= S_{Tr}(\omega,T)+ S_{HM}(T)$, where the
318: Heikes-Mott term is $S_{HM}(T)= \frac{ \mu(0)- \mu(T)}{q_e T} $, and
319: the transport term is
320: \begin{equation}
321: S_{Tr}^*(T)=\lim_{\omega \rightarrow \infty} S_{Tr}(\omega,T) = \frac{q_e
322: (\Delta(T)-\Delta(0))}{T \langle \tau^{xx} \rangle}.
323: \end{equation}
324: Here $q_e = -|e|$ is the electronic charge and $\hat{\tau}_{xx}$ is
325: the diagonal part of the stress tensor\cite{sss,shastry_1}. The term
326: $S_{HM}$ is entropic in origin\cite{beni_chaikin}. Detailed
327: expressions for $\Delta$ as an expectation of a many-body operator in
328: the $t$-$J$ model are given in equation (83) of ref.~\cite{shastry_1}, and in a
329: longer work\cite{long_paper}.
330:
331: The main approximation in ref.~\cite{shastry_1} is to use the high
332: $\omega$ limit of $S(\omega,T)$ (called $S^*$), and is expected to be numerically quite
333: reasonable, in parallel to the behavior of the Hall constant reported
334: in this work. At low $T$ the two contributions to $S$ vanish
335: separately as we have written them, and in general we find that
336: for $t<0$ (the case of NCO) the entropic part is by far the dominant term. For the
337: opposite case ($t>0$) the transport term comes into play in a dominant
338: way, and leads to very interesting behavior as we shall show.
339:
340: We first discuss the thermopower in the absence of magnetic field.
341: Fig. 2(b) shows the $T$ and $J$
342: dependence of $S^*$ for $x$ relevant to the Curie-Weiss phase. The
343: Heikes-Mott term dominates over the transport term and the
344: frequency dependence of the thermopower (evaluated via Kubo formulae)
345: is found to be quite weak in this range of doping for $t<0$ relevant to NCO\cite{long_paper}, and thus the high
346: frequency approximation is as good as exact. We find our results
347: compare well with the experimental data\cite{terasaki,ong_nature}.
348: We now sharpen the prediction
349: (from Eq. (88) of ref.~\cite{shastry_1}) that reversing the sign of
350: hopping leads to a maximum in $S^*$ as a function of $T$. Fig. 2(c)
351: indeed shows such a behavior and provides an estimate of the
352: expected enhancement in $S^*$, a factor of 2 to 3 over the frustrated
353: case of NCO and also of the expected temperature scale $T_{max} \sim
354: |t|$. It would be interesting to check this experimentally. A part of the enhancement for $t>0$ arises
355: from the prominent peak in the single particle density of states coming close to the
356: fermi level. Our numerics show that interactions further amplify this effect
357: considerably (a factor of $\sim 2$).
358:
359: \begin{figure}
360: \begin{picture}(0,0)(0,0)
361: \put(-.4,4.){{\large (a)}}
362: \put(7.6,6.){{\large (b)}}
363: \put(7.6,3.7){{\large (c)}}
364: \put(7.6,1.4){{\large (d)}}
365: \end{picture}
366: \includegraphics[width=4.5cm]{fig3a.ps}
367: \includegraphics[width=3cm]{fig3b-d.ps}
368: \caption{(color online) Magnetic field dependence of the thermopower. (a), $S_{HM}(B)$ normalised to
369: $S_{HM}(0)$ shows the relative suppression of $S_{HM}$ with $B$
370: for temperatures $T=(100,120,140,160,180,200)$ K (blue to red), horizontal axis
371: is rescaled by $1/T$ as done in ref.~\cite{ong_nature}. (b)-(d), $S_{HM}(T)$ for several doping values around $x=0.71$ and several values of
372: $B$ from 0 T to 13 T (red to orange). Note that for
373: $x>0.75$ the field-dependence becomes very weak. The insets display the full
374: temperature behavior of $S_{HM}$ out to $T=400$ K. All data for $J=0$.
375: Finite $J$ reduces the field suppression.}
376: \label{Heikes_T}
377: \end{figure}
378:
379: Next we consider the effects of a magnetic field
380: of strength $B$ (perpendicular to
381: the plane) on the thermopower. The field induced change of the
382: transport term arises from the Peierls factors, and is found to
383: be a very small fraction ($\sim 0.001$) of the change in the Heikes-Mott term.
384: The field dependence of the chemical potential is significant and responsible
385: for the overall change. In Fig. 3(a)
386: we show the normalised $S_{HM}(B,T)$ for several dopings as function of $B/T$. This scaling
387: with $B/T$ is very similar to the one found in ref.~\cite{ong_nature}, both
388: qualitatively and quantitatively. In Fig. 3(b)-(d) we find a crossover for $x\sim
389: 0.75$ from a weakly $B$-dependent $S(B,T)$ at $x\geq 0.75$ to the
390: Curie-Weiss phase where $S$ is greatly suppressed by $B$. This
391: crossover is similar to the one seen in the $T$ dependence of the
392: spin-susceptibility. These results confirm the interpretation\cite{ong_nature} in terms
393: of spin-entropy as the leading contribution to the field-suppression
394: and also provide a guide to what one can expect at high magnetic
395: fields that are not accessible experimentally.
396:
397: \begin{figure}
398: \begin{picture}(0,0)(0,0)
399: \put(-0.75,-1.3){{\large (a)}}
400: \put(-0.75,-3.5){{\large (b)}}
401: \end{picture}
402: \includegraphics[width=5.0cm,angle=-90]{fig4ab.ps}
403: \caption{(color online) Hall coefficient and thermopower as a function of filling $n$.
404: (a), Filling dependence of $R^*_H(T=0)$ for a $12$-site torus (black
405: circles) and icosahedron (red diamonds) as well as
406: $R_H(T=0,\omega=0)$ for the $12$-site torus (blue triangles) all for
407: $J/|t|=0$. (b), Filling dependence of $S^*(T)$
408: for $T=100$ K, $200$ K, and $1000$ K (blue triangle, red star, black triangle)
409: for $J/|t|=0$ on a $12$-site torus. The solid black line is $S_{HM}$ at infinite $T$. The
410: dotted lines in both panels are guides to the eye indicating zero
411: $R_H$ and $S$ and half filling ($n=1$).}
412: \label{RH_S_x}
413: \end{figure}
414:
415: We next make a few comments about the overall behavior of the
416: thermopower and Hall constant as a function of filling in strong
417: coupling theories. As noted earlier\cite{sss,phillips}, while band
418: theory predicts a single zero crossing of the Hall constant as $n$
419: increases from $0 \rightarrow 2$, a strongly correlated electron
420: system has three crossings. In precisely the same sense the
421: thermopower also has three zero crossings for a correlated band
422: system. We demonstrate this for the triangular lattice. In
423: Fig. 4(a)-(b) we show the filling dependence of both the Hall
424: constant and the thermopower. The divergence of the Hall constant at
425: half filling is forced by the carrier freeze out accompanying the Mott-Hubbard gap,
426: and is less pronounced in the triangular lattice than in
427: the square lattice\cite{sss}. For the thermopower we expect similar
428: results for the square lattice.
429:
430: The hopping parameter $t$ ($\sim 100$ K) used here to reconcile
431: a considerable body of experimental data with theory, is much smaller
432: than that in high $T_c$ systems $t \sim 3000$ K. This value
433: is also smaller, by a factor of 10, than the one found in LDA studies \cite{singh,pickett},
434: and should motivate further studies of the difficult problem of
435: projecting a multi-band system onto a single band model
436: systematically. In conclusion, our work shows
437: that strong correlations account for the observed dramatic behavior
438: of the Curie-Weiss metallic phase in considerable quantitative detail.
439:
440: % figures should be put into the text as floats.
441: % Use the graphics or graphicx packages (distributed with LaTeX2e)
442: % and the \includegraphics macro defined in those packages.
443: % See the LaTeX Graphics Companion by Michel Goosens, Sebastian Rahtz,
444: % and Frank Mittelbach for instance.
445: %
446: % Here is an example of the general form of a figure:
447: % Fill in the caption in the braces of the \caption{} command. Put the label
448: % that you will use with \ref{} command in the braces of the \label{} command.
449: % Use the figure* environment if the figure should span across the
450: % entire page. There is no need to do explicit centering.
451:
452: % \begin{figure}
453: % \includegraphics{}%
454: % \caption{\label{}}
455: % \end{figure}
456:
457: % Surround figure environment with turnpage environment for landscape
458: % figure
459: % \begin{turnpage}
460: % \begin{figure}
461: % \includegraphics{}%
462: % \caption{\label{}}
463: % \end{figure}
464: % \end{turnpage}
465:
466: % tables should appear as floats within the text
467: %
468: % Here is an example of the general form of a table:
469: % Fill in the caption in the braces of the \caption{} command. Put the label
470: % that you will use with \ref{} command in the braces of the \label{} command.
471: % Insert the column specifiers (l, r, c, d, etc.) in the empty braces of the
472: % \begin{tabular}{} command.
473: % The ruledtabular enviroment adds doubled rules to table and sets a
474: % reasonable default table settings.
475: % Use the table* environment to get a full-width table in two-column
476: % Add \usepackage{longtable} and the longtable (or longtable*}
477: % environment for nicely formatted long tables. Or use the the [H]
478: % placement option to break a long table (with less control than
479: % in longtable).
480: % \begin{table}%[H] add [H] placement to break table across pages
481: % \caption{\label{}}
482: % \begin{ruledtabular}
483: % \begin{tabular}{}
484: % Lines of table here ending with \\
485: % \end{tabular}
486: % \end{ruledtabular}
487: % \end{table}
488:
489: % Surround table environment with turnpage environment for landscape
490: % table
491: % \begin{turnpage}
492: % \begin{table}
493: % \caption{\label{}}
494: % \begin{ruledtabular}
495: % \begin{tabular}{}
496: % \end{tabular}
497: % \end{ruledtabular}
498: % \end{table}
499: % \end{turnpage}
500:
501: % Specify following sections are appendices. Use \appendix* if there
502: % only one appendix.
503: %\appendix
504: %\section{}
505:
506: \begin{acknowledgments}
507: We thank N. P. Ong and Y. Wang for valuable discussions. This work is
508: supported by Grant No. NSF-DMR0408247.
509: \end{acknowledgments}
510:
511: % Create the reference section using BibTeX:
512: %\bibliography{references}
513:
514: \begin{thebibliography}{22}
515: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
516: \expandafter\ifx\csname bibnamefont\endcsname\relax
517: \def\bibnamefont#1{#1}\fi
518: \expandafter\ifx\csname bibfnamefont\endcsname\relax
519: \def\bibfnamefont#1{#1}\fi
520: \expandafter\ifx\csname citenamefont\endcsname\relax
521: \def\citenamefont#1{#1}\fi
522: \expandafter\ifx\csname url\endcsname\relax
523: \def\url#1{\texttt{#1}}\fi
524: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
525: \providecommand{\bibinfo}[2]{#2}
526: \providecommand{\eprint}[2][]{\url{#2}}
527:
528: \bibitem[{\citenamefont{Levi}(2003)}]{nco_general}
529: \bibinfo{author}{\bibfnamefont{B.~G.} \bibnamefont{Levi}},
530: \bibinfo{journal}{Physics Today} \textbf{\bibinfo{volume}{56}},
531: \bibinfo{pages}{15} (\bibinfo{year}{2003}).
532:
533: \bibitem[{\citenamefont{Wang et~al.}(2003{\natexlab{a}})\citenamefont{Wang,
534: Rogado, Cava, and Ong}}]{ong_nature}
535: \bibinfo{author}{\bibfnamefont{Y.~Y.} \bibnamefont{Wang}},
536: \bibinfo{author}{\bibfnamefont{N.~S.} \bibnamefont{Rogado}},
537: \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Cava}}, \bibnamefont{and}
538: \bibinfo{author}{\bibfnamefont{N.~P.} \bibnamefont{Ong}},
539: \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{423}},
540: \bibinfo{pages}{425} (\bibinfo{year}{2003}{\natexlab{a}}).
541:
542: \bibitem[{\citenamefont{Kumar and Shastry}(2003)}]{kumar}
543: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Kumar}} \bibnamefont{and}
544: \bibinfo{author}{\bibfnamefont{B.~S.} \bibnamefont{Shastry}},
545: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{68}},
546: \bibinfo{pages}{104508} (\bibinfo{year}{2003}).
547:
548: \bibitem[{\citenamefont{Baskaran}(2003)}]{baskaran}
549: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Baskaran}},
550: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{91}},
551: \bibinfo{pages}{097003} (\bibinfo{year}{2003}).
552:
553: \bibitem[{\citenamefont{Wang et~al.}(2004)\citenamefont{Wang, Lee, and
554: Lee}}]{lee}
555: \bibinfo{author}{\bibfnamefont{Q.-H.} \bibnamefont{Wang}},
556: \bibinfo{author}{\bibfnamefont{D.-H.} \bibnamefont{Lee}}, \bibnamefont{and}
557: \bibinfo{author}{\bibfnamefont{P.~A.} \bibnamefont{Lee}},
558: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{69}},
559: \bibinfo{pages}{092504} (\bibinfo{year}{2004}).
560:
561: \bibitem[{\citenamefont{Hasan et~al.}(2004)\citenamefont{Hasan, Chuang, Qian,
562: Li, Kong, Kuprin, Fedorov, Kimmerling, Rotenberg, Rossnagel et~al.}}]{hasan}
563: \bibinfo{author}{\bibfnamefont{M.~Z.} \bibnamefont{Hasan}},
564: \bibinfo{author}{\bibfnamefont{Y.-D.} \bibnamefont{Chuang}},
565: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Qian}},
566: \bibinfo{author}{\bibfnamefont{Y.~W.} \bibnamefont{Li}},
567: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Kong}},
568: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Kuprin}},
569: \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Fedorov}},
570: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kimmerling}},
571: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Rotenberg}},
572: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Rossnagel}},
573: \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rev. Lett.}
574: \textbf{\bibinfo{volume}{92}}, \bibinfo{pages}{246402}
575: (\bibinfo{year}{2004}).
576:
577: \bibitem[{\citenamefont{Yang et~al.}(2005)\citenamefont{Yang, Pan, Sekharan,
578: Sato, Souma, Takahashi, Jin, Sales, Mandrus, Fedorov et~al.}}]{yang_1}
579: \bibinfo{author}{\bibfnamefont{H.-B.} \bibnamefont{Yang}},
580: \bibinfo{author}{\bibfnamefont{Z.-H.} \bibnamefont{Pan}},
581: \bibinfo{author}{\bibfnamefont{A.~K.~P.} \bibnamefont{Sekharan}},
582: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Sato}},
583: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Souma}},
584: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Takahashi}},
585: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Jin}},
586: \bibinfo{author}{\bibfnamefont{B.~C.} \bibnamefont{Sales}},
587: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Mandrus}},
588: \bibinfo{author}{\bibfnamefont{A.~V.} \bibnamefont{Fedorov}},
589: \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rev. Lett.}
590: \textbf{\bibinfo{volume}{95}}, \bibinfo{eid}{146401}
591: (\bibinfo{year}{2005}).
592:
593: %\bibitem[{lon()}]{long_paper}
594: % To be published elsewhere.
595:
596: \bibitem[{\citenamefont{Haerter et~al.}(2005)\citenamefont{Haerter, Peterson,
597: and Shastry}}]{long_paper}
598: \bibinfo{author}{\bibfnamefont{J.~O.}~\bibnamefont{Haerter}},
599: \bibinfo{author}{\bibfnamefont{M.~R.}~\bibnamefont{Peterson}}, \bibnamefont{and}
600: \bibinfo{author}{\bibfnamefont{B.~S.}~\bibnamefont{Shastry}},
601: \bibinfo{journal}{condmat/0608005}
602: (\bibinfo{year}{2006}), and to be published.
603:
604:
605: \bibitem[{\citenamefont{Shastry et~al.}(1993)\citenamefont{Shastry, Shraiman,
606: and Singh}}]{sss}
607: \bibinfo{author}{\bibfnamefont{B.~S.} \bibnamefont{Shastry}},
608: \bibinfo{author}{\bibfnamefont{B.~I.} \bibnamefont{Shraiman}},
609: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.~R.~P.}
610: \bibnamefont{Singh}}, \bibinfo{journal}{Phys. Rev. Lett.}
611: \textbf{\bibinfo{volume}{70}}, \bibinfo{pages}{2004} (\bibinfo{year}{1993}).
612:
613: \bibitem[{\citenamefont{Hatsugai and Kohmoto}(1990)}]{kohmoto}
614: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Hatsugai}} \bibnamefont{and}
615: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kohmoto}},
616: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{42}},
617: \bibinfo{pages}{8282} (\bibinfo{year}{1990}).
618:
619:
620: \bibitem[{lon()}]{footnote1}
621: At $x\sim0.8$, the data in ref.~\cite{fujimori} suggest a value $t\sim -300$K, this conflicts with the thermopower data of ref.~\cite{terasaki} taken to measure $\mu(T)$ by the Heikes-Mott formula $S_{HM}$.
622:
623: \bibitem[{\citenamefont{Ishida et~al.}(2005)\citenamefont{Ishida, Ohta,
624: Fujimori, and Hosono}}]{fujimori}
625: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Ishida}},
626: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ohta}},
627: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Fujimori}}, \bibnamefont{and}
628: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Hosono}},
629: \bibinfo{journal}{condmat/0511149}
630: (\bibinfo{year}{2005}).
631:
632: \bibitem[{\citenamefont{Terasaki et~al.}(1997)\citenamefont{Terasaki, Sasago,
633: and Uchinokura}}]{terasaki}
634: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Terasaki}},
635: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Sasago}}, \bibnamefont{and}
636: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Uchinokura}},
637: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{56}},
638: \bibinfo{pages}{R12685} (\bibinfo{year}{1997}).
639:
640: \bibitem[{\citenamefont{Haerter and Shastry}(2005)}]{counter_nagaoka}
641: \bibinfo{author}{\bibfnamefont{J.~O.} \bibnamefont{Haerter}} \bibnamefont{and}
642: \bibinfo{author}{\bibfnamefont{B.~S.} \bibnamefont{Shastry}},
643: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{95}},
644: \bibinfo{pages}{087202} (\bibinfo{year}{2005}).
645:
646: \bibitem[{\citenamefont{Anderson}(1950)}]{anderson_nagaoka}
647: \bibinfo{author}{\bibfnamefont{P.~W.} \bibnamefont{Anderson}},
648: \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{79}},
649: \bibinfo{pages}{350} (\bibinfo{year}{1950}).
650:
651: \bibitem[{\citenamefont{Wang et~al.}(2003{\natexlab{b}})\citenamefont{Wang,
652: Rogado, Cava, and Ong}}]{wang_hall}
653: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Wang}},
654: \bibinfo{author}{\bibfnamefont{N.~S.} \bibnamefont{Rogado}},
655: \bibinfo{author}{\bibfnamefont{R.~J.} \bibnamefont{Cava}}, \bibnamefont{and}
656: \bibinfo{author}{\bibfnamefont{N.~P.} \bibnamefont{Ong}},
657: \bibinfo{journal}{condmat/0305455}
658: (\bibinfo{year}{2003}{\natexlab{b}}).
659:
660: \bibitem[{\citenamefont{Shastry}(2006)}]{shastry_1}
661: \bibinfo{author}{\bibfnamefont{B.~S.} \bibnamefont{Shastry}},
662: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{73}},
663: \bibinfo{pages}{085117} (\bibinfo{year}{2006}).
664:
665: \bibitem[{\citenamefont{Wang et~al.}(2006)\citenamefont{Wang, Li, and
666: Ong}}]{torque}
667: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Wang}},
668: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Li}}, \bibnamefont{and}
669: \bibinfo{author}{\bibfnamefont{N.~P.} \bibnamefont{Ong}},
670: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{73}},
671: \bibinfo{pages}{024510} (\bibinfo{year}{2006}).
672:
673: \bibitem[{\citenamefont{Chaikin and Beni}(1976)}]{beni_chaikin}
674: \bibinfo{author}{\bibfnamefont{P.~M.} \bibnamefont{Chaikin}} \bibnamefont{and}
675: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Beni}},
676: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{13}},
677: \bibinfo{pages}{647} (\bibinfo{year}{1976}).
678:
679: \bibitem[{\citenamefont{Stanescu and Phillips}(2004)}]{phillips}
680: \bibinfo{author}{\bibfnamefont{T.~D.} \bibnamefont{Stanescu}} \bibnamefont{and}
681: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Phillips}},
682: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{69}},
683: \bibinfo{pages}{245104} (\bibinfo{year}{2004}).
684:
685: \bibitem[{\citenamefont{Singh}(2000)}]{singh}
686: \bibinfo{author}{\bibfnamefont{D.~J.} \bibnamefont{Singh}},
687: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{61}},
688: \bibinfo{pages}{13397} (\bibinfo{year}{2000}).
689:
690: \bibitem[{\citenamefont{Lee et~al.}(2004)\citenamefont{Lee, Kunes, and
691: Pickett}}]{pickett}
692: \bibinfo{author}{\bibfnamefont{K.-W.} \bibnamefont{Lee}},
693: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kunes}}, \bibnamefont{and}
694: \bibinfo{author}{\bibfnamefont{W.~E.} \bibnamefont{Pickett}},
695: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{70}},
696: \bibinfo{pages}{045104} (\bibinfo{year}{2004}).
697:
698: \end{thebibliography}
699:
700:
701: \end{document}
702: %
703: % ****** End of file template.aps ******
704:
705:
706: