1: \documentclass[prl,aps,twocolumn,amsmath,showpacs,amssymb]{revtex4}
2: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
3: %TCIDATA{Created=Fri Jun 29 10:42:43 2001}
4: %TCIDATA{LastRevised=Mon Oct 22 09:58:51 2001}
5: %TCIDATA{Language=American English}
6:
7: \usepackage{graphicx}% Include figure files
8: \usepackage{dcolumn}% Align table columns on decimal point
9: \usepackage{bm}% bold math
10:
11: \begin{document}
12: \title{Anisotropic Magnetoresistance in Charge-Ordering $Na_{0.34}(H_3O)_{0.15}CoO_2$:
13: Strong Spin-Charge Coupling and Spin Ordering}
14:
15:
16: \author{C. H. Wang$^1$}
17: \author{ X. H. Chen$^1$}
18: \altaffiliation{Corresponding author} \email{chenxh@ustc.edu.cn}
19: \author{G. Wu$^1$, T. Wu$^1$, H. T. Zhang$^1$, J. L. Luo$^2$, G. T. Liu$^2$ and N. L. Wang$^2$}
20: \affiliation{ 1. Hefei National Laboratory for Physical Science at
21: Microscale and Department of Physics, University of Science and
22: Technology of China, Hefei, Anhui 230026, People's Republic of
23: China} \affiliation{2. Beijing National Laboratory for Condensed
24: Matter Physics, Institute of Physics, Chinese Academy of Science,
25: Beijing 100080, People's Republic of China}
26:
27:
28: \date{\today}
29:
30:
31:
32: \begin{abstract}
33: Angular-dependent in-plane magnetoresistance (AMR) for single
34: crystal $Na_{0.34}(H_3O)_{0.15}CoO_2$ with charge ordering is
35: studied systematically. The anisotropic magnetoresistance shows a
36: twofold symmetry at high temperature with rotating H in the Co-O
37: plane, while a sixfold symmetry below a certain temperature
38: ($T_\rho$). At $T_\rho$, the symmetry of AMR changes from twofold
39: to fourfold with rotating magnetic field (H) in the plane
40: consisting of the current and c-axis. The variation of AMR
41: symmetry with temperature arises from the subtle changes of the
42: spin structure. These results give a direct evidence for the
43: itinerant electrons directly coupled to the localized spins.
44: \end{abstract}
45:
46: \pacs{71.27.+a, 74.70.-b, 75.25.+z}
47:
48: \maketitle
49: \newpage
50:
51: The layered sodium cobaltate $Na_xCoO_2$ has become one of the
52: focus in research due to the discovery of superconductivity with
53: $T_c\sim 5$ K in $Na_{0.35}CoO_2\cdot1.3H_2O$.\cite{Takada} In
54: fact, the phase diagram in this system is very
55: rich\cite{foo,Sakurai,Milne} and one has observed charge order
56: and several complicated magnetic orders in addition to
57: superconductivity.\cite{Sugiyama,Luojl,Mendels} Recently, many
58: anomalous properties were observed in this system. The
59: angle-resolved photoemission spectroscopy(ARPES) experiments
60: indicates that in $Na_{0.7}CoO_2$ the well-defined quasiparticle
61: peaks only exist in low temperature when a coherent transport
62: forms along c-axis.\cite{hasan} Such peculiar behaviors make one
63: believe that $Na_xCoO_2$ could be regarded as a new prototype for
64: studying the doped Mott insulator.
65:
66: $Na_xCoO_2$ system shows complicated magnetic structure. The bulk
67: antiferromagnetism (AF) with the ordering moment perpendicular to
68: Co-O plane was observed in $Na_{0.82}CoO_2$\cite{Bayrakci}. In
69: $Na_{0.75}CoO_2$, the ferromagnetic correlation within the cobalt
70: layer and AF correlation between the cobalt layers were observed
71: with spin direction along c-axis.\cite{Boothroyd,Helme} Neutron
72: diffraction and NMR studies indicate that there exist two distinct
73: Co sites with different magnetic moments in $Na_{0.5}CoO_2$ and
74: antiferromagnetic ordering occurs below $T_{c1}\sim 87$
75: K.\cite{yokoi,gasparovic} However, two different magnetic
76: structures are proposed by Yokoi et al.\cite{yokoi} and Gasparovic
77: et al.\cite{gasparovic}, respectively. Gasparovic et al. proposed
78: that magnetic moment of $Co^{4+}$ ions that is magnetic (in the
79: S=1/2 low spin state) aligns antiferromagnetically, the rows of
80: $Co^{4+}$ ions alternate with rows of $Co^{3+}$ ions (in the S=0
81: spin state) that are nonmagnetic. Yokoi et al. believed that the
82: spin structure of $Na_{0.5}CoO_2$ has three axis, two axes are
83: $Co^{3.5-\delta}-Co^{3.5-\delta}$ and the other one is
84: $Co^{3.5+\delta}-Co^{3.5+\delta}$. The large moments of
85: $Co^{3.5+\delta}$ sites align antiferromagnetically with spin
86: direction in Co-O plane. The small moments of $Co^{3.5-\delta}$
87: sites is along c-axis and their interplane correlation is
88: antiferromangetic. The magnetic structure for $Na_{0.5}CoO_2$ is
89: open question.
90:
91: The oxonium ions, $(H_3O)^+$, can occupy the same crystallographic
92: sites of the Na ions when the $Na_xCoO_2$ crystal is immersed in
93: distilled water.\cite{takada1,goodenough} Intercalation of
94: $(H_3O)^+$ makes the Co valence to be lower than theoretical value
95: estimated directly from Na content. Much research focused on the
96: effect of oxonium ions\cite{goodenough,Milne} in $Na_xCoO_2$
97: system, so that superconducting phase diagram is revised.
98: Oxidation state of Co for superconducting sample is very close to
99: 3.5 of $Na_{0.5}CoO_2$ due to intercalation of oxonium ions
100: instead of about 3.7.\cite{karppinen} One should take the charge
101: ordering insulator with Co oxidation state of $\sim3.5$ as parent
102: compound for the superconductor. Therefore, understanding on
103: microscopic magnetic structure of $Na_{0.5}CoO_2$ is very
104: important to understand the charge ordering behavior and
105: superconducting mechanism.
106:
107: Magnetoresistance (MR) provides new insight into the coupling
108: between charges and background magnetism. This is particularly
109: valuable because, as shown in this work, small magnetic moments
110: order in the background of magnetic ordering with large magnetic
111: moments, and such ordering is difficultly detected by
112: magnetization measurement. In this paper, we studied AMR of
113: $Na_{0.34}(H_3O)_{0.15}CoO_2$ crystal with the same charge
114: ordering as that in $Na_{0.5}CoO_2$.\cite{chenxh} Change of AMR
115: symmetry from twofold to sixfold is observed with rotating H in
116: the Co-O plane at a certain temperature ($T_\rho$). At $T_\rho$,
117: symmetry of AMR changes from twofold to fourfold with rotating
118: magnetic field (H) in the plane consisting of the current (I) and
119: c-axis. Such symmetry change of AMR can be well understood by spin
120: ordering of $Co^{3.5-\delta}$ sites with small magnetic moments
121: below $T_\rho$. Our results support the model proposed by Yokoi et
122: al.\cite{yokoi} These results give a direct evidence for the
123: itinerant electrons directly coupled to the localized spins.
124:
125: High quality single crystal $Na_{0.34}(H_3O)_{0.15}CoO_2$ was
126: obtained through $Na_{0.41}CoO_2$ crystal (20 mg) immersed in
127: distilled water (120 ml) for about 120 hours. The actual Na
128: concentration was determined by inductively coupled plasma
129: spectrometer (ICP) chemical analysis. The detailed description on
130: the synthesis and characterization of
131: $Na_{0.34}(H_3O)_{0.15}CoO_2$, which is only sample to show the
132: charge ordering behavior except for $Na_{0.5}CoO_2$, has been
133: reported elsewhere.\cite{chenxh} The near-normal incident
134: reflectance spectra were measured on the freshly cleaved surface
135: by a Bruker 66 v/s spectrometer in the frequency range from
136: 40$\sim$2900 $cm^{-1}$, as described in our early
137: report.\cite{wangnl} Standard Kramers-Kronig transformations were
138: employed to drive the frequency-dependent conductivity spectra.
139: Susceptibility and magnetoresistance were measured in Quantum
140: Design SQUID and PPMS system, respectively.
141:
142: Fig. 1(a) shows temperature dependence of in-plane resistivity and
143: susceptibility for $Na_{0.34}(H_3O)_{0.15}CoO_2$. As shown in
144: Fig.1(a), similar insulating behavior to that in $Na_{0.5}CoO_2$
145: was observed below 50 K. It indicates that charge ordering
146: behavior occurs in $Na_{0.34}(H_3O)_{0.15}CoO_2$. Susceptibility
147: shows two transitions at 84 and 47 K. It is quite similar to that
148: observed in $Na_{0.5}CoO_2$ although the transition temperatures
149: are slightly lower compared to $Na_{0.5}CoO_2$. Oxidation state
150: (3.49) of Co ions in $Na_{0.34}(H_3O)_{0.15}CoO_2$ is almost the
151: same as that (3.50) in $Na_{0.5}CoO_2$. It suggests that the
152: charge ordering in $Na_xCoO_2$ system strongly depends on the
153: valence of Co ion. In comparison with $Na_{0.5}CoO_2$,
154: differential curve $d\rho/dT$ was shown in Fig.1(b) and (c). A dip
155: of the $d\rho/dT$ was observed at $T_{c1}\sim 84$ K, and the
156: obvious change in the slope $d\rho/dT$ was also observed at
157: $T_{c2} \sim 50 $ K and $T_{\rho} \sim 10$ K, respectively. These
158: features are quite similar to that in $Na_{0.5}CoO_2$, in which
159: the dip is observed at $T_{c1} \sim 87$ K and the slope of
160: $\rho(T)$ increases abruptly at $T_{c2} \sim 53$ K and $T_{\rho}
161: \sim 20 $ K.\cite{foo,huang,wangch} It is commonly believed that
162: the anomaly at $T_{c1} \sim 87$ K arises from an antiferromagnetic
163: ordering.\cite{yokoi,gasparovic} The anomaly at $T_{c2} \sim 53$ K
164: is not clear although a kink is observed in
165: susceptibility\cite{foo,huang} due to no change in magnetic
166: structure drastically.\cite{yokoi} Dramatic change in $d\rho/dT$
167: at $T_{\rho}$ is believed to arise from spin
168: reorientation.\cite{huang} The behavior in
169: $Na_{0.34}(H_3O)_{0.15}CoO_2$ crystal is similar to
170: $Na_{0.5}CoO_2$ despite these anomalous temperatures are lower.
171: Therefore, the origin for these anomalies in
172: $Na_{0.34}(H_3O)_{0.15}CoO_2$ and $Na_{0.5}CoO_2$ should be the
173: same.
174:
175: \begin{figure}[h]
176: \includegraphics[width=8cm]{Fig1.eps}\vspace{-5mm}
177: \caption{\label{fig:epsart} (a): Temperature dependence of
178: in-plane resistivity and susceptibility with H $\|$ Co-O plane for
179: $Na_{0.34}(H_3O)_{0.15}CoO_2$; Temperature dependence of the slope
180: $d\rho/dT$ for the data shown in (a), (b): temperature range from
181: 30 to 100 K; (c): the temperatures below 30 K; (d): In-plane
182: optical conductivity spectra of $Na_{0.34}(H_3O)_{0.15}CoO_2$.}
183: \end{figure}
184:
185: To further characterize the charge ordering behavior in
186: $Na_{0.34}(H_3O)_{0.15}CoO_2$, in-plane optical conductivity
187: spectra are measured. As shown in Fig.1(d), in-plane optical
188: conductivity spectra of $Na_{0.34}(H_3O)_{0.15}CoO_2$ are quite
189: similar to that of $Na_{0.5}CoO_2$.\cite{wangnl} A broad hump
190: appears below 100 K at about 800 $cm^{-1}$ was
191: observed.\cite{wangnl} A sharp suppression of the conductivity
192: spectra below charge ordering temperature indicates an opening of
193: charge gap. The gap value is close to the reported value of $\sim
194: 125$ $cm^{-1}$ with $2\Delta/k_BT_{co}=3.5$.\cite{wangnl} It
195: further indicates that all features observed in $Na_{0.5}CoO_2$
196: occur in $Na_{0.34}(H_3O)_{0.15}CoO_2$.
197:
198: Fig. 2(a) shows isothermal in-plane MR at 5 K and 20 K with
199: different $\alpha$. The $\alpha$ is the angle between magnetic
200: field H and current I. It can be seen that MR are negative at 5K
201: and 20 K with $H \| I$ (within Co-O plane), while positive with $H
202: \bot$ the Co-O plane. One should note that MR is almost zero with
203: $\alpha=90^o$ at 20 K. Such behavior has been observed in
204: $Na_{0.5}CoO_2$.\cite{wangch}
205:
206: \begin{figure}[h]
207: \includegraphics[width=8cm]{Fig2.eps}\vspace{-5mm}
208: \caption{\label{fig:epsart} (a): The isothermal MR at 5 K and 20 K
209: with different angles between H and I; (b): The angular dependent
210: isothermal MR and polar plot for $Na_{0.34}(H_3O)_{0.15}CoO_2$
211: under H=12 Tesla at different temperatures. The magnetic field H
212: is kept in the plane consisting of the current and c-axis when the
213: sample is rotated, as shown in schematic diagram.}
214: \end{figure}
215:
216: In order to study effect of spin ordering on the charge transport
217: (especially below $T_{\rho}$) and to understand the behavior
218: observed in Fig.2(a), we investigated the angular dependent
219: in-plane MR behavior at different temperatures by rotating H in
220: the plane consisting of the current and c-axis, as shown in
221: schematic diagram. The magnetic field is fixed at 12 T during the
222: measurement. As shown in Fig.2, the MR increases with decreasing
223: temperature, and the magnitude of MR $\Delta\rho_{ab}/\rho_{ab}$
224: at 2 K is almost five times larger than that at 45 K. The data
225: reveal that the AMR has two-fold symmetry above 15 K, and the MR
226: is always negative and the maximum negative MR occurs at
227: $\alpha$=0 and 180$^o$, that is: the H lies in the Co-O plane.
228: While the MR is almost zero with H perpendicular to the Co-O plane
229: above 15 K. A striking feature is that the AMR shows clearly a
230: symmetry evolution from p-wave-like to d-wave-like with decreasing
231: temperature. One can observe only two negative arms above 15 K.
232: Two small positive arms begin to be observed at 10 K. The two
233: positive arms become large with further decreasing temperature,
234: and eventually dominate at 2 K. With the growth of positive arms,
235: one can find that the negative arms shrink slightly. It should be
236: pointed out that the AMR symmetry transition from twofold to
237: fourfold occurs at $\sim 15$ K close to $T_{\rho}$. Such
238: fascinating transition was also observed at $T_{\rho}\sim 20$ K in
239: $Na_{0.5}CoO_2$.\cite{wangch} Such d-wave like behavior in AMR has
240: been observed in antiferromagnetic $YBa_2Cu_3O_{6+x}$, and is
241: believed as a consequence of the rotation of stripe direction with
242: respect to the current direction.\cite{ando1} Therefore, we
243: believe that the evolution of AMR symmetry observed in charge
244: ordering state should arise from change of magnetic structure
245: under magnetic field.
246:
247: AMR is also studied with rotating H in the Co-O plane.
248: Fig.3 shows isothermal in-plane MR as a function of angle and the
249: polar plot for $Na_{0.34}(H_3O)_{0.15}CoO_2$ at H=12 Tesla at
250: different temperatures. The magnitude of MR increases
251: approximately by 5 times at low temperature relative to that at 40
252: K. Similar to the case with rotating H in the plane consisting of
253: the current and c-axis, symmetry of AMR is twofold above 15 K, but
254: changes to sixfold at 2 K. The variation of MR with angle follows
255: a sine wave above 15 K. The data at 15 K clearly show that peak
256: and valley of the sine wave broaden and become asymmetry. It
257: suggests that symmetry of AMR begins to change to sixfold at 15 K,
258: which is consistent with the temperature corresponding to twofold
259: to fourfold symmetry in AMR as shown in Fig.2. It implies that
260: symmetry change of AMR in Fig.2 and Fig.3 arises from the same
261: origin. Nevertheless, the data in Fig.3 exhibit some distinct
262: difference from that in Fig. 2. Firstly, the MR remains negative
263: at all angles in the whole temperature regime, and the magnitude
264: of MR at each fixed angle always increases with decreasing
265: temperature. Secondly, AMR shown in Fig.3 changes to six-fold at 2
266: K and exhibits obviously asymmetric, while AMR at various
267: temperatures shown in Fig.2 is symmetric.
268:
269: \begin{figure}[h]
270: \includegraphics[width=8cm]{Fig3.eps}
271: \caption{\label{fig:epsart} The isothermal MR as a function of
272: angle for the $Na_{0.34}(H_3O)_{0.15}CoO_2$ at H=12 Tesla and the
273: polar plot. H is kept in Co-O plane when the sample is rotated.}
274: \end{figure}
275:
276: Another striking feature is that the angle difference between
277: neighboring peaks in AMR at 2 K shown in Fig.3 is exactly 60$^o$.
278: Therefore, such nice sixfold symmetry of AMR at 2 K seems to be
279: related to the crystallographic structure because the crystal
280: structure of $Na_xCoO_2$ is hexagonal. However, the sixfold
281: symmetry of AMR with rotating H in Co-O plane is only observed
282: below 15 K as shown in Fig.3. Therefore, such sixfold symmetry of
283: AMR does not originate directly from the crystal structure. It
284: should be pointed out that no such sixfold symmetry of AMR can be
285: observed in other $Na_xCoO_2$ without charge ordering. It suggests
286: that the symmetric feature of AMR observed in Fig.2 and Fig.3
287: could be related to the spin structure. Yokoi et al. proposed a
288: magnetic structure shown in Fig.4 (a) and (b) for $Na_{0.5}CoO_2$.
289: The spin structure has three axis, two axes are
290: $Co^{3.5-\delta}-Co^{3.5-\delta}$ and the other one is
291: $Co^{3.5+\delta}-Co^{3.5+\delta}$.\cite{yokoi} The large moments
292: of $Co^{3.5+\delta}$ sites align antiferromagnetically at
293: $T_{c1}\sim 87$ K with spin direction in Co-O plane. The small
294: moments of $Co^{3.5-\delta}$ sites is along c-axis and their
295: interplane correlation is antiferromangetic. It is difficult to
296: distinguish their in-plane correlation, so that there exist two
297: possibilities for the in-plane correlation as shown in Fig.4(a)
298: and (b).\cite{yokoi} But they remain an open question that at what
299: temperature the small moments of $Co^{3.5-\delta}$ sites order.
300: The evolution of the AMR symmetry with temperature shown in Fig.2
301: and Fig.3 can be well understood by the spin structure proposed by
302: Yokoi et al. At high temperature, the symmetry of AMR is twofold
303: for both of the cases with rotating H in the Co-O plane and in the
304: plane consisting of the c-axis and the current. This is because
305: only the large moments of $Co^{3.5+\delta}$ sites order
306: \begin{figure}[h]
307: \includegraphics[width=8cm]{Fig4.eps}
308: \caption{\label{fig:epsart} (a) and (b): Possible magnetic
309: structures proposed by Yokoi et al.\cite{yokoi} for
310: $Na_{0.5}CoO_2$; (c) and (d): A possible configuration for the
311: in-plane matching between the magnetic structure and the AMR
312: symmetry at 15 and 2 K, respectively; (e) and (f): Another
313: possible configuration for the in-plane matching between the
314: magnetic structure and the AMR symmetry at 15 and 2 K,
315: respectively.}
316: \end{figure}
317: antiferromagnetically at high temperature. Therefore, only a
318: "stripe"-like spin ordering along
319: $Co^{3.5+\delta}-Co^{3.5+\delta}$ exists in the Co-O plane. It
320: naturally explains the twofold symmetry of AMR shown in Fig.2 and
321: Fig.3. For the case with rotating H in Co-O plane at high
322: temperatures, we proposed a model for the matching between the AMR
323: symmetry and the magnetic structure shown in Fig.4(c) and (e).
324: There exist two possibilities for the relation between the AMR
325: symmetry and magnetic structure: one is that the maximum MR occurs
326: with H along the $Co^{3.5+\delta}-Co^{3.5+\delta}$ axis, the other
327: is perpendicular to the $Co^{3.5+\delta}-Co^{3.5+\delta}$ axis. At
328: low temperatures, the spin ordering at $Co^{3.5-\delta}$ sites
329: with small moment occurs, and consequently the magnetic structure
330: is hexagonal in the Co-O plane. Such hexagonal magnetic structure
331: leads to the sixfold symmetry of AMR with rotating H in Co-O plane
332: shown in Fig.3. The two possible correlations between the AMR
333: symmetry and the magnetic structure are proposed in Fig.4(d) and
334: (f) at low temperatures. Based on the observation of sixfold
335: symmetry of AMR at 2 K shown in Fig.3, the magnetic structure
336: prefers to that in-plane correlation for the magnetic moments at
337: $Co^{3.5-\delta}$ sites is antiferromagnetic shown in Fig.4(b).
338: This is because in-plane symmetry of the magnetic structure is not
339: hexagonal if all the $Co^{3.5-\delta}$ sites are exactly
340: equivalent as shown in Fig.4(a). The fourfold and twofold
341: symmetries of AMR shown in Fig.2 can be also understood with the
342: magnetic structures at low and high temperatures, respectively.
343: Therefore, present results support the magnetic structure proposed
344: by Yokoi et al.\cite{yokoi} The magnetic structure proposed by
345: Gasparovic et al.\cite{gasparovic} cannot explain the sixfold
346: symmetry of AMR observed in Fig.3 because only a "stripe"-like
347: spin ordering along $Co^{4+}-Co^{4+}$ occurs in Co-O plane in
348: their model. A strong scattering of electrons from spins should
349: occur with the current perpendicular to the antiferromagnetic
350: "stripe" along $Co^{3.5+\delta}-Co^{3.5+\delta}$. However, it is
351: difficult to distinguish which set of the configurations shown in
352: Fig.4(c) and (d), Fig.4(e) and (f) is more realistic because we
353: cannot determine the $Co^{3.5+\delta}-Co^{3.5+\delta}$ axis. The
354: evolution of the AMR symmetry with temperature shown in Fig.2 and
355: Fig.3 suggests that the spin ordering of $Co^{3.5-\delta}$ sites
356: with small moments occurs at $\sim 15$ K, which is close to the
357: $T_{\rho}$. This is explained why the resistivity increases
358: rapidly at $T_{\rho}$. It should be pointed out that symmetry
359: change of AMR observed in $Na_{0.34}(H_3O)_{0.15}CoO_2$ can be
360: also observed in $Na_{0.5}CoO_2$, only $T_{\rho}$ $\sim 20$ K at
361: which the ARM symmetry changes from twofold to sixfold is slightly
362: different in $Na_{x}CoO_2$ x$\sim$ 0.5.\cite{hu}. These results
363: give a good understanding on the transport and magnetic properties
364: for the charge ordering $Na_xCoO_2$, especially on the anomalies
365: at $T_{\rho}$, $T_{c1}$ and $T_{c2}$.
366:
367: In this letter, the angular-dependent in-plane magnetoresistance
368: (AMR) for $Na_{0.34}(H_3O)_{0.15}CoO_2$ is systematically studied.
369: The symmetry of AMR is twofold for both of the case with rotating
370: H in the Co-O plane and in the plane consisting of the current and
371: c-axis above $T_{\rho}$, while changes to sixfold and fourfold
372: below $T_{\rho}$, respectively. These intriguing results support
373: the magnetic structure proposed by Yokoi et al. We solve the open
374: question for the spin ordering temperature ($\sim$ $T_{\rho}$) of
375: the small moments at $Co^{3.5-\delta}$ sites. The present work
376: shows that the AMR is an auxiliary but powerful tool to study the
377: subtle change of the magnetic structure. The magnetic and
378: transport properties of $Na_{0.34}(H_3O)_{0.15}CoO_2$ are the same
379: as those of $Na_{0.5}CoO_2$. It indicates intercalation of oxonium
380: ions occurred and its charge compensation to make the oxidation
381: state of Co to be close to $\sim3.5$ when $Na_xCoO_2$ crystal was
382: hydrated. Therefore, one should take the charge ordering insulator
383: with Co oxidation state of $\sim3.5$ as parent compound of the
384: superconductor to think about superconducting mechanism and phase
385: diagram.
386:
387: {\bf Acknowledgement:} This work is
388: supported by the Nature Science Foundation of China and by the
389: Ministry of Science and Technology of China (973 project No:
390: 2006CB601001), and by the Knowledge Innovation Project of Chinese
391: Academy of Sciences.
392:
393: \begin{thebibliography}{00}
394:
395: \bibitem{Takada} K. Takada et al., Nature \textbf{422}, 53(2003).
396: \bibitem{foo} M. L. Foo et al., Phys. Rev. Lett. \textbf{92}, 247001(2004).
397: \bibitem{Sakurai} H. Sakurai et al., J. Phys. Soc. Jpn. \textbf{74}, 2909(2005).
398: \bibitem{Milne} C. J. Milne et al., Phys. Rev. Lett. \textbf{93},
399: 247007(2004).
400: \bibitem{Sugiyama} J. Sugiyama et al., Phys. Rev. Lett. \textbf{92}, 017602(2004).
401: \bibitem{Luojl} J. L. Luo et al., Phys. Rev. Lett. \textbf{93}, 187203(2004).
402: \bibitem{Mendels} P. Mendels et al., Phys. Rev. Lett. \textbf{94}, 136403(2005).
403: \bibitem{hasan} M. Z. Hasan et al., Phys. Rev. Lett. \textbf{92}, 246402(2004)
404: \bibitem{Bayrakci} S. P. Bayrakci et al., Phys. Rev. B \textbf{69}, 100410(2005)
405: \bibitem{Boothroyd} A. T. Boothroyd et al., Phys. Rev. Lett. \textbf{92},
406: 197201(2004).
407: \bibitem{Helme}
408: L. M. Helme et al., Phys. Rev. Lett. \textbf{94}, 157206(2005).
409: \bibitem{yokoi} M. Yokoi et al., J. Phys. Soc. Jpn {\bf 74},
410: 3046(2005).
411: \bibitem{gasparovic}G. Gasparovic et al., Phys. Rev. Lett. {\bf 96}, 046403(2006).
412: \bibitem{takada1} K. Takada et al., J. Mater. Chem. \textbf{14}, 1448(2004).
413: \bibitem{goodenough} M. Banobre-Lopez et al., Chem. Mater. \textbf{17},
414: 1965(2005).
415: \bibitem{karppinen} M. Karppinen et al., Chem. Mater. {\bf 16},
416: 1693(2004); P. W. Barnes et al., Phys. Rev. B {\bf 72},
417: 134515(2005).
418: \bibitem{chenxh} C. H. Wang et al., Solid State Commun. {\bf 138}, 169(2006).
419: \bibitem{huang} Q. Huang et al., J. Phys.: Condens. Matter \textbf{16}, 5803(2004)
420: \bibitem{wangch} C. H. Wang et al., Phys. Rev. B \textbf{71}, 224515(2005).
421: \bibitem{wangnl} N. L. Wang et al., Phys. Rev. Lett. \textbf{93}, 147403(2004).
422: \bibitem{ando1}
423: Y. Ando, A. N. Lavrov, and S. Komiya, Phys. Rev. Lett. {\bf 83},
424: 2813(2003).
425: \bibitem{hu}
426: F. Hu et al., Phys. Rev. B {\bf 73}, 212414(2006).
427:
428:
429:
430:
431:
432:
433:
434:
435:
436: \end{thebibliography}
437:
438:
439:
440: \end{document}
441: