1: %last edited 06/14/01

2: % a double-spaced preprint version

3: %\documentstyle[preprint,aps,epsf]{revtex}

4: % a good-looking double column galley style output

5: \documentstyle[aps,prl,multicol,floats,epsf]{revtex}

6: 

7: \def\thefootnote{\fnsymbol{footnote}}

8: 

9: \begin{document}

10: \draft

11: \wideabs{

12: 

13: \title{Magnetotransport properties of untwinned 

14: YBa$_{2}$Cu$_{3}$O$_{y}$ single crystals: \\

15: Novel 60-K-phase anomalies in the charge transport}

16: 

17: \author{Yoichi Ando\cite{corr} and Kouji Segawa}

18: \address{Central Research Institute of Electric Power Industry, 

19: Komae, Tokyo 201-8511, Japan}

20: 

21: \date{\today}

22: \maketitle

23: 

24: \begin{abstract}

25: We present the result of our accurate measurements of 

26: the  $a$- and $b$-axis resistivity ($\rho_a$ and $\rho_b$), 

27: magnetoconductivity $\Delta\sigma/\sigma$, 

28: Hall coefficient $R_H$, and the $a$-axis thermopower $S_a$ 

29: in untwinned YBa$_{2}$Cu$_{3}$O$_{y}$ single crystals in a 

30: wide range of doping ($6.45 \le y \le 7.0$).  

31: The systematics of our data reveals a number of 

32: novel 60-K-phase anomalies in the charge transport: 

33: (i) Temperature dependences of $\rho_a$ show anomalous 

34: overlap below $\sim$130 K for $6.65 \le y \le 6.80$,

35: (ii) Hall mobility $\mu_{H}$ shows an enhancement 

36: near $y \simeq 6.65$, which is reflected in an 

37: anomalous $y$ dependence of $\sigma_{xy}$, 

38: (iii) With decreasing temperature $R_H$ shows a marked drop 

39: upon approaching $T_c$ only in samples with $6.70 \le y \le 6.85$, 

40: (iv) Superconducting fluctuation magnetoconductivity is 

41: anomalously enhanced near $y \simeq 6.7$, and 

42: (v) $H_{c2}$ is anomalously reduced near $y \simeq 6.70$. 

43: We discuss that the fluctuating charge stripes might be 

44: responsible for these anomalies in the charge transport.

45: \end{abstract}

46: 

47: \pacs{Keywords: Transport Properties, 60-K phase, Superconducting 

48: Fluctuations, Coherence Length}

49: }

50: \narrowtext

51: 

52: \section{Introduction}

53: 

54: In YBa$_{2}$Cu$_{3}$O$_{y}$ (YBCO) compound, 

55: an increase in the oxygen content $y$ from 6 to 7 causes 

56: the hole doping into the CuO$_2$ planes (above $y \simeq 6.2$) 

57: and leads to superconductivity (above $y \simeq 6.4$).

58: The dependence of $T_c$ on $y$ is non-trivial 

59: and there is a well-known plateau with $T_c \simeq$ 60 K 

60: for $y$ of around 6.7 (``60-K plateau" or ``60-K phase").

61: Both an oxygen-ordering scenario \cite{Veal} and an 1/8-anomaly 

62: scenario \cite{Tallon} have been discussed for the 

63: origin of the 60-K plateau, but the case remains controversial.

64: To understand the true nature of the 60-K phase in YBCO, 

65: we have conducted systematic measurements of the transport 

66: properties across the 60-K phase, which reveal

67: clear electronic anomalies in the 60-K phase and thus indicate 

68: an electronic origin of the 60-K plateau. 

69: In this paper, we summarize the novel 60-K-phase anomalies 

70: we found in the charge transport in YBCO and discuss their 

71: possible origin in conjunction with the self-organization of 

72: the holes into stripes.

73: 

74: \begin{figure}[t!]

75: \epsfxsize=0.95\columnwidth

76: \centerline{\epsffile{Fig1.eps}}

77: \vspace{0.2cm}

78: \caption{$T$ dependences of (a) $\rho_a$ and (b) $\rho_b$ for 

79: untwinned YBCO crystals in 0 T.  

80: Inset in (a): Phase diagram of zero-resistance $T_c$ vs. $y$.}

81: \label{fig1}

82: \end{figure}

83: 

84: \section{Experimental}

85: 

86: It should be noted that the oxygen arrangement in the Cu-O 

87: chain layers is essentially random, causing 

88: complications to the study of YBCO; for example, for a given 

89: $y$ the actual hole doping can differ 

90: depending on the arrangement of the O atoms, and 

91: the O atoms in the Cu-O chains can rather easily rearrange at 

92: room temperature, which causes the ``room-temperature (RT) 

93: annealing effect" \cite{Lavrov}.  

94: For this work, the crystals are always quenched at the 

95: end of the high-temperature annealing (which tunes the oxygen 

96: content) and detwinning is performed at temperatures below 

97: 220$^{\circ}$C after the annealing.

98: The samples are left at room temperature for 

99: at least a week for the oxygen arrangement to equilibrate before 

100: the measurements; therefore, the oxygen atoms on the chain sites 

101: are expected to be locally ordered (because of the RT annealing) 

102: but macroscopically uniform (because of the quenching) \cite{Lavrov}.

103: (Note that a long-time annealing at relatively low temperature 

104: ($\sim$100$^{\circ}$C) causes a macroscopic phase separation in 

105: heavily-underdoped samples \cite{Radaelli} and messes up the 

106: transport properties.)

107: Our procedure ensures very good reproducibility of the 

108: transport properties, as has been demonstrated in Ref. \cite{60K}.  

109: In particular, we have determined the absolute values 

110: of the resistivity and the Hall coefficient for a given $y$ with 

111: the accuracy of 5\%; this gives us confidence in discussing the 

112: systematics of the transport properties across the 60-K phase.

113: Details of the measurement techniques are described in Refs. 

114: \cite{60K,Abe,stripeMR}. 

115: 

116: 

117: \section{Results and Discussions}

118: 

119: Let us start with the resistivity behavior.  

120: Figure 1(a) shows the temperature dependences of $\rho_a$ for a wide 

121: range of oxygen contents, $y = 6.45-7.0$.  Remember that 

122: the Cu-O chains run along the $b$-axis and thus $\rho_a$ is not 

123: complicated with the conductivity of the chains \cite{Gagnon}.

124: We emphasize that at least 3 samples are measured for each 

125: $y$ and the data are reproducible within 5\%; in fact, 

126: Fig. 1(a) is a summary of the measurements of more than 30 samples.

127: One can immediately notice in Fig. 1(a) that the $\rho_a(T)$ data for 

128: $y = 6.65-6.80$ show clear overlap below $\sim$130 K.

129: Note that in the underdoped YBCO the pseudogap opening 

130: can be inferred from a downward deviation from the 

131: high-temperature $T$-linear dependence \cite{Ito}, 

132: and thus the overlapping of $\rho_a(T)$ is observed 

133: in the pseudogap state. 

134: Figure 1(b) shows the corresponding evolution of $\rho_b(T)$ 

135: with $y$; $\rho_b$ is generally smaller than 

136: $\rho_a$ for the same $y$, which is believed to be due to 

137: the finite chain conductivity \cite{Gagnon}.

138: The $\rho_b(T)$ data do not show as clear overlap in the 

139: 60-K phase as the $\rho_a(T)$ data. 

140: 

141: \begin{figure}[t!]

142: \epsfxsize=0.8\columnwidth

143: \centerline{\epsffile{Fig2.eps}}

144: \vspace{0.2cm}

145: \caption{$y$ dependence of the $a$-axis thermopower at 290 K.}

146: \label{fig2}

147: \end{figure}

148: 

149: The overlap of the $\rho_a(T)$ data is very unusual.  

150: Unless the effective mass $m^*$ of the charge carrier in YBCO 

151: is anomalously changing with $y$ (which is very unlikely), 

152: the $y$-independence of $\rho_a$ can have only 

153: two possible origins: (i) both the carrier concentration $n$ 

154: and the scattering time $\tau$ remain unchanged with $y$, or 

155: (ii) a change in $n$ is compensated by a change in $\tau$.  

156: To clarify which of the two is actually the case, we measured 

157: the room-temperature thermopower $S(290{\rm K})$, which is 

158: generally believed to reflect the change in the hole concentration 

159: and thus may be used as a guide to estimate $n$ \cite{Tallon}. 

160: Figure 2 shows the $y$ dependence of the thermopower measured 

161: along the $a$-axis at 290 K, $S_a(290{\rm K})$, which is expected 

162: to be free from the contribution of the Cu-O chain transport.

163: The $S_a(290{\rm K})$ data show a continuous change across the 

164: 60-K phase, which strongly suggests that $n$ is continuously 

165: changing with $y$ in our samples.  Therefore, among the two 

166: possibilities listed above, it is more likely that a change in 

167: $n$ is somehow compensated by a change in $\tau$ in the 60-K phase.

168: 

169: \begin{figure}[t!]

170: \epsfxsize=0.8\columnwidth

171: \centerline{\epsffile{Fig3.eps}}

172: \vspace{0.2cm}

173: \caption{$y$ dependences of (a) raw $R_H$ and (b) Hall conductivity 

174: $\sigma_{xy}$ (calculated for $B$=1 T) at 125 and 290 K.}

175: \label{fig3}

176: \end{figure}

177: 

178: Examination of the Hall channel in the in-plane transport 

179: gives us a clue to the (phenomenological) origin of the 

180: anomaly in $\rho_a(T)$.  

181: In YBCO, it is expected that $\sigma_{xy}$ is governed by the 

182: properties of the planes (since the Cu-O chains contribute 

183: little to $\sigma_{xy}$ because of their one-dimensionality), 

184: while the Hall resistivity $\rho_{xy}$ [which is expressed 

185: as $\rho_{xy} \simeq \sigma_{xy}/(\sigma_{xx}\sigma_{yy})$ with 

186: $\sigma_{xx} \simeq 1/\rho_a$ and $\sigma_{yy} \simeq 1/\rho_b$] 

187: is affected by the properties of the chains through $\sigma_{yy}$.

188: Therefore, $\sigma_{xy}$ is a better indicator of the 

189: properties of the planes compared to $R_H$. 

190: Perhaps reflecting this situation, $\sigma_{xy}$ at 125 K shows 

191: a clearly anomalous $y$ dependence while such anomalies are 

192: smeared out in the $y$ dependence of $R_H$ [Figs. 3(a) and 3(b)].

193: The nature of this anomaly in the Hall channel is 

194: best understood by the plot of the Hall mobility in the planes,

195: $\mu_H = \sigma_{xy}/(B\sigma_{xx})$ (Fig. 4), 

196: which reflects $\tau /m^*$ and does not include $n$ in the 

197: Drude picture.  One can clearly see that $\mu_H$ is anomalously 

198: enhanced near $y$=6.65, particularly at 125 K 

199: [where the overlap of $\rho_a(T)$ is observed].

200: Therefore, it appears that the scattering time $\tau$ gets enhanced 

201: upon reducing $y$ from 6.80 to 6.65 and this change in $\tau$ 

202: cancels the change in $n$, causing the $y$-independent resistivity 

203: in the CuO$_2$ planes for $y = 6.65-6.80$.  

204: It should be noted that this anomalous enhancement of $\tau$ takes 

205: place only when the pseudogap opens, and thus the overlap 

206: of $\rho_a(T)$ is observed only below $\sim$130 K.

207: 

208: \begin{figure}[b!]

209: \epsfxsize=0.8\columnwidth

210: \centerline{\epsffile{Fig4.eps}}

211: \vspace{0.2cm}

212: \caption{$y$ dependence of the Hall mobility at 125 and 290 K.}

213: \label{fig4}

214: \end{figure}

215: 

216: \begin{figure}[t!]

217: \epsfxsize=1\columnwidth

218: \centerline{\epsffile{Fig5.eps}}

219: \vspace{0.2cm}

220: \caption{(a) $T$ dependences of $R_H$ for various $y$. 

221: (b), (c): $T$ dependences of $R_H$ and $\sigma_{xy}$ (calculated 

222: for $B$=1 T) for samples with $y=7.00$, 6.75, and 6.50; 

223: the $R_H$ data are normalized by the maximum values for each $y$ 

224: in (b).}

225: \label{fig5}

226: \end{figure}

227: 

228: The Hall effect reveals yet another aspect of the anomalies in 

229: the 60-K phase.

230: Figure 5(a) shows the temperature dependences of $R_H$ (measured with 

231: the current along the $a$-axis) for a wide range of $y$; the data 

232: points in Fig. 3(a) are extracted from this series of data.

233: One may notice in Fig. 5(a) that only the data for $y = 6.65-6.85$ 

234: show noticeable drop in $R_H(T)$ with lowering temperature upon 

235: approaching $T_c$; this situation becomes 

236: obvious in Fig. 5(b), where the $R_H(T)$ data for $y=7.0$, 6.75, 

237: and 6.50 are compared, with clearly indicated $T_c$.

238: It is apparent that only the 60-K-phase sample shows a marked drop 

239: in $R_H(T)$ from well above $T_c$ and the data seem to be heading 

240: towards zero.  To demonstrate that this anomaly is not extrinsically 

241: caused by the effect of the Cu-O chains on $R_H$, we show similar 

242: comparison of $\sigma_{xy}(T)$ for the three compositions [Fig. 5(c)], 

243: which again shows that the anomalous $T$ dependence is observable 

244: (though a bit weakened) only in the 60-K phase sample.

245: This remarkable drop in $R_H$ is somewhat reminiscent of the 

246: $R_H(T)$ behavior in Nd-doped La$_{2-x}$Sr$_{x}$CuO$_{4}$ \cite{Noda}, 

247: where it has been discussed that some peculiarity of the 

248: charge transport \cite{Noda,Emery,Prelovsek} imposed by the 

249: stripes \cite{Tranquada} is responsible for the drop in $R_H$.

250: Thus, the anomaly observed in the $R_H(T)$ behavior is suggestive 

251: of a role of the stripes in the 60-K phase of YBCO.

252: 

253: \begin{figure}[t!]

254: \epsfxsize=1\columnwidth

255: \centerline{\epsffile{Fig6.eps}}

256: \vspace{0.2cm}

257: \caption{(a) Evolution of the fluctuation MC in the 

258: $(T-T_c)/T_c$ vs. $y$ plane. (b) $y$ dependence of 

259: $\xi_{ab}$ obtained from the ALO fits. (c) $y$ dependence of 

260: the mean-field upper critical field calculated from $\xi_{ab}$.}

261: \label{fig6}

262: \end{figure}

263: 

264: We also measured the magnetoresistance in $\rho_a$ in both 

265: the transverse and the longitudinal geometries, from which we 

266: extract orbital magnetoconductivity (MC).  The orbital MC at 

267: high temperatures, whose temperature dependence is very well fitted 

268: with $[bT^2+c]^{-2}$ \cite{Harris,MR}, is due to the normal-state 

269: contribution; the data show additional contribution at lower 

270: temperatures, which can be identified as the Aslamazov-Larkin orbital 

271: (ALO) component of the superconducting fluctuation MC \cite{Harris}.  

272: The full details of the data and their analysis will be 

273: published elsewhere \cite{MR}, and here we show only the fluctuation 

274: MC ($\Delta\sigma_{\rm FL}/\sigma$) obtained after the analysis.  

275: Figure 6(a) shows how the fluctuation MC evolves in the 

276: $(T-T_c)/T_c$ vs. $y$ plane.  

277: One can see that there is a general growth of 

278: $\Delta\sigma_{\rm FL}/\sigma$ with decreasing $y$, but on top of 

279: this trend there is a marked enhancement of 

280: $\Delta\sigma_{\rm FL}/\sigma$ near $y \simeq 6.7$.  

281: An enhancement of $\Delta\sigma_{\rm FL}/\sigma$ means that 

282: the characteristic magnetic-field scale to suppress the 

283: superconducting fluctuations is smaller, which suggests that 

284: $H_{c2}$ is reduced in the 60-K phase. 

285: In fact, when we extract the in-plane coherence length $\xi_{ab}$

286: by fitting the $\Delta\sigma_{\rm FL}/\sigma$ data to the 

287: ALO formula, obtained $\xi_{ab}$ [Fig. 6(b)] shows an 

288: anomalous enhancement in its $y$ dependence near $y \simeq 6.7$;

289: this gives rise 

290: to a reduction of the mean-field upper critical field 

291: $H_{c2}^{\rm MF}$ [$\equiv \Phi_0/(2\pi\xi_{ab}^2)$] as 

292: shown in Fig. 6(c).

293: 

294: \section{Stripes?}

295: 

296: All the above results indicate that the {\it electronic} state 

297: in the CuO$_2$ planes near $y \simeq 6.7$ is somewhat anomalous 

298: compared to other composition and that the 60-K-phase anomalies 

299: are definitely not simply due to oxygen ordering. 

300: The origin of the electronic anomaly in the 60-K phase is 

301: not clear at this stage, but an intriguing possibility is the 

302: charge stripes, because the drop of $R_H$ with decreasing 

303: temperature observed in the 60-K-phase samples is most easily 

304: associated with the stripe physics \cite{Noda,Emery,Prelovsek}.

305: Below we propose a highly speculative scenario, which just 

306: describes one possibility:

307: Suppose that the physics in the whole phase diagram is governed 

308: by the fluctuating stripes or ``electronic liquid crystals" 

309: (as is proposed by Kivelson {\it et al.} \cite{Kivelson}) and that 

310: the superconductivity is caused by the fluctuating stripes.

311: If there is a quantum critical point (QCP) associated with the 

312: stripes near $y \simeq 6.7$ \cite{Kivelson2}, 

313: the superconductivity may be weakened at such QCP (which leads to 

314: a reduction in $H_{c2}$ and a plateau in $T_c$). 

315: Since the fluctuations are always enhanced near QCP, the charge 

316: mobility may also be enhanced.  

317: Thus, though highly speculative, it is possible to 

318: qualitatively understand the observed anomalies in the 60-K phase 

319: in terms of the electronic liquid crystal picture.

320: In the above scenario, the drop in $R_H(T)$ should be due to 

321: the particle-hole symmetry \cite{Emery,Prelovsek} rather than the 

322: suppression of the transverse motion of charges \cite{Noda}. 

323: 

324: \section{Summary}

325: 

326: We have systematically measured the transport 

327: properties of untwinned YBCO single crystals in a wide range of 

328: doping and found that the behaviors of $\rho_a(T)$, 

329: $\mu_{H}(y)$, $R_H(T)$,  $\Delta\sigma_{\rm FL}(y)$, and 

330: $H_{c2}^{\rm MF}(y)$ all show novel anomalies in the 60-K 

331: phase.  These anomalies are clearly of electronic origin, and 

332: possibly related to the stripe physics in this compound.

333: 

334: \section{Acknowledgments}

335: 

336: We thank S. A. Kivelson and A. N. Lavrov for helpful discussions.

337: 

338: % Place here the list of the references:

339: %

340: \medskip

341: \vfil

342: \vspace{-0.8cm}

343: 

344: \begin{references}

345: 

346: \vspace{-1.6cm}

347: 

348: \bibitem[*]{corr}

349: Corresponding author. Fax: +81-3-3480-3401. \\

350: {\it E-mail address:} ando@criepi.denken.or.jp (Y. Ando).

351: 

352: \vspace{0.4cm}

353: 

354: \bibitem{Veal}

355: B. W. Veal and A. P. Paulikas, Physica C 184 (1991) 321, 

356: and references therein.

357: 

358: \bibitem{Tallon}

359: J. L. Tallon, G. V. M. Williams, N. E. Flower, and C. Bernhard, 

360: Physica C 282-287 (1997) 236.

361: 

362: \bibitem{Lavrov}

363: A. N. Lavrov and L. P. Kozeeva, Physica C 253 (1995) 313, 

364: and references therein.

365: 

366: \bibitem{Radaelli}

367: P. G. Radaelli, C. U. Segre, D. G. Hinks, and J. D. Jorgensen,

368: Phys. Rev. B 45 (1992) 4923.

369: 

370: \bibitem{60K}

371: K. Segawa and Y. Ando, Phys. Rev. Lett. 86 (2001) 4907.

372: 

373: \bibitem{Abe}

374: Y. Abe, K. Segawa, and Y. Ando, Phys. Rev. B 60 (1999) R15055.

375: 

376: \bibitem{stripeMR}

377: Y. Ando, A. N. Lavrov, and K. Segawa, Phys. Rev. Lett. 83 (1999) 2813.

378: 

379: \bibitem{Gagnon}

380: R. Gagnon, C. Lupien, and L. Taillefer, Phys. Rev. B 50 (1994) 3458.

381: 

382: \bibitem{Ito}

383: T. Ito, K. Takenaka, and S. Uchida, Phys. Rev. Lett. 70 (1993) 

384: 3995.

385: 

386: \bibitem{Noda}

387: T. Noda, H. Eisaki, and S. Uchida, Science 286 (1999) 265.

388: 

389: \bibitem{Emery}

390: V. J. Emery, E. Fradkin, S. A. Kivelson, and T. C. Lubensky,

391: Phys. Rev. Lett. 85 (2000) 2160.

392: 

393: \bibitem{Prelovsek}

394: P. Prelovsek, T. Tohyama, and S. Maekawa, cond-mat/0102418.

395: 

396: \bibitem{Tranquada}

397: J. M. Tranquada, B. J. Sternlieb, J. D. Axe, Y. Nakamura, and 

398: S. Uchida, Nature 375 (1995) 561.

399: 

400: \bibitem{Harris}

401: J. M. Harris {\it et al.}, Phys. Rev. Lett. 75 (1995) 1391.

402: 

403: \bibitem{MR}

404: Y. Ando and K. Segawa, cond-mat0108054.

405: 

406: \bibitem{Kivelson}

407: S. A. Kivelson, E. Fradkin, and V. J. Emery, Nature 393 (1998) 550.

408: 

409: \bibitem{Kivelson2}

410: S. A. Kivelson (private communication).

411: 

412: \end{references}

413: 

414: \end{document}

415: 

416: