1: %\documentclass[aps,prb,preprint,groupedaddress,showpacs]{revtex4}

2: \documentclass[aps,prb,twocolumn,superscriptaddress]{revtex4}

3: \usepackage{graphicx}

4: 

5: \begin{document}

6: 

7: \title{A local metallic state in globally insulating $La_{1.24}Sr_{1.76}Mn_2O_7$ well above the metal-insulator transition}

8: 

9: \author{Z. Sun}

10: \altaffiliation{zsun@lbl.gov}\affiliation{Department of Physics,

11: University of Colorado, Boulder, CO 80309, USA}\affiliation{Advanced

12: Light Source, Lawrence Berkeley National Laboratory, Berkeley, CA

13: 94720, USA}

14: \author{J. F. Douglas}

15: \affiliation{Department of Physics, University of Colorado, Boulder,

16: CO 80309, USA}

17: \author{ A. V. Fedorov}

18: \affiliation{Advanced Light Source, Lawrence Berkeley National

19: Laboratory, Berkeley, CA 94720, USA}

20: \author{Y. -D. Chuang}

21: \affiliation{Advanced Light Source, Lawrence Berkeley National

22: Laboratory, Berkeley, CA 94720, USA}

23: \author{H. Zheng}

24: \affiliation{Materials Science Division, Argonne National

25: Laboratory, Argonne, IL 60439, USA}

26: \author{J. F. Mitchell}

27: \affiliation{Materials Science Division, Argonne National

28: Laboratory, Argonne, IL 60439, USA}

29: \author{D. S. Dessau}

30: \altaffiliation{Dessau@colorado.edu} \affiliation{Department of

31: Physics, University of Colorado, Boulder, CO 80309, USA}

32: 

33: 

34: \date{\today}

35: 

36: 

37: 

38: \begin{abstract}

39: 

40: \end{abstract}

41: 

42: \pacs{71.18.+y, 79.60.-i}

43: 

44: \maketitle

45: 

46: \textbf{In the spectacularly successful theory of solids, the

47: distinction between metals, semiconductors, and insulators is based

48: upon the behavior of the electrons nearest the Fermi level $E_F$,

49: which separates the occupied from unoccupied electron energy levels.

50: A metal has $E_F$ in the middle of a band of electronic states,

51: while $E_F$ in insulators and semiconductors lies in the gap between

52: states. The temperature-induced transition from a metallic to an

53: insulating state in a solid is generally connected to a vanishing of

54: the low energy electronic excitations \cite{Imada}.  Here we show

55: the first direct evidence of a counter example, in which a

56: significant electronic density of states at the Fermi energy exists

57: in the insulating regime. In particular, angle-resolved

58: photoemission data from the ``colossal magnetoresistive" oxide

59: $La_{1.24}Sr_{1.76}Mn_2O_7$ show clear Fermi edge steps both below

60: the $T_C$ when the sample is globally metallic, as well as above

61: $T_C$ when it is globally insulating. Further, small amounts of

62: metallic spectral weight survive up to the temperature scale $T^*$

63: more than twice the $T_C$ of the system. Such behavior also may have

64: close ties to a variety of exotic phenomena in correlated electron

65: systems including in particular the pseudogap scale T* in underdoped

66: cuprates \cite{Timusk}.}

67: 

68: 

69: As shown in figure \ref{fig1}a, the colossal magnetoresistive (CMR)

70: oxide $La_{2-2x}Sr_{1+2x}Mn_2O_7$ (x=0.38) exhibits a metal

71: insulator transition at a $T_C$ of about 130K, at which point the

72: system also switches from being a ferromagnet (low $T$) to a

73: paramagnet (high $T$) \cite{LiQA}.  We performed angle-resolved

74: photoemission spectroscopy (ARPES) experiments on cleaved single

75: crystals of these materials, with an experimental arrangement as

76: described elsewhere \cite{Sun}. ARPES is an ideal experimental probe

77: of the electronic structure since it gives the momentum-resolved

78: single-particle excitation spectrum. As discussed in ref. 4 the

79: x=0.38 compound studied here does not contain the low energy

80: pseudogap of the x=0.4 samples \cite{DessauPRL, ChuangScience,

81: Saitoh, Mannella}(see supplementary material for more details on

82: this, the possible issue of surface sensitivity of ARPES, and of

83: possible intergrowths at the surface). The much larger metallic

84: spectral weight of these non-pseudogapped compounds also allows us

85: to study the electronic behavior in greater detail.

86: 

87: \begin{figure}[tbp]

88: \begin{center}

89: \includegraphics[width=1\columnwidth,angle=0]{fig1.eps}

90: \end{center}

91: \caption{Overview of features of $La_{1.24}Sr_{1.76}Mn_2O_7$. (a)

92: Resistivity vs. temperature, after ref 3. (b) A representative Fermi

93: surface. (c) Low temperature (20K) ARPES data over a

94: large-energy-scale taken along the black cut near the zone boundary,

95: as shown in (b).} \label{fig1}

96: \end{figure}

97: 

98: Figure \ref{fig1}c shows a large-energy-scale experimental picture

99: of a low temperature $d_{x^2-y^2}$ symmetry band taken along the

100: black cut near the zone boundary, as shown in figure \ref{fig1}b. We

101: are able to get clean data by isolating the various bilayer-split

102: bands using different photon energies, as described in ref. 4. In

103: particular, in this paper we only show data from the antibonding

104: bilayer-split band which has Fermi crossings at $k_x$=$\pm$0.17

105: $\pi/a$, $k_x$=0.9 $\pi/a$, corresponding to the solid Fermi surface

106: in figure \ref{fig1}b. The energy distribution curves (EDCs) at

107: $k_F$ (indicated by the red line in figure \ref{fig1}c) taken at a

108: series of temperatures are shown in figure \ref{fig2}a. Figure

109: \ref{fig2}b shows the identical spectra and identical scaling, but

110: offset vertically for clarity. All spectra have been normalized only

111: to the incident photon flux.

112: 

113: \begin{figure}[tbp]

114: \begin{center}

115: \includegraphics[width=1\columnwidth,angle=0]{fig2.eps}

116: \end{center}

117: \caption{Energy Distribution Curves (EDCs) as a function of

118: temperature at $k_F$ (red line of figure \ref{fig1}c), indicating

119: metallic spectral weight above $T_C$. (a, b) the same data set

120: scaled by the incident flux and are taken while warming. (c) EDCs

121: from a different sample taken at the high temperature range.  Clear

122: breaks are seen in the spectral intensity near $E_F$ for all but the

123: highest temperature, indicating finite metallic spectral weight and

124: a $T^*$ just above 285K (see figure \ref{fig21} for details of the

125: $T^*$ determination).} \label{fig2}

126: \end{figure}

127: 

128: At low temperature, the EDCs clearly show a structure of

129: peak-dip-hump, where the peak and the hump would nominally be

130: considered the coherent part (quasiparticle) and ``incoherent" part

131: of the single particle spectrum respectively, as has been discussed

132: for the spectra of the high $T_C$ cuprate superconductors

133: \cite{Dessau, ShenDessau, Damascelli}. One sees that the near-$E_F$

134: spectral weight diminishes with increasing temperature, while the

135: high binding energy ($>$700meV) part is less affected by

136: temperature. Contrary to the general picture of the metal-insulator

137: transition, in which a gap develops in the single particle spectrum

138: when an electronic system becomes insulating \cite{Imada}, the EDCs

139: here still exhibit a sharp Fermi cutoff indicating metallic behavior

140: at temperatures in which the macroscopic DC conductivity is

141: characteristic of insulation (e.g. the spectra at 135, 150 and

142: 180K). To our knowledge, this unusual behavior, a metallic Fermi

143: edge in a globally insulating system, has not been previously

144: observed on the insulating side of a metal-insulator transition. The

145: opposite, in which a metallic system shows a lack of a Fermi cutoff,

146: is on the other hand expected in exotic low-dimensional systems such

147: as the Luttinger Liquids \cite{Voit}, and has likely been observed

148: \cite{Allen}.  The other situation most likely to show a metallic

149: Fermi edge in a globally insulating system is that of an

150: Anderson-localized system beyond the mobility edge. However, in such

151: systems a Coulomb gap is expected to remove the metallic weight near

152: the Fermi energy \cite{Varma}. Our data could only be consistent

153: with such a scenario if the Coulomb gap were extremely small - on

154: the order of a few meV or less.  Such a picture also would not

155: naturally explain the metallic spectral weight dependence with

156: temperature, to be discussed in more detail later in the letter.

157: 

158: \begin{figure}[tbp]

159: \begin{center}

160: \includegraphics[width=0.7\columnwidth,angle=0]{fig21.eps}

161: \end{center}

162: \caption{Determination of $T^*$.  Zero intensity intercepts as a

163: function of temperature from the linear fits to the data shown in

164: figure \ref{fig2}c.  These intercepts go to zero at 305K, which we

165: label as $T^*$ $\sim$ 305K.} \label{fig21}

166: \end{figure}

167: 

168: 

169: On a different sample we have done higher temperature scans, looking

170: for a possible temperature scale at which the metallic spectral

171: weight disappears. These data are shown in figure \ref{fig2}c and

172: show a clear discontinuity in the slope near the Fermi energy for

173: all but the 285K data, indicating a finite metallic spectral weight.

174: This effect is emphasized by an extrapolation of the spectral weight

175: using a simple linear fit to the data between -0.3 and -0.05 eV, as

176: shown by the dotted lines in the figure. Upon raising the sample

177: temperature we see that the intercept of these dotted lines with the

178: horizontal axis decreases at an approximately linear rate (figure

179: \ref{fig21}).  As shown in this figure these zero intensity

180: intercepts reach the Fermi energy at 305 $\pm$10K. We thus indicate

181: 305K as the temperature at which the first bits of metallic weight

182: become apparent, which we indicate as the temperature $T^*$.

183: Technical reasons including sample aging and excessive manipulator

184: drift preclude us from making the full range of measurements on a

185: single cleave. We therefore used different samples to study the

186: electronic excitations in different temperature regimes.

187: 

188: 

189: Figure \ref{fig3}a shows the electronic dispersion of the near-Fermi

190: states as a function of temperature obtained from an analysis of

191: momentum distribution curves (MDCs).  This data indicates that the

192: main properties of the metal, such as the Fermi wave vector $k_F$,

193: the Fermi velocity $v_F$, the electron phonon coupling parameter

194: $\lambda$, and the effective mass $m^*$ don't change significantly

195: as a function of temperature, even as the metal-insulator transition

196: temperature $T_C$ is traversed.  This is an unexpected behaviour for

197: a metal-insulator transition in which these parameters would vary

198: dramatically with temperature, and likely even diverge \cite{Imada}.

199: 

200: \begin{figure}[tbp]

201: \begin{center}

202: \includegraphics[width=1\columnwidth,angle=0]{fig3.eps}

203: \end{center}

204: \caption{Properties of the metallic portion of the sample.  (a)

205: Electronic dispersion showing a similar $k_F$, $v_F$ and

206: electron-phonon coupling   as a function of temperature.  (b, c)

207: Metallic EDCs or M-EDCs obtained by subtracting the 180K EDC from

208: all lower temperature data. (b) shows the raw scaling while (c)

209: scales each spectrum to have a similar max intensity. (d) MDC widths

210: (green triangles) and integrated M-EDC spectral weights (blue

211: squares) as a function of temperature.} \label{fig3}

212: \end{figure}

213: 

214: 

215: Our data can be understood by invoking a model of disconnected local

216: ferromagnetic metallic regimes above  $T_C$ up to approximately the

217: temperature $T^*$. This suggestion is consistent with earlier

218: studies which have found significant ferromagnetic signals far above

219: $T_C$ \cite{ArgyriouJAP,Osborn,Rosenkranz1,Rosenkranz2}, since

220: metallicity and ferromagnetism should have a connection in these

221: systems via the double-exchange interaction. In general, the

222: metallic regions may be either phase separated (and possibly static)

223: domains \cite{Uehara,Dagotto} or they may be dynamic fluctuations of

224: the ferromagnetic metallic state, which in a two-dimensional system

225: may persist to quite high temperatures

226: \cite{ArgyriouJAP,Osborn,Rosenkranz1,Rosenkranz2}. We will discuss

227: these two possibilities later in this letter. Here we show that we

228: can study the metallic portions further by our ability to

229: approximately deconvolve the spectrum into the components which

230: arise from the metal and non-metal portions. We do this by

231: subtracting the 180K EDC from all other EDCs as shown in figure

232: \ref{fig2}b to create ``metallic EDCs" or M-EDCs as shown in figure

233: \ref{fig3}b. It should be pointed out that the slight variation of

234: spectra from sample to sample, which has been commonly observed in

235: ARPES, imperils the practice of extracting data of one sample from

236: that of another. Therefore, we don't use the higher temperature data

237: of figure \ref{fig2}c to do the subtraction as this is from a

238: different sample. Figure \ref{fig3}c shows the same M-EDCs but

239: scaled to all have the same amplitude. Within the noise, all the

240: M-EDCs have similar lineshapes with coherent peaks near $E_F$ and an

241: incoherent background at high binding energy, though the widths of

242: the M-EDC coherent peak (or low energy MDC peak) become broader with

243: increasing temperature (figure \ref{fig3}d). The integrated spectral

244: weight of the M-EDCs varies smoothly as a function of temperature,

245: with no clear break at $T_C$ (figure \ref{fig3}d). This, as well as

246: the approximate temperature-independence of the M-EDC lineshape

247: indicates that the electrons in the metallic regions have similar

248: properties above and below $T_C$, and that temperature has

249: surprisingly little effect on the behavior or interactions of

250: electrons in the metallic regions. This is consistent with the

251: approximate independence of $v_F$, $\lambda$, and $m^*$ in the

252: metallic regions shown in figure \ref{fig3}a. The experimentally

253: determined MDC width of the electrons at the Fermi energy (green

254: triangles of figure \ref{fig3}d) does broaden with increasing

255: temperature. The inverse of this quantity, the mean free path of the

256: electrons, thus decreases with increasing temperature, consistent

257: with a decreased size of metallic regions or increased scattering

258: events at higher temperatures.

259: 

260: While many aspects of our data are consistent with either the phase

261: separation or magnetic fluctuation picture, certain aspects of it

262: can address the question of whether the metallic regions above $T_C$

263: are phase-separated out from a more insulating environment

264: \cite{Uehara,DagottoBook}, or if they are just fluctuations from a

265: lower temperature ordered environment which is otherwise homogeneous

266: \cite{Rosenkranz1,Rosenkranz2}. In particular, the smooth dependence

267: of the spectral weight of the metallic regions as a function of

268: temperature across $T_C$ (blue squares of figure \ref{fig3}d) is

269: more consistent with phase separation, as we would likely expect a

270: clear drop in the metallic weight near $T_C$ if the metallic

271: portions were just fluctuations of the ordered lower temperature

272: environment. At other doping levels (for example $x$=0.4),

273: experiments do observe a sharp drop in the metallic weight at $T_C$

274: to zero or almost zero \cite{Mannella}, and so those samples may be

275: more consistent with the fluctuation physics.

276: 

277: Within the picture of phase separation, we imagine that the metallic

278: islands arise at a temperature $T^*$ near room temperature, which

279: also may be related to the temperature scale at which polaronic

280: correlations freeze \cite{ArgyriouPRL}.  As the temperature is

281: lowered the size and proportion of metallic portions grows until a

282: critical ratio of metallic to insulating portions is reached.  At

283: that point electrons can percolate from one metallic region to

284: another, bringing about the macroscopic metallic \cite{Uehara} and

285: ferromagnetic states, as well as being consistent with the

286: ``colossal" decrease in resistivity with an applied magnetic field.

287: In certain models this behavior is expected from a competition

288: between different phases, for example between the ferromagnetic

289: metal phase and the charge-ordered antiferromagnetic insulating

290: phase \cite{Uehara, Dagotto, Tokura}, though in contrast to ref 19,

291: the materials used here are far away from the charge-ordered doping

292: level. Theoretical arguments predict both the phase separation and

293: the existence of a higher temperature scale $T^*$  \cite{Burgy},

294: with ideas similar to the Griffiths singularity \cite{Griffiths}  in

295: which $T^*$ would be the critical temperature of the associated

296: clean system in the absence of disorder, and which have recently

297: been discussed in the context of manganite physics \cite{Burgy,

298: Salamon}. We are presently undertaking a more thorough study of the

299: full Fermi surface to test this percolation model quantitatively.

300: 

301: A $T^*$ scale is one of the key properties of the high $T_C$

302: superconductors, and has for years been the subject of intense

303: controversy \cite{Timusk}. In these compounds, disorder also appears

304: to be highly relevant, especially in the underdoped regime where the

305: $T^*$ scale exists. In that case it signals the emergence of the

306: pseudogap, which may be the precursor to the long range

307: superconducting order which forms at $T_C$ \cite{Emery} -- a clear

308: analogy to the manganites where $T^*$ signals the emergence of the

309: metallic domains which become long range at $T_C$. Also similar to

310: the cuprates, it appears that the $T^*$ temperature scales may not

311: be universal to all doping levels of the manganites. Pinning these

312: details down and then understanding their implications will

313: certainly be an area of intense study in the near future.

314: 

315: 

316: It is becoming increasingly clear that some of the most dramatic

317: responses in modern materials occur in systems in which multiple

318: phases or orders with similar energy scales compete with each other

319: \cite{Dagotto, Tokura, Burgy, Murakami}.  It is then natural that in

320: at least some of these systems spatial heterogeneities will occur,

321: and small perturbations can cause drastic macroscopic alterations to

322: the physical properties or even new types of  ``emergent" behavior.

323: The key is finding which aspects of the inhomogeneity are intrinsic

324: and what is their role in determining the key physical properties of

325: the system.

326: 

327: %In summary, using ARPES we show a new type of metal-insulator

328: %transition in solids in the colossal magnetoresistive oxide

329: %$La_{1.24}Sr_{1.76}Mn_2O_7$. In contrast to known cases, the unusual

330: %finding is that while $La_{1.24}Sr_{1.76}Mn_2O_7$ is globally

331: %insulating it is locally still a metal, with the local metal

332: %behavior forming at an even higher temperature -- $T^*$. Our data

333: %indicates the important roles of phase separation and percolation

334: %effects.

335: 

336: The authors thank Y. Tokura and T. Kimura for providing preliminary

337: samples and are grateful to D. N. Argyriou, A. Bansil, E. Dagotto,

338: K. Gray, A. Moreo, R. Osborn, L. Radzihovsky, D. Reznik, S.

339: Rosenkranz, Y. Tokura, and M. Veillette for helpful discussions.

340: This work was supported by the U.S. Department of Energy under grant

341: DE-FG02-03ER46066 and by the U.S. National Science Foundation grant

342: DMR 0402814. The ALS is operated by the Department of Energy, Office

343: of Basic Energy Sciences. Argonne National Laboratory, a U.S.

344: Department of Energy Office of Science Laboratory, is operated under

345: Contract No. DE-AC02-06CH11357. The U.S. Government retains for

346: itself, and others acting on its behalf, a paid-up nonexclusive,

347: irrevocable worldwide license in said article to reproduce, prepare

348: derivative works, distribute copies to the public, and perform

349: publicly and display publicly, by or on behalf of the Government.

350: 

351: \begin{thebibliography}{99}

352: 

353:  \bibitem{Imada}M. Imada, A. Fujimori, and Y. Tokura, Rev. Mod. Phys. \textbf{70}, 1039 (1998).

354: 

355:  \bibitem{Timusk}T. Timusk and B. Statt, Rep. Prog. Phys. \textbf{62}, 61 (1999).

356: 

357:  \bibitem{LiQA}Q. A. Li, et al., Phys. Rev. B \textbf{59}, 9357

358:  (1999).

359: 

360:  \bibitem{Sun}Z. Sun et al., Phys. Rev. Lett. \textbf{97}, 056401 (2006).

361: 

362:  \bibitem{DessauPRL}D. S. Dessau et al., Phys. Rev. Lett. \textbf{81}, 192 (1998).

363: 

364:  \bibitem{ChuangScience}Y.-D. Chuang et al., Science \textbf{292}, 1509 (2001).

365: 

366:  \bibitem{Saitoh}T. Saitoh et al., Phys. Rev. B \textbf{62}, 1039 (2000).

367: 

368:  \bibitem{Mannella}N. Mannella et al., Nature \textbf{438}, 474 (2005).

369: 

370:  \bibitem{Dessau}D. S. Dessau et al.,  Phys. Rev. Lett. \textbf{66}, 2160 (1991).

371: 

372:  \bibitem{ShenDessau}Z. -X. Shen and D. S. Dessau, Phys. Rep. \textbf{253}, 1 (1995).

373: 

374:  \bibitem{Damascelli}A. Damascelli, Z. Hussain, and Z. -X. Shen, Rev. Mod. Phys. \textbf{75}, 473 (2003).

375: 

376:  \bibitem{Voit}J. Voit, Rep. Prog. Phys. \textbf{57}, 977 (1995).

377: 

378:  \bibitem{Allen}J.W. Allen, Sol. State Comm. \textbf{123}, 469 (2002).

379: 

380:  \bibitem{Varma}C.M. Varma, Phys. Rev. B \textbf{54}, 7328 (1996).

381: 

382:  \bibitem{ArgyriouJAP}D. N. Argyriou, et al., J. Appl. Phys. \textbf{83}, 6374 (1998).

383: 

384:  \bibitem{Osborn}R. Osborn, et al., Phys. Rev. Lett. \textbf{81}, 3964

385:  (1998).

386: 

387:  \bibitem{Rosenkranz1}S. Rosenkranz, et al., cond-mat/9909059.

388: 

389:  \bibitem{Rosenkranz2}S. Rosenkranz, et al., Physica B \textbf{312-313}, 763-765 (2002).

390: 

391:  \bibitem{Uehara}M. Uehara, et al., Nature \textbf{399}, 560 (1999).

392: 

393: \bibitem{DagottoBook}E. Dagotto, Nanoscale Phase Separation and Colossal Magnetoresistanc (Springer Verlag, 2003).

394: 

395:  \bibitem{ArgyriouPRL}D.N. Argyriou, et al., Phys. Rev. Lett. \textbf{89}, 36401 (2002).

396: 

397:  \bibitem{Dagotto}E. Dagotto, Science \textbf{309}, 257 (2005).

398: 

399:  \bibitem{Tokura}Y. Tokura, Rep. Prog. Phys. \textbf{69}, 797 (2006).

400: 

401:  \bibitem{Burgy}J. Burgy, et al., Phys. Rev. Lett. \textbf{87}, 277202 (2001).

402: 

403:  \bibitem{Griffiths}R.B. Griffiths, Phys. Rev. Lett. \textbf{23}, 17 (1969).

404: 

405:  \bibitem{Salamon}M. B. Salamon, et al., Phys. Rev. Lett. \textbf{88}, 197203 (2002).

406: 

407:  \bibitem{Emery}V.J. Emery, and S.A. Kivelson, Nature \textbf{374}, 434 (1995).

408: 

409:  \bibitem{Murakami}S. Murakami, and N. Nagaosa, Phys. Rev. Lett. \textbf{90}, 197201 (2003).

410: 

411: 

412: \end{thebibliography}

413: 

414: \textbf{Supplementary discussion:}

415: 

416: 1. \emph{The difference between $x$=0.38 and $x$=0.4 samples.} It

417: should be pointed out that there is a remarkable difference between

418: the ARPES spectra of $La_{1.24}Sr_{1.76}Mn_2O_7$ and

419: $La_{1.2}Sr_{1.8}Mn_2O_7$ samples, even though many macroscopic

420: properties are similar. Quasiparticles have been found near the zone

421: boundary at the doping levels of $x$=0.36 and 0.38 in

422: $La_{2-2x}Sr_{1+2x}Mn_2O_7$, while there exists a large energy

423: pseudogap in $x$=0.40 samples

424: \cite{Sun,DessauPRL,ChuangScience,Saitoh,Mannella}. Temperature

425: dependent studies have also been performed on $x$=0.4 samples and

426: have not shown evidence for metallic spectral weight above $T_C$

427: \cite{ChuangScience,Saitoh,Mannella}. Similar to high-$T_C$

428: cuprates, physical properties exhibit strong variations with doping

429: in manganites. The cause of the difference between

430: $La_{2-2x}Sr_{1+2x}Mn_2O_7$ ($x$=0.38) and

431: $La_{2-2x}Sr_{1+2x}Mn_2O_7$ ($x$=0.40) samples is not understood

432: yet, though it could have to do with the increased lattice anomalies

433: for the 0.4 samples [J. Mitchell et al., J. Phys. Chem. B 105, 10731

434: (2001).], the onset of spin canting between ferromagnetic layers

435: which starts at the doping level of 0.4 [M.Kubota et.al., J. Phys.

436: Soc. Jpn 69, 1606 (2000).],  or even something extrinsic such as a

437: surface issue.

438: 

439: 2. \emph{The issue of surface sensitivity.} Because of the shallow

440: probing depth of the ARPES experiment ($\sim$ 5-10 Angstroms), we

441: cannot completely rule out the potential that a surface phase whose

442: properties do not follow those of the bulk gives rise to some of the

443: phenomena reported here. For ARPES on the layered manganites we are

444: relatively well off since the samples cleave readily between the

445: $La,Sr-O$ bilayers, which are ionically (not covalently) bonded.

446: High quality LEED pictures without any evidence of surface

447: reconstruction are obtained from these surfaces. The doping level at

448: the surfaces, as obtained from the Fermi surface volume also appears

449: to be correct for these samples - for example the $d_{x^2-y^2}$

450: bonding band Fermi surface nesting vector of $0.27\times(2\pi/a)$

451: for the $x=0.38$ samples used in this study \cite{Sun} exactly

452: matches that obtained from neutron scattering measurements

453: \cite{ArgyriouPRL}. The nesting vector of 0.4 samples is slightly

454: larger at $0.3\times(2\pi/a)$ \cite{ChuangScience} and also matches

455: the results of scattering measurements [L. Vasiliu-Doloc, et al.

456: Phys. Rev. Lett. 83, 4393 (1999).].

457: 

458: 3. \emph{The issue of intergrowths.} One should consider whether it

459: might be possible for the metallic spectral weight far above $T_C$

460: to have originated from small bits of intergrowth (IG) of a higher

461: $T_C$ sample left near the surface after cleaving.  Here we discuss

462: why that is inconsistent with our data. Such intergrowths should not

463: arise from a layered manganite, as the maximum temperature at which

464: bulk metallic behavior is found among all known layered manganites

465: is $\sim$ 160K. A small amount of (non-layered) perovskite-like IG

466: with a $T_C$ = 300K could exist at a cleaved surface, though would

467: not show the bilayer splitting since the perovskite samples have

468: only one $MnO_2$ plane per unit cell. Both our high and low

469: temperature data display this bilayer band splitting (this paper

470: only presents the data from the antibonding component), which is a

471: direct consequence of having two $MnO_2$ planes per unit cell. In

472: addition to being able to vary the intensity of the bilayer split

473: bands (relative and overall) by taking advantage of the

474: photoemission matrix elements, we can follow the dispersion in $E$

475: and $k$ of each of the bilayer bands, including tracking them all

476: the way to $E_F$. Therefore we know the origin of the metallic

477: weight as explicitly originating from these bilayer split bands, and

478: therefore, from the bilayer manganite.

479: 

480: \end{document}

481: