1: % INJ

2: 

3: %\documentclass[aps,superscriptaddress,showpacs,preprint]{revtex4}

4: \documentclass[aps,superscriptaddress,showpacs,twocolumn]{revtex4}

5: %\documentclass[aps,prb,superscriptaddress,showpacs,twocolumn]{revtex4}

6: %\documentclass[aps,superscriptaddress,showpacs,preprint]{revtex4}

7: \usepackage{graphicx}

8: \usepackage{color}

9: \usepackage{amssymb}

10: \usepackage{amsmath}

11: 

12: \newcommand{\bk}{{\bf k}}

13: \newcommand{\bq}{{\bf q}}

14: \newcommand{\bQ}{{\bf Q}}

15: \newcommand{\bp}{{\bf p}}

16: \newcommand{\bx}{{\bf x}}

17: \newcommand{\br}{{\bf r}}

18: \newcommand{\Li}{{\mathop{\rm{Li}}\nolimits}}

19: \renewcommand{\Im}{{\mathop{\rm{Im}}\nolimits\,}}

20: \renewcommand{\Re}{{\mathop{\rm{Re}}\nolimits\,}}

21: \newcommand{\sgn}{{\mathop{\rm{sgn}}\nolimits\,}}

22: \newcommand{\Tr}{{\mathop{\rm{Tr}}\nolimits\,}}

23: \newcommand{\EF}{E_{\mathrm{F}}}

24: \newcommand{\kB}{k_{\mathrm{B}}}

25: \newcommand{\Green}{{\mathcal G}}

26: \newcommand{\dSC}{{\mathrm{dSC}}}

27: \newcommand{\dPG}{{\mathrm{dPG}}}

28: \newcommand{\dDW}{{\mathrm{dDW}}}

29: \newcommand{\Ret}{{\mathrm{R}}}

30: \newcommand{\Tau}{T_\tau}

31: 

32: %\newcommand{\modified}[1]{\textcolor{red}{#1}}

33: %\newcommand{\added}[1]{\textcolor{red}{#1}}

34: %\newcommand{\deleted}[1]{\textcolor{yellow}{#1}}

35: %\newenvironment{emodified}{\color{red}}{\normalcolor}

36: 

37: \newcommand{\modified}[1]{\relax #1}

38: \newcommand{\added}[1]{\relax #1}

39: \newcommand{\deleted}[1]{\relax #1}

40: \newenvironment{emodified}{\relax}{\relax}

41: 

42: 

43: \begin{document}

44: 

45: \title{Comment on ``Variation of the superconducting transition

46:    temperature of hole-doped copper oxides''.}

47: 

48: 

49: 

50: \author{G. G. N. Angilella}

51: \email{Giuseppe.Angilella@ct.infn.it}

52: \affiliation{Dipartimento di Fisica e Astronomia, Universit\`a di

53:    Catania,\\ and Istituto Nazionale per la Fisica della Materia,

54:    UdR di Catania,\\ Via S. Sofia, 64, I-95123 Catania, Italy}

55: \author{R. Pucci}

56: \affiliation{Dipartimento di Fisica e Astronomia, Universit\`a di

57:    Catania,\\ and Istituto Nazionale per la Fisica della Materia,

58:    UdR di Catania,\\ Via S. Sofia, 64, I-95123 Catania, Italy}

59: \author{A. Sudb\o}

60: \affiliation{ Department of Physics, Norwegian University of Science

61:    and Technology, N-7491 Trondheim, Norway}

62: 

63: \date{\today}

64: 

65: 

66: \begin{abstract}

67: \medskip

68: We point out the incorrect derivation of the gap equation in X.-J. Chen

69:    and H. Q. Lin [Phys. Rev. B {\bf 69}, 104518 (2994)] within the

70:    interlayer tunneling (ILT) model for multilayered cuprates. 

71: There, the \emph{local} structure in $\bk$-space of the ILT effective

72:    interaction has not been taken into due account when the ILT model

73:    is generalized to the case of $n$ layers per unit cell. 

74: This is a specific characteristic of the ILT model that, apart from giving

75:    rise to a highly nontrivial $\bk$-dependence of the gap function, is

76:    known to enhance the critical temperature $T_c$ in a natural way.

77: As a consequence, we argue that Chen and Lin's results cannot be

78:    employed, in their present form, for a quantitative interpretation

79:    of the high-pressure dependence of $T_c$ in Bi-2212, as is done by

80:    X.-J. Chen \emph{et al.} [{\tt cond-mat/0408587}, to appear in

81:    Phys. Rev. B].

82: Moreover, when the generalization of  X.-J. Chen \emph{et al.} [{\tt

83:    cond-mat/0408587}] is applied to the case $n=2$, it fails to

84:    reproduce the original ILT gap equation. 

85: However, a more careful analysis of the ILT model for multilayered

86:    cuprates, taking into account the nonuniform hole distribution

87:    among inequivalent layers, has been earlier suggested to describe 

88:    the observed pressure dependence of $T_c$ in homologous series of

89:    high-$T_c$ cuprates.

90: \\

91: \pacs{% 

92: 74.62.-c, % Tc variations

93: 74.72.-h, % Cuprate superconductors (high-Tc and insulating parent compounds)

94: %74.72.Hs, % Bi-based cuprates

95: 74.62.Fj, % pressure effects

96: 74.20.-z  % Theories and models of superconducting state

97: %Theories and models of superconducting state

98: }

99: \end{abstract} 

100: 

101: \maketitle

102: 

103: %\section{Introduction}

104: 

105: In Ref.~\onlinecite{Chen:04}, Chen and Lin reconsider the dependence

106:    of $T_c$ on doping and on the number of layers in a homologous series of

107:    multilayered high-$T_c$ cuprates within the interlayer tunneling

108:    (ILT) model \cite{Chakravarty:93}. 

109: However, in deriving their gap equation, Chen and Lin erroneusly

110:    neglect the intrinsic \emph{local} structure in momentum ($\bk$)

111:    space of the effective ILT coupling.

112: This is a specific characteristic of the ILT model, which is known to give

113:    rise to highly nontrivial features in the $\bk$-dependence of the

114:    gap function already for a bilayer complex

115:    \cite{Angilella:99,Angilella:00}.

116: Moreover, a local term in the gap equation has been shown to provide a

117:    lower bound for $T_c$ at all dopings, which is the precise way in

118:    which the ILT mechanism enhances $T_c$ \cite{Angilella:99}.

119: The consequences of such an incorrect analysis of the ILT model are

120:    both qualitative and quantitative.

121: Therefore, the recent use of Chen and Lin's results to interpret the

122:    high-pressure dependence of $T_c$ in Bi-2212 \cite{Chen:04a} can be

123:    questioned.

124: In this context, we point out that a more careful analysis of the ILT

125:    model for layered cuprates has been presented elsewhere

126:    \cite{Sudboe:94c}, and successfully applied to study the pressure

127:    dependence of $T_c$ in homologous series of layered cuprates, by

128:    explicitly taking 

129:    into account the inhomogeneous hole-doping in inequivalent

130:    layers \cite{Angilella:99b,Wijngaarden:99}.

131: 

132: Superconductivity in the high-$T_c$ layered cuprates is characterized

133:    by \emph{(i)} a non-monotonic dependence of $T_c$ on the overall hole-doping

134:    $\delta$; \emph{(ii)} a monotonic increase of $T_c$ with the number of

135:    layers $n$, for moderately low $n$ ($n\lesssim 3$).

136: While \emph{(i)} is a generic consequence of the quasi-bidimensional nature

137:    of these compounds (see \emph{e.g.} Ref.~\onlinecite{Angilella:01}), the

138:    latter fact has suggested that coherent tunneling of

139:    superconducting pairs between adjacent CuO$_2$ layers may

140:    considerably enhance $T_c$ \cite{Chakravarty:93}.

141: Within the ILT model, it is postulated that strong in-plane

142:    correlations forbid coherent hopping of single

143:    particles between adjacent CuO$_2$ planes.

144: Such a restriction is removed when accessing the superconducting

145:    state, where interlayer Josephson tunneling of Cooper pairs is

146:    allowed.

147: This results in a net gain in kinetic energy, as compared to the

148:    normal state.

149: Thus, within the ILT model, superconductivity is stabilized \emph{via} a

150:    kinetic mechanism, as opposed to conventional BCS

151:    superconductivity, where the enhancement in kinetic energy is

152:    overcompensated by a reduction in the potential energy

153:    \cite{Chakravarty:98a}.

154: 

155: However, after its original formulation more than a decade ago

156:    \cite{Chakravarty:93}, the relevance of the ILT mechanism at least

157:    for single-layer cuprates has been called into question by

158:    experiments \cite{Anderson:98,Moler:98}.

159: Recently, Chakravarty \emph{et al.} \cite{Chakravarty:04} have revived

160:    the ILT model in connection with multilayered cuprates.

161: There, ILT needs not be the sole source of superconducting

162:    condensation energy.

163: Charge carriers require a `seed' in-plane interaction to form Cooper

164:    pairs in a given symmetry channel, before they can actually tunnel

165:    between adjacent layers \cite{Angilella:99}.

166: Such in-plane interaction would then provide the missing condensation

167:    energy \cite{Chakravarty:04}.

168: 

169: Moreover, it has been suggested that the competition with a `hidden'

170:    order parameter, such as a $d$-density-wave (dDW)

171:    \cite{Chakravarty:01}, could be responsible for the downturn of $T_c$

172:    with $n$, for $n\gtrsim 3$.

173: Indeed, in multilayered cuprates, due to the different proximity to

174:    the `charge reservoir' blocks, experiments \cite{Trokiner:91} as

175:    well as density functional theory calculations

176:    \cite{Ambrosch-Draxl:04} revealed a nonuniform hole-content

177:    distribution between inner and outer layers.

178: Since this usually places inner (outer) layers in the underdoped

179:    (overdoped) region of the cuprate phase diagram, competition with

180:    the dDW order would be stronger in inner layers than in outer

181:    layers, thus depressing $T_c$ with increasing $n$.

182: Hydrostatic pressure could then be used to tune both the overall

183:    hole-content content and its distribution among inequivalent

184:    layers, thus inducing an `exchange of roles' between inner and outer

185:    layers with respect to the onset of superconductivity

186:    \cite{Angilella:99b}, which is observed as `kinks' in the

187:    pressure dependence of $T_c$ in layered cuprates

188:    \cite{Wijngaarden:99}.

189: 

190: The effective Hamiltonian considered by Chen and Lin \cite{Chen:04}

191:    (see also \cite{Angilella:99,Angilella:99b}) is

192: \begin{eqnarray}

193: H &=& \sum_{\ell \bk\sigma} \xi_\bk c_{\bk\sigma}^{\ell\dag}

194:    c_{\bk\sigma}^\ell - \sum_{\ell\bk\bk^\prime} V_{\bk\bk^\prime}

195:    c_{\bk\uparrow}^{\ell\dag} c_{-\bk\downarrow}^{\ell\dag}

196:    c_{-\bk^\prime \downarrow}^\ell c_{\bk^\prime \uparrow}^\ell

197: \nonumber\\

198: &&+ \sum_{\langle\ell\ell^\prime \rangle} \sum_\bk T_J (\bk)

199:    c_{\bk\uparrow}^{\ell\dag} c_{-\bk\downarrow}^{\ell\dag}

200:    c_{-\bk\downarrow}^{\ell^\prime} c_{\bk\uparrow}^{\ell^\prime} ,

201: \label{eq:H}

202: \end{eqnarray}

203: where $\xi_\bk$ is the in-plane quasiparticle dispersion measured with

204:    respect to the chemical potential $\mu^\ell$ ($\mu^\ell \equiv \mu$

205:    for all layers, in Ref.~\onlinecite{Chen:04}), and

206:    $c_{\bk\sigma}^{\ell\dag}$ 

207:    is a quasiparticle creation operator with wave-vector $\bk$ and spin

208:    $\sigma$ on layer $\ell$.

209: It should be emphasized that in Eq.~(\ref{eq:H}) the first interaction

210:    term ($V_{\bk\bk^\prime}$) pertains to a single layer and governs the

211:    overall symmetry of the order parameter (\emph{i.e.}, $d$-wave, if

212:    $V_{\bk\bk^\prime} = V g_\bk g_{\bk^\prime}$, with $g_\bk =

213:    \frac{1}{2} (\cos   k_x - \cos k_y )$) \cite{Sudbo:95b}, while the

214:    second term applies to adjacent layers 

215:    $\langle\ell\ell^\prime \rangle$, and is \emph{local} in

216:    $\bk$-space, with $T_J (\bk) = \frac{1}{16} T_J (\cos k_x - \cos

217:    k_y )^4$ \cite{Chakravarty:93}. 

218: This enforces momentum conservation for the interlayer pair tunneling

219:    process.

220: (The effect of $\bk$-space broadening of the ILT kernel, \emph{e.g.}

221:    due to impurities, has been considered in

222:    Ref.~\onlinecite{Fjaerestad:98}.)

223: 

224: A straightforward mean-field analysis of Eq.~(\ref{eq:H}) for a

225:    bilayer complex and an in-plane superconducting instability in the

226:    $d$-wave channel yields the gap equation \cite{Angilella:99}:

227: \begin{equation}

228: \Delta_\bk = \frac{\Delta_0 g_\bk}{1-T_J (\bk)\chi_\bk} ,

229: \label{eq:gap1}

230: \end{equation}

231: where $\Delta_0$ is determined self-consistently from

232: \begin{equation}

233: 1 = \frac{V}{N} \sum_{\bk^\prime} g_{\bk^\prime}^2

234:    \frac{\chi_{\bk^\prime}}{1-T_J (\bk^\prime )\chi_{\bk^\prime} }.

235: \label{eq:gap2}

236: \end{equation}

237: Here, $\chi_\bk = (2E_\bk )^{-1} \tanh (\beta E_\bk /2)$ is the pair

238:    susceptibility at the inverse temperature $\beta = (\kB T)^{-1}$,

239:    $E_\bk = \sqrt{\xi_\bk^2 + |\Delta_\bk |^2}$ is the upper branch of

240:    the superconducting spectrum, and $N$ is the number of lattice sites.

241: 

242: Eqs.~(\ref{eq:gap1}) and (\ref{eq:gap2}) should be immediately compared and

243:    contrasted with Eq.~(9) in Ref.~\onlinecite{Chen:04} (for a

244:    multilayered complex) and Eq.~(1) in Ref.~\onlinecite{Chen:04a}

245:    (for a bilayer complex).

246: Even without going into the subtleties of the more general derivation

247:    for an $n$-layered complex (for which, see

248:    Refs.~\onlinecite{Sudboe:94c,Angilella:99b}), or with the

249:    competition among several in-plane pairing channels

250:    (Ref.~\onlinecite{Angilella:99}), it is apparent that the gap

251:    function within 

252:    the ILT model, Eq.~(\ref{eq:gap1}), is characterized by a

253:    \emph{local} prefactor $[1-T_J (\bk)\chi_\bk ]^{-1}$ which, albeit

254:    linked self-consistently to $\Delta_0$ \emph{via}

255:    Eq.~(\ref{eq:gap2}), is responsible of most of the quantitative and

256:    qualitative features of the model.

257: Such a structure is missing in Refs.~\onlinecite{Chen:04,Chen:04a}.

258: Even though the actual symmetry of the gap function is independent of

259:    the ILT kernel, and is rather determined by the $d$-wave nature of

260:    the in-plane coupling, the ILT mechanism endows the gap function

261:    with a nontrivial structure in $\bk$-space \cite{Angilella:99},

262:    which has been shown to be consistent with ARPES results

263:    \cite{Angilella:00}.

264: Moreover, the `renormalized' pair susceptibility in the summand of

265:    Eq.~(\ref{eq:gap2}), \emph{viz.} $\chi_\bk \mapsto \chi_\bk /

266:    [1-T_J (\bk)\chi_\bk ]$, which is due to the local ILT tunneling

267:    amplitude, gives rise to additional, algebraic divergences

268:    in the energy dependence of the integrated pair susceptibility, as

269:    opposed to the logarithmic one, typical of BCS theory \cite{AGD}.

270: This is directly responsible of the enhancement of $T_c$ within the

271:    ILT model.

272: In particular, in the case of a bilayer complex, one analytically

273:    finds a lower bound for $T_c$ as

274: \begin{equation}

275: \kB T^\ast (\mu) =

276: \begin{cases}

277: \displaystyle

278: \frac{T_J}{64} \left( \frac{\mu_\perp - \mu}{\mu_\perp + 2t}

279:    \right)^4 , & \mu_\perp \leq \mu < \mu_{\mathrm{VH}} ,\\

280: \displaystyle

281: \frac{T_J}{64} \left( \frac{\mu_\top - \mu}{\mu_\top - 2t}

282:    \right)^4 , & \mu_{\mathrm{VH}} \leq \mu \leq \mu_\top ,

283: \end{cases}

284: \end{equation}

285: where nearest ($t$) and next-nearest ($t^\prime$) hopping have been

286:    assumed, and $\mu_\perp = -4t+4t^\prime$, $\mu_\top =

287:    4t+4t^\prime$, and $\mu_{\mathrm{VH}} = -4t^\prime$

288:    denote the bottom, the top of the band, and the location of the

289:    Van~Hove singularity, respectively \cite{Angilella:99}.

290: 

291: On the contrary, the ILT kernel enters Chen \emph{et al.}'s gap

292:    equation in Eq.~(9) of Ref.~\onlinecite{Chen:04} and Eq.~(1) of

293:    Ref.~\onlinecite{Chen:04} as an additional contribution to the

294:    non-local in-plane coupling term, \emph{i.e.} it amounts to

295:    defining another in-plane interaction, with no reference to

296:    interlayer tunneling.

297: \emph{This same observation applies to the general case of an

298:    $n$-layered complex.} 

299: In that case, the gap equation for each layer should also contain a

300:    local contribution due to the ILT mechanism between adjacent

301:    layers (again, absent in Ref.~\onlinecite{Chen:04}), with an ILT

302:    renormalized pair susceptibility $\chi_\bk^\ell / [1-T_J

303:    (\bk)\hat{\chi}_\bk^\ell ]$ for each layer

304:    \cite{Sudboe:94c,Angilella:99b}, with 

305: \begin{eqnarray}

306: \hat{\chi}_\bk^\ell &=& \left[\sin\left( \frac{\ell\pi}{n+1}

307:    \right)\right]^{-1} \left[ \chi_\bk^{\ell+1} \sin \left(

308:    \frac{(\ell+1)\pi}{n+1} \right) \right. \nonumber\\

309: &&+ \left. \chi_\bk^{\ell-1} \sin \left(

310:    \frac{(\ell-1)\pi}{n+1} \right) \right],

311: \end{eqnarray}

312: which can be further simplified in the limit of uniform hole-content

313:    in all layers (as is tacitly assumed in

314:    Ref.~\onlinecite{Chen:04}).

315: In analogy to the bilayer case, the condition

316: \begin{equation}

317: \min_\bk [1-T_J (\bk)\chi_\bk^\ell ] = 0

318: \end{equation}

319: then implicitly defines a lower bound $T_c^{\ast\ell}$ for the

320:    critical temperature 

321:    corresponding to the onset of superconductivity \emph{in the given

322:    layer $\ell$.}

323: Therefore, for nonuniform hole-content among inequivalent layers, as

324:    is the case for the multilayered cuprates

325:    \cite{Trokiner:91,Ambrosch-Draxl:04}, one can estimate a lower

326:    bound to $T_c$ as $\max_\ell T_c^{\ast\ell}$.

327: A nonuniform distribution of the overall hole-content among

328:    inequivalent layers can be conveniently described by means of

329:    appropriate models \cite{Angilella:99b}.

330: This then enables us to identify whether the superconducting instability

331:    first sets in in inner or outer layers.

332: One finds a crossover as function of the overall hole-content

333:    \cite{Angilella:99b}, which has been related to the observed

334:    `kinks' in the pressure dependence of $T_c$ in several layered

335:    cuprates \cite{Wijngaarden:99}.

336: 

337: In conclusion, we have pointed out an incorrect derivation of the gap

338:    equation(s) for layered cuprates within the ILT model

339:    \cite{Chen:04,Chen:04a} for the general case of $n$ superconducting

340:    layers per unit cell.

341: This in turn leads to a failure to capture most of the

342:    qualitative and quantitative features of the theory, both for

343:    bilayered and multilayered compounds.

344: As a consequence, the theoretical analysis of the high pressure data

345:    in Ref.~\onlinecite{Chen:04a} is not consistent with the ILT

346:    mechanism.

347: On the other hand, a more careful analysis of the ILT model

348:    \cite{Sudboe:94c}, when taking into account a nonuniform

349:    hole-content distribution among inequivalent layers

350:    \cite{Angilella:99b}, is indeed able to reproduce the observed

351:    pressure dependence of $T_c$ in multilayered cuprates

352:    \cite{Wijngaarden:99}.

353: 

354: 

355: 

356: \begin{acknowledgments}

357: We thank S. Chakravarty and J. O. Fj\ae{}restad for useful discussions

358:    and correspondence.

359: \end{acknowledgments}

360: 

361: \begin{small}

362: \bibliographystyle{apsrev}

363: \bibliography{a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,zzproceedings,Angilella}

364: \end{small}

365: 

366: \end{document}

367: