1: %       Submitted to Canadian Journal of Physics (HFM Conference)

2: 

3: %       Started writing on 24th July 2000

4: %

5: %       Benjamin Canals and Dmitri Alexandrovitch Garanin

6: 

7: 

8: %\documentclass[aps,preprint,groupedaddress]{revtex4}

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

10: %\documentclass[aps,twocolumn,groupedaddress]{revtex4}

11: \documentclass[aps,twocolumn,showpacs]{revtex4b5}

12: 

13: \usepackage{graphicx} 

14: \usepackage{dcolumn}

15: \usepackage{amsmath}

16: 

17: 

18: \begin{document}

19: 

20: \title{Spin Liquid Phase in the Pyrochlore Antiferromagnet}

21: 

22: 

23: \author{Benjamin Canals}

24: \email[]{canals@polycnrs-gre.fr}

25: \homepage[]{http://ln-w3.polycnrs-gre.fr/pageperso/canals/}

26: \affiliation{Laboratoire Louis N\'eel, CNRS, 25 avenue

27: des Martyrs, Boite Postale 166, 38042 Grenoble Cedex 9, France}

28: 

29: \author{D. A. Garanin}

30: \email[]{dgaranin@yahoo.com}

31: \homepage[]{http://www.mpipks-dresden.mpg.de/~garanin}

32: \affiliation{Max-Planck-Institut f\"ur Physik komplexer Systeme, 

33: N\"othnitzer Strasse 38, D-01187 Dresden, Germany}

34: 

35: %

36: \date{\today}

37: 

38: \begin{abstract}

39: \vspace{0.4cm}

40: %

41: Correlation functions (CFs) of the classical

42: Heisenberg antiferromagnet on the pyrochlore lattice are studied 

43: by solving exactly the infinite-component spin-vector model.

44: %

45: As in many Fully Frustrated Lattices, the constraint due to the minimization of the

46: energy and the particular structure based on corner sharing tetrahedra

47: both contribute to the creation of local degrees of freedom.

48: %

49: The resulting degeneracy destroys any magnetic order at all temperature and we

50: obtain no sign of criticality, even at $T = 0$.

51: %

52: Calculated neutron scattering 

53: cross sections have their maxima beyond the first Brillouin

54: Zone and reproduce experimental results obtained on Y(Sc)Mn$_2$ and

55: CsCrNiF$_6$ as well as theoretical predictions previously obtained

56: by classical Monte Carlo simulations.

57: %

58: Evidences for thermal and spatial decoupling of the magnetic modes 

59: are found so that the magnetic fluctuations in this system can be approximated by 

60: $S(q,T) \approx f(q) h(T)$.

61: %

62: \end{abstract}

63: %

64: \pacs{75.10.Hk, 75.50.Ee, 75.40.Cx, 75.40.-s}

65: 

66: \maketitle

67: 

68: 

69: \section{Introduction}

70: 

71: Motivated by the search and understanding of new quantum or classical disordered ground

72: states in magnetic systems, recent years have seen a renewal of interest in

73: the properties of frustrated Heisenberg antiferromagnets.

74: %

75: Special attention focused on intrinsically frustrated models, where frustration is

76: mainly coming from the connectivity of the underlying lattice and not only

77: from competing interactions.

78: %

79: From this point of view, two lattices distinguished themselves~:~the kagom\'e lattice 

80: (made of corner sharing triangles) and the pyrochlore lattice (made of corner sharing

81: tetrahedra).

82: %

83: Experimental and theoretical studies \cite{diep94} related to both lattices

84: have shown that materials as well as models exhibit unconventionnal magnetic properties, 

85: involving noncollinear or incommensurate orderings, apparently broken ergodicity 

86: without structural disorder and in general, deviations from canonical behaviors.

87: %

88: For most geometrically frustrated systems, the absence of ordering at finite temperature

89: or the tendancy to a reduction of the real interactions seems to be a common characteristic.

90: %

91: An experimental signature of this property is a persistent paramagnetic state down to low

92: temperature without any sign of ordering \cite{ballelfak96} and sometimes a transition to a frozen state at

93: low temperature \cite{ginstarajgaugre97} ($T_F << |\Theta_{CW}|$), 

94: with an unconventional slowing down of the dynamics \cite{uemura941,dunsiger96,uemura942}.

95: %

96: The theoretical understanding of these properties is in progress but it seems that several

97: properties, expected to be generic, are not systematically reproduced and depend on the

98: material or the model.

99: %

100: One point is at least widely shared: due to frustration, short range correlations develop

101: at a temperature much lower than the interactions. This phenomenon has been proven by mean

102: field analysis \cite{rei91} and classical monte-carlo simulations \cite{rei921,rei922,moecha98}.

103: %

104: On the other hand, there are some properties much more dependent on the studied system.

105: %

106: A major theme of study has been the effect of ``order by disorder'' previously quoted by

107: Villain and co-workers \cite{villain80} and also by Shender \cite{shender82}.

108: %

109: It was noticed that switching on any source of fluctuation (like thermal or quantum fluctuations) 

110: has a tendancy to lift the degeneracy between the low lying classical ground states.

111: %

112: Indeed, thermal fluctuations give rise to entropic selection in the kagom\'e case

113: \cite{chalker92,ritchey93} but this kind of selection was shown to be fluctuation-like

114: dependent \cite{doucot98}, and sometimes absent \cite{huse92,chandra94}.

115: %

116: Other types of perturbation have been taken into account, such as anisotropy 

117: \cite{bra94}, structural disorder \cite{shender93}, 

118: longer range interactions \cite{rei91,harris92}

119: and quantum fluctuations 

120: \cite{canals98,sobral97,harris91,zeng91,sachdev92,chubukov92,leche97}, 

121: leading in general to a magnetic mode selection.

122: %

123: 

124: 

125: The pyrochlore lattice is a three dimensional arrangement of corner sharing tetrahedra

126: as depicted in Fig.~\ref{pyr_str}.

127: %

128: According to Lacorre definition of

129: frustration \cite{lac87}, it is probably the most frustrated structure.

130: %

131: Experimentally, several families of compounds are known to crystallize with this geometry.

132: %

133: The first one is an oxide family with the general formula

134: A$_2^{3+}$B$_2^{4+}$O$_7$ (Refs. \onlinecite{gardner99,greedan91,dunsiger96}) 

135: (A is a Rare Earth ion and B a transition metal).

136: %

137: The second one concerns spinels of general formula $AB_2O_4$ or $A_2B_2O_4$

138: (Refs. \onlinecite{wills99,greedan98,fujiwara98}) (A is a Rare Earth ion and B a transition metal).

139: %

140: The third one concerns fluorides of general formula AB$^{2+}$C$^{3+}$F$_6$ 

141: (Refs. \onlinecite{harris97,harzintunwanswa94}) (A is an alkali metal, B and C are transition metals).

142: %

143: The last one is the intermetallic Laves Phases compounds of general formula AB$_2$ 

144: (Refs. \onlinecite{shiga94,shiga88}) (A is a Rare Earth ion and B is a transition metal).

145: %

146: Strictly speaking, the only candidate that can be rigorously described by a localized, uniform

147: lattice model is the first family, as this system is well ordered with a

148: uniform distribution of magnetically localized ions.

149: %

150: The second and third sets often possess positional or chemical disorder 

151: \cite{wills99,harris97} that should be taken into account in the model 

152: and the last one is better described within an itinerant model \cite{ballac96}.

153: %

154: 

155: In this paper, we will focus only on the isotropic classical Heisenberg model on the pyrochlore

156: lattice, neglecting all other possible contributions that might be necessary for 

157: an understanding of a peculiar material.

158: %

159: This choice is supported by our hope to deduce intrinsic properties which come 

160: exclusively from geometrical structure.

161: %

162: Our goal is to compute correlation functions using an approximation

163: of the Heisenberg model, the infinite component spin-vector model.

164: %

165: 

166: \section{Lattice structure and model}

167: %

168: 

169: The ``next simplest'' approximation for classical spin systems, which follows the 

170: Mean Field Approximation (MFA), 

171: consists in generalizing the Heisenberg Hamiltonian for the $D$-component spin vectors:

172: %\marginpar{dham}

173: %

174: \begin{equation}\label{dham}

175: %

176: {\cal H} = - \frac{1}{2} \sum_{i,j} J_{ij} \, {\bf s}_{i} \cdot {\bf s}_{j}  ,  \qquad |{\bf s}_i|=1

177: %

178: \end{equation}

179: %

180: and taking the limit $D\to\infty$.

181: %

182: In this limit, it is possible to take into account exactly the effect of the

183: Goldstone modes or would be Goldstone modes, which are the most

184: important modes in frustrated magnets. 

185: %

186: We will only give the outline of the calculation here and give in a further article

187: the details of the technique.

188: Nevertheless, readers who are interested in the full derivation can find a

189: description of the method in Ref. \onlinecite{garcan99}, where 

190: the antiferromagnetic model on the kagome lattice was considered.

191: %

192: 

193: %

194: \begin{figure}[t]

195: \unitlength1cm

196: \vspace{-1.1cm}

197: %\centerline{\includegraphics[width=2.7in]{pyr_str.eps}}

198: \centerline{\includegraphics[width=8cm]{pyr_str}}

199: \vspace{-4cm}

200: \caption{ \label{pyr_str} 

201: Structure of the pyrochlore lattice.

202: %

203: The elementary tetrahedra are labeled by $i=1,\ldots,N$, 

204: and the sites on tetrahedra (the

205: sublattices) are labeled by $l=1,2,3,4$.}

206: %

207: \end{figure}

208: %

209: %

210: %

211: The pyrochlore lattice shown in Fig. \ref{pyr_str} consists of corner-sharing

212: tetrahedra.

213: %

214: Each node of the underlying Bravais lattice (face centered cubic) is numbered by $i=1,\ldots, N$ and

215: corresponds to a tetrahedron.

216: %

217: Each site of elementary tetrahedra is labeled by the index $l=1,2,3,4$.

218: %

219: 

220: %%%%%%%%%%%%%

221: %%%%%%%%%%%%%

222: 

223: It is convenient to put the Hamiltonian (\ref{dham}) into a diagonal form.

224: %

225: First, one goes to the Fourier representation according to

226: %\marginpar{fourier}

227: %

228: \begin{equation}\label{fourier}

229: %

230: {\bf s}_{\bf q}^l = \sum_i {\bf s}_i^l e^{-i {\bf q} r_{i}^{l}}, 

231: \qquad {\bf s}_i^l = \frac 1N \sum_{\bf q} {\bf s}_{\bf q}^l e^{i {\bf q} r_{i}^{l}}, 

232: %

233: \end{equation}

234: %

235: where the wave vector ${\bf q}$ belongs to the $1^{\rm st}$ Brillouin zone of the

236: underlying face centered cubic lattice (see Fig. \ref{pyr_bri}).

237: %

238: The Fourier-transformed Hamiltonian reads

239: %\marginpar{dhamq}

240: %

241: \begin{equation}\label{dhamq}

242: %

243: {\cal H} = \frac{1}{2N} \sum_{ll'\bf q} V_{\bf q}^{ll'} {\bf s}_{\bf q}^l 

244:  \cdot \rm s_{-{\bf q}}^{l'} , 

245: %

246: \end{equation}

247: %

248: where the interaction matrix is given by

249: %\marginpar{Vq}

250: %

251: \begin{eqnarray}\label{Vq}

252: %

253: \hat V_{\bf q} = 2J \nonumber 

254: \left( \begin{array}{cccc} \nonumber 

255: 0 & b & c & a \nonumber \\

256: b & 0 & f & d \nonumber \\

257: c & f & 0 & e \nonumber \\

258: a & d & e & 0 \nonumber \\

259: \end{array} \right) , \nonumber

260: %

261: \qquad

262: \textrm{where} \\

263: \\

264: a \equiv \cos( \frac{q_y + q_z}{4})

265: \qquad

266: d \equiv \cos( \frac{q_x - q_y}{4}) \nonumber \\

267: %

268: b \equiv \cos( \frac{q_x + q_z}{4})

269: \qquad

270: e \equiv \cos( \frac{q_x - q_z}{4}) \nonumber \\

271: %

272: c \equiv \cos( \frac{q_x + q_y}{4})

273: \qquad

274: f \equiv \cos( \frac{q_y - q_z}{4}) \nonumber

275: %

276: %

277: \end{eqnarray}

278: %

279: %

280: %

281: At the second stage, the Hamiltonian (\ref{dhamq}) is finally diagonalized to the form

282: %\marginpar{dhamdiag}

283: %

284: \begin{equation}\label{dhamdiag}

285: % 

286: {\cal H} =  -\frac{1}{2N} \sum_{n\bf q} \tilde V_{\bf q}^n 

287: \sigma_{\bf q}^n \cdot \sigma_{-\bf q}^n , 

288: %

289: \end{equation}

290: %

291: where $ \tilde V_{\bf q}^n = 2J \nu_n({\bf q})$ are the eigenvalues of the 

292: matrix $V_{\bf q}^{ll'} $ taken with a negative sign,

293: %\marginpar{nudef}

294: %

295: \begin{equation}\label{nudef}

296: %

297: \nu_{1,2} = 1, \qquad \nu_{3,4} = - 1 \pm \sqrt{1+Q}.

298: %

299: \end{equation}

300: %

301: where

302: %

303: \begin{eqnarray}\label{Q}

304: %

305: Q=\frac{1}{2} ( & \cos( & \frac{q_y + q_z}{2}) + 

306:                       \cos( \frac{q_x + q_z}{2}) + \nonumber \\

307:                 & \cos( & \frac{q_x + q_y}{2}) +

308:                       \cos( \frac{q_x - q_y}{2}) + \\

309:                 & \cos( & \frac{q_x - q_z}{2}) +

310:                       \cos( \frac{q_y - q_z}{2})      

311:               ) \nonumber

312: %

313: \end{eqnarray}

314: %

315: The diagonalizing transformation has the explicit form 

316: %\marginpar{Vtrans}

317: %

318: \begin{equation}\label{Vtrans}

319: %

320: U_{nl}^{-1}({\bf q}) V^{ll'}_{\bf q} U_{l'n'}({\bf q}) = \tilde V^n_{\bf q} \delta_{nn'} ,

321: %

322: \end{equation}

323: %

324: where the summation over the repeated indices is implied and $\hat U$ is the real unitary matrix,

325: $\hat U^{-1} = \hat U^T$, i.e., $U_{nl}^{-1} = U_{ln}$.

326: %

327: 

328: The Fourier components of the spins ${\bf s}_{\bf q}^l$ and the 

329: collective spin variables $\sigma_{\bf q}^n$ are related by

330: %\marginpar{s-sigma}

331: %

332: \begin{equation}\label{s-sigma}

333: %

334: {\bf s}_{\bf q}^l = U_{l n}({\bf q}) \sigma_{\bf q}^n, 

335: \qquad \sigma_{\bf q}^n = {\bf s}_{\bf q}^l U_{l n}({\bf q}).

336: %

337: \end{equation}

338: % 

339: 

340: %

341: %

342: %

343: \begin{figure}[t]

344: \unitlength1cm

345: \vspace{0cm}

346: %\centerline{\includegraphics[width=2.7in]{pyr_bri.ps}}

347: \centerline{\includegraphics[width=7cm]{pyr_bri.ps}}

348: \vspace{0cm}

349: \caption{ \label{pyr_bri} 

350: Brillouin zone of the pyrochlore lattice.

351: %

352: Letters $\Gamma$, $X$, $W$, $U$ and $L$ are high symmetry points

353: using the usual crystallographic conventions.}

354: %

355: \end{figure}

356: %

357: 

358: 

359: %

360: \begin{figure}[t]

361: \unitlength1cm

362: \vspace{-0.5cm}

363: %\centerline{\includegraphics[width=2.7in]{pyr_str.eps}}

364: \centerline{\includegraphics[width=9cm]{pyr_mod}}

365: \vspace{-0.5cm}

366: \caption{\label{pyr_mod} 

367: Reduced eigenvalues of the interaction matrix, $ \nu_n({\bf q}) = \tilde

368: V_{\bf q}^n/(2J)$ of Eq. (\protect\ref{dhamdiag}). They are plotted such that

369: they cover the high symmetry axis of the face centered cubic Brillouin

370: Zone (see Fig. \ref{pyr_bri}).

371: }

372: %

373: \end{figure}

374: %

375: 

376: The largest dispersionless eigenvalues $\nu_{1,2}$ of the interaction matrix 

377: [see Eq. (\ref{nudef}) and Fig. \ref{pyr_mod}] are a manifestation of frustration in the system 

378: which precludes an extended short-range order even in the limit $T\to 0$.

379: %

380: Since $\nu_{1,2}$ are independent of ${\bf q}$, 1/2 of all spin degrees of freedom 

381: are local and can rotate freely.

382: %

383: The eigenvalue $\nu_3$ which becomes degenerate with $\nu_{1,2}$ in the limit

384: $q\to 0$ is related, to the usual would be Goldstone mode destroying the

385: long-range order in low-dimensional magnets with a continuous symmetry whereas

386: the eigenvalue $\nu_4$ is positioned,  in the long-wavelength region, 

387: much lower than the first two ones, and it is tempting to call it the ``optical'' eigenvalue.

388: %

389: 

390: \section{Results}

391: 

392: %

393: Following the derivation introduced in Ref. \onlinecite{garcan99}, we compute

394: all CF's at any temperature

395: and in particular, the static magnetic neutron scattering cross section :

396: %\marginpar{defcrsec}

397: %

398: \begin{equation}\label{defcrsec}

399: %

400: \frac{ d\sigma }{ d\Omega } \propto \sum_{{\bf rr}'} 

401: \langle S_{\bf r}^\perp S_{{\bf r}'}^\perp \rangle e^{i{\bf Q}({\bf r} - {\bf r}')},  

402: %

403: \end{equation}

404: %

405: where {\bf Q} is a vector of the reciprocal space and  $r$ and $r^{'}$ design 

406: spin belonging to the pyrochlore lattice, i.e.,

407: $S_{\bf r} = S_{i}^{l}$ is a spin located on site $i$ of the Bravais lattice

408: in the position $l$ of the unit cell.

409: %

410: Two peculiar planes of the reciprocal space have been investigated as shown

411: in Fig.~\ref{pyr_neuhex} and Fig.~\ref{pyr_neucar}.

412: %

413: The major feature is that there is almost no magnetic signal in the 

414: first Brillouin zone, all the weight

415: being recovered in soft cusps outside. In both planes, we observe patterns

416: very similar to the one obtained by Monte Carlo simulations in 

417: Ref.~\onlinecite{zinharzei97}

418: or by neutron diffraction on $Y(Sc)Mn_2$ \cite{ballelfak96}. 

419: %

420: We also note that the cross section in the ([hh0],[00l]) plane is very

421: close to the one previously calculated in the kagome case by the same

422: method \cite{garcan99}.

423: %

424: The main implications of these results are that the quantum or classical

425: nature of the spins in the pyrochlore lattice does not seem to be of

426: relevance for understanding the liquid nature in terms of

427: spin-spin correlations.

428: %

429: Indeed, different methods and measures, applied to differents spin

430: species (classical or quantum), give the same magnetic responses.

431: %

432: Besides, the very strong resemblance between the neutron diffraction

433: pattern in the ([hh0],[00l]) plane and the one obtained in the kagome

434: antiferromagnet is evidence that these two lattices are very similar.

435: %

436: This is not surprising as the kagome lattice is a cut of the pyrochlore

437: lattice perpendicular to the (111) axis.

438: %

439: As previously noted in Ref. \onlinecite{canals98} there are two magnetic modes coexisting 

440: in this system (see Fig. \ref{pyr_neuhex} and Fig. \ref{pyr_neucar}).

441: %

442: The one related to the ([hh0],[00l]) plane

443: ($Q_1 = [220] \pm [\frac{3}{4} \frac{3}{4} 0]$ or

444:  $Q_{1}^{'} = [002] \pm [\frac{3}{4} \frac{3}{4} 0]$) would coincide

445: in an ordered phase with a structure where consecutive tetrahedra are

446: in phase, while the one in the ([h00],[0l0]) plane 

447: ($Q_2 = [210]$) correspond to a $\pi$ phase between tetrahedra.

448: %

449: Whereas in the quantum approach\cite{canals98} there is a difference of

450: amplitude between these two modes, within our approach both have

451: the same amplitude so that the system is unable to select a 

452: particular one.

453: %

454: %This means that the system should be unstable and that a phase transition

455: %associated to a selection of a mode could be first order.

456: %

457: %

458: The powder average of the neutron cross section 

459: $\langle d\sigma / d\Omega \rangle$, i.e., the average of 

460: \ref{defcrsec} over the directions of $q$ is shown 

461: in Fig. \ref{pyr_powd_comp}.

462: %

463: The continuous line is obtained from the $D = \infty$ component classical

464: antiferromagnet approach while the dots

465: correspond to Monte Carlo simulations performed in 

466: Ref.\onlinecite{rei92}.

467: %

468: There is a good agreement between these two methods and they do not

469: show any sign of magnetic transition.

470: %

471: We have calculated the effect of the temperature on the powder average

472: cross section.

473: %

474: In Fig. \ref{pyr_powd_multi} we have plotted the variation of the profiles with reduced

475: temperature ($\theta = T/T_{\rm MFA}$ where $T_{\rm MFA} = 2J/3$).

476: %

477: One can see that at temperature below $T_{C}^{\rm MFA}$, i.e., below 

478: $\theta = 1$, these profiles have approximatively the same shape.

479: %

480: In order to control that quantity, we have rescaled each profile and

481: plotted them one onto the others.

482: %

483: As they almost coincide, it means that at low temperature we can

484: decouple the $q$ dependance and the $\theta$ dependance so that 

485: $\langle d\sigma / d\Omega \rangle (Q,\theta) \approx 

486:  \langle d\sigma / d\Omega \rangle (Q,\theta = 0) f(\theta)$.

487: %

488: The same type of decoupling was assumed in Ref. \onlinecite{canals98} where the authors

489: proposed that the spatial and the time dependance could be

490: independent at low temperatures ($\chi (q ,\omega) \approx f(q) g(\omega)$).

491: %

492: These properties are original and certainly come from the high frustration

493: of the pyrochlore lattice and its related dynamics.

494: %

495: At first approximation, we can state that spatial and time as well as

496: spatial and thermal

497: dependance are decoupled at energy scales lower than the mean field

498: temperature.

499: %

500: 

501: 

502:  

503: %

504: \begin{figure}

505: \unitlength1cm

506: \hspace{0cm}

507: \centerline{\includegraphics[width=8cm]{pyr_neu_hexa}}

508: \caption{ \label{pyr_neuhex} 

509: Neutron scattering cross section from the large-$D$ pyrochlore antiferromagnet

510: at $T=0$ in the ([hh0],[00l]) plane.

511: (cf. Fig. \protect\ref{pyr_bri}).

512: Dark regions are low intensity ones while white regions correspond to the

513: maximum of the scattering.

514: }

515: %

516: \end{figure}

517: %

518: 

519: 

520: %

521: \begin{figure}

522: \unitlength1cm

523: \hspace{0cm}

524: \centerline{\includegraphics[width=8cm]{pyr_neu_carr}}

525: \caption{ \label{pyr_neucar} 

526: Neutron scattering cross section from the large-$D$ pyrochlore antiferromagnet

527: at $T=0$ in the ([h00],[0l0]) plane.

528: (cf. Fig. \protect\ref{pyr_bri}).

529: Dark regions are low intensity ones while white regions correspond to the

530: maximum of the scattering.

531: }

532: %

533: \end{figure}

534: %

535: 

536: 

537: %

538: \begin{figure}[t]

539: \unitlength1cm

540: \centerline{\includegraphics[width=8cm]{pyr_powd_comp}}

541: \caption{ \label{pyr_powd_comp} 

542: Powder average of the neutron scattering cross section from the pyrochlore

543: antiferromagnet. 

544: %

545: The straight line correspond to the $T=0$ case in the $D = \infty$ component classical

546: antiferromagnet

547: approach while dots are extracted points from the Monte Carlo results obtained

548: in Ref.\protect\onlinecite{rei92}

549: % 

550: }

551: %

552: \end{figure}

553: %

554: 

555: %

556: \begin{figure}[t]

557: \unitlength1cm

558: \centerline{\includegraphics[width=8cm]{pyr_powd_multi}}

559: \caption{ \label{pyr_powd_multi} 

560: Powder average of the neutron scattering cross section from the pyrochlore

561: antiferromagnet at different temperatures ($\theta \in [0,1]$ ).

562: %

563: Each curve is rescaled to be compared with the $T=0$ one.

564: % 

565: }

566: %

567: \end{figure}

568: %

569: 

570: 

571: 

572: 

573: \section{Discussion}

574: \label{secDiscussion}

575: 

576: In this article, we have presented the structure of the neutron diffraction

577: cross sections obtained from the exact

578: solution for the $D=\infty$ component classical antiferromagnet on the

579: pyrochlore lattice.

580: %

581: 

582: In contrast to conventionnal three dimensional magnets, there is no ordering

583: and even no extended short range order at $T=0$.

584: %

585: This anomalous behavior is due to the existence of a would be Goldstone mode

586: at low temperature but also to the presence of two

587: dispersionless excitations modes that inhibit any magnetic transition.

588: %

589: 

590: The magnetic cross sections are very close to what has been obtained in 

591: theoretical approaches (Monte Carlo simulations with Heisenberg spins \cite{zinharzei97},

592: density matrix expansion for spin $S=1/2$ \cite{canals98}) and experimental results

593: (Y(Sc)Mn$_2$ \cite{ballelfak96}, CsNiCrF$_6$ \cite{harris97,harzintunwanswa94}).

594: %

595: We have reproduced the existence of two dominant magnetic modes that could

596: explain a first order transition induced by magnetoelastic effects \cite{ballelfak96}.

597: %

598: It is remarkable that the nature of the spins (classical or quantum) as well

599: as the type of magnetism (localized or itinerant) does not seem to play

600: a crucial role in the pyrochlore system.

601: %

602: Even the geometry, three dimensional here, seem to have a very little effect on the

603: struture of the magnetic response.

604: %

605: In fact, we obtain results similar to the ones previously observed on the kagome

606: lattice.

607: %

608: Even if this point is not solved, it is clear that the close relation between these

609: two lattices, and the short range scale of the geometry are the relevant 

610: ingredients in the pyrochlore and the kagome antiferromagnets.

611: %

612: 

613: Taking into account assumptions of Ref .\onlinecite{canals98} we can say that

614: spatial and thermal as well as spatial and time dependences 

615: of the magnetic cross sections

616: are decoupled at low energy scale (i.e. $T \lesssim J$) as if 

617: $S(q,T) \approx f(q) h(T)$ and $S(q,\omega) \approx f(q) g(\omega)$. This striking

618: result confirms that this model describes an unconventionnal antiferromagnet

619: and has some similarities with a collective paramagnet.

620: %

621: 

622: All these points strongly support that isotropic antiferromagnetic

623: Heisenberg models (quantum or classical) on the pyrochlore lattice are 

624: spin liquids.

625: %

626: 

627: Although this point is now clearer, it is still an open problem

628: to understand the experimental results that suggest the appearance of

629: a topological spin glass like behavior.

630: %

631: There must certainly be other interactions or processes involved in this

632: case as up to now, no sign of glassiness has ever been seen within theoretical

633: approaches concerning the Heisenberg model on the pyrochlore lattice.

634: 

635: 

636: 

637: %\vspace{15mm}

638: 

639: \begin{thebibliography}{10}

640: 

641: 

642: %\bibitem[$\dagger$]{e-can}

643: %Electronic address: canals@polycnrs-gre.fr \\

644: 

645: %\bibitem[*]{e-gar}

646: %Electronic addresses: dgaranin@yahoo.com \\

647: 

648: %%%%%%%%%%%%

649: %%%%%%%%%%%%

650: 

651: \bibitem{diep94}

652: For a collection of articles, see

653: {\it Magnetic systems with Competing Interactions : Frustated Spin System}, 

654: edited by H. T. Diep (World Scientific, Singapore, 1994); for Ising systems, 

655: see R. Liebmann, 

656: {\it Statistical Mechanics of Periodic Frustrated Ising Systems} 

657: (Springer, Berlin, 1986); for reviews, see 

658: A. P. Ramirez, Annu. Rev. Mater. Sci. {\bf 24}, 453, (1994); P. Schiffer

659: and A. P. Ramirez, Comments Cond. Mat. Phys. {\bf 18}, 21, (1996);

660: M. J. Harris and M. P. Zinkin, Mod. Phys. Lett. B {\bf 10}, 417, (1996).

661: 

662: \bibitem{ballelfak96}

663: R. Ballou, E. Leli\`{e}vre-Berna and B. F\aa k,

664: Phys. Rev. Lett. {\bf 76}, 2125, (1996).

665: 

666: \bibitem{ginstarajgaugre97}

667: M. J. P. Gingras, C. V. Stager, N. P. Raju, B. D. Gaulin and J. E. Greedan,

668: Phys. Rev. Lett. {\bf 78}, 947, (1997).

669: 

670: \bibitem{uemura941}

671: Y. J. Uemura, A. Keren, L. P. Le, G. M. Luke, B. J. Sternlieb

672: and W. D. Wu, Hyperfine Interact. {\bf 85}, 133 (1994).

673: 

674: \bibitem{uemura942}

675: Y. J. Uemura, A. Keren, K. Kojima, L. P. Le, G. M. Luke, W. D. Wu,

676: Y. Ajiro, T. Asano, Y. Kuriyama, M. Mekata, H. Kikuchi and K. Kakurai,

677: Phys. Rev. Lett. {\bf 73}, 3306 (1994).

678: 

679: \bibitem{dunsiger96}

680: S. R. Dunsiger, R. F. Kiefl, K. H. Chow, B. D. Gaulin, M. J. P. Gingras, 

681: J. E. Greedan, A. Keren, K. Kojima, G. M. Luke, W. A. MacFarlane, N. P. Raju, 

682: J. E. Sonier, Y. J. Uemura, and W. D. Wu, 

683: Phys. Rev. B {\bf 54}, 9019 (1996).

684: 

685: \bibitem{rei91} 

686: J. N. Reimers, A. J. Berlinsky and A.-C. Shi., Phys. Rev. B

687: {\bf 43} (1991) 865.

688: %Mean-field approach to magnetic ordering in

689: %highly frustrated pyrochlores.

690: 

691: \bibitem{rei921}

692: J. N. Reimers, J. E. Greedan, and M. Bj\"orgvinsson, Phys. Rev. B {\bf 45},

693:   7295  (1992).

694: 

695: \bibitem{rei922} 

696: J. N. Reimers, Phys. Rev. B {\bf 45}, 7287 (1992).

697: 

698: \bibitem{moecha98} 

699: R. Moessner and J. T. Chalker, Phys. Rev. Lett. {\bf 80}, 2929 (1998);

700: R. Moessner and J. T. Chalker, Phys. Rev. B {\bf 58}, 12049 (1998).

701: 

702: \bibitem{villain80}

703: J. Villain, R. Bidaux, J. P. Carton, R. Conte, 

704: J. Phys. {\bf 41}, 1263 (1980).

705: 

706: \bibitem{shender82}

707: E. F. Shender, Sov. Phys.-JETP, {\bf 56}, 178 (1982).

708: 

709: \bibitem{chalker92}

710: J. T. Chalker, P. C. W. Holdsworth and E. F. Shender,

711: Phys. Rev. Lett. {\bf 68}, 855 (1992).

712: 

713: \bibitem{ritchey93}

714: I. Ritchey, P. Chandra and P. Coleman,

715: Phys. Rev. B{\bf 47}, 15342 (1993).

716: 

717: \bibitem{doucot98}

718: B. Dou\c{c}ot and P. Simon, 

719: J. Phys. A: Math. Gen. {\bf 31}, 5855 (1998).

720: 

721: \bibitem{huse92}

722: D. A. Huse and A. D. Rutenberg, 

723: Phys. Rev. B {\bf 45}, 7536 (1992).

724: %CLASSICAL ANTIFERROMAGNETS ON THE KAGOME LATTICE

725: 

726: \bibitem{chandra94}

727: P. Chandra and B. Dou\c{c}ot,

728: J. Phys. A, {\bf 27}, 1541 (1994).

729: %Spin liquids on the Husimi cactus

730: 

731: \bibitem{bra94} 

732: S. T. Bramwell, M. J. P. Gingras, and J. N. Reimers, J. Appl. P\

733: hys. {\bf 75}, 5523 (1994).

734: 

735: \bibitem{shender93}

736: E. F. Shender, V. B. Cherepanov, P. C. W. Holdsworth and

737: A. J. Berlinsky, 

738: Phys. Rev. Lett. {\bf 70}, 3812 (1993).

739: %Kagome Antiferromagnet with Defects: Satisfaction, Frustration, 

740: %and Spin Folding in a Random Spin System.

741: 

742: \bibitem{harris92}

743: A. B. Harris, C. Kallin and A. J. Berlinsky,

744: Phys. Rev. B {\bf 45}, 2899 (1992).

745: %Possible Neel orderings of the Kagome

746: %antiferromagnet.

747: 

748: \bibitem{canals98}

749: B. Canals and C. Lacroix, Phys. Rev. Lett. {\bf 80},  2933  (1998);

750: B. Canals and C. Lacroix, Phys. Rev. B {\bf 61},  1149  (2000).

751: 

752: \bibitem{sobral97}

753: R. R. Sobral and C. Lacroix, 

754: Solid State Comm. {\bf 103}, 407 (1997).

755: %Order by disorder in the pyrochlore antiferromagnets

756: 

757: \bibitem{harris91}%

758: A. B. Harris, A. J. Berlinsky and C. Bruder,

759: J. Appl. Phys. {\bf 69}, 5200 (1991).

760: %ORDERING BY QUANTUM FLUCTUATIONS IN A

761: %STRONGLY FRUSTRATED HEISENBERG- ANTIFERROMAGNET

762: 

763: \bibitem{zeng91}

764: J. B. Marston and C. Zeng, 

765: J. Appl. Phys. {\bf 69}, 5962 (1991).

766: %Spin-Peierls and

767: %spin-liquid phases of Kagome quantum antiferromagnets

768: 

769: \bibitem{sachdev92}

770: S. Sachdev, 

771: Phys. Rev. B {\bf 45}, 12377 (1992).

772: %Kagomé lattice and

773: %triangular lattice Heisenberg antiferromagnets -- ordering from quantum

774: %fluctuations and quantum-disordered groundstates with unconfined

775: %bosonic spinons

776: 

777: \bibitem{chubukov92}

778: A. Chubukov, 

779: Phys. Rev. Lett. {\bf 69}, 832 (1992).

780: %Order from disorder in a

781: %Kagome antiferromagnet

782: 

783: \bibitem{leche97}

784: P. Lecheminant, B. Bernu, C. Lhuillier, L. Pierre and P. Sindzingre,

785: Phys. Rev. B {\bf 56}, 2521 (1997).

786: %Order versus Disorder in the Quantum Heisenberg Antiferromagnet

787: %on the Kagome lattice using exact spectral analysis} 

788: %,cond-mat/9706167.

789: 

790: \bibitem{lac87}

791: P. Lacorre, J. Phys. C: Solid State Phys. {\bf 20}, L775, (1987).

792: 

793: \bibitem{gardner99}

794: J. S. Gardner, S. R. Dunsiger, B. D. Gaulin, M. J. P. Gingras, J. E. Greedan, 

795: R. F. Kiefl, M. D. Lumsden, W. A. MacFarlane, N. P. Raju, J. E. Sonier, I. Swainson, 

796: and Z. Tun, 

797: Phys. Rev. Lett. {\bf 82}, 1012 (1999).

798: 

799: \bibitem{greedan91}

800: J. E. Greedan, J. N. Reimers, C. V. Stager, and S. L. Penny, 

801: Phys. Rev. B {\bf 43}, 5682 (1991); 

802: J. N. Reimers, J. E. Greedan, C. V. Stager, M. Bjorgvinssen, and M. A. Subramanian, 

803: {\it ibid} {\bf 43}, 5692 (1991).

804: 

805: \bibitem{wills99}

806: A. S. Wills, N. P. Raju, and J. E. Greedan, 

807: Chem. Mater. {\bf 11}, 1510 (1999).

808: 

809: \bibitem{greedan98}

810: J. E. Greedan, N. P. Raju, A. S. Wills, C. Morin, and S. M. Shaw, 

811: Chem. Mater. {\bf 10}, 3058 (1998).

812: 

813: \bibitem{fujiwara98}

814: N. Fujiwara, H. Yasuoka, and Y. Ueda, 

815: Phys. Rev. B {\bf 57}, 3539 (1998).

816: 

817: \bibitem{harris97}

818: M. J. Harris and M. P. Zinkin, and T. Zeiske,

819: Phys. Rev. B {\bf 56}, 11786 (1997).

820: 

821: \bibitem{harzintunwanswa94}

822: M. J. Harris, M. P. Zinkin, Z. Tun, B. M. Wankly and I. P. Swainson,

823: Phys. Rev. Lett. {\bf 73}, 189, (1994).

824: 

825: \bibitem{shiga94}

826: M. Shiga, 

827: J. Magn. Magn. Mater. {\bf 129}, 17 (1994).

828: 

829: \bibitem{shiga88}

830: M. Shiga, 

831: Physica B {\bf 149}, 293 (1988).

832: 

833: \bibitem{ballac96}

834: R. Ballou, C. Lacroix, and M. D. Nunez-Regueiro,

835: Phys. Rev. Lett. {\bf 66}, 1910 (1996).

836: 

837: \bibitem{zinharzei97}

838: M. P. Zinkin, M. J. Harris, T. Zeiske,

839: Phys. Rev. B {\bf 56}, 11786, (1997).

840: 

841: \bibitem{garcan99}

842: D. A. Garanin and B. Canals, Phys. Rev. B {\bf 59},  443  (1999).

843: 

844: \bibitem{rei92}

845: J. N. Reimers, Phys. Rev. B {\bf 46},  193  (1992).

846: 

847: %\bibitem{gar94jsp}

848: %{D. A. Garanin}, J. Stat. Phys. {\bf 74},  275  (1994).

849: 

850: 

851: 

852: \end{thebibliography}

853: 

854: 

855: 

856: \end{document}

857: 

858: 

859: tar -cvzf hfm.tar.gz hfm.tex pyr_str.eps pyr_bri.ps pyr_mod.eps pyr_neu_carr.eps pyr_neu_hexa.eps pyr_powd_comp.eps pyr_powd_multi.eps 

860: 

861: 

862: 

863: 

864: 

865: