1: % ****** Start of file apssamp.tex ******
2: %
3: % This file is part of the APS files in the REVTeX 4 distribution.
4: % Version 4.0 of REVTeX, August 2001
5: %
6: % Copyright (c) 2001 The American Physical Society.
7: %
8: % See the REVTeX 4 README file for restrictions and more information.
9: %
10: % TeX'ing this file requires that you have AMS-LaTeX 2.0 installed
11: % as well as the rest of the prerequisites for REVTeX 4.0
12: %
13: % See the REVTeX 4 README file
14: % It also requires running BibTeX. The commands are as follows:
15: %
16: % 1) latex apssamp.tex
17: % 2) bibtex apssamp
18: % 3) latex apssamp.tex
19: % 4) latex apssamp.tex
20: %
21: \documentclass[amsmath,amssymb,prb,floatfix,showpacs]{revtex4}
22: %\documentclass[preprint,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
23: \newcommand{\figwid}{0mm}
24: %Some other (several out of many) possibilities
25: %\documentclass[preprint,aps]{revtex4}
26: %\documentclass[preprint,aps,draft]{revtex4}
27: %\documentclass[prb]{revtex4}% Physical Review B
28:
29: \newcommand{\booo}{BO$_3$}
30: \newcommand{\cm}{cm$^{-1}$}
31: \newcommand{\rfb}{RFe$_3$(BO$_3$)$_4$}
32: \newcommand{\nfb}{NdFe$_3$(BO$_3$)$_4$}
33: \newcommand{\gfb}{GdFe$_3$(BO$_3$)$_4$}
34: \newcommand{\tfb}{TbFe$_3$(BO$_3$)$_4$}
35: \newcommand{\efb}{ErFe$_3$(BO$_3$)$_4$}
36: \newcommand{\yfb}{YFe$_3$(BO$_3$)$_4$}
37:
38: \usepackage{graphicx}% Include figure files
39: \usepackage{dcolumn}% Align table columns on decimal point
40: \usepackage{bm}% bold math
41: \renewcommand{\includegraphics}{\relax}
42:
43: %\nofiles
44:
45: \begin{document}
46:
47: \preprint{APS/123-QED}
48:
49: \title{Raman scattering from phonons and magnons in RFe$_{3}$(BO$_{3}$)$_{4}$}
50: \author{Daniele Fausti}
51: \email{d.fausti@rug.nl}
52: \author{Agung Nugroho}
53: \author{Paul H.M. van Loosdrecht}
54: \email{P.H.M.van.Loosdrecht@rug.nl}
55:
56: \affiliation{Material Science Centre, University of Groningen,
57: 9747 AG Groningen, The Netherlands.}
58:
59: \author{Sergei A.Klimin}
60: \author{Marina N. Popova}
61: \affiliation{Institute of Spectroscopy, RAS, 142190, Troitsk,
62: Moscow Region, Russia.}
63:
64: \author{Leonard N. Bezmaternykh}
65: \affiliation{L.V. Kirensky Institute of Physics, Siberian Branch
66: of RAS, Krasnoyarsk, 660036, Russia.}
67:
68:
69:
70:
71:
72:
73:
74: \date{\today}
75:
76: \begin{abstract}
77: Inelastic light scattering spectra of several members of the
78: RFe$_{3}$(BO$_{3})_{4}$ family reveal a cascade of phase
79: transitions as a function of temperature, starting with a
80: structural, weakly first order, phase transition followed by two
81: magnetic phase transitions. Those consist of the ordering of the
82: Fe-spin sublattice revealed by all the compound, and a subsequent
83: spin-reorientational transition for GdFe$_{3}$(BO$_{3})_{4}$. The
84: Raman data evidence a strong coupling between the lattice and
85: magnetic degrees of freedom in these borates. The Fe-sublattice
86: ordering leads to a strong suppression of the low energy magnetic
87: scattering, and a multiple peaked two-magnon scattering continuum
88: is observed. Evidence for short-range correlations is found in the
89: `paramagnetic' phase by the observation of a broad magnetic
90: continuum in the Raman data, which persists up to surprisingly
91: high temperatures.
92:
93: \end{abstract}
94:
95: \pacs{75.40.Gb (Dynamic properties ~dynamic susceptibility, spin
96: waves, spin diffusion, dynamic scaling, etc.) 63.22.+m (Phonons or
97: vibrational states in low-dimensional structures and nanoscale
98: materials) 75.30.-m (Intrinsic properties of magnetically ordered
99: materials) 64.70.p (Solid–solid transitions) 78.30.-j (Infrared and
100: Raman spectra)}
101:
102: \maketitle
103: \section{Introduction}
104: The family of RM$_{3}$(BO$_{3})_{4}$ crystals (R=Y or a rare
105: earth, M=Al, Ga, Sc, Cr, Fe) has attracted considerable attention
106: during the last years. Different combinations of R and M lead to a
107: large variety of physical properties that, together with their
108: excellent physical characteristics and chemical stability, make
109: these crystals extremely interesting both from application and
110: fundamental points of view. The lack of inversion symmetry (they
111: crystallize in a trigonal space group) has stimulated different
112: applications in the field of optical and optoelectronic devices.
113: Crystals of YAl$_{3}$(BO$_{3})_{4}$ and GdAl$_{3}$(BO$_{3})_{4}$
114: doped with Nd have been widely studied and used in optical
115: devices, for self-frequency doubling and self-frequency summing
116: lasers \cite{jaq01,bre04,che01}. Concentrated
117: NdAl$_{3}$(BO$_{3})_{4}$ crystals are efficient media for
118: mini-lasers \cite{che01}. Rare-earth borates with magnetic ions (M
119: = Cr, Fe) are much less studied. Recently, some iron borates were
120: reported to possess multiferroic features, demonstrating the
121: coexistence of magnetic and ferroelectric order parameters
122: \cite{zve05}; they may therefore be considered for optoelectronic
123: applications. These magnetic borates are also expected to exhibit
124: interesting magnetic properties because of the presence of two
125: different kinds of magnetic ions (3d and 4f elements), and in
126: particular because their structure features isolated helicoidal
127: chains of magnetic 3d atoms.
128:
129: The crystal structure of RM$_{3}$(BO$_{3})_{4}$ borates belongs to
130: the structural type of the natural mineral huntite
131: CaMg$_{3}$(CO$_{3})_{4}$ that crystallizes in the space group
132: $R$32 of the trigonal system \cite{jou68,bel79,cam97,kli05}. The
133: primitive unit cell contains one formula unit. Three kinds of
134: coordination polyhedra are present, trigonal prisms for RO$_{6}$,
135: octahedra for MO$_{6}$, and two types of planar triangular
136: BO$_{3}$ groups: equilateral B1O$_{3}$ and isosceles B2O$_{3}$
137: (the numbering of inequivalent ions position is the same as in
138: Ref. \cite{kli05}). The MO$_{6}$ octahedra share edges forming
139: mutually independent helicoidal chains which run parallel to the
140: $c$-axis (see Fig.\ref{fig1}). The RO$_{6}$ prisms are isolated
141: polyhedra, each of them connects three helicoidal MO$_{6}$ chains,
142: while there are no direct R-O-R links between different RO$_{6}$
143: prisms. This structure can also be seen as a succession of
144: alternating BO$_{3}$ and R+M layers perpendicular to the c axis.
145: \begin{figure}
146: \includegraphics[width=\figwid]{fig1.eps}% Here is how to import EPS art
147: \caption{\label{fig1} (Color online) The structure of
148: RM$_{3}$(BO$_{3}$)$_{4}$ incorporates helicoidal chains of
149: M$^{3+}$ ions. The picture shows the structure of iron borates
150: (M=Fe) viewed from different angles a) perpendicular to the
151: C$_{3}$ axis and b) parallel to it. The possible exchange paths
152: are shown.}
153: \end{figure}
154:
155: Recently, a systematic study of rare-earth iron borates
156: RFe$_{3}$(BO$_{3})_{4}$ (R = Y, La-Nd, Sm-Ho) was undertaken for
157: polycrystalline samples \cite{hin03}. The measurements of
158: magnetization, specific heat, and $^{57}$Fe M\"{o}ssbauer spectra
159: revealed an antiferromagnetic ordering at low temperatures. The
160: magnetic ordering temperature increases with decreasing R$^{3 + }$
161: ionic radius, from 22~K for LaFe$_{3}$(BO$_{3})_{4}$ to 40~K for
162: TbFe$_{3}$(BO$_{3})_{4}$. In addition to this, X-ray diffraction,
163: specific heat measurements, and differential thermal analysis have
164: indicated a structural phase transition for
165: RFe$_{3}$(BO$_{3})_{4}$ compounds with R = Eu-Ho, Y. Its
166: temperature increases with decreasing R$^{3 + }$ ionic radius,
167: from 88~K for EuFe$_{3}$(BO$_{3})_{4}$ to 445~K for
168: YFe$_{3}$(BO$_{3})_{4}$. Magnetic ordering of the La- and Nd-iron
169: borates has been confirmed by spectroscopic and magnetic
170: measurements on non oriented single crystals \cite{cam97,chu04}.
171: Studies on oriented single crystals have only been carried out for
172: GdFe$_{3}$(BO$_{3})_{4}$, using a variety of methods
173: \cite{bal03,lev04,pan04,zve05,kli05,gav04}. Recent specific heat,
174: Raman scattering, and Nd$^{3 + }$ probe absorption measurements
175: revealed a cascade of three spontaneous phase transitions in
176: GdFe$_{3}$(BO$_{3})_{4}$ \cite{lev04}. A structural, weakly
177: first-order, phase transition into a less symmetric
178: low-temperature phase was found at $T_{S}$=156~K (this in contrast
179: to the powder result $T_{S}$=174 K \cite{hin03}). At lower
180: temperature, two magnetic transitions have been observed. A
181: second-order phase transition occurs at 37~K, resulting in an
182: antiferromagnetic ordering of the Fe spin sublattice. This
183: transition is accompanied by a polarization of the Gd spin
184: subsystem. The second magnetic transition is observed at 9~K, and
185: has been identified as a first-order Fe spin-reorientation
186: transition \cite{lev04}. A detailed picture of magnetic phase
187: transitions in GdFe$_{3}$(BO$_{3})_{4}$ as a function of
188: temperature and magnetic field has recently been suggested in Ref.
189: \cite{pan04} on the basis of antiferromagnetic resonance studies.
190:
191: Quite recently, the low-temperature structural phase of
192: GdFe$_{3}$(BO$_{3})_{4}$ has been identified as a $P$3$_{1}$21
193: trigonal structure \cite{kli05}. To our knowledge, no information
194: exists on the low-temperature structure of other rare-earth iron
195: borates. Such information can be obtained from the comparison of
196: low-temperature Raman spectra of GdFe$_{3}$(BO$_{3})_{4}$ and
197: other compounds from the RFe$_{3}$(BO$_{3})_{4}$ family. Raman
198: studies of RFe$_{3}$(BO$_{3})_{4}$ were performed for R=La and Nd
199: at room temperature \cite{and97}. Preliminary temperature
200: dependent results on the R=Gd compound have also been reported
201: recently \cite{lev04}.
202:
203: The present work investigates two different aspects of the
204: rare-earth iron borates. First, inelastic light scattering studies
205: of the vibrational and structural properties of different
206: RFe$_{3}$(BO$_{3})_{4}$ are presented. Amongst others they reveal
207: the weak first-order phase transition from the high-temperature
208: $R$32 structure to the low-temperature $P$3$_{1}$21 one. The
209: second part of this paper focuses on the magnetic properties.
210: Apart from observing changes in phonon frequencies at the
211: antiferromagnetic ordering transition arising from magneto-elastic
212: couplings, the spectra also exhibit relatively strong two-magnon
213: scattering features. The magnetically ordered phases exhibits well
214: defined magnon excitations resulting in a multiple peaked
215: scattering continuum at finite energy. In the paramagnetic phase a
216: strong low energy continuum is found, which results from the
217: persistence of short range correlations in the helicoidal Fe spin
218: chains up to quite high temperatures.
219:
220: \section{EXPERIMENTAL DETAILS}
221: Different samples of RFe$_{3}$(BO$_{3})_{4}$, with R = Gd, Nd, Tb,
222: Er, and Y were grown using a Bi$_{2}$Mo$_{3}$O$_{12}$ - based flux,
223: as described in Ref. \cite{bezm1, bezm2}. Unlike the Bi$_{2}$O$_{3}$
224: based fluxes \cite{cam97}, in Bi$_{2}$Mo$_{3}$O$_{12}$ - based flux
225: Bi$_{2}$O$_{3}$ is strongly bonded to MoO$_{3}$ which excludes a
226: partial substitution of bismuth for a rare earth during
227: crystallization. Spontaneous nucleation from the flux resulted in
228: small single crystals (~1x1x1 mm). We used these crystals as seeds
229: to grow large (~10x5x5 mm) single crystals of Gd, Nd, and Tb iron
230: borates. As for ErFe$_{3}$(BO$_{3})_{4}$ and
231: YFe$_{3}$(BO$_{3})_{4}$, only small single crystals of these
232: compounds were studied in this work. All the crystals were green in
233: color. The samples were oriented either by X-ray diffraction
234: technique or by using their morphology and optical polarization
235: methods.
236:
237: The Raman measurements were performed in a backscattering
238: configuration, using a three-grating micro-Raman spectrometer
239: (T64000 Jobin Yvon) equipped with a liquid nitrogen cooled charged
240: coupled device (CCD) detector. The frequency resolution was better
241: that 1~cm$^{ - 1}$ for the frequency region considered. The
242: samples were placed in a optical microscope cryostat. The
243: temperature was varied from 2.7 to 500~K, with a stability of $\pm
244: $0.02~K. The scattering was excited by the second harmonic light
245: of a Nd:YVO$_{4}$ laser (532~nm), focused down to $\sim $50~$\mu
246: $m$^{2}$ with the power density on the sample kept below
247: 0.1~mW/$\mu $m$^{2}$. The polarization was controlled both on the
248: incoming and outgoing beams giving access to all relevant
249: polarization combinations.
250:
251: Most of the experiments were performed with light incident
252: perpendicular to the C$_{3}$-axis of the crystal,
253: $\vec{k}\perp{c}$ (as a rule, precise orientation of $\vec {k}$ in
254: the (\textit{ab})-plane was not known), however, when the shape of
255: the samples allowed, also the $\vec{k}\vert \vert {c}$
256: configuration has been employed. Spectra are marked using the
257: Porto notation \cite{rou81}. For $\vec {k}\perp{c}$, only the
258: polarization of incident and scattered light are indicated. The
259: conventions z$\vert\vert $C$_{3}$, x$\vert \vert $C$_{2 }$ are
260: used to describe the geometry of the Raman experiments. We were
261: able to obtain only partially polarized Raman spectra of
262: ErFe$_{3}$(BO$_{3}$)$_{4}$ and YFe$_{3}$(BO$_{3}$)$_{4}$.
263:
264: Preliminary dielectric measurements were performed for
265: RFe$_{3}$(BO$_{3})_{4}$ on non oriented crystals of typical size
266: 1x1x0.2~mm$^{3}$. The surfaces were polished and covered with Ag
267: paste and the electrical contacts were made using Pt wires
268: connected to the surface by additional Ag paste. The samples were
269: mounted on a home-made insert for a Quantum Design PPMS system.
270: The capacitance was measured using a Andeen-Hagerling AH2500A
271: capacitance bridge operating at a fixed frequency of 1kHz.
272:
273: \begin{figure}
274: \includegraphics[width=\figwid]{fig2.eps}% Here is how to import EPS art
275: \caption{\label{fig2} Polarized Raman spectra of
276: GdFe$_{3}$(BO$_{3}$)$_{4}$ at room temperature. The bars at the
277: top indicate the frequency regions for the internal vibrations of
278: BO$_{3}$ groups.}
279: \end{figure}
280: \begin{figure}
281: \includegraphics[width=\figwid]{fig3.eps}% Here is how to import EPS art
282: \caption{\label{fig3} (a, b, c) Polarized Raman spectra for
283: RFe$_{3}$(BO$_{3}$)$_{4}$, R=Gd, Nd, Tb, at room temperature. (d,e)
284: Partially polarized Raman spectra for (d) ErFe$_{3}$(BO$_{3}$)$_{4}$
285: at 390 K and (e) YFe$_{3}$(BO$_{3}$)$_{4}$ at 420 K.}
286: \end{figure}
287:
288: \section{EXPERIMENTAL RESULTS}
289: The room-temperature (RT) spectra of GdFe$_{3}$(BO$_{3})_{4}$ for
290: all studied experimental configurations are presented in
291: Fig.\ref{fig2}. The regions where the internal BO$_{3}$ vibrations
292: are expected are indicated by the horizontal bars in the top of
293: the panel. The spectra show a strong dependence on the
294: polarization of the incoming and scattered light, but also show a
295: distinct difference between (xx+yy+xy) polarized spectra for $\vec
296: {k}$ parallel to ${c}$ and for $\vec{k}$ perpendicular to ${c}$.
297: The spectra for different rare earth iron borates are, as
298: expected, qualitatively quite similar, as is shown in
299: Fig.\ref{fig3}.
300:
301: The phonon lines narrow progressively with lowering the
302: temperature and several new modes appear abruptly in
303: GdFe$_{3}$(BO$_{3})_{4}$, TbFe$_{3}$(BO$_{3})_{4}$,
304: ErFe$_{3}$(BO$_{3})_{4}$, and ErFe$_{3}$(BO$_{3})_{4}$, at
305: $T_{S}$=155~K, 198~K, 340~K, and 350~K, respectively. Figures
306: \ref{fig4} and \ref{fig5} and Tab. \ref{tab3} give a detailed
307: comparative picture of the phonon modes observed above and below
308: $T_{S}$ in RFe$_{3}$(BO$_{3})_{4}$. The intensities of the new
309: modes show a hysteresis as a function of temperature (see
310: Fig.\ref{fig6}). The lowest-frequency intense new mode appears
311: abruptly below the phase transition, but then exhibits a
312: soft-mode-like behavior with further lowering of the temperature;
313: i.e. for \gfb~ and \tfb~ the low frequency mode shifts from
314: 26~cm$^{ - 1}$ at $T_{S}$ to respectively 56~cm$^{ - 1}$ and
315: 57~cm$^{ - 1}$ at 3~K, and for \efb~ and \yfb~ it goes
316: respectively from 25 to 72~cm$^{ - 1}$ and from 27 to 76~cm$^{ -
317: 1}$. At the same time its linewidth decreases from 12 to 3~cm$^{ -
318: 1}$ (see Figs.\ref{fig7} and 4b). The spectra in (xz+yz)
319: polarization of all the compounds demonstrate a low-frequency
320: scattering continuum that gradually changes its shape with
321: lowering the temperature and exhibits the opening of a ''gap'' at
322: $T_{N}$ (see Fig.\ref{fig8}). For NdFe$_{3}$(BO$_{3})_{4}$ at the
323: lowest temperatures a well-defined narrow peak at about 10~cm$^{ -
324: 1}$ is observed within the gap. This mode shifts to zero frequency
325: and strongly broadens upon approaching $T_{N}$ from below
326: (Fig.\ref{fig8}(c)). For the Gd compound a similar mode is
327: observed below $T_{N}$ at about 18~cm$^{-1}$.
328:
329: \begin{figure}
330: \includegraphics[width=\figwid]{fig4.eps}% Here is how to import EPS art
331: \caption{\label{fig4} Raman spectra of GdFe$_{3}$(BO$_{3}$)$_{4}$
332: above and below $T_{s}$=156 K, namely, at 161 K (the lowest traces
333: in each panel), 150 K (middle traces), and 30 K (upper traces).
334: (a) (zx+zy) polarization; (b) (zz) polarization. New modes that
335: appear below $T_{s}$ are indicated by arrows. Asterisks mark the
336: lines from another polarization.}
337: \end{figure}
338:
339: \begin{figure}
340: \includegraphics[width=\figwid]{fig5.eps}% Here is how to import EPS art
341: \caption{\label{fig5} Comparison of several low-temperature new
342: modes in different RFe$_{3}$(BO$_{3}$)$_{4}$. Spectra in
343: (xx+xy)-polarization for R=Gd and Tb and partially polarized spectra
344: for R=Y and Er.}
345: \end{figure}
346:
347: \begin{figure}
348: \includegraphics[width=\figwid]{fig6.eps}% Here is how to import EPS art
349: \caption{\label{fig6} The lowest-frequency A$_{1}$-mode appearing in
350: low-temperature $P$3$_{1}$21 structure demonstrates hysteresis in a
351: plot of its intensity vs temperature. The data for Gd(Tb)-compound
352: are presented by open (black) triangles. Inset shows this mode
353: during cooling down for temperatures 154.91, 155.03, 155.08, 155.11,
354: 155.16, 155.20, 155.45 K in Gd-compound. The lower temperature axis
355: refer to R=Gd, the upper one - to Tb.}
356: \end{figure}
357:
358: \section{GROUP-THEORETICAL ANALYSIS}
359: \subsection{High-temperature structure $R$32 ($D_3^7 $)}
360: The primitive cell of the high-temperature structure $R$32 of
361: RFe$_{3}$(BO$_{3})_{4}$ contains 20 atoms which results in 57
362: vibrational normal modes. Knowing the local symmetry of all the
363: atomic positions, one can perform the factor-group analysis to
364: find the symmetries of these modes \cite{mav80,rou81}. R and B1
365: atoms reside in highly symmetric $D_{3}$ positions
366: \cite{cam97,kli05} and generate $A_{2}\oplus E$ modes each. Fe,
367: B2, O1, and O2 atoms are in $C_{2}$ symmetry positions, which
368: results in $A_{1} \oplus $2$A_{2}\oplus$3$E$ modes for each of
369: them. O3 atoms occupy general $C_{1}$ positions and give 3$A_{1}
370: \oplus $3$A_{2} \oplus$6$E$ modes. Summing all these modes and
371: subtracting the $A_{2}\oplus E$ acoustic modes one gets the
372: following optical vibrational modes of the crystal, in accordance
373: with the results of Ref.\cite{and97}.
374:
375: \begin{eqnarray}\label{eq1}
376: \Gamma _{vibr} = 7A_1 \mathrm{(xx,yy,zz)} \oplus 12A_2 (E //\mathrm{z})\nonumber\\
377: \oplus 19E(E / / \mathrm{x},E / / \mathrm{y,xz,yz,xy)}
378: \end{eqnarray}
379:
380: Notations in parentheses refer to the allowed components of
381: electric dipole moment (infrared (IR) activity) and the
382: polarizability tensor (Raman activity). Doubly degenerated E modes
383: are polar and both IR and Raman active. The light propagating
384: along the $c$-axis ($\vec {k}\vert \vert c$) probes TO modes,
385: while the $\vec {k} \bot c$ configuration gives pure LO modes
386: \cite{and97}.
387:
388: The strongest bonds in the structure of RFe$_{3}$(BO$_{3})_{4}$
389: borates are B-O bonds within the planar BO$_{3}$ triangles.
390: Considering them as molecular units that occupy $D_{3}$ (for
391: B1O$_{3})$ and $C_{2}$ (for B2O$_{3})$ positions, and adding the
392: modes of translational character coming from R and Fe ions
393: (respectively in $D_{3}$ and $C_{2}$ position), we obtain the
394: following optical modes:
395:
396: \begin{equation}
397: \label{eq2} \Gamma _{vib}^{tr} = 2A_1 \oplus 5A_2 \oplus 7E; \quad
398: \Gamma _{vib}^{lib} = A_1 \oplus 3A_2 \oplus 4E,
399: \end{equation}
400: where the librational ones (right equation) come from the BO$_{3}$
401: units. The total number of external optical vibrations is, thus,
402:
403: \begin{equation}
404: \label{eq3} \Gamma _{vib}^{ext} = 3A_1 \oplus 8A_2 \oplus 11E
405: \end{equation}
406:
407: Subtracting (\ref{eq2}) from (\ref{eq1}), one gets the internal
408: vibrational modes related to the vibrations of BO$_{3}$ groups:
409:
410: \begin{equation}
411: \label{eq4} \Gamma _{vib}^{int} = 4A_1 \oplus 4A_2 \oplus 8E
412: \end{equation}
413:
414: The correlation scheme of Fig.\ref{fig9} helps to get a deeper
415: insight into the origin of the vibrational modes of
416: RFe$_{3}$(BO$_{3})_{4}$. It follows from this scheme that two of
417: the seven $A_1$ modes originate from the fully symmetric vibration
418: of the B1O$_{3}$ and B2O$_{3}$ groups, two correspond to the
419: B2O$_{3}$ $\nu _{3}$ and $\nu _{4}$ vibrations, and two to the
420: rotational and translational modes of the B2O$_{3}$ group. The
421: remaining $A_{1}$ mode results from displacements of the iron
422: atom. Table \ref{tab1} indicates the origin of crystal modes in
423: the frequency ranges of internal vibrations of the BO$_{3}$
424: molecules.
425:
426: \subsection{Low-temperature structure $P$3$_{1}$21($D_3^4 $)}
427: The primitive cell of the low temperature structure contains three
428: formula units \cite{kli05} (60 atoms) which gives 177 vibrational
429: modes. Unique positions for the R and B1 atoms remain but their
430: symmetry lowers from $D_{3}$ in the $R$32 structure to $C_{2}$ in
431: the $P$3$_{1}$21 one. On the contrary, instead of single threefold
432: $C_{2}$ symmetry positions for Fe, B2, O1, and O2 there are now
433: two different positions, namely, a threefold $C_{2}$ and a sixfold
434: $C_{1}$ positions for each of them. O3 oxygen atoms occupy now
435: three different general ($C_{1})$ positions \cite{kli05}. The
436: factor-group analysis of the low-temperature structure
437: $P$3$_{1}$21 gives the following optical vibrational modes:
438:
439: \begin{equation}
440: \label{eq5} \Gamma_{vib}(P3_121)=27A_1\oplus32A_2\oplus59E
441: \end{equation}
442:
443: Carrying out the same procedure as in the previous subsection one
444: finds what modes correspond to the external (translational and
445: librational) motions of the R and Fe atoms and of the BO$_{3}$
446: entities and to the internal vibrations of the BO$_{3}$
447: "molecules". The results are summarized in Table \ref{tab2}.
448:
449: \section{DISCUSSION}
450: \subsection{Phonons of RFe$_{3}$(BO$_{3}$)$_{4}$ in the $R$32 and $P$3$_{1}$21 phases}
451: First of all, we consider the polarized room-temperature Raman
452: spectra of RFe$_{3}$(BO$_{3})_{4}$, R=Nd, Gd, and Tb
453: (Fig.\ref{fig3}). Table \ref{tab3} summarizes the observed Raman
454: modes of these iron borates at RT. It has been constructed taking
455: into account the results of the group-theoretical analysis for the
456: $R$32 structure and the frequency ranges for normal modes of the
457: regular planar free ion (BO$_{3})^{3 - }$ (they are indicated in
458: Fig.\ref{fig2}; see also Tables \ref{tab1} and \ref{tab2}). The
459: spectra of ErFe$_{3}$(BO$_{3})_{4}$ and YFe$_{3}$(BO$_{3})_{4}$
460: were only partially polarized. Comparing the 390~K spectra of
461: ErFe$_{3}$(BO$_{3})_{4}$ and the 420~K spectra of
462: YFe$_{3}$(BO$_{3})_{4}$ with the RT spectra of other iron borates
463: we complete Table \ref{tab3} for these compounds as well (at room
464: temperature, ErFe$_{3}$(BO$_{3})_{4}$ and YFe$_{3}$(BO$_{3})_{4}$
465: are already in the low-temperature phase).
466:
467: The low-frequency part of the Raman spectrum is dominated by
468: external modes arising from translational motions of the R and Fe
469: atoms and from librational and translational modes of BO$_{3}$
470: groups. The number and polarization properties of the modes observed
471: below 500~cm$^{-1}$ agree with the group-theoretical predictions.
472: All the modes observed in the frequency ranges of the $\nu _{4}$,
473: $\nu _{2}$, and $\nu _{1}$ vibrations of BO$_{3}$ correspond well to
474: those derived using the correlation analysis and summarized in Fig.
475: \ref{fig9}. In contrast, there is one extra mode in the region of
476: the $\nu _{3}$ vibration. Possibly, the strong doublet at about
477: 1220~cm$^{ - 1}$ in the (zx+zy)-polarization arises due to the Fermi
478: resonance \cite{her45} between the $\nu _{3}$ vibrations and an
479: overtone of the $\nu _{4}$ vibration.
480:
481: The highest vibrational frequency (originating from the $\nu_3$
482: vibration of the BO$_{3}$ group) observed in
483: ErFe$_{3}$(BO$_{3})_{4}$ differs markedly from the highest
484: frequencies of both TO ($\vec {k}\parallel c$) and LO ($\vec
485: {k}\perp c$) phonons (1280 and 1414 cm-1, respectively, in
486: GdFe$_{3}$(BO$_{3})_{4}$). This is because in our particular non
487: oriented ErFe$_{3}$(BO$_{3})_{4}$ sample the direction of $\vec
488: {k}$ was arbitrary, while the frequencies of polar E modes depend
489: on the direction of $\vec {k}$.
490:
491: Summarizing the RT results, it is concluded that (i) the number
492: and symmetries of the Raman modes observed for the
493: RFe$_{3}$(BO$_{3})_{4}$ in the high-temperature phase are fully
494: consistent with the predictions of group-theoretical analysis
495: based on the $R$32 space group; (ii) the modes above $\sim
496: $550~cm$^{ - 1}$ originate from the internal vibrations of the
497: BO$_{3}$ groups, while those below $\sim $550~cm$^{ - 1}$ can be
498: considered as external modes.
499:
500: In the low-temperature $P$3$_{1}$21 phase, 20~$A_{1}$ (12 external
501: and 8 internal) and 40~$E$ (24 external an 16 internal) new modes
502: should appear, according to the results of the group-theoretical
503: analysis (see Table \ref{tab2}). These new modes arise due to two
504: reasons. First, the symmetries of the local positions for some of
505: the atoms reduce. The intensity of these new modes should be
506: proportional to the deviation from the former symmetric position.
507: Second, the primitive cell contains now not one but three formula
508: units, leading to an additional Davydov (factor-group) splitting
509: proportional to the strength of interaction between equivalent atoms
510: inside the new primitive cell. Both these effects are, as a rule,
511: small. That is why the number of the observed new modes
512: (10$A_{1}\oplus18E$ in GdFe$_{3}$(BO$_{3})_{4}$) is lower than the
513: number predicted by group-theoretical analysis. The same new modes
514: appear below $T_{s}$ in all the compounds RFe$_{3}$(BO$_{3})_{4}$,
515: with R=Gd, Tb, Er, and Y (see Fig.~5 and the bottom of Table III),
516: which strongly suggests the same $P$3$_{1}$21 low-temperature
517: structure in all of them.
518:
519: \subsection{Weak first-order structural phase transition}
520: The most characteristic Raman feature that announces the phase
521: transition from the high-temperature $R$32 structure to the
522: low-temperature $P$3$_{1}$21 one is the sudden appearance of a
523: strong new low-frequency mode (see Figs.\ref{fig4}, \ref{fig6} and
524: \ref{fig7}) in the Raman spectra. As the intensities of new modes
525: are proportional to the squares of atomic displacements, ${I}\sim
526: \delta ^{2}$, the strongest new mode is believed to be associated
527: with the biggest displacements. A detailed analysis, according to
528: Ref. \cite{kli05}, of the structural changes shows that the
529: biggest displacements are those associated with the BO$_{3}$
530: "molecules". In particular, BO$_{3}$ triangles, perpendicular to
531: the $C_{3}$ axis in the $R$32 structure, tilt by $\sim $ 7$^{o}$
532: in the $P$3$_{1}$21 phase; the B1 atoms shift by $\sim $0.03~{\AA}
533: from the centers of regular triangles. The shifts of boron ions
534: relative to neighboring oxygen ions create local dipole moments;
535: their triangular arrangement corresponds to an antiferroelectric
536: ordering below the temperature of the structural phase transition.
537: This ordering manifests itself via a strong dielectric anomaly at
538: $T_{S}$ observed in our preliminary dielectric measurements (Fig.
539: \ref{fig10}).
540:
541: \begin{figure}
542: \includegraphics[width=\figwid]{fig7.eps}
543: \caption{\label{fig7} (Color online) The lowest-frequency mode
544: appears at the phase transition at a finite frequency
545: (characteristic of a first-order transition), but exhibits a
546: high-frequency shift with further lowering the temperature
547: (characteristic of a second-order transition). The inset shows the
548: derivative of the peak frequency in the vicinity of the magnetic
549: ordering temperature, evidencing the coupling of vibrational and
550: magnetic degrees of freedom (see text for discussion).}
551: \end{figure}
552:
553: The structural changes considered above give rise to many new
554: Raman active vibrational modes connected with the BO$_{3}$ groups
555: (see Table \ref{tab2}), in particular, to 4$A_{1}$ additional
556: librational modes. The energy value of the intense excitation
557: measured is in the typical range of the molecular librations. At 3
558: K, it is 56, 57, 72, and 76 cm$^{-1}$ for Gd(157), Tb(159),
559: Er(167), and Y(89) compounds, respectively. As the values for the
560: Er and Y compounds are very close, notwithstanding a big
561: difference in atomic masses of Er and Y, we conclude that the rare
562: earth does not take part in this lowest-frequency vibration. A
563: difference between the values for the compounds with relatively
564: big (Gd, Tb) and small (Er, Y) ions comes from the difference in
565: interatomic distances and, hence, force constants. The ensemble of
566: these observations strongly suggests that the phase transition in
567: RFe$_{3}$(BO$_{3})_{4}$ is correlated to a rotational mode of the
568: BO$_{3}$ groups.
569:
570: The abrupt appearance of new phonon modes and the hysteresis
571: observed for their intensities when approaching $T_{S}$ from
572: different sides indicate the first-order character of the phase
573: transition. The energy of the low frequency mode strongly
574: increases upon lowering the temperature below $T_{S}$
575: (Fig.\ref{fig7}). At the lower temperature (3K) the frequency of
576: the mode has more than doubled. Such an anomalously large shift is
577: typical for soft modes that announce a second-order structural
578: phase transition. Apparently, we deal with a so-called "weak
579: first-order" phase transition. It is worth noting that the
580: lowering of the symmetry from $R$32 ($D_3^7 )$ to $P$3$_{1}$21
581: ($D_3^4 )$ is indeed also compatible with a second-order phase
582: transition \cite{lan80}. Thus, the observed first-order character
583: of the phase transition does not depend on symmetry changes but
584: rather can arise from the third order therm in the Landau
585: expansion of the crystal free energy $h$, allowed by the crystal
586: symmetry, or from the negative sign of the fourth order
587: coefficient $u$:
588:
589: \begin{equation}
590: \label{eq6} G = G_0 + rQ^2 + hQ^3 + uQ^4 +...,
591: \end{equation}
592: where Q is the order parameter \cite{bru81}.
593:
594: Figure~\ref{fig11} shows the temperatures of the structural phase
595: transition in RFe$_{3}$(BO$_{3})_{4}$ plotted versus ionic radii of
596: R$^{3 + }$, as determined in Ref.~\cite{hin03} for powder samples
597: and in the present study for single crystals. A mismatch between our
598: data and those of Ref.~ \cite{hin03} could arise from a rather poor
599: quality of the Er and Y iron borate single crystals. A further study
600: is necessary to clarify this question.
601:
602: Finally we note that NdFe$_{3}$(BO$_{3})_{4}$ preserves the $R$32
603: structure down to the lowest temperatures, while
604: RFe$_{3}$(BO$_{3})_{4}$ with R$^{3 + }$ smaller than Tb$^{3 + }$
605: have the low-temperature $P$3$_{1}$21 structure already at room
606: temperature.
607: \begin{figure}
608: \includegraphics[width=\figwid]{fig8.eps}% Here is how to import EPS art
609: \caption{\label{fig8} (Color online) Raman spectra of different
610: RFe$_{3}$(BO$_{3}$)$_{4}$ collected in (xz+yz) polarization. The
611: upper picture (a) shows the opening of a gap in the magnon Raman
612: scattering at $T_{N}$ = 37K for GdFe$_{3}$(BO$_{3}$)$_{4}$ (the
613: intensity scale of the left bottom corner of the picture is
614: optimized to evidence the low-frequency mode, and the dotted lines
615: are guides for evidencing the scattering features). The central
616: picture (b,c) show the magnetic scattering around $T_{N}$ for
617: YFe$_{3}$(BO$_{3}$)$_{4}$ and TbFe$_{3}$(BO$_{3}$)$_{4}$. The lower
618: one (d,e) shows the two-magnon spectra in the full temperature range
619: (2.7K$<T<$300K) for NdFe$_{3}$(BO$_{3}$)$_{4}$ on the left and for
620: GdFe$_{3}$(BO$_{3}$)$_{4}$ on the right. For the Gd sample a phonon
621: line is evident in proximity of the magnon feature. LFP indicates
622: the low-frequency peak, see the text for the discussion. Above
623: $T_{N}$ the paramagnon scattering survives up to at least nearly ten
624: times the ordering temperature. Different temperature measurements
625: in the panels b,c,d,e are depicted with an arbitrary offset for
626: clarity.}
627: \end{figure}
628:
629: \subsection{Magnetoelastic coupling}
630: The inset of Fig.\ref{fig7} shows the derivative of the peak
631: frequency of the strongest new $A_{1}$ mode as a function of
632: temperature in the vicinity of the magnetic ordering temperature
633: $T_{N}$. The coupling of vibrational and magnetic degrees of
634: freedom is evident. The discontinuity in the phonon frequency
635: observed at $T=T_{N}$ is ascribed to magneto-elastic coupling. The
636: magnetic ordering causes spontaneous magnetostriction, that is,
637: atomic displacements. The latter, evidently, influence vibrational
638: frequencies. Recently, Ref.\cite{zve05} reported on the study of
639: magnetoelectric and magnetoelastic properties of
640: GdFe$_{3}$(BO$_{3})_{4}$. A strict correlation between both has
641: been shown experimentally and discussed theoretically. The authors
642: of Ref.\cite{zve05} assumed that, below $T_{N}$, the crystal
643: symmetry lowers, due to magnetoelastic coupling with spontaneous
644: magnetic moments lying in the ab-plane. This breaks the symmetry
645: of the antiferroelectric arrangement of the electric dipole
646: moments in the $P$3$_{1}$21 structure and leads to the appearance
647: of an electric polarization which could be responsible for the
648: weak growth of the dielectric constant in the temperature region
649: $T_{R}<T<T_{N}$ \cite{zve05}, (see also Fig. \ref{fig10}). The
650: proposed lowering of crystal symmetry below $T_{N}$, however,
651: should result in the appearance of new vibrational modes. No
652: indications for additional modes below $T_{N}$ have been found.
653: This does not necessarily mean that the model by Zvezdin
654: \textit{et al.} is in error. It might very well be that the
655: structural changes are simply too small to be detected by Raman
656: scattering.
657:
658: \begin{figure}
659: \includegraphics[width=\figwid]{fig9.eps}% Here is how to import EPS art
660: \caption{\label{fig9} Correlation scheme for vibrational modes of
661: RFe$_{3}$(BO$_{3}$)$_{4}$, space group $R$32.}
662: \end{figure}
663:
664: As Reference \cite{zve05} suggests, the spin-reorientation
665: first-order phase transition at $T_{R}$ into the antiferromagnetic
666: configuration of spins parallel to the c-axis recovers the
667: $P$3$_{1}$21 structure (that does not allow the presence of the
668: electric dipole moment) and, therefore, is expected to be
669: accompanied by a jump of the dielectric constant. As is clear from
670: Fig.~\ref{fig10}, the dielectric function indeed exhibits such a
671: jump, thereby strongly supporting the suggested scenario.
672:
673: \subsection{Magnetic scattering}
674: As discussed in the introduction, due to the structural properties
675: of the RFe$_{3}$(BO$_{3})_{4}$ family, Fe atoms (S=5/2) are
676: arranged in helicoidal chains. As a result, the Fe-Fe distance
677: inside the chain is substantially smaller than the distance
678: between different chains. The main exchange interaction is
679: therefore expected to be between Fe ions on the same chain ($J_{ /
680: / } \rangle \rangle J_ \bot )$, suggesting that low dimensionality
681: could play a crucial role in the magnetic properties of the
682: family. This, together with the presence of two coupled magnetic
683: sublattices makes the compounds with magnetic R$^{3 + }$ ions good
684: candidates for exotic magnetic ground states.
685:
686: \begin{figure}
687: \includegraphics[width=\figwid]{fig10.eps}% Here is how to import EPS art
688: \caption{\label{fig10} Dielectric measurements of
689: GdFe$_{3}$(BO$_{3}$)$_{4}$. The structural phase transition is made
690: evident by a step in the capacitance at $T_{S}$. No evidence is
691: found of the Fe ordering phase transition at $T_{N}$, while the
692: capacitance shows a discontinuity at the spin-reorientation phase
693: transition at $T_{R}$ (displayed in more detail in the Inset).}
694: \end{figure}
695:
696: \begin{figure}
697: \includegraphics[width=\figwid]{fig11.eps}% Here is how to import EPS art
698: \caption{\label{fig11} (Color online) Temperatures of the structural
699: phase transitions for different RFe$_{3}$(BO$_{3}$)$_{4}$ as a
700: function of the ionic radius. Our data on single crystals are shown
701: by circles, those of Ref. \cite{hin03} on powder samples are
702: presented by squares.}
703: \end{figure}
704:
705: A complex magnetic behavior has indeed been found and widely
706: discussed in GdFe$_{3}$(BO$_{3})_{4}$. The first magnetization
707: measurements on oriented single crystals were interpreted under
708: the assumption that both Fe and Gd subsystems order
709: antiferromagnetically at $T_{N}$=40 K~\cite{bal03}. In this model,
710: below 10~K, the total magnetic moment of the three Fe sublattices
711: (S~=~5/2) oriented at polar angles $\sim $ 60\r{ } to the $c$ axis
712: is compensated by the Gd moment (S~=~7/2) oriented along the $c$
713: axis. Sharp singularities in magnetization in the magnetic field
714: H$\vert \vert c$ observed below 10 K were attributed to spin-flop
715: transitions in the Fe subsystem. It was assumed further that in
716: the range of temperatures between 10 and~40 K the magnetic moments
717: of every Fe sublattice fall into the \textit{ab} plane,
718: perpendicular to the $c$ axis, preserving the 120\r{ } azimuthal
719: orientation.
720:
721: This model was later reconsidered in Ref.~\cite{lev04}. It has been
722: suggested, on the basis of the analysis of the Nd-probe spectra and
723: of the specific heat data, that the Gd subsystem does not undergo a
724: magnetic phase transition at $T_{N}$ but only gets polarized by the
725: Fe subsystem. The following model has been put forward. At
726: $T_{N}$=37~K, the iron magnetic moments order antiferromagnetically
727: in the direction perpendicular to the c-axis and polarize the Gd
728: subsystem. Below $T_{R}$=9~K, the magnetic moments order
729: antiferromagnetically along the c-axis. The temperature $T_{R}$=9~K
730: corresponds to a first-order spin-reorientation magnetic phase
731: transition \cite{lev04}. Simultaneously and independently, the same
732: conclusion has been drawn in Ref.\cite{pan04} where magnetic
733: structures and magnetic phase transitions in
734: GdFe$_{3}$(BO$_{3})_{4}$ were studied using antiferromagnetic
735: resonance experiments. This study revealed also a detailed picture
736: of the magnetic structure and anisotropy for different magnetic
737: phases in GdFe$_{3}$(BO$_{3})_{4}$ as a function of both temperature
738: and magnetic field. At $T_{N}$=38~K, the Fe magnetic subsystem
739: orders into a two-sublattice collinear easy-plane antiferromagnet
740: and polarizes the Gd spins, which also form a two-sublattice
741: antiferromagnetic subsystem. The anisotropy constant of the Gd
742: subsystem has an opposite sign to that of the Fe subsystem. The Gd
743: contribution to the total anisotropy energy grows with lowering the
744: temperature, in conjunction with the growing polarization of the Gd
745: subsystem, and becomes appreciable below $\sim $20~K. At
746: $T_{R}$=10~K the total energy of anisotropy changes its sign which
747: results in the spontaneous spin reorientation transition.
748:
749: The long-range order of the spins on the iron subsystem manifests
750: itself in the inelastic light scattering experiments with two main
751: features. The first one, as reported in Fig.\ref{fig8}, is the
752: arising of a low-frequency peak (LFP) at $\sim $10~cm$^{ - 1}$ in
753: NdFe$_{3}$(BO$_{3})_{4}$ and at $\sim $18~cm$^{ - 1}$ in
754: GdFe$_{3}$(BO$_{3})_{4}$. These peaks soften and broaden upon
755: approaching $T_{N}$. The energy of 10~cm$^{ - 1}$ in
756: NdFe$_{3}$(BO$_{3})_{4}$ has been identified, by absorption
757: spectroscopy of Nd$^{3 + }$ crystal-field levels, as the exchange
758: splitting energy of the Nd$^{3 + }$ ground Kramers doublet at 5 K
759: arising from the interaction with the ordered Fe sublattice
760: \cite{chu04}. The temperature dependences of the frequency and the
761: linewidth of the Raman LFP are the same as those of the ground-state
762: splitting and, respectively, the ground-level width found from
763: optical spectroscopy \cite{chu04}. Therefore the 10~cm$^{ - 1}$
764: Raman mode observed in NdFe$_{3}$(BO$_{3})_{4}$ is assigned to spin
765: flip scattering on a single Nd$^{3 + }$ moment in the effective
766: field created by the Fe sublattice. Most likely, the 18~cm$^{ - 1}$
767: excitation observed in GdFe$_{3}$(BO$_{3})_{4}$ has the same origin.
768: The absence of LFP in Raman scattering of \yfb ~with nonmagnetic
769: Y$^{3+}$ (see Fig. \ref{fig8}) is in favor of such interpretation.
770: The crystal field of \rfb ~splits the ground-state multiplet of an
771: ion R$^{3+}$ with odd number of electrons into $\Gamma_{4}$ and
772: $\Gamma_{56}$ Kramers doublets within the D$_3$ site symmetry. The
773: analysis of Raman selection rules for the ground-state spin flip
774: scattering shows that in the case of the $\Gamma_{4}$ ground state
775: all the components of the Raman tensor are allowed while in the case
776: of the $\Gamma_{56}$ ground state only the antisymmetric xy-yx
777: component is Raman active. Thus, a relatively strong (xz) Raman LFP
778: of \nfb~ points to the $\Gamma_{4}$ symmetry of the Nd$^{3+}$ ground
779: state.
780:
781: The second major feature of the iron spin ordering is the arising
782: of a broad structured scattering band around 60~cm$^{ - 1}$
783: ascribed to two-magnon Raman scattering involving the creation of
784: a pair of magnons with wave vectors $\vec{k}$ and -$\vec{k}$.
785: Below the N\'{e}el temperature this is a characteristic feature
786: for all the different compound investigated of the
787: RFe$_{3}$(BO$_{3})_{4}$ family (R=Y, Er, Tb, Gd and Nd). Figure
788: \ref{fig8} shows the temperature evolution of this broad signature
789: of magnetic scattering. The spectra reported in Fig.\ref {fig8}
790: are found in the (xz+yz) polarization only. No evidence of
791: magnetic scattering is found when the $\vec{k}$ vector of light is
792: parallel to the c-axis (z(xx+yy+xy)z).
793:
794: \begin{figure}
795: \includegraphics[width=\figwid]{fig12.eps}% Here is how to import EPS art
796: \caption{\label{fig12} (Color online) First moment $<\omega>$ of the
797: magnetic excitation vs temperature calculated from 5 to 85~cm$^{-1}$
798: for GdFe$_{3}$(BO$_{3}$)$_{4}$ (circles) and
799: NdFe$_{3}$(BO$_{3}$)$_{4}$ (squares). The inset shows the integrated
800: intensity of the magnetic signal at different temperature, compared
801: to the magnetic heat capacity of YFe$_{3}$(BO$_{3}$)$_{4}$ (solid
802: line)\cite{hin03}.}
803: \end{figure}
804:
805:
806: As discussed in the previous section, NdFe$_{3}$(BO$_{3})_{4}$
807: does not undergo a structural phase transition and the crystal
808: space group remains the high temperature one ($R$32), while the
809: symmetry of all the other compounds investigated (for R=Gd, Tb, Y,
810: and Er) is reduced to the space group $P$3$_{1}$21. It is evident
811: from Fig.\ref{fig8} that the magnetic scattering spectra of all
812: iron borates are quite similar. So, the presence of inequivalent
813: Fe chains \cite{kli05} in the low-temperature structure of
814: RFe$_{3}$(BO$_{3})_{4}$, R~=~Gd, Tb, Y, and Er does not strongly
815: affect the magnetic excitation spectra, and the spectra can be
816: analyzed taking into account only the four different magnetic ions
817: present in the $R$32 unit cell (three Fe$^{3 + }$ ions and one
818: R$^{3 + }$ ion). Still, a rather complicated two-magnon spectrum
819: is expected, strongly depending on the anisotropy parameters. At
820: the lowest temperature reached (T=2.7 K) the two-magnon spectra
821: indeed show a complex structure exhibiting at least three main
822: peaks (Fig.\ref{fig8}). As the most efficient mechanism of
823: two-magnon scattering in antiferromagnets is usually the
824: exchange-scattering mechanism, and the strongest exchange
825: interaction is between the iron atoms along the helicoidal chains,
826: these peaks presumably arise from three magnon branches
827: representing spin excitations on the iron chains.
828:
829: The broad magnetic scattering feature of Gd, Tb, Er and Y
830: compounds is approximately centered around $\simeq$70~cm$^{ - 1}$,
831: while the one of Nd is centered at lower frequency
832: ($\simeq$60~cm$^{ - 1}$). This suggests that the bigger ionic
833: radius of Nd$^{3 + }$, causes the interatomic Fe-Fe distances to
834: be larger, and therefore the exchange interaction to be smaller,
835: this would be consistent with the scaling of $T_{N}$ with the rare
836: earth's ionic radii \cite{hin03}. Unfortunately, due to the
837: presence of phonon modes at the same frequency of the magnetic
838: excitation, it is not possible to determine the frequency scaling
839: of the magnetic scattering among the other compounds (Gd, Tb, Er
840: and Y) to confirm it.
841:
842: At the temperature of the spin-reorientational transition
843: ($T_{R})$, no drastic changes are observed in the magnon spectra.
844: This observation confirms that the observed scattering is mainly
845: due to magnetic excitations on the Fe$^{3 + }$ chains, and sheds
846: light on the proposed exotic magnetic ordering of Fe spins for the
847: GdFe$_{3}$(BO$_{3})_{4}$ compound. The configuration with
848: 120$^\circ$ angle between nearest neighbor Fe$^{3 + }$ spins for
849: $T_{N}>T>T_{R}$ which then reorient at $T_{R}$ into an easy axis
850: antiferromagnet along the c-axis as proposed in \cite{bal03} is
851: not consistent with our observation. Such a drastic change in the
852: spin configuration would induce a major change in the two-magnon
853: dispersion and inelastic scattering spectra. The picture of a
854: reorientational phase transition at $T_{R}$, as proposed in
855: \cite{pan04,lev04}, survives assuming an easy-axis/easy-plane
856: anisotropy perpendicular to the c-axis in the temperature region
857: $T_{R}<T<T_{N}$. Indeed, considering a simple Heisenberg-type
858: Hamiltonian, where the interaction depends only on the nearest
859: neighbor scalar product, an easy axis anisotropy perpendicular or
860: parallel to the c-axis would produce similar magnetic excitation
861: spectra and therefore similar two-magnon spectra.
862:
863: A striking feature of the magnetic scattering, as shown in Fig.
864: \ref{fig8}, is the persistence of the magnetic scattering in an
865: anomalous shape up to extremely high temperature above $T_{N}$. In
866: the three dimensional antiferromagnet NiF$_{2}$ this ``paramagnon
867: scattering'' has been observed well into the paramagnetic phase up
868: to 4$T_{N}$ \cite{fle69}. The first moment of the frequency
869: $<\omega>$ as a function of temperature in NiF$_{2}$ and MnF$_{2}$
870: exhibited a fast decrease to lower frequency ($\simeq10\%$ of the
871: ''0 temperature value") at $T_{N}$ and a subsequent smooth
872: decrease to zero for T $>$ $T_{N}$, in excellent qualitative
873: agreement with the theory prediction for the moments \cite{cot86}.
874: Also in the case of RFe$_{3}$(BO$_{3})_{4}$ the two-magnon
875: spectrum softens continuously its frequency approaching the
876: N\'{e}el temperature (Fig.\ref{fig8}). Yet, surprisingly, the
877: paramagnon scattering survives at all measured temperatures above
878: $T_{N}$ ($i.e.$ up to about 10 times $T_{N})$. As reported in Fig.
879: \ref{fig12}, the first moment of the excitation decreases at
880: $T_{N}$ less than expected and approaches zero only very slowly
881: above $T_{N}$. Considering NdFe$_{3}$(BO$_{3})_{4}$, the minimum
882: at $T_{N}$ and subsequent increases with temperature up to
883: 2.5~$T_{N}$ of the first moment is possibly due to the presence of
884: luminescence lines overlapping with features of the magnetic
885: scattering. This is likely also the origin of the temperature
886: shift to high frequency (lower absolute energy) of the excitation
887: observed at around 80cm$^{-1}$.
888:
889: \begin{figure}
890: \includegraphics[width=\figwid]{fig13.eps} \caption{(Color
891: online) Low frequency Raman scattering above the N\'eel temperature
892: for GdFe$_{3}$(BO$_{3})_{4}$. The different temperature are depicted
893: with an offset for clarity. The paramagnon scattering is present up
894: to almost $4T_{N}$. The fit is obtained with an fenomenological
895: curve $f(\omega)=\frac{G}{1+aG}+L$, where $G$ is a Gaussian
896: distribution centered at 0 frequency, and $L$ is a Lorenzian
897: distribution to take into account the presence of the vibrational
898: mode (see text for discussion).} \label{fig13}
899: \end{figure}
900:
901: From a theoretical point of view the presence of quasi-elastic
902: scattering well into the paramagnetic phase of three-dimensional
903: antiferromagnets has been ascribed to two possible origins,
904: namely, spin diffusion \cite{rich74} and spin density fluctuations
905: or energy diffusion \cite{hall78}. In the scenario of spin
906: diffusion, absent in a perfect one-dimensional system, the
907: four-spin correlation function (describing the Raman response) is
908: well approximated by a Gaussian shape centered at zero energy. In
909: contrast, for the energy diffusion scenario one expects a
910: lorentzian lineshape in the absence of strong spin-lattice
911: interactions. The integrated intensity $I_m$ in the energy
912: diffusion scenario is expected to follow the magnetic contribution
913: to the heat capacity $C_m$ as $I_m\propto C_m
914: T^2$\cite{rich74,hall78}. In the presence of strong spin-lattice
915: coupling this later relation is expected to remain valid, even
916: though the low energy scattering is expected to broaden
917: significantly due to the rapid relaxation of the magnetic
918: excitation energy into the lattice.
919:
920: The shape of the low-energy scattering feature in the paramagnetic
921: phase is different from any previously reported. Neither a
922: Lorenzian nor a Gaussian distribution can fit our data. This
923: anomalous shape of the paramagnon scattering continuum is a common
924: feature of all the compounds investigated; Fig.\ref{fig13} depicts
925: the observed spectra of GdFe$_{3}$(BO$_{3})_{4}$ as an example.
926: The observation that the spectra can not be modelled with either a
927: Gaussian nor a Lorentzian response might evidence the presence of
928: important spin-lattice interactions, in line with the observations
929: discussed in the previous section. A comparison of the integrated
930: intensity for \gfb\ and \nfb\ to the magnetic heat capacity
931: determined by Hinatsu {\em et al.}\cite{hin03} (see inset fig. 12)
932: for \yfb\ does not show a fair agreement in the `paramagnetic'
933: phase, even though there is a good correlation below the phase
934: transition. Apparently energy diffusion, combined with
935: spin-lattice coupling does not explain the observed scattering. On
936: the other hand, as shown in Figure \ref{fig13} the spectra for
937: GdFe$_{3}$(BO$_{3})_{4}$ can be fitted nicely at all temperature
938: with a broad phenomenological function consisting of a
939: renormalized Gaussian distribution
940: $I(\omega)=\frac{G(\omega)}{1+\alpha G(\omega)}+L$, where
941: $G(\omega)$ is a pure Gaussian, and $L(\omega)$ is a lorentzian to
942: account for the phonon line observed in the spectra. This equation
943: is reminescent of the response function of interacting magnons in
944: low dimensional quantum spin systems, although it would be unclear
945: why this would still hold for a $S=5/2$ system at high
946: temperatures. At present one can not draw a definite conclusion,
947: but the results do strongly suggest that short range order
948: correlations in the spin system survives up to quite high
949: temperatures. Most likely an interplay between the low
950: dimensionality and strong spin-lattice interactions are
951: responsible for the observed unique behavior of the magnetic
952: scattering up to several times $T_{N }$, and deserves further
953: studies.
954:
955: \section{Summary}
956: In summary, structural, magnetic, and magneto-elastic properties
957: of various members of the RFe$_{3}$(BO$_{3})_{4}$ family (R~=~Gd,
958: Tb, Nd, Er, and Y) have been studied using primarily inelastic
959: light scattering. The compounds show a weak first-order structural
960: phase transition which results in the activation of a strong mode
961: of BO$_{3}$ librational nature, and its subsequent temperature
962: evolution. The scattering spectra observed in the low-temperature
963: phase are fully consistent with the low-temperature structure
964: reported earlier. A detailed analysis of the vibrational spectra
965: at the temperature of the magnetic ordering transition showed a
966: strong magneto-elastic coupling for all the compounds. The
967: analysis of the two-magnon Raman scattering in the magnetic phases
968: showed qualitatively similar scattering spectra for the five
969: compounds investigated, evidencing that the structural differences
970: between them do not strongly affect the magnetism and the magnetic
971: excitation spectra. Finally an unprecedented an intriguing
972: paramagnon scattering continuum up to quite high temperatures. The
973: origin of this scattering does not seem to be the usual energy
974: diffusion observed in a variety of other magnetic system, and
975: seems to indicate the presence of short-range order spin-spin
976: correlations arising from the low-dimensional nature of these
977: compounds.
978:
979:
980: \begin{acknowledgments}
981: The authors are grateful to M. Mostovoy and D. Khomskii for
982: valuable discussions. This work was partially supported by the
983: Stichting voor Fundamenteel Onderzoek der Materie [FOM,
984: financially supported by the Nederlandse Organisatie voor
985: Wetenschappelijk Onderzoek (NWO)]. SAK, MNP, and LNB acknowledge
986: the support of the Russian Foundation for Basic Research, grants ¹
987: 04-02-17346 and ¹ 03-02-16286, and the Russian Academy of Sciences
988: under the Programs for Basic Research.
989: \end{acknowledgments}
990:
991:
992:
993: \begin{thebibliography}{26}
994: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
995: \expandafter\ifx\csname bibnamefont\endcsname\relax
996: \def\bibnamefont#1{#1}\fi
997: \expandafter\ifx\csname bibfnamefont\endcsname\relax
998: \def\bibfnamefont#1{#1}\fi
999: \expandafter\ifx\csname citenamefont\endcsname\relax
1000: \def\citenamefont#1{#1}\fi
1001: \expandafter\ifx\csname url\endcsname\relax
1002: \def\url#1{\texttt{#1}}\fi
1003: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
1004: \providecommand{\bibinfo}[2]{#2}
1005: \providecommand{\eprint}[2][]{\url{#2}}
1006:
1007: \bibitem[{\citenamefont{Jaque}(2001)}]{jaq01}
1008: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Jaque}}, \bibinfo{journal}{J.
1009: of Alloys and Compounds} \textbf{\bibinfo{volume}{323-324}},
1010: \bibinfo{pages}{204} (\bibinfo{year}{2001}).
1011:
1012: \bibitem[{\citenamefont{Brenier et~al.}(2004)\citenamefont{Brenier, Tu, Zhu,
1013: and Wu}}]{bre04}
1014: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Brenier}},
1015: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Tu}},
1016: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Zhu}}, \bibnamefont{and}
1017: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Wu}}, \bibinfo{journal}{Appl.
1018: Phys. Lett.} \textbf{\bibinfo{volume}{84}}, \bibinfo{pages}{2034}
1019: (\bibinfo{year}{2004}).
1020:
1021: \bibitem[{\citenamefont{Chen et~al.}(2001)\citenamefont{Chen, Luo, Jaque,
1022: Romero, {Garcia~Sol{\'{e}}}, Huang, Jiang, and Tu}}]{che01}
1023: \bibinfo{author}{\bibfnamefont{X.}~\bibnamefont{Chen}},
1024: \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Luo}},
1025: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Jaque}},
1026: \bibinfo{author}{\bibfnamefont{J.~J.} \bibnamefont{Romero}},
1027: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{{Garcia~Sol{\'{e}}}}},
1028: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Huang}},
1029: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Jiang}}, \bibnamefont{and}
1030: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Tu}}, \bibinfo{journal}{J.
1031: Phys. :Cond. Mat.} \textbf{\bibinfo{volume}{13}}, \bibinfo{pages}{1171}
1032: (\bibinfo{year}{2001}).
1033:
1034: \bibitem[{\citenamefont{Zvezdin et~al.}(2005)\citenamefont{Zvezdin, Krotov,
1035: Kadomtseva, Vorob'ev, Popov, Pyatakov, Bezmaternykh, and Popova}}]{zve05}
1036: \bibinfo{author}{\bibfnamefont{A.~K.} \bibnamefont{Zvezdin}},
1037: \bibinfo{author}{\bibfnamefont{S.~S.} \bibnamefont{Krotov}},
1038: \bibinfo{author}{\bibfnamefont{A.~M.} \bibnamefont{Kadomtseva}},
1039: \bibinfo{author}{\bibfnamefont{G.~P.} \bibnamefont{Vorob'ev}},
1040: \bibinfo{author}{\bibfnamefont{Y.~F.} \bibnamefont{Popov}},
1041: \bibinfo{author}{\bibfnamefont{A.~P.} \bibnamefont{Pyatakov}},
1042: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1043: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Popova}},
1044: \bibinfo{journal}{JETP Lett.} \textbf{\bibinfo{volume}{81}},
1045: \bibinfo{pages}{272} (\bibinfo{year}{2005}).
1046:
1047: \bibitem[{\citenamefont{Joubert et~al.}(1968)\citenamefont{Joubert, White, and
1048: Roy}}]{jou68}
1049: \bibinfo{author}{\bibfnamefont{J.~C.} \bibnamefont{Joubert}},
1050: \bibinfo{author}{\bibfnamefont{W.~B.} \bibnamefont{White}}, \bibnamefont{and}
1051: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Roy}}, \bibinfo{journal}{J.
1052: Appl. Cryst.} \textbf{\bibinfo{volume}{1}}, \bibinfo{pages}{318}
1053: (\bibinfo{year}{1968}).
1054:
1055: \bibitem[{\citenamefont{Belokoneva et~al.}(1979)\citenamefont{Belokoneva,
1056: Alshinskaya, Simonov, Leonyuk, Timchenko, and Belov}}]{bel79}
1057: \bibinfo{author}{\bibfnamefont{E.~L.} \bibnamefont{Belokoneva}},
1058: \bibinfo{author}{\bibfnamefont{L.~I.} \bibnamefont{Alshinskaya}},
1059: \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Simonov}},
1060: \bibinfo{author}{\bibfnamefont{N.~I.} \bibnamefont{Leonyuk}},
1061: \bibinfo{author}{\bibfnamefont{T.~I.} \bibnamefont{Timchenko}},
1062: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{N.~V.} \bibnamefont{Belov}},
1063: \bibinfo{journal}{Journal Sructurnoi Khimii (Russian Journal of Structural
1064: Chemistry)} \textbf{\bibinfo{volume}{20}}, \bibinfo{pages}{542}
1065: (\bibinfo{year}{1979}).
1066:
1067: \bibitem[{\citenamefont{Campa et~al.}(1997)\citenamefont{Campa, Cascales,
1068: Gutierrez-Puebla, Monge, Rasines, and Ruiz-Valero}}]{cam97}
1069: \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Campa}},
1070: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Cascales}},
1071: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Gutierrez-Puebla}},
1072: \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Monge}},
1073: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Rasines}}, \bibnamefont{and}
1074: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Ruiz-Valero}},
1075: \bibinfo{journal}{Chem. Mater.} \textbf{\bibinfo{volume}{9}},
1076: \bibinfo{pages}{237} (\bibinfo{year}{1997}).
1077:
1078: \bibitem[{\citenamefont{Klimin et~al.}(2005)\citenamefont{Klimin, Fausti,
1079: Meetsma, Bezmaternykh, van Loosdrecht, and Palstra}}]{kli05}
1080: \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Klimin}},
1081: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Fausti}},
1082: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Meetsma}},
1083: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1084: \bibinfo{author}{\bibfnamefont{P.~H.~M.} \bibnamefont{van Loosdrecht}},
1085: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{T.~T.~M.}
1086: \bibnamefont{Palstra}}, \bibinfo{journal}{Acta Cryst. B}
1087: \textbf{\bibinfo{volume}{61}}, \bibinfo{pages}{481} (\bibinfo{year}{2005}),
1088: \urlprefix\url{http://xxx.lanl.gov/abs/cond-mat/0502423}.
1089:
1090: \bibitem[{\citenamefont{Hinatsu et~al.}(2003)\citenamefont{Hinatsu, Doi, Ito,
1091: Wakeshima, and Alemi}}]{hin03}
1092: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Hinatsu}},
1093: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Doi}},
1094: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ito}},
1095: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Wakeshima}},
1096: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Alemi}},
1097: \bibinfo{journal}{J. Solid State Chem.} \textbf{\bibinfo{volume}{172}},
1098: \bibinfo{pages}{438} (\bibinfo{year}{2003}).
1099:
1100: \bibitem[{\citenamefont{Chukalina et~al.}(2004)\citenamefont{Chukalina,
1101: Kuritsin, Popova, Bezmaternykh, Kharlamova, and Temerov}}]{chu04}
1102: \bibinfo{author}{\bibfnamefont{E.~P.} \bibnamefont{Chukalina}},
1103: \bibinfo{author}{\bibfnamefont{D.~Y.} \bibnamefont{Kuritsin}},
1104: \bibinfo{author}{\bibfnamefont{M.~N.} \bibnamefont{Popova}},
1105: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1106: \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Kharlamova}},
1107: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{V.~L.}
1108: \bibnamefont{Temerov}}, \bibinfo{journal}{Phys. Lett. A}
1109: \textbf{\bibinfo{volume}{322}}, \bibinfo{pages}{239} (\bibinfo{year}{2004}).
1110:
1111: \bibitem[{\citenamefont{Balaev et~al.}(2003)\citenamefont{Balaev, Bezmaternyh,
1112: Gudim, Temerov, Ovchinnikov, and Kharlamova}}]{bal03}
1113: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Balaev}},
1114: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternyh}},
1115: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Gudim}},
1116: \bibinfo{author}{\bibfnamefont{V.~L.} \bibnamefont{Temerov}},
1117: \bibinfo{author}{\bibfnamefont{S.~G.} \bibnamefont{Ovchinnikov}},
1118: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.~A.}
1119: \bibnamefont{Kharlamova}}, \bibinfo{journal}{JMMM}
1120: \textbf{\bibinfo{volume}{258-259}}, \bibinfo{pages}{532}
1121: (\bibinfo{year}{2003}).
1122:
1123: \bibitem[{\citenamefont{Levitin et~al.}(2004)\citenamefont{Levitin, Popova,
1124: Chtsherbov, Vasiliev, Popova, Chukalina, Klimin, van Loosdrecht, Fausti, and
1125: Bezmaternykh}}]{lev04}
1126: \bibinfo{author}{\bibfnamefont{R.~Z.} \bibnamefont{Levitin}},
1127: \bibinfo{author}{\bibfnamefont{E.~A.} \bibnamefont{Popova}},
1128: \bibinfo{author}{\bibfnamefont{R.~M.} \bibnamefont{Chtsherbov}},
1129: \bibinfo{author}{\bibfnamefont{A.~N.} \bibnamefont{Vasiliev}},
1130: \bibinfo{author}{\bibfnamefont{M.~N.} \bibnamefont{Popova}},
1131: \bibinfo{author}{\bibfnamefont{E.~P.} \bibnamefont{Chukalina}},
1132: \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Klimin}},
1133: \bibinfo{author}{\bibfnamefont{P.~H.~M.} \bibnamefont{van Loosdrecht}},
1134: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Fausti}}, \bibnamefont{and}
1135: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1136: \bibinfo{journal}{JETP Lett.} \textbf{\bibinfo{volume}{79}},
1137: \bibinfo{pages}{423} (\bibinfo{year}{2004}).
1138:
1139: \bibitem[{\citenamefont{Pancratz et~al.}(2004)\citenamefont{Pancratz,
1140: Petrakovskii, Bezmaternykh, and Bayukov}}]{pan04}
1141: \bibinfo{author}{\bibfnamefont{A.~I.} \bibnamefont{Pancratz}},
1142: \bibinfo{author}{\bibfnamefont{G.~A.} \bibnamefont{Petrakovskii}},
1143: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1144: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{O.~A.}
1145: \bibnamefont{Bayukov}}, \bibinfo{journal}{JETP}
1146: \textbf{\bibinfo{volume}{99}}, \bibinfo{pages}{766} (\bibinfo{year}{2004}).
1147:
1148: \bibitem[{\citenamefont{Gavriliuk et~al.}(2004)\citenamefont{Gavriliuk,
1149: Kharlamova, Lyubutin, Troyan, Ovchinnikov, Potsjelujko, Eremets, and
1150: B\"{o}hler}}]{gav04}
1151: \bibinfo{author}{\bibfnamefont{A.~G.} \bibnamefont{Gavriliuk}},
1152: \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Kharlamova}},
1153: \bibinfo{author}{\bibfnamefont{I.~S.} \bibnamefont{Lyubutin}},
1154: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Troyan}},
1155: \bibinfo{author}{\bibfnamefont{E.~S.} \bibnamefont{Ovchinnikov}},
1156: \bibinfo{author}{\bibfnamefont{A.~M.} \bibnamefont{Potsjelujko}},
1157: \bibinfo{author}{\bibfnamefont{M.~I.} \bibnamefont{Eremets}},
1158: \bibnamefont{and}
1159: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{B\"{o}hler}},
1160: \bibinfo{journal}{JETP Lett.} \textbf{\bibinfo{volume}{80}},
1161: \bibinfo{pages}{426} (\bibinfo{year}{2004}).
1162:
1163: \bibitem[{\citenamefont{{de~Andres} et~al.}(1997)\citenamefont{{de~Andres},
1164: Agullo-Rueda, Taboada, Cascales, Campa, Ruiz-Valero, and Rasines}}]{and97}
1165: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{{de~Andres}}},
1166: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Agullo-Rueda}},
1167: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Taboada}},
1168: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Cascales}},
1169: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Campa}},
1170: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Ruiz-Valero}},
1171: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Rasines}},
1172: \bibinfo{journal}{J. Alloys and Compounds} \textbf{\bibinfo{volume}{250}},
1173: \bibinfo{pages}{396} (\bibinfo{year}{1997}).
1174:
1175: \bibitem[{\citenamefont{Bezmaternykh et~al.}(2005)\citenamefont{Bezmaternykh,
1176: Temerov, Gudim, , and Stolbovaya}}]{bezm1}
1177: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1178: \bibinfo{author}{\bibfnamefont{V.~L.} \bibnamefont{Temerov}},
1179: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Gudim}}, ,
1180: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{N.~A.}
1181: \bibnamefont{Stolbovaya}}, \bibinfo{journal}{Crystallography Reports}
1182: \textbf{\bibinfo{volume}{50, Suppl. 1}}, \bibinfo{pages}{S97}
1183: (\bibinfo{year}{2005}).
1184:
1185: \bibitem[{\citenamefont{Bezmaternykh et~al.}(2004)\citenamefont{Bezmaternykh,
1186: Kharlamova, and Temerov}}]{bezm2}
1187: \bibinfo{author}{\bibfnamefont{L.~N.} \bibnamefont{Bezmaternykh}},
1188: \bibinfo{author}{\bibfnamefont{S.~A.} \bibnamefont{Kharlamova}},
1189: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{V.~L.}
1190: \bibnamefont{Temerov}}, \bibinfo{journal}{Crystallography Reports}
1191: \textbf{\bibinfo{volume}{49}}, \bibinfo{pages}{855} (\bibinfo{year}{2004}).
1192:
1193: \bibitem[{\citenamefont{Rousseau et~al.}(1981)\citenamefont{Rousseau, Bauman,
1194: and Porto}}]{rou81}
1195: \bibinfo{author}{\bibfnamefont{D.~L.} \bibnamefont{Rousseau}},
1196: \bibinfo{author}{\bibfnamefont{R.~P.} \bibnamefont{Bauman}},
1197: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.~P.~S.}
1198: \bibnamefont{Porto}}, \bibinfo{journal}{J. of Raman Spectroscopy}
1199: \textbf{\bibinfo{volume}{10}}, \bibinfo{pages}{253} (\bibinfo{year}{1981}).
1200:
1201: \bibitem[{\citenamefont{Mavrin}(1980)}]{mav80}
1202: \bibinfo{author}{\bibfnamefont{B.~N.} \bibnamefont{Mavrin}},
1203: \bibinfo{journal}{Optika i Spektroskopiya} \textbf{\bibinfo{volume}{49}},
1204: \bibinfo{pages}{79} (\bibinfo{year}{1980}).
1205:
1206: \bibitem[{\citenamefont{Herzberg}(1945)}]{her45}
1207: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Herzberg}},
1208: \emph{\bibinfo{title}{Infrared and Raman spectra of polyatomic molecules}}
1209: (\bibinfo{publisher}{Van Nostrand Reinhold}, \bibinfo{address}{New York},
1210: \bibinfo{year}{1945}).
1211:
1212: \bibitem[{\citenamefont{Landau and Lifshitz}(1980)}]{lan80}
1213: \bibinfo{author}{\bibfnamefont{L.~D.} \bibnamefont{Landau}} \bibnamefont{and}
1214: \bibinfo{author}{\bibfnamefont{E.~M.} \bibnamefont{Lifshitz}},
1215: \emph{\bibinfo{title}{Statistical Physics}} (\bibinfo{publisher}{Pergamon
1216: Press}, \bibinfo{address}{Oxford}, \bibinfo{year}{1980}).
1217:
1218: \bibitem[{\citenamefont{Bruce and Cowley}(1981)}]{bru81}
1219: \bibinfo{author}{\bibfnamefont{A.~D.} \bibnamefont{Bruce}} \bibnamefont{and}
1220: \bibinfo{author}{\bibfnamefont{R.~A.} \bibnamefont{Cowley}},
1221: \emph{\bibinfo{title}{Structural Phase Transitions}}
1222: (\bibinfo{publisher}{Taylor \& Francis}, \bibinfo{address}{London},
1223: \bibinfo{year}{1981}).
1224:
1225: \bibitem[{\citenamefont{Fleury}(1969)}]{fle69}
1226: \bibinfo{author}{\bibfnamefont{P.~A.} \bibnamefont{Fleury}},
1227: \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{180}},
1228: \bibinfo{pages}{591} (\bibinfo{year}{1969}).
1229:
1230: \bibitem[{\citenamefont{Cottam and Lockwood}(1986)}]{cot86}
1231: \bibinfo{author}{\bibfnamefont{M.~G.} \bibnamefont{Cottam}} \bibnamefont{and}
1232: \bibinfo{author}{\bibfnamefont{D.~J.} \bibnamefont{Lockwood}},
1233: \emph{\bibinfo{title}{Light Scattering in Magnetic Solids}}
1234: (\bibinfo{publisher}{JohnWilley \& Sons}, \bibinfo{address}{New York},
1235: \bibinfo{year}{1986}).
1236:
1237: \bibitem[{\citenamefont{Richards and Brya}(1974)}]{rich74}
1238: \bibinfo{author}{\bibfnamefont{P.~M.} \bibnamefont{Richards}} \bibnamefont{and}
1239: \bibinfo{author}{\bibfnamefont{W.~J.} \bibnamefont{Brya}},
1240: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{9}},
1241: \bibinfo{pages}{3044} (\bibinfo{year}{1974}).
1242:
1243: \bibitem[{\citenamefont{Halley}(1978)}]{hall78}
1244: \bibinfo{author}{\bibfnamefont{J.~W.} \bibnamefont{Halley}},
1245: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{41}},
1246: \bibinfo{pages}{1605} (\bibinfo{year}{1978}).
1247:
1248: \end{thebibliography}
1249:
1250:
1251:
1252:
1253:
1254:
1255: \begin{table*}%[H] add [H] placement to break table across pages
1256: \caption{\label{tab1} Frequency range and assignment of the vibrations of the \booo-group.
1257: The second column lists the assignments for the free ions.
1258: The third column shows the expected internal modes and their symmetries in the high temperature
1259: phase of the rare-earth iron borates (the superscripts 1 and 2 refer to the different borate groups
1260: in the rare-earth iron borates).}
1261: \begin{ruledtabular}
1262: \begin{tabular}{lll}
1263: Frequency range (\cm)& \booo\ vibrations (D$_{3\mathrm{h}}$) & Crystal vibrations\\\hline
1264: 600 & $\nu_4(E')$: in-plane bending & $\nu_4^{(2)}(A_1)+\nu_4^{(1)}(E)+2\nu_4^{(2)}(E)$ \\
1265: 700-800 & $\nu_2(A''_2)$ & $\nu_2^{(2)}(E)$ \\
1266: 950 & $\nu_1(A'_1)$: symmetric breathing & $\nu_1^{(1)}(A_1)+\nu_1^{(2)}(A_1)+\nu_1^{(2)}(E)$ \\
1267: 1250-1400 & $\nu_3(E')$: asymmetric breathing & $\nu_3^{(2)}(A_1)+\nu_3^{(1)}(E)+\nu_3^{(2)}(E)$
1268: \end{tabular}
1269: \end{ruledtabular}
1270: \end{table*}
1271:
1272:
1273: \begin{table*}%[H] add [H] placement to break table across pages
1274: \caption{
1275: The origin of vibrations for two structural modifications of \rfb\ and the new modes that
1276: should appear below the temperature of the structural phase transition $T_s$.
1277: \label{tab2}
1278: }
1279: \begin{ruledtabular}
1280: \begin{tabular}{llll}
1281: species & $T>T_s$: $R$32 ($D^7_3$) & $T<T_s$: $P$3$_1$21 ($D^4_3$) & New modes \\\hline
1282: R & $A_2 \oplus E$ & $A_1 \oplus 2A_2 \oplus 3E$& $A_1\oplus 2E\oplus A_2$ \\
1283: Fe & $A_1 \oplus 2A_2 \oplus 3E$ & $4A_1 \oplus 5A_2 \oplus 9E$ & $3A_1 \oplus 6E\oplus 3A_2$\\
1284: external \booo & & & \\
1285: \hspace*{5mm} translational & $A_1 \oplus 3A_2 \oplus 4E$
1286: & $5A_1 \oplus 7A_2 \oplus 12E$ & $4A_1 \oplus 8E\oplus 4A_2$\\
1287: \hspace*{5mm} librational & $A_1 \oplus 3A_2 \oplus 4E$
1288: & $5A_1 \oplus 7A_2 \oplus 12E$ & $4A_1 \oplus 8E\oplus 4A_2$\\
1289: internal \booo & $4A_1 \oplus 4A_2 \oplus 8E$ & $12A_1 \oplus 12A_2 \oplus 24E$
1290: & $8A_1 \oplus 16E\oplus 8A_2$\\\hline
1291: Total & $7A_1 \oplus 13A_2 \oplus 20E$ & $27A_1 \oplus 33A_2 \oplus 60E$ & $20A_1 \oplus 40E\oplus 20A_2$
1292: \end{tabular}
1293: \end{ruledtabular}
1294: \end{table*}
1295:
1296: \begin{table*}%[H] add [H] placement to break table across pages
1297: \caption{
1298: Observed vibrational modes for the high-temperature phase of several rare-earth iron borates.
1299: \label{tab3}
1300: }
1301: \begin{ruledtabular}
1302: \begin{tabular}{llllllllllll}
1303: \multicolumn{12}{l}{External modes ($3A_1 \oplus 11E$)}\\
1304: \hline
1305: \multicolumn{3}{l}{\nfb}&\multicolumn{3}{l}{\gfb}&\multicolumn{2}{l}{\tfb}&\multicolumn{2}{l}{\efb}&\multicolumn{2}{l}{\yfb}\\
1306: $A_1$& $E_{TO}$ & $E_{LO}$ & $A_1$ & $E_{TO}$ & $E_{LO}$ & $A_1$ & $E$ & $A_1$ & $E$ & $A_1$ & $E$\\
1307: 180& 89& 93&180& 84& 93&180& 93&181&84&180&107\\
1308: 298& &159&307&160&160&309&159&312&160&312&158\\
1309: 473& &193&475&195&198&476&198&472&199&468&200\\
1310: & &232& & &232& &230& &225& & \\
1311: &260&266& &270&270& &270& &272& & 268\\
1312: &272&281& &273&287& &287& & & & \\
1313: &312&332& &315&330& &330& &325& & 327\\
1314: &354&356& &352&357& &355& &350&&354\\
1315: & &384& &391&391& &392& &395&&390\\
1316: & &439& &443&443& &444& &442&&441\\
1317: &475&488& & &488& &480& &475&&470\\
1318: \hline\hline
1319: \multicolumn{12}{l}{Internal modes ($4A_1 \oplus 8E$)}\\
1320: \hline
1321: \multicolumn{3}{l}{\nfb}&\multicolumn{3}{l}{\gfb}&\multicolumn{2}{l}{\tfb}&\multicolumn{2}{l}{\efb}&\multicolumn{2}{l}{\yfb}\\
1322: $A_1$& \multicolumn{2}{l}{$E$} & $A_1$ & $E_{TO}$ & $E_{LO}$ & $A_1$ & $E$ & $A_1$ & $E$ & $A_1$ & $E$\\
1323: 636&\multicolumn{2}{l}{ 579}& 638& & 580& 635& 579& 632& 579&631&576\\
1324: 950&\multicolumn{2}{l}{ 625}& 957& 631& 633& 957& 632& 960& 632&&\\
1325: 990&\multicolumn{2}{l}{ 669}& 990& 670& 676& 988& 676& 988& 675&984&672\\
1326: 1220&\multicolumn{2}{l}{ 734}&1230& & 735&1220& 734&1230& 730&&732\\
1327: &\multicolumn{2}{l}{ 968}& & & 968& & 966& & 960&&959\\
1328: &\multicolumn{2}{l}{1195}& & & & & & & &&1219\\
1329: &\multicolumn{2}{l}{1218}& & &1212& &1214& &1220&&1230\\
1330: &\multicolumn{2}{l}{1244}& & &1229& &1230& &1233&&1408\\
1331: &\multicolumn{2}{l}{1260}& &1198&1250& &1254& & &&\\
1332: &\multicolumn{2}{l}{1413}& &1280&1414& &1411& &1342&&\\
1333:
1334: \hline\hline \multicolumn{12}{l}{Modes appearing in the low T phase: $20A_{1}$ (12 ext. and 8 int.) $\oplus$ $40E$ (24 ext. an 16 int.)}\\
1335: \hline
1336: \multicolumn{3}{l}{\nfb}&\multicolumn{3}{l}{\gfb}&\multicolumn{2}{l}{\tfb}&\multicolumn{2}{l}{\efb}&\multicolumn{2}{l}{\yfb}\\
1337: \multicolumn{3}{c}{-} & $A_1$ & \multicolumn{2}{l}{$E$} & $A_1$ & $E$ & $A_1$ & $E$ & $A_1$ & $E$\\
1338: \multicolumn{3}{c}{-} & 53 &\multicolumn{2}{l}{101}& 54 & &72 & &76 &\\
1339: \multicolumn{3}{c}{-} &144 &\multicolumn{2}{l}{114}& & &149 & &150 &111 \\
1340: \multicolumn{3}{c}{-} &203 &\multicolumn{2}{l}{167}& & 169 &202 & 170&206 &173 \\
1341: \multicolumn{3}{c}{-} &233 &\multicolumn{2}{l}{206}&232 & 207 &234 & 208& &211 \\
1342: \multicolumn{3}{c}{-} &244 &\multicolumn{2}{l}{254}&247 & 256 & & 250&246 &253 \\
1343: \multicolumn{3}{c}{-} &263 &\multicolumn{2}{l}{276}&265 & 278 & & 278& &280 \\
1344: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{281}& & 282 & & 284& &287 \\
1345: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{305}& & 306 & & 305& &307 \\
1346: \multicolumn{3}{c}{-} &368 &\multicolumn{2}{l}{310}&368 & 311 & & & & \\
1347: \multicolumn{3}{c}{-} &385 &\multicolumn{2}{l}{337}& & 337 & & & &335 \\
1348: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{374}& & 375 & & 373& &370 \\
1349: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{ }& & & & & &378 \\
1350: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{395}& & 398 & & & &\\
1351: \multicolumn{3}{c}{-} &677 &\multicolumn{2}{l}{472}&679 & 473 & & 470& &\\
1352: \multicolumn{3}{c}{-} &724 &\multicolumn{2}{l}{596}&723 & 597 & & & &\\
1353: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{654}&953 & 654 & & & &\\
1354: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{669}& & 670 & & & &\\
1355: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{955}& & 954 & & & &\\
1356: \multicolumn{3}{c}{-} & &\multicolumn{2}{l}{968}& & 966 & & & &\\
1357: \end{tabular}
1358: \end{ruledtabular}
1359: \end{table*}
1360:
1361: \end{document}
1362:
1363: % ****** End of file apssamp.tex ******
1364: