Phonon Raman scattering of RCrO3 perovskites

(R = Y, La, Pr, Sm, Gd, Dy, Ho , Yb, Lu)

M. Weber1, J. Kreisel1,2,*, P.A. Thomas1, M. Newton1, K. Sardar3, R.I Walton3

1
Department of Physics, University of Warwick, Coventry CV4 7AL, United Kingdom
2
Laboratoire Matériaux et Génie Physique, Minatec, CNRS, Grenoble Institute of Technology, 3,
parvis Louis Néel, 38016 Grenoble, France
3
Department of Chemistry, University of Warwick, Coventry CV4 7AL, United Kingdom

Abstract
We report a systematic investigation of orthorhombic perovskite-type RCrO3 powder samples by Raman
scattering for nine different rare earth R3+ cations (R = Y, La, Pr, Sm, Gd, Dy, Ho, Yb, Lu). The room-
temperature Raman spectra and the associated phonon mode assignment provide reference data for
structural investigation of the whole series of RCrO3 orthochromites and phonon ab-initio calculations.
The assignment of the chromite spectra and comparison with Raman data on other orthorhombic
perovskites allows correlating the phonon modes with the structural distortions in the RCrO3 series. In
particular, two Ag modes are identified as octahedra rotation soft modes as their positions scale linearly
with the octahedra tilt angle of the CrO6 octahedra.

*
to whom correspondence should be addressed (E-mail: Jens.Kreisel@grenoble-inp.fr)
 I. INTRODUCTION

The understanding of functional ABO3 perovskite-type oxides is a very active research area with
relevance to both fundamental- and application-related issues. A particular aspect of perovskites is its
capacity to adopt a multitude of different structural distortions due to the possible incorporation of almost
every element of the periodic table into its structure [1]. Such distortions can be driven by external
parameters like temperature, pressure or chemical composition, which leads to an extraordinary richness
of physical properties within the family of perovskites. Multiferroic perovskites, which possess
simultaneously several so-called ferroic orders such as ferromagnetism, ferroelectricity and/or
ferroelasticity, currently attract a specific interest [2-4]. Structural distortions in perovskites can be
described by separating three main features with respect to its ideal cubic structure [1, 5-6]: (i) a rotation
(tilt) of essentially rigid BO6 octahedra, (ii) polar cation displacements which often lead to ferroelectricity
(iii) distortions of the octahedra such as the Jahn-Teller distortion. Past investigations of these instabilities
have been a rich source for the understanding of structural properties not only in perovskites but also in
oxides in general. The most common distortion in perovskites is the tilt of octahedra. Although many
different octahedral tilt systems exist [1, 5-7], the R-3c structure with tilt system a-a-a- (Glazer’s notation
[6]) and the Pnma with a-b+a- are by far the most common structures found in perovskites.
We are particularly interested in perovskites with the orthorhombic Pnma structure which is for
instance adopted by RFeO3 orthoferrites, RMnO3 orthomanganites, RNiO3 orthonickelates or RScO3
orthoscandates (R = Rare earth). In all these perovskites the orthorhombic distortion (octahedral tilt angle)
can be continuously tuned by the ionic size rR of the R3+ rare earth. This rotation is equivalent to modify
the B-O-B bond angle and the B-O orbital overlap, which in turn leads to a rich magnetic phase diagram
for magnetic B-cations, accompanied by intriguing metal insulator phase transitions when B = Mn, Ni. In
this study we focus on the family of RCrO3 orthochromites, which have in the past attracted considerable
interest for their complex magnetic phases at low temperature [8-11] and, more recently, for their potential
magnetoelectric or multiferroic properties [11-15], similarly to manganites and nickelates.
The aim of our study is to provide a better understanding of the phonon spectra of RCrO3 and to
correlate individual phonon modes to the structural distortions of the structure. For this we have
investigated RCrO3 powder samples with a large number of different rare earth (Y, La, Pr, Sm, Gd, Dy,
Ho, Yb, Lu) by phonon Raman spectroscopy, which is known to be a versatile technique for the
investigation of both phase transitions [16-23] and more subtle structural distortions in perovskites [24-
29]. While the Raman signature of orthomanganites [24-27, 30-31] or orthoferrites [32-34] is well known
and understood, investigations of orthochromites remain limited to Raman studies for R = Y, La, Nd, Gd,
Ho, Er [22, 24, 35-38], out of which the single crystal work on LaCrO3 [22] and YCrO3 [24] is of
particular relevance for this work. However, the whole RCrO3 series has not yet been systematically
investigated.

2
 II. EXPERIMENTAL

Powder samples of RCrO3 (R = Y, La, Pr, Sm, Gd, Dy, Ho, Yb, Lu) chromites were synthesized
by a direct hydrothermal synthesis that gave polycrystalline powders of a greenish colour. More details
about the synthesis of samples can be found in ref. [10]. The high-quality of the samples has been attested
by previous X-ray diffraction measurements and the characterization of their magnetic properties [10].
Throughout this manuscript we will discuss the different structural details of orthochromites, Table 1
presents a resume of these characteristics for both the chromites investigated here and others discussed in
the text. A particular attention has been paid to the calculation of the octahedral tilt angles. This is not
straightforward for the Pnma structure as the angles can be calculated either from cell parameters or from
the atomic coordinates [1], both results are compared in Table 1. Typically, the use of cell parameters
introduces an error in the tilt angles by about 15 % as the octahedra are slightly distorted and we thus
rather rely on the tilt angles which we have calculated from the atomic coordinates according to the
formalism in ref. [39]. In order to illustrate the distortion throughout the series, Figure 1 presents the
evolution of the lattice parameters and the tolerance factor t, defined as t = (rR3+ +rO2-) /2 (rCr3+ +rO2-).

Figure 1
Variation of the pseudo-cubic cell parameters and the tolerance factor for RCrO3 as a function of the R3+VIII ionic
radius. The lattice parameters for NdCrO3 are taken from ref. [38], for all other chromites from ref. [10].

3
 Lattice parameters CrO6 octahedra tilt angles

From lattice parameters From atomic positions

rR3+ (Å) a (Å) b (Å) c (Å) V (Å3) t [010] (°) θ[101] (°) [010] (°) θ[101] (°)

LaCrO3 1.160 5.494 7.777 5.520 235.814 0.903 2.50 5.57 5.25 10.92
PrCrO3 1.126 5.473 7.717 5.496 232.104 0.891 5.20
NdCrO3 1.109 5.480 7.699 5.425 228.8 0.885 4.79 8.12 7.55 13.86
SmCrO3 1.079 5.524 7.637 5.366 226.328 0.874 6.45 13.73
GdCrO3 1.053 5.529 7.605 5.313 223.395 0.864 8.91 16.06 10.96 14.39
DyCrO3 1.027 5.526 7.547 5.262 219.428 0.855 9.61 17.77 12.09 17.19
YCrO3 1.019 5.520 7.536 5.255 218.606 0.852 9.56 17.85 12.20 16.98
HoCrO3 1.015 5.533 7.534 5.243 218.554 0.851 10.21 18.62
ErCrO3 1.004 5.503 7.504 5.214 215.303 0.847 10.70 18.66 12.23 17.32
YbCrO3 0.985 5.525 7.497 5.202 215.462 0.840 11.13 19.70
LuCrO3 0.977 5.524 7.494 5.186 214.667 0.837 11.89 20.16

Table 1:
Structural characteristics of RCrO3 samples: lattice parameters in the orthorhombic setting, R3+ ionic radii (rR3+
values given in an eight-fold environment from [40]) and octahedral tilt angles ([010], θ[101]). The tilt angles are
determined both from lattice parameters [1] and, where available, from atomic coordinates [39]. The lattice
parameters for NdCrO3 are taken from ref. [38], for ErCrO3 from [41] and for all other chromites from ref. [10]. The
atomic positions for the calculation of tilt angles are taken from refs. [41-45].

Raman spectra were recorded with a Renishaw inVia Reflex Raman Microscope with a low-
wavenumber spectral cutoff at about 120 cm-1. Experiments were conducted in micro-Raman mode at
room temperature by using 633 nm He-Ne laser as exciting wavelength. It is well-known that Raman
spectra recorded on transition-metal oxides such as orthomanganites [30, 46] or orthonickelates [29, 47]
often show a strong dependence on the exciting laser power leading to structural modifications, phase
transitions or even locally decomposed material. Contrary to the above samples, our RCrO3 powders are
not black where laser heating is naturally expected. Nevertheless, we have observed that experiments with
a laser power above 10 mW lead to a significantly modified spectral signature RCrO3 in terms of band
position, width and intensity due to local laser heating. As a consequence, our experiments have been
carried out using laser powers of less than 1 mW under the microscope, and we have carefully verified
that no structural transformations or overheating take place.

III. RESULTS AND DISCUSSION

The ideal cubic structure of ABO3 perovskites is rather simple, with corner-linked anion octahedra
BO6, the B cations at the centre of the octahedra and the A cations in the space between the octahedra. In
this cubic perovskite structure Raman scattering is forbidden by symmetry.
RCrO3 chromites crystallize in an orthorhombically distorted perovskite structure with space group
Pnma (alternative setting Pbnm). With respect to the ideal cubic Pm3̄m perovskite structure this

4
orthorhombic Pnma structure is obtained by an anti-phase tilt of the adjacent CrO6 octahedra (a-b+a- in
Glazer’s notation [6]). The tilting of the CrO6 octahedra necessarily induces a distortion of the RO12
polyhedra. In the Pnma structure, the A-cation is usually considered to be in eight-coordination and also
antiparallel A-cation displacements are permitted by symmetry. Cr3+ is a Jahn-Teller (JT) in-active cation,
leading to almost no dispersion in the Cr-O bonds, typically an order of magnitude lower than what is
observed in JT-active orthomanganites. All three distortions (tilt, octahedra distortion and A-cation
displacement) break the cubic symmetry and thus activate zone centre Raman modes. According to group
theory, the orthorhombic Pnma structure with four formula units per unit cell give rise to 24 Raman-active
modes [30]: 7 Ag + 5 B1g + 7 B2g + 5 B3g.
Raman spectra of orthochromites with R = La, Pr, Sm, Gd , Dy , Ho ,Yb, Lu are presented in
Figure 2 and show 11 to 14 modes depending on the rare earth. The other predicted modes are either of
too low intensity to be observed or below our experimental cut-off. Our powder Raman spectra of LaCrO3,
GdCrO3 and HoCrO3 are consistent with literature work on single crystals of the same materials [22, 24,
35-36], for the other five chromites there is to date no data in the literature, to the best of our knowledge.
Table 2 presents the position and the hereafter discussed assignment of the individual modes, together
with their notation used in the manuscript.

Figure 2:
Raman spectra at 300K of RCrO3 powders (R = La, Pr, Sm, Gd , Dy, Ho ,Yb, Lu) collected with a 632.8 nm He-Ne
laser line. I, II, III, and IV denote specific spectral regions discussed in the text. The red arrows illustrate for a
specific mode, how the phonons can be followed though the series of RCrO3.

Qualitatively, the overall spectral signature of the chromites is similar, in agreement with their
shared space group where the orthorhombic distortion varies continuously from the smallest rare earth Lu
(rLu,VIII = 0.98 Å) to the largest rare earth La (rLa,VIII = 1.16 Å) [10, 40]. According to Figure 1, LaCrO3 and
PrCrO3 occupy a special place in the orthochromite series in the sense that their orthorhombic distortion is
significantly smaller than that of other orthochomites with very close lattice parameters a, b and c.

5
Consistently, the spectra of LaCrO3 and PrCrO3 differ the most from the other spectra of the series,
namely in the low- and mid-wavenumber range, suggesting that the Raman signature can be used to
follow the orthorhombic distortion.

Table 2
Band positions and assignment of the observed Raman modes in RCrO3. While the symmetry assignment is
discussed in text, the activating distortions and main atomic motions are extended from refs. [22, 24, 26].

Activating
Symmetry LuCrO3 YbCrO3 HoCrO3 YCrO3 DyCrO3 GdCrO3 SmCrO3 PrCrO3 LaCrO3 Main atomic motions
distortion

Ag(1) 149.0
Ag (2) 137.6 138.5 141.2 184.5 141.0 142.2 140.8 143.5 175.2 rot[101] A(z) out of plane
Ag (3) 285.9 278.0 274.3 281.8 270.2 255.5 231.5 193.2 104.0 rot[010], JT BO6 in‐phase y rotations
Ag (4) 353.4 351.7 341.3 338.8 335.6 324.9 309.1 279.9 255.3 A shift O1(x), A(‐x)
Ag (5) 446.8 441.7 419.6 421.1 409.5 392.4 365.7 320.9 274.7 rot[101] BO6 out‐of‐phase x rotations
Ag (6) 508.2 501.9 493.3 489.1 486.1 477.0 465.3 449.7 436.6 rot[101] BO6 bendings
Ag (7) 558.3 561.2

B1g (1) 265.6 260.7 260.3 273.1 260.5 245.3 244.7 ‐ ‐

B1g (2) 411.8 408.8 401.6 405.4 397.2 388.7 382.4 365.4 351.4 A shift A(z), O1(‐z)

B2g (1) 161.3 161.4 162.1 216.8 160.4 159.3 155.2 149.1 151.7 rot[101] A(x)
B2g (2) 339.8 332.4 315.2 316.7 305.5 284.5 252.7 201.1 124.6 A shift A(z), O1(z)
B2g (3) 519.2 514.1 493.3 498.3 490.6 473.6 460.9 425.5 405.4 rot[101] BO6 out‐of‐phase bendings

B3g (1) 129.2 171.5 130.5 137.1 rot[101] A(y)

B3g (2) 497.4 504.7 486.9 482.3 482.2 473.6 454.7 443.8 424.5 rot[101] out‐of‐phase O2 "scissors"‐like
B3g (3) 565.1 563.8 562.9 540.5 560.0 568.7 567.2 577.7 586.8 rot[101] O2,O1 antistretching

A. Symmetry assignment of Raman modes in RCrO3

A symmetry assignment of the different Raman features is the sine qua non condition for a later
discussion of the correlation between structural distortions and the Raman signature. Generally speaking,
the use of powder samples with a random orientation of crystallites, as in our study, inhibits the
assignment of Raman modes by using polarized configurations. As a consequence, the hereafter presented
assignment is in a first step based on single crystal measurements that have been reported in the literature
for LaCrO3 [22] GdCrO3 [35], YCrO3 [24, 35] and HoCrO3 [36]. In a second step, we have taken
advantage of the continuous changes in the Raman spectra across the whole RCrO3 series, which has
allowed verifying the reported literature assignments, assigning the modes for the not yet reported samples
and investigating the crossing of several modes. For the sake of clarity and given the important number of
both materials and Raman bands, we will successively discuss four spectral regions as defined in Figure 2:
 Region I (below 200 cm-1) characterized by two sharp and intense bands.
 Region II (200 to 380 cm-1) characterized by four partly overlapping bands.
 Region III (380 to 500 cm-1) with two bands of similar intensity, except for GdCrO3 where only
one band is observed.
 Region IV (500 to 600 cm-1) with two features, each characterized by overlapping bands.

6
 1. Region I (100 to 200 cm-1)

In the approximation of an harmonic oscillator  = (k/µ)1/2 (k, force constant; µ, reduced mass) the
heaviest atom of the structure is expected to vibrate in the low-wavenumber region. Among the materials
shown in Figure 2 the mass increases only by about 25 % from La to Yb, so that the frequency associated
with A-cation vibrations is expected to increase only little with decreasing mass. Bands which are
dominated by vibrations of the A-cation are expected to be only little affected by changes in the
orthorhombic distortion.
The two sharp low-wavenumber bands at 137 and 161 cm-1 (values for LuCrO3) show only small
variation despite the large changes in distortion throughout the series. It is thus natural to assign these two
bands to A-cation vibrations.

FIG. 3:
Comparison of Raman spectra for of HoCrO3 (bottom, black), YCrO3 (middle, red) and DyCrO3 (top, black)
illustrating a significant shift of the low wavenumber modes for YCrO3. The assignment of Ag, B1g, B2g and B3g
modes is based on ref. [24] for YCrO3. The inset shows the mass-dependent evolution of the two low-wavenumber
bands.

The most convincing argument for the assignment of the two low-wavenumber bands to A-cation
vibrations comes from Figure 3 which compares the Raman spectrum of YCrO3 to that of DyCrO3 and
HoCrO3. The three rare earth have close ionic radii for Dy (rDy,VIII = 1.03 Å), Y (rY,VIII = 1.02 Å) and Ho
(rHo,VIII = 1.02 Å) leading to a very similar orthorhombic distortion, but the mass of the Y (mY = 89) is very
different to that of Dy (mDy = 162.5) and Ho (mHo = 165). It can be seen from Figure 3 that the sharp bands
shift for YCrO3 significantly by about 45 cm-1, while the other bands are only little or not affected. The
inset of Figure 3 shows that this shift corresponds for the band at 137 cm-1 well to the mass-induced shift
expected from a harmonic oscillator. This observation provides conclusive evidence that the two sharp
low-wavenumber modes are largely dominated by vibrations of the A-cation. Following the symmetry

7
assignment from single crystal work on YCrO3 [24], we assign the lower mode to Ag and the higher mode
to B2g symmetry. Beyond clarifying the band assignment it is interesting to note that Figure 3 illustrates
that materials like YCrO3 or HoCrO3, which present an almost identical distortion and thus a rather similar
XRD pattern, can be easily differentiated by Raman scattering.

2. Region II (200 to 380 cm-1)

Region II is characterized by two doublets, which can be assigned by extension of single crystal
literature data [24, 35-36] to bands of B1g symmetry at 265 cm-1, Ag at 285 cm-1, B2g at 339 cm-1 and Ag at
353 cm-1 (values for LuCrO3). Different spectral features should be commented: (i) Figure 3 shows that
the B1g-Ag doublet around 275 cm-1 is affected by changes in the mass of the A-cation, while the doublet
around 340 cm-1 is not. (ii) The B2g mode shows a very significant low-wavenumber shift with increasing
radii of the rare earth (decreasing distortion), making its attribution difficult for LaCrO3 and PrCrO3. (iii)
As an overall trend, we note that this spectral region is greatly affected by changes in the orthorhombic
distortion, which becomes namely obvious for the materials with large rR (thus reduced orthorhombicity),
as will be discussed later.

3. Region III (380 to 500 cm-1)

Region III is characterized by two bands of similar intensity, the evolution of their band position
across the RCrO3 series is presented in more detail in Figure 4. For the smallest R3+ in LuCrO3 the two
bands at 411 and 446 cm-1 are well separated and can be assigned to B1g and Ag symmetry, respectively.
Due to a pronounced low-wavenumber-shift of the Ag(5) band with increasing rR, the B1g(2) and Ag(5)
modes gradually approach until only a single and more intense band is observed for GdCrO3. Upon further
increase of rR the two modes gradually split again. This continuous evolution of both bands and the
observation of a single intense band for GdCrO3 strongly suggest a mode-crossing. As the two modes have
different B1g and Ag symmetry such a mode crossing is allowed by symmetry. The low-wavenumber-shift
of the Ag(5) band throughout the whole series of nine rare earth is remarkable and the most pronounced
shift observed in the spectra. For the case of LaCrO3 we note that the Ag(5) band has shifted to such an
extent that it closely approaches a lower lying Ag(4) mode of region II (see Figure 2). Such a closeness of
modes with the same symmetry can result in a mixing of the modes, as reported in the similar
orthomanganites [26], although we note that the intensity of the modes suggests that a transfer of the
character has not yet been occurred. The origin of the pronounced shift of the Ag(5) band will be discussed
later in more detail.

8
Figure 4:
Detailed view of the mid-wavenumber spectral region III. (a) Raman modes and (b) Evolution of the band positions
with rR for RCrO3.

4. Region IV (500 to 600 cm-1)

Region IV looks at first sight rather simple with only two broad spectral features, but a closer
inspection of Figure 2 shows that it is in fact a rather complex region with an important number of
overlapping and crossing bands. The presence of three overlapping features at 497, 508 and 519 cm-1 is
best observed for the most distorted LuCrO3 and can be assigned to B3g, Ag and B2g symmetry,
respectively. The different symmetries allow these three modes to cross or take the same frequency as for
instance for GdCrO3 where only a single intense band is observed around 475 cm-1. Due to the closeness
and intercrossing of the three bands, a band assignment of the individual shoulders is only possible on the
basis of polarized Raman data on single crystals (available in literature for some rare earths [22, 24, 35-
36]) and on the hypothesis that the Ag mode remains throughout the series the most intense among the
three modes. Based on this, we observe that the B2g mode shows the most pronounced low-wavenumber
shift and even crosses the two other modes from the high-wavenumber-side in LuCrO3 to the low-
wavenumber-side of the three-mode feature in LaCrO3.
The broadness of the massif at around 580 cm-1 suggests two or three underlying bands for this
feature, which can unfortunately not be distinguished by our powder Raman experiment, although we note
that their presence is suggested by single crystal data in the literature [22, 24, 35-36]. Modes in this region

9
involve mainly stretching vibrations of the CrO6 octahedra, which in turn explains their insensitivity to the
orthorhombic distortion, similar to observations for equivalent modes in manganites [26].

5. The particular case of LaCrO3, PrCrO3 and SmCrO3

Figure 2 and the discussion in the previous sections show that the Raman spectra of LaCrO3, PrCrO3
and SmCrO3 present in the low-wavenumber region a spectral signature that is significantly different from
the other RCrO3. This difference, which is in agreement with the reduced distortion of these materials
when compared to the other rare earth (see Figure 1 and Table 1), complicates the assignment of the
individual modes as the spectral changes are no more smooth. Figure 5 presents a zoom into the 100 to
450 cm-1 region for LaCrO3, PrCrO3 and SmCrO3, which illustrates that the assignment of the 300 to 400
cm-1 region is still straightforward. However, the region below 300 cm-1 is more complex with two Ag
modes and two B2g modes and where modes of the same symmetry are likely to interact and mix.

Figure 5:
Detailed view and assignment of the Raman modes in the low-wavenumber and mid-wavenumber region for
LaCrO3, PrCrO3 and SmCrO3. The band at 160 cm-1, marked with * is a spurious line, likely related to an impurity
phase. The assignment of LaCrO3, specifically the position and assignment of the 104 cm-1 Ag band marked by a red
line, is from ref. [22].

We first discuss the behavior of the B2g modes. It is tempting to relate the lower B2g modes to one type
of vibrational mode and the higher lying B2g to a second mode. However, Figure 5 shows that the lower
B2g mode in LaCrO3 is weak, which is in contrast to its intense and sharp profile throughout the whole
series of RCrO3 (including SmCrO3) and the higher B2g mode of LaCrO3. This observation suggests a
transfer of intensity which is the classic behavior for mixing of modes of the same symmetry and which

10
are close in wavenumber, resulting in transfer of their vibrational character. To support this scenario we
consider the mode assignments of lattice-dynamical calculations (LDC) reported for LaCrO3 [22] and
YCrO3 [24] which lead to the following scenario: For all RCrO3 (except LaCrO3 and PrCrO3) the higher
B2g(2) mode is activated by the A-cation shift [24] while the lower B2g(1) mode is activated by octahedral
rotations around [101]. For PrCrO3 we expect that the two modes mix and transfer energy although the
overlap of the modes inhibits a detailed analysis. Finally, for LaCrO3 the modes have clearly exchanged
their character as evidenced by the intensity transfer, but our data does not allow judging to what extend
the modes are still partly mixed. As a consequence, we assign the 124 cm-1 mode in LaCrO3 to B2g(2) and
the 154 cm-1 mode to B2g(1).
The discussion of the Ag modes on the basis of our own measurements is more difficult as their
difference in terms of intensity is less marked and namely because our spectral cut-off does not allow us
the observation of the low lying Ag mode of LaCrO3, which was reported to be located at 104 cm-1 [22].
Our discussion thus has to rely on the mode assignments of lattice-dynamical calculations (LDC) reported
for LaCrO3 [22] and YCrO3 [24]. According to [24], the higher Ag(3) mode of RCrO3 (except LaCrO3 and
PrCrO3) has the pattern of octahedra rotation, while the lower Ag(2) is dominated by A-cation vibrations.
From our data with overlapping modes it is difficult to judge if the two modes mix for PrCrO3. However,
according to the assignment presented in ref. [22] the two modes have clearly exchanged their character
for LaCrO3 so that we assign the 104 cm-1 mode in LaCrO3 to Ag(3) and the 175 cm-1 mode to Ag(2).
As a consequence, the low wavenumber region of LaCrO3, PrCrO3 and SmCrO3 is characterized by a
complex crossing in-between two Ag modes and in-between two B2g modes. Further understanding of the
mode mixing could be gained from investigations of solid solutions such as La1-xPrxCrO3 or La1-xSmxCrO3.

B. Phonon Raman modes vs. ionic radii & structural distortions

Figure 6 presents the evolution of the band position for the different chromites as a function of the
rare earth R3+ ionic radii. The proposed assignment in terms of symmetry and mode crossing is based on
the above discussion of the individual spectral regions. The figure illustrates the overall trend of
decreasing band positions with increasing rR which naturally correlates with the increase of most bond-
lengths. The few exceptions to this general trend have been discussed in the previous section.

11
Figure 6:
Raman phonon wavenumbers of RCrO3 as a function of the rare earth R3+ ionic radius. All lines are guides to the eye
only. The dashed lines for the low-wavenumber modes indicate schematically the region of mode mixing, discussed
in the text.

It can be seen that the rR-dependent shift in wavenumber varies significantly among the different
Raman modes and it is natural to expect that the modes with a significant shift involve O-Cr-O bending
(thus octahedra tilting), as this is the main structural distortion that changes throughout the series. Before
discussing this in more detail, we shall recall early work by Scott and co-workers [16-17, 19] which
demonstrated that Raman spectroscopy allows investigating the distortion of perovskites that is caused by
the rotation of octahedra. By studying temperature-dependent Raman spectra of SrTiO3 [16] and LaAlO3
[17] across their phase transition from a non-cubic structure with octahedra tilts, to the undistorted cubic
structure, they have observed that the phase transition is driven by soft modes which scale with the angle
of rotation of the octahedra. This Raman soft-mode picture has later shown to describe also the behavior
of the same model materials under high-pressure [20, 23, 48-49]. As this soft mode concept is of
importance for the interpretation of our data, we will shortly illustrate its application to the relatively
simple model system LaAlO3 (LAO). LAO crystallizes in a rhombohedral R-3c structure that is
characterized by an anti-phase a-a-a- tilt of the adjacent AlO6 octahedra about the [111]p pseudo-cubic
diagonal which gives rise to five Raman-active modes Raman = A1g + 4 Eg. The A1g and one Eg mode of the
rhombohedral structure are mainly related to the collective modes of the octahedral network. The vibration
of the A1g mode has specifically the pattern of the single rotation of adjacent AlO6 about the [111] axis [17,
27, 50-51]. Experimentally, the position of the A1g modes shows a linear scaling behavior with the AlO6

12
octahedral rotation angle with a slope of about 23.5 cm-1/degree [50-51]. LaNiO3, which adopts the same
structure but has a different chemistry presents interestingly a similar slope of 23 cm-1/degree [47, 52].
The observed linear relationship is in agreement with Landau theory which predicts for the simplest
second-order phase transition that the soft-mode frequency  should vary in the distorted phase (T < Tc)
with temperature as ² = A(Tc-T) ~ 2, where  is the octahedral rotational order parameter [53].
The situation is more complicated in perovskites with a Pnma structure, because several soft
modes take part in the lattice distortion. The structure can either be described in terms of three orthogonal
tilt angles tilt angles (leading to the a-b+a- Glazer’s notation) or by rotations ,  and  around the pseudo-
cubic [101]pc, [010]pc, and [111]pc axes, respectively [1]. Under the assumption that the orthogonal
rotations a-x and a-z are approximately equal in the Pnma structure, it has been shown by Megaw [54] that
the tilt angles  and  alone are sufficient for describing the octahedral tilting of the structure while the
angle  can be obtained via cos  = cos cos [1]. The necessary requirement for correlating the two tilt
angles  and  to the observed Raman signature is the knowledge of the distortion-dependent soft modes,
for which Iliev et al. [26] have shown that they are of Ag symmetry. This identification of this mode is not
trivial, given the important number of Raman-active modes in the Pnma structure. Also the Pnma
structure is among the most stable structures within perovskites, so that usually only a weak softening is
observed, complicating the identification of the vibrational modes which present the pattern of the
distortion.
In the past, the empirical comparison of several materials with the same type of structural
distortion has been used to overcome this problem and has been a rich source of understanding Raman
signatures in oxide materials. For the specific case of orthorhombic perovskites we mention the systematic
Raman work on orthorhombic RMnO3 orthomanganites [26, 30-31] and RScO3 orthoscandates [55-57], all
of which crystallize in the same Pnma symmetry as orthochromites. These studies have allowed
identifying two Ag modes with a octahedra-rotation vibrational pattern of which the frequency scale
linearly with the rotation angle of the BO6 octahedral tilting [26, 55]. Interestingly the observed slopes,
23.5 cm-1/degree for manganites [26] and 20 cm-1/degree for scandates [55] are close to the value observed
for rhombohedral perovskites (see above).
We now discuss the observed spectral changes in our orthochromites. The inspection of Figure 6
shows that the Ag(5) mode presents a particularly pronounced shift in position with increasing rA, it is thus
natural to assign this mode to one of the two Ag octahedral soft modes which are expected to scale with
one of the octahedra tilt angles. Figure 7 plots the Ag(5) phonon positions as a function of octahedra tilt
angles for those RCrO3 where available atomic coordinates allow a reliable calculation of the octahedra tilt
angles (see tables 1 and 2 for explicit values). It can be seen that the positions of the Ag(5) mode in RCrO3
shows the same tendency as the manganite and scandate modes, i.e. they scale linearly with the octahedra
tilt angles at a slope of  24.3 cm-1/degree, adding further support to the assignment of the Ag(5) mode.
The visual assignment of the second soft mode is less straightforward as no other Ag Raman mode shifts
for the small A-cations in a similarly strong way as Ag(5). However, the earlier work on manganites and

13
scandates suggests that both Ag soft modes lie on the same slope (see [26, 55] and Figure 7). As a
consequence of this, we have plotted the evolution of all Ag modes against the second tilt mode by
assuming the 24.3 cm-1/degree slope observed for Ag (5). This plot (not shown) illustrates that only the
evolution of the Ag (3) mode follows the behavior of the Ag (5), allowing assigning the Ag (3) mode as the
second soft mode, in agreement with earlier assignments of LaCrO3 [22] and YCrO3 [24] As a matter of
fact, this identification has also provided confidence to our assignment of the low-frequency region (see
section III.A.5).
On the basis of these findings, we conclude that the Ag (5) soft mode is an excellent signature for
determining the magnitude of the octahedral rotation in orthorhombic RCrO3 and thus for its deviation
from the cubic structure. The Ag (5) is preferred to the Ag (3) on the basis of its easy identification
throughout the whole series and because the Ag (3) mode is for large rR likely affected by mode repulsion
and intermixing with the Ag (2) mode, as discussed in section III.A.5. This observation in orthochromites,
thus in a further family to orthomanganite and orthoscandate perovskites, adds further support to the
earlier proposition [26, 55] that a general relationship between the angle of octahedral tilting and the
frequency of the rotational soft modes is viable in a vast number of perovskites and thus suggests a general
soft-mode type relation of the type ² = B(rR,c – rR) ~ 2.

Figure 7
Phonon wavenumbers as a function of octahedra tilt angles for Ag(3) [red squares] and Ag(5) [black
circles] modes of those RCrO3 for which the tilt angle can be calculated from available atomic positions.
Raman band positions for NdCrO3 [38] and ErCrO3 [37] are from literature.

14
 IV. CONCLUSION

We have presented a Raman scattering investigation of a series of nine RCrO3 (R = Y, La, Pr, Sm,
Gd , Dy, Ho ,Yb, Lu) powder samples. A symmetry assignment of the observed modes has been proposed
on the basis of earlier published single crystal data for some orthochromites and by taking advantage of
the continuous changes in the Raman spectra across the whole RCrO3 series. This careful assignment has
allowed verifying the reported literature assignments, assigning the modes for the not yet reported
samples, investigating the complex crossing of several modes and, in particular, relating several modes to
the structural distortion of the orthochromites. We have shown that the frequency of the Ag (5) correlates
in a soft mode fashion with the magnitude of the octahedral rotation in RCrO3 and can thus be used to
follow the distortion of the structure.
The room-temperature Raman spectra and the associated phonon mode assignment provide
reference data for structural investigation of the whole series of RCrO3 chromites, including the
investigation of strain (via phonon shifts) in thin films. Similarly to previous systematic work on
manganites [25] it will be interesting to investigate the A-cation dependent spin-phonon coupling in
orthochromites via Raman scattering through the magnetic phase transition. We also note that recent ab-
initio calculations of phonon properties have shown to be a great source of understanding for structural
and physical properties of perovskites-type oxides, our study provides benchmark data for such
calculations on orthochromites, similarly to recent work in the field [52, 58-59].

Acknowledgements

JK is grateful for a visiting fellowship from the Institute of Advanced Study (IAS) Warwick and
acknowledges financial support from the department of Physics of the University of Warwick during his
sabbatical stay at Warwick. Some of the equipment used at the University of Warwick was obtained
through the Science City Advanced Materials project “Creating and characterizing Next Generation
Advanced Materials”. O. Chaix, Grenoble (F), is acknowledged for a careful reading of the manuscript.

15
 References

[1] R. H. Mitchell, Perovskites: Modern and ancient (Almaz Press, Ontario (C), 2002).
[2] M. Fiebig, Journal of Physics D: Applied Physics 38, R123 (2005).
[3] R. Ramesh, and N. A. Spaldin, Nature Materials 6, 21 (2007).
[4] J. Kreisel, B. Noheda, and B. Dkhil, Phase Transitions 82, 633 (2009).
[5] A. M. Glazer, Acta Cristallographica A 31, 756 (1975).
[6] A. M. Glazer, Acta Crystallographica B 28, 3384 (1972).
[7] A. M. Glazer, Phase Transitions 84, 405 (2011).
[8] E. F. Bertaut, Marescha.J, G. Devries, R. Aleonard, R. Pauthene, J. Rebouill, and Zarubick.V,
IEEE Trans. Magn. MAG2, 453 (1966).
[9] K. Tsushima, K. Aoyagi, and S. Sugano, J. Appl. Phys. 41, 1238 (1970).
[10] K. Sardar, M. R. Lees, R. J. Kashtiban, J. Sloan, and R. I. Walton, Chem. Mat. 23, 48 (2011).
[11] J. R. Sahu, C. R. Serrao, N. Ray, U. V. Waghmare, and C. N. R. Rao, Journal of Materials
Chemistry 17, 42 (2007).
[12] Y. L. Su, J. C. Zhang, L. Li, Z. J. Feng, B. Z. Li, Y. Zhou, and S. X. Cao, Ferroelectrics 410, 102
(2010).
[13] A. Durán, A. M. Arévalo-López, E. Castillo-Martínez, M. García-Guaderrama, E. Moran, M. P.
Cruz, F. Fernández, and M. A. Alario-Franco, Journal of Solid State Chemistry 183, 1863 (2010).
[14] C. R. Serrao, A. K. Kundu, S. B. Krupanidhi, U. V. Waghmare, and C. N. R. Rao, Physical
Review B 72, 220101 (2005).
[15] Z. X. Cheng, X. L. Wang, S. X. Dou, H. Kimura, and K. Ozawa, J. Appl. Phys. 107, 09D905
(2010).
[16] P. A. Fleury, J. F. Scott, and J. M. Worlock, Physical Review Letters 21, 16 (1968).
[17] J. F. Scott, Physical Review 183, 823 (1969).
[18] G. Burns, and B. A. Scott, Physical Review B 7, 3088 (1973).
[19] J. F. Scott, Reviews of Modern Physics 46, 83 (1974).
[20] P. Bouvier, and J. Kreisel, Journal of Physics: Condensed Matter 14, 3981 (2002).
[21] J. Kreisel, P. Jadhav, O. Chaix-Pluchery, M. Varela, N. Dix, F. Sánchez, and J. Fontcuberta,
Journal of Physics: Condensed Matter 23, 342202 (2011).
[22] M. N. Iliev, A. P. Litvinchuk, V. G. Hadjiev, Y. Q. Wang, J. Cmaidalka, R. L. Meng, Y. Y. Sun,
N. Kolev, and M. V. Abrashev, Physical Review B 74, 214301 (2006).
[23] M. Guennou, P. Bouvier, J. Kreisel, and D. Machon, Phys. Rev. B 81, 054115 (2010).
[24] N. D. Todorov, M. V. Abrashev, V. G. Ivanov, G. G. Tsutsumanova, V. Marinova, Y. Q. Wang,
and M. N. Iliev, Physical Review B 83, 224303 (2011).
[25] J. Laverdière, S. Jandl, A. A. Mukhin, V. Y. Ivanov, V. G. Ivanov, and M. N. Iliev, Physical
Review B 73, 214301 (2006).
[26] M. N. Iliev, M. V. Abrashev, J. Laverdière, S. Jandl, M. M. Gospodinov, Y.-Q. Wang, and Y.-Y.
Sun, Physical Review B 73, 064302 (2006).
[27] M. V. Abrashev, A. P. Litvinchuk, M. N. Iliev, R. L. Meng, V. N. Popov, V. G. Ivanov, R. A.
Chakalov, and C. Thomsen, Physical Review B (Condensed Matter) 59, 4146 (1999).
[28] C. Girardot, J. Kreisel, S. Pignard, N. Caillault, and F. Weiss, Physical Review B 78, 104101
(2008).
[29] M. Zaghrioui, A. Bulou, P. Lacorre, and P. Laffez, Physical Review B 64, 081102 (2001).
[30] M. N. Iliev, M. V. Abrashev, H. G. Lee, V. N. Popov, Y. Y. Sun, C. Thomsen, R. L. Meng, and C.
W. Chu, Physical Review B (Condensed Matter) 57, 2872 (1998).
[31] M. N. Iliev, M. V. Abrashev, H. G. Lee, V. N. Popov, Y. Y. Sun, C. Thomsen, R. L. Meng, and C.
W. Chu, Journal of the Physics and Chemistry of Solids 59, 1982 (1998).
[32] R. M. White, R. J. Nemanich, and C. Herring, Physical Review B 25, 1982 (1982).
[33] S. Venugopalan, M. Dutta, A. K. Ramdas, and J. P. Remeika, Physical Review B 31, 1490 (1985).
[34] J. Kreisel, G. Lucazeau, and H. Vincent, Journal of Solid State Chemistry 137, 127 (1998).
[35] M. Udagawa, K. Kohn, N. Koshizuka, T. Tsushima, and K. Tsushima, Solid State
Communications 16, 779 (1975).
[36] W. Kaczmarek, and I. Morke, Journal of Magnetism and Magnetic Materials 58, 91 (1986).

16
[37] D. Ullrich, R. Courths, and C. Vongrundherr, Physica B & C 89, 205 (1977).
[38] Y. Du, Z. X. Cheng, X.-L. Wang, and S. X. Dou, J. Appl. Phys. 108, 093914 (2010).
[39] Y. S. Zhao, D. J. Weidner, J. B. Parise, and D. E. Cox, Phys. Earth Planet. Inter. 76, 17 (1993).
[40] R. D. Shannon, Acta Crystallographica A32, 751 (1976).
[41] J. Prado-Gonjal, R. Schmidt, D. Ávila, U. Amador, and E. Morán, J. European Ceram. Soc. 32,
611 (2012).
[42] N. Sakai, H. Fjellvag, and B. C. Hauback, Journal of Solid State Chemistry 121, 202 (1996).
[43] K. Ramesha, A. Llobet, T. Proffen, C. R. Serrao, and C. N. R. Rao, Journal of Physics-Condensed
Matter 19, 102202 (2007).
[44] Z. A. Zaitseva, and A. L. Litvin, Dopovidi Akademii Nauk Ukrainskoi Rsr Seriya B-Geologichni
Khimichni Ta Biologichni Nauki 1, 27 (1979).
[45] B. Vanlaar, and J. B. A. Elemans, Journal De Physique 32, 301 (1971).
[46] J. Kreisel, G. Lucazeau, C. Dubourdieu, M. Rosina, and F. Weiss, Journal of Physics: Condensed
Matter 14, 5201 (2002).
[47] N. Chaban, M. Weber, J. Kreisel, and S. Pignard, Appl. Phys. Lett. 97, 031915 (2010).
[48] A. Grzechnik, G. H. Wolf, and P. F. McMillan, Journal of Raman Spectroscopy 28, 885 (1997).
[49] M. Guennou, P. Bouvier, G. Garbarino, and J. Kreisel, Journal of Physics: Condensed Matter 23,
395401 (2011).
[50] S. A. Hayward et al., Physical Review B 72, 054110 (2005).
[51] A. Dubroka, J. Humlíček, M. V. Abrashev, Z. V. Popović, F. Sapiña, and A. Cantarero, Physical
Review B 73, 224401 (2006).
[52] G. Gou, I. Grinberg, A. M. Rappe, and J. M. Rondinelli, Physical Review B 84, 144101 (2011).
[53] E. K. H. Salje, Phase Transitions in Ferroelastic and Co-elastic Crystals (Cambridge University
Press, Cambridge, 1993).
[54] H. D. Megaw, Crystal structures: a working approach (Saunders Co, Philadelphia, 1973).
[55] O. Chaix-Pluchery, and J. Kreisel, Phase Transitions 48, 542 (2011).
[56] O. Chaix-Pluchery, D. Sauer, and J. Kreisel, Journal of Physics: Condensed Matter 22, 165901
(2010).
[57] O. Chaix-Pluchery, and J. Kreisel, Journal of Physics: Condensed Matter 21, 175901 (2009).
[58] P. Hermet, M. Veithen, and P. Ghosez, Journal of Physics-Condensed Matter 21, 215901 (2009).
[59] P. Hermet, M. Goffinet, J. Kreisel, and P. Ghosez, Physical Review B 75, 220102(R) (2007).

17