Electronic and crystal structure, EPR and magnetic investigations of YF3:1%RE (RE = Pr, Ho, Er and Tm) and LaF3:1%Pr nanocrystals

Article abstract of DOI:10.1016/j.jallcom.2014.07.074

A new chemical synthesis route by a thermal decomposition of nitrates and acetates, preceded by solution displacement reaction, was successfully applied to obtain the YF₃:1%RE (RE = Pr, Ho, Er and Tm) and LaF ₃:1%Pr nanocrystals. The samples were characterized by the following methods: X-ray diffraction, scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), magnetic susceptibility thermal dependence measurements and electron paramagnetic resonance (EPR). The obtained YF ₃:1%RE (RE = Pr, Ho, Er and Tm) materials crystallized with orthorhombic symmetry, whilst LaF₃:1%Pr exhibits hexagonal structure. Chemical composition determined by EDX, XPS and magnetic measurements was close to nominal formula. SEM images show the nanometric size of the grains. Surface contamination can be suggested to be responsible for the deterioration of the optical properties of the composites.

Full text of DOI:10.1016/j.jallcom.2014.07.074

Journal of Alloys and Compounds 616 (2014) 556–568
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jalcom
Electronic and crystal structure, EPR and magnetic investigations
of YF3:1%RE (RE = Pr, Ho, Er and Tm) and LaF3:1%Pr nanocrystals
a
a
a
b
b
E. Talik a, , P. Zajdel , A. Guzik , D. Skrzypek , L. Lipinska , M. Michalska
⇑
´
a Institute of Physics, University of Silesia, ul. Uniwersytecka 4, 40-007 Katowice, Poland
b Institute of Electronic Materials Technology, ul. Wólczyn´ska 133, 01-919 Warszawa, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 May 2014
Received in revised form 2 July 2014
Accepted 8 July 2014
Available online 16 July 2014
A new chemical synthesis route by a thermal decomposition of nitrates and acetates, preceded by solu-
tion displacement reaction, was successfully applied to obtain the YF3:1%RE (RE = Pr, Ho, Er and Tm) and
LaF3:1%Pr nanocrystals. The samples were characterized by the following methods: X-ray diffraction,
scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), magnetic susceptibility
thermal dependence measurements and electron paramagnetic resonance (EPR). The obtained YF3:1%RE
(RE = Pr, Ho, Er and Tm) materials crystallized with orthorhombic symmetry, whilst LaF3:1%Pr exhibits
hexagonal structure. Chemical composition determined by EDX, XPS and magnetic measurements was
close to nominal formula. SEM images show the nanometric size of the grains. Surface contamination
can be suggested to be responsible for the deterioration of the optical properties of the composites.
Ó 2014 Elsevier B.V. All rights reserved.
Keywords:
Nanomaterials
Yttrium compounds
Phosphors
1. Introduction
combined with all advantages of the polymer host. In contrast YF3:
Pr3+ composites exhibited decreased luminescence intensity and
Fluoride materials are useful as crystalline matrices for rare
earth ions which are popular activators in optoelectronics [1]. YF3
belongs to the group of fluoride crystals widely studied for years,
as an attractive host for active ions emitting in the UV and visible
spectral range. In general, the attractiveness of fluorides results
from their low phonon energies, good chemical stability and
excellent optical properties [2]. Recently, the set of active oxide
and fluoride nanopowders, obtained by a sol–gel method has been
characterized by means of laser spectroscopy [3]. In particular – the
emission and excitation spectra have been investigated together
with fluorescence dynamics profiles, which in turn enabled optimi-
zation of developed manufacturing technology. Improved polymer
composites, based on Polymethyl methacrylate (PMMA) doped
with LaAlO3:Pr3+ and YF3:Pr3+ nanocrystallites, have been obtained
and characterized in respect of their luminescent properties in the
visible spectral range. The results confirmed that, in general,
LaAlO3:Pr3+ nanocrystallites incorporated into a polymer matrix
maintained their original optical properties, both with respect of
emission characteristics and fluorescence kinetics behavior. This
suggests that the active ions were effectively shielded from the
interactions with the phonons of PMMA matrix, keeping the
luminescent and lasing potential of the crystalline active medium
significant shortening of the fluorescence lifetime. This may be
interpreted in terms of impurities in the polymer matrix (for exam-
ple the presence of OHꢀ groups). However, given the excellent
luminescent properties are maintained in the LaAlO3:Pr3+ compos-
ite materials this argument seems unlikely. In order to fully under-
stand these materials further studies are essential. What is
important, unlike in the case of bulk single crystals, the manufac-
turing of fluoride nanopowders is relatively easy and does not
require a fluorinating atmosphere [4].
The aim of this work is characterization of the YF3:1%RE
(RE = Pr, Ho, Er and Tm) and LaF3:1%Pr nanocrystals obtained by
new chemical synthesis – a thermal decomposition of nitrates
and acetates, preceded by solution displacement reaction for the
above mentioned applications.
2. Experimental
The nanopowders were prepared using new chemical ‘‘wet’’ method as previ-
ously reported by us [3]. This simple and cost effective chemical synthesis is a
two stage process. The first stage, carried out in solution, relies on a displacement
reaction. Owing to good diffusion, the homogenization of compounds proceeds on
molecular level. The next step is the thermal decomposition of nitrates and acetates
contained within a concentrated solution. Crystallization of the fluoride takes place
during the thermal treatment.
YF3:1%RE materials were produced as follows. Firstly, Y2O3 was dissolved in
0.4 mol/l acetic acid with rare earth oxides dissolved in diluted 0.5 mol/l nitric acid
and NH4F in deionized water. RE oxides were 4N purity. All other chemicals were
⇑
Corresponding author. Tel.: +48 32 359 1187.
E-mail address: talik@us.edu.pl (E. Talik).
http://dx.doi.org/10.1016/j.jallcom.2014.07.074
0925-8388/Ó 2014 Elsevier B.V. All rights reserved.

E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
557
analytical grade. Next, all solutions were combined together and mixed for 6 h at
70 °C using a magnetic stirrer with a stirring speed of 200 rpm resulting in slow
evaporation of the mixture. The volume was decreased by approximately 5 times.
The concentrated solution was subsequently dried at 150 °C for total removal of
water. The dry powder was then calcined in air at 400 °C for 2 h.
the ADPs for the doped LaF3 material resulted in negative values for
fluorine, so in the final model a single isotropic ADP was used for all
sites. Atomic parameters obtained in the refinement are collected
in Tables 1 and 2 and agreement factors in Table 3.
The microstructural observations of the nanopowders as well as the microcom-
positional analysis were conducted using the JEOL-7600F scanning electron
microscope.
The structures are in a good agreement with those published
earlier by Cheetham and Norman [12] for YF3 materials and by
De Rango et al. [15] for LaF3. Additional phases were only observed
in two samples: YF3:1%Pr has a single reflection located at d-spac-
ing 3.02 Å and LaF3:1%Pr exhibits two broad peaks with inter-pla-
nar spacings of d = 3.35 Å and d = 2.91 Å. In the second one, the
d = 3.35 Å impurity was overlapping with the 111 reflection, so
for the final refinement the angular range (25.8–29.0) was
excluded from fit. The single observed impurity line in YF3:1%Pr
Powder diffraction studies were performed on a PW1050 diffractometer in
Bragg–Brentano geometry. A nickel filtered Cu Ka1,2 source was used at 30 kV/
30 mA. The patterns were collected in step mode in the angular range 10–135° with
a step size of 0.02°. The powders were attached to a glass holder using double sided
Scotch tape. One of the samples (1% Ho) had bad adhesion to the tape, which
resulted in appearance of weak sample holder reflections at 14.1°, 17.0° and 25.6°.
The Rietveld method was applied to refine collected patterns using the Fullprof
[5] software. As the width of the diffraction peaks was larger than the machine
broadening, size and strain models appropriate for Laue classes mmm (size = 3,
strain = 18) for YF3 and 6/mmm (size = 12, strain = 19) for LaF3 were added to the
refinement. The anisotropic strain broadening was introduced in the quartic form
[6] with symmetry constraints and adaptation for the Thompson, Cox and
Hastings profile function [7,8]. The anisotropic Lorentzian size broadening was
modeled using the spherical harmonic approach developed earlier for preferred ori-
entation Jarvinen [9]. In all cases the ionic form factors of Y3+, La3+ and Fꢀ were
used in the refinement. Visualization of the structures was done using the VESTA
software [10].
is likely
a 200 reflection of trigonal PrF3 (P-3c1, a = 7.08 Å,
c = 7.24 Å) but there are no other peaks available to corroborate
this finding. Additionally, the finely powdered samples had bad
adhesion to the tape, which resulted in appearance of weak and
broad sample holder reflections at 14.1°, 17.0° and 25.6° (see Sup-
plementary Information).
The machine resolution was estimated from a reference pattern
of silicon. The profile broadening of diffraction peaks due to size
and strain was quantified using general models implemented in
the Fullprof suite. The refinement, which included only the size
broadening was found to be not sufficient for the YF3 series, so
strain broadening was introduced following the formalism pro-
posed by Stephens [7]. Multiple strain parameters SHKL quantifying
strains and fluctuations of metric parameters had to be included:
XPS spectra were obtained using a PHI 5700/660 Physical Electronics Photoelec-
tron Spectrometer with monochromatized Al K X-ray radiation (1486.6 eV). A
a
hemi-spherical mirror analyzer measured the energy of electrons with an energy
resolution of about 0.3 eV. The photoelectron emission from a surface area of
800
lm ꢁ 2000 lm was recorded. All measurements were performed in the condi-
tions of 10ꢀ8 Pa. In every case, a charge neutralizer was used because of the charge
effect which occurs for non-conducting samples. The binding energy was deter-
mined with reference to the C1s component set at 285 eV. Each peak of the
recorded spectrum is characteristic for a certain electron energy level of the mea-
sured elements. However, the measured binding energies are not absolutely con-
stant but depend on the chemical environments, where functional groups are
located due to modification of the valence electron distribution. These differences
in relation to pure elements in the electron binding energies are called chemical
shifts. The angle between the X-ray source and the sample surface was 45°.
For XPS investigations, it is important to determine the relative concentrations
of the various constituents. The Multipak Physical Electronics program enables
quantification of the XPS spectra utilizing peak area and sensitivity factor. The stan-
dard atomic concentration calculation provides a ratio of each component to the
sum of all elements taken into account in the data. Elemental identification, all ele-
ments except H and He are detectable with a detection limit of the order 0.1 at.% for
good quality of the spectra. Only those elements are considered for which the spe-
cific line is clearly visible in the spectrum. For these lines, the background individ-
ually selected in the region limited to the particular line is subtracted and after that
integration of the peak area is performed [11]. The Gaussian–Lorentzian functions
were used to fit the XPS core level spectra.
S400, S040, S004, S220, S202, S022 for the orthorhombic Pnma and S400
,
S004 and S112 for the hexagonal P63cm refinements. The obtained
results are summarized in Tables 4 and 5.
3.1.1. Discussion of strain and size contributions to line profile
The increased surface to bulk ratio in small crystallites, together
with the chemical pressure caused by different ionic radii, may
lead to the appearance of internal stress in the studied materials.
Therefore assessment of the size should be accompanied by the
analysis of the strain. The coherent domain sizes (CDS) obtained
from the Lorentzian component of the width, primarily varying
as 1/cosh, were found to lie between 60 and 85 nm with small
anisotropy of shape of less than 1 nm. During the refinement of
strain – connected primarily with the Gaussian part varying as
tanh – two different behaviors were observed. No strain (within
Magnetic susceptibility was measured using the Quantum Design MPMS-XL-
7AC SQUID magnetometer in the temperature range 2–400 K.
The EPR spectra were collected by the X-band spectrometer with a TE011 rect-
angular cavity and 100 kHz field modulation, equipped with an Oxford Instruments
ESR 910 helium flow cryostat. The microwave frequency was measured using a
Hewlett Packard 534 microwave frequency counter.
3r) was found in the hexagonal, Pr doped LaF3, despite a complete
deviation of the refinement towards Lorentzian contribution,
which had to be fixed at its maximum allowed value of 1. On the
other hand, all samples based on YF3 revealed a non-zero strain
contribution (Table 4), which was mostly confined to the h_l planes
as evidenced by significant values of S400, S004 and S202 parameters
3. Results and discussion
3.1. X-ray diffraction
Table 1
Atomic parameters of the doped YF3 compounds. All uncertainties are statistical on
The starting model for the refinement of the YF3:RE samples
was based on the structure obtained from neutron diffraction by
Cheetham and Norman [12], using the Pnma space group (SG 62),
which was earlier found by Zalkin and Templeton [13].
1r
level.
Parameter
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
Y position
x
4c (x, 1/4, z) .m.
0.36772(7)
0.05922(11)
0.66(2)
0.36760(8)
0.05904(11)
1.06(2)
0.36768(6)
0.05892(8)
0.41(1)
0.36763(8)
0.05909(10)
0.59(2)
For the LaF3, two different space groups in ambient conditions
have been reported: a centrosymmetric P-3c1 (SG 165) by Mans-
mann [14] or non-centrosymmetric P63cm (SG 185) by De Rango
et al. [15]. Both have successfully been applied to a neutron pow-
der diffraction [16] but a later single crystal neutron diffraction
study [17] found the latter to be more appropriate. The chosen
space group P63cm has a floating origin along the c crystallographic
axis and one of the atomic coordinates had to be fixed, so following
Gregson et al. [17], the z coordinate of La was set at z_La = 0.25.
In the case of the YF3 series, atoms were refined with individual
isotropic atomic displacement parameters (ADP). The refinement of
z
Biso (Å2)
F1 position
x
4c (x, 1/4, z)
0.5253(4)
0.5905(6)
0.94(6)
0.5225(4)
0.5914(6)
1.38(6)
0.5231(3)
0.5917(5)
0.70(5)
0.5232(3)
0.5924(6)
0.91(6)
z
Biso (Å2)
F2 position
x
8d (x, y, z)
0.1636(3)
0.0628(2)
0.3766(4)
0.93(5)
0.1646(3)
0.0623(2)
0.3758(4)
1.209(5)
0.1648(3)
0.0626(2)
0.3753(4)
0.61(4)
0.1643(3)
0.0633(2)
0.3756(4)
0.80(5)
y
z
Biso (Å2)

558
E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
Table 2
Atomic parameters of the doped LaF3. Single isotropic atomic displacement parameter was used for all atoms Biso = 0.05(2). Site symmetries for Wyckoff positions are: (2a) 3.m,
(4b) 3.., (6c) ..m.
Parameter
La
F1
F2
F3
F4
Wyckoff
x
z
6c (x, 0, z)
0.6710(4)
0.25(fix)
2a (0, 0, z)
0
0.320(4)
4b (1/3, 2/3, z)
1/3
0.212(3)
6c (x, 0, z)
0.277(2)
0.075(4)
6c (x, 0, z)
0.371(2)
0.421(3)
Table 3
Background corrected agreement factors of the refinement of all non-excluded points
and global v2. Names of the samples containing a second phase are marked in italics.
Sample
R-factors (%)
v2
F
Bragg
p
wp
exp
LaF3:1%Pr
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
4.32
5.01
1.39
1.30
1.28
0.997
15.6
5.86
7.14
5.51
6.14
15.2
8.11
8.33
6.73
7.81
12.23
5.99
7.66
5.29
6.27
1.53
1.83
1.18
1.62
1.55
0.803
0.883
0.734
0.681
(see Table 5). The largest value was found to be generated by the
uniaxial strain along c axis (S004 factor), being around 10 times lar-
ger than the corresponding values along a (S400) and another order
of magnitude higher than along b (S040). Additionally, it was
accompanied by high absolute but negative values of S202, which
indicated anti-correlations of variations along the a and c axes.
To understand this fact, one must recall that the orthorhombic
structure of YF3 is built of 9-fold coordinated Y atoms (Fig. 1),
which are stacked in such a way that yttrium ions form flat layers
perpendicular to the b axis (Fig. 2). In the Pnma space group only
the F2 atom is allowed to move along the c axis (Table 1) (it is 1
out of 7 positional parameters), therefore any stress will be primar-
ily relaxed by displacements in the a–c plane. This is in agreement
with the result that the distortions are confined in a–c layers and
the Y planes remain flat.
Fig. 1. The main building blocks of YF3 structure are edge connected polyhedra. The
grey vertices are occupied by F1 and pink by F2. Site symmetry is determined by
mirror plane (green) perpendicular to axis .b (m.). There are 8 Y–F distances close to
2.3 Å with one longer around 2.5 Å. The values given in the picture are calculated for
lattice parameters of YF3:1%Tm. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of this article.)
comes at a cost of the non-physical values of the atomic displace-
ment parameters (ADP) obtained in the refinement of LaF3, when
each site is treated separately. The physical and stable solution
was possible only with single spherical ADP for all atoms.
It must be noted that, in a general case, the profile broadening
model proposed by Stephens [7] allows for mixing 1/cosh and tanh
contributions to Lorentzian and Gaussian shapes by refining one
The planar alignment of the cations is also present in the LaF3
(Table 2) (Fig. 3) but contrary to the YF3 structure, there are 4 posi-
tional fluorine parameters, which can relax in the direction per-
pendicular to the layer. Therefore, the internal stress can be more
easily accommodated by atomic displacements. However, this
Table 4
Main profile parameters of the compounds in study. CDS is a coherent domain size estimated from profile broadening with the anisotropy given in curly braces. All reported
uncertainties are statistical on 1 level.
r
Sample
Lattice parameter (Å)
V (Å3)
CDS (nm)
Max. strain (%)
a
b
c
LaF3:1%Pr
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
7.1868(1)
6.3606(1)
6.3576(1)
6.3554(1)
6.3563(1)
–
7.3541(1)
4.4035(1)
4.4019(1)
4.4045(1)
4.4025(1)
328.95(1)
192.23(1)
191.98(1)
191.97(1)
191.97(1)
81{1}
63{1}
70{1}
85{1}
62{1}
3(1)
19(7)
15(6)
16(6)
13(6)
6.8633(1)
6.8602(1)
6.8580(1)
6.8600(1)
Table 5
Anisotropic strain and size parameters. S400, S040 and S004 measure strain along main crystallographic axes, whereas S220, S202, S022 and S122 quantify cross-correlations between
axial strains. Parameter Lor. is a measure of Lorenztian shape contribution to the strain.
Sample
Strain parameters
Lor.
S400
S040
S004
S220
S202
S022
LaF3:1%Pr
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
0.004(2)
0.89(4)
0.39(4)
0.57(2)
0.39(2)
–
0.0002(8)
8.4(3)
4.6(4)
6.7(2)
5.2(2)
–
S112 ->
0.04(1)
1.3(2)
0.53(10)
0.66(9)
0.31(8)
1(fix)
0.07(1)
0.033(6)
0.053(7)
0.007(3)
0.48(6)
0.14(3)
0.241(3)
0.09(2)
ꢀ2.9(2)
ꢀ1.5(2)
ꢀ2.4(2)
ꢀ2.0(2)
0.49(2)
0.69(4)
0.47(1)
0.47(2)

E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
559
additional parameter reported in Table 5 as Lor. Purely Lorenztian
shape should give Lor = 1 as it was used in the case of LaF3. In the
refinements of the YF3 series, its value was close to 0.5 with a 20%
larger result observed for Ho doped compound.
In other words, the strain dependence obtained from profile
broadening and ADPs indicated that effects of small domain size
and doping were relaxed differently in LaF3 and YF3. In LaF3 the dis-
persion of lattice parameters was smaller as evidenced by negligi-
ble strain parameters SHKL and a more stable lattice but at a cost of
less defined average coordinates and non-physical individual ADPs.
In YF3, the stress was accommodated mostly by the lattice (high
values of S004) with more stable atomic parameters within the
average cell (well defined individual ADPs). Here, one exception
should be raised in the case of holmium doped YF3, which has con-
sistently higher ADPs (Table 1) and Lor parameter (Table 5), which
would point toward much higher disorder than in the remaining
compounds. The YF3:1%Ho also stands out as its high ADPs are
counter-weighted by the smallest strain parameters SHKL in the
YF3 series, which puts it in the middle between yttrium and lantha-
num compounds.
Fig. 2. Anisotropic character of the YF3 confines the distortion and resulting strain
in the a–c plane. The distance between Y layers is b/2 = 3.4 Å.
It must be noted that in the case when the refinement of indi-
vidual isotropic ADPs was unsuccessful, an anisotropic approach
to ADP might have been more appropriate, as reported in previous
studies with neutrons [17]. However, refining from powder XRD
data ADPs of a large (1.3 Å) low-Z (only 8 electrons) ion like fluo-
rine Fꢀ is difficult due to the weak contribution of fluorine to the
total structure factor. Especially, when the compounds contain
high-Z ions like Y3+ or La3+
.
3.1.2. Discussion of structural details and Bond Valence Sum analysis
The lattice parameters as well as the unit cell volume of yttrium
based samples were found to decrease with the increasing atomic
number of rare-earth (RE). This is an understandable result of lan-
thanide contraction as the effective ionic radii of Y and RE elements
in the 3+ oxidation state with ninefold coordination are (Å):Y
1.042, Pr 1.179, Ho 1.072, Er 1.062, Tm 1.052, with 10-fold coordi-
nated La3+ 1.27 Å and fluorine Fꢀ 1.3 Å [18]. The changes in lattice
Fig. 3. Projection of the LaF3 structure along c axis (63) with different colors used to
visualize crystallographically nonequivalent fluorine atoms. (For interpretation of
the references to color in this figure legend, the reader is referred to the web version
of this article.)
Table 8
Multiplicity of La–F bond distributions up to 3 Å with bond valence (BV) calculated
using BVS for LaF3:1%Pr.
Distance (Å)
BV
Bond
Multi.
Bond
Multi.
Table 6
2.421
2.423
2.427
2.437
2.495
2.548
2.826
3.111
0.367
0.366
0.361
0.353
0.305
0.267
0.133
0.065
La1–F1
La1–F3
La1–F2
La1–F4
La1–F4
La1–F3
La1–F4
La1–F3
1
1
2
1
1
2
2
1
F1–La1
F3–La1
F2–La1
F4–La1
F4–La1
F3–La1
F4–La1
F3–La1
3
1
3
1
1
2
2
1
Structural stability analysis obtained by Bond Valence Sum formalism implemented
in BondStr program. Two sets of parameters for LaF3:Pr were calculated using 10-fold
(left) and 11-fold (right) coordination of La.
Sample
LaF3:1%Pr
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
Old GII (%)
GII(c) (%)
9.01|7.05
8.85|6.23
5.45
6.31
7.21
6.44
7.13
6.39
7.00
6.30
Table 7
Bond Valence Sum Parameters for LaF3:1%Pr. Coordination number, calculated and predicted distances, polyhedral distortionꢁ104 [20], nominal valence, Bond
Valence Sum, deviation from BVS, deviation from Equal Valence. The upper part is calculated using 10-fold coordination of La, whereas the bottom one contains
differences generated by with 11-fold coordination of La.
Atom
Coord.
davg (Å)
dpred (Å)
Distortion (*)
Nom. Val.
BVS
Dev. BVS (%)
Dev. EqV (%)
10-Fold coordination of La
La
F1
F2
F3
F4
10
3
3
3
4
2.538
2.420
2.427
2.506
2.646
2.502
2.459
2.459
2.459
2.574
35.7
0.0
0.0
5.6
46.9
3+
1ꢀ
1ꢀ
1ꢀ
1ꢀ
2.91
1.10
1.08
0.90
0.92
2.9
10.2
8.4
10.0
7.6
8.7
0.0
0.0
4.6
9.9
11-Fold coordination of La
La
F3
11
4
2.590
2.657
2.540
2.575
71.7
100.9
3+
1ꢀ
2.98
0.97
0.7
3.5
10.6
10.9

560
E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
Table 9
Bond Valence Sum Parameters for YF3 series. Coordination number, calculated and predicted distance, polyhedral distortionꢁ104 [20], nominal valence, Bond Valence Sum,
deviation from BVS, deviation from Equal Valence.
Atom
Coord.
davg (Å)
dpred (Å)
Distortion (*)
Nom. Val.
BVS
Dev. BVS (%)
Dev. EqV (%)
YF3:1%Pr
Y
F1
F2
9
3
3
2.323
2.372
2.299
2.310
2.310
2.310
11.5
26.8
0.03
3+
1ꢀ
1ꢀ
2.96
0.89
1.03
1.5
10.9
3.2
5.4
8.5
0.4
YF3:1%Ho
Y
F1
F2
9
3
3
2.323
2.372
2.299
2.310
2.310
2.310
11.1
25.7
0.02
3+
1ꢀ
1ꢀ
2.95
0.89
1.03
1.6
11.1
3.2
5.3
8.4
0.3
YF3:1%Er
Y
F1
F2
9
3
3
2.323
2.372
2.299
2.310
2.310
2.310
11.6
26.8
0.03
3+
1ꢀ
1ꢀ
2.96
0.89
1.03
1.5
11.0
3.2
5.4
8.5
0.4
YF3:1%Tm
Y
F1
F2
9
3
3
2.323
2.372
2.298
2.310
2.310
2.310
11.8
27.3
0.06
3+
1ꢀ
1ꢀ
2.96
0.89
1.03
1.4
10.9
3.4
5.5
8.6
0.5
parameters are followed by the behavior of the unit cell volume,
which is consistent with earlier results by Zalkin and Templeton
obtained for binary REF3 [13].
In order to estimate the stability of the structure, a Bond
Valence Sum (BVS) analysis of the refinement was done using the
BondStr program within the Fullprof suite. The Bond Valence
Parameters for La, Y and F were taken from Zachariesen [19] and
the polyhedral distortion index was calculated on the basis of Balic,
Zunic and Vickowic [20]. The results are presented in Tables 6–10.
Table 6 presents Global Instability Indexes (GII) for all refined
structures. Old GII measures the root mean square deviation of
BVS calculated for all non-equivalent sites, whereas GII(c) includes
site multiplicities. In all cases, they were found to lie below 10%
indicating stable structures. The worst agreement (largest instabil-
ity) was found in cases of YF3:1%Ho and LaF3:1%Pr, which is in
agreement with the high ADPs and the lowest strain broadening
observed in these materials.
The 9-fold coordination of yttrium in YF3 is an established fact
(Fig. 1) [12,13,21]. Eight out of 9 Y–F bonds are around 2.3 Å with
one being over 2.5 Å (see Table 10). The most distant Y–F1 bond is
over 10% longer than the expected Y–F distance of 2.3 Å [19]. It is
however necessary to obtain the balanced structure as its formal
charge transfer of 0.17 electrons is around 50% of the 0.34 charge
provided by each of the other bonds (see Table 10).
A less clear situation is encountered in the LaF3 structure, where
lanthanum is usually reported to be 10 or 11 coordinated [17]
(Fig. 4). The BVS analysis was therefore performed using both coor-
dinations, which differed by inclusion of a distant (3.11 Å) F3 site.
The addition of the 11th fluorine resulted in a slight improvement
of the GIIs (Table 6) but increased distortion index [20] of the La
polyhedron from 35 to 71. Additionally, the stability of the sur-
roundings of the F3 ion worsened by a factor of over 20 times
(Table 7).
Fig. 4. 11-Fold coordination of La atom is formed by
4
crystallographically
nonequivalent fluorine sites. Green surface marks ..m mirror plane. All distances
are given in Ångströms. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)
Comparison of the effective bond valences in the LaF3 structure
(Table 8) indicated that the most distant bond had minimal (0.07)
contribution to bond valence (BV), which was over 50% smaller
than the 9th and 10th La-F4 2.8 Å bonds (BV 0.13). The remaining
8 La-F connections had formal BVs between 0.27 and 0.36.
Table 10
0
Multiplicity of Y–F bond distributions up to 2.6 ÅA with bond valence (BV) calculated using BVS for YF3 series.
Bond type and multiplicity
YF3:1%Pr
Distance
YF3:1%Ho
Distance
YF3:1%Er
Distance
YF3:1%Tm
Distance
BV
BV
BV
BV
Y1–F1|F1–Y11|1
Y1–F1|F1–Y11|1
Y1–F2|F2–Y12|1
Y1–F2|F2–Y12|1
Y1–F2|F2–Y12|1
Y1–F1|F1–Y11|1
2.276
2.294
2.294
2.298
2.304
2.545
0.366
0.348
0.348
0.345
0.339
0.177
2.282
2.292
2.295
2.298
2.304
2.542
0.360
0.351
0.347
0.345
0.340
0.178
2.283
2.288
2.293
2.300
2.303
2.546
0.359
0.354
0.350
0.343
0.340
0.176
2.280
2.289
2.292
2.297
2.305
2.548
0.362
0.354
0.351
0.345
0.338
0.176

E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
561
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
LaF3:1%Pr
Fig. 5. SEM images of the nanopowder YF3:Er. The nano-sized grains forming agglomerates are visible.
Altogether, the 10-fold coordination of lanthanum was
sufficient to provide good charge stability of the structure
and did not require to accept significant distortions of local
polyhedra.
particles, as was found for microarea of LaF3:1%Pr. These grains
were enriched in oxygen and depleted of lanthanum (Fig. 6a).
The microanalysis of the chosen areas are shown in Fig. 6b and
their composition in Table 11. The differences between grain sizes
suggest that during short calcination times (2 h), the temperature
distribution in the sample was not homogeneous.
3.2. SEM/EDX
As reported previously for LaAlO3:Pr3+ [3], nanocrystallites
incorporated into a polymer matrix keep their original optical
properties, while YF3:Pr3+ composites showed decreased intensity
of luminescence and significant shortening of the fluorescence life-
time. This might be explained by some surface contribution. For
LaF3:1%Pr the ratio La/F is 2.9 while for YF3:1%Pr the ratio Y/F is
only 2.2. For the rest of the compounds Y/F is 2.5. The ratio C/O
is the highest for YF3:1%Pr and 1%Tm and is about 4.5; for the com-
pounds doped with Er, Ho and for LaF3:1%Pr it is about 2. The
SEM images show the morphologies of grains in the examined
compounds at magnifications ꢁ50,000 (Fig. 5). The crystalline par-
ticles YF3:1%RE and LaF3:1%Pr show the distribution of the grain
size. The obtained powders can be said to be included to nanocrys-
talline materials since a large number of the grains are below/close
to 100 nm. The observed bigger grains probably came as a result of
localized region subjected to higher calcination temperatures.
Moreover, in some areas this has lead to heterogeneity in the

562
E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
Fig. 6. (a) EDX element maps for LaF3:1%Pr. Contamination enriched with oxygen and depleted with lanthanum is visible. Local elemental analysis was performed in areas
shown in part (b). Results of analysis are given in Table 11.
Table 11
Elemental microanalysis of LaF3:1%Pr. Areas are shown in Fig. 6b.
3.3. XPS
Areas
C
O
F
La
The XPS spectra in the wide energy range 0–1400 eV of LaF3:
1%Pr and YF3:1%Ho are shown as examples in Fig. 7. Besides the
lines related to the yttrium or lanthanum and fluorine i.e., the
elements forming the compounds, the additional lines of carbon
and oxygen were observed in all investigated compounds. They
are connected with the contamination of the surface.
Whole area
Spectrum 2
Spectrum 3
Spectrum 4
10.12
10.90
11.43
12.07
8.55
60.41
54.00
59.89
59.54
19.43
10.21
20.43
20.41
24.25
6.81
6.64
All results in atomic %.
Fig. 8 shows the Y3d spin–orbit split doublets for YF3:1%RE. The
binding energies for each compound are collected in Table 12.
These Y3d lines were fitted with synthetic peaks. Besides the main
chemical composition of the surface can be significant for the cou-
pling with the polymer matrix.

E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
563
lines, additional lines at lower biding energy are also observed.
Their intensities are close to 10% of the intensity of the main dou-
blet line.
LaF3:1%Pr
F1s lines are shown in Fig. 9. The measured lines positions of the
main line are comparable with the data published by Wang et al.
Kirikova et al. and Chong et al. [22–24]. Moreover, the additional
low binding energy lines are visible.
d
3
r
P
The positions of core levels and the valence band are collected
in the Table 12. Some distribution of the binding energies can be
related to the complexity of the carbon line C1s. The most intensive
C1s line was assumed for calibration of the spectra.
d
4
r
P
The additional low intensity lines (A) at lower binding energies,
can be related to surface contribution. A similar effect was visible
for LaF3 spectra measured by Chong et al. [24]. The additional line
of La3d at 834.4 eV, present also in this work, was ascribed to
La2O3. Such oxide is not present in these XRD measurements. Some
extra phase observed for the samples doped with praseodymium is
related to the presence of PrF3 detected in the magnetic
measurements.
YF3:1%Ho
Another reason for the presence of the low intensity lines may be
considered to be due to a difference in the binding energy of the lon-
gest bonds between the fluorine and the rare earth ion (see Figs. 1
and 4). Such kind of bonds can be weaker giving a lower binding
energy with the intensities around 10%. The indication of such con-
tribution is asymmetry of the REF3 XPS lines, observed towards low
binding energy but mostly publish without deconvolution.
Fig. 10a and b shows valence bands consisting fluorine F2p and
Y4d or La5d states. The width of energy gap estimated from this
measurement is ꢂ10 eV, taking the Fermi level location about
the center of the energy gap. Similar values of the energy gap were
estimated in the literature for YF3 (10.3 eV) [25] and for LaF3
(9.7 eV) [26].
0
200
400
600
800
1000
1200
1400
Binding energy (eV)
Fig. 7. Overview example XPS spectra in the wide energy range 0–1400 eV of
YF3:1%Ho and LaF3:1%Pr. Additional lines of C and O are related to the surface
contamination.
The observed chemical shifts are positive and equal about +3 eV
for Y4p3/2, in relation to the elemental position 24 eV; +1 eV for F2s
(30 eV elemental position) and about +3 eV for La5p3/2 (17 eV for
elemental position) (Table 12) [27]. It indicates the covalent contri-
bution to the bonds in these compounds.
Fig. 11 presents the La3d spin–orbit doublet of LaF3:1%Pr. Each
of La3d line has a double peak structure corresponding to the
La3d94f0 (M) and the satellite lines S attributed to La3d94f1L (L is
a hole in valence band) states which are visible at higher binding
energy (Table 12). Moreover, additional lines (A), similarly to those
earlier described for the F1s and Y3d, are observed. A change of a
shape of the La3d line in fluoride in relation to the oxide material
was observed. For comparison similar line measured in SrLaAlO4
was added [28]. The change of the ratios of intensities IS/IM show
the same tendency as the theoretical calculations for oxide and
fluorine materials [29] which is shown in Fig. 11.
YF3:1%Pr
YF3:1%Ho
Y3d
Ho4d
3.4. Magnetic susceptibility
YF3:1%Er
In order to estimate the concentration of the RE3+ ions, mag-
netic measurements were done. The magnetic susceptibility tem-
perature dependence of YF3:1%RE and LaF3:1%Pr were measured
in a magnetic field of 0.5 T and over temperature range 1.9–
400 K (Fig. 12). The obtained curves were fitted to the Curie–Weiss
law in the temperature range 200–400 K. The range of the fit
depends on the sample due to crystal field effect.
Er4d
YF3:1%Tm
1
x
1
¼
ð1Þ
Tm4d
C
x0 þ Tꢀh
Using Eq. (1), the Curie constant C, the Curie–Weiss tempera-
140
150
160
170
180
190
ture h and the temperature-independent diamagnetic susceptibil-
ity vo were obtained. The Curie constant value was used for the
calculation of the effective magnetic moment of the nanocrystals.
Binding energy (eV)
Fig. 8. XPS core lines of Y3d.

564
E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
Table 12
Parameters of XPS lines of investigated compounds.
Compound band
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
LaF3:1%Pr
EB (eV)
9.9
EB (eV)
CHMFW (eV)
EB (eV)
CHMFW (eV)
EB (eV)
CHMFW (eV)
EB (eV)
CHMFW (eV)
CHMFW (eV)
Fꢀ 2p (VB)
Y3+ 4p3/2
Y3+ 4p1/2
Fꢀ 2s
9.5
27.2
28.2
31.4
160.4
159.0 (A)
12%
162.4
161.0 (A)
686.0
684.6 (A)
13%
2.9
2.9
2.9
2.9
2
9.4
27.0
28.1
31.1
160.3
158.4 (A)
16%
162.4
160.4 (A)
685.8
683.9 (A)
14%
3.3
3.3
3.3
3.3
2.4
9.4
26.7
27.7
31.0
160.5
158.4 (A)
7%
162.6
160.4 (A)
686.5
684.4 (A)
7%
2.5
2.5
2.5
2.5
2.3
9.0
26.6
27.7
30.8
159.6
157.2 (A)
9%
161.6
159.2 (A)
685.2
683.3 (A)
9%
2.9
2.9
2.9
2.9
2.3
2.7
30.9
3.5
Y3+ 3d5/2
Y3+ 3d3/2
Fꢀ1 1s
2
2.4
2.3
2.3
2.5
2.3
2.5
2.2
686.0
684.0 (A)
16%
2.3
2.3
La3+ 5p3/2
19.9
21.0 (S)
14%
17.6 (A)
15%
22.5
23.5 (S)
19.7 (A)
837.8 (M)
841.6 (S)
16%
La3+ 5p1/2
La3+ 3d5/2
2.3
3.3
834.4 (A)
16%
854.5 (M)
858.7 (S)
851.1 (A)
La3+ 3d3/2
3.3
Binding energy for elements: F2s – 30 eV; F1s – 685 eV; La5p – 17 eV; La3d5/2 – 835.8 eV; La3d3/2 – 853 eV; La M4N45N45 – 854 eV; Y4p – 24 eV; Y3d5/2 – 156 eV; Y3d3/2
158 eV [27].
–
LaF3:1%Pr
YF3:1%Pr
F1s
(a)
LaF3:1%Pr
La5p
F2s
F2p
A
S
YF3:1%Ho
YF3:1%Er
0
10
20
La5p
30
(b)
Y4p
F2s
F2p
LaF3:1%Pr
YF3:1%Pr
YF3:1%Ho
YF3:1%Er
YF3:1%Tm
YF3:1%Tm
0
10
20
30
Binding energy (eV)
670
675
680
685
690
695
700
Binding energy (eV)
Fig. 10. (a) XPS valence band of LaF3:1%Pr together with La5p and F2s states with
deconvolution and (b) comparison of valence bands all measured compounds and
deconvolution of the Tm doped sample as example.
Fig. 9. XPS core lines of F1s.

E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
565
YF3 and LaF3 elements was done. The total diamagnetic contribu-
tion was ꢀ3.8 ꢁ 10ꢀ5 emu/mole for YF3 and ꢀ4.8 ꢁ 10ꢀ5 emu/mole
for LaF3. These values are close to the fitted values.
La3d5/2
M
SrLaAlO4
LaF3:1%Pr
3.5. EPR
La3d3/2
M
Because of the high spin-lattice rate, the EPR spectra of the tri-
valent rare earth ions are observed at a low temperatures. In addi-
tion, no EPR signal from the non-Kramers ions: Ho, Pr and Tm was
observed [31]. However, the EPR measurements exhibit the pres-
ence of impurity phases in all studied systems.
3.5.1. EPR study of YF3 nanopowder doped with Er
S
A
The low temperature EPR spectrum of YF3:1%Er shown in
Fig. 13, reveals the presence of three broad signals in the range
10–500 mT.
A
S
It is known that in trivalent erbium ions with the configuration
4f11 (L = 6, S = 3/2) the spin–orbit coupling produces an array of
states, of which the lowest is 4I15/2. This is split by the cubic crystal
field into two doublets, C6 and C7, with an effective spin S = 1/2
and three C8 quartets with an effective spin S = 3/2. In crystal fields
with symmetries lower than cubic, the Er3+ ion ground term splits
into the eight Kramers doublets [31].
820 825 830 835 840 845 850 855 860 865 870
Binding energy (eV)
Fig. 11. XPS core lines of La3d5/2 and La3d3/2 of LaF3:1%Pr compared with La3d lines
The spin Hamiltonian characterizing the EPR spectra of the rare
earth Kramers ions (when the quadrupole coupling and nuclear
Zeeman interactions are negligible) can be expressed as [32]:
of SrLaAlO4. A decrease of S lines intensity for fluorine compounds is visible.
Using the calculated effective magnetic moment and assuming the
theoretical value of the effective magnetic moment for the free
RE3+ ion, the concentration of Ho, Er and Tm ions was found as
close to 1% (Table 13). A discrepancy is observed for the samples
doped with Pr ions. Additionally for these samples the magnetic
upturn is observed at about 54 K which suggests the presence of
the PrF3 phase ascribed to the Van Vleck paramagnetism [30]. This
additional phase is visible in the XRD pattern.
H ¼
l
BSgB þ SAI
ð2Þ
where
l
B is the Bohr magneton; S and I are the electron and nuclear
spins, respectively; B is the external Zeeman field; g and A are ten-
sors characterizing Zeeman and hyperfine interactions, respectively.
Erbium has one stable odd isotope, of mass number 167, with
22.94% abundance and I = 7/2. As a result, the EPR spectrum of
Er3+ should show a central signal corresponding to the fine-
structure transition M = +1/2 M M = ꢀ1/2 Kramers doublet
transition for the 166Er isotope with no nuclear magnetic moments
A comparison of the experimental vo values and the total of the
vo obtained as a sum of the Slater diamagnetic susceptibilities of
(a)
(d)
(b)
(c)
(e)
Fig. 12. Magnetic susceptibility and inverse of magnetic susceptibility vs temperature. Solid line presents the Curie–Weiss fit for the inverse of susceptibility. The magnetic
susceptibilities of LaF3:1%Pr and YF3:1%Pr are negative near room temperature. In order to draw the inverse susceptibilities, their values were corrected of +5 ꢁ 10ꢀ5 emu/
mol.

566
E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
Table 13
Magnetic properties of investigated compounds. Diamagnetic susceptibility, effective magnetic moment, Curie–Weiss temperature, effective magnetic moment calculated from
the Hund’s rules and computed concentration of the rare earth ions is given.
v
0*105 (emu/mol)
leff
(lB
)
H
(K)
ltheor
(lB
)
Concentration (%)
YF3:1%Er
YF3:1%Ho
YF3:1%Tm
YF3:1%Pr
LaF3:1%Pr
ꢀ3.78
ꢀ0.74
ꢀ1.72
ꢀ8.05
ꢀ6.50
0.92
1.07
0.73
0.33
0.32
3
9.58
10.61
7.56
3.58
3.58
0.93
1.02
0.93
0.85
0.80
ꢀ10
9
0
ꢀ3.4
YF3:Er
experimental spectrum
simulated spectrum
T=Troom
LaF3:Pr
YF3:Pr
YF3:Tm
YF3:Ho
YF3:Er
4000
2000
0
YF3:Er
experimental spectrum
simulated spectrum
T=4.2K
0,30
0,32
0,34
0,36
0,38
T=Troom
-2000
-4000
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Magnetic field induction (T)
S1
Fig. 13. X-band EPR spectrum of YF3:Er at T = 4.2 K (gain: 1 ꢁ 102).
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Magnetic field induction (T)
Fig. 14. X-band EPR spectra of LaF3:Pr, YF3:Pr, Tm, Ho, Er at a room temperature
(gain: 1 ꢁ 104).
(abundance 77.06%) and eight hyperfine transitions distributed
about the central signal. It can be seen (Fig. 14) that the hyperfine
structure is not resolved in the EPR spectrum of the YF3:1%Er
compound.
Taking into account the crystal structure of YF3 we assume that
trivalent erbium substitutes for Y3+ in YF3 preserving the mono-
clinic (.m.) local site symmetry of Y-ion. The EPR spectrum was
variations present: the crystalline field in the clusters [36] and
magnetic interactions.
3.5.2. Identification of Mn2+ impurity
fitted to the effective Hamiltonian H = lBSgB with S = 1/2 and
The EPR measurements exhibit the presence of Mn2+ ions in the
all investigated compounds. Figs. 14 and 15 show the X-band EPR
spectra collected at a room and liquid helium temperatures,
respectively. There can be seen the sextet centered at g ꢃ 2.0
(marked as S1) and the single line resonance near g ꢃ 4.3 (marked
as S2), which are the signature of isolated Mn2+ in sites with differ-
ent local symmetry.
g-tensor exhibiting monoclinic symmetry. The
g values are:
g1 = 14.60; g2 = 2.99 and g3 = 1.71.
According to Ref. [31] the g-values associated with transitions
within the C6 and C7 states in cubic symmetry are 6.00 and 6.8,
respectively. If the overall crystal field symmetry is lower than
cubic, the principal components of the g-tensor can be related to
the g-value calculated for cubic symmetry gc by:
As the Mn2+ ions belong to d5 configuration, in the axial distor-
tion of octahedral symmetry of crystal field, it gives rise to three
Kramers doublets. Therefore, five fine structure transitions are pos-
sible. Each fine structure transition will be split into six hyperfine
components due to 55Mn hyperfine coupling (I = 5/2; 100% abun-
dance). For powder and polycrystalline samples, all components
of the fine structure except that with MS = +1/2 M MS = ꢀ1/2 are
broadened to such an extent that they cannot be observed. Thus,
only one six line pattern is recorded in the experimental spectrum
[31].
gc ¼ gav ¼ ð1=3Þðg1 þ g2 þ g3Þ
ð3Þ
This approach is only valid if the lower symmetry crystal field
is small when compared to cubic crystal field [33] and has been
used successfully in the analysis of rare earth EPR spectra
[34,35]. However, if the difference between the value of gc and
gav is large (>0.1), the assumption that the low symmetry
components of the crystal field are small compared to those
when the cubic crystal field is present, breaks down [35]. In
our case, the gav value is 6.43 and is significantly distant from
the value predicted for cubic symmetry. It would indicate that
the low-symmetry crystal field around the erbium ion may be
strong.
The low temperature EPR spectrum for YF3:1%Er compound
revealed only broad resonance signals with the linewidth in the
range: 8–25 mT and the absence of any hyperfine components.
This means the formation of clusters (i.e., aggregation) of erbium
ions. The broadening of resonance lines is the result of both
The spin Hamiltonian for Mn2+ in low symmetric crystal field is
given by:
H ¼
l
BgBS þ ASI þ DðS2z ꢀ 35=12Þ þ ð1=2ÞðEðSþ2 þ S2ꢀÞ
ð4Þ
where D and E are the axial and rhombic fine structure parameters.
As Mn2+ is the S-state ion, the g- and A-tensors are isotropic. The
inset in Fig. 14 shows the spectrum calculated with g = 2.00 and
A = 84 ꢁ 10ꢀ4 cmꢀ1. As the ionic radius of Mn2+ is 0.96 Å and from

E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
567
of clusters of activator ions. No rare earth related EPR signal was
observed for samples: YF3:1%RE (RE = Ho, Pr, Tm). Contrary to the
erbium doped sample, XRD results indicate absence of holmium
clustering. EPR measurements exhibit the presence of Mn2+ ions
in the all investigated compounds. We suppose that these uncon-
trolled impurities originate from the chemical ingredients during
the technology procedure. From EPR it is observed that the Mn2+
ions in the all investigated compounds occupy more than one sites:
the first one is characterized by weakly deformed cubic sites with
distribution spin Hamiltonian parameters around their average
values and the other has completely rhombic symmetry sites.
YF3:Ho
LaF3:Pr(2)
YF3:Pr
YF3:Tm
T=4.2K
S2
S1
Acknowledgments
We express our thanks to the Director of Institute of Electronic
´
Materials Technology, Dr Z. Łuczynski for support of this work.
SQUID magnetometer was financed by European Regional Devel-
opment Fund whereas SEM/EDX Jeol was financed by Regional
Fund for Environmental Protection in Katowice.
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Magnetic field induction (T)
Fig. 15. X-band EPR spectra of LaF3:Pr, YF3:Ho,Pr,Tm at T = 4.2 K (gain: 1 ꢁ 104).
Appendix A. Supplementary material
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.jallcom.2014.
07.074.
the magnitude of spin-Hamiltonian parameters [37], it is reasonable
to assume that manganese occupy yttrium site and the valence of
manganese remains as Mn2+. Simultaneously, a complete smearing
of the hyperfine forbidden lines and significant broadening of
References
hyperfine allowed lines (linewidth of each component is
DB = 5 mT)
become visible (see Figs. 14 and 15). These are the characteristic
features of Mn2+ EPR spectra in disordered systems [38]. They are
due to effects of distribution of zero field splitting (D, E) parameters
around their average values.
[1] B.P. Sobolew, The Rare Earth Trifluorides, Institut d’Estudis Catalans, Barcelona,
Spain, 2000.
[2] A. Kaminskii, Laser Crystals, Springer, Berlin, 1990.
[3] A. Jusza, L. Lipinska, M. Baran, P. Mergo, A. Millan, F. Dieleman, R.
Piramidowicz, Proc. SPIE 8435, Organic Photonics V, 84351T, June 1, 2012,
doi:10.1117/12.922143.
[4] B.P. Sobolev, Crystallogr. Rep. 47 (1) (2002) S63–S75.
[5] J. Rodriguez-Carvajal, Recent developments of the program FULLPROF, in
commission on powder diffraction (IUCr) Newsletter, vol. 26, 2001, pp. 12–19
<http://journals.iucr.org/iucr-top/comm/cpd/Newsletters/>.
[6] J. Rodriguez-Carvajal, M.T. Fernández-Díaz, J.L. Martínez, J. Phys, Condens.
Matter 3 (1991) 3215–3234.
[7] P.W. Stephens, J. Appl. Cryst. 32 (1999) 281–289.
[8] P. Thompson, D.E. Cox, J.B. Hastings, J. Appl. Cryst. 20 (1987) 79–83.
[9] M. Jarvinen, J. Appl. Cryst. 26 (1993) 527–531.
[10] K. Momma, F. Izumi, J. Appl. Crystallogr. 44 (2011) 1272–1276.
[11] Multipak Software Version 6, edited by Physical Electronics, Minnesota, 1998.
[12] A.K. Cheetham, N. Norman, Acta Chem. Scand. Ser. A 28 (1974) 55–60.
[13] A. Zalkin, D.H. Templeton, J. Am. Chem. Soc 75 (1953) 2453.
[14] M. Mansmann, Zeit. Anorg. und Allgem. Chem. 331 (1964) 98–101.
[15] C. de Rango, G. Tsoucaris, C. Zelwer, Compt. Rend. Hebd. Seances Acad. Sci., Ser.
C, Sci. Chim. 263 (1966) 64–66.
The investigations of the EPR spectra of Mn2+ in a variety of dis-
ordered materials (glasses and polycrystalline powders) show that
the spectra have been characterized in addition hyperfine sextet,
the absorption around g ꢃ 4.3 [39]. These resonances arise from
the isotropic transitions within the lowest or middle Kramers dou-
blets. In our case, the intensity ratio I2.0/I4.3 decreases with decreas-
ing temperature, therefore the resonance near g = 4.3 arises from
the transition within the lowest Kramers doublet. An isotropic g-
factor of 30/7 from a high-spin d5 ion in a site of rhombic symme-
try can only appear if the symmetry is completely rhombic (k = E/
D = 1/3). Such interpretation was supported also by successful sim-
ulation of the EPR spectra for d5 ions by Yahiaoui and Kliava
[40,41].
[16] T.J. Udovic, Q. Huang, A. Santoro, J.J. Rush, Z. Krist. 223 (1979) 697–705.
[17] D. Gregson, C.R.A. Catlow, A.V. Chadwick, G.H. Lander, A.N. Cormack, B.E.F.
Fender, Acta Cryst. B 39 (1983) 687–691.
4. Concluding remarks
[18] R.D. Shannon, Acta Cryst. A32 (1976) 751–767.
The thermal decomposition method was successfully applied
for the first time in order to obtain the good quality YF3:1%RE
materials. The SEM investigations showed that the particles are
nanometric in size as also confirmed by XRD measurements. The
synthesized samples have a stoichiometry close to the nominal
what composition as confirmed by XRD and magnetic measure-
ments. For the samples doped with praseodymium, PrF3 was also
observed. Surface contamination can be suggested to be responsi-
ble for the deterioration of the optical properties of the composites.
Positions of XPS lines are in agreement with the literature data. The
energy gap is about 10 eV. The chemical shift of the XPS lines indi-
cate covalent type of bonds. To the best of our knowledge, no EPR
measurements have been done on YF3:1%Er material. We conclude
that trivalent erbium preserves the local site symmetry of the
Y-site in the investigated fluoride material. The absence of any
hyperfine structure and broad resonance signals which are
observed in the EPR spectrum of YF3:1%Er indicate the formation
[19] F.W.H. Zachariesen, J. Less Common Metals 62 (1978) 1–7.
[20] T. Balic Zunic, I. Vickovic, J. Appl. Cryst. 29 (1996) 305–306.
[21] B.E.G. Lucier, K.E. Johnston, D.C. Arnold, J.-L. Lemyre, A. Beaupre, M. Blanchette,
A.M. Ritcey, R.W. Schurko, J. Phys. Chem. C 118 (2014) 1213–1228.
[22] M. Wang, Q. Huang, J. Hong, X. Chen, Mater. Lett. 61 (2007) 1960–1963.
[23] N.Yu. Kirikova, J.C. Krupa, N.V. Makhov, C. Severac, J. Elect, J. Electron
Spectrosc. Relat. Phenom. 122 (2002) 85–89.
[24] L. Chong, J. Zou, X. Zeng, W. Ding, J. Mat. Chem. A 1 (2013) 3983–3991.
[25] E. Sarantopoulou, Z. Kollia, A.C. Cefalas, Opt. Mater. 18 (2001) 23–26.
[26] H.D. Wiemhöfer, W. Göpel, Frasenius, J. Anal. Chem. 341 (1991) 106–111.
[27] J. Moulder, W. Stickle, P. Sobol, K. Bomben, Handbook of X-ray photoelectron
spectroscopy, Phys. Electron. (1995).
[28] E. Talik, A. Paja˛czkowska, A. Guzik, P. Zajdel, J. Kusz, A. Kłos, A. Szysiak, Mater.
Sci. Eng. B 182 (2014) 74–80.
[29] C. Suzuki, T. Mukoyama, J. Kawai, H. Adachi, Phys. Rev. B 51 (1998) 9507–9514.
[30] C. Leycuras, H. Le Gall, M. Guillot, A. Marchand, J. Appl. Phys. 55 (1984) 2161–
2163.
[31] A. Abragam, B. Bleaney, Electron Paramagnetic Resonance of Transition Ions,
Clarendon Press, Oxford, 1970.
[32] S.K. Misra, Physica B 253 (1998) 111.
[33] J.D. Carey, R.C. Barklie, J.F. Donegan, F. Priolo, G. Franzo, S. Coffa, Phys. Rev. B 59
(1999-II) 2773.

568
E. Talik et al. / Journal of Alloys and Compounds 616 (2014) 556–568
[34] J. Dziesiaty, St. Muller, R. Boyn, Th. Buhrow, A. Klimakow, J. Kreissl, J. Phys.
:Condens. Matter 7 (1995) 4271.
[35] R.K. Watts, W.C. Holton, Phys. Rev. 173 (1968) 417.
[36] S.A. Kazanskii, Sov. Phys. JETP 62 (1985) 727.
[37] V.K. Jain, G. Lehman, Phys. Status Solidi B 159 (1990) 495. and references
therein.
[38] S.K. Misra, Physica B 203 (1994) 193.
[39] V. Cerny, B. Frumarova-Petrova, J. Rosa, I.L. Licholit, M. Frumar, J. Non-Cryst.
Solids 192&193 (1995) 165.
[40] E.M. Yahiaoui, R. Berger, Y. Servant, J. Kliava, L. Cugunov, A. Mednis, J. Phys.:
Condens. Matter 6 (1994) 9415.
[41] J. Kliava, R. Berger, Mol. Phys. Rep. 39 (2004) 130.