Journal of Structural Chemistry. Vol. 48, No. 5, pp.

831-839, 2007
Original Russian Text Copyright © 2007 by R. I. Mashkovtsev

PARAMAGNETIC CENTERS RELATED TO Al, Sc, In,

AND Nb IMPURITIES IN KTiOAsO4 CRYSTALS

R. I. Mashkovtsev UDC 538.113;548.3

Electron paramagnetic resonance (EPR) is applied to study Al-, Sc-, In-, and Nb-doped KTiOAsO4 (KɌȺ)
crystals. Paramagnetic hole centers Ɉ– are observed after ionizing irradiation of KTA crystals. These
centers are, as a rule, unstable at room temperature and are slowly annealed for about two weeks. Oxygen
ions are bridging two cations in KɌȺ. Near the impurity, two ɪ-orbitals of oxygen atoms participate in
covalent bonding with cations, whereas the third ɪ-orbital remains free and under the radiation effect
captures the hole thus forming the paramagnetic center of Mn+–O––M(n–1)+ (here Mn+ is the lattice cation and
M(n–1)+ is the impurity cation of Al, In, Sc, or Nb). In the centers investigated the specific principal direction
of the g-factor g a 2 is normal to the Mn+–O––M(n–1)+ plane, and the main value of gmax falls in this plane.
The direction of the O––M(n–1)+ bond is close to the selected direction of the hyperfine interaction with the
impurity ion. The models of six hole centers and the found parameters of EPR spectra are discussed.
Keywords: EPR, hole centers, KTiOAsO4 (KɌȺ).

Potassium titanyl phosphate KTiOPO4 (KTP) and its analogues are of particular interest because they are nonlinear
optical crystals. Crystals of KTiOAsO4 (KɌȺ) as well as KɌɊ belong to the orthorhombic symmetry mm2 (space group
Pna21) with lattice parameters as follows: a = 13.13 Å, b = 6.58 Å, and c = 10.78 Å [1, 2]. Titanium ions, denoted as Ti1 and
Ti2, are arranged in two crystallographically inequivalent sites of the KTA structure. Oxygen atoms comprise a distorted
octahedral coordination of each titanium ion. Four octahedral oxygen atoms are common with AsO4 tetrahedra, whereas two
other, denoted as OT1 ɢ OT2, are involved into an infinite chain of titanium and oxygen atoms (Fig. 1). These Ti1 and Ti2
ions differ in that the two chain oxygen atoms bonded to Ti1 form the OT1–Ti1–OT2 angle, close to 90q, while the two
oxygen atoms bonded to Ti2 form the OT1–Ti2–OT2 angle, close to 180q. Each titanium ion Ti2 has a short bond (a1.74 Å)
with one oxygen atom in the Ti–O chain, whereas the Ti–O trans-bond is long (a2.1 Å). There are also two different sites for
alkaline ions and two for As ions. Oxygen ions occupy 10 crystallographically inequivalent sites. Together, these ions occupy
16 different sites in the KTA crystal whose unit cell contains 64 ions.
The survey [3] describes the preparation of these crystals, their properties and the application of isomorphic KTP
compounds.
Paramagnetic centers in the KTP crystal family. EPR is known as an effective tool for investigating impurities
and defects in crystals. As far as we know, apart from our works [4-7], only Fe3+ impurity in KTA crystals was studied by
EPR spectroscopy [8]. By EPR we studied hole centers related to Al and In impurities [4,6], and Rh2+ ion [7]. Since the

Institute of Geology and Mineralogy, Siberian Division, Russian Academy of Sciences, Novosibirsk;
[email protected]. Translated from Zhurnal Strukturnoi Khimii, Vol. 48, No. 5, pp. 890-898, September-October, 2007.
Original article submitted December 6, 2006.

0022-4766/07/4805-0831 © 2007 Springer Science+Business Media, Inc. 831

 Fig. 1. Projection of KTiOAsO4 structure in ac
plane.

structures of KɌɊ and KɌȺ are virtually identical, all paramagnetic centers studied in KɌɊ (see [9-21] and references
therein) are likely to be observed also in KTA crystals. Though we have not been intended to prove this possibility, still it
may be useful to enlist main paramagnetic centers (PC) observed in KTP. As EPR spectroscopy is suitable for studying
defects, it seems promising in revealing the nature of the so-called gray tracks (dark colored regions) which are induced in
KTP crystals by high-power lasers. Therefore, part of the works was devoted to the analysis of the nature of gray tracks by
means of EPR and optical absorption spectra. Dark regions in the crystal bulk also emerge when crystals are electrothermally
treated.
Four PC bound with Ti3+ ion, which probably cause crystals to darken, were obtained both by the electrochemical
treatment and by annealing the crystal in a hydrogen atmosphere at 800qɋ [9]. Scripsick [10] observed three radiation centers
stable only at low temperatures — Ti3+, Pt3+ ions and the hole-type center. Andreev and Efimov [11] found a hole aluminum
center (this center is examined in detail in [12]) and Ti3+ ions previously observed in [9]. Edwards et al. [13] studied the main
radiation hole center inherent in hydrothermal and flux grown crystals. They proposed the model of this center as the hole
captured by bridging oxygen ion OT2 between two titanium ions. Hole center stabilization confirmed the presence of K+ ion
vacancies in the lattice. In the same work [13] it was suggested that this Ɉ– center can be essentially important for the
formation mechanism of gray tracks in KTP crystals. A similar hole center was obtained as a result of X-ray irradiation of the
KɌɊ crystal at 77 K [14]. When heated to 160 K, this center disappeared, but there occurred two new hole centers, which was
related to the beginning of the diffusion of potassium ion vacancy.
EPR spectroscopy was also applied to investigate isomorphic impurities in KɌɊ crystals. Three paramagnetic
centers are related to platinum entering into the sites of Ti4+ and K+ ions [15]. The Pt (A) center (Pt3+ ion) is generated by X-
ray irradiation due to electron capturing by Pt4+ ion substituting for titanium one. It is assumed that Pt (B) and Pt (C) centers
are Pt+ ions substituting for K+ ions in two crystallographically inequivalent sites. They are generated when Pt0 atom captures
the hole. Platinum embeds itself in the KTP lattice as an uncontrolled impurity during the crystal growth in platinum
crucibles. Rhodium impurities can enter into the KTP structure too, when crystals are grown in platinum-rhodium crucibles
[16]. Two paramagnetic Rh2+ centers are studied, which originate from Rh3+ ions substituting for titanium ions in two
inequivalent sites after reducing annealing of KTP crystals in vacuum. Also Fe3+, Cr3+ [17], V4+ [18], Mo5+ [19], and W5+ [20]
ions in substitution positions for titanium ions are examined by EPR spectroscopy in KɌɊ.

832
 In this work we present the results obtained in EPR studies of hole centers in KɌȺ crystals. A part of the work is
aimed at correcting the interpretation of results, which is caused by an erroneous determination of crystallographic axes in
crystals doped with Al and In. Therefore the refined data for Al and In impurity centers are given along with new results of
the investigations of hole centers in Sc- and Nb-doped KɌȺ crystals.

EXPERIMENTAL

We have examined a large number of samples. KTA crystals were grown by flux method in platinum crucibles on
crystal seeds oriented along the [010], [001], or [100] axes. The temperature of crystal growth was varied from 800qɋ to
1000qɋ. In specific experiments crystals were grown with admixtures of oxides of those metals which were supposed to be
introduced in the crystal structure [5]. In some crystals uncontrolled impurities were found. As a rule, in order to obtain PC
we needed to expose crystals to ionizing radiation (Ȗ-irradiation from a 60ɋɨ source or hard X-rays) of 2 Mrad. EPR spectra
were recorded on a Radiopan SE/X 2543 spectrometer operating in the X-range. The angular dependence of spectra in the
principal crystallographic planes was measured using a goniometer to determine the principal values of the spin Hamiltonian
tensors. The technique of calculating the parameters of EPR spectra was described previously [4, 7].

RESULTS AND DISCUSSION

As experience in studying crystals of the KTP family shows, the observation of EPR spectra in them is due to either
paramagnetic ions with an unfilled d-shell, or with PC that form near impurities under ionizing radiation at room temperature
and at 77 K. In the latter case the paramagnetic centers are thermally unstable.
In oxide compounds hole capturing is observed, as a rule, on oxygen atoms near substituting cations that have lower
valence than the network-forming metal. Centers with holes captured on oxygen atoms were observed in KTP as a result of
irradiation by laser, X-, and J-rays [9-14]. After J-irradiation at 77 K and also at room temperature we found the hole centers
almost in all KTA samples examined. Further we consider similar centers in the crystals which were specially doped with In
and Sc impurities because it is known that the introduction of a small amount of oxides such as In2O3, Sc2O3, and Fe2O3 in the
flux during the growth stabilizes the single-domain structure of KTA crystals [21]. We also observed the hole center related
to Al impurity that uncontrollably entered into the structure of the crystal grown from the flux with WO3, probably, to
compensate for the excessive positive charge of tungsten which could enter into the KTA crystal structure. All the centers
found by us turned out to be unstable and were annealed at room temperature for two weeks, nearly a 50% decrease of the
number of centers being observed on the first day after J-irradiation. Figs. 2 and 3 display EPR spectra taken at 77 K in the
irradiated crystals. To analyze the EPR spectra a spin Hamiltonian of the following form is used:
= ȕSgH + SAI,
where the first term describes the Zeeman energy of electron, and the second is the energy of the hyperfine interaction (HFI)
of the electron and nuclear spin.
In oxide compounds oxygen atoms are often bridging the lattice cation and the impurity. In this case two ɪ-orbitals
of oxygen atoms participate in covalent bonding with two ions, while the third ɪ-orbital remains free and can capture the
hole. According to the theory of the hole center Ɉ– [23], the direction of this ɪ-orbital, normal to the Mn+–O––M(n–1)+ plane
(here Mn+ is the lattice cation and M(n–1)+ is the impurity ion), indicates the specific principal direction of the g-factor (g a 2).
Another principal value of gmax falls in the Mn+–O––M(n–1)+ plane and has the selected direction along the Mn+–M(n–1)+ line.
The selected direction of the hyperfine interaction (HFI) with the impurity ion must be close to the direction of the O––M(n–1)+
bond. It is these typical features that a well-known “aluminum” center in quartz has [24].
The hole center O– related to Al impurity has been well studied in many compounds. The EPR spectrum of this
center is characterized by HFS due to the interaction between the electron spin and the magnetic moment of 27Al nucleus

833
 Fig. 2. EPR spectra of radiation hole centers Fig. 3. EPR spectra of radiation hole
associated with Al (ɚ) and In (b) impurities centers associated with Sc (ɚ) and
in KTA crystals. Nb (b) impurities in KTA crystals.

(nuclear spin I = 5/2, natural abundance 100%). Here the HFS value for all compounds varies in narrow limits from 5G to
7G. The “aluminum” hole center is examined in KTP crystals specially activated by aluminum impurity [12]. Hole centers O–
induced by Sc and In impurities are more rare; we know only one work [25] devoted to the investigation of tetragonal crystals
of GeO2 with the addition of Sc, Y, and In. A multicomponent HFI is typical of EPR spectra of hole centers O– induced by Sc
and In impurities (45Sc isotope has a nuclear spin of I = 7/2 with a natural abundance of 100%, and 115In and 113In isotopes
have I = 9/2 with a natural abundance of 95.5 and 4.5%, respectively). Figs. 2 and 3 show the observed hyperfine structure in
the EPR spectra of KTA crystals for O––In3+ and O––Sc3+ centers.
The angular dependence of EPR spectra for all centers was studied in three crystallographic planes. The correctness
of signs for the non-diagonal elements in g2 and (gA)2 matrices was additionally checked by measuring EPR spectra in
oblique orientations. The results obtained are given in Tables 1 and 2. Though two different O– centers related to Al impurity
were observed, Tables 1 and 2 enlist the results for only one center that we managed to examine (HFS lines for this center are
schematically shown in Fig. 2ɚ). The spectra superimposed on each other were complicated by additional lines due to the
quadrupole interaction in the second center, which prevented from tracing its angular dependence, Moreover, because of a
relatively large width of the line (varying in the limits from 3.5 G to 6 G) and small anisotropy of HFS we failed to measure
the anisotropic value b for the O––Sc3+ center, therefore Table 2 included only the mean value of the HFI isotropic constant
with the 45Sc nucleus.
In order to determine the supposed structure of defects, it is necessary to compare the specific principal directions of
g and A matrices with the structural directions of KTA crystal. Unlike simple oxides, the structure of KTA is more complex.
There are ten structurally different oxygen atoms, and the sites of both anions and cations have only trivial symmetry at that.
Based on the comparison of the magnitudes of ionic radii, it is possible to assume that the cations in question can occupy the
sites of Ti ions. Finally, Table 3 presents the structural directions that fit the hole centers (the angle of deviation between the

834
 TABLE 1. g Matrices for O––Al3+, O––In3+, and O––Sc3+ centers in KɌȺ

g2 in crystal coordinates Principal values Direction cos

a b c of g±0.0005 l m n

O––Al3+
a 4.0590 g1 2.0736 0.343 0.559 0.755
b 0.04 4.1270 g2 2.0129 –0.513 0.784 –0.348
c 0.08 0.11 4.1821 g3 2.0039 0.786 0.269 –0.557
O––In3+
a 4.0556 g1 2.0588 0.272 0.605 0.748
b 0.02 4.1186 g2 2.0145 –0.670 0.676 –0.306
c 0.05 0.09 4.1477 g3 2.0062 0.688 0.422 –0.590
O––Sc3+
a 4.0868 g1 2.0626 0.450 0.404 0.796
b 0.07 4.1046 g2 2.0226 –0.552 0.827 –0.107
c 0.009 0.09 4.1669 g3 2.0039 0.703 0.389 –0.595

TABLE 2. HFI A Matrices for O–—Al3+, O–—In3+, and O–—Sc3+ Centers in KTA
g2A2 in crystal coordinates Principal values Direction cos
a b c of Ⱥ±0.3 MHz l m n

O––Al3+
a 2443 ~A1~ 20.5 0.411 0.569 0.712
b –300 2189 ~A2~ 24.2 0.480 0.799 0.362
c –250 –100 2193 ~A3~ 25.7 0.775 0.193 0.602
a = –23.4, b = 1.47, ɟ = 0.7
O––In3+
a 50272 ~A1~ 128.2 0.066 0.728 0.682
b 320 59739 ~A2~ 112.7 0.415 0.642 0.645
c 1776 8234 59731 ~A3~ 110.1 0.908 0.239 0.344
a = 117, b = 5.6, e = 1.3
O––Sc3+
a | –20.7

structural directions and the orientation of the principal values of g and A matrices is given in the last column) chosen from a
large number of those available.
It follows from Table 3 that the hole is captured on the ɈɌ2 oxygen atom which is bridging between two structurally
different Ti ions, the impurity ion being located in the Ti2 site (Fig. 1). A similar conclusion the authors of [12] made, who
analyzed only the HFI principal direction with Al3+ ion. It should be noted that unlike KTP, in KTA crystals we observed two
hole centers with Al3+. Here it is possible to suggest that the second center is induced by the substitution for titanium ion in
the Ti1 site.
With the help of relations [26] A1 = a + 2b, A2 = a – b + e, A3 = a – b – e, the principal values of HFI are usually
divided into isotropic Fermi contact interaction a, dipole-dipole interaction b, and orthorhombic dipole interaction e. As
M(n–1)+ ion is at the node of the ɪ-orbital of the unpaired electron, then the direct overlapping between metal and oxygen
orbitals does not occur, and hence, there shall be no electron distribution on the metal s-shell. However, Fermi interaction is
observed experimentally, and its appearance is explained by the exchange polarization of closed s-shells that results in a

835
 TABLE 3. Comparison of Structural Directions in KTA with Principal Directions of g and A Matrices for O––M3+ Centers
Direction cos
Structural direction Deviation angle, deg
l m n

A to Ti1–OT2–Ti2 0.6753 0.3669 –0.6401

O––Al3+
g3 0.786 0.269 –0.557 9.6
– 3+
O –In
g3 0.688 0.422 –0.590 4.4
– 3+
O –Sc
g3 0.703 0.389 –0.595 3.3
Ti1–Ti2 0.4789 0.4416 0.7587
O––Al3+
g1 0.343 0.559 0.755 10.3
– 3+
O –In
g1 0.272 0.605 0.748 15.2
– 3+
O –Sc
g1 0.450 0.404 0.796 3.8

Ti2–OT2 0.2547 0.6981 0.6691

– 3+
O –Al
A1 0.411 0.569 0.712 11.9
– 3+
O –In
A1 0.066 0.728 0.682 11.1

difference between densities of the wave functions of electrons with opposite spin directions (up (+) and down (–)) [27, 28].
The polarization parameter F is determined as
4S 2 2
F{ ¦ ª M (0)
2 S ¬«
 M (0) º ,
¼»
i

where S = 1/2 and the summation is over all internal s-orbitals of the impurity ion. Then Fermi contact interaction is
determined by the formula
2
a g eµ B g N µ N Ȥ.
3
For light Al and Sc ions the negative value of the a contact HFI is observed, while for heavy Ga and Y it is positive [25, 27].
For light ions, the spin exchange interaction of 2ɪ Ɉ– attracts (+) spins of external s-orbitals of the impurity ion leaving
negative (–) spin density closer to the nucleus. In heavy ions, the internal shells in which the opposite spin relaxation occurs
start to play an important role, which results in a positive value of F [26, 28].
The negative sign for isotropic contact interaction with Al3+ is obtained experimentally from the
relation A1 < A2 , A3 at the positive sign of dipole-dipole interaction. The positive sign of the isotropic parameter of HFI
with In3+ is obtained from EPR spectra: the observed splitting between hyperfine lines in the high-field region of the spectrum
is larger than the splitting in its low-field region. This space tapering in HFS is explained by taking into account second order
terms in the determination of the spin Hamiltonian energy by perturbation theory [29] and is consistent with the spectrum
observed, if the isotropic component ɚ of HFI is positive. Isotropic HFI for Sc3+ should be negative in accordance with theory
confirmed by the experimental result [25]. Determined in the same work [25] very weak anisotropy, b | 0.15 G for Sc3+, is
corroborated by our observations. Unfortunately, we were not able to compare the HFI constants obtained for In3+ with any

836
data because the HFS observed in the work [25] for the hole center with In3+ was so complicated, due to quadrupole
interaction, that it was not possible to determine the HFI tensor.
It seems worth noting that, unlike in the Al-doped crystals, we observed the introduction of In3+ and Sc3+ impurities
only in one Ti2 structural site in KTA:In and KTA:Sc crystals. Probably, it is this asymmetry of the introduction of impurities
that determines the stabilization of the single-domain structure of KɌȺ crystals grown with In3+ and Sc3+ added as dopants
[22].
One more type of centers observed in different KTA crystals is hole centers with a complex HFS which
demonstrates the properties of HFI with a proton when the value of hyperfine splitting is comparable with nuclear Zeeman
frequency QH | 0.5 mT. In this case, four HFS lines are observed [30] instead of the two expected in EPR spectra for the
defects which have a nuclear spin of I = 1/2 in their environment. For the centers with a strong anisotropy of the g matrix and
HFI, the HFS lines change their intensity up to their disappearance or they superimpose on each other when the crystal
rotates, which prevents from determining the tensor of HFI with a proton in the analysis of the angular dependence of the
EPR spectrum. These cases were described previously for hole centers in apatite [31], and we often observed them when
studying the KTA crystals. Therefore, for O–/H+ centers we determined only the g matrix, its principal values and directions
given in Table 4.
To find the structure of the centers we compared the g matrix principal directions with the directions of bonds and
normals to the planes formed by atoms in the KTA structure. As seen from the data of Table 5, satisfactory results are
obtained if we suppose that the hole is captured by the bridging ɈɌ1 oxygen atom. At the present stage we can only guess

TABLE 4. Parameters of O–/H+ Centers

Principal values of g Direction cos

l m n

O–/H+(I)
g1 2.0376 0.373 0.502 –0.779
g2 2.0207 0.898 0.403 –0.175
g3 2.0028 0.232 0.764 0.602
O–/H+(II)
g1 2.0459 0.332 0.512 –0.791
g2 2.0192 0.938 0.261 –0.225
g3 2.0041 0.099 0.830 0.549

TABLE 5. Comparison of Structural Directions in KTA with Principal Directions of g Matrices for O–/H+ Centers

Structural direction Direction cos Deviation angle, deg

l m n

A to Ti1–OT1–Ti2 0.4102 0.6941 0.5915

O–/H+(I)
g3 0.232 0.764 0.602 11
O /H+(II)
–

g3 0.099 0.830 0.549 19.7

Ti2–Ti1 0.4789 0.4416 –0.7587
O–/H+(I)
g1 0.373 0.502 –0.779 7.5
O–/H+(II)
g1 0.332 0.512 –0.791 9.9

837
 TABLE 6. Parameters of the Spin Hamiltonian for the OT2 —Nb5+ Center Observed at 77 K in KTA and
Comparison of Principal Directions of the g Matrix with Structural Directions

Principal values of g matrix Principal directions

l m n
g1 2.0320r0.0005 0.281 0.566 0.775
g2 2.0229r0.0005 –0.982 0.036 0.183
g3 2.0051r0.0005 0.033 –0.813 0.582
Comparison of structural directions in KTA with principal axes of the g tensor
Structural direction Principal g-axes
T, deg M, deg T, deg M, deg
OT2–Ti2 48 69.9 g1 39.2 63.6
A to the 42 252.4 g3 54.4 272.3
K2–OT2–Ti2
plane
Note. The g matrix is given in coordinates (a, b, c). In the diagonal form the orientation of principal axes is given by
direction cosines l, m, n and polar angles (T, M) relative to the system of (a, b, c) axes.

why the centers associated with the hole capture in the same site of ɈɌ1oxygen atom have different parameters of EPR
spectra. Either it is caused by different positions of a proton, or additional impurities or vacancies of atoms occur in the
second sphere surrounding the O–/H+ centers. We have noticed that the value of HFI with a proton changes for the O–/H+(I)
center when the crystals are stored at room temperature, which indicates the reorganization of this center.
One more hole-type center was observed in crystals grown with niobium and magnesium oxides added to the flux.
Two crystals were investigated. The first sample of KTA:Nb1 had a composition of K0.95(Ti0.96Mg0.01Nb0.05O)AsO4, and the
second sample of KTA:Nb2 was K0.96(Ti0.91Mg0.03Nb0.09O)AsO4. After the KTA:Nb1 sample was Ȗ-irradiated at 77 K, the
O–/H+(II) center was observed, but no paramagnetic centers were observed after Ȗ-irradiation at room temperature. In the
KTA:Nb2 sample after Ȗ-irradiation at room temperature, we observed the spectrum unstable at room temperature which
consisted of ten lines of the hyperfine structure (Fig 3b). The cause of observing ten lines seems to be HFI with 93Nb isotope
(I = 9/2, natural abundance 100%). The value of the g-factor along the ɚ axis was ga = 2.0233 and the corresponding
hyperfine splitting Aa = 1.7 mT. For two other crystallographic axes gc = 2.0141 and gb = 2.0227, while HFI Ac = 2.07 mT
and Ab = 2.03 mT. The ion of Nb4+ cannot cause the observed EPR spectrum because the value of the g-factor for Nb4+ ion
(4d1) should be less than two, and HFI should be larger than that observed by the order of magnitude [32]. A small positive
shift of the g-factor and weak HFI for this center suggest that the unpaired spin is captured on the oxygen atom in the
immediate environment of the substituting Nb ion. The centers with similar values of g and A were observed in MoO3 [33]
and ɋdWO4 crystals [34] and were assigned to Ɉ– centers near the substituting Nb5+ ion.
The principal values and directions of the g matrix were determined from the angular dependence of spectra
(Table 6). Weak anisotropy and rather broad lines (0.6-1.4 mT) made the determination of exact values of the A matrix
impossible, but we determined that Amax = 2.23r0.05 mT was directed approximately along the gmax. Though the principal
values of the g-factor have the directions which are well consistent with some directions in the KTA crystal (Table 6),
however, it is difficult to say something definite about the structure of the defect, apart from that the hole is captured by the
oxygen atom in the ɈɌ2 site. It is difficult to explain the hole center formation in the assumption of a single Nb5+ion
substituting for Ti4+ ion. In this case it would be logical to suppose a change in the niobium ion valence state when d1 Nb4+
ion appears. Magnesium ion is likely to play some role because the paramagnetic OT2 –Nb5+ center does not form in the
KTA:Nb1 sample with a low concentration of Mg. The studies of K[Nb,Mg]OAsO4 crystals [35] suggest that Ti4+ is
substituted in the adjacent sites by pairs of ions Nb5+–Mg2+. In this case the local positive charge is in deficiency, and the hole

838
center may form under irradiation. In the nearest environment of the center there is also the Ɉɇ group because the splitting in
two components (AHa3 G) of the narrowest HFS lines is well seen in the EPR spectrum (Fig. 3b, central part).
The author is grateful to L. I. Isaenko for the samples kindly provided for this investigation.

REFERENCES

1. M. El Brahimi and J. Durand, Rev. Chim. Min., 23, 146-153 (1986).

2. S. C. Mayo, P. A. Thomas, S. J. Teat, et al., Acta Crystallogr. B, 50, 655-662 (1994).
3. M. N. Satyanarayan, A. Deepthy, and H. L. Bhat, Crit. Rev. Solid St. Mater. Sci., 24, 103-191 (1999).
4. R. I. Mashkovtsev and L. I. Isaenko, Solid State Commun., 95, 739-743 (1995).
5. L. Isaenko, A. Merkulov, R. Mashkovtsev, et al., J. Cryst. Growth., 166, 502-506 (1996).
6. R. I. Mashkovtsev and L. I. Isaenko, Phys. Stat. Sol. (b), 198, 577-585 (1996).
7. R. I. Mashkovtsev and L. I. Isaenko, Ferroelectrics, 330, 85-92 (2006).
8. Y. Dusausoy, V. Lorrain, F. Ribert, et al., Appl. Magn. Res., 5, 331-337 (1993).
9. M. G. Roelofs, J. Appl. Phys., 65, 4976-4980 (1989).
10. M. P. Scripsick, G. J. Edwards, L. E. Halliburton, and R. F. Belt, J. Appl. Phys., 70, 2991-2994 (1991).
11. B. V. Andreev and V. N. Efimov, Mod. Phys. Lett. B, 6, 177-180 (1992).
12. J.-T. Yu, C.-H. Lee, K.-T. Liu, et al., J. Phys. Chem. Solids, 56, 233-240 (1995).
13. G. J. Edwards, M. P. Scripsick, L. E. Halliburton, and R. F. Belt, Phys. Rev. B, 48, 6884-6891 (1993).
14. M. A. Laruhin, V. N. Efimov, and V. A. Nazarova, Appl. Magn. Res., 12, 517-527 (1997).
15. N. Y. Garces, K. T. Stevens, and L. E. Halliburton, J. Appl. Phys., 87, 8682-8687 (2000).
16. D. Bravo, M. J. Martin, J. Gavalda, et al., Phys. Rev. B, 50, 16224-16231 (1994).
17. J. M. Gaite, J. F. Stenger, Y. Dusausoy, et al., J. Phys.: Condens. Matter., 3, 7877-7886 (1991).
18. J. Han, J. Wang, Y. Xu, et al., ibid., 4, 6009-6014 (1992).
19. I. N. Geifman, A. N. Usov, and P. G. Nagornyi, Phys. Stat. Solid. (b), 172, K73-K76 (1992).
20. D. Bravo, X. Ruiz, F. Diaz, and F. J. Lopez, Phys. Rev. B, 52, 3159-3169 (1995).
21. K. T. Stevens, L. E. Halliburton, M. Roth, et al., J. Appl. Phys., 88, 6239-6244 (2000).
22. L. T. Cheng, L. K. Cheng, J. D. Bierlein, et al., Appl. Phys. Lett., 62, 346-348 (1993).
23. A. E. Hughes, B. Henderson, in: Point Defects in Solids, J. H. Crawford and L. M. Slifkin (eds.), Vol. 1, Chap. 7,
Plenum, New York (1972).
24. R. H. D. Nutall and J. A. Weil, Can. J. Phys., 59, 1696-1708 (1981).
25. R. B. Bossoli, T. J. Welsh, O. R. Gilliam, and M. Stapelbroek, Phys. Rev. B, 19, 4376-4381 (1979).
26. O. F. Schirmer, K. W. Blazey, W. Berlinger, and R. Diehl, ibid., 11, 4201-4211 (1975).
27. M. Stapelbroek, O. R. Gilliam, and R. H. Bartram, ibid., 16, 37-43 (1977).
28. R. E. Watson and A. J. Freeman, in: Hyperfine Interactions, A. J. Freeman and R. B. Frankel (eds.), Academic, New
York (1967).
29. A. Abraham and B. Bleaney, Electron Paramagnetic Resonance of Transition Ions, Clarendon Press, Oxford (1970).
30. C. P. Poole and H. A. Farach, The Theory of Magnetic Resonance, Wiley-Interscience, New York (1972).
31. M. Mengeot, R. H. Bartram, and O. R. Gilliam, Phys. Rev. B, 11, 4110-4124 (1975).
32. V. M. Vinokurov, M. M. Zaripov, V. G. Stepanov, et al., Fiz. Tverd. Tela, 5, 2034-2039 (1963).
33. V. A. Ioffe, I. B. Patrina, and E. V. Zelenetskaya, Phys. Stat. Solid., 37, 425-432 (1970).
34. H. J. Murphy, K. T. Stevens, N. Y. Garces, et al., Radiation Eff. Def. Solids, 149, 273-278 (1999).
35. V. I. Chani, K. Shimamura, S. Endo, and T. Fukuda, J. Cryst. Growth., 173, 117-122 (1997).

839