The phase diagram GdF3-LuF3

Article abstract of DOI:10.1016/j.jssc.2008.02.017

The phase diagram gadolinium fluoride-lutetium fluoride was determined by differential scanning calorimetry (DSC) and X-ray powder diffraction analysis. Both pure components undergo a reversible first order transformation to a high temperature phase. The mutual solubility of both components is unlimited in the orthorhombic room temperature phase. The maximum solubility of Lu in the high temperature phase of GdF₃ (tysonite type) is about 20% and the maximum solubility of Gd in LuF₃ (α-YF₃ type) is about 40%. Intermediate compositions of the low temperature phase decompose upon heating in a peritectoid reaction to a mixture of both high temperature phases.

Full text of DOI:10.1016/j.jssc.2008.02.017

ARTICLE IN PRESS
Journal of Solid State Chemistry 181 (2008) 1070–1074
www.elsevier.com/locate/jssc
The phase diagram GdF3–LuF3
Ã
I.M. Ranieria, S.L. Baldochia, D. Klimmb,
aCenter for Lasers and Applications, Inst. Pesquisas Energeticas & Nucl., CP 11049, Butanta˜ 05422-970, Sa˜o Paulo, SP, Brazil
bInstitute for Crystal Growth, Max-Born-Str. 2, 12489 Berlin, Germany
Received 3 December 2007; accepted 21 February 2008
Available online 29 February 2008
Abstract
The phase diagram gadolinium fluoride–lutetium fluoride was determined by differential scanning calorimetry (DSC) and X-ray
powder diffraction analysis. Both pure components undergo a reversible first order transformation to a high temperature phase. The
mutual solubility of both components is unlimited in the orthorhombic room temperature phase. The maximum solubility of Lu in the
high temperature phase of GdF3 (tysonite type) is about 20% and the maximum solubility of Gd in LuF3 (a-YF3 type) is about 40%.
Intermediate compositions of the low temperature phase decompose upon heating in a peritectoid reaction to a mixture of both high
temperature phases.
r 2008 Elsevier Inc. All rights reserved.
Keywords: Phase diagram; Solid solution; Phase transformation
1. Introduction
Another characteristic pointed out by Mansmann for the
change of the structure from trigonal to orthorhombic was
due to a relation between the size of the metal and its
coordination number. When the ionic radius of the rare
earth rRE decreases, the fluorine ions tend to touch each
other, resulting in a repulsive energy. A critical ratio
rRE=rF ¼ 0:94 was proposed for the change from the LaF3
to the a-YF3 structure, taking into account the ionic radius
calculated by Ahrens [8]. Furthermore, the trigonal
structure could be stabilized if the REF3-orthorhombic
structure becomes deficient in fluorine ions or at high
temperatures [5].
Rare earth trifluorides (REF3, RE ¼ La; Ce; . . . ; Lu),
including yttrium fluoride, were extensively studied in the
past, instead of this many questions about the high-
temperature polymorphic transformations and its structure
so far remain. Oftedal [1,2] and Schlyter [3], determined the
LaF3 structure as hexagonal with the P63=mcm space
group and two formula units per elementary cell, using
synthesized powder and the lanthanum mineral tysonite
( ¼ fluocerite). Mansmann [4,5] and at same time Zalkin
et al. [6] established the current accepted structure of the
LaF3. Using smaller X-ray wavelength ðMoKaÞ it was
possible to observe weaker additional reflections on a
single crystal, and the structure of the LaF3 tysonite
Thoma et al. [9] studied the behavior of the REF3 from
SmF3 to LuF3 taking into account X-ray diffraction at
high temperature and differential thermal analysis (DTA)
experiments. It was observed that the REF3 from Sm to Ho
¯
structure was described as trigonal with space group P3c1
¯
have the trigonal P 3c1 high-T structure of LaF3 ‘‘tysonite’’
and Z ¼ 6. The light rare earths fluorides from La to Nd
crystallize in this structure. All other rare earth fluorides
crystallize at room temperature in the orthorhombic
structure determined by Zalkin and Templeton [7] for
YF3, also referred as b-YF3, space group P nma and Z ¼ 4.
type mentioned above. For the small rare earths from Er to
Lu upon heating, the b-YF3 structure changes to a
hexagonal not well identified a-YF3 structure. This
structure was considered tentatively isostructural with the
trigonal a-UO3 structure by Sobolev et al. [10,11],
3
¯
belonging to P3m1 (D3d) space group, Z ¼ 1 [12]. Never-
Ã
Corresponding author. Fax: +49 30 6392 3003.
E-mail addresses: iranieri@ipen.br (I.M. Ranieri),
baldochi@ipen.br (S.L. Baldochi), klimm@ikz-berlin.de (D. Klimm).
theless, the a-UO3 structure is not yet well understood too.
Either superstructures or a lower (e.g. orthorhombic)
0022-4596/$ - see front matter r 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jssc.2008.02.017

ARTICLE IN PRESS
I.M. Ranieri et al. / Journal of Solid State Chemistry 181 (2008) 1070–1074
1071
symmetry were discussed, as single crystal show a biaxial
optical interference figure [13,14].
fluorination process. These samples were used in the DSC
measurements.
Jones and Shand [15] proved that it was possible to grow
crystals of the four tysonites, LaF3, CeF3, PrF3 and NdF3,
but only the orthorhombic DyF3 and HoF3 using CdF2 as
scavenger. After Garton [16] and Pastor [17] it was
established that GdF3, TbF3, DyF3 and HoF3 crystals
can be grown from oxygen free compounds and in a
reactive atmosphere, confirming the thermodynamic stu-
dies by Spedding et al. [18].
Thermoanalytic measurements were performed with a
NETZSCH STA 409CD with DSC/TG sample carrier
(thermocouples type S). The sample carrier was calibrated
for T and sensitivity at the phase transformation points of
BaCO3 and at the melting points of Zn and Au. Sample
powders (50–70 mg) were placed in graphite DSC crucibles
with lid. As graphite is not wetted by the molten fluorides,
the melt forms one single almost spherical drop (diameter
d ꢁ 3 mm) that could be used for the subsequent X-ray
phase analysis.
For the intermediate SmF3, EuF3 and GdF3 two sub-
ð1Þ
ð2Þ
¯
sequent transformations P nma ! P 3c1 ! P 63=mmc
were discussed by Greis [19] (the number in brackets
indicates the order of the phase transformation (PT)).
These PT were inferred from electron diffraction experi-
ments with LaF3, where small synthesized crystals pre-
sented also sub-cells reflections as observed by Schlyter [3]
and Maximov [20]. Stankus [21] claimed that at high T the
The vapor pressures of both fluorides at their melting
points are high. Fortunately, the evaporation rate for pure
GdF3 or LuF3, respectively, was found to be almost
identical. Thus it can be assumed that the partial
evaporation does not lead to a considerable concentration
shift. The inhomogeneous powder samples were homo-
genized in a first heating/cooling cycle with ꢂ40 K=min.
Here the heating was performed to that Tmax where the
DSC melting peak was just finished and the molten sample
could homogenize. Depending on x, this was the case for
1145 ꢃCpTmaxp1280 ꢃC and due to the large heating rate
the mass loss did never exceed 4% in these preliminary
mixing cycles. Without intermediate opening of the
apparatus, the mixing cycle was followed by a measuring
run with a heating rate of 10 K/min. Although the crucibles
were covered by lids, the evaporation of ꢁ10–15% sample
mass during this DSC/TG run cannot be avoided. Cooling
curves showed often supercooling and were not used for
the construction of the phase diagram. In total, 14 different
compositions ranging from pure GdF3 ðx ¼ 0Þ to pure
LuF3 ðx ¼ 1Þ were measured.
¯
b-YF3 type ðP nmaÞ and the LaF3 type (tysonite, P 3c1)
become practically identical. Sobolev et al. [22] constructed
the phase diagrams of the systems GdF3–LnF3 (Ln ¼ Tb,
Ho, Er, Yb), solid solutions regions were proposed without
phase transitions in all systems, when the cation mean ionic
radius was between that of the Tb3þ and Er3þ
.
Recently, the present authors have published a phase
diagram study of the system GdF3–YF3 [23]. It was
established that both components undergo a solid-state
phase transformation of first order before melting.
Additionally, a l-shaped maximum of cpðTÞ being char-
acteristic for a second order transformation was found
for GdF3. Both low-T and high-T phases exhibit un-
limited mutual solubility. This observation raises the
question, whether the high-T structures of YF3 (reported
Other samples were melted under a flux of hydrogen
fluoride gas, then pulverized to be analyzed by powder
X-ray diffraction, using a Bruker AXS diffractometer,
model D8 Advance, operated at 40 kV and 30 mA, in the 2y
range of 22.5–68.51. The diffraction patterns where treated
with the Rietveld Method [26] using the GSAS program to
calculate the lattice parameters [27].
¯
as P 3m1) and GdF3 (reported as P 63=mmc) may really be
different [10].
In the current paper, the phase diagram of the system
GdF3–LuF3 is reported for the first time. The interest in
this phase diagram derived from DTA studies regarding
the phase diagram of the system LiF2Gd1ꢀxLuxF3, which
is interesting to develop new solid-solution crystals of the
type LiGd1ꢀxLuxF4 to be used as laser host.
3. Results and discussion
2. Experimental
It turned out that GdF3 as well as LuF3 showed similar
DSC heating curves: A first endothermal peak due to the
first order PT (T
tively) is followed by a second endothermal peak due to
GdF3
PT
Mixtures of Gd1ꢀxLuxF3 with x ¼ 0:2; 0:4; 0:6 and
0:8, respectively, were prepared using commercial LuF3
(AC Materials, 6N purity) and GdF3 synthesized from
commercial Gd2O3 powder (Alfa, 5N purity) by hydro-
fluorination. The oxide was placed in a platinum boat
inside a platinum tube, and slowly heated in a stream of
argon gas (White Martins, purity 99.995%) and HF gas
(Matheson Products, purity 99.99%) up to 850 1C. This
process is described in detail elsewhere [24,25]. Conversion
rates 499:9% of the theoretical value calculated for the
reaction Gd2O3 þ 6HFꢀ! 2GdF3 þ 3H2O were measured
by comparing the masses prior to and after the hydro-
¼ 902 ꢃC or TPLTuF3 ¼ 946 ꢃC, respec-
LuF3
melting (TGf dF ¼ 1252 ꢃC or Tf
¼ 1182 ꢃC, respec-
3
tively). The values for GdF3 were measured and compared
with literature data recently [23], and for lutetium fluoride
LuF3
PT
one finds values 943pT
p963 and 1180pTLf uF3 p1188
(T in 1C) in the literature [9,19,21]. Only Jones and Sand
LuF3
PT
[15] reported the very low value T
ꢁ 874 ꢃC. Like in the
recent study [23] weak l shaped bends were found in the
DSC curves of GdF3 rich mixtures with xo0:2. This could
be an indication for a second order transformation between
TPT and Tf but will not be discussed here.

ARTICLE IN PRESS
I.M. Ranieri et al. / Journal of Solid State Chemistry 181 (2008) 1070–1074
1072
0.0
-1.0
-2.0
x
x
exo
LuF3
x
x
Gd0.3Lu0.7F3
Gd0.5Lu0.5F3
x
Gd0.7Lu0.3F3
GdF3
1051°C
1092°C
900
950
1000
1050
1100
1150
1200
Temperature (°C)
25
30
35
40
45
50
55
60
65
Fig. 1. DSC curves (2nd heating with 10 K/min) of LuF3 (x ¼ 1:0) and of
2ꢀ (deg)
four mixtures Gd1ꢀxLuxF3 with 0:4pxp0:8. TPT ¼ 1051 ꢃC and Teut
¼
1092 ꢃC are the two isothermal phase boundaries in Fig. 4.
Fig. 2. Diffractions patterns of some Gd1ꢀxLuxF3 samples utilized to
determine the lattice parameters at room temperature.
Fig. 1 shows DSC heating curves for five compositions
starting from pure LuF3 (x ¼ 1:0) to Gd0:6Lu0:4F3. The PT
peak for x ¼ 1:0 is disturbed by a shoulder on the low T
side that is assumed to result from some contamination of
the sample, as no additional peak that could be related to
the presence of oxyfluorides was detected on cooling. With
increasing GdF3 content the melting peak shifts to lower T
7.00
6.50
until it reaches for x ꢁ 0:6 the constant value Teut
¼
1092 ꢃC. On the contrary, the PT peak shifts with
increasing GdF3 content to higher T and reaches for x ꢁ
0:5 the constant value TPT ¼ 1051 ꢃC. The same TPT and
Teut are reached if one starts with pure GdF3 and adds
LuF3.
b0
6.00
4.50
a0
c0
4.40
The lattice constants of nine samples with initial LuF3
concentrations 0px0p1 were measured by X-ray diffrac-
tion. All diffraction patterns were fitted considering the
orthorhombic structure with the P nma space group, no
parasitic peaks that would indicate the presence of other
phases were observed at room temperature (Fig. 2). The
experimental errors fDa0; Db0; Dc0g were always p5 ꢄ
10
5
10ꢀ4 A but were found to be maximum in the region
˚
0
0.0
0.2
0.4
0.6
0.8
1.0
0:2px0p0:4. Fig. 3 shows the total error s ¼ Da0 þ Db0 þ
Dc0 versus the LuF3 concentration x0 of the sample,
together with a0; b0; c0.
x0
Fig. 3. Top: lattice constants for the P nma phase (b-YF3 type) of
Gd1ꢀxLuxF3 solid solutions at room temperature. Bottom: experimental
error s ¼ Da0 þ Db0 þ Dc0.
Table 1 reports the results for calculated values of a0, b0
and c0, the volume of the unit cell V, the mass density R and
the mean rare-earth ionic radius in the solid solution
½8ꢅ
½8ꢅ
½8ꢅ
½8ꢅ
r ¼ ð1 ꢀ x Þr þ x0 r . r and rLu are the ionic radii of
fluorides due to the lanthanides contraction, there is a
˚
linear decreasing in the c0 parameter value from 4.40 A
¯
0
Lu
Gd
Gd
Gd3þ and Lu3þ, respectively, with coordination number 8
as determined by Shannon et al. [28].
˚
(SmF3) down to 4.376 A ðDyF3) and then a nearly
˚
The lattice parameters a0 and b0, drop linearly with x0
and c0, is almost constant for x0p0:4 and starts then to rise
slightly up to the value of pure LuF3 (Fig. 3). The smooth
variation of a0, b0 and c0 over the whole concentration
range is obviously the result of unlimited mutual solubility
of GdF3 and LuF3 in their low-T phases. The variation of
exponential increase up to 4.467 A (LuF3) [7]. Taking into
account the ionic radius of Ho3þ, namely 1.015 A (with a
˚
coordination number of 8) [28], and the mean ionic radius
of the composition Gd0:5Lu0:5F3, one has the same value.
The DSC and XRD results are summarized in Fig. 4
where full circles represent experimental DSC points that
could be well determined by extrapolated onsets of sharp
peaks. Such points represent the lower boundary of PT
~
c0 in this system is very similar to the c axis variation of the
pure rare earth fluorides. In the orthorhombic rare-earth

ARTICLE IN PRESS
I.M. Ranieri et al. / Journal of Solid State Chemistry 181 (2008) 1070–1074
1073
Table 1
Lattice parameters for the P nma phase of Gd1ꢀxLuxF3 ð0pxp1Þ
R ðg=cm3Þ
˚
r (A)
˚
a0 (A)
˚
b0 (A)
˚
c0 (A)
3
˚
Lu content (x,
molar)
¯
V ðA Þ
0.00
0.00
0.20
0.25
0.30
0.40
0.50
0.60
0.70
1.00
1.00
6.571a
6.984a
4.390a
201.47a
7.063a
7.045
7.157
7.331
7.359
7.588
7.667
7.773
7.967
8.283
8.291b
1.053
1.038
1.034
1.030
1.023
1.015
1.007
1.000
0.977
6.5758(2)
6.5143(6)
6.4908(4)
6.4671(5)
6.4102(3)
6.3838(2)
6.3421(2)
6.2855(2)
6.1481(2)
6.150b
6.9898(2)
6.9517(5)
6.9429(3)
6.9209(4)
6.8867(2)
6.8770(2)
6.8615(2)
6.8222(1)
6.7640(2)
6.760b
4.3947(2)
4.3907(3)
4.3964(2)
4.3918(2)
4.3885(1)
4.4025(1)
4.4154(2)
4.4140(1)
4.4732(1)
4.468b
201.999(11)
198.833(31)
198.125(18)
196.569(28)
193.733(13)
193.275(8)
192.143(11)
189.282(10)
186.021(11)
185.83b
aPDF 49-1804.
bPDF 32-0612.
and TPT both constituents are mutually saturated. In the
b-phase both components exhibit unlimited mutual solu-
bility. The phase transformation occurring at TPT for
medium concentrations x
tysonite þ a ꢀ YF3$b ꢀ YF3
is a peritectoid reaction.
(1)
Near the left or right rims of the phase diagram, single
phase Lu-doped GdF3 (tysonite structure) as well as single
phase Gd-doped LuF3 ( ¼ a-Gd:LuF3) transform upon
cooling to b-(Gd,Lu)F3, crossing a 2-phase field where
a- and b-phases are mixed and the upper and lower limit of
this 2-phase field depend on x. Both 2-phase fields
‘‘tysonite + liq.’’ and ‘‘tysonite + b-(Gd,Lu)F3’’ are
broader than the corresponding fields on the LuF3 rich side
where tysonite is replaced by a-Gd:LuF3. The larger widths
of the 2-phase fields on the GdF3 rich side may result in
stronger phase segregation of such compositions, resulting
after cooling to room temperature in microscopic grains
with a wider variation of compositions. As different
compositions of the Gd1ꢀxLuxF3 grains result in different
lattice constants, the larger experimental error s that is
observed for Gd-rich mixtures (but not for pure GdF3) can
be explained by variations of x resulting from segregation.
Fig. 4. Phase diagram of GdF3–LuF3. ꢆ—sharp peaks. ꢃ— small peaks or
from offsets. a; b mean high-T phase or low-T phase, respectively.
regions or, in the case of melting, the solidus line. The
higher boundary of PT regions or the liquidus line,
respectively, are less remarkable as here the PT process
or melting just finishes and the DSC curve returns to the
basis line. Such more vague determined events are marked
by hollow circles in Fig. 4.
Two horizontal lines at 1051 ꢂ 2 and 1092 ꢂ 2 ꢃC are the
most prominent feature of the experimental phase diagram
Fig. 4. Obviously, two different processes with strong
thermal effects start independent on composition x for a
4. Conclusion
Solid solutions Gd1ꢀxLuxF3 are formed over the whole
concentration range 0pxp1 and crystallize in the same
b-YF3 type structure with space group P nma like the pure
components GdF3 and LuF3. Upon heating GdF3 under-
goes a first order phase transformation to a tysonite type
structure whereas LuF3 undergoes a first order transforma-
tion to a a-YF3 type structure that is not isomorphous to
tysonite. As a result, a miscibility gap separates the high-T
phases of GdF3 and LuF3. For intermediate compositions
0:2txt0:6 the phase transformation from single phase
Gd1ꢀxLuxF3 to a mixture of the high-T phases of GdF3
broad range 0:2txt0:5 . . . 0:6 always at the same TPT
¼
1051 ꢃC or Teut ¼ 1092 ꢃC, respectively. The components
GdF3 and LuF3 show complete miscibility in the liquid
phase. If the melt is cooled, either Gd-saturated LuF3
¯
(a-YF3 type, P 3m1) or Lu-saturated GdF3 (tysonite type,
¯
P 63=mmc or P 3c1) crystallize from the melt. Both high-T
phases belong to different space groups and unlimited
miscibility is therefore prohibited—instead a eutectic is
formed below 1092 ꢃC. In the eutectic mixture between Teut

ARTICLE IN PRESS
I.M. Ranieri et al. / Journal of Solid State Chemistry 181 (2008) 1070–1074
1074
(saturated with Lu) and LuF3 (saturated with Gd) is a
peritectoid decomposition.
[7] A. Zalkin, D.H. Templeton, J. Am. Chem. Soc. 75 (1953) 2453–2458.
[8] L.H. Ahrens, Geochim. Cosmochim. Acta 2 (1952) 155–169.
[9] R.E. Thoma, G.D. Brunton, Inorg. Chem. 5 (1966) 1937–1939.
[10] B.P. Sobolev, P.P. Fedorov, Sov. Phys. Crystallogr. 18 (1973) 392.
[11] B.P. Sobolev, P.P. Fedorov, D.B. Shteynberg, B.V. Sinitsyn,
G.S. Shakhkalamian, J. Solid State Chem. 17 (1976) 191–199.
[12] W.H. Zachariasen, Acta Crystallogr. 1 (1948) 256–268.
[13] C. Greaves, B.E.F. Fender, Acta Crystallogr. B 28 (1972) 3609–3614.
[14] S. Siegel, H.R. Hoekstra, Inorg. Nucl. Chem. Lett. 7 (1971) 497–504.
[15] D.A. Jones, W.A. Shand, J. Crystal Growth 2 (1968) 361–368.
[16] G. Garton, P.J. Walker, Mater. Res. Bull. 13 (1978) 129–133.
[17] R.C. Pastor, M. Robinson, Mater. Res. Bull. 9 (1974) 569–578.
[18] F.H. Spedding, B.J. Beaudry, D.C. Henderson, J. Moorman,
J. Chem. Phys. 60 (1974) 1578–1588.
The behavior of the system GdF3–LuF3 is somewhat
uncommon: Usually the mutual solubility of components
rises with T, but in the present case both components show
unlimited miscibility in the low-T phase only. It should be
noted that Nafziger et al. [29] obtained similar results
(constant eutectic and phase transformation temperatures
Teut ¼ 1073 ꢃC and TPT ¼ 1045 ꢃC, respectively, over an
extended composition range \50%) with DTA measure-
ments in the LaF3–YF3 system. Nafziger et al. speculated
that this is an indication for a low-T intermediate phase,
but the present authors think that a peritectoid decom-
position (1) as shown in Fig. 4 of this work would be a
more simple explanation and that the postulation of an
(otherwise not proven) intermediate phase is not necessary.
[19] O. Greis, M.S.R. Cader, Thermochim. Acta 87 (1985) 145–150.
[20] B. Maximov, H. Schulz, Acta Crystallogr. B 41 (1985) 88–91.
[21] S.V. Stankus, R.A. Khairulin, K.M. Lyapunov, High Temp. High
Pressures 32 (2000) 467–472.
[22] D.F. Sobolev, V.S. Sidorov, P.P. Fedorov, D.D. Ikrami, Sov. Phys.
Crystallogr. 22 (1977) 574–577.
Acknowledgments
[23] D. Klimm, I.M. Ranieri, R. Bertram, S.L. Baldochi, Mater. Res.
Bull. 42 (2008) 676–681.
[24] H. Guggenheim, J. Appl. Phys. 34 (1963) 2482–2485, doi:10.1063/
1.1702768, see also: R. Bougon, J. Ehretsmann, J. Portier, A.
Tressaud, in: P. Hagenmuller (Ed.), Preparative Methods in Solid
State Chemistry, Academic Press, New York, 1972, p. 401.
The authors acknowledge financial support from
CAPES (Grant no. 246/06) and DAAD (Grant no. D/05/
30364) in the framework of the PROBRAL program.
[25] I.M. Ranieri, Growth of LiY1ꢀxTRxF4 : Nd (TR
crystals for optical applications, Ph.D. Thesis, IPEN/University of
Sao Paulo, 2001.
¼
Lu or Gd)
References
˜
[26] H.M. Rietveld, J. Appl. Crystallogr. 2 (1969) 65–71.
[27] A.C. Larson, R.B.V. Dreele, General structure analysis system
(GSAS), Technical Report LAUR 86-748, Los Alamos National
Laboratory, 2007.
[1] I. Oftedal, Z. Phys. Chem. B 5 (1929) 272–291.
[2] I. Oftedal, Z. Phys. Chem. 13 (1931) 190–200.
[3] K. Schlyter, Ark. Kemi 5 (1953) 73–82.
[4] M. Mansmann, Z. Anorg. Allg. Chem. 331 (1964) 98–101.
[5] M. Mansmann, Z. Kristallogr. 122 (1965) 375–398.
[6] A. Zalkin, D.H. Templeton, T.E. Hopkins, Inorg. Chem. 5 (1966)
1466–1468.
[28] R.D. Shannon, Acta Crystallogr. A 32 (1976) 751–767.
[29] R.H. Nafziger, R.L. Lincoln, N. Riazance, J. Inorg. Nucl. Chem. 35
(1973) 421–426.