Full text of DOI:10.1016/0022-0248(68)90114-0

146
Journal of Crystal Growth 3, 4 (1968) 146—149 C North-Holland Publishing Co., Amsterdam
HIGH TEMPERATURE VAPOUR PHASE GROWTH OF
EUROPIUM SULPHIDE, SELENIDE AND TELLUIUDE
E. KALDIS
Laboratoriumfür Festkörperphysik,
Eidgenossische Technische Hochschule,
Zurich, Switzerland
Preliminary results are reported for the synthesis, crystal growth
and characterisation of EuTe, EuSe and EuS. The crystals were
grown by chemical transport (EuTe, EuSe) and sublimation
(EuTe, EuS) in evacuated and sealed molybdenum crucibles. The
temperature of the iodine transport was 1700 °C and that of
sublimation 2050 °C .Reaction between iodine and molybdenum
was avoided by working at temperatures where the molybdenum
iodides are not stable (T > 1500 °C ).Optical measurements show
that the absorption edge of the iodine grown crystals agrees with
the values given in the literature. The rest absorption and the
refractive index of EuTe and EuSe is lower for transported crys-
tals and the lowest reported up to now in the literature. The
magnetic properties of EuTe grown by iodine transport have been
measured and found comparable to the ones reported in the
literature.
The compounds have been synthesised by direct re-
action of the elements in evacuated and sealed silica
1. Introduction
Physical and chemical investigations on high temper- ampoules. The reaction with tellurium begins at much
ature materials have been seriously hindered up to now higher temperatures and is not completed even by heat-
due to the difficulties of growing crystals at high tern- ing for many days at 880 °C , as can be seen by the
peratures. High melting points, non-compatibility of faint yellow vapour in the arnpoule (table 1). Heating
the wall materials, high degree of imperfection and at higher temperatures in silica ampoules is not possible
non-stoichiometry are some difficulties frequently met because of the reaction with silica.
when high temperature crystal growth is tried. How-
The reaction of europium with sulfur is almost corn-
ever, it could be expected that if vapor phase methods pleted between 90 and 180 ° Cin 48 hours. The temper-
in closed systems and particularly chemical transport ature of 90 ° Ciscritical. When it is reached, the heating
could be used, some ofthese difficulties could be solved, rate must be slowed down to 2 °C /hrin order to avoid
We have used the sealed molybdenum crucible explosions. To ensure complete reaction, the ampoule
method’), which has been successfully used up to now is heated for one day at 880 °C .At this temperature no
in our laboratory2), for crystal growth both by chemical reaction with the silica wall can be seen.
transport and sublimation. Single crystals of EuTe and
The contents of the silica ampoules were used as
EuSe havebeen grown by chemical transport at 1700 ° C starting material for sublimation experiments. The
and those of EuTe and EuS by sublimation at 2000 °C . powder was transferred into a degassed molybdenum
In the following the synthesis, crystal growth and char-
acterisation of these compounds are reported.
TABLE 1
Characteristic temperatures of the synthesis of europium chalco-
genides
2. Synthesis and sublimation of EuTe and EnS
The difficulties of synthesising bivalent europium
Temperature of Temperature of Max heating
compounds arise from the high reactivity of europium
metal and the sensitivity of Eu2 + to oxidation. The
handling, therefore, of the europium metal and the
compounds in powder form took place in an argon
filled dry box. A 30% Na—70% K liquid alloy has been EuSe+I2
used as getter.
reaction of Eu
reaction of
temperature
(C )
with 12
the elements
(°C )
(°C )
+12
175-270
175—210
—
600—730
185—210
90—180
880
750
880
EuS
II
—
8

H IG H TEMPERATURE VAPOUR PHASE GROWTH
147
C
culated3). Though they are different from the actual
undercoolings in the crucible, they can be used in order
N.
to increase the reproducibility of the experiments.
<CI)
3. Synthesis and chemical transport of EnTe and EuSe
For chemical transport the same molybdenum cruci-
—
~
bles were used. To avoid any reaction of the iodine
with the crucible walls the experiments were performed
at temperatures where the molybdenum iodides are not
—
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
stable4) (T> 1500 °C).The high temperatures during
the sealing of the crucible make difficult the selection
of the transporting agent. Even with intensive cooling,
the temperature of the charge reaches a few hundred
degrees Celsius. Thistemperature makes impossible the
use of most conventional transporting agents, under
high vacuum, due to their high vapour pressure. There-
f
=
filament
crucible
c
—
fore, to keep high vacuum during the sealing of the
crucibles, Eu12 has been used as a source of iodine and
the synthesis procedure was modified. Stoichiometrical
I[
amounts of the elements together with the necessary
quantity of iodine were sealed and heated in the am -
H2o
poule, where the following reaction takes place:
Fig. 1. Principle ofthe apparatus used for sealing molybdenum
crucibles under high vacuum.
Eu + ~ X2 + n ~
(1—n) EuX + n Eu12 + 4n X2
(X Te,Se).
=
crucible (30 mm diameter and 120 mm length) and the Details about the way to introduce the iodine in the
crucible was put in the sealing apparatus. To seal the evacuated silica ampoule have been given earlier3).
molybdenum crucibles, any commercial electron beam
As it can be seen from table 1 for EuSe, the reaction
welding apparatus can be used. In our work, we have of europium with iodine and selenium takes place in
used a simple apparatus based on the same principle the same temperature range (175—210 °C).To ensure
(fig. I). A dc voltage of 2000 V is applied between the completion of the reaction, the ampoule is heated at
hot filament and the crucible. Part of the emitted elec- 750 °C .However, the temperature of the reaction of
trons heat the thin edge of the lid and seal it with the europium with tellurium is not influenced by the pre-
crucible. To avoid decomposition of the starting mate- sence of iodine.
na!, effective cooling must be used. Before the sealing
The contents of the silica ampoules were transferred,
starts, the vacuum in the sealing chamber is l0 6 Torr; under argon atmosphere, in the molybdenum crucible,
during the sealing, the vacuum falls to l 0 ~ Torr. At evacuated and sealed as above.
pressures higher than 10 ~ Torr, discharges in the
chamber make sealing impossible.
It was assumed that chemical transport will take
place according to the equilibrium
The sealed crucible is put in a high temperature fur-
nace for the crystal growth. We have used both HF-
high vacuum, silica double-walled, water cooled fur-
EuX + ‘2
~— Eul
+ 4X2
“X Te S ’ ~
—
‘
naces and resistance heated, high vacuum, graphite The EuI2 and chalcogen necessary to start the trans-
furnaces. In these furnaces temperature profiles have port are provided by the synthesis reaction.
been measured with tungsten—rhenium (97% W—3 %
A problem of the synthesis of the bivalent europium
Re versus 75 % W—25 % Re) thermocouples. From chalcogenides has been the Eu3 + content. Investiga-
these profiles apparent undercoolings, AT, can be cal- tions with Mossbauer-effect5 ) have shown that some of
II
—
8

148
F. KALDIS
the first synthesised powders of EuTe6) contained up
to 25% Eu3 ~• With recrystallisation at higher temper-
atures, it was possible to suppress the Eu3 + content.
çio3
An advantage of the use of equilibrium (1) for the
crystal growth of bivalent europium compounds is that
in the gas phase only Eu2~is stable. Eu13 decomposes
at temperatures lower than 1000 °C .Any Eu3 + content
of crystals grown under these conditions would prob-
ably result from a secondary reaction of defects or
from subsequent oxidation.
~
~
10
-
-
EuTe subL
2.684
247
-
-
~
~
EuSe trans
.
(1) 10
-
EuTe transp. 2.654
-
4. Results of the crystal growth
0.5
10
1.5
W AVELENGTH
2.0
2.5
At this first stage of our investigations, with which
this paper is dealing, it was tried primarily to improve,
( ii
Fig. 2. Optical absorption as a function of the wave length for
as far as possible, the quality of the crystals and not EuTe (iodine transported and sublimated), and EuSe (iodine
transported); n refractive index.
=
their size. The crystals grew as rather thin layers (2 to
3 mm thick) on the lid of the crucible because for the (1580 °C),in order to have high vapor pressure of the
first experiments only small amounts of starting mate- transporting agent in the gas phase.
rials were used. It is clear that if greater amounts of The EuTe crystals are wine red and transparent. Micro-
starting material are used, the crystals can be much scopic investigation with polarised light showed high
thicker. The conditions of the crystal growth experi- homogeneity and almost no stresses. This is partly the
ments together with some properties of the crystals are result of the lower growth temperatures which are only
summarised in table 2.
possible with chemical transport.
d) Iodine transport of EuSe: Microscopicinvestigations
showed some inclusions of another phase. X-ray dif-
TABLE 2
Crystal growth data for europium chalcogenides
fraction photographs showed this phase to be molyb-
denum.
Dimensions
f h
(°C ) (°C ) (°C ) (mg/cm3) (mg/h) °tals ~YS
Ti
T2
iT
12
5. Physical properties of the crystals
EuTe+12
EuTe
EuSe+12
EuS
1722 1627
2000 1857 143
1687 1619
2050 1950
95
1.35
—
18.6
14.7
9.3
9 >< 8 >< 3
5 x4x2
4x4x3
2x2x2
For the characterisation of the crystals, optical and
magnetic measurements were performed. Fig. 2 shows
the optical absorption as a function of the wave length
68
96
1.0
—
16
for single crystals of EuTe both transported and subli-
mated and of transported EuSe. The absorption edges
a) Sublimation o f EuTe: The crystals were a dark shown by these crystals are the same as the ones given
red colour and to some degree transparent. Micro- in the literature7’8’9)• Further they show no appreciable
scopic examination showed that the crystals had small free carrier absorption beyond the absorption edge.
dark inclusions and mechanical stresses.
The rest absorption and refractive index of transported
b) Sublimation of EuS: The colour of the crystals was EuTe are much lower than the one of the sublimated
green. The crystals are not transparent because the ab- and even lower than any other value mentioned in the
sorption edge is at 1.645 eV. Microscopic examination literature’ O)• The rest absorption of transported EuSe
showed that the crystals are homogeneous but have is also much lower than the value previously given in
stresses.
the literature8). Therefore, it is clear that the trans-
c) Iodine transport ofEuTe: The transport temperatures ported crystals are optically pure. A thorough investi-
were chosen to fulfil the following conditions: 1) higher gation of the refractive indexes as criteria for the purity
than 1550 °C ,in order to avoid reaction of iodine with of these crystals will be given elsewhere’ 0 ).
molybdenum, 2) higher than the boiling point of Eu12
Three types of magnetic measurements have been
II
—
15

HIGH TEMPERATURE VAPOUR PHASE GROWTH
149
performed with iodine transported EuTe crystals. The perimental confirmation of Fisher’s relation’2 ) which
paramagnetic susceptibility was measured in a field of states that the peak ofthe specific heat should be located
about 13 kOe between 100 and 300 °K . The effective at the same temperature as the maximum slope of the
number of Bohr magnetons performula unit was found susceptibility.
to be 7.65. This is compared to the theoretical value
g(J(J+ l))~
=
7.94forthefreeEu24 ionwitbj
=
s ~ 6. Conclusions
Using the method of Busch et a!.”), the induced
Crystal growth of EuTe and EuSe by iodine trans-
ferromagnetic moment of the sample has been deter- port in a closed system can take place at 1700 ° Cusing
mined in a pulsed field. Both experiments show that the sealed molybdenum crucibles. A comparison of the
number of the effective Bohr magnetons is lower than physical properties of the transported crystals with
the theoretical, as it was also previously found6). One values in the literature shows that the transported crys-
possible reason is that the crystals used for these mea- tals have better optical properties and similar magnetic
surements had some condensed Eu12 on their surface ones.
which hydrolyses in the humidity of the ambient atmo-
The high temperature transport method described is
sphere to the hygroscopic EuI3. In this case, the for- probably not limited to these compounds but can be
mula EuTe used for the calculations was not correct, used also for other systems at temperatures up to
Another explanation would be the presence of euro- 2000 ° C and higher.
pium, due to the partial decomposition of Eu12 or the
incomplete reaction of europium with tellurium.
Under the assumption that part of the europium
Acknowledgement
The author is deeply indebted to the Director of the
ions is present as Eu3 ~ it can be calculated from these Laboratonium für Festkorperphysik, Prof. G. Busch,
experiments that the amount of Eu3 + is 3—5 %. who suggested and supported this investigation. He is
To determine the ordering temperature, the initial also indebted to Dr. 0. Vogt for discussions about the
susceptibility Xo was measured in an alternating mag- construction of apparatuses. Many thanks are due to
netic field of about 10 Oe and 21 c/s. At this smallfield Mr. P. Schwob for the magnetic and to Mr. R. Verrault
any increase of the Néel point due to induced order- for the optical measurements. It is also a pleasure to
ing is avoided. The susceptibility versus temperature acknowledge the skilled technical assistance of Mr. G.
curve (fig. 3) shows a maximum of d~o/dTat T
=
Sampietrowhoperfomedthecrystalgrowthexperiments.
9.58 ± 0.1 °K .This result represents a very good ex-
References
1) M. W. Shafer, J. AppI. Phys. 36 (1965) 1145.
20
2) P. Schwob and 0. Vogt, Phys. Letters 22 (1966) 374;
G. Busch, J. AppI. Phys. 38 (1967) 1386.
3) E. Kaldis and R. Widmer, J. Phys. Chem. Solids 26 (1965)
1697.
~0~
1 5
> ‘ 10
5
0
4) R. F. Roisten, Iodide Metals and Metal Jodides (Wiley, New
York, 1961).
0
0
5) P. Brinx, S. Hüffner, P. Kienle and D. Quitman, Phys. Letters
0
,
13 (1964) 140.
6) G. Busch, P. Junod and 0. Vogt, in: Colloque International
du C.N.R.S. Ed. R. Kern, Sept. 1965 (C.N.R.S., Paris).
7) G. Busch, P. Junod and P. Wachter, Phys. Letters 12 (1964)
11.
T0 9.6 ~ 0.1 K
H0,= 10 Oe
8) B. E. Argyle, J. C. Suits and M. J. Freiser, Phys. Rev. Letters
15 (1965) 822.
9) F. Holtzberg, T. R. McGuire and S. Methfessel, J. Appi.
Phys. 37 (1966) 976.
10) G. Busch and R. Verrault, to be published.
I
0
5
10
15
2I0
T(K )
11) G. Busch, P. Junod, P. Schwob, 0. Vogt and F. Hulliger,
Fig. 3. Initial susceptibility versus temperature curve for iodine
transported EuTe.
Phys. Letters 9 (1964) 7.
12) M. E. Fisher, Phil. Mag. 7 (1962) 1731.
11—8