Thermal decomposition of Ti5(Se,Te)8 in argon and nitrogen atmospheres

Article abstract of DOI:10.1007/s11176-005-0361-4

Chalcogenides TiSe_1.60-x Te_x (0 ≤.x ≤ 1.60), forming a continuous series of hexagonal solid solutions, were prepared by the direct ampule procedure. The thermal decomposition of TiSe_1.60-x Te_x was studied for the samples with x = 0, 0.16, 0.80, 1.44, and 1.60 in Ar and N₂ atmospheres in the course of heating from 25 to 1000°C. The selenide undergoes no weight loss under Ar, in contrast to the telluride which disproportionates and loses weight owing to the formation of volatile TiTe₂. At high temperatures, tellurides are more sensitive than selenides to the presence of nitrogen: The disproportionation is accompanied by the reaction of TiTe₂ with N₂, yielding low-volatile titanium nitride and free tellurium. Titanium selenide and telluride as components of the solid solutions behave similarly to the corresponding individual chalcogenides. 2005 Pleiades Publishing, Inc.

Full text of DOI:10.1007/s11176-005-0361-4

Russian Journal of General Chemistry, Vol. 75, No. 7, 2005, pp. 1005 1009. Translated from Zhurnal Obshchei Khimii, Vol. 75, No. 7, 2005,
pp. 1063 1067.
Original Russian Text Copyright
2005 by Pankratova, Zvinchuk, Suvorov, Hatanpaa, Kozlov, Leskela.
Thermal Decomposition of Ti5(Se, Te)8
in Argon and Nitrogen Atmospheres
O. Yu. Pankratova*, R. A. Zvinchuk*, A. V. Suvorov*,
T. Hatanpaa**, V. I. Kozlov*, and M. Leskela**
* St. Petersburg State University, St. Petersburg, Russia
** University of Helsinki, Helsinki, Finland
Received July 19, 2004
Abstract Chalcogenides TiSe1.60 xTex (0
x
1.60), forming a continuous series of hexagonal solid
solutions, were prepared by the direct ampule procedure. The thermal decomposition of TiSe1.60 xTex was
studied for the samples with x = 0, 0.16, 0.80, 1.44, and 1.60 in Ar and N2 atmospheres in the course of
heating from 25 to 1000 C. The selenide undergoes no weight loss under Ar, in contrast to the telluride which
disproportionates and loses weight owing to the formation of volatile TiTe2. At high temperatures, tellurides
are more sensitive than selenides to the presence of nitrogen: The disproportionation is accompanied by the
reaction of TiTe2 with N2, yielding low-volatile titanium nitride and free tellurium. Titanium selenide and
telluride as components of the solid solutions behave similarly to the corresponding individual chalcogenides.
In the binary systems Ti Se and Ti Te, in the
interval from sesqui- to dichalcogenides, a series of
nonstoichiometric phases of the CdI2 structure are
formed: Ti2Ch3, Ti5Ch8, and TiCh2. Data on the
boundaries of the homogeneity, continuity, and dis-
creteness areas and on the structures of these phases,
mostly obtained by powder X-ray diffraction, are still
ambiguous and sometimes contradictory. The best
method for crystallization of nonstoichiometric chal-
cogenides is chemical gas transport. In particular,
Rimmington and Balchin [1] have grown series of
Ti(S,Se)2, Ti(S,Te)2, and Ti(Se,Te)2 single crystals
by iodide gas transport.
N2 atmospheres, which is particularly important when
performing chemical transport reactions for growing
nonstoichiometric transition metal chalcogenides [8].
According to X-ray phase analysis, TiSe1.60 xTex
forms a continuous series of hexagonal solid solutions
from TiSe1.60 to TiTe1.60. Typical diffraction patterns
of the samples with x = 0.16, 0.80, and 1.44 are
shown in Fig. 1.
The unit cell parameters a 3.845 and c 6.388 for
TiTe1.60 are consistent with those reported in [2]. The
powder X-ray diffraction data and unit cell parameters
for TiSe1.60 are given in the table. Only the monoclin-
In the Ti Te system, in the examined composition
range, formation of Ti5Te8 (TiTe1.60) was observed;
this compound was considered either as an individual
compound [2] or as the extreme member of the series
Ti2Te3 Ti5Te8 [3]. In the Ti Se system, Gronvold
and Langmyhr [4] and Bernusset [5] revealed mono-
clinic Ti5Se8 (TiSe1.60). Previously [6] we proved the
existence of a continuous hexagonal series of sele-
nides Ti2Se3 Ti5Se8, in which the hexagonal unit cell
parameters were the smallest for Ti5Se8 (TiSe1.60).
The fact that Ti5Se8 is an individual compound is also
supported by formation of its crystals (Laue class 3m,
superstructure 2a, 2c) in various versions of the gas
transport reaction, including that with TiSe2 as seleni-
um source [7].
Powder X-ray diffraction data for TiSe1.60 (a 3.754, c
5.984
)
I/I0, %
, deg
hkl
I/I0, %
, deg
hkl
57
2
7.45
14.42
14.95
16.25
21.00
22.72
25.58
27.35
30.15
31.00
31.00
34.10
001
100
002
101
102
003
110
103
112
201
004
202
<1
<1
5
19
<1
13
4
2
6
5
1
34.90
35.40
39.05
42.12
42.25
43.60
44.95
48.30
49.70
50.55
52.15
104
113
203
211
114
105
212
300
213
006
302
86
77
66
27
17
33
6
100
100
5
The goal of this study was to determine the homo-
geneity area of the Ti5(Se,Te)8 (TiSe1.60 xTex) solid
solution and to evaluate its thermal stability in Ar and
1070-3632/05/7507-1005 2005 Pleiades Publishing, Inc.

1006
PANKRATOVA et al.
4000
3000
2000
1000
0
3
2
1
10
20
30
40
50
60
70
2 , deg
Fig. 1. Powder X-ray diffraction patterns of TiSe
Te . x: (1) 0.16, (2) 0.80, and (3) 1.44.
1.60
x
x
were heated in an Ar or N2 atmosphere from 25 to
1000 C and then cooled to 100 C. The TG curves for
all the ten runs are shown in Fig. 3.
120
7
6
2
3
110
100
Data on the thermal dissociation of titanium sele-
nides and tellurides are mainly limited to higher com-
pounds TiChy with y 2. Suzuki and Waldbeck [9]
studied the Ti Te system by the Knudsen method and
found that titanium tellurides vaporize incongruently,
with the release of tellurium as the only gaseous prod-
uct. No such studies were made for the Ti Se system.
There are only indications [10] that thermal dissocia-
tion of higher titanium selenides (from sesqui- to di-
selenide) is a route to lower selenides.
5
4
1
90
80
3
2
1
70
60
0
0.16 0.48 0.80 1.12
1.44
x
Fig. 2. Hexagonal cell parameters (1) a and (2) c, and
(3) unit cell volume V of chalcogenides TiSe
as functions of x.
Te
x
Since the Ar atmosphere is inert, the weight loss
of the chalcogenide during the experiment can be as-
sumed to be exclusively due to its thermal dissocia-
tion.
1.60
x
ic structure was reported for this compound previously
[4, 5].
The selenide TiSe1.60 shows virtually no weight
loss in an Ar atmosphere, and the composition of the
selenide remains unchanged. The powder patterns of
the selenide before and after heating in an Ar atmos-
phere are identical.
Figure 2 shows how the hexagonal cell parameters
and unit cell volume of chalcogenides TiSe1.60 xTex
depend on the composition. The parameter a shows
a small negative deviation, and the parameter c, posi-
tive deviation from the linearity.
On the contrary, TiTe1.60 shows significant weight
loss, and its powder pattern changes. The thermal
dissociation of the chalcogenide can be described by
one of the two reactions. The first alternative is the
The thermal decomposition of TiSe1.60 xTex was
studied systematically in Ar and N2 atmospheres.
Chalcogenides with x = 0, 0.16, 0.80, 1.44, and 1.60
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 75 No. 7 2005

THERMAL DECOMPOSITION OF Ti5(Se, Te)8
dissociation into the gaseous chalcogen and chalco-
1007
genide enriched in the metal relative to the starting
compound:
2
1
100
90
4
3
TiTe1.60(solid)
TiTe1.60 z(solid) + z/2Te2(gas). (1)
5
In this case, the composition of the chalcogenide
should approach TiTe1.14, which contradicts the re-
sults of X-ray phase analysis, according to which the
composition of the chalcogenide formed is close to
6
7
80
8
70
TiTe1.50
.
Another alternative is the disproportionation into
a higher chalcogenide and a chalcogenide with a lower
chalcogen content than that in the starting compound.
Then, taking into account the composition of the re-
sulting chalcogenide determined by X-ray diffraction,
the following reaction can be assumed to occur in an
Ar atmosphere at 900 1000 C:
60
9
50
40
10
Heating
500
Cooling
1000
0
500
Temperature, C
TiTe1.60(solid)
0.8TiTe1.50(solid) + 0.2TiTe2(gas). (2)
Fig. 3. TG curves of TiSe
Te in Ar and N
1.60
x
x
2
Such a hypothesis is well consistent with the quan-
titative data. The weight loss in the course of this TG
run was 23.14%, and Eq. (2) implies 24%.
atmospheres: (1) TiSe
Te
/Ar, (2) TiSe /N ,
1.44 0.16
1.6
2
(3) TiSe
Te
/N , (4) TiSe /Ar, (5) TiSe
1.44 0.16 2 1.6 0.80
Te
/Ar, (6) TiSe
Te
/Ar, (7) TiTe /Ar,
0.80
0.16 1.44 1.6
Thus, the thermal decomposition of the tellurides
under the conditions of our experiment cardinally
differs from that observed in [9] under the effusion
conditions.
(8) TiSe
(10) TiTe
Te
/N , (9) TiSe
Te
/N , and
0.80 0.80
2
0.16 1.44 2
/N .
1.60
2
powder pattern (Fig. 4) indicate that the sample con-
tains an amorphous phase. This result indicates that
the reaction could be incomplete and that TiN could
also be amorphous. In the experiments involving
longer heating of TiTe1.60 under N2 (additional heat-
ing for 15 min at 950 C), the weight loss increased.
Our preliminary experiments showed that the nitro-
gen atmosphere is not inert toward the solid solutions
under consideration. Only the selenide TiSe1.60 did
not change its composition, as in the experiment
under Ar. For the compounds containing Te, thermal
dissociation (2) can be accompanied by the successive
reaction of the gaseous titanium dichalcogenide with
nitrogen, with the formation of low-volatile titanium
nitride and free tellurium:
Titanium selenide and telluride as components of
the solid solutions (x = 0.16, 0.80, 1.44) behave simi-
larly to the corresponding individual chalcogenides
(Fig. 3). For chalcogenides of the same composition,
the difference in the weight loss in different atmos-
pheres, determined by the occurrence of reactions (2)
and (3), grows with an increase in the tellurium con-
tent. In a nitrogen atmosphere, at a low Te content
(x = 0.16), reaction (2) is the major process, i.e., only
the thermal dissociation of the chalcogenide occurs.
At a higher Te content (x = 1.44), replacement of ni-
trogen for tellurium [reaction (3)] becomes the major
process. The diffraction patterns of TiSe0.16Te1.44
before and after heating under Ar and N2 are shown
as examples in Fig. 4.
TiTe2(gas) + 1/2N2(gas)
TiN(solid) + Te2(gas). (3)
The formation of TiN could be clearly seen in the
powder patterns of the selenotellurides with x 0.8
(Fig. 4).
The powder patterns of TiTe1.60 before and after
heating under N2 differ insignificantly, whereas the
weight loss in the course of the experiment is as large
as 55%. This fact additionally counts in favor of the
occurrence of reaction (2) rather than (1). The calcu-
lations show that the TiN : TiTe1.50 weight ratio in
the final mixture should be 2 : 3. Hence, 3/4 of the
TiTe1.60 decomposed under the conditions of the TG
experiment, and the resulting TiTe2 fully reacted with
N2. The diffuse reflections at small angles of the
Thus, TiSe1.60 xTex forms a continuous series
of hexagonal solid solutions with a CdI2 structure.
Nevertheless, titanium selenides and tellurides cardi-
nally differ in the thermal behavior in Ar and N2 at-
mospheres. On the whole, the selenides are considera-
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 75 No. 7 2005

1008
PANKRATOVA et al.
1000
800
600
400
3
TiN
TiN
TiN
2
200
0
1
CaF2
30
10
20
40
50
60
2 , deg
Fig. 4. Powder diffraction patterns of TiSe
Te
: (1) before TGA, (2) after TGA under Ar, and (3) after TGA under N .
0.16 1.44 2
bly more stable than the tellurides. We found that the
thermal behavior of titanium tellurides differs from
the thermal dissociation described previously [9].
On the one hand, at temperatures of up to 1000 C, in
an inert atmosphere, tellurides show a much stronger
tendency for thermal dissociation than selenides. This
fact may be responsible for the faster growth of tita-
nium telluride crystals, compared to selenides. On the
other hand, titanium tellurides, compared to selenides,
appeared to be more reactive toward nitrogen. There-
fore, when performing gas transport reactions involv-
ing titanium tellurides, it is necessary to provide as
high vacuum as possible, to avoid side reactions. We
observed previously the binding of residual nitrogen
[11] when growing single crystals of titanium disele-
nide by a chemical gas transport reaction in the diffu-
sion mode. Similar trends are shown by intermediate
titanium selenotellurides in both atmospheres.
spectral analysis, was as follows: Fe 0.03, Si 0.01, Al
0.01, and Cr 0.001. To avoid oxidation, the metal was
stored in evacuated ampules. Stoichiometric amounts
of the starting substances were placed in ampules
made from vacuum compression quartz; the ampules
were evacuated to 0.13 Pa and sealed. The syntheses
were performed in a muffle furnace in the stepwise
heating mode: 550 C, 170 h (primary reaction) and
then 850 C, 720 h, followed by quenching in water at
0 C. Two samples of each composition were prepared.
The composition of titanium selenide was deter-
mined by gravimetric analysis (calcination to TiO2 in
an oxygen flow at 900 C [6]). The phase composition
of the synthesis products and the unit cell parameters
were determined by X-ray phase analysis (DRON-3,
graphite, CuK radiation, external reference -Al2O3).
The unit cell parameters were determined by the least-
squares method from the peaks at 20
52 ,
0.005 . After the thermogravimetric analysis (TGA),
the phase composition of titanium selenotellurides
was determined with a Philips Analytical PW-1710/00
EXPERIMENTAL
Polycrystalline samples of TiSe1.60 xTex (0
x
diffractometer from the peaks at 1.5
2
154 .
1.60, x = 0.16) were prepared by a direct ampule
procedure. The starting substances were selenium
(ultrapure V grade), tellurium (ultrapure T-A1 grade),
and titanium prepared from titanium hydride by vacu-
um dehydrogenation at 850 C. The content of impuri-
ties (wt %) in this titanium sample, as determined by
Samples after TGA (10 mg) were mixed with CaF2
(internal reference, 40 mg).
The thermal behavior of TiSe1.60 xTex was studied
for five compositions: x = 0, 0.16, 0.80, 1.44, and
1.60, by TGA on a Mettler-Toledo Ta-8000 device
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 75 No. 7 2005

THERMAL DECOMPOSITION OF Ti5(Se, Te)8
1009
equipped with a TGA-850 thermal balance. The ex-
periments were performed in an Ar or N2 flow; the
sample weight was 10 mg, and the heating rate,
5 deg min . The samples were heated from 25 to
1000 C and then cooled to 100 C.
4. Gronvold, F. and Langmyhr, F.J., Acta Chem. Scand.,
1961, vol. 15, no. 10, p. 1949.
5. Bernusset, P., C. R. Acad. Sci. Paris, 1963, vol. 257,
1
no. 19, p. 2840.
6. Pankratova, O.Yu., Vladimirova, V.A., and Zvin-
chuk, R.A., Zh. Neorg. Khim., 1991, vol. 36, no. 4,
p. 1050.
ACKNOWLEDGMENTS
7. Pavlova, G.V., Pankratova, O.Yu., Zvinchuk, R.A.,
and Suvorov, A.V., Vestn. S.-Peterb. Gos. Univ.,
Ser. 4: Fiz., Khim., 2000, issue 2, no. 11, p. 117.
The study was financially supported by the Minis-
try of Education of the Russian Federation (grant no.
E02-5.0-226).
8. Pankratova, O.Yu., Grigor’eva, L.I., Zvinchuk, R.A.,
and Suvorov, A.V., Vestn. S.-Peterb. Gos. Univ.,
Ser. 4: Fiz., Khim., 1997, issue 1, no. 4, p. 119.
REFERENCES
1. Rimmington, H.P.B. and Balchin, A.A., J. Cryst.
Growth, 1974, vol. 21, no. 2, p. 171.
9. Suzuki, A. and Wahlbeck, P.G., J. Phys. Chem., 1966,
vol. 70, no. 6, p. 1914.
2. Brunie, S. and Chevreton, M., Mater. Res. Bull., 1968,
10. McTaggart, F.K. and Wadsley, A.D., Aust. J. Chem.,
vol. 3, no. 4, p. 309.
1958, vol. 11, no. 4, p. 445.
3. Pankratova, O.Yu., Grigor’eva, L.I., Zvinchuk, R.A.,
and Suvorov, A.V., Zh. Neorg. Khim., 1993, vol. 38,
no. 3, p. 410.
11. Pankratova, O.Yu., Grigor’eva, L.I., Zvinchuk, R.A.,
and Suvorov, A.V., Zh. Obshch. Khim., 1998, vol. 68,
no. 10, p. 1598.
RUSSIAN JOURNAL OF GENERAL CHEMISTRY Vol. 75 No. 7 2005