I.V. Fedorchenko et al. / Journal of Alloys and Compounds 599 (2014) 121–126
123
samples we observed to be two phases. The differential thermal
analysis was done on derivatograph LB-2. The accuracy of the data
is ±10°. To support these data the samples were studied with
EDXRF, X-ray and SEM.
D-5000 diffractometer (E = 40 keV, I = 25 mA, t = 8 s/point,
step = 0.02°). The calculations of lattice parameters and existing
phases were made by TRIOR program, Table 3. All samples from
2 2
x = 0.1–0.9 had two phases, namely ZnGeAs and CdGeAs . The lat-
The EDXRF analyses were done with Tracor X-ray Spectrace
000 Spectrometer equipped with Si(Li) detector. Circular (1.5–
mm thickness) discs cut from the middle of the samples were
tice parameters for single phase samples NN 1 and 11 are in good
agreement with literature data (ICDD). The solid solutions with
5
2
2 2
chemical content near both ZnGeAs and CdGeAs alloy have
used for the measurements. The EDXRF spectra obtained for each
individual sample were fitted to the theoretical model assuming
the presence of fluorescence Ka lines of Zn, Cd, Ge and As atoms.
No other atoms than the ones mentioned above were observed in
the EDXRF spectra. It indicated that (if they were present) the con-
centration of unintentional dopants was lower than 1015 cm . The
relative maximum uncertainty of this method with respect to the
calculation of molar fractions of the elements does not exceed
changes of the lattice parameters depending from the concentra-
tion of the opposite component. As the chosen compounds have
the same crystal structure, close values of the lattice parameters,
difference of atomic radii is less than 12% and approximate valence
shell of the atom and the difference of the electronegativity is 0.04
(Pauling scale), it is logical to suppose that these compounds will
form substitutional solid solutions which have to obey the Vegard’s
law. Vegard law for the ‘‘a’’ parameter can be written as:
ꢀ3
1
0%. The EDXRF results are gathered in Table 1. The dispersion of
a
Zn Cd GeAs ¼ aZnGeAs2 ꢂ x þ ð1 ꢀ xÞaCdGeAs2
ð1Þ
the data suggests that the liquation process is present in the sam-
ples. It encourages us to make more careful investigations.
1ꢀx
x
2
On the other hand, the calculation of the cell parameters was
The Scanning electron microscope (SEM) investigations were
made by two different apparatus. First, Carl Zeiss NVision40
three-beam workstation, equipped with Oxford Instruments
X-Max analyzer, at the Center for collective use of physical methods
of study Inorganic Chemistry, RAS. The range of accelerating volt-
ages was up 1–20 kV. X-ray microanalysis was performed in scan-
made through X-ray patterns present in Table 3. Comparing these
data, the conclusions about solubility of the components can be
correct. In Fig. 4 the hypothetical Vegard’s law line for ZnGeAs
2
–
CdGeAs solid solutions and the biggest changes of the lattice
2
parameter calculated with X-ray data are presented. The crossing
of these lines determines the maximum solubility of the
components (can be calculated from Eq. (1). According to Fig. 4
ning mode and
a full line of mapping. Fig. 1 shows that
distribution of the components is homogeneous except for Zn and
Cd. In Fig. 1f it is clearly visible that samples consist of three differ-
ent areas, two light compounds separated with dark one. Fig. 1e
shows that dark area consists both type of zinc and cadmium atoms.
Table 2 presents the concentrations along the marked lines in Fig. 2.
It is clearly visible that the dark area satisfied uninterrupted com-
pound of solid solutions and at the same time, more bright areas
have a sharp border with it, with concentration from 16 till
the border compounds have compositions Zn0,08Cd0,92GeAs
2
and
Zn0,81Cd0,19GeAs . These data are different from early work [24].
2
The results of the EDXRF, microstructures, X-ray and DTA are
listed in Table 4. The recalculation of the X-ray data to the concen-
tration of zinc and cadmium ions were done taking in account the
phase relations and concentration of the component in solid solu-
tions determined by the Vegard law. The presence of the cubic
structure in the solid solutions was reported in the literature
1
9 at.% from both sides. These data show some difference with early
work. The authors [24] defined the borders of the compounds as not
more than 10 wt.% from both sides, these are Zn0.16Cd0.84GeAs and
Zn0.94Cd0.06GeAs respectively. This demands more careful study, as
[
28]. The difference of the data of the composition of the samples
can be explained by the nanosized (20–40 nm) inclusions into
2
2
the observed cadmium solubility is bigger than three times from our
SEM observation.
Table 2
SEM data of surface in Fig. 2.
A more detailed investigation of the dark area were made by the
field emission scanning electron microscopy with the use of Hit-
achi SU-70 Analytical UHR FE-SEM equipped with the Thermo
Fisher NSS 312 energy dispersive X-ray spectrometer system
Specter
Zn (at.%)
Cd (at.%)
Ge (at.%)
As (at.%)
4
4
4
4
.0
.4
.6
.5
21.4
20.6
20.6
20.1
23.6
23.8
23.9
24.2
51.1
51.2
50.9
51.1
1
2
3
(
EDS) equipped with SDD-type detector in the Institute of Physics
4
5
6
7
8
9
PAS. The imagine of the dark area is presented in Fig. 3. In Fig. 3a it
is clearly visible, that dark area consists of the mixture of the small
dispersion granules of solid solutions with both sides. More dark
color of this zone is explained by the borders between grains. In
Fig 3b and c we show the SEM image at which the small grains
are included into the homogeneous material. The formation of
these grains is explained by dissipation of the solid solution at
the solid state. They were segregated from the saturated solutions
with decreasing temperature. The process of segregation goes
along the binodal dissipation curve, that is the reason of the differ-
ent size of the granules. Moreover, as it is visible from Fig. 3c, the
distribution of the inclusions is random and the SEM data cannot
be correct, as the underlying inclusions will make influence on
the SEM data. The same effect is described in the literatures
5.0
6.7
7.4
8.1
9.8
11.3
15.1
11.0
7.8
8.1
10.0
9.4
19.8
18.3
17.7
17.7
15.2
13.1
9.2
13.9
17.5
16.9
15.5
16.1
16.9
16.1
14.0
13.9
5.4
24.0
23.7
24.1
23.0
24.0
23.8
23.6
23.8
23.2
23.7
23.3
24.4
23.7
23.6
24.2
23.8
23.8
51.2
51.3
50.8
51.2
51.0
51.7
52.1
51.2
51.4
51.2
51.1
50.1
50.7
50.5
50.4
49.9
50.9
10
11
12
1
1
1
1
3
4
5
6
17
18
9.0
9.7
11.4
12.4
1
2
9
0
19.9
2
2
1
2
[
25–27]. The report [25] described the formation of nanometer-
5.1
4.9
6.2
7.6
10.1
13.5
20.0
20.3
18.9
17.7
15.4
12.7
5.1
24.0
24.3
23.7
23.5
22.7
21.0
22.8
51.0
50.5
51.3
51.2
51.8
52.9
51.7
sized (ꢁ70 nm) grains in equimolar ZrC–NbC solid solutions. The
23
24
authors reported, that the nanostructure remains stable during
25
long-term (100–700 h) annealing between 670 and 1270 K. This
way, SEM data requires support the solubility of AII cations.
26
27
28
29
The powder X-ray diffraction method was used to correct the
solubility data. The diffraction patterns were obtained with
Cu-radiation ((Cu Ka1) = 1.54056 Å) with the use of the Siemens
20.4
21.2
4.2
23.1
51.5