Differential thermal analyses of Cd0.95-xMnxZn0.05Te alloys V. Kopach1, O. Kopach1, L. Shcherbak1, P. Fochuk1, A. E. Bolotnikov2, R. B. James2

1 - Chernivtsi National University, Ukraine 2 - Brookhaven National Laboratory, USA

Abstract

Peculiarities of Cd0.95-xMnxZn0.05Te (x=0.05-0.25) alloys melting and crystallization kinetics were investigated by

the differential thermal analysis at various heating/cooling rates. We synthesized Cd0.95-xMnxZn0.05Te alloys from elemental materials in a double-zone furnace. Their melting-crystallization rate was 5- and 10-K/min with a melt-dwell time of 1-, 10-, 20-, and 30-minutes. We found that the melting temperature, Tm, declined with increasing “x”, while the crystallization temperature, Ts, rose with increasing “x”. We detailed the dependencies of the crystallization rate versus the melt’s crystallization temperature. In most cases, the dependencies dropped with the rise of the crystallization temperature, while the crystallization temperature fell with decreasing melt-dwell time.

Keywords: Differential Thermal Analyses, Cd0.95-xMnxZn0.05Te, melts, supercooling, superheating, solid-state

volume fraction.

Introduction Detailed investigations of known semiconductors and the development and synthesis of new and better ones are

ongoing activities in many fields. In recent years, the semiconductors AIIBVI have seen widespread use in the IR- and γ-detector industry, in optoelectronic devices, and in high-temperature electronics. Thus, there is a continuing need for better single crystals grown from CdTe- and ZnTe-alloys. Alloys of the ternary system, CdTe-ZnTe-MnTe, are diluted magnetic semiconductors (DMS). Among these compounds, the ones most developed are the Cd1-xMnxTe alloys that are crystallized from a zinc blend structure up to x=0.82 [1]. The difference between DMSs and other semiconductors is that one of its atoms, which has magnetic properties, has replaced an atom of the component (the Mn atom in our compound). Thus, DMS materials have a large magneto-optical effect, which is dominated by the specific properties of the magnetic component, viz., a strong interaction between the electrons’ spin, or the band-gap holes, with a localized magnetic moment caused by d-electron charged magnetic ions.

The quaternary compounds, Cd1-x-yMnxZnyTe, contain two types of ternary compounds, Cd1-xMnxTe and Cd1-xZnxTe [2]. Therefore, it is important to study a variety of sp-d exchange interactions, which are similar to those of d-d, as a nonmagnetic semiconductor changes from CdTe to CdZnTe, and then to ZnTe. To conduct these experiments, a major problem lies in growing single crystals of Cd1-x-yMnxZnyTe, which requires our undertaking a detailed study of the system’s liquidus- and solidus-curves, data on which is quite sparse [3]. Hence, the aim of our work is to elucidate the processes of melting and crystallization of a solid solution of the CdTe-ZnTe-MnTe system, and clarify the behavior of these alloys during their heating- and cooling-processes at the melting point.

Experiments

Cd0,,95-xMnxZn0.05Te alloys (x=0.05-0.25) were synthesized from individual materials in a glass-graphite crucible in a vertical furnace with a high-gradient temperature that prevented the sublimation of the components. The mass of the synthesized alloys was 5 g. The maximum temperature of synthesis was 1428 K, and the melt homogenization process took 5 hours following by slow cooling. For further investigation by differential thermal analyses (DTA), we loaded 500 mg of the synthesized alloys into a graphite-coated quartz ampoule, 8 mm in diameter. A quartz rod was placed under the melt to minimize the free space and assure the standardization of the DTA conditions. The ampoule then was evacuated to 10-3 mbar.

The DTA was carried out in an automatic system constructed in the Inorganic Chemistry Department. The quasi-equilibrium heating/cooling rates were specified equal to 5- and 10-K/minute, and the dwell times were 10-, 30-, and 60-minutes. High-purity quartz was used as the etalon, and a Pt and Pt-Rh thermocouple served as the temperature register. The temperature was measured to better than ±1.5 K. Recording of the thermal effects caused by different methods of heating was automated, as was the temperature control in the furnace.

Hard X-Ray, Gamma-Ray, and Neutron Detector Physics XV, edited by Michael Fiederle, Arnold Burger,Larry Franks, Ralph B. James, Proc. of SPIE Vol. 8852, 88521D · © 2013 SPIE

CCC code: 0277-786X/13/$18 · doi: 10.1117/12.2022837

Proc. of SPIE Vol. 8852 88521D-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 10/09/2013 Terms of Use: http://spiedl.org/terms

Results and discussion The DTA were processed in two different ways. In the first method, the sample was heated to the maximum temperature of about 1423 K. After a dwell time of the

melt at this temperature (10-, 30- or 60-minutes), it next was cooled to 1123 K. In the following cycles, the temperatures during the dwell periods were lowered in steps of 10 K down to the alloy’s melting temperature, Tm. This allowed us to determine the areal extent of full melting of the alloy, and to investigate its supercooling versus superheating dependencies.

The DTA thermograms of melting and crystallization obtained from the first thermal-treatment method are shown in Fig. 1 for the Cd0.90Mn0.05Zn0.05Te alloys. The smoothness of the melting process is typical for solid solutions case. The melts’ crystallization temperature usually was lower than the melting effect’s baseline bend temperature, a feature that is significant for the melts’ supercooling behavior. As one can see in Fig.1., larger dwell temperature followed by lower crystallization effect temperature.

1260

1300

1340

1380

1420

1460

40 60 80 100 120

Time, min

T sam

ple,

K

-160

-140

-120

-100

-80

-60

ΔU

, μV

1368 K

1391 K

1405 K

1341 K

1342 K

1319 K

1423 K

1315

1340

1365

1390

40 50 60 70 80 90

Time, min

T sam

ple,

K

-135

-120

-105

-90

ΔU

, μV

1380 K1367 K

1383 K

1380 K

1375 K

1362 K

1347 K

а) b)

Fig. 1. Typical thermograms of the melting and crystallization of Cd0.90Mn0.05Zn0.05Te alloys under the maximum (a) and minimum (b) dwell temperatures (Vh/c = 5 K/min).

We investigated the melt’s supercooling dependence versus its superheating. In Figs. 2a and 2b, we show the

melt’s supercooling versus its superheating for the alloys of Cd0.90Mn0.05Zn0.05Te and Cd0.80Mn0.15Zn0.05Te. The melts of both alloys crystallize with the supercooling, but not in the full-range of their superheating; the so-called melt’s “negative” supercooling effect is evident for both alloys when the melts are superheated only to ∼ 20 K and ∼ 15 K,

-15

0

15

30

45

60

0 20 40 60 80ΔT+, K

ΔT- , K

time dwell = 10 mintime dwell = 30 mintime dwell = 60 min

-10

0

10

20

30

40

50

0 20 40 60 80ΔT+, K

ΔT- , K

time dwell=10 mintime dwell=30 mintime dwell=60 min

а) b)

Fig. 2. The supercooling versus superheating dependences for (a) the Cd0.90Mn0.05Zn0.05Te alloy, and, (b) the Cd0.80Mn0.15Zn0.05Te alloy (Vh/c = 5 K/min).

Proc. of SPIE Vol. 8852 88521D-2

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 10/09/2013 Terms of Use: http://spiedl.org/terms

respectively. That is, the melts’ crystallization temperatures are higher than the alloys’ thaw points. In addition, Fig. 2 illustrates that melt (a), Cd0.90Mn0.05Zn0.05Te, became supercooled a little more (ΔT- = 25-55 K), than did (b) Cd0.80Mn0.15Zn0.05Te (ΔT- = 15-40 K) after superheating them above the critical point.

Corresponding to the melt crystallization effect area in the thermograms (Scrystallization) versus dwell-temperature dependence (Fig. 3) shows that the full crystallization occurs when the superheating of the Cd0.80Mn0.15Zn0.05Te melt is higher than 1385 K. In the case of lower dwell temperatures the process of the melt crystallization released less heat, therefore the melt is in a semi-liquid state.

0

10

20

30

40

50

60

1370 1380 1390 1400 1410 1420 1430 1440T dwell, K

S cr

ysta

lliza

tion

time dwell = 10 mintime dwell = 30 mintime dwell = 60 min

Fig. 3. Crystallization effect area versus dwell temperature for the Cd0.90Mn0.05Zn0.05Te alloy.

The second DTA procedure is characterized by series of intermediate dwells in the temperature range from the

starting melting point of the alloy up to the temperature of the top of melting record while heating the sample. After intermediate dwells, the alloys were heated to maximum temperature of 1423 K. Endothermic effect which was recorded during heating after the dwell, reflected melting of part of solid phase (probably, clusters of the tellurides with shortest and, correspondingly, strongest bonds) that had remained in the melt after the dwell. A reduction of this affected area indicates a decrease in the clusters volume in the melt.

Thermograms of the Cd0.70Mn0.25Zn0.05Te and Cd0.75Mn0.20Zn0.05Te alloys’ melting and crystallization obtained via our second way of thermal treatment of the samples are shown in Figs. 4a and 4b,. Overheating the alloys after an intermediate dwell clearly fixed the melting of the solid phase, which remained in the melt after the dwell.

1250

1290

1330

1370

1410

1450

35 50 65 80 95 110

Time, min

T sam

ple,

K

-135

-120

-105

-90

-75

-60

ΔU

, μV1358 K

1365 K

1382 K

1390 K

1343 K

1346 K

1331 K1355 K

1290

1320

1350

1380

1410

1440

40 50 60 70 80 90 100

Time, min

T sam

ple,

K

-130

-115

-100

-85

-70

ΔU, μ

V1363 K

1377 K

1377 K 1376 K

1334 K

1337 K

1326 K1375 K

а) b)

Fig. 4. Typical DTA thermograms of the melting and crystallization of the alloys of (a) Cd0.70Mn0.25Zn0.05Te, and (b) Cd0.75Mn0.20Zn0.05Te with intermediate holding during the heating process, and a maximum melt temperature of

1423 K (Vh/c = 5 K/min).

The dependencies of the solid-state volume fraction (φsolid state) versus the intermediate dwell temperature of the alloys during the heating process for Cd0.70Mn0.25Zn0.05Te and Cd0.75Mn0.20Zn0.05Te, respectively, are presented in

Proc. of SPIE Vol. 8852 88521D-3

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 10/09/2013 Terms of Use: http://spiedl.org/terms

0 10 15 20

MnTe content, %

25

F

1370

1360 -

1350 -

1340 -

1330 -

1320

V hie =10K/minV hie =5K/min

0 10 15 20 25 30

MnTe content, %

Figs. 5a and 5b. They show that increasing of the melt-dwell temperature led to the melts full homogenization only near 1377±1 K. During the melting of the Cd0.80Mn0.15Zn0.05Te alloy, the influence of changing the heating/cooling rate from 10- to 5-K/min on the volume of the quasi-solid clusters is greater. Solid-phase melting at a heating/cooling rate of 5 K/min occurs at temperatures higher than at the heating/cooling rate of 10 K/min (Fig. 5a).

0

20

40

60

80

100

1356 1359 1362 1365 1368 1371 1374 1377T dwell, К

ϕ so

lid st

ate,

%

V h/c = 5 K/minV h/c = 10 K/min

0

20

40

60

80

1360 1363 1366 1369 1372 1375 1378T dwell, К

ϕ so

lid st

ate,

%

V h/c = 10 К/minV h/c = 5 К/min

а) b)

Fig. 5. The temperature-dependencies of the volume fraction of the solid phase remaining in (a) Cd0.70Mn0.25Zn0.05Te, and (b) Cd0.75Mn0.20Zn0.05Te melts after an isothermal dwell of 30 minutes.

Moreover, such dependence for Cd0.70Mn0.25Zn0.05Te (Fig. 5a) displays a convex upwards behavior and is similar to

such dependencies for melts of the Cd1-xZnxTe system. Besides the characteristics of the ϕsolid phase ∼ f(Tdwell) curve for Cd0.75Mn0.20Zn0.05Te alloy is similar to such dependencies as those for pure CdTe and CdTe-In system alloys [5]. Thus, these experimental results confirm the conclusions in [6] about the influence of anionic-lattice on the melting and crystallization processes of alloys based on tellurides. Increasing the mole fraction of Mn in alloys of this quaternary system reduces the influence of Zn on the characteristics of these processes because the content of the structural parts that contain Zn-Te bonds decreases.

а) b)

Fig. 6. The (a) melting- and (b) crystallization-temperature versus the MnTe content dependencies for Cd0.95-xMnxZn0.05Te alloys.

Thus, the thermograms allow us to determine the starting temperatures of melting and crystallization of the

synthesized alloys. The beginning melting temperature of alloys declined with increasing MnTe content (Fig. 6a). It is interesting that at lower melting/crystallization rate of 5 K/min the initial melting temperatures are greater in most cases. The crystallization temperature, Ts, increased with MnTe content rising. We studied the dependencies of the crystallization rate versus the melt’s crystallization temperature. In most cases, they declined with the rise of the crystallization temperature, while the crystallization temperature fell with decreasing melt-dwell time.

Proc. of SPIE Vol. 8852 88521D-4

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 10/09/2013 Terms of Use: http://spiedl.org/terms

References 1. Hwang Y., Chung S., Um Y., “Weak ferromagnetism in CdMnZnTe single crystal,” Phys. Stat. Sol. 12, 4457-

4460 (2007). 2. W. C. Chou, F. R. Chen, T. Y. Chiang, H. Y. Shin, “Growth and characterization of Cd1-x-yZnxMnyTe crystals,”

J. Crystal Growth 169, 747-751 (1996). 3. Beckett D. J. S., Bernard P., Lamarhe G., et al., “Growth of Bulk Single Crystals of CdxZnyMnzTe Alloys,”

J. Solid State Chem. 73, 585-587 (1988). 4. Larysa P. Shcherbak, Oleh V. Kopach, Petro M. Fochuk, Andriy I. Kanak, Aleksey E. Bolotnikov, and Ralph B.

James, “Melting and cooling processes in CdTe-ZnTe near the CdTe-rich side,” Proc. of SPIE, Vol. 8142, 8142L1-L7 (2011).

5. L. Shcherbak, P. Feychuk, O. Kopach, O. Panchuk, E. Hayer, H. Ipser, “Fine Structure of the Melting Process in Pure CdTe and in CdTe with 2 mol % of Ge or Sn,” J. Alloys Comp. 349, 145-149 (2003).

6. L. Shcherbak, P. Feychuk, O. Kopach, “Self-Assembly in Molten CdTe Doped by In (Sn, Sb),” in Proceedings of 6th International School-Conference, “Phase Diagrams in Material Science”, PDMS VI-2001, Octоber 14-20, 2001, Kiev, Ukraine, Editor Tamara Ya. Velikanova – Stuttgart (Materials Science International Services, GmbH, 199-209, (2004).

Proc. of SPIE Vol. 8852 88521D-5

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 10/09/2013 Terms of Use: http://spiedl.org/terms