The influence of cooling rate on the formation of an amorphous phase in Si-based multicomponent alloys and its thermal stability

Article abstract of DOI:10.1016/S0025-5408(99)00099-9

By adding Ni, Cr, and Zr to the Si-Al-Fe system and using a higher surface velocity of the copper wheel of 63 m/s, the composition range of an amorphous single phase was extended to 60 at% Si. An amorphous Si₆₀Al₁₈Fe₁₂Ni₄Cr₃Zr ₃ alloy was produced. The amorphous phase in the Si-Al-TM (TM = transition metals) alloys shows higher thermal stability than in the Al-Si-TM alloys. By the shift of differential scanning calorimetry (DSC) peaks with change of heating rate, the activation energy for the transformation was determined to be 343 kJ/mol for Si₅₅Al₂₀Fe₁₀Ni₅Cr₅Zr ₅ alloy and 374 kJ/mol for Si₅₀Al₂₅Fe₁₀Ni₅Cr₅Zr ₅ alloy.

Full text of DOI:10.1016/S0025-5408(99)00099-9

Materials Research Bulletin, Vol. 34, No. 7, pp. 1165–1172, 1999
Copyright © 1999 Elsevier Science Ltd
Printed in the USA. All rights reserved
0025-5408/99/$–see front matter
PII S0025-5408(99)00099-9
THE INFLUENCE OF COOLING RATE ON THE FORMATION OF AN
AMORPHOUS PHASE IN Si-BASED MULTICOMPONENT ALLOYS AND ITS
THERMAL STABILITY
Dmitri V. Louzguine* and Akihisa Inoue
Institute for Materials Research, Tohoku University, Katahira 2-1-1, Aoba-Ku,
Sendai 980-8577, Japan
(Refereed)
(Received July 6, 1998; Accepted August 21, 1998)
ABSTRACT
By adding Ni, Cr, and Zr to the Si–Al–Fe system and using a higher surface
velocity of the copper wheel of 63 m/s, the composition range of an amor-
phous single phase was extended to 60 at% Si. An amorphous
Si60Al18Fe12Ni4Cr3Zr3 alloy was produced. The amorphous phase in the
Si–Al–TM (TM ϭ transition metals) alloys shows higher thermal stability
than in the Al–Si–TM alloys. By the shift of differential scanning calorimetry
(DSC) peaks with change of heating rate, the activation energy for the
transformation was determined to be 343 kJ/mol for Si55Al20Fe10Ni5Cr5Zr5
alloy and 374 kJ/mol for Si50Al25Fe10Ni5Cr5Zr5 alloy. © 1999 Elsevier Science Ltd
KEYWORDS: A. amorphous materials, C. differential scanning calorimetry
(DSC), C. electron microscopy
INTRODUCTION
Positive influence of multicomponent alloying on the formation of the amorphous phase has
been declared in some works (for example, ref. 1). It has also been shown that the addition
of one component lowers the probability of the formation of any concentration fluctuations
in liquid alloys, preventing formation of crystalline nuclei [2]. The presence of several types
of atoms in the amorphous phase impedes diffusion and helps to prevent crystallization of the
amorphous phase on heating. Mutual alloying of Al–Si–Fe by different transition metals
*To whom correspondence should be addressed.
1165

1166
D.V. LOUZGUINE and A. INOUE
Vol. 34, No. 7
causes an increase in the degree of the satisfaction of the empirical rules for achievement of
high glass-forming ability, i.e., (1) multicomponent alloy system, (2) large atomic size ratios
among the main constituent elements, and (3) negative heats of mixing among the constituent
elements [1]. The Si–Al–TM (TM ϭ transition metals) system satisfies these requirements.
On the basis of these principles, Si-based amorphous alloys containing 50–55 at% Si have
been produced [3] by adding several transition metals (namely, Ni, Cr, and Zr) to the ternary
Al–Si–Fe alloys studied earlier [4–6]. The compositional ranges for the formation of the
amorphous single phase at the surface velocity of the copper wheel of 42 m/s were found to
be Si50Al22–28Fe10–18Ni3–7Cr3–7 and Si55Al20–22Fe8–10Ni5–7Cr3–5Zr3–5 [3]. The stability of
the amorphous phase, however, has remained poorly studied. In the present paper, we
describe the influence of cooling rate on the formation of the amorphous phase and thermal
stability of multicomponent amorphous Si–Al–Fe–TM alloys.
EXPERIMENTAL
Ingots of Si–Al–Fe–Ni–Cr–Zr alloys were prepared by arc melting a mixture of pure Al
(99.99%), Fe (99.9%), Cr (99.9%), Ni (99.9%), Zr (99.7%), and Si (99.9999%) in an argon
atmosphere. From these alloys, ribbon samples of about 0.015–0.02 mm in thickness and
0.9–1.0 mm in width were prepared by rapid solidification of the melt on a single copper
roller at different surface velocities of the copper wheel. The amorphous structure of ribbon
samples was examined by X-ray diffraction with monochromatic Cu K␣ radiation. Trans-
mission electron microscopy (TEM) observation was carried out using a JEOL JEM 2010
electron microscope operating at 200 kV. Samples for TEM were polished electrolytically in
a solution of 10 vol% perchloric acid and 90 vol% methanol at 208–213 K. Crystallization
temperature and heat of crystallization were examined by differential scanning calorimetry
(DSC) at different heating rates.
RESULTS
Influence of the Cooling Rate on Amorphization. In addition to Si50Al26Fe10Ni7Cr7 and
Si55Al20Fe10Ni5Cr5Zr5 amorphous alloys, which were produced earlier [3], we observed the
formation of an amorphous single phase in Si50Al20–26Fe8–12Ni5–7Cr3–5Zr3–5 and
Si60Al18Fe12Ni4Cr3Zr3 alloys. Amorphization of the alloy containing 60% Si (Fig. 1) was
achieved by using a surface velocity of the copper wheel of 63 m/s. Alloys with 50 and 55
at% Si were found to have a definite compositional area of homogeneity [3] of the amorphous
phase, whereas a slight change in composition of the Si60Al18Fe12Ni4Cr3Zr3 alloy led to
precipitation of a crystalline phase. Conditions for the formation of the amorphous phase
depending on the surface velocity of the copper wheel are summarized in Table 1.
DSC curves of Si50Al25Fe10Ni5Cr5Zr5, Si55Al20Fe8Ni7Cr5Zr5 and Si60Al18Fe12Ni4Cr3Zr3
alloys are shown in Figure 2. The curves for the Si50Al26Fe10Ni7Cr7 and
Si55Al20Fe10Ni5Cr5Zr5 alloys have been reported in ref. 3. The Si50Al25Fe10Ni5Cr5Zr5,
Si55Al20Fe10Ni5Cr5Zr5, and Si55Al20Fe8Ni7Cr5Zr5 amorphous alloys transform to a crystal-
line phase through a one-stage exothermic reaction leading to the formation of the multi-
component Si10Al4Fe2NiCrZr [3] and a small amount of Fd3m Si phase.
Transformation Diagrams. Thermal stability of the amorphous phase in the
Si55Al20Fe10Ni5Cr5Zr5 alloy was examined by isothermal annealing of as-solidified samples

Vol. 34, No. 7
SILICON-BASED ALLOYS
1167
FIG. 1
Bright-field electron micrograph (a) and selected-area electron diffraction pattern (b) of a
rapidly solidified Si60Al20Fe8Ni4Cr4Zr4 alloy.
at different temperatures. A transformation diagram is shown in Figure 3a. At higher
temperatures, a diagram for ribbon samples, as plotted in Figure 3b, was constructed by using
incubation time from isothermal calorimetry curves. The ribbon samples were annealed
isothermally at temperatures of 667, 671, 675, and 679 K (Fig. 4a). Heating up to each
annealing temperature was done at a heating rate of 3.3(3) K/s. Points on the curve
correspond to less than 0.5% of the volume fraction transformed.
Kinetics of Crystallization. The study of the kinetics of crystallization of the
Si55Al20Fe10Ni5Cr5Zr5 alloy was performed using the Johnson–Mehl–Avrami isothermal
analysis for volume fraction x transformed as a function of time t on the basis of the
following equation [7]:
x(t) ϭ 1 Ϫ exp[Ϫ(k(T)t)n]
(1)
TABLE 1
Microstructure of Si-Based Alloys Depending on the Cooling Rate Controlled by the
Copper Wheel Surface Velocity
Copper wheel surface velocity, m/s
Alloy composition
16
21
31.5
42
63
Si50Al26Fe10Ni7Cr7
Si55Al20Fe10Ni5Cr5Zr5
Si60Al20Fe8Ni4Cr4Zr4
am. ϩ cryst.a
cryst.c
am.b
am. ϩ cryst.
cryst.
am.
am. ϩ cryst.
cryst.
am.
am.
am. ϩ cryst.
am.
am.
am.
cryst.
aAmorphous ϩ crystalline.
bFully amorphous.
cCrystalline.

1168
D.V. LOUZGUINE and A. INOUE
Vol. 34, No. 7
FIG. 2
DSC curves of the amorphous Si50Al25Fe10Ni5Cr5Zr5, Si55Al20Fe8Ni7Cr5Zr5, and
Si60Al20Fe8Ni4Cr4Zr4 alloys.
The samples were annealed isothermally at several temperatures: 667, 671, 675, and 679 K
(see Fig. 4a). The volume fraction transformed vs. the annealing time plot is shown in Figure
4b. The volume fraction transformed at the time t is assumed to scale with the fraction of the
total heat released. Eq. (1) can be written as
ln[ln(1/(1 Ϫ x)] ϭ n(ln(k) ϩ nln(t))
(2)
where k is an effective rate constant and n is Avrami exponent. The Avrami plot of
ln[Ϫln(1 Ϫ x) vs. ln(t) yields a straight line with slope n and intercept nln(k). Figure 4c
shows this plot at different temperatures. The plots are reasonably linear. The average slope
corresponds to n ϭ 3.4.
Energy Barrier Opposing Crystallization. In differential scanning calorimetry, the tem-
perature at which the maximum deflection is observed varies with heating rate. By making
a number of differential calorimetry measurements at different heating rates, the kinetic
constant can be obtained directly from the DSC data. The activation energy E (the energy
barrier opposing the crystallization) and a frequency factor A (a measure of the probability
that an atom having energy E will participate in the reaction) in the following equation for
the fraction transformed x [8]:
dx/dt ϭ A(1 Ϫ x)exp(ϪE/RT)
(3)
can be obtained directly from the temperature Tm (peak temperature) at which the ratio dx/dt
attains the maximum [8]. The corresponding equation is:

Vol. 34, No. 7
SILICON-BASED ALLOYS
1169
FIG. 3
Transformation diagrams for the Si55Al20Fe10Ni5Cr5Zr5 alloy at different temperature
ranges. Tx ϭ crystallization temperature, measured by DSC at a heating rate of 0.042 K/s.
2
ϪE/R ϭ d(lnVh/Tm )/d(1/Tm)
(4)
where Vh is the heating rate and R is the gas constant. The data plotted as indicated by the
relation (lnVh/Tm2) vs. (1/Tm) for Si55Al20Fe10Ni5Cr5Zr5 are shown in Figure 5a. Heating
rates used were 0.042, 0.083, 0.17, 0.67, and 1.33 K/s. From the slope, the activation energy
was estimated to be 343 kJ/mol (or 3.55 eV). The same Kissinger plot (lnVh/Tm2) vs. (1/Tm)
for the Si50Al25Fe10Ni5Cr5Zr5 alloy is shown in Figure 5b. The slope value is close to that
for the Si55Al20Fe10Ni5Cr5Zr5 alloy (see Fig. 5a) and the activation energy is estimated to be
374 kJ/mol.
DISCUSSION
Changing the surface velocity of the copper wheel is a widely used procedure to influence the
cooling rate during rapid solidification. It is seen from Table 1 that an increase of surface

1170
D.V. LOUZGUINE and A. INOUE
Vol. 34, No. 7
FIG. 4
(a) Isothermal differential calorimetry curves, (b) the fraction transformed vs. the annealing
time, and (c) the Avrami plot for the Si55Al20Fe10Ni5Cr5Zr5 alloy.
velocity, which means an increase of cooling rate, enables the formation of amorphous alloys
with higher Si content. Compositional ranges for the formation of the amorphous phase are
Si50Al22–28Fe10–18Ni3–7Cr3–7 [3], Si50Al20–26Fe8–12Ni5–7Cr3–5Zr3–5, Si55Al20–22Fe8–10Ni5–7
Cr3–5Zr3–5 [3], and Si60Al18Fe12Ni4Cr3Zr3. From the data obtained and the narrowing of the
composition range for the formation of the amorphous phase with an increase in Si content,
we can conclude that 60 at% is the highest Si content at which an amorphous single phase
can be obtained in Si–Al–Fe–Cr–Ni–Zr and some other Si–Al–TM systems by the melt
spinning technique. To the reasons enumerated in our previous work [3], for the formation
of the amorphous phase in the Si-rich alloys, we can add a large difference in the atomic radii
among the Si, Al, and TM elements. Al, Zr, and other TM atomic radii are 1.22, 1.62, and
about 1.07 times, respectively, larger than Si atomic radius.
The amorphous phase in the Si–Al–TM alloys exhibits a relatively high thermal stability
in comparison with that in Al–Si–TM alloys [9]. Figure 3a shows that at temperatures lower

Vol. 34, No. 7
SILICON-BASED ALLOYS
1171
FIG. 5
Kissinger plot for (a) Si55Al20Fe10Ni5Cr5Zr5 and (b) Si50Al25Fe10Ni5Cr5Zr5 alloys.
than 600 K, the Si55Al20Fe10Ni5Cr5Zr5 alloy was stable for more than 100 h. At lower
temperatures, the C curve becomes practically parallel to the axis of time.
Si50Al25Fe10Ni5Cr5Zr5, Si55Al20Fe10Ni5Cr5Zr5, and Si55Al20Fe8Ni7Cr5Zr5 amorphous al-
loys transformed into a crystalline state through a one-stage exothermic reaction, leading to
the formation of the multicomponent Si10Al4Fe2NiCrZr and a small amount of Fd3m Si
phase. Crystalline particles formed during crystallization are shown in Figure 6. This figure
shows that the transformation was not completed and the amorphous phase still remained.
The homogeneous distribution of the particles observed in different parts of the ribbon
sample leads us to believe that each crystallite grew from its own nuclei. We consider that
the crystallization of the amorphous phase in Si55Al20Fe10Ni5Cr5Zr5 is close to polymor-
phous transformation into a multicomponent Si10Al4Fe2NiCrZr phase. The value of Avrami
exponent 3.4 is usual for a transformation involving an overlap three-dimensional linear
growth on quenched-in nuclei and on new nuclei. The activation energy as the energy barrier
opposing the crystallization of Si55Al20Fe10Ni5Cr5Zr5 alloy of 343 kJ/mol is close to the
value of 345 kJ/mol for the polymorphic crystallization of Ti2Ni alloy [10].

1172
D.V. LOUZGUINE and A. INOUE
Vol. 34, No. 7
FIG. 6
Microstructure of the Si55Al20Fe10Ni5Cr5Zr5 alloy after annealing for 6000 sec at 667 K.
CONCLUSIONS
In conclusion, it can be said that by adding Ni, Cr, and Zr to the Si–Al–Fe system and using
a higher speed of the copper wheel, an amorphous Si60Al18Fe12Ni4Cr3Zr3 alloy was pro-
duced. Si50Al25Fe10Ni5Cr5Zr5, Si55Al20Fe10Ni5Cr5Zr5, and Si55Al20Fe8Ni7Cr5Zr5 amor-
phous alloys transform into a crystalline state through a one-stage exothermic reaction. At
temperatures lower than 600 K, the Si55Al20Fe10Ni5Cr5Zr5 alloy was stable for more than
100 h. The crystallization kinetics of the Si55Al20Fe10Ni5Cr5Zr5 alloy was analyzed on the
basis of the Johnson–Mehl–Avrami relation. The Avrami exponent n was found to be 3.4.
The activation energy for the transformation determined by the shift of DSC peaks with
change of heating rate to be 343 kJ/mol for Si55Al20Fe10Ni5Cr5Zr5 alloy is close to the value
of 345 kJ/mol for the polymorphic crystallization of Ti2Ni alloy.
REFERENCES
1. A. Inoue, Mater. Trans., JIM 36, 866 (1995).
2. P.J. Desre, Mater. Trans., JIM 38, 583 (1997).
3. D.V. Louzguine and A. Inoue, Mater. Trans., JIM 38, 1095 (1997).
4. O. Suzuki, Y. Komatsu, K.E. Kobayashi, and P.H. Shingu, J. Mater. Sci. 18, 1195 (1983).
5. A. Inoue, M. Yamamoto, H.M. Kimura, and T. Masumoto, J. Mater. Sci. Lett. 6, 194 (1987).
6. A. Inoue, Y. Bizen, H.M. Kimura, and T. Masumoto, J. Mater. Sci. 23, 3640 (1988).
7. J.W. Christian, The Theory of Transformations in Metals and Alloys, Pergamon Press Ltd., Oxford
(1975).
8. H.E. Kissinger, J. Res. Natl. Bur. Stand. 57, 217 (1956).
9. R.A. Dunlap and K. Dini, J. Mater. Res. 3, 415 (1986).
10. M. Buchwitz, R. Adlwarth-Diedall, and P.L. Ryder, Acta Metall. Mater. 41, 1885 (1993).