Evolution of solidification microstructure in undercooled Co 80 Pd 20 alloys

Abstract

Applying fluxing method, Co₈₀Pd₂₀ alloy melts were undercooled up to 340 K. Evolution of the solidification microstructure of the undercooled alloys was investigated. Within the range of the achieved undercooling (ΔT), two grain refinement events were observed. The first event happens in the range of 50 K<ΔT<265 K, while the second event takes place at ΔT>280 K. Based on experimental investigations and theoretical calculations, the mechanisms of grain refinements were illuminated. The first grain refinement event is induced by the remelting effects involved in the post-recalescence period, resulting in the formation of a spherical grained microstructure. In the narrow transition range of 265 K<ΔT<280 K, the microstructure is occupied by a mixture of coarse dendrites with equiaxed grains. When ΔT is higher than 280 K, which is rather above the hypercooling limit, recrystallisation of the rapidly solidified solid happens and leads to the formation of an equiaxed grained microstructure.

Keywords

Solidification Undercooling Defects recrystallisation Alloys TEM

Introduction

Evolution of the solidification microstructure of undercooled metals has been extensively investigated for many years.1,2 Grain refinement as an interesting phenomenon of undercooled metals has aroused much attention.3^–7 A number of models were proposed to explain the grain refinement occurring when the initial undercooling of a melt (ΔT) exceeds a critical value, such as copious homogeneous nucleation induced by the pressure pulse generated from the collapse of shrinkage cavities,8 remelting of dendrite skeleton by chemical superheating,7 dendrite breakup driven by the instability of solid/liquid interface in post-recalescence period9 and stress induced recrystallisation.10^–14 So far, it seems to be widely accepted that the grain refinement happening at a range of relatively low ΔT is caused by the remelting effect during a post-recalescence period, e.g. the effect of chemical superheating7 and the effect of dendrite breakup driven by the solid/liquid surface tension induced structural instability, as suggested by Karma;15 for the grain refinement occurring at a range of relatively high ΔT, the corresponding mechanism is still in controversy. For example, for Ni–Cu systems, some authors believed that both of the two grain refinement events can be ascribed to the remelting effect during a post-recalescence period on the dendrite skeleton formed during recalescence,16,17 while others suggested that a stress induced recrystallisation should be responsible for the grain refinement in the range of high ΔT.4,18 If an alloy is hypercooled, the whole alloy melt will solidify after recalescence, and the post-recalescence period will disappear. As a consequence, the remelting effect involved in the post-recalescence period will not influence the microstructure formation in the hypercooled alloy. Then, there will be an opportunity to testify whether the dendrite remelting or the stress induced recrystallisation can be responsible for the grain refinement at high ΔT.

Single phase Co–Pd alloy system has a low enthalpy of fusion and a narrow liquidus–solidus interval.19 The hypercooling limit of this alloy system is relatively low and in the order of 300 K,19 which can be easily reached at a reduced extent of ΔT.20 Therefore, this system can be a suitable model system to clarify the mechanisms of grain refinements in single phase alloy systems. Since when the alloy is hypercooled, the remelting effect involved in the post-recalescence period will be excluded. Previous investigations on the undercooled Co–Pd alloys focused mainly on the hypercooling behaviour and the paramagnetic–ferromagnetic transition at Curie temperature,21,22 but the microstructure evolution of these alloys as a function of ΔT has not been systematically reported yet.

In the present work, Co₈₀Pd₂₀ alloy was selected as a model alloy and was undercooled by fluxing method. The maximum undercooling up to 340 K was achieved. The evolution of solidification structures was systematically investigated. The formation mechanisms of the solidification structures were investigated in detail.

Experimental

Alloy specimens, each weighing ∼5 g, were prepared by in situ melting pure Co pieces (99·9 purity) and pure Pd pieces (99·95 purity) under the protection of Ar. Afterwards, a high purity quartz crucible containing an alloy specimen and ∼1·5 g dehydrated B₂O₃ was placed in the centre of an induction coil. The melting process was conducted in a vacuum chamber. The vacuum chamber was first evacuated to a pressure of 2×10⁻³ Pa, and then ultrapure Ar was backfilled into the chamber until the pressure of the chamber reached 0·05 Pa. The alloy was melted and cooled in the chamber cyclically until a desired undercooling was achieved. The temperature of the specimen was monitored by an infrared pyrometer with an accuracy of ± 5 K and a response time of 10 ms. In order to investigate the recrystallisation mechanism of grain refinement, some hypercooled specimens were quenched rapidly into Ga–In liquid alloy after recalescence. The as solidified microstructures were revealed by etching the polished specimens with a mixture of nitric acid water solution (30 mL HCl+10 mL HNO₃+40 mL H₂O). Then, the microstructures were observed by optical microscopy (Olympus GX71). The microstructures were further studied using transmission electron microscopy (TEM, Technai F30 G2, 300 kV) in order to show the morphologies of the substructures, e.g. dislocations. The TEM specimens were prepared by the standard procedures of mechanical grinding and ion milling.

Results and discussion

Recalescence characteristics and achievement of hypercooling in Co₈₀Pd₂₀ alloy melts

Three typical temperature profiles corresponding to the solidification of Co₈₀Pd₂₀ alloy melts in different undercooling regimes are shown in Fig. 1. It was found that when ΔT<265 K, the post-recalescence period presents in the temperature profiles, as seen in the plateau marked by the plateau duration time (Δt_pl) in Fig. 1a. The plateau duration time Δt_pl is the time required for the residual liquid to transform into solid after recalescence. Δt_pl can be obtained by subtracting the time point of recalescence from the time point at which the first derivative of cooling curve after recalescence abruptly changes. When ΔT reaches 265 K, the post-recalescence period does not appear, and the maximum recalescence temperature is lower than the solidus temperature of this alloy (see Fig. 1b and c). Since the post-recalescence period will only vanish when the hypercooling limit of a metal is reached,1 the hypercooling limit of the Co₈₀Pd₂₀ alloy can be determined as 265 K.

Figure 1

Measured temperature profiles of Co₈₀Pd₂₀ alloys undercooled by a 135 K, b 265 K and c 340 K: insets in b and c show magnified recalescence peaks

Evolution of solidification microstructures

Within the range of the achieved ΔT, the solidification microstructures show several morphological transitions, as seen in Fig. 2. When ΔT is smaller than 50 K, the microstructure consists of well developed dendritic skeleton, as seen in Fig. 2a. While ΔT falls into the range of 50 K<ΔT<265 K, a spherical grained microstructure forms, accompanying an abrupt reduction in grain size down to 20–30 μm (see Figs. 2b and c and 3). Further increasing ΔT into a narrow slit of 265 K<ΔT<280 K, an elongated grained microstructure appears, leading to a rise of grain size up to 120 μm, as seen in Figs. 2d and 3. Compared to the dendritic microstructure in the range of ΔT<50 K, the elongated grains observed in this range are not well developed and are mainly dendrite trunks. As ΔT exceeds ΔT_hyp, the second grain refinement occurs (see Fig. 3). The microstructures of the hypercooled alloys are characterised by equiaxed grains with rather straight grain boundaries, as seen in Fig. 2f. Except for the rather elongated grained microstructure observed in a narrow slit of 265 K<ΔT<280 K, these results are basically in agreement with Willnecker et al.'s results.21 The experimental methods in the present study and in Willnecker et al.'s study are the same, i.e. fluxing method. The elongated grained microstructure observed in the narrow slit of 265 K<ΔT<280 K is a new finding for the solidification structure evolution of Co₈₀Pd₂₀ alloys.

Figure 2

Optical micrographs of Co₈₀Pd₂₀ alloys undercooled by a 20 K, b 85 K, c 160 K, d 265 K, e 290 K and f 340 K

Figure 3

Measured average grain size as function of ΔT

Formation mechanisms of solidification microstructures

Microstructure formation of alloys with ΔT<ΔT_hyp

When an alloy melt solidifies at a small undercooling, e.g. ΔT <50 K in the case of undercooled Co₈₀Pd₂₀, the advancement of solid/liquid (S/L) interface is controlled by solute diffusion ahead of the S/L interface. Solidification of the melt proceeds rather slowly and results in the formation of well developed dendritic microstructures23 (see Fig. 2a). With increasing undercooling, a grain refinement event is observed in the undercooled Co₈₀Pd₂₀ alloy at 50–265 K. The first grain refinement in an undercooled single phase alloy is generally believed to be caused by the remelting effect during the post-recalescence period. In order to demonstrate the remelting effect during post-recalescence period, Karma15 proposed a model describing the relationship between the necessary breakup time (Δt_bu) of dendrite and the duration time of post-recalescence period (Δt_pl). According to Karma's model, when Δt_bu<Δt_pl, the dendrite skeleton formed on recalescence can be broken up completely and vice versa. Δt_bu can be calculated according to the following equation15 (1) where α_L is the thermal diffusivity in liquid, D is the solute diffusivity in liquid, m_L is the equilibrium liquidus slope, k_e is the equilibrium solute partition coefficient, C₀ is the solute concentration of the alloy, d₀ ( = ΓC_p/ΔH_f) is the capillary length, ΔH_f is the heat of fusion, C_p is the specific heat of the liquid and Γ is the Gibbs–Thomson coefficient. R is the dendrite trunk radius that is proportional to the dendrite tip radius R_tip,24 which can be calculated by the dendrite growth model proposed by Boettinger et al.25 The value of R/R_tip is ∼15, which was determined by comparing the measured R from the micrograph to the calculated R_tip at a same ΔT, as stated in Ref. 24. Employing the physical parameters listed in Table 1, Δt_bu can be calculated by equation (1). Δt_pl can be measured directly from the recorded temperature profiles upon solidification, e.g. Fig. 1a. The calculated Δt_bu and the measured Δt_pl as a function of ΔT are shown in Fig. 4. It can be seen that Karma's model predicts that there would be two grain refinements happening in the range of ∼40 K<ΔT<70 K and 140 K<ΔT<265 K for the undercooled Co₈₀Pd₂₀ alloy. Apparently, the refined structure observed in the range of 70 K<ΔT<140 K cannot be explained by this model. It should be noted that besides the effect of the liquid/solid interface tension induced structural instability as proposed by Karma,15 the other remelting effect that is due to chemical superheating, as suggested by Li et al.,7 should as well play a crucial role in grain refinement. Since after recalescence the temperature of the system is raised to a temperature between solidus and liquidus of the phase diagram, the solid formed in recalescence will therefore be subjected to partial remelting, which will cause the occurrence of the grain refinement.7 According to Li et al.'s model,7 the remelted fraction of the solid formed in recalescence can be calculated by the following equation (2) where T_R is the maximum recalescence temperature, T_s is the equilibrium solidus temperature corresponding to the compositions C_s of the central part in the dendrite trunks and ΔT₀ is the equilibrium crystallisation temperature interval of the alloy with compositions C_s. Assuming solidification within recalescence to be an adiabatic process,1 T_R can be calculated according to the laws of conservation of mass and energy.26 On this basis, f_L can be calculated by equation (2) with the physical parameters listed in Table 1, as seen in Fig. 5. According to the calculated result, in the range of 70 K<ΔT <140 K, f_L is rather large and could be responsible for the formation of the refined spherical grained structure. On this basis, the grain refinement observed in the wide range of 50 K<ΔT<265 K can be ascribed to a combined effect of breakup of dendrite skeleton driven by solid–liquid interfacial tension and chemical superheating induced dendrite remelting.

Figure 4

Δt_bu (solid line) of Co₈₀Pd₂₀ calculated by Karma's model15 and Δt_pl versus ΔT

Figure 5

Calculated remelted fraction of primary dendrite at highest recalescence temperature for Co₈₀Pd₂₀ alloys

Table 1

Physical parameters used for Co₈₀Pd₂₀ alloys13,19

Parameter	Value
Heat of fusion ΔH_f/kJ mol⁻¹	12·11
Specific heat of the liquid C_p/J⁻¹ mol⁻¹	49·2
Dynamic viscosity of the liquid μ/Pa s	6×10⁻³
Molar volume of the liquid /m³ mol⁻¹	8·13×10⁻⁶
Molar volume of the solid /m³ mol⁻¹	7·25×10⁻⁶
Liquidius T_L/K	1610
Solidus T_S/K	1563
Solute diffusivity in the liquid D_L/m² s⁻¹	10⁻⁹
Thermal diffusivity α_L/m² s⁻¹	6×10⁻⁶
Atomic space a₀/m	3·0×10⁻¹⁰
Interfacial energy σ_S,L/J m⁻²	0·25
Speed of sound in the melt V₀/m s⁻¹	4000
Equilibrium liquidus slope m_L/K at-⁻¹	−6·5
Equilibrium solute partition coefficient k_e	0·66
Solidification time t_f/s	0·1
Size of the mushy zone a/m	0·01
Gibbs–Thomson coefficient Γ	2·7×10⁻⁷
Solid fraction at the dendrite coherency point	0·15
Solidification shrinkage of the primary phase β_s	0·1082
Atomic diffusive speed V_D/m s⁻¹	10

Microstructure formation of hypercooled alloys

When ΔT reaches the range of 265–280 K, which is above the hypercooling limit, due to a zero Δt_pl and a zero f_L, the dendrite trunks cannot be remelted and therefore can be preserved in the solidification microstructure, leading to the presence of the elongated grained structure. The appearance of elongated grained structure in the hypercooled Co₈₀Pd₂₀ alloy is apparently ascribed to the vanished remelting effect in the post-recalescence period. This is, in principle, in agreement with Herlach et al.'s explanation.16

As ΔT is higher than 280 K, the second grain refinement event occurs. The equiaxed grains with rather straight grain boundaries (see Fig. 2e and f) strongly suggest that the formation of the solidification microstructure is due to recrystallisation. The recrystallised microstructures in hypercooled alloys have been observed by Willnecker et al. in Co₈₀Pd₂₀ alloy21 and by Lu et al. in Ni₇₅Pd₂₅ alloy.14 In the present work, the occurrence of recrystallisation in the hypercooled Co₈₀Pd₂₀ alloy with ΔT>280 K was considered to be induced by the accumulation of internal strain/stress caused by the dramatic shrinkage stress resulting from rapid solidification. For rapid solidification taking place at a rather high undercooling, the advance of the solid/liquid interface proceeds extremely fast. As a consequence, a large volume contraction owing to rapid solidification will cause a high shrinkage stress developing in the system.13 Liu and Yang13 proposed a model to calculate the shrinkage stress developed during rapid solidification. According to this model, this shrinkage stress can be expressed by the following equation (3) where . g_s is the volume fraction of primary solid, is the solid fraction after the dendrite coherency point during recalescence, g_l is the volume fraction of liquid, μ is the dynamic viscosity of the liquid, a is the size of the mushy zone in which primary growth or recalescence occurs, λ₂ is the secondary dendrite arm spacing that can be measured experimentally and t_f is the solidification time. β_s is the solidification shrinkage of the primary dendrite, which can be calculated by the following equation , where and are molar volumes of the liquid and solid respectively. Assuming the recalescence period is adiabatic and that the specific heats of the solid and liquid are constant and equal, the solid fraction formed in recalescence can be calculated by ,26 and T_n is the nucleation temperature. Using the essential parameters listed in Table 1, as a function of ΔT can be calculated by equation (3) (see Fig. 6). It can be seen from Fig. 6 that, with increasing ΔT, continuously increases. The yielding strength of Co₈₀Pd₂₀ alloy at 1093°C is ∼83 MPa,27 which corresponds to the value marked by the dash line in Fig. 6. Once exceeds such a value, the dendrites formed will be plastically deformed, leading to a storage of microstrain in the microstructure. As is known, the stored microstrain will act as the driving force for recrystallisation.28 Further increasing ΔT causes a continuous increase in the stored microstrain, resulting in a continuous increase in driving force for recrystallisation. Once the driving force is high enough to initiate the nucleation, the recrystallisation will take place. As a consequence, a refined recrystallised microstructure can be resultant. Since the annealing twin is a typical characteristic of the recrystallised metal with fcc structure, the appearance of annealing twins in the microstructure (Fig. 7a) further evidences the occurrence of recrystallisation.28

Figure 6

Calculated stress accumulated in dendrite skeleton upon rapid solidification as function of initial undercooling: dash line represents yielding strength of Co₈₀Pd₂₀ alloys

Figure 7

Optical images of microstructure of hypercooled 340 K Co₈₀Pd₂₀ alloys

The driving force for recovery and recrystallisation is an extra free energy of the cold worked state that is difficult to be measured. However, a major part of the stored energy will be attributed to the introduction of extra dislocations in the system.29 In order to provide more reliable evidence for the recrystallisation mechanism of grain refinement at high undercooling for the alloy, a hypercooled specimen with ΔT = 340 K was quenched into Ga–In liquid alloy immediately after recalescence. Interestingly, compared to the specimen with the same ΔT, which was cooled to room temperature naturally, the quenched specimen exhibits an even coarser solidification microstructure, as seen in Fig. 7a and b. Apparently, the quenched specimen is as well subjected to recrystallisation, since the solidification microstructure shows similar characteristics with the specimen subjected to natural cooling. For example, the solidification microstructure also consists of straight grain boundaries and twins; however, different from the naturally cooled specimen that shows very few dislocations (see Fig. 7e), rather dense dislocation networks were observed in the quenched specimen (see Fig. 7c and d). In the quenched specimen, a rather high cooling rate suppresses the completion of the recrystallisation. In this case, a large fraction of dislocations in the quenched specimen cannot be consumed by recrystallisation; therefore, the dense dislocation networks can be preserved in the microstructure. Owing to the partial recrystallisation, a coarser solidification microstructure compared to the specimen subjected to natural cooling can be resultant. This experiment provides additional evidence for the occurrence of recrystallisation in the specimen with ΔT>280 K and proves that the second grain refinement event at high undercooling in this alloy should be ascribed to the recrystallisation rather than dendrite remelting.

Conclusions

Co₈₀Pd₂₀ alloy was undercooled by means of fluxing method. The maximum undercooling of 340 K was achieved in the present work. Two grain refinement events of the solidification microstructures were observed in the ranges of 50–265 K and ΔT>280 K. The grain refinement in the range of 50–265 K is attributed to a combined effect of breakup of dendrite skeleton driven by solid–liquid surface tension induced structural instability and chemical superheating induced dendrite remelting. The second grain refinement at ΔT>280 K is ascribed to hypercooling and the recrystallisation of the rapidly solidified solid.

Footnotes

Acknowledgements

The authors are grateful to the financial support of the Free Research Fund of State Key Laboratory of Solidification Processing (09-QZ-2008 and 24-TZ-2009), the Natural Science Foundation of China (51071127 and 51134011), the Fundamental Research Fund of Northwestern Polytechnical University (JC200801), China National Funds for Distinguished Young Scientists (51125002) and National Basic Research Program of China (973 Program, 2011CB610403), and the 111 project B08040.

References

Herlach

: ‘Non-equilibrium solidification of undercooled metallic melts’, Mater. Sci. Eng. R , 1994, R12, 177–272.

Liu

, Yang

: ‘Rapid solidification of highly undercooled bulk liquid superalloy: recent developments, future directions’, Int. Mater. Rev. , 2006, 51, 145–170.

, Liu

, Yang

: ‘Grain refinement in solidification of highly undercooled eutectic Ni–Si alloy’, Mater. Lett. , 2007, 61, 987–990.

Zhang

, Liu

, Wang

, Yang

: ‘Grain refinement in highly undercooled solidification of Ni₈₅Cu₁₅ alloy melt: direct evidence for recrystallisation mechanism’, Scr. Mater. , 2010, 63, 43–46.

Yang

, Yu

, Wang

: ‘Application of dendrite fragmentation to fabricate the homogeneous dispersed structure in undercooled Cu–Co immiscible alloy’, J. Alloys Compd , 2011, 509, 9675–9678.

Zhou

, Li

: J. Alloys Compd , 2008, 461, 113–116.

, Liu

, Lu

, Yang

, Zhou

: ‘Structural evolution of undercooled Ni–Cu alloys’, J. Cryst. Growth , 1998, 192, 462–470.

Horvay

: ‘The tension field created by a spherical nucleus freezing into its less dense undercooled melt’, Int. J. Heat Mass Transfer , 1965, 8, 195–243.

Mullis

, Cochrane

: ‘Grain refinement and the stability of dendrites growing into undercooled pure metals and alloys’, J. Appl. Phys. , 1997, 82, 3783–3797.

10.

Powell

GLF

, Hogan

: Trans. TMS-AIME , 1968, 242, 2133–2138.

11.

Powell

GLF

, Hogan

: Trans. TMS-AIME , 1969, 245, 407–412.

12.

Powell

GLF

: Comments on ‘Undercoolability of copper bulk samples’, J. Mater. Sci. Lett. , 1991, 10, 745–746.

13.

Liu

, Yang

: ‘Stress-induced recrystallisation mechanism for grain refinement in highly undercooled superalloy’, J. Cryst. Growth , 2001, 231, 295–305.

14.

, Li

, Zhou

: ‘Grain refinement in the solidification of undercooled Ni–Pd alloys’, J. Cryst. Growth , 2007, 309, 103–111.

15.

Karma

: ‘Model of grain refinement in solidification of undercooled melts’, Int. J. Non Equilibr. Process , 1998, 11, 201–233.

16.

Herlach

, Eckler

, Karma

, Schwarz

: ‘Grain refinement through fragmentation of dendrites in undercooled melts’, Mater. Sci. Eng. A , 2001, A304–A306, 20–25.

17.

Wang

, Liu

, Yang

: J. Mater. Res. , 2010, 25, 1963–1974.

18.

, Zhou

, Yang

: ‘Effect of solidification time on the structural evolution of undercooled single phase alloys’, J. Cryst. Growth , 1999, 206, 141–146.

19.

Volkmann

, Wilde

, Willnecker

, Herlach

: ‘Nonequilibrium solidification of hypercooled Co–Pd melts’, J. Appl. Phys. , 1998, 83, 3028–3034.

20.

Wilde

, Görler

, Willnecker

: ‘Hypercooling of completely miscible alloys’, Appl. Phys. Lett. , 1996, 69, 2995–2997.

21.

Willnecker

, Görler

, Wilde

: ‘Appearance of a hypercooled liquid region for completely miscible alloys’, Mater. Sci. Eng. A , 1997, A226–A228, 439–442.

22.

Reske

, Herlach

, Keuser

: ‘Evidence for the existence of long-range magnetic ordering in a liquid undercooled metal’, Phys. Rev. Lett. , 1995, 75, 737–739.

23.

Liu

, Guo

, Yang

: Mater. Res. Bull. , 2001, 36, 181–192.

24.

Schwarz

, Karma

, Eckler

, Herlach

: Phys. Rev. Lett. , 1994, 73, 1380–1384.

25.

Boettinger

, Coriell SR , Trivedi

: ‘Rapid solidification processing: principles and technologies IV’, (ed. , Mehrabian

, Parrish

P A

); 1988, Baton Rouge, Claitor's.

26.

Piccone

, Wu

, Shiohara

, Flemings

: ‘Dendritic growth of undercooled nickel–tin. II’, Metall. Trans. A , 1987, 18A, 925–932.

27.

Zhou

, Hu

, Li

: ‘Stress induced deformation in the solidification of undercooled Co₈₀ Pd₂₀ alloys’, Mater. Sci. Eng. A , 2011, A528, 973–977.

28.

Humphreys

, Hatherly

: ‘recrystallisation and related annealing phenomena’, 2nd edn; 2004, New York, Pergamon Materials Series.

29.

Christian

: ‘The theory of transformation in metals and alloys. Recovery, recrystallisation and grain growth’, 832–858; 2002, Oxford, Pergamon Press.

Evolution of solidification microstructure in undercooled Co 80 Pd 20 alloys

Abstract

Keywords

Introduction

Experimental

Results and discussion

Recalescence characteristics and achievement of hypercooling in Co80Pd20 alloy melts

Evolution of solidification microstructures

Formation mechanisms of solidification microstructures

Microstructure formation of alloys with ΔT<ΔThyp

Microstructure formation of hypercooled alloys

Conclusions

Footnotes

Acknowledgements

References

Recalescence characteristics and achievement of hypercooling in Co₈₀Pd₂₀ alloy melts

Microstructure formation of alloys with ΔT<ΔT_hyp