Materials and Design 84 (2015) 48–52

Contents lists available at ScienceDirect

Materials and Design

j ourna l homepage: www.e lsev ie r .com/ locate / jmad

Electrical and thermal conductivity in Mg–5Sn Alloy at differentaging status

Chunming Wang a, Zhiming Cui b, Hongmei Liu c, Yungui Chen a,b,⁎, Wucheng Ding a, Sufen Xiao a

a College of Materials Science and Engineering, Sichuan University, Chengdu 610064, Chinab School of Aeronautics and Astronautics, Sichuan University, Chengdu 610065, Chinac College of Materials Science and Engineering, Southwest Jiaotong University, Chengdu 611756, China

⁎ Corresponding author at: School of Materials ScieUniversity, Chengdu 610065, China.

E-mail address: ygchen60@aliyun.com (Y. Chen).

http://dx.doi.org/10.1016/j.matdes.2015.06.1100264-1275/© 2015 Elsevier Ltd. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 22 April 2015Received in revised form 14 June 2015Accepted 15 June 2015Available online 26 June 2015

Keywords:Mg–Sn alloyAging treatmentMicrostructureElectrical conductivityThermal conductivity

The electrical conductivity, thermal conductivity and its relationship with the microstructure in Mg–5Sn alloyaged at 513 K for different aging times were investigated systematically in this paper. The results show thatthe electrical conductivity and thermal conductivity obviously increase with the increasing aging time, and itsvalues increase from 10.25 × 106 S·m−1 to 13.7 × 106 S·m−1, 87.5 W·m−1·K−1 to 122 W·m−1·K−1 afteraging treatment for 120 h, respectively. Meanwhile, it is found that there exist quite different relationships be-tween unit cell volume and thermal conductivity in early and later aging stages.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Magnesium alloys are the lightest weight commercially availablestructural materials, which have high specific strength and stiffness,good thermal and electrical conductivity [1,2]. Therefore, magnesiumalloys are attractive candidates for heat dissipation materials in LEDetc., compared with aluminum alloys. The Mg–Sn alloys, known as atypical precipitation hardening system [3], have great potential applica-tions in heat dissipation materials at elevated temperatures (b423 K)due to the Mg2Sn phase, which has a high melting temperature of1043 K [4–6]. For Mg–Sn alloys, most researchers have given more at-tention to the improvement of aging hardening on the performance ofthe alloys [7–10]. Thermal conductivity is an important thermophysicalproperty for heat dissipation materials. The higher the thermal conduc-tivity, the more effective the cooling is [11]. Some researchers havestudied the thermal conductivity of as-cast and as-extruded Mg-basedalloys [11–18]. A. Rudajevova et al. have found that thermal conductiv-ity is sensitive to the microstructure of as-cast Mg alloys [11–14]. M.Y.Zheng et al. recently have studied the thermal conductivity of as-castand as-extrudedMg–Al/Zn alloys [15,16] and found that the anisotropyof thermal conductivity resulted from microstructure anisotropybecause of the texture formed during extrusion [17]. The phasetransformation and the different effects of Zn and Al on the thermal/

nce and Engineering, Sichuan

electrical conductivity have also been studied by F.S. Pan et al. [18]. Itis well known that the conductivity of the aging-treated materials isbetter than that of as-cast and solution-treated materials. This is mainlydue to the precipitation of the dissolved elements out of the solutions.However, few papers pay attention to the thermal properties of agedMg-based alloys, especially the relationship between the precipitatesand thermal conductivity. The paper mainly focuses on the effects ofthe aging status on the electrical conductivity and thermal conductivityof aged Mg–Sn alloys, and tries to discuss the relationship between theprecipitates and thermal conductivity for different aging status.

2. Experimental

The Mg–5 wt.% Sn alloys were prepared from high purity Mg(99.95%) and pure Sn (99.98%), melted in a low-carbon steel crucibleunder the protection of N2 + SF6 mixed gas. The melt was stirred toensure homogeneity and held at 993 K for about 30 min, and thencast in steel molds preheated up to 533 K. The specimens for electricaland thermal conductivity measurement were cut into slices with14.0 mm × 14.0 mm × 3.0 mm and discs with a diameter of 12.7 mmand a thickness of 2.88 mm, respectively. The samples were solutiontreated for 28 h at 733 K and quenched intowater at room temperature.Subsequently, the specimenswere isothermally aged at 513 K for differ-ent times.

The samples for microstructural characterization were polishedand etched with a solution of 4 vol.% nital. The microstructures andphase characterizations were investigated by scanning electron

Fig. 1. XRD patterns of Mg–5Sn alloys with different aging times at 513 K, (a) 0 h, (b) 6 h,(c) 20 h, (d) 64 h, and (e) 120 h.

Table 1The lattice parameters of α-Mg phase of the aged Mg–5Sn alloys.

Aging time,h

Lattice parameters,nm

Volume,nm3

Mg2Sn,vol. %

a c

0 0.3211 0.5209 7.752 × 10−3 –6 0.3210 0.5208 7.746 × 10−3 –20 0.3208 0.5207 7.735 × 10−3 0.364 0.3208 0.5205 7.732 × 10−3 1.5120 0.3207 0.5203 7.724 × 10−3 2.9

49C. Wang et al. / Materials and Design 84 (2015) 48–52

microscopy (SEM, JEOL, JSM-6490LV) and X-ray diffraction (XRD,DanDongFangYuan, DX-2600) with Cu Kα radiation, respectively.The precipitate microstructures after aging treatment were observed

Fig. 2. SEM images of Mg–5Sn alloys after aging treatment

using transmission electron microscopy (TEM, JEM-2010UHR)equipped with Energy Dispersive X-ray (EDS, Noran Vantage DS)operating at 200 kV. The foils were prepared by ion-milling usingthe Precision Ion Polishing System (GATAN691). The age hardeningresponses were measured by the Vickers hardness tester (HVS-1000)under a load of 25 g.

The electrical conductivity of the aging samples wasmeasured by aneddy-current device (FIRST FD-101) at room temperature. The probe(probe diameter,Φ: 8 mm) of the device was put in the smooth surfaceof the samples, and led to eddy current by forming the loop. Accordingto the International Annealed Copper Standard (% IACS) with a±1% ac-curacy, the range of electrical conductivity measurements was from6.9% IACS (4.0 MS·m−1) to 110% IACS (64 MS·m−1). The thermaldiffusivity was measured at room temperature (298 K) with aNETZSCH model LFA447 Flash Analyzer. The surface of the specimendiscs was blackened by carbon-coating in order to improve the absorp-tion of the light pulse. The density of the samples at room temperaturewas determined by the Archimedes method. The specific heat capacityof the alloy was calculated using the Neumann–Kopp rule [19,20].Thus, the thermal conductivity (λ) of the samples was calculated bythe following equation [21]:

λ ¼ αρc ð1Þ

where λ is the thermal conductivity, α is the thermal diffusivity, ρ is thedensity and c is the specific heat capacity.

3. Results and discussion

3.1. The microstructure of aged Mg–5Sn alloy

The XRD patterns of Mg–5Sn alloys with different aging times at513 K are shown in Fig. 1. According to the indexed results, there areonly α-Mg (PDF: No. 35-0821) and Mg2Sn (PDF: No. 07-0274) phasesin aged Mg–5Sn alloy, and the diffraction peak intensities of the

for 6 h (a), 20 h (b), 64 h (c) and 120 h (b) at 513 K.

Fig. 3. Bright-field TEMmicrographs of aged Mg–5Sn alloys along the [0001] zone at 513 K for 2 h (a) and 16 h (b).

50 C. Wang et al. / Materials and Design 84 (2015) 48–52

Mg2Sn phase gradually increase with the increasing aging time, indicat-ing that the amount of Mg2Sn increases with aging time increasing.Meanwhile, the dependence of the lattice parameters of the α-Mgphase on different aging times is given in Table 1. The values of a andc of the lattice parameters of α-Mg phase gradually decrease with theincreasing aging time. The values of a and c of the lattice parametersof pure Mg are 0.3202 nm and 0.5199 nm, respectively, and the valueof lattice volume is 7.694 × 10−3 nm3 [22]. The lattice parameters (a,c) ofα-Mg inMg–5Sn alloy are closer to pureMg and the lattice volumedecreases from 7.752 × 10−3 nm3 to 7.724 × 10−3 nm3 after a 120 haging treatment. Meanwhile, the volume fraction of the Mg2Sn phaseis calculated by the Rietveld refinement, the value of the Mg2Sn phaseranges from 0 vol.% to 2.9 vol.% at different aging times.

Fig. 2 shows the SEM images of Mg–5Sn alloys after aging treatmentfor 6 h, 20 h, 64 h and 120 h at 513 K, respectively. As can be seen fromFig. 2, the amount of the Mg2Sn phase increases with the increase ofaging time, and the size of Mg2Sn particles also grows gradually,which can be obviously seen in Fig. 2(c, d). The above results are consis-tent with Fig. 1. To observe further the morphology of precipitates in ashort time after aging treatment, the bright-filed TEM micrographs ofaged Mg–5Sn alloys for 2 h (a) and 16 h (b) are shown in Fig. 3. As canbe seen from Fig. 3, the precipitates are not observed in Fig. 3(a), and itis found that the morphology of the precipitates tends to be rod- andplate-shaped in Fig. 3(b), the results are consistent with Refs. [8,9].

3.2. Electrical and thermal conductivity of aged Mg–5Sn alloy

Fig. 4 shows the electrical conductivity of Mg–5Sn alloy with differ-ent aging times at 513 K. The electrical conductivity in aged Mg–5Sn

Fig. 4. Electrical conductivity of Mg–5Sn alloy after different aging times at 513 K.

alloy increases with the increasing aging time. The values of electricalconductivity of aged Mg–5Sn alloys increase from 10.25 × 106 S·m−1

for aging treatment for 0 h to 13.7 × 106 S·m−1 for aging treatmentfor 120 h. Meanwhile, the growth rate of the electrical conductivity be-comes slower with aging time increasing, suggesting that thedesolventizing driving force of Sn atoms in the α-Mg matrix decreases.

The measured thermal diffusivity of Mg–5Sn alloys after differentaging times is shown in Fig. 5. According to the Neumann–Kopp rule,the corresponding specific heat capacity (Cp) is shown as the following[20]:

Cp Tð Þ ¼X

Cp;i Tð Þxi ð2Þ

where Cp,i is the specific heat capacity of the i pure component; xi is theatomic ratio of the i pure component. The specific heat capacity of thealloy is acquired by Eq. (2), and the value is 1.026 J·g−1·K−1 at roomtemperature where the values of specific heat capacity are plugged in,24.9 J·K−1·mol−1 for Mg and 26.989 J·K−1·mol−1 for Sn at 298 K[23]. It is thought that the specific heat capacity of the alloys is almostunchanged through the Neumann–Kopp rule due to the invariablealloy composition. Therefore, it is negligible that the different agingtimes have an influence on the specific heat capacity. According toEq. (1), the thermal conductivity of the samples is calculated at roomtemperature by introducing the values of density (1.818 g·cm−3),specific heat capacity (1.026 J·g−1·K−1) and thermal diffusivity inFig. 5, which is also shown in Fig. 5. As can be seen from Fig. 5, the ther-mal diffusivity and thermal conductivity increase with the aging timeincreasing. The thermal conductivity of aged Mg–5Sn alloys increasesfrom 87.5 W·m−1·K−1 for aging 0 h to 122 W·m−1·K−1 for aging120 h. The thermal conductivity increases by 39%, compared with Mg–

Fig. 5. Thermal diffusivity and conductivity of Mg–5Sn alloys after different aging times at513 K.

Fig. 6. The relationship between the unit cell volume and thermal conductivity of α-Mg.

51C. Wang et al. / Materials and Design 84 (2015) 48–52

5Sn alloys after solution treatment. The Mg2Sn precipitates increasewith the increasing aging time, as shown Figs. 1 and 2. It is suggestedthat the Sn solute atoms in Mg matrix decrease gradually, which makesthe lattice parameters ofα-Mg tend to be closer to pureMg, andweakensthe lattice distortion of the α-Mg matrix, and result in the improvement

Fig. 7. TheHRTEM image of the interface between rod-shapedmorphology (a), plate-shapedmo(b), the edge dislocation of the interface between rod-shaped precipitates and α-Mg matrix (c

of the thermal conductivity of Mg–5Sn alloys improves. Meanwhile, itis implied that the lattice distortion has significant influences on the ther-mal conductivity compared with the precipitates in Mg alloys.

3.3. The relationship between the microstructure and thermal conductivityin aged Mg–5Sn alloys

To elucidate the effect of different aging times on the thermalconductivity of the α-Mg matrix, the thermal conductivity of the α-Mg matrix (λm) is calculated simply by the Maxwell model [24]. TheMaxwell model for the thermal conductivity (λ) of aged Mg–5Sn alloyscan be written in the following form:

λ ¼ λm

2Vdλd

λm−1

� �þ λd

λmþ 2

� �

1−λd

λm

� �Vd þ

λd

λmþ 2

2664

3775 ð3Þ

where λm is the thermal conductivity of the continuous matrix (α-Mgmatrix at different aging times), λd is the thermal conductivity ofuniformly distributed dispersions (Mg2Sn precipitates), and Vd is thevolume fraction of dispersions (Mg2Sn particle). According to the resultsabout the total thermal conductivity (λ) and the volume fraction ofMg2Sn phases (Vd), the thermal conductivity of the α-Mg matrix (λm)is acquired by Eq. (3) at different aging times, where the thermalconductivity of Mg2Sn is 6.04 W·m−1·K−1 [25].

The relationship between the unit cell volume of α-Mg and thethermal conductivity (λm) of α-Mg is shown in Fig. 6. The thermalconductivity (λm) increases with the unit cell volume of α-Mg

rphology (d) andα-Mgmatrix at 513 K for 16 h aging, HRTEM image by the FFT transform).

52 C. Wang et al. / Materials and Design 84 (2015) 48–52

decreasing.When the aging time ranges from 20 h to 120 h, the volumeshrinkage of unit cell is about 0.14%, the thermal conductivity (λm)increases by 36.4% correspondingly, and namely, the thermal conduc-tivity (λm) increases by 26% per shrinkage of thousandth. When theaging time ranges from 0 h to 20 h, the volume shrinkage of unit cellis about 0.21%, the thermal conductivity (λm) only increases by 8.9%correspondingly, and namely, the thermal conductivity (λm) increasesby 4.2% per shrinkage of thousandth. Meanwhile, the variation inVickers hardness (HV0.25) as a function of aging time at 513 K isshown in the inset of Fig. 6. The hardness increases with the increasingaging time, and reaches the peak hardness for 16 h.

Combined with the dependence of Vickers hardness on aging time inFig. 6, itmay imply that the agingmechanism is different after aging treat-ment for 16 h, which may result from the different slope of thermalconductivity aged Mg–5Sn alloys after aging treatment for 20 h. Heneset al. have reported that the Mg2Sn in the binary Mg–Sn alloy has theorientation relationships of (0001)α‖(110) β, [1120]α‖[001]β and(0001)α‖(110)β, [1120]α‖[111]β, at an aging temperature of 403 to473 K and (0001)α‖(111)β, [1120]α‖[112]β and (0001)α‖(111)β,[1120]α‖[101]β at an aging temperature of 473 to 573 K [26]. In general,the distortion caused by the orientation relationship is more seriousthan that caused by an incoherent relationship, due to the interface rela-tionship shifts from the coherent towards the incoherent with the in-creasing aging time. Therefore, the HV hardness corresponding to thedistortion energy caused by the orientation relationship at 513 K in-creaseswith aging time increasing until the peak and then drops [27]. Ac-cording to themorphology of aged-Mg–5Sn alloy for 16 h in Fig. 3(b), theHRTEMmicrograph is showed in Fig. 7, including theHRTEM image of theinterface between rod-shapedmorphology (a), plate-shapedmorphology(d) andα-Mgmatrix at 513 K for aging 16 h. The rod-shaped precipitatesare semi-coherent with theα-Mgmatrix along the [220] and [101] direc-tions in Fig. 7(a). Fig. 7(b) shows the HRTEM image by the FFT trans-formed with a series of parallel edge dislocation (Fig. 7(c)).Fig. 7(d) shows the incoherent interface between the plate-shaped pre-cipitates and α-Mg matrix, which has a transition layer of disorder. It isimplied that the nanoscale Mg2Sn particles distributed homogeneouslyin the α-Mg matrix are coherent with the α-Mg matrix before the peakhardness, which leads to the relatively serious lattice distortion of theα-Mg matrix around the Mg2Sn particles, and has an obvious influenceon the thermal conductivity of Mg–5Sn alloys. After the peak hardness,the Mg2Sn particles grow gradually in size and are incoherent with α-Mg, which weakens the distortion of the α-Mg matrix around theMg2Sn particles, so the thermal conductivity increases obviously.

4. Conclusions

1). The electrical conductivity and thermal conductivity ofMg–5Sn alloysaged at 513 K increase with the increasing aging time. The value ofwhich increases from 10.25 × 106 S·m−1 to 13.7 × 106 S·m−1,87.5W·m−1·K−1 to 122W·m−1·K−1 after aging120h, respectively.

2). The thermal conductivity and the amount of precipitate Mg2Snincrease with the increasing aging time. It is implied that the latticedistortion has a significant influence on the thermal conductivity,compared with the precipitates in Mg alloys.

3). The coherent relationship of Mg2Sn precipitates with the α-Mgmatrix has an obvious influence on the thermal conductivity of theα-Mg matrix.

References

[1] B.L. Mordike, L. Ebert, Magnesium: properties — application — potential, Mater. Sci.Eng. A 302 (2001) 37–45.

[2] H.C. Pan, F.S. Pan, R.M. Yang, J. Peng, C.Y. Zhao, J. She, Z.Y. Gao, A.T. Tang, Thermaland electrical conductivity of binary magnesium alloys, J. Mater. Sci. 49 (2014)3107–3124.

[3] A.A. Nayeb-Hashemi, J.B. Clark, Phase Diagrams of Binary Magnesium Alloys, ASMInternational, Materials Park, OH, 1988.

[4] P. Poddar, K.L. Sahoo,Microstructure andmechanical properties of conventional castand rheocast Mg–Sn based alloys, Mater. Sci. Eng. A 556 (2012) 891–905.

[5] G. Nayyeri, Mahmudi, Effects of Ca additions on the microstructural stability andmechanical properties of Mg–5%Sn alloy, Mater. Des. 32 (2011) 1571–1576.

[6] D.Q. Fang, N.Ma, K.L. Cai, X.C. Cai, Y.S. Chai, Q.M. Peng, Age hardening behaviors, me-chanical and corrosion properties of deformed Mg–Mn–Sn sheets by pre-rolledtreatment, Mater. Des. 54 (2014) 72–78.

[7] M.A. Gibson, X. Fang, C.J. Bettles, C.R. Hutchinson, The effect of precipitate state onthe creep resistance of Mg–Sn alloys, Scr. Mater. 63 (2010) 899–902.

[8] T.T. Sasaki, K. Oh-ishi, T. Ohkubob, K. Hono, Enhanced age hardening response bythe addition of Zn in Mg–Sn alloys, Scr. Mater. 55 (2006) 251–254.

[9] T.T. Sasaki, K. Oh-ishi, T. Ohkubob, K. Hono, Effect of double aging andmicroalloyingon the age hardening behavior of a Mg–Sn–Zn alloy, Mater. Sci. Eng. A 530 (2011)1–8.

[10] B. Rashkovaa, W. Prantlb, R. Gorgl, J. Keckesb, S. Cohend, M. Bamberger, G. Dehma,Precipitation processes in a Mg–Zn–Sn alloy studied by TEM and SAXS, Mater. Sci.Eng. A 494 (2008) 158–165.

[11] A. Rudajevova, F. von Buch, B.L. Mordike, Thermal diffusivity and thermal conductiv-ity of Mg–Sc alloys, J. Alloys Compd. 292 (1999) 27–30.

[12] A. Rudajevová, M. Stanek, P. Lukác, Determination of thermal diffusivity and thermalconductivity of Mg–Al alloys, Mater. Sci. Eng. A 341 (2003) 152–157.

[13] A. Rudajevová, S. Kudela, M. Stanek, P. Lukác, Thermal properties of Mg–Li and Mg–Li–Al alloys, Mater. Sci. Technol. 19 (2003) 1097–1100.

[14] A. Rudajevová, P. Lukác, Comparison of the thermal properties of AM20 and AS21magnesium alloys, Mater. Sci. Eng. A 397 (2005) 16–21.

[15] T. Ying, M.Y. Zheng, Z.T. Li, X.G. Qiao, Thermal conductivity of as-cast and as-extruded binary Mg–Al alloys, J. Alloys Compd. 608 (2014) 19–24.

[16] T. Ying, M.Y. Zheng, Z.T. Li, X.G. Qiao, S.W. Xu, Thermal conductivity of as-cast andas-extruded binary Mg–Zn alloys, J. Alloys Compd. 621 (2015) 250–255.

[17] J.W. Yuan, K. Zhang, T. Li, X.G. Li, Y.J. Li, M.L. Ma, P. Luo, G.Q. Luo, Y.G. Hao, Anisotropyof thermal conductivity and mechanical properties in Mg–5Zn–1Mn alloy, Mater.Des. 40 (2012) 257–261.

[18] H. Pan, F. Pan, R. Yang, J. Peng, A. Tang, Q. Huang, K. Song, Z. Gao, Thermal and elec-trical conductivity of Mg–Zn–Al alloys, Mater. Sci. Technol. 30 (2014) 988–994.

[19] S. Tian, Physical Properties of Materials, Beijing University of Aeronautics andAstronautics Press, Beijing, 2004. 211–218.

[20] A. Lindemann, J. Schmidt, M. Todte, T. Zeuner, Thermal analytical investigations ofthe magnesium alloys AM 60 and AZ 91 including the melting range, Thermochim.Acta 382 (2002) 269–275.

[21] M. Yamasaki, Y. Kawamura, Thermal diffusivity and thermal conductivity of Mg–Zn–rare earth element alloys with long-period stacking ordered phase, Scr. Mater.60 (2009) 264–267.

[22] R.W. Cahn, P. Haasen, E.J. Kramer, Materials science and technology— a comprehen-sive treatment, in: K.H. Matucha (Ed.), Structure and Properties of NonferrousAlloys, vol. 8, VCH, Weinheim; New York, 1996.

[23] I. Barin, Thermochemical Data of Pure Substances, third ed. VCH Publishers, NewYork, 1995. 993, 1538.

[24] E.T. Swartz, R.O. Pohl, Thermal boundary resistance, Rev. Mod. Phys. 61 (1989)605–668.

[25] Y.S. Touloukian, R.W. Powell, C.Y. Ho, P.G. Klemens, Thermophysical properties ofmatter, Thermal Conductivity Metallic Elements and Alloys, vol. 1, Plenum Press,New York, 1970. 1317–1319.

[26] S. Henes, V. Gerold, Radiographical analysis of segregation process in Mg–Pb andMg–Sn welded alloys, Z. Met. 53 (1962) 743–751.

[27] D. Feng, Metal physics, Crystal Structures and Defects, vol. 1, Science Press, Beijing,1998. 375–466.