- Email: [email protected]
Investigating the effect of isothermal aging on the morphology and shear strength of Sn-5Sb solder reinforced with carbon nanotubes
Journal of Alloys and Compounds 649 (2015) 368e374
Contents lists available at ScienceDirect
Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom
Investigating the effect of isothermal aging on the morphology and shear strength of Sn-5Sb solder reinforced with carbon nanotubes T.T. Dele-Afolabi a, *, M.A. Azmah Hanim a, d, M. Norkhairunnisa b, d, H.M. Yusoff c, M.T. Suraya a a
Department of Mechanical and Manufacturing Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia Department of Aerospace Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia d Laboratory of Biocomposite Technology, Institute of Tropical Forestry and Forest Products (INTROP), Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 April 2015 Received in revised form 8 June 2015 Accepted 4 July 2015 Available online 14 July 2015
An analysis of the role played by the addition of carbon nanotubes (CNTs) to the solder matrix of conventional Sn-5Sb lead-free solder was performed. In a bid to determine the potential of this new solder system, the powder metallurgy approach was used to synthesise a plain Sn-5Sb solder system and CNTs reinforced composite solder formulations of Sn-5Sb-xCNT; x ¼ 0.01wt.%, 0.05wt.% and 0.1wt.%. Isothermal aging study was conducted on the solder joints, to examine the evolution of the interfacial intermetallic compound (IMC) layer between solder and the adjoining copper (Cu) substrate. Similarly, shear strength analysis was performed on as-reflow and aged solder joints. A considerable improvement in the wetting properties, the microstructural evolution, and the interfacial intermetallic compound (IMC) layer growth was observed in the composite solder joints. Owing to the excellent mechanical properties of CNTs, the shear strength assessment revealed that the composite solder joints gave a superior shear strength property, especially the Sn-5Sb-0.01CNT solder joint sample. © 2015 Elsevier B.V. All rights reserved.
Keywords: Lead-free solders Carbon nanotubes Composite solders Shear strength
1. Introduction The enactment of legislations barring the utilisation of lead in electronic products [1e3] has instigated the sensitivity of researchers towards conformity with the alternative lead-free standards for electronic systems packaging. Although rigorous efforts have been channeled towards the development of these alternative solder candidates, the Sn-5Sb solder system has received less priority despite its great potential in step soldering and high temperature applications [4e7]. However, with the recent trend of miniaturisation and the IC terminal upsurge in electronic hardwares, critical evaluation of the lifetime operational capability in interconnection joints should be of utmost concern in order to inspect and curb the reliability issues associated with the evolving lead-free solder candidates. Several attempts to enhance the performance of these solder materials has
* Corresponding author. E-mail addresses: [email protected] (T.T. Dele-Afolabi), azmah@ upm.edu.my (M.A. Azmah Hanim). http://dx.doi.org/10.1016/j.jallcom.2015.07.036 0925-8388/© 2015 Elsevier B.V. All rights reserved.
moved researchers towards the development of a composite approach in promoting the mechanical properties of interconnection joints, particularly the limitation in grain boundary sliding, creep rate mitigation and thermo-mechanical fatigue resistance [8,9]. Thus, introducing suitable reinforcement particles to the Sn5Sb solder matrix is pivotal in fabricating a revamped solder system which can thrive well under extreme service conditions and can function decently in densely packed electronic devices. Thus far, carbon nanotubes (CNTs) an outstanding nanomaterial have aroused great interest for the production of better solder joints when doped with the conventional lead-free solders due to the exceptional mechanical properties demonstrated by this group of materials. Reports from previous studies have revealed a significant enhancement in the wettability results and the microstructural evolution of the CNT reinforced solders [10e13]. Studies have shown that the IMC layer at the solder/Cu interface increased with aging time and the growth was faster for higher aging temperatures [14,15]. Ko et al. found that the IMC thickness formation of the CNT-SnAg solder was slower than the SnAg solder and the IMC growth difference between both solder samples
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
further increased as the aging time increased [16]. Furthermore, shear strength of as-reflow and aged composite Sn-Ag-Cu solder joints was improved with CNTs reinforcement [17]. Accordingly, various researchers [18e25] have adopted the single-lap joint system in conducting experimental evaluations of the shear strength property of solder materials. A bid by Han et al. [26] to develop a composite solder via the incorporation of Nicoated carbon nanotubes proved rewarding. Although, both the monolithic and the composite solder joints experienced a decline in the shear strength values as the thermal cycle increased. Nonetheless, the composite solders emerged better than the monolithic solder counterpart, irrespective of the thermal cycling period. Nearly all studies on the development of composite solders through the incorporation of carbon nanotubes have centered mainly on the SAC solder candidate. Nevertheless, the influence of CNTs in the Sn-5Sb solder system is yet to be investigated. Pitching this study on the rattling records of CNTs doped solders, it is imperative to investigate the effect of isothermal aging on the shear strength property of composite and plain Sn-5Sb solder joints. In the present study, triad formulations of the Sn-5Sb composite solder systems were synthesised. The wettability, microstructural evolution and shear strength property of the synthesised solder joints were evaluated under different aging times. 2. Experimental procedures 2.1. Material processing and sample preparation The powder metallurgy approach was used to synthesise four different solder formulations: Sn-5Sb, Sn-5Sb-0.01CNT, Sn-5Sb0.05CNT and Sn-5Sb-0.1CNT. Pure Sn and Sb micron-sized particles of 30e45 mm and 3e7 mm respectively were used as the host alloy, while multi-walled carbon nanotubes (MWCNTs) with an outer diameter of 15e20 nm was purchased from Cheap Tubes Inc, USA. For the raw materials, carbon and oxide particulates were identified from the EDS analysis as the main impurities. Using a Planetary Mono Mill, the solder formulations were dry milled for 6 h in a container with a ball to powder ratio of 20:1 at 800 rpm. After mixing, 2 g of the powder blends were compacted at 80 MPa to produce a solder pellet with 20 mm diameter and 2 mm thickness. Fig. 1(a) and Fig. 1(b) respectively present the FE-SEM micrographs of the MWCNTs used in this study and the embedded MWCNTs at the interstices between the compositional elements. 2.2. Microstructural characterization For the microstructural characterisation studies, solder joints were prepared by the reflow soldering process on a copper (Cu)
369
Fig. 2. Optical micrograph of the contact angle between the solder and the Cu substrate.
substrate. Following the reflow process, the solder joint samples were subjected to isothermal aging at 170 C for 500 h, 1000 h and 1500 h. After the aging process, the samples were cold mounted using the epoxy resin system, cross-sectioned and subjected to standard polishing procedures using a 0.05 mm Master Prep alumina suspension. Prior to the electron microscopy analysis, images of the wetting angle and microstructural morphology were taken using the optical micrograph (OM). Also, the contact angle and the interfacial IMC layer of the respective solder joints were measured using the streams essential software integrated with the OM (see Fig. 2). Finally, the solder samples were observed using the HITACHI SU8000 field emission scanning electron microscope (FESEM) equipped with energy dispersive spectroscope (EDS) to evaluate the microstructural morphology and elemental compositions respectively. 2.3. Shear test To conduct the shear strength evaluation, single-lap joint samples were utilised. The samples were fabricated in such a way as to mimic actual solder joints by filling the overlap area between the Cu substrates with solder pellets and flux application prior to the reflow process. To ensure the proper alignment of all the components that make up the lap joint setup, the samples were placed in an alumina boat jig and thus subjected to the reflow soldering process. The geometry of the single-lap joint sample is shown in Fig. 3. As pictured, each of the copper substrates has a dimension of 52 11 1 mm3 and a lap joint area of 10 11 mm2. After the as-reflow soldering operation, some of the samples were held back for the as-reflow shear strength evaluation, while the other samples were subjected to isothermal aging at a
Fig. 1. FE-SEM micrographs of (a) the MWCNTs used in this study and (b) the embedded MWCNTs at the interstices of the compositional elements after ball milling process.
370
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
as the formation of such nucleating particles has a high propensity of impeding the free flow of the solder melt. 3.2. Microstructural evolution
Fig. 3. Geometry of the single-lap solder joint.
temperature of 170 C for 500, 1000 and 1500 h. The shear test was carried out on the universal testing machine (Instron 3366, 10 KN capacity) with an extension rate of 0.1 mm/min at room temperature. Lastly, five samples were tested for both the plain and composite single-lap solder joints with varying weight proportions of the MWCNTs. 3. Results and discussion 3.1. Wettability measurement The wettability of a solder system on a substrate is a key factor in an interconnection joint formation. A substrate is said to be decently wetted by a liquid when the value of the contact angle ðqÞ gets smaller [10,27]. Fig. 4 presents the graphical representation and the numerical enumeration of the wetting angles respectively. From the representation, it is evident that the contact angle decreased with increasing CNTs addition, whereas with 0.1wt.% CNTs solder reinforcement, the contact angle had a leap up to 29.2 C. It is noteworthy that the Sn-5Sb-0.05CNT solder candidate exhibited the best wettability, with the contact angle decreasing by 14% when compared to the plain solder. It should be taken into consideration that the fluidity of solder alloys can be influenced by the emanating IMC phases from the solder-substrate chemical reaction. Salleh et al. [28] reported the formation of a planar IMC layer at the instance of contact between the solder melt and the Cu substrate. Thus it is inferred from their report that the formation of this IMC layer could be pivotal in controlling the wetting process at the solder-substrate interface. Meanwhile, the existence of CNTs at the interstices of the solder alloy is capable of curtailing the diffusion of the Sn atoms needed for intermetallic compound formation. Hence, it is proposed in this study that the suppression in the volumetric proportion of the precipitating float IMC (Cu6Sn5) and the interfacial IMCs rendered by the CNT reinforced solders led to a decrease in the wetting angle,
The bulk solder microstructures for the as-reflow and aging conditions are presented in Fig. 5 (a) and (b) respectively. Taking both conditions into consideration, a dendritic, dark grey precipitate is dispersed within the bulk solder matrix of both the plain and composite solder for the as-reflow condition, while for the aged samples, a coarsening effect was spotted in all samples which increased with the aging time. A detailed investigation of the consistent features (light grey and dark grey areas) as observed in both micrographs were further characterised with the EDS analysis. From the EDS spectrums shown in Fig. 5(cee), the solder microstructure is observed to be composed of three phases which are the Sn rich, Sb-Sn and the Cu-Sn phases. The XRD results reported in the study of El-Daly et al. gave an insight to the phases formed [4]. The authors identified the phases to be the b-Sn, SbSn and Cu6Sn5 phases, which is consistent with the findings in the present study. Meanwhile, further investigation was performed for the interfacial intermetallic compound layer. FE-SEM micrographs of interfacial IMC layers for both plain and 0.05CNT reinforced composite solder joints subjected to as-reflow and isothermal aging conditions are presented in Fig. 6. From the representation, scallop-shape interfacial IMC layer profile was observed in both plain and composite solder joints for the as-reflow condition. The IMC layer was spotted as the Cu6Sn5 IMC phase by EDS analysis. Sb element was not detected at the interfacial IMC layer as the element only dissolved into the solder matrix. A plot of the average total IMC layer thickness against the aging hour is presented in Fig. 7. From the plot, an appreciable degree of IMC layer growth retardation can be observed in the composite solder joints, most especially in the Sn-5Sb-0.05CNT sample as compared with the plain solder counterpart. Hence, it is proposed in this study that the plausible IMC layer growth observed in the composite solder joints could be linked with the diffusion inhibiting role of the CNTs in the solder matrix, which restricted the number of Sn atoms needed for further IMC growth as the aging period increases. A similar finding was reported in other investigations [17,29]. As discussed earlier, CNTs particulates were embedded within the interstices of the host alloy (Sn and Sb) after the ball milling process and thus altering the microstructural morphology of the composite solders, hence during the reflow soldering process, the established physical bond between CNTs and host alloy hindered the spontaneous reaction between Sn and Cu atoms at the instance of contact between the solder melt and Cu substrate by restraining the diffusion path of Sn atoms in the solder matrix. Meanwhile, after subjecting the solder joints to isothermal aging process, a new Cu3Sn IMC layer in contrast to the already existing Cu6Sn5 IMC layer emerged. Mayapan et al. proposed that Cu6Sn5 IMC layer is thermodynamically unstable due to the lower activation energy of this layer relative to the Cu3Sn IMC layer [30]. Thus, the Cu3Sn IMC layer evolved over time with respect to the aging temperature by depleting the Cu6Sn5 IMC. To further substantiate the formation and growth of the interfacial intermetallic compound layers, the highlighted equations proffer the possible reactions for the evolution of Cu6Sn5 and Cu3Sn IMCs respectively [17,31].
6Cu þ 5Sn ¼ Cu6 Sn5 Fig. 4. Graphical plot of the contact angle relative to the respective solder samples.
(1)
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
371
Fig. 5. (a) FE-SEM micrograph of as-reflow solder microstructure (b) FE-SEM micrograph of 500 h aged solder microstructure (cee) EDS spectrums of elemental phases present in the solder microstructure.
3Cu þ Sn ¼ Cu3 Sn
3.3. Shear strength test
(2)
9Cu þ Cu6 Sn5 ¼ 5Cu3 Sn
(3)
Cu6 Sn5 3Sn ¼ 2Cu3 Sn
(4)
Equation (1) illustrates the reaction mechanism controlling the formation of the Cu6Sn5 IMC layer at the instance of the reflow soldering process while equations (2) and (3) are largely responsible for the IMC layer growth at the initiation of the aging process as the source of Cu atoms is directly from the underlying substrate. However, over the time, equations (2) and (3) remained dormant as the IMC scallops got planar and the vacancies at the valley of the scallops got filled up by the growing Cu3Sn IMC layer thereby impeding the penetration of Cu atoms from the Cu substrate to the solder bulk. With increasing aging period, equation (4) became fully activated and with CNTs exhibiting a diffusion inhibiting role within the composite solder bulk, supply of Sn atoms from solder bulk to the interfacial IMC layer was effectively controlled. Hence, it is proposed in this study that the plausible IMC layer growth retardation observed in the composite solder joints can help remedy the reliability issues related to the conventional Sn-5Sb lead-free solder interconnect.
During their service lives, solder joints are often subjected to mechanical loadings emanating from thermal fluctuations and shock loads. Thus, ample shear strength in the solder joints is essential for electronic packaging. The shear strength investigations were conducted by replicating the actual solder joints in electronic gadgets with a single-lap solder joint system. Fig. 8 presents the shear strength measurement of the plain and composite solder joints subjected to the as-reflow and aging conditions. In view of assessing the shear capacity of the joints, shear strength investigations were conducted on the plain and composite solder joints aged at 170 C for 500 h, 1000 h and 1500 h. From Fig. 8(b), a downward slope can be observed in all samples as the aging hours increased. This trend can be ascribed to the increasing growth of the intermetallic compound layer as all the samples fractured at the IMC layer sites. A similar finding was reported in previous studies, where the IMC layer growth and solder microstructure coarsening played key roles in degrading the mechanical property of the solder joints [32,33]. Despite the solid-state aging process, the shear strength values of the composite solder joints were higher compared with the plain solder joints, except for the Sn-5Sb-0.1CNT sample.
372
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
Fig. 6. Representative FE-SEM micrograph showing the interfacial IMC layer of (a) Sn-5Sb (b) Sn-5Sb-0.05CNT solder joints. Subscripts 0 and 1 represents the as reflow and 1500 h (Aging 170 C) conditions respectively.
Contrary to the most significant interfacial IMC layer retardation recorded in the Sn-5Sb-0.05CNT composite joints, the Sn-5Sb0.01CNT single-lap joint system has higher shear strength than every other solder joints across board. As presented in Fig. 8, the shear strength of the 0.01wt.% CNT reinforcement increased up to 29% over the plain solder joint counterpart. Hence, it is proposed that the superior shear strength property of the composite solder joints can be attributed to the exceptional mechanical property of CNTs, proper dispersion of CNTs in the solder matrix, and IMC layer growth cushioning offered by CNTs' existence in the solder matrix. Also, it is noteworthy that during the shear strength investigation, stress concentration was largely localized between the solder/ IMC interface region and due to the weakness of the brittle IMC layer, it became the preferred site for failure propagation. Meanwhile, it appears that CNTs presence within the solder matrix of the composite solder joint played a critical role in strengthening the joint during axial loading by resisting the parallel dislocation of the protruding scallop shaped IMC layer against the solder matrix.
Fig. 7. Average total IMC layer thickness against the isothermal aging hour.
Fig. 8. Relationship of Shear strength results with (a) Solder joint samples and (b) Aging time.
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
373
Fig. 9. (a) FE-SEM fracture surface of the as-reflow sample; (b) FE-SEM fracture surface of the 1500 h aged sample; (c) Representative OM cross-sectional view of the fractured single-lap joint; (d) Schematic illustration of the fracture mode; (e) FE-SEM micrograph of broken Cu6Sn5 IMC (f) FE-SEM micrograph of exposed Cu3Sn IMC.
Therefore, the decline observed in the shear strength of higher composite solder formulations (0.05wt.% and 0.1wt.%) can reasonably be attributed to poor dispersion of CNTs within the solder matrix, thus resulting to agglomeration and dysfunctional behavior of CNTs to effectively bond the host solder particles. 3.4. Fracture analysis of samples In view of arriving at an intuitive conclusion concerning the particles and mechanisms related to sample failure, fractured single-lap joints were randomly selected and investigated using the OM and FE-SEM techniques. The representative FE-SEM micrographs of the as-reflow fractured sample and the isothermally-aged fractured sample at 170 C for 1500 h are presented in Fig. 9 (a) and (b). Dimple-like features are evidently observed on the fractured surface of the samples, regardless of the elemental compositions and the aging conditions. Furthermore, the dimples are aligned such that they reflect the axial loading direction along the sample surface. With the features observed, it can be adjudged that all samples experienced a ductile fracture mode. A similar observation was reported in other studies [22,29]. Consequently, it can be observed that the dimple area increased with increasing aging time. The dimple size variation in the as-reflow and the 1500 h aged samples could be ascribed to the sharp scallop-shaped IMC layer and the coarse, planar IMC layer formed with increasing aging time respectively. From the test carried out, fracture occurred in-between the solder bulk and the underlying IMC layer. The debonding of the IMC layer from the bulk solder suggests that the brittle IMC layer is a
favourable site for the initiation of crack propagation resulting from the axial loading process, see Fig. 9 (c) and (d). To further ascertain the role of the IMC layer in the failure mechanism, Fig. 9 (e) and (f) illustrates some of the notable features that contribute to sample failure. Consequently, broken Cu6Sn5 IMCs were spotted on the fracture surfaces, as well as images of exposed Cu3Sn IMC located on the detached Cu piece with only the IMC layer. In view of analysing the strengthening mechanism in the composite solder joints, it is noteworthy that during the reflow process, Sn ions become highly reactive and are adsorbed onto the CNTs through the oxide layer covering the CNTs' surface. Thus, the morphology of the bond is such that it depicts strands and clusters of CNTs encapsulated by Sn ions. This finding synchronises with the investigation of Li et al. concerning Pb2þ adsorption onto functionalised and non-functionalised MWCNTs [34]. Fig. 10 (a) and (b) demonstrates the morphology of Sn-CNTs in the Sn-5Sb-0.01CNT and Sn-5Sb-0.05CNT solder joint sample respectively, which therefore suggests the pull-out sites in the composite solder joints. Carbon nanotubes tend to agglomerate due to the relatively weak Van der Waals forces binding the CNTs together. Thus, increasing the weight content of the CNTs in the solder matrix could result in a poor distribution and wider interfacial volume of the CNTs within the solder matrix, thereby generating a weak bond between the CNTs and the solder matrix. This phenomenon therefore justifies the CNTs' shape variation in the composite solder formulation shown in the figure, and also the drawback in the shear strength values of the 0.05CNT and 0.1CNT solder reinforcements as the clustered CNTs failed to bond the solder matrix effectively.
374
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
Fig. 10. FE-SEM fractographs of the morphology of Sn-CNT in the solder matrix of (a) Sn-5Sb-0.01CNT and (b) Sn-5Sb-0.05CNT.
4. Conclusions
References
In comparison with the plain solder, the effect of the multiwalled carbon nanotubes in the composite solder was substantiated through a detailed evaluation of the wettability property, the microstructural evolution and the shear strength property. The results from the investigation are hereby outlined as follows:
[1] M. Abtew, G. Selvaduray, Mater. Sci. Eng. A 27 (2000) 95e141. [2] Http//www.jeita.or.jp. assessed 14 January 2014. [3] I.P.C. Roadmap, A Guide for Assembly of Lead-free Electronics, IPC, Northbrook, IL, 2000. [4] A.A. El-Daly, A. Fawzy, A.Z. Mohamad, A.M. El-Taher, J. Alloys Comp. 509 (2011) 4574e4582. [5] A.A. El-Daly, Y. Swilem, A.E. Hammad, J. Alloys Comp. 471 (2009) 98e104. [6] C. Leinenbach, F. Valenza, D. Giuranno, H.R. Elsener, S. Jin, R. Novakovic, J. Electron. Mater 40 (2011) 1533e1541. [7] R.M. Shalaby, J. Alloys Comp. 480 (2009) 334e339. [8] F. Guo, J. Mater, Sci.:Mater. Electron 18 (2007) 129e145. [9] J. Shen, Y.C. Chan, Microelectron. Reliab 49 (2009) 223e234. [10] S.M.L. Nai, J. Wei, M. Gupta, Mater. Sci. Eng. A 423 (2006) 166e169. [11] V.L. Niranjani, B.S. Rao, V. Singh, S.V. Kamat, Mater. Sci. Eng. A 529 (2011) 257e264. [12] K.M. Kumar, V. Kripesh, A.A.O. Tay, J. Alloys Comp. 455 (2008) 148e158. [13] Y.D. Han, S.M.L. Nai, H.Y. Jing, L.Y. Xu, C.M. Tan, J. Wei, J. Mater. Sci.:Mater. Electron 22 (2010) 315e322. [14] J. Zhao, C. Cheng, L. Qi, C. Chi, J. Alloys Comp. 473 (2009) 382e388. [15] J. Yoon, S. Kim, S. Jung, Mater. Trans. 45 (2004) 727e733. [16] Y. Ko, S. Kwon, Y. Lee, J. Kim, C. Lee, S. Yoo, J. Alloys Comp. 583 (2014) 155e161. [17] S. Xu, Y. Cheong, K. Zhang, K.C. Yung, J. Alloys Comp. 595 (2014) 92e102. [18] C. Andersson, Z. Lai, J. Liu, H. Jiang, Y. Yu, Mater. Sci. Eng. A 394 (2005) 20e27. [19] Y. Shen, Acta Mater 53 (2005) 2633e2642. [20] X. Deng, R.S. Sidhu, P. Johnson, N. Chawla, Mettalurg. Mater. Trans. A 36 (2005) 55e64. [21] H. Lee, S. Hu, T. Hong, Y. Chen, J. Electron, Mater 37 (2008) 9e12. [22] Y. Lee, H. Lee, Mater. Sci. Eng. A 444 (2007) 75e83. [23] M.A.A. Salleh, A.M.M. Al Bakri, M.H.Z. Hazizi, F. Somidin, N. Farhani, M. Alui, Z. Arifin, Mater. Sci. Eng. A 556 (2012) 633e637. [24] H. Fallahi, M.S. Nurulakmal, A. Fallahi, J. Abdullah, J. Mater, Sci. Mater. Electron 23 (2012) 1739e1749. [25] A. Kantarc, Y.E. Kalay, Mater. Sci. Eng. A 593 (2014) 79e84. [26] Y.D. Han, H.Y. Jing, S.M.L. Nai, L.Y. Xu, C.M. Tan, J. Wei, Intermetallics 31 (2012) 72e78. [27] M.F. Arenas, V.L. Acoff, J. Electron, Mater 33 (2004) 1452e1458. [28] M.A.A. Salleh, S.D. McDonald, H. Yasuda, A. Sugiyama, K. Nogita, Scr. Mater 100 (2015) 17e20. [29] S.M.L. Nai, J. Wei, M. Gupta, J. Alloys Comp. 473 (2009) 100e106. [30] R. Mayappan, I. Yahya, N.A.A. Ghani, H.A. Hamid, J. Mater. Sci. Mater. Electron 25 (2014) 2913e2922. [31] X. Hu, Y. Li, Z. Min, J. Alloys Comp. 582 (2014) 341e347. [32] K.S. Kim, S.H. Huh, K. Suganuma, J. Alloys Comp. 352 (2003) 226e236. [33] H.T. Lee, M.H. Chen, H.M. Jao, T.L. Liao, Mater. Sci. Eng. A 358 (2003) 134e141. [34] Y.H. Li, S. Wang, J. Wei, X. Zhang, C. Xu, Z. Luan, B. Wei, Chem. Phys. Letts 357 (2002) 263e266.
(1) It was evident from the wettability results computed that the composite solder joints showed better results when compared with the plain solder. This can be ascribed to the volumetric suppression of the floating Cu6Sn5 IMC and the underlying IMC layer which has a high propensity to inhibit the free flow of the solder melt. (2) A coarsening effect was observed in the microstructure of all the solder systems which increased with aging time. Also, for the interfacial IMC layer evolution, an appreciable retardation in IMC layer growth was observed in the composite solder joints as against the plain solder joint. (3) The shear strength values of both the plain and composite single-lap solder joints were observed to decrease with increasing aging time. However, the composite solder joints showed better shear strength, with the 0.01CNT solder reinforcement exhibiting the best shear strength value across board. (4) The significant improvement in the shear strength of the composite solder joints was attributed to the exceptional mechanical properties of CNTs and the proper dispersion of CNTs in the solder matrix. The drawback recorded in the other composite solder formulation can be linked to the clustering tendency of CNTs with increasing quantity, which resulted in poor distribution and weak bonding between the CNTs and the solder matrix. In addition, all the samples experienced a ductile failure mode as dimple-like features were evident on the fractured surface; these grew wider with increasing aging time.
Contents lists available at ScienceDirect
Journal of Alloys and Compounds journal homepage: http://www.elsevier.com/locate/jalcom
Investigating the effect of isothermal aging on the morphology and shear strength of Sn-5Sb solder reinforced with carbon nanotubes T.T. Dele-Afolabi a, *, M.A. Azmah Hanim a, d, M. Norkhairunnisa b, d, H.M. Yusoff c, M.T. Suraya a a
Department of Mechanical and Manufacturing Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia Department of Aerospace Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia d Laboratory of Biocomposite Technology, Institute of Tropical Forestry and Forest Products (INTROP), Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 21 April 2015 Received in revised form 8 June 2015 Accepted 4 July 2015 Available online 14 July 2015
An analysis of the role played by the addition of carbon nanotubes (CNTs) to the solder matrix of conventional Sn-5Sb lead-free solder was performed. In a bid to determine the potential of this new solder system, the powder metallurgy approach was used to synthesise a plain Sn-5Sb solder system and CNTs reinforced composite solder formulations of Sn-5Sb-xCNT; x ¼ 0.01wt.%, 0.05wt.% and 0.1wt.%. Isothermal aging study was conducted on the solder joints, to examine the evolution of the interfacial intermetallic compound (IMC) layer between solder and the adjoining copper (Cu) substrate. Similarly, shear strength analysis was performed on as-reflow and aged solder joints. A considerable improvement in the wetting properties, the microstructural evolution, and the interfacial intermetallic compound (IMC) layer growth was observed in the composite solder joints. Owing to the excellent mechanical properties of CNTs, the shear strength assessment revealed that the composite solder joints gave a superior shear strength property, especially the Sn-5Sb-0.01CNT solder joint sample. © 2015 Elsevier B.V. All rights reserved.
Keywords: Lead-free solders Carbon nanotubes Composite solders Shear strength
1. Introduction The enactment of legislations barring the utilisation of lead in electronic products [1e3] has instigated the sensitivity of researchers towards conformity with the alternative lead-free standards for electronic systems packaging. Although rigorous efforts have been channeled towards the development of these alternative solder candidates, the Sn-5Sb solder system has received less priority despite its great potential in step soldering and high temperature applications [4e7]. However, with the recent trend of miniaturisation and the IC terminal upsurge in electronic hardwares, critical evaluation of the lifetime operational capability in interconnection joints should be of utmost concern in order to inspect and curb the reliability issues associated with the evolving lead-free solder candidates. Several attempts to enhance the performance of these solder materials has
* Corresponding author. E-mail addresses: [email protected] (T.T. Dele-Afolabi), azmah@ upm.edu.my (M.A. Azmah Hanim). http://dx.doi.org/10.1016/j.jallcom.2015.07.036 0925-8388/© 2015 Elsevier B.V. All rights reserved.
moved researchers towards the development of a composite approach in promoting the mechanical properties of interconnection joints, particularly the limitation in grain boundary sliding, creep rate mitigation and thermo-mechanical fatigue resistance [8,9]. Thus, introducing suitable reinforcement particles to the Sn5Sb solder matrix is pivotal in fabricating a revamped solder system which can thrive well under extreme service conditions and can function decently in densely packed electronic devices. Thus far, carbon nanotubes (CNTs) an outstanding nanomaterial have aroused great interest for the production of better solder joints when doped with the conventional lead-free solders due to the exceptional mechanical properties demonstrated by this group of materials. Reports from previous studies have revealed a significant enhancement in the wettability results and the microstructural evolution of the CNT reinforced solders [10e13]. Studies have shown that the IMC layer at the solder/Cu interface increased with aging time and the growth was faster for higher aging temperatures [14,15]. Ko et al. found that the IMC thickness formation of the CNT-SnAg solder was slower than the SnAg solder and the IMC growth difference between both solder samples
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
further increased as the aging time increased [16]. Furthermore, shear strength of as-reflow and aged composite Sn-Ag-Cu solder joints was improved with CNTs reinforcement [17]. Accordingly, various researchers [18e25] have adopted the single-lap joint system in conducting experimental evaluations of the shear strength property of solder materials. A bid by Han et al. [26] to develop a composite solder via the incorporation of Nicoated carbon nanotubes proved rewarding. Although, both the monolithic and the composite solder joints experienced a decline in the shear strength values as the thermal cycle increased. Nonetheless, the composite solders emerged better than the monolithic solder counterpart, irrespective of the thermal cycling period. Nearly all studies on the development of composite solders through the incorporation of carbon nanotubes have centered mainly on the SAC solder candidate. Nevertheless, the influence of CNTs in the Sn-5Sb solder system is yet to be investigated. Pitching this study on the rattling records of CNTs doped solders, it is imperative to investigate the effect of isothermal aging on the shear strength property of composite and plain Sn-5Sb solder joints. In the present study, triad formulations of the Sn-5Sb composite solder systems were synthesised. The wettability, microstructural evolution and shear strength property of the synthesised solder joints were evaluated under different aging times. 2. Experimental procedures 2.1. Material processing and sample preparation The powder metallurgy approach was used to synthesise four different solder formulations: Sn-5Sb, Sn-5Sb-0.01CNT, Sn-5Sb0.05CNT and Sn-5Sb-0.1CNT. Pure Sn and Sb micron-sized particles of 30e45 mm and 3e7 mm respectively were used as the host alloy, while multi-walled carbon nanotubes (MWCNTs) with an outer diameter of 15e20 nm was purchased from Cheap Tubes Inc, USA. For the raw materials, carbon and oxide particulates were identified from the EDS analysis as the main impurities. Using a Planetary Mono Mill, the solder formulations were dry milled for 6 h in a container with a ball to powder ratio of 20:1 at 800 rpm. After mixing, 2 g of the powder blends were compacted at 80 MPa to produce a solder pellet with 20 mm diameter and 2 mm thickness. Fig. 1(a) and Fig. 1(b) respectively present the FE-SEM micrographs of the MWCNTs used in this study and the embedded MWCNTs at the interstices between the compositional elements. 2.2. Microstructural characterization For the microstructural characterisation studies, solder joints were prepared by the reflow soldering process on a copper (Cu)
369
Fig. 2. Optical micrograph of the contact angle between the solder and the Cu substrate.
substrate. Following the reflow process, the solder joint samples were subjected to isothermal aging at 170 C for 500 h, 1000 h and 1500 h. After the aging process, the samples were cold mounted using the epoxy resin system, cross-sectioned and subjected to standard polishing procedures using a 0.05 mm Master Prep alumina suspension. Prior to the electron microscopy analysis, images of the wetting angle and microstructural morphology were taken using the optical micrograph (OM). Also, the contact angle and the interfacial IMC layer of the respective solder joints were measured using the streams essential software integrated with the OM (see Fig. 2). Finally, the solder samples were observed using the HITACHI SU8000 field emission scanning electron microscope (FESEM) equipped with energy dispersive spectroscope (EDS) to evaluate the microstructural morphology and elemental compositions respectively. 2.3. Shear test To conduct the shear strength evaluation, single-lap joint samples were utilised. The samples were fabricated in such a way as to mimic actual solder joints by filling the overlap area between the Cu substrates with solder pellets and flux application prior to the reflow process. To ensure the proper alignment of all the components that make up the lap joint setup, the samples were placed in an alumina boat jig and thus subjected to the reflow soldering process. The geometry of the single-lap joint sample is shown in Fig. 3. As pictured, each of the copper substrates has a dimension of 52 11 1 mm3 and a lap joint area of 10 11 mm2. After the as-reflow soldering operation, some of the samples were held back for the as-reflow shear strength evaluation, while the other samples were subjected to isothermal aging at a
Fig. 1. FE-SEM micrographs of (a) the MWCNTs used in this study and (b) the embedded MWCNTs at the interstices of the compositional elements after ball milling process.
370
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
as the formation of such nucleating particles has a high propensity of impeding the free flow of the solder melt. 3.2. Microstructural evolution
Fig. 3. Geometry of the single-lap solder joint.
temperature of 170 C for 500, 1000 and 1500 h. The shear test was carried out on the universal testing machine (Instron 3366, 10 KN capacity) with an extension rate of 0.1 mm/min at room temperature. Lastly, five samples were tested for both the plain and composite single-lap solder joints with varying weight proportions of the MWCNTs. 3. Results and discussion 3.1. Wettability measurement The wettability of a solder system on a substrate is a key factor in an interconnection joint formation. A substrate is said to be decently wetted by a liquid when the value of the contact angle ðqÞ gets smaller [10,27]. Fig. 4 presents the graphical representation and the numerical enumeration of the wetting angles respectively. From the representation, it is evident that the contact angle decreased with increasing CNTs addition, whereas with 0.1wt.% CNTs solder reinforcement, the contact angle had a leap up to 29.2 C. It is noteworthy that the Sn-5Sb-0.05CNT solder candidate exhibited the best wettability, with the contact angle decreasing by 14% when compared to the plain solder. It should be taken into consideration that the fluidity of solder alloys can be influenced by the emanating IMC phases from the solder-substrate chemical reaction. Salleh et al. [28] reported the formation of a planar IMC layer at the instance of contact between the solder melt and the Cu substrate. Thus it is inferred from their report that the formation of this IMC layer could be pivotal in controlling the wetting process at the solder-substrate interface. Meanwhile, the existence of CNTs at the interstices of the solder alloy is capable of curtailing the diffusion of the Sn atoms needed for intermetallic compound formation. Hence, it is proposed in this study that the suppression in the volumetric proportion of the precipitating float IMC (Cu6Sn5) and the interfacial IMCs rendered by the CNT reinforced solders led to a decrease in the wetting angle,
The bulk solder microstructures for the as-reflow and aging conditions are presented in Fig. 5 (a) and (b) respectively. Taking both conditions into consideration, a dendritic, dark grey precipitate is dispersed within the bulk solder matrix of both the plain and composite solder for the as-reflow condition, while for the aged samples, a coarsening effect was spotted in all samples which increased with the aging time. A detailed investigation of the consistent features (light grey and dark grey areas) as observed in both micrographs were further characterised with the EDS analysis. From the EDS spectrums shown in Fig. 5(cee), the solder microstructure is observed to be composed of three phases which are the Sn rich, Sb-Sn and the Cu-Sn phases. The XRD results reported in the study of El-Daly et al. gave an insight to the phases formed [4]. The authors identified the phases to be the b-Sn, SbSn and Cu6Sn5 phases, which is consistent with the findings in the present study. Meanwhile, further investigation was performed for the interfacial intermetallic compound layer. FE-SEM micrographs of interfacial IMC layers for both plain and 0.05CNT reinforced composite solder joints subjected to as-reflow and isothermal aging conditions are presented in Fig. 6. From the representation, scallop-shape interfacial IMC layer profile was observed in both plain and composite solder joints for the as-reflow condition. The IMC layer was spotted as the Cu6Sn5 IMC phase by EDS analysis. Sb element was not detected at the interfacial IMC layer as the element only dissolved into the solder matrix. A plot of the average total IMC layer thickness against the aging hour is presented in Fig. 7. From the plot, an appreciable degree of IMC layer growth retardation can be observed in the composite solder joints, most especially in the Sn-5Sb-0.05CNT sample as compared with the plain solder counterpart. Hence, it is proposed in this study that the plausible IMC layer growth observed in the composite solder joints could be linked with the diffusion inhibiting role of the CNTs in the solder matrix, which restricted the number of Sn atoms needed for further IMC growth as the aging period increases. A similar finding was reported in other investigations [17,29]. As discussed earlier, CNTs particulates were embedded within the interstices of the host alloy (Sn and Sb) after the ball milling process and thus altering the microstructural morphology of the composite solders, hence during the reflow soldering process, the established physical bond between CNTs and host alloy hindered the spontaneous reaction between Sn and Cu atoms at the instance of contact between the solder melt and Cu substrate by restraining the diffusion path of Sn atoms in the solder matrix. Meanwhile, after subjecting the solder joints to isothermal aging process, a new Cu3Sn IMC layer in contrast to the already existing Cu6Sn5 IMC layer emerged. Mayapan et al. proposed that Cu6Sn5 IMC layer is thermodynamically unstable due to the lower activation energy of this layer relative to the Cu3Sn IMC layer [30]. Thus, the Cu3Sn IMC layer evolved over time with respect to the aging temperature by depleting the Cu6Sn5 IMC. To further substantiate the formation and growth of the interfacial intermetallic compound layers, the highlighted equations proffer the possible reactions for the evolution of Cu6Sn5 and Cu3Sn IMCs respectively [17,31].
6Cu þ 5Sn ¼ Cu6 Sn5 Fig. 4. Graphical plot of the contact angle relative to the respective solder samples.
(1)
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
371
Fig. 5. (a) FE-SEM micrograph of as-reflow solder microstructure (b) FE-SEM micrograph of 500 h aged solder microstructure (cee) EDS spectrums of elemental phases present in the solder microstructure.
3Cu þ Sn ¼ Cu3 Sn
3.3. Shear strength test
(2)
9Cu þ Cu6 Sn5 ¼ 5Cu3 Sn
(3)
Cu6 Sn5 3Sn ¼ 2Cu3 Sn
(4)
Equation (1) illustrates the reaction mechanism controlling the formation of the Cu6Sn5 IMC layer at the instance of the reflow soldering process while equations (2) and (3) are largely responsible for the IMC layer growth at the initiation of the aging process as the source of Cu atoms is directly from the underlying substrate. However, over the time, equations (2) and (3) remained dormant as the IMC scallops got planar and the vacancies at the valley of the scallops got filled up by the growing Cu3Sn IMC layer thereby impeding the penetration of Cu atoms from the Cu substrate to the solder bulk. With increasing aging period, equation (4) became fully activated and with CNTs exhibiting a diffusion inhibiting role within the composite solder bulk, supply of Sn atoms from solder bulk to the interfacial IMC layer was effectively controlled. Hence, it is proposed in this study that the plausible IMC layer growth retardation observed in the composite solder joints can help remedy the reliability issues related to the conventional Sn-5Sb lead-free solder interconnect.
During their service lives, solder joints are often subjected to mechanical loadings emanating from thermal fluctuations and shock loads. Thus, ample shear strength in the solder joints is essential for electronic packaging. The shear strength investigations were conducted by replicating the actual solder joints in electronic gadgets with a single-lap solder joint system. Fig. 8 presents the shear strength measurement of the plain and composite solder joints subjected to the as-reflow and aging conditions. In view of assessing the shear capacity of the joints, shear strength investigations were conducted on the plain and composite solder joints aged at 170 C for 500 h, 1000 h and 1500 h. From Fig. 8(b), a downward slope can be observed in all samples as the aging hours increased. This trend can be ascribed to the increasing growth of the intermetallic compound layer as all the samples fractured at the IMC layer sites. A similar finding was reported in previous studies, where the IMC layer growth and solder microstructure coarsening played key roles in degrading the mechanical property of the solder joints [32,33]. Despite the solid-state aging process, the shear strength values of the composite solder joints were higher compared with the plain solder joints, except for the Sn-5Sb-0.1CNT sample.
372
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
Fig. 6. Representative FE-SEM micrograph showing the interfacial IMC layer of (a) Sn-5Sb (b) Sn-5Sb-0.05CNT solder joints. Subscripts 0 and 1 represents the as reflow and 1500 h (Aging 170 C) conditions respectively.
Contrary to the most significant interfacial IMC layer retardation recorded in the Sn-5Sb-0.05CNT composite joints, the Sn-5Sb0.01CNT single-lap joint system has higher shear strength than every other solder joints across board. As presented in Fig. 8, the shear strength of the 0.01wt.% CNT reinforcement increased up to 29% over the plain solder joint counterpart. Hence, it is proposed that the superior shear strength property of the composite solder joints can be attributed to the exceptional mechanical property of CNTs, proper dispersion of CNTs in the solder matrix, and IMC layer growth cushioning offered by CNTs' existence in the solder matrix. Also, it is noteworthy that during the shear strength investigation, stress concentration was largely localized between the solder/ IMC interface region and due to the weakness of the brittle IMC layer, it became the preferred site for failure propagation. Meanwhile, it appears that CNTs presence within the solder matrix of the composite solder joint played a critical role in strengthening the joint during axial loading by resisting the parallel dislocation of the protruding scallop shaped IMC layer against the solder matrix.
Fig. 7. Average total IMC layer thickness against the isothermal aging hour.
Fig. 8. Relationship of Shear strength results with (a) Solder joint samples and (b) Aging time.
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
373
Fig. 9. (a) FE-SEM fracture surface of the as-reflow sample; (b) FE-SEM fracture surface of the 1500 h aged sample; (c) Representative OM cross-sectional view of the fractured single-lap joint; (d) Schematic illustration of the fracture mode; (e) FE-SEM micrograph of broken Cu6Sn5 IMC (f) FE-SEM micrograph of exposed Cu3Sn IMC.
Therefore, the decline observed in the shear strength of higher composite solder formulations (0.05wt.% and 0.1wt.%) can reasonably be attributed to poor dispersion of CNTs within the solder matrix, thus resulting to agglomeration and dysfunctional behavior of CNTs to effectively bond the host solder particles. 3.4. Fracture analysis of samples In view of arriving at an intuitive conclusion concerning the particles and mechanisms related to sample failure, fractured single-lap joints were randomly selected and investigated using the OM and FE-SEM techniques. The representative FE-SEM micrographs of the as-reflow fractured sample and the isothermally-aged fractured sample at 170 C for 1500 h are presented in Fig. 9 (a) and (b). Dimple-like features are evidently observed on the fractured surface of the samples, regardless of the elemental compositions and the aging conditions. Furthermore, the dimples are aligned such that they reflect the axial loading direction along the sample surface. With the features observed, it can be adjudged that all samples experienced a ductile fracture mode. A similar observation was reported in other studies [22,29]. Consequently, it can be observed that the dimple area increased with increasing aging time. The dimple size variation in the as-reflow and the 1500 h aged samples could be ascribed to the sharp scallop-shaped IMC layer and the coarse, planar IMC layer formed with increasing aging time respectively. From the test carried out, fracture occurred in-between the solder bulk and the underlying IMC layer. The debonding of the IMC layer from the bulk solder suggests that the brittle IMC layer is a
favourable site for the initiation of crack propagation resulting from the axial loading process, see Fig. 9 (c) and (d). To further ascertain the role of the IMC layer in the failure mechanism, Fig. 9 (e) and (f) illustrates some of the notable features that contribute to sample failure. Consequently, broken Cu6Sn5 IMCs were spotted on the fracture surfaces, as well as images of exposed Cu3Sn IMC located on the detached Cu piece with only the IMC layer. In view of analysing the strengthening mechanism in the composite solder joints, it is noteworthy that during the reflow process, Sn ions become highly reactive and are adsorbed onto the CNTs through the oxide layer covering the CNTs' surface. Thus, the morphology of the bond is such that it depicts strands and clusters of CNTs encapsulated by Sn ions. This finding synchronises with the investigation of Li et al. concerning Pb2þ adsorption onto functionalised and non-functionalised MWCNTs [34]. Fig. 10 (a) and (b) demonstrates the morphology of Sn-CNTs in the Sn-5Sb-0.01CNT and Sn-5Sb-0.05CNT solder joint sample respectively, which therefore suggests the pull-out sites in the composite solder joints. Carbon nanotubes tend to agglomerate due to the relatively weak Van der Waals forces binding the CNTs together. Thus, increasing the weight content of the CNTs in the solder matrix could result in a poor distribution and wider interfacial volume of the CNTs within the solder matrix, thereby generating a weak bond between the CNTs and the solder matrix. This phenomenon therefore justifies the CNTs' shape variation in the composite solder formulation shown in the figure, and also the drawback in the shear strength values of the 0.05CNT and 0.1CNT solder reinforcements as the clustered CNTs failed to bond the solder matrix effectively.
374
T.T. Dele-Afolabi et al. / Journal of Alloys and Compounds 649 (2015) 368e374
Fig. 10. FE-SEM fractographs of the morphology of Sn-CNT in the solder matrix of (a) Sn-5Sb-0.01CNT and (b) Sn-5Sb-0.05CNT.
4. Conclusions
References
In comparison with the plain solder, the effect of the multiwalled carbon nanotubes in the composite solder was substantiated through a detailed evaluation of the wettability property, the microstructural evolution and the shear strength property. The results from the investigation are hereby outlined as follows:
[1] M. Abtew, G. Selvaduray, Mater. Sci. Eng. A 27 (2000) 95e141. [2] Http//www.jeita.or.jp. assessed 14 January 2014. [3] I.P.C. Roadmap, A Guide for Assembly of Lead-free Electronics, IPC, Northbrook, IL, 2000. [4] A.A. El-Daly, A. Fawzy, A.Z. Mohamad, A.M. El-Taher, J. Alloys Comp. 509 (2011) 4574e4582. [5] A.A. El-Daly, Y. Swilem, A.E. Hammad, J. Alloys Comp. 471 (2009) 98e104. [6] C. Leinenbach, F. Valenza, D. Giuranno, H.R. Elsener, S. Jin, R. Novakovic, J. Electron. Mater 40 (2011) 1533e1541. [7] R.M. Shalaby, J. Alloys Comp. 480 (2009) 334e339. [8] F. Guo, J. Mater, Sci.:Mater. Electron 18 (2007) 129e145. [9] J. Shen, Y.C. Chan, Microelectron. Reliab 49 (2009) 223e234. [10] S.M.L. Nai, J. Wei, M. Gupta, Mater. Sci. Eng. A 423 (2006) 166e169. [11] V.L. Niranjani, B.S. Rao, V. Singh, S.V. Kamat, Mater. Sci. Eng. A 529 (2011) 257e264. [12] K.M. Kumar, V. Kripesh, A.A.O. Tay, J. Alloys Comp. 455 (2008) 148e158. [13] Y.D. Han, S.M.L. Nai, H.Y. Jing, L.Y. Xu, C.M. Tan, J. Wei, J. Mater. Sci.:Mater. Electron 22 (2010) 315e322. [14] J. Zhao, C. Cheng, L. Qi, C. Chi, J. Alloys Comp. 473 (2009) 382e388. [15] J. Yoon, S. Kim, S. Jung, Mater. Trans. 45 (2004) 727e733. [16] Y. Ko, S. Kwon, Y. Lee, J. Kim, C. Lee, S. Yoo, J. Alloys Comp. 583 (2014) 155e161. [17] S. Xu, Y. Cheong, K. Zhang, K.C. Yung, J. Alloys Comp. 595 (2014) 92e102. [18] C. Andersson, Z. Lai, J. Liu, H. Jiang, Y. Yu, Mater. Sci. Eng. A 394 (2005) 20e27. [19] Y. Shen, Acta Mater 53 (2005) 2633e2642. [20] X. Deng, R.S. Sidhu, P. Johnson, N. Chawla, Mettalurg. Mater. Trans. A 36 (2005) 55e64. [21] H. Lee, S. Hu, T. Hong, Y. Chen, J. Electron, Mater 37 (2008) 9e12. [22] Y. Lee, H. Lee, Mater. Sci. Eng. A 444 (2007) 75e83. [23] M.A.A. Salleh, A.M.M. Al Bakri, M.H.Z. Hazizi, F. Somidin, N. Farhani, M. Alui, Z. Arifin, Mater. Sci. Eng. A 556 (2012) 633e637. [24] H. Fallahi, M.S. Nurulakmal, A. Fallahi, J. Abdullah, J. Mater, Sci. Mater. Electron 23 (2012) 1739e1749. [25] A. Kantarc, Y.E. Kalay, Mater. Sci. Eng. A 593 (2014) 79e84. [26] Y.D. Han, H.Y. Jing, S.M.L. Nai, L.Y. Xu, C.M. Tan, J. Wei, Intermetallics 31 (2012) 72e78. [27] M.F. Arenas, V.L. Acoff, J. Electron, Mater 33 (2004) 1452e1458. [28] M.A.A. Salleh, S.D. McDonald, H. Yasuda, A. Sugiyama, K. Nogita, Scr. Mater 100 (2015) 17e20. [29] S.M.L. Nai, J. Wei, M. Gupta, J. Alloys Comp. 473 (2009) 100e106. [30] R. Mayappan, I. Yahya, N.A.A. Ghani, H.A. Hamid, J. Mater. Sci. Mater. Electron 25 (2014) 2913e2922. [31] X. Hu, Y. Li, Z. Min, J. Alloys Comp. 582 (2014) 341e347. [32] K.S. Kim, S.H. Huh, K. Suganuma, J. Alloys Comp. 352 (2003) 226e236. [33] H.T. Lee, M.H. Chen, H.M. Jao, T.L. Liao, Mater. Sci. Eng. A 358 (2003) 134e141. [34] Y.H. Li, S. Wang, J. Wei, X. Zhang, C. Xu, Z. Luan, B. Wei, Chem. Phys. Letts 357 (2002) 263e266.
(1) It was evident from the wettability results computed that the composite solder joints showed better results when compared with the plain solder. This can be ascribed to the volumetric suppression of the floating Cu6Sn5 IMC and the underlying IMC layer which has a high propensity to inhibit the free flow of the solder melt. (2) A coarsening effect was observed in the microstructure of all the solder systems which increased with aging time. Also, for the interfacial IMC layer evolution, an appreciable retardation in IMC layer growth was observed in the composite solder joints as against the plain solder joint. (3) The shear strength values of both the plain and composite single-lap solder joints were observed to decrease with increasing aging time. However, the composite solder joints showed better shear strength, with the 0.01CNT solder reinforcement exhibiting the best shear strength value across board. (4) The significant improvement in the shear strength of the composite solder joints was attributed to the exceptional mechanical properties of CNTs and the proper dispersion of CNTs in the solder matrix. The drawback recorded in the other composite solder formulation can be linked to the clustering tendency of CNTs with increasing quantity, which resulted in poor distribution and weak bonding between the CNTs and the solder matrix. In addition, all the samples experienced a ductile failure mode as dimple-like features were evident on the fractured surface; these grew wider with increasing aging time.