Creep Mechanism Fractography Analysis on SnPb Eutectic Solder Joint - - PDF document

Creep Mechanism Fractography Analysis on SnPb Eutectic Solder Joint Failure

Chulmin Oh*, Changwoon Han, Nochang Park, Byungsuk Song, and Wonsik Hong

Physics-of-Failure Research Center, Korea Electronic Technology Institute, Seongnam, 463-816, Republic of Korea *82-31-789-7288, 82-31-789-7059 and cmoh@keti.re.kr Abstract Microstructural fracture mode observed in creep can be divided into intergranular and transgranular fracture. Depending on temperature and stress condition, creep fracture mode is decided. To design an accelerated life test, it should be confirmed that the failure mode in the accelerated test is identical to the mode in real field

• condition. Selecting optimal conditions of temperature and stress in accelerated creep rupture test requires

extensive fractography analysis. In this study, SnPb eutectic solder joints for holding an anchor from heat sink system are subjected to creep rupture tests. After the test, failed solder joints are investigated and analyzed to identify creep fracture mode. Fracture microstructures of solder joints are analyzed using SEM and FIB. It is

• bserved that transgranular fractures are predominant in the condition of low temperature and high stress and

intergranular fractures are predominant in the condition of high temperature and low stress. Analysis results confirmed creep deformation mechanism map made by X.Q.Shi et al. and suggested optimal conditions of temperature and stress for accelerated creep rupture test with SnPb eutectic solder joints. Key words: creep fracture mode, fractography analysis, accelerated life test, SnPb solder joint Introduction SnPb eutectic solder has been applied in electronics for a long time because the solder has several following properties;(i) good wettability with an aid of mildly active fluxes, (ii) no brittle intermetallic compound formation in solder itself, (iii) low melting point to permit the design of components that can endure the high temperature associated with the soldering process, and (iv) few problem occurrence of tin oxide film problems compared with the oxide films of other solder alloys[1]. SnPb eutectic solder is still used in specified area such as military, aviation and health due to an exception of the environmental regulations although Pb-free solder has been applied in electronic industry several years ago corresponding with environmental regulations. SnPb eutectic solder shows a slow plastic strain(creep) under low permanent stresses at ambient temperature because the ambient operating temperature is above 0.5 times melting point of eutectic solder[1]. At high elevated temperature, the mobility of atom and dislocation increases and the concentration of vacancy increases with temperature. Creep deformation comes into play at evaluated temperature[2]. The deformation of SnPb eutectic solder at operating temperature has followed the creep deformation including the matrix diffusion, dislocation glide, dislocation climb and grain boundary sliding. The deformation of SnPb eutectic solder is increased by these thermal activitated processes and causes various failures of solder joint responsible for interconnecting the electronic components and making the robust structural reliability of the electronic package[3]. Therefore, it is necessary to understand the creep behavior of SnPb eutectic solder with stress and temperature in

• rder to predict the lifetime of SnPb solder joint in

electronic assembly. A creep deformation map for SnPb solder alloy was established by X.Q.shi et al. and divided into two regions of dislocation-controlled creep and diffusion-controlled creep[4]. Diffusion–controlled creep is favored at high temperatures and low stresses, while dislocation-controlled creep is more dominant at low temperature and high stresses. In intermediate temperature regime(in between 0.4 and 0.6 Tm) the creep deformation mechanism can be a mix of those in the low temperature and high temperature regime[5]. The creep rate of this regime is plotted as a power(n) function of stress and an Arrhenius-type expression with characteristic activation energy(Q). The values of n and Q are variable with respect to temperature and stress. A break in the isothermal curve is shown between low stress and high stress[5]. The break in the curve

• ccurs at stress at which the fracture mode changes

from intergranular to transgranular facture[5]. The facture mechanism map is useful for indentifying the fracture mode at any conditions of temperature and stress. Solder joints in electronics play role on not

• nly making an electrical path by connecting the

electronic components and board but also holding the anchor from heat sink pin of microprocessor

• package. The field temperature is above 0.5Tm and

the applied stress in solder joints is constant so that the creep damage of solder joint would be accumulated during the usage and brings about the failure of solder joints. It is needed to establish an acceleration model based on creep deformation mechanism for SnPb eutectic solder joint in order to predict the lifetime for SnPb eutectic solder joint. In our study, accelerated test was conducted

• n several conditions of temperature and stress

corresponding to creep mechanism to acquire proper acceleration creep model for SnPb eutectic solder

• joint. The failure mode of failed solder joint is

analyzed to indentify the failure mode in accelerated

• conditions. It should be confirmed that the failure

mode in the accelerated test is identical to the mode in real field condition. We discussed the results for fracture analysis and provided a guideline to the accelerated test for creep rupture test. Experimental procedures The test board for accelerated creep rupture test is shown Fig. 1. Specimens were soldered with SnPb eutectic solder on specific coupon board with which was attached steel bar for preventing the board warpage during creep rupture test and involved Cu layer patterns for acquiring the failure time. (a) (b) Figure 1 Test board for accelerated creep rupture test: (a) front side of board, and (b) back side of board. Considering the operation environment for solder joint of an anchor to hold heat sink pin, it was found that the operation temperature is about ambient temperature (above 0.5Tm) and the applied load to anchor is around 1 kg. Therefore, the experimental conditions for creep rupture accelerated test are planned in Table 1. The specimen size is 10 units per each condition. Table 1 Experimental conditions for accelerated creep rupture test Temp.( ℃) Load(kg) 35 55 75 4.0 10 ea 10 ea 10 ea 6.0g 10 ea 10 ea 10 ea 8.0 10 ea 10 ea 10 ea As shown Fig. 2, the weight assigned to each condition was hung on every specimen and inserted into chambers capable of keeping constant

• temperature. It was able to acquire the failure time

by measuring the resistance of solder joint

• continuously. When the creep fracture was happened

during accelerated creep rupture test, microstructure fractography of solder joint was analyzed with scanning electron microscope(SEM) and focused ion beam(FIB) in order to identify the failure mode of fracture specimen. (a) (b) Figure 2 Experimental configurations with (a) weights used for creep rupture test, (b) test fixture with weights and specimens. Results and discussions

• Fig. 3 shows the SEM micrographs of

microstructure of creep failure on solder joint of anchor at the condition of low stress(4kg) and high temperature(75℃). The microstructure of SnPb eutectic solder was composed of Pb-rich α phase and Sn-rich β phase. As indicated by arrows in Fig. 3(b), micro cracks were observed between Pb-rich α phase and Sn-rich β phase or along Sn-rich β phases with different grain orientation. Voids and cavities liked on grain boundary normal to principal stress were seen in Fig. 3(c),(d). It was reported that the voids and cavities could be formed at the early stage

• f creep deformation especially perpendicular to

tensile stress and grow and coalesce to become micro cracks[3,5]. Although test temperature was decreased to minimum value under constant low stress, the failure mode was very similar with Fig. 3. Grain boundary sliding was prevalent and micro crack grows on the grain boundary under low applied stress. In this condition that applied load was relative low, it was revealed that intergranular fracture is predominant as creep fracture mode. (a) (b)

(c) (d) Figure 3 Cross-sectional SEM micrographs of (a) creep fracture on solder joint of anchor at the condition of low stress and high temperature, (b) magnified image of A region in (a), (c) magnified image of B region in (a) and (d) magnified image

• f interested area of (b)

Under the condition of medium load(6kg), microstructure of failed solder joint on board was shown in Fig. 4. As an applied load was increased, it was observed that a mircocrack grew along grain boundary and propagated inside grain at final stage. It was also observed that microcrack was penetrating into Sn-rich β phase. Voids were observed near crack tip as supplier on nucleation and growth of microcrack during creep process. Considering the failure mode with increased applied load of creep test, experiment condition shown in Fig. 4 seemed to be a transition region from intergranular fracture to transgranular fracture. (a) (b) (c) (d) Figure 4 Cross-sectional optical micrograph of (a) solder joint on board at the condition of medium load, SEM micrographs of (b) interested area of (a), (c) magnified image of A region in (b) and (d) magnified image of B region in (b) Micro crack penetrated into Pb-rich α phase and Sn-rich β phase along tensile stress axis under high stress(8kg), as shown in Fig. 5. With

• bservation for the failure mode with temperatures

under high constant stress it was revealed that a transglanular mode was predominant under relative high stress and the effect of temperature was more negligible than that of stress. According to previous studies, fracture mode was transgranular in limited range of creep where stress flow was stabilized and void coalescence was postponed due to the strain- rate dependence of creep[2]. As mentioned on previous clause, the break in the creep rate-stress curve represented in power-law creep occurred at stress at which the fracture mode changed from intergranular to transgranular fracture[5]. In power- law creep regime, there are different n and Q between high-temperature(HT) creep and low- temperature(LT) creep, that is, failure mode is changed with the shift of failure mechanism by increasing the applied stress. Therefore, it was recognized that the failure mode of high stress condition was different with that of low and medium stress condition. Figure 5 Cross-sectional optical micrograph of (a) solder joint on board at the condition of high stress, cross-sectional SEM micrographs of (b) interested area of (a), (c) magnified image of A region in (b). Considering with

• verwriting
• ur

test conditions on SnPb eutectic solder creep mechanism map established by X. Q. shi et al,[4] as shown in

• Fig. 6, it was found that the diffusion creep was

involved on condition of low and medium stress whereas dislocation creep was involved on the condition of high stress. Although creep mechanism boundary between each region might be changed by grain size or by other process, it was found that our results listed in Table 2 corresponded with previous work[4].

Figure 6 Creep deformation map for 63Sn/37Pb eutectic solder alloy by X.Q.Shi et al.[4] with tested conditions Table 2 Results of fracture analysis with applied load and temperature Applied Load Temp. (℃) Fracture Mode Creep Mechanism 4kg 35 Intergranular Diffusion 55 Intergranular Diffusion 75 Intergranular Diffusion 6kg 35 Intergranular & Transgranular Diffusion & Dislocation 55 Intergranular & Transgranular Diffusion & Dislocation 75 Intergranular & Transgranular Diffusion & Dislocation 8kg 35 Transgranular Dislocation 55 Transgranular Dislocation 75 Transgranular Dislocation Conclusion In order to build an accelerated life model for creep rupture of SnPb eutectic solder, the accelerated tests were conducted on several conditions of temperature and stress considering creep mechanism. As solder joints were failed, fractograpghy of solder joint was analyzed with SEM and FIB. We found that the failure mode was shifted from intergranular fracture to transgranular fracture with increased applied stress. We suggested

• ptimal conditions of temperature and stress for

accelerated creep rupture test on SnPb eutectic solder joints. Acknowledgements Authors would like to thank Heesun Lee, Jiho Kim for supporting failure analysis work. References [1] R.J.Klein Wassink, “Soldering in Electronics”, electrochemical publication, second edition, Bristol, chapter 4, pp135, 1989. [2] George E. Dieter, “Mechanical Metallurgy”, McGraw-Hill, SI Metric edition, London, chapter13, pp.423, 1988 [3] Chin-Kuang Lin, De-You Chu, “Creep rupture

• f lead-free Sn-3.5Ag and Sn-3.5Ag-0.5Cu

Solders”, Journal

• f

Materials Science: Materials in Electonics, Vol 16, pp355-365, 2005. [4] X. Q. Shi, Z.P. Wang, Q,J, Yang, and H.L.Pang,”Creep behavior and Deformation Mechanism Map of Sn-Pb Eutectic Solder Alloy”, Journal of Engineering Materials and Technology, Vol.125, pp 81-88, January, 2003. [5] Alan Liu, “Mechanics and Mechanisms of Fracture”, ASM international, first edition, chapter 2, pp105-111, 2005