Figure 7 - uploaded by David D. Hillman
Content may be subject to copyright.
Source publication
The electronics industry is undergoing a materials evolution due to the pending Restriction of Hazardous Substances (RoHS) European Directive. Printed wiring board laminate suppliers, component fabricators, and printed wiring assembly operations are engaged in a multitude of investigations to determine what leadfree (Pbfree) material choices best f...
Citations
... The peak temperature of a reflow profile depends on solder ball composition, ball/solder paste ratio, dwell time, and component size. A majority of studies found that full mixing in solder joints is directly related to the reliability of mixed assemblies [7]- [9], [19], and [20]. D. Hillman et al. [20] claimed that partial mixing led to early failure in the ATC test. ...
... A majority of studies found that full mixing in solder joints is directly related to the reliability of mixed assemblies [7]- [9], [19], and [20]. D. Hillman et al. [20] claimed that partial mixing led to early failure in the ATC test. ...
Plastic ball grid array packages (Sn<sub>3.0</sub>Ag<sub>0.5</sub>Cu) were assembled with Sn<sub>3.0</sub>Ag<sub>0.5</sub>Cu and Sn<sub>37</sub>Pb solder paste to form Pb-free and mixed BGA assemblies. The Pb-free and mixed assemblies were subjected to one (1X) rework cycle and three (3X) rework cycles. All the reworked Pb-free and mixed assemblies were then subjected to a temperature cycling test with a temperature range of -55°C-125°C. The cycles to 0.1% failure of 1X reworked Pb-free assemblies was 18% smaller than that of nonreworked Pb-free assemblies, and the cycles to 0.1% failure of 3X reworked Pb-free assemblies was approximately 100% smaller. However, the cycles to 0.1% failure 1X reworked mixed BGA assemblies was approximately 200% larger than that of the nonreworked mixed BGA assemblies, and the cycles to 0.1% failure 3X reworked mixed assemblies was 35% larger. Detailed microstructural analysis and geometry analysis were provided to explain the temperature cycling reliability differences between the nonreworked and reworked assemblies, as well as the differences between the Pb-free assemblies and the mixed assemblies. The increase in the temperature cycling reliability of reworked mixed assemblies was found to be related to more homogenous Pb-rich phase distribution in the bulk solder.
... In these cases, the degree of mixing was controlled by adjusting the peak reflow temperature and the dwell at the peak temperature. Further examples of poor mixing can be found in the literature [8][9][10]. Maintaining solder joint reliability is the key challenge in the transition to lead free, and questions always occur when the mixing of tin-lead and lead free materials and processes occur. ...
The transition from tin-lead to lead free soldering in the electronics manufacturing industry has been in progress for the past 10 years. In the interim period before lead free assemblies are uniformly accepted, mixed formulation solder joints are becoming commonplace in electronic assemblies. For example, area array components (BGA/CSP) are frequently available only with lead free Sn-Ag-Cu (SAC) solder balls. Such parts are often assembled to printed circuit boards using traditional 63Sn-37Pb solder paste. The resulting solder joints contain unusual quaternary alloys of Sn, Ag, Cu, and Pb. In addition, the alloy composition can vary across the solder joint based on the paste to ball solder volumes and the reflow profile utilized. The mechanical and physical properties of such Sn-Ag-Cu-Pb alloys have not been explored extensively in the literature. In addition, the reliability of mixed formulation solder joints is poorly understood. In this work, we have explored the physical properties and mechanical behavior of mixed formulation solder materials. Seven different mixture ratios of 63Sn-37Pb and SAC305 solder materials have been formed, which include five carefully controlled mixtures of the two solder alloys (by weight percentage) and the two extreme cases (pure Sn-Pb and pure SAC). For the various percentage mixtures, the melting point, pasty range, stress-strain curves, mechanical properties (modulus, strength), and creep curves have been characterized. The variations of the mechanical properties and creep rates with aging at room temperature (25degC) and elevated temperature (100degC) have also been measured. Finally, the microstructures realized with the various mixtures have been found and correlated to the mechanical measurements and microstructures found in actual mixed formulation BGA solder joints. The results for the mechanical and physical properties show a very complicated dependence on the mixture ratio.
... There are not enough data on mixed Pb-free components / Sn-Pb solder available and the existing data are inconsistent [1][2][3][4][5][6][7][8][9][10][11][12]. It is generally believed that it is desirable to have a SAC ball that is completely molten, in order to form a well-mixed alloy upon solidification. ...
... Interfacial Failure. Unlike the data obtained in Celestica RIA projects, there are data reported in literature on interfacial failure in mixed joints [1,2,5,11]. We also observed this type of failure working with our customers on their assemblies, and in different consortium projects. ...
The results of an intensive reliability study on Pb-free ball grid array (BGA)/Sn-Pb solder assemblies as well as some lessons
learnt dealing with mixed assembly production at Celestica are described in this paper. In the reliability study, four types
of Pb-free ball grid array components were assembled on test vehicles using the Sn-Pb eutectic solder and typical Sn-Pb reflow
profiles with 205°C to 220°C peak temperatures. Accelerated thermal cycling (ATC) was conducted at 0°C to 100°C. The influence
of the microstructure on Weibull plot parameters and the failure mode will be shown. Interconnect defects such as nonuniform
phase distribution, low-melting structure accumulation, and void formation are discussed. Recommendations on mixed assembly
and rework parameters are given.
... Several studies [7], [8], [13]- [18] have been conducted to investigate the reliability of the mixed solder joints subjected to various loading conditions and metallurgical combinations. For Pb-free ball-grid-array (BGA) components assembled with Pb-based solder, the reliability has been shown to be equivalent to completely Pb-free assemblies, provided that the Pb is distributed evenly throughout the joint. ...
... For Pb-free ball-grid-array (BGA) components assembled with Pb-based solder, the reliability has been shown to be equivalent to completely Pb-free assemblies, provided that the Pb is distributed evenly throughout the joint. However, significantly, earlier failures can occur if the Pb is not distributed in the mixed solder joint [13]. ...
The global transition to lead-free (Pb-free) electronics has led component and equipment manufacturers to transform their tin-lead (SnPb) processes to Pb-free. At the same time, Pb-free legislation has granted exemptions for some products whose applications require high long-term reliability. However, due to a reduction in the availability of SnPb components, compatibility concerns can arise if Pb-free components have to be utilized in a SnPb assembly. This compatibility situation of attaching a Pb-free component in a SnPb assembly is generally termed "backward compatibility." This paper presents the results of microstructural analysis of mixed solder joints which are formed by attaching Pb-free solder balls (SnAgCu) of a ball-grid-array component using SnPb paste. The experiment evaluates the Pb phase coarsening in bulk solder microstructure and the study of intermetallic compounds formed at the interface between the solder and the copper pad.
... A significant number of experimental studies have been done recently on investigating the solder joint reliability of BGA/CSP backward compatibility using various reflow profiles [5,6,7,8,9,11,12,13,14,15,16]. ...
... It is evident that the reliability of solder joint interconnections in backward compatibility assemblies degrades significantly if SnAgCu solder spheres are only partly melted in backward compatibility. Hillman et al. evaluated the reliability of a BGA package assembled using a peak reflow tempera ture of 215°C with the duration time above 200°C at 40 seconds [12]. They observed partial mixing of Pb in the joint microstructure. ...
... Hunt and Wickham concluded that there should be few solder joint reli ability problems when mixing SnPb and lead-free components and solder alloys (with lead contamination in the range of 1 to 10%) [26]. Therefore, it is difficult to draw a conclusion regarding the effect of Pb content on backward compatibility reliability although there is evidence to show that it could be detrimental [12]. For the case where the lead-free SnAgCu paste is assembled with SnPb BGA/CSP components, if the voiding is excessive, this may lead to reliability issues from excess voiding which reduce the effective solder cross-sectional area or if the ball size and pitch is small, bridging may occur between adjacent spheres. ...
In response to the European Union (EU) Restriction of Hazardous Substances (RoHS) and other countries’ impending lead-free
directives, the electronics industry is moving toward lead-free soldering. Total lead-free soldering requires not only lead-free
solder paste but also lead-free printed circuit board (PCB) finish and lead-free component/packages. Transitioning tin-lead
(SnPb) soldering to totally lead-free soldering is a complex issue and involves movement of the whole electronics industry
supply chain. In reality, there is a transition period.
... Studies [1][2][3][4] show that the reliability of solder interconnections degraded significantly when the SnAgCu ball is only partially mixed with the SnPb paste. Gregorich & Holmes [1] reported that the reliability of backward compatibility assembly when the mixed assembly was reflowed at peak temperature of 200°C was much poorer than that of the control SnAgCu ball with SnAgCu paste in both the accelerated temperature cycling test from -40°C to +125°C and the mechanical shock test. ...
... The reliability of backward compatibility assembly improves as the reflow temperature increased to 225°C. Hillman, et al. [2] evaluated the reliability of a BGA package assembled using a peak reflow temperature of 215°C with the duration of time above 200°C at 40 seconds. They observed partial mixing of Pb in the joint microstructure. ...
... Although a considerable amount of work [1][2][3][4][5][6][7][8][9][10][11][12] has been done so far on the backward compatibility assembly and its reliability, the minimum temperature able to achieve the full mixing is still unknown. The key in backward compatibility assembly is to develop a reflow profile with the peak temperature high enough to be able to achieve full mixing of the SnPb paste and the SnAgCu ball, and the peak temperature low enough (prefer below 220°C) so that SnPb components won't be damaged. ...
Recently, the soldering of lead-free components with SnPb paste, or lead-free backward compatibility, is becoming a hot topic. One of the major challenges in backward compatibility assembly is the development of a right reflow profile for the soldering of SnAgCu ball grid array (BGA)/chip scale package (CSP) components with SnPb paste. If the SnAgCu reflow profile is used, the reflow temperature may be too high for other SnPb components in the same board during assembly according to the component rating per IPC/JEDEC J-STD-020C. In addition, the flux in SnPb solder paste may not function properly in such a high reflow temperature. On the other hand, if the SnPb reflow profile is used, SnAgCu solder ball may only partially melt. The incomplete mixing of the solder paste with the BGA/CSP ball raises serious reliability concern. Therefore, it is important to know the minimum reflow peak temperature that is able to achieve complete mixing of SnPb paste with lead-free components. This paper presents a method to estimate the liquidus temperature of mixed compositions when SnAgCu BGA/CSP components are soldered with SnPb paste. The liquidus temperature is the minimum reflow peak temperature able to achieve complete mixing of SnPb paste with lead-free components. It is shown that the liquidus temperature depends on the Pb ratio in the mixed composition and the liquidus temperature is below 217degC, which is the liquidus temperature of SnAg3.0Cu0.5 solder. The liquidus temperatures of several experimental studies in literature are estimated and it is found that the estimated temperatures are consistent with experimental results. A user interface is designed using visual basic for application in the Microsoft Excel environment to facilitate the estimation of the liquidus temperature. It is expected that the estimation of the mixed compositions liquidus temperature are able to guide process engineers to develop a right reflow profile in backward compatibility assembly
... Studies [1][2][3][4] show that the reliability of solder interconnections degraded significantly when the SnAgCu ball is only partially mixed with the SnPb paste. Gregorich & Holmes [1] reported that the reliability of backward compatibility assembly when the mixed assembly was reflowed at peak temperature of 200°C was much poorer than that of the control SnAgCu ball with SnAgCu paste in both the accelerated temperature cycling test from -40°C to +125°C and the mechanical shock test. ...
... The reliability of backward compatibility assembly improves as the reflow temperature increased to 225°C. Hillman, et al. [2] evaluated the reliability of a BGA package assembled using a peak reflow temperature of 215°C with the duration of time above 200°C at 40 seconds. They observed partial mixing of Pb in the joint microstructure. ...
... Although a considerable amount of work [1][2][3][4][5][6][7][8][9][10][11][12] has been done so far on the backward compatibility assembly and its reliability, the minimum temperature able to achieve the full mixing is still unknown. The key in backward compatibility assembly is to develop a reflow profile with the peak temperature high enough to be able to achieve full mixing of the SnPb paste and the SnAgCu ball, and the peak temperature low enough (prefer below 220°C) so that SnPb components won't be damaged. ...
Recently, the soldering of lead-free components with SnPb paste, or lead-free backward compatibility, is becoming a hot topic. One of the major challenges in backward compatibility assembly is the development of a right reflow profile for the soldering of SnAgCu Ball Grid Array (BGA)/Chip Scale Package (CSP) components with SnPb paste. If the SnAgCu reflow profile is used, the reflow temperature may be too high for other SnPb components in the same board during assembly according to the component rating per IPC/JEDEC J-STD-020C. In addition, the flux in SnPb solder paste may not function properly in such a high reflow temperature. On the other hand, if the SnPb reflow profile is used, SnAgCu solder ball may only partially melt. The incomplete mixing of the solder paste with the BGA/CSP ball raises serious reliability concern. Therefore, it is important to know the minimum reflow peak temperature that is able to achieve complete mixing of SnPb paste with leadfree components. This paper presents a method to estimate the liquidus temperature of mixed compositions when SnAgCu BGA/CSP components are soldered with SnPb paste. The liquidus temperature is the minimum reflow peak temperature able to achieve complete mixing of SnPb paste with lead-free components. It will be shown that the liquidus temperature depends on the Pb ratio in the mixed composition and the liquidus temperature is below 217°C, which is the liquidus temperature of SnAg3.OCuO.5 solder. The liquidus temperatures of several experimental studies in literature are estimated and it is found that the estimated temperatures are consistent with experimental results. A user interface is designed using Visual Basic for Application in the Microsoft Excel environment to facilitate the estimation of the liquidus temperature. It is expected that the estimation of the mixed compositions liquidus temperature will be able to guide process engineers to develop a right reflow profile in backward compatibility assembly.
Accelerated temperature cycling (ATC) was used to assess the thermal fatigue reliability of a Pb free, 37.5 mm fully populated, 1295 I/O ball grid array (BGA) package assembled with backward compatible or mixed alloy (Pb free BGA/SnPb paste) processing. The Pb-free BGA components were fabricated with Sn-4.0Ag-0.5Cu (SAC 405) solder balls. The surface mount assembly was done using several custom SnPb eutectic soldering profiles designed to produce different levels of Pb mixing in the BGA solder balls. The test program also included SAC405-SAC405 assemblies for reliability comparisons. The temperature cycling and metallographic data indicate that assembly parameters that produce the lowest level of Pb mixing degrade the solder joint fatigue reliability. Failure analysis demonstrates that the lower reliability is due to poorly formed solder joints resulting from incomplete reflow. The failed solder joints exhibit inadequate BGA ball collapse, minimal Pb mixing, and fatigue failure at the printed circuit board (PCB) side of the joints instead of the package- side fatigue failures that are typical for ATC testing. The test cells with more complete Pb mixing and the pure Pb-free test cell exhibited better fatigue reliability. However, failure analysis of these samples revealed that the data were compromised by the occurrence of a second failure mode in plated through hole vias that were incorporated into the PCB test vehicle. The presence of mixed mode failures (solder joint and via) precluded direct quantitative comparisons among all the test cells. The results are discussed within the context of previously published test results from the literature, including the impact of the contribution of via failures. Recommendations are outlined for additional testing to quantify the relationships between Pb mixing level and Pb- free reliability for this high density BGA package.
