-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueIPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
Boost Your Sales
Every part of your business can be evaluated as a process, including your sales funnel. Optimizing your selling process requires a coordinated effort between marketing and sales. In this issue, industry experts in marketing and sales offer their best advice on how to boost your sales efforts.
The Cost of Rework
In this issue, we investigate rework's current state of the art. What are the root causes and how are they resolved? What is the financial impact of rework, and is it possible to eliminate it entirely without sacrificing your yields?
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Isothermal Aging of Lead-free Joints
December 31, 1969 |Estimated reading time: 11 minutes
In the age of lead-free, many assemblers have chosen one of the SAC alloys recommended for lead-free soldering. This change also has prompted a change in the choice of board surface finishes. This article details a study designed to understand the formation of IMC phases at the interface and in the volume of surface mount joints soldered with SAC 305 and tin/lead alloys using several surface finishes.
By Mercedes Chacon and Jorge A. Manriquez, Ph.D., Tecnologico de Monterrey and Jose L. Mendoza and Brian Toleno, Ph.D, the electronics group of Henkel
ith the RoHS-compliance deadline a thing of the past and lead-free implementation in full swing, many assemblers have selected one of the SAC alloys recommended for lead-free soldering. In addition to a change in solder alloy, this implementation also has necessitated a change in the choice of board surface finishes.1-3 For tin/lead assemblies, the most common surface finish used is tin/lead hot-air solder leveling (HASL). With lead removed from their assembly materials, many are moving from a HASL finish to either organic solder preservative (OSP), immersion silver (ImAg), immersion tin (ImSn), or electroless nickel/immersion gold (ENIG).
Figure 1. Test vehicle for the isolthermal aging study.
The practical relevance of this chemistry change has to do with potential harmful effects that a new soldering system (alloy and process) might have on the reliability of the formed joints. The prevailing industry view is that the weakest link in a solder joint is the solder/substrate interface. Moreover, it is known that the presence of a thin and continuous intermetallic compound (IMC) layer at such interface is necessary to ensure a strong metallurgical joint. However, a thick IMC might compromise this by statistically providing more opportunities for crack initiation/propagation, and may reduce the volume of ductile material available to sustain the strains that arise due to differences in coefficient of thermal expansion (CTE) across the solder joining plane.3-6 Consequently, the evolution of the microstructure through extended aging must be slow enough, and in such a direction that it would not cause a fast reduction of the overall mechanical strength of the joint.7-8
This article describes the formation of IMC phases at the interfaces and in the volume of surface mount joints soldered with both 96.5%Sn 3.0%Ag 0.5%Cu (SAC 305) and 63%Sn37%Pb alloys on PCBs with OSP, ImAg, ImSn, and ENIG coating finishes. The test vehicles were designed with 0603 and 0402 resistors, QFPs, and 0.5-mm-pitch BGA components, arranged separately according to daisy-chain circuits. The boards were subjected to isothermal aging at 125°C for 48, 192, 432, and 768 hours; the growth of various intermetallic phases was studied. The chemical nature of the phases, as well as the morphology of the microstructure formed, were marked by the solder-coating system. The growth rate of the total aggregate phases depended on the system too, but in a less marked way. IMC thickness of formed species is presented for the distinct alloys and PCB finishes. Overall thickness was 2-6 μm. Structural damage on the joints as induced by aging also was present.
Figure 2. Reflow profile used to produce the samples for the isothermal aging study. This diagram illustrates the lead-free solder system profile. Peak temperature and reflow time were optimized for the best reflow behavior.
Previous studies on lead-free assemblies have shown that tin is still the chemical element defining IMC formations at solder-pad interfaces in a joint. A layer of Cu6Sn5 remains the main compound forming at the interface when a SAC or Sn0.7Cu solder is reflowed over a tin- or OSP-finished copper pad.9-12 Prolonged residence times at high temperatures may cause interdiffusion of copper and tin across the interface to form a copper-rich Cu3Sn layer next to the pad. The small amount of silver contained in the SAC solder does not seem to play a major role in interface-compound formation, as it rapidly interacts with the tin to form Ag3Sn particles that are distributed homogenously throughout the entire joint volume.
Soldering with SAC over silver-finished pads has been shown to produce the same Cu6Sn5 and Cu3Sn layers as with tin finishes. However, small Ag3Sn platelets were observed protruding from the Cu/Sn IMC layers toward the bulk joint. These platelets reportedly form during reflow, and disappear upon further aging - having little effect on the interface microstructure.13 The silver from the finish gets dissolved into the liquid solder, and eventually contributes to the formation of the Ag3Sn particles inside the solder volume.
When a SAC solder is reflowed over an ENIG finish, gold reportedly dissolves rapidly into the liquid tin-rich solder, allowing the formation of Ni3Sn4 in the same way as the Cu3Sn layer forms on the tin-finished pads. The dissolved gold interacts with tin to form large AuSn4 regions distributed close to the interface.14
With respect to the growth rate of IMC layers, studies show that all observed interface layers grow to a thickness range of 4-8 μm when aged at 125°C for up to 400 hours.9-15 These studies have shown a continuous and uniform copper/tin layer to have the lowest growth among the observed IMCs. Conversely, the IMC layers forming on the silver finish appear to have an irregular growth front compared to others. Studying the evolution of bulk and interface microstructure is of great importance because it is a simple way to find out about possible detrimental effects normally observed with complete reliability studies, such as thermal cycling.
Sample Preparation and Acceptance
The test vehicle for this study is shown in Figure 1. The SMT board was designed to carry eight 0.5-mm-pitch CABGAs, two 0.5-mm-pitch QFPs, and 50 0603 and 0402 resistors daisy-chain connected to a 32-pin-thru-hole (PTH) connector. All components used in the study were lead-free finished. According to an experimental design (Table 1), four board finishes were studied: 0.5-μm-thick OSP, 20-μm-thick ImAg, 40-μm-thick ImSn, and 200-μm-thick ENIG. Solder chemistries used to assemble test boards included conventional tin/lead Sn63Pb37 alloy and SAC 305 lead-free system. Figure 2 shows the reflow profile used to produce samples for the study with the lead-free system. In both the lead-free and tin/lead samples, the reflow window (peak temperature and reflow time) was optimized to provide the best reflow behavior for a board containing four different components. Table 2 shows actual process conditions and standards used to produce and accept valid study samples. Twenty-four test boards were assembled in an automated surface mount line according to the experimental design (Figure 3). The assembled boards were accepted according to standard IPC criteria for fillet wetting and geometry. The set of boards was then placed in a box oven at 125°C (±3°C), and a set of eight boards (two chemistries × four finishes) was progressively removed after aging times of 48, 192, 432, and 768 hours.
Figure 3. Extended display of the set of 24 samples used in the study.
Every board, corresponding to one of the eight experimental conditions, was cut and resin-fixed to provide planar cross-sections suitable for thickness measurement of IMCs formed at solder-pad interfaces. Each cross-section was then ground, mirror-polished, and carbon-coated in preparation for analysis with a scanning electron microscope (SEM-EDX). When required, polished surfaces were treated with a 3%HCl/97% ethanol solution. This preparation was performed for each of the four distinct components from each of the eight distinct experimental conditions. Analysis of the microstructure and IMC-thickness measurements in the SEM were conducted with the backscattered electron unit at 20 KeV and calibrated working distances to provide best-chemical-phase resolution. Chemical analysis was conducted using an X-ray energy dispersive spectrometer unit (EDX).
Intermetallic growth
Figures 4-7 show the growth behavior of IMC layers formed at the solder-pad interface for each of the eight conditions studied. The values in the figures correspond to the average thickness of the aggregate layer of IMC formed, regardless of whether the layer consisted of one or several distinct chemical species. Because the fragile nature of the IMC affects the reliability of the interface to some degree, the global thickness should provide an indication of the possible degrading nature of such a compound layer.
The data collected from all conditions indicate that the growth rate is significantly slower for the joints soldered with the tin/lead system, regardless of pad finish. This behavior may be due to an increased activation energy needed for chemical species interdiffusing when significant amounts of a non-participating element such as lead are present - as opposed to SAC systems composed mainly of tin. Because most of the IMC layers contain tin, the difference in activation energy for diffusion of tin between the lead base and the lead-free systems should be the dominant factor for the growth of the IMCs.
Figures 4-7. Growth rate of the global IMC layer at the solder-pad interface of 0402 and 0603 resistors, QFPs, and BGAs, aged at 125°C for up to 432 hours.
Results also show that the growth rate of the global IMC layer in the tin/lead ENIG system is slower than all other cases. This is in agreement with previous reports; however, it appears that when lead is removed from the solder, the activation energy for tin and nickel interdiffusion reduces considerably, and the growth rate rises. Regardless of the volume and type of chemical species formed, the thickness of the global IMC up to 768 hours at 125°C was between 2-6 μm, which is consistent with previous findings.
IMC Formation and Evolution
The morphology of the IMCs formed at the interface clearly depends on the chemical nature of the solder-finish system. For example, the Cu6Sn5 in the tin/lead OSP system was smooth and continuous. Conversely, irregular layers of both Ni/Sn and Ni/Sn/Au combine to form an IMC aggregate layer at the interfaces of a SAC 305 ENIG-coated BGA system aged to 432 hours.
A peculiar nickel-rich intermetallic-like band that spanned the diameter of the BGA balls close to the solder/component interface was observed. These bands were present in the balls before isothermal aging was applied, but became more defined with the aging process and, apparently, induced damage by providing a favorable propagation path for cracks. The appearance of flat cracks was observed in some cases.
Conclusion
The chemical nature of the plating finish on PCB pads affects the formation of specific IMCs at the relevant interfaces. Pb37Sn63 solder over tin- and OSP-finished pads tend to form continuous and uniform Cu6Sn5 IMC layers after extended aging at 125°C. This is true on open joints such as 0402s, 0603s, and QFPs. The use of SAC 305 to solder the same components yielded a somewhat regular IMC layer when reflowed over tin- and OSP-finished substrates (Cu6Sn5); when SAC 305 was applied to ENIG or silver substrates, IMC layers became clearly irregular.
Regardless of plating-pad finish and type of IMC formed, IMC layers that were formed when using SAC 305 solder grew faster than when using Pb37Sn63. However, when a global measurement of IMC layers formed at the solder-pad interface was taken (regardless of the volume or type formed), a global average thickness grew no larger than 6 μm after isothermal aging at 125°C, and for up to 768 hours. Therefore, no detrimental effect on the overall amount of brittle layers is observed when using combinations of tin/lead SAC 305 solders with ImSn, ImAg, OSP, and ENIG finishes.
Isothermal aging at 125°C caused some structural damage (cracks), particularly to BGA joints at specific locations. Those locations were solder-pad interfaces with the Pb37Sn63 ENIG system containing a peculiar intemetallic-like Ni/Cu/Sn-rich band that formed near the solder-component interface of BGAs. An ongoing 125°C/-55°C thermal-cycling study should yield information on the possible effects of the observed cracks over the reliability of joints and the entire assembly.
REFERENCES
- Nat’l Center for Manufacturing Sciences, Lead-free Solder Project Final Report, August 1997; and SMT, pp. 3-8, June 2000.
- Greg Jones, “Are You Ready for Lead-free Assembly?” SMT, pp. 60-62, June 2000.
- Laura Turbini, Gregory C. Munie, Dennis Bernier, Jurgen Gamalski, and David W. Bergman, “Examining the Environmental Impact of Lead-free Soldering Alternatives,” IEEE Transactions on Components and Packaging Technologies, Vol. 24, No. 1, Jan. 2001.
- Solders and Soldering (3rd ed.), Howard H. Manko, McGraw-Hill, 1992; and (2nd ed.), Howard Manko, New York, 1979.
- Surface Mount Technology with Fine Pitch Components, Hans Danielsson (1st ed.), Chapman & Hall, 1995.
- J.S. Hwang and R.M. Vargas, Soldering and Surface Mount Tech, No.5, 38, 1990.
- J. S. Hwang, “Lead-free Solder: Sn/Ag/Cu and Sn/Ag/Bi Systems, SMT pp. 18-21, July 2000.
- Zhenfeng Guo, Holger Koenigsmann, Jennie S. Hwang, “High-strength and High-fatigue-resistant Lead-free Solder,” SMT, pp. 12-16, July 2000.
- Joreg A. Manriquez, Juan C. Cárdenas, and Brian Toleno, “Reliability of Tin Terminated Components in a Pb-free System,” SMTAI, Sept. 2005.
- K.N. Tu, et al., “Kinetics of Interfacial Reactions in Bimetallic Cu-Sn Systems,” Acta Metall., 30 (1982), pp. 947-952.
For a complete list of references (11-15) and figures, contact the authors.
ACKNOWLEDGEMENTS
This article was originally presented at iNEMI “International Conference on Lead-free Soldering,” Canada, May 2006. The authors thank Fernando Vega for technical assistance with board assembly and sample preparation.
Mercedes Chacon, associate engineer, Electronic Manufacturing Laboratory, Tecnologico de Monterrey, may be contacted via e-mail: mchacon@itesm.mx. Jorge A. Manriquez, Ph.D., professor and head of the Electronic Manufacturing Laboratory, Tecnologico de Monterrey, may be contacted via e-mail: jmanriquez@itesm.mx. Jose L. Mendoza, application engineer, the electronics group of Henkel, may be contacted via e-mail: joseluis.mendoza@mx.henkel.com. Brian Toleno, Ph.D., application engineering team leader, the electronics group of Henkel, may be contacted via e-mail: brian.toleno@us.henkel.com.