Industry demands environmentally friendly processes
EDWIN BRADLEY
JASBIR BATH
GORDON WHITTEN
SRINIVAS CHADA
There has been increasing interest in the electronics assembly community regarding environmentally friendly assembly processes, including lead-free assembly, because of regulatory and consumer pressure from Europe and Japan. In the past year, leading Japanese original equipment manufacturers (OEMs) have made numerous pronouncements about their goals to reduce and eliminate lead in their products, and they have demonstrated that such “green” products can increase market share. In Europe, the proposed Waste Electrical and Electronic Equipment (WEEE) directive would require the elimination of several substances, including lead solders, in many electronic products by 2004. Final voting is not due until later this year, so some revisions to WEEE are possible, but it appears to have substantial political backing. The result of this legislation, coupled with substantial activity in Japan to eliminate lead, means that this issue cannot be ignored.
In response to these initiatives, the National Electronics Manufacturing Initiative (NEMI) formed a lead-free task force in early 1999 to investigate process and material considerations for lead-free electronics assemblies. More than 30 organizations participated in this task force, including leading OEMs, electronics manufacturing services (EMS) providers, component providers, solder suppliers, government agencies and universities. The group focused on the technical issues surrounding lead-free implementation, purposely avoiding any efforts to influence legislative mandates requiring lead-free assemblies.
This group came to the consensus that the North American electronics industry should develop the capability to produce lead-free products by 2001, with an eye toward total elimination of lead by 2004. The task force also mapped out a plan of action to achieve this goal.
Given the enormity of the task, the NEMI project has been divided into several specific activity-driven groups: solder alloy, components, solder reliability and process development. Additional subgroups, such as wave soldering, inspection/test and equipment, will be formed as necessary, while existing groups will be dissolved once deliverables are completed. Current group chairs and co-chairs are listed in Table 1.
Lead-free Solder Alloy
The lead-free solder alloy subgroup, which consists of representatives from major solder manufacturers, end users and various governmental organizations, will make recommendations to the reliability and component subgroups for a standardized lead-free solder alternative.
Candidate materials have been assessed according to relative strengths and weaknesses so the subgroup can recommend a lead-free alloy for reflow soldering and two alloys for wave soldering. These alloys will be evaluated by the component and reliability subgroups alongside the Sn37Pb control.
The subgroups` activities have focused on the SnAgCu solder family of alloys because the SnAgBi family have the following issues:
– Lead-contamination from plating: Sn-Bi-Pb eutectic with a melting point of 96∞C can form in solder joints containing those elements,1 causing major reliability concerns.
– Recycling: Copper smelters have a major problem with bismuth contamination of copper.
– Fillet lifting: SnAgBi solders have problems with fillet lifting in wave-soldered joints, and it was deemed the advantage of the less than 7∞C difference in melting temperature for SnAgBi over SnAgCu and SnAgBi solders may only be needed in niche applications.
SnAgCu Eutectic Alloy Composition Determination: Melting point data produced at the National Institute of Science and Technology using a variety of SnAgCu alloy compositions was compared to data from Marquette University and Northwestern University to determine that the ternary eutectic had a melting temperature of 216-217∞C with a composition approximately Sn-3.74Ag-0.85Cu.2 Alloys with compositions within the range Sn3.5 4wt%Ag0.5-1wt%Cu have a melting point of 217∞C with similar microstructures and mechanical properties.
Material Property Data: Solderability data for the lead-free alloys of interest. Various solder manufacturers contributed. Solderability of the candidate alloys, as measured by the wetting balance, was adequate, although not as good as for Sn37Pb solder. Additions of less than 1 percent antimony to SnAgCu alloys did not have a significant effect on solderability. Reliability data for the SnAgCu lead-free alloys indicated that they would have similar or better reliability than Sn-37Pb.
Alloy Selection: After the alloy review was completed, a straw poll was sent to 14 companies engaged in electronics assembly for their opinions about the leading replacement alloys identified. The companies were asked to consider use of these alloys for both reflow soldering and wave soldering. They ranked their first, second and third choices (5, 3 and 1 points respectively) for reflow and wave soldering in order of preference for further tests to be conducted by the component and reliability subgroups. Two SnAgCu alloys were placed in the ballot: the Sn-3.9Ag-0.6Cu (a compromise between two alloys that have been sampled widely, Sn-3.8Ag-0.7Cu and Sn-4Ag-0.5Cu) and Sn-3.5Ag-0.7Cu, the primary ternary alloy being investigated in Japan3 (Tables 2 and 3).
Results from the straw poll identified Sn3.9Ag0.6Cu as the primary choice for reflow soldering. This alloy is being subjected to further trials by the component and reliability subgroups using Sn-37Pb as a control. The Sn3.9Ag0.6Cu alloy also fits into the alloy range recommended by the International Tin Research Institute4 in the United Kingdom.
First and second choices for wave soldering were Sn0.7Cu and Sn3.5Ag respectively. The Sn0.7Cu solder was a candidate primarily because of its low metal cost (which is a factor when considering wave solder alloys) and because the effect of temperature on the components in wave soldering is considered to be less of an issue than for reflow soldering because the components do not reach temperatures as high as in reflow sol dering and the duration at these temperatures is much shorter. However, the reliability of Sn0.7Cu solder in through-hole applications is an issue the reliability subgroup will address.
As part of ongoing work, a process development team is being formed to help determine standard reflow profiles and wave soldering profiles for effective lead-free soldering.
Components Subgroup
A major challenge in lead-free soldering with SnAgCu alloys with a 217∞C melting point is the survivability of components and printed circuit boards (PCB) within the higher process (reflow and wave) temperature environment. Additionally, compatibility is an issue of lead-free solders with solderable surfaces on the components and PCBs.
Temperature/Moisture Survivability: Components and PCBs will be evaluated in two phases. In the first phase, test-ing high temperature and moisture sensitivity/worthiness of existing PCBs and components will be carried out using a reflow profile based on J standard 020, but with a peak temperature of 260∞C followed by a liquid-to-liquid thermal shock (LLTS) ranging from negative 55∞C to +125∞C for up to 2,000 cycles. It is imperative that PCB components maintain functionality, structure and shape without individual layer delamination and barrel cracking of through-holes.
Moisture level performance testing is planned for a variety of components, such as small-outline integrated circuits, thin quad flat packs, chip scale packages (CSP), ball grid arrays (BGA) and leadless chip carriers. In the first phase, the components will be tested by component manufacturers within NEMI, based on the profile shown in Table 4. The test will also be repeated with peak temperatures of 250∞C and 240∞C in the event of very high failure rates from the 260∞C test. Once established, these test parameters will be proposed to replace the existing J standard 020 to reflect the changes required to accommodate lead-free soldering.
Leadframe/PCB Finishes: There are currently a number of lead-free PCB surface finishes, including organic solderability preservatives, organo-silver, immersion gold, lead-free hot-air soldering leveling (HASL) and electroless tin. All of these, except for HASL, are in production and are not expected to pose major problems. For lead frames, the leading lead-free metallizations include pure tin, tin-copper alloy, nickel-palladium and nickel-gold. Besides exhibiting good wetting characteristics with various combinations of lead-free solders, it is critical for the integrity of the package that the lead finishes produce reliable wire bonds and superior adhesion to molding compounds.
The wetting characteristics of PCB and lead finishes will be established utilizing wetting balance and solder spread tests. Concurrent to that testing, the component and process development subgroups will evaluate solder joints assembled using the primary PCB surface finishes and component lead finishes with the Sn-3.9Ag-0.6Cu alloy chosen by the alloy subgroup; this will be followed by solder joint evaluation and ranking of the solder, finish and termination combination based on the joint appearance and cross sections. A reflow profile that reflects the processing peak temperature (based on results of the first phase) and parameters will be used for this part of the evaluation.
The PCB characterization will be carried out by the Interconnection Technology Research Institute working with the component subgroup.
Solder Joint Reliability Subgroup
Given the long history of using SnPb solders in assemblies, there is a great deal of reliability data on this system. Earlier work has demonstrated that the reliability of the SnAgCu family of alloys is superior to SnPb in many applications,5,6 but there are gaps in the data set, especially for BGA, CSP and flip chip components. This group is chartered to:
– Identify current tests used to determine reliability.
– Develop information on the performance of the chosen lead-free alloy(s) through literature review, internal testing and external testing.
– Identify new tests that may be required by the new processing and materials.
— Identify industry reliability standards that may be impacted by lead-free processing.
– Propose changes for the standards identified.
A major reliability risk identified by earlier work was “fillet lifting.” It describes the separation of the fillet from the pad for a through-hole joint (Figure 1); this is believed to be caused by “hot tearing,” which results from the solidification of the high temperature phase of a wide pasty range alloy during the cool-down and resulting contraction of the higher CTE FR4.7 In truth, the fillet does not lift, but the pad drops away because of the contraction.8 This effect is aggravated further by the rapid increase in the Z-axis coefficient of thermal expansion above the Tg.
A number of solder reliability tests are planned for a cross-section of component types.
Thermal Shock: This is defined as a thermal ramp rate that exceeds 20∞C/min, typically done in a two-chamber test with a 15- to 20-minute soak at the extremes.
Thermal Cycling: This is defined as a thermal ramp rate that is less than 15∞C/min and is typically conducted in a test chamber with a tightly controlled thermal profile. The proposed JEDEC JESD22-A104-A test method B4 with -40∞C to +125∞C and a 15-minute soak is very interesting to project members.
Bend Test: This test is intended to simulate the damage that occurs in the field by physical handling and out-of-plane deformation, especially related to BGA components. A board with a component mounted on it is suspended between two supports, a specified deflection is applied to the board, and the part is inspected for any cracks or electrical defects (Figure 2).
High Temperature Soak: This test explores the growth of intermetallics, such as Cu3Sn and Cu6Sn5 and their affect on reliability.
Vibration: This test identifies any resonance modes that may be excited in the field. It is a useful test when conducted with high temperature soak or thermal cycle testing.
Electromigration: The Bellcore GR-78 and the IPC-TM-650 method 2.6.3.3 are being considered.
Conclusions
NEMI is committed to develop the capabilities for North American companies to initiate lead-free assembly by the year 2001, with an eye toward total lead elimination by 2004. The consortium has successfully focused attention on the technical issues surrounding lead-free electronics assemblies and built a strong team to drive and evaluate the solutions that are promoted by the industry.
References
1. Z. Mei, F. Hua, J. Glazer, “Thermal Reliability of 58Bi-42Sn Solder Joints on Pb-Containing Surfaces,” Proceedings of Design and Reliability of Solders and Solder Interconnects, The Minerals, Metals & Materials Society, 1997, p. 229-239.
2. C. Handwerker, NEMI Alloy Presentation, Oct. 1999.
3. K. Suganuma, “Researches and Developments for Lead-Free Soldering in Japan,” IPC Works 99 Proceedings, Oct. 23-28, 1999. pp. S-01-5-1 -S-01-6-8.
4. “Lead-free Alloys – The Way Forward,” Soldertec (ITRI) Limited, October 1999. www.lead-free.org.
5. National Center for Manufacturing Sciences Lead-Free Solder Project Final Report, Aug. 1997, NCMS, 3025 Boardwalk, Ann Arbor, MI 48108-3266.
6. J.H. Vincent and M.R. Harrison, “Improved Design Life and Environmentally Aware Manufacturing of Electronic Assemblies by Lead-Free Soldering,” IMAPS `99.
7. C. Handwerker, NIST, NEMI Meeting, Oct. 1999.
8. G. Whitten, “Lead Free Solder for Automotive Electronics,” Proceedings of the SAE, March 1998.
Acknowledgements
The authors would like to thank Elizabeth Bennedetto (Compaq), Mark Kwoka (Intersil), Carol Handwerker (NIST), Ken Snowden (Nortel Networks), and Bob Smith and Ron Gedney (NEMI), as well as the many participants in the NEMI task group.
EDWIN BRADLEY, senior staff engineer, can be contacted at Motorola, 8000 W. Sunrise Blvd. Rm. 2333, Plantation, FL 33351; 954-723-3865; E-mail: [email protected]. JASBIR BATH, manufacturing engineer, can be contacted at Solectron Corp., Solectron Technical Center, 637 Gibraltar Court, Building 1, Milpitas, CA 95035; 408-957-2935; Fax: 408-956-6083; E-mail: [email protected]. GORDON C. WHITTEN, Ph.D., advanced packaging, process and materials, can be contacted at M/S: 8186, Delphi Delco Electronics Systems, One Corporate Center, P.O. Box 9005, Kokomo, IN 46904-9005; 765-451-1168; Fax: 765-451-1728; E-mail: [email protected]. SRINIVAS CHADA, senior materials scientist, can be contacted at Motorola, 8000 West Sunrise Blvd. Rm. 2329, Plantation, FL 33322; 954-723-5293; Fax: 954-723-5584; E-mail: [email protected].
 |
 |
 |
 |
 |
Figure 1. Fillet lifting or separation from pad with Sn3.5Ag solder (Source: NCMS Lead-Free Solder Project Final Report).
 |
Figure 2. Schematic of the bend test.