spacer
spacer spacer spacer
spacer spacer spacer
spacer
LeadFree Banner
spacer
spacer spacer spacer
 
Mike Bixenman   Erik Miller   Fernando Rueda  

Ten lessons learned in cleaning Pb-free flux residues

by Mike Bixenman, Erik Miller and Fernando Rueda
Kyzen Corporation

Kyzen Corporation 430 Harding Industrial Drive, Nashville, TN 37211
phone: 615-831-0888
www.kyzen.com

 

Introduction
Over the past two years, Kyzen has performed numerous Pb-free cleaning studies and evaluations at the company’s Application Testing Lab and at Cleaning Equipment Demonstration Centers. The evaluations consisted of Pb-free customer applications and designed experiments with solder paste and cleaning equipment companies. The results are a series of lessons learned, shared here to assist engineers in considering the process window for cleaning Pb-free circuit assemblies.


Application Cleaning Lab
Solder paste and equipment manufacturers are strategic in that one group makes the soils that assemblers clean and the other builds the equipment that make up a critical element of the cleaning process. Kyzen’s application cleaning lab’s cleaning equipment represents both inline and batch designs targeted to electronic assemblers (Figures 1 & 2) including aqueous inline and batch cleaning equipment; batch dishwasher, centrifugal, spray under immersion, and ultrasonic cleaning equipment; semi-aqueous spray under immersion inline and spray in air Microline (trade name of Speedline Technologies) cleaning equipment; centrifugal spray under immersion and ultrasonic cleaning equipment; stencil cleaning spray in air and ultrasonic cleaning equipment, vapor degreasing two and three sump cleaning machines; advanced vapor degreasing equipment; and board building, desoldering and inspection capabilities.


Figure 1: Kyzen’s cleaning application testing lab


Figure 2: The lab is fully equipped for cleaning studies

The analytical side of the lab performs quantitative and qualitative testing to determine board cleanliness and reliability analyses. Instruments include digital high definition microscopes, ion chromatography, SMT 600 omegameter, FT infra-red spectrophotometer, FT infra-red microscope, gas chromatography mass spec detector, gas chromatography flame ionization detector, ultraviolet visible diode-array spectrophotometer, and auto carl fischer titration. The analytical and application testing labs collaborate to develop best practices to innovate cleaning fluids that match up with industry needs.

The application lab averages from two to five new projects on a weekly basis. Project requests vary depending on the customer requirement. As engineers work to develop Pb-free electronic assembly processes, requests for cleaning process development projects have steadily increased. Over the last five years a number of Pb-free cleaning studies have been conducted, and the top ten lessons derived from those studies are presented here.

Lesson 1: The level of flux residue and its appearance is greater for most Pb-free materials.
Compatibility of Pb-free alloys with the flux composition is critical for performance in printing, shelf life, tack time, solder balling, wetting and solder joint appearance. The reduced wetting effect of high Sn (tin) alloys require more aggressive flux chemistries to achieve wetting comparable with eutectic Sn/Pb alloys. Solder paste fluxes must be formulated to address the specific alloy(s) of choice.

Pb-free flux compositions require increased oxygen barrier and flux capacity when reflowing in air. Thermally stable flux compositions require higher molecular weight resins to prevent decomposition of organic acids. Higher thermal reflow increase side reactions within the flux such as oxidation or cross-linking reaction. With the flux component representing upwards to 50% of the solder paste composition, the appearance and degree of residue increase.

Figures 3-6 are illustrations from customer parts sent to the application lab for cleaning process development. Figures 3 and 4 illustrate the level of flux residue and appearance for select Pb-free water soluble flux test at the application lab. Figures 5 and 6 illustrate flux residue and appearance from select low-residue no-clean solder pastes tested. The level of flux residue for both water soluble and low residue no-clean Pb-free solder pastes represent more flux residue as compared to eutectic Sn/Pb comparable assemblies. The appearance of the solder flux was a light amber and noticeable to the naked eye as compared to eutectic Sn/Pb that is transparent and not highly visible to the naked eye.


Figure 3: Water soluble solder paste flux residue solder


Figure 4: Water soluble solder paste flux residue reflowed in nitrogen


Figure 5: Low residue no-clean solder paste flux residue reflowed in air


Figure 6: Low residue no-clean solder paste residue in air reflowed in nitrogen

The lesson learned is that increased levels of flux residue force assembly houses to consider four essential wash process parameters: 1) wash time; 2) wash temperature; 3) wash concentration; and 4) impingement energy. Additionally, increased levels of flux correlate to higher levels of flux under area array components, chip caps, and fine pitch components. This does not mean that the parts are not cleanable, but it does mean that the wash process parameters may need to be increased.

Lesson 2: Thermal profile and time above liquidus increase the cleaning challenge.
The liquidus temperature of the SAC Pb-free alloy is 217-220ºC, which represents an increase of 34ºC above the melting point of Sn63Pb37. This higher melting range requires peak temperatures of 235-245ºC to wet and wick to achieve a good solder filet. Heavily populated boards increase the thermal mass differential across the board1. When developing a process window, higher peak reflow temperatures and longer time above liquidus increases cleaning difficulty. Notice from Figure 5 the thin layer of flux that flows and wets the board laminate. Extended thermal profiles cause the thin flux layers to decompose, oxidize and potentially polymerize. Figure 7 illustrates dry and hardened flux residue on the board laminate as the flux flows away from the solder filet. The flux penetrates into the pores of the solder mask rending a residue that is very difficult to clean.


Figure 7: Polymerized flux residue

The time from reflow to clean represents a variable that must be considered. The heating gradients during reflow drive the solvent molecules toward the center of the flux, creating a solvent rich zone. This solvent rich zone is softer and more easily penetrated and dissolved. Conversely, a solvent depleted zone is created next to the heated and exposed surfaces. When a zone is solvent depleted it becomes harder and more crystalline and thereby significantly more difficult for cleaning agents to soften and dissolve.

The eutectic Sn/Pb flux innovations over the past ten years provided low residue, cleanable flux, and multiple reflow cycles before cleaning the assemblies. Instead of cleaning after each reflow cycle, assemblers populated both sides of the board and completed touch-up work before final cleaning. The engineered cleaning fluid advancements over the past ten years opened the process window that allowed assemblers to complete upfront process steps and only clean one time.

With the higher Pb-free reflow temperatures and time above liquidus to achieve superior Pb-free soldering, cleaning after each reflow cycle may be necessary. The lesson learned is that cleaning difficulty increases reflow temperature and time above liquidus is increased. To address this difficulty, increased cleaning temperature, cleaning time and cleaning concentration may be required.

Lesson 3: Pb-free flux technology evolves to improve wetting, reduce voiding, reduce solder balls and improve solder joint appearance.
The early introduction of Pb-free solder pastes used flux chemistries designed for eutectic Sn/Pb. With the higher reflow temperature and time above liquidus the early flux compositions decomposed during reflow. This caused popcorning, dewetting, and solder-balling. To address this issue, the solder paste formulators continuously improved the flux compositions. As flux packages improve, solder joint appearance and ease of cleaning are steadily improved as well. Figures 8 & 9 illustrate solder joint appearance from of these recent flux innovations.

I
Figure 8: Improved solder joint appearance due to recent flux innovations


Figure 9: Another example of improved solder joint appearance

The dull, grainy, and partially wetted solder filet of the typical Pb-free solder filet has been improved with new solder pastes innovations; the solder filet takes on a similar appearance to eutectic Sn/Pb solder filets. The critical question from a cleaning perspective is “Do these new paste innovations improve cleanability?” The answer is yes and no. The testing data does support improved cleanability from some solder paste innovations, but some of the new pastes are more difficult to clean. Figure 10 illustrates white residue after cleaning fine pitch leads, which indicates that some of the new Pb-free flux formulations are difficult to clean.

The lesson learned here is that the constituents in flux packages are different. The cleaning efficacy is also different for different flux compositions. Some are easier to clean, others more difficult. We also find the ease of cleaning is dependent on the cleaning fluid composition. The solvent selection within the cleaning agent and the degree of reactivity influence cleaning efficacy.


Figure 10: White residue left behind after cleaning fine pitch leads

Lesson 4: Wash process parameters increase.
The wash process parameters are influenced by the soil, the static cleaning rate and the dynamic cleaning rate2&3. All flux residues do not clean at the same rate. Minor ingredients and the type of flux used affect the static cleaning rate. With highly dense assemblies, the thickness of the solder paste printed on the board correlates to the amount of flux residue under components and adjacent to leads. To remove the flux residue, the cleaning fluid must wet and must soften the residue to allow the full impact of mechanical impingement sources.

When considering the static cleaning rate, cleaning fluid formulators focus on two critical properties: 1.) The rate at which the cleaning fluid dissolves the flux at a preset concentration and temperature; and 2.) The surface tension properties of the cleaning fluid. Lower surface tension reduces the mechanical energy required to push a fluid through a tight opening, which allows the cleaning fluid to penetrate and clean at a more rapid rate. Surface tension is an important property when cleaning highly dense assemblies and low standoff devices.

Impingement sources improve the dynamic cleaning rate. Nozzles are used in spray systems to create fluid jets that carry energy to the surface of the part. The design and layout of the nozzles is intended to optimize the mechanical driving forces. Fan nozzles provide excellent coverage and flow while coherent nozzles provide less pressure drop and directional impingement under components. Optimizing the nozzle selection and the number of manifolds improve cleaning performance.

Many data points on several Pb-free evaluations show a strong correlation between time under the spray manifolds and successful removal of all flux residues under low standoff devices. Increased cleaning time in the form of soak and impingement may be needed on some Pb-free flux residues to achieve 100% cleanliness.

Elevated wash bath temperature typically improves the static cleaning rate. The resin structure of the flux residue often softens at higher wash temperatures, improving the rate of flux removal. Increased temperature in modern cleaning fluids employing surfactant technology can lower surface tension to improve penetration under low standoff components. While elevating wash bath temperature may improve cleaning performance, there are compatibility and consumption tradeoffs.

Figures 11, 12, 13, and 14 illustrate how adjusting the wash process parameters improve cleaning performance. This problem surfaced when an electronic circuit assembly house was not able to clean a Pb-free flux residue using their current cleaning process designed to clean eutectic Sn/Pb circuit assemblies. Increasing the cleaning fluid concentration and the time in the wash section improves flux removal. The problem points to the need for either longer wash time, which perhaps correlates to longer wash sections in machines or improved cleaning fluid designs.


Figure 11: Heavy residue
Process parameters:

  1. Wash Chemistry Concentration ~ 15%
  2. Wash Temperature ~ 150ºF
  3. Belt Speed ~ 2 feet/min
  4. Impingement ~ Spray in Air using Fan nozzles

 


Figure 12: Moderate residue
Process parameters:

  1. Wash Chemistry Concentration ~ 25%
  2. Wash Temperature ~ 150ºF
  3. Belt Speed ~ 2 feet/min
  4. Impingement ~ Spray in Air using Fan nozzles


Figure 13: Trace residue
Process parameters:

  1. Wash Chemistry Concentration ~ 25%
  2. Wash Temperature ~ 150ºF
  3. Belt Speed ~ 1 feet/min
  4. Impingement ~ Spray in Air using Fan nozzles


Figure 14: No residue
Process parameters:

  1. Wash Chemistry Concentration ~ 25%
  2. Wash Temperature ~ 150ºF
  3. Belt Speed ~ .5 feet/min
  4. Impingement ~ Spray in Air using Fan nozzles

These illustrations suggest that increased time and concentration improve cleaning efficacy to the point of a residue-free part. The lesson learned is the need to determine process parameters before buying cleaning equipment. Longer wash time and concentration can be planned in the equipment design.

Lesson 5: Water soluble may no longer be water-soluble.
Solder paste formulators must consider deposition, printing, rheology, tack life, wetting, solder joint appearance and many other properties when developing a solder paste. Water soluble pastes are formulated with oxygenated materials, organic acids, and surfactants that clean with DI-water only. Higher reflow temperatures and increased activation force solder paste formulators to reconsider solvency, vehicle, acid content, and rheological additives used within the water soluble paste formulation.

Higher reflow temperatures and the dwell at and above liquidus have a tendency to drive off the solvent and vehicle within the water soluble flux formulation. When this occurs, the organic acid in the water soluble solder paste and wave flux forms a tenacious white residue that is very difficult to clean. Residue formation occurs at the leading edge of the flux residue. Additionally, increased thermal energy may harden the residue to a point where DI-water and dynamic energy no longer remove 100% of the residue. Figure 15 illustrates white residue when cleaning the water soluble flux residue with DI water in an inline spray-in-air cleaning process.


Figure 15: Tenacious white residue from water soluble flux

To resolve the residue issues when using water soluble Pb-free formulations, aqueous cleaning agents are used to wet and dissolve the flux residue. Aqueous cleaning agents provide solvency, reactivity and wetting to dissolve organic acids and tin-salts that formed when the vehicle decomposes during thermal reflow. The lesson learned is that water soluble flux materials on highly dense assemblies may require an aqueous cleaning material, at dilute concentration, to completely remove residue.

Lesson 6: Some hard residue no-clean flux compositions contain flux materials that cross-link at high temperatures.
One of the difficult challenges for solder paste formulators is the ability to wet and improve solder appearance of Pb-free alloys. As solder paste formulators innovate improved materials to address challenges such as wetting; no-clean, OA, and RMA flux compositions become more sensitive to the reflow process. Choosing between the two traditional profiles, ramp to spike and soak require careful consideration. With eutectic Sn/Pb, the soak profile aided the flux activation process, however with Pb-free the soak allows more carriers/solvents to evaporate, potentially affecting cleaning4. As the profile reaches the spike zone, the delta T and time above liquidus becomes far more critical to how, and sometimes if the residue can be successfully cleaned. These higher heat and exposure time conditions convert the flux residue into a hard clear shell.

The straight ramp profile appears to be somewhat more favorable for cleanable Pb-free post soldering residues. The gradual increase in temperature does not allow the carriers to be boiled off as quickly, yielding softer and more cleanable residue. Figure 16 illustrates a Pb-free ramp to spike reflow profile.


Figure 16: Pb-free ramp to spike reflow profile


Figure 17: Pb-free soak reflow profile

The soak profile increases the time at or above liquidus. Figure 17 illustrates the soak profile.

Some of the materials used in advanced flux formulations appear to polymerize when exposed to excessive heat. Figure 18 illustrates the ribbon around the solder legs and component. The flux residue cross-links and forms a plastic-like clear residue. Polymerization is a cross-linking reaction that forms a longer chain molecule that is no longer removable with wet chemical cleaning processes. The lesson learned suggests the importance of the reflow profile and the time above liquidus.


Figure 18: Cross linking polymerization

Lesson 7: Low standoff components may require increased cleaning time, direct impingement and directional sprays.
With advanced packaging I/O (number of bumps under the die) increasing, and tiny chip caps sitting flush mount to the board, reliability concerns move design engineers to study the beneficial properties of removing flux residue. Cleaning flux residue under advanced packages and chip caps require advanced process design considerations in the form of mechanical impingement and the cleaning fluid. Understanding the balance between the static chemical and mechanical driving forces is fundamental to predicting and optimizing process variables.

Advanced packages and chip caps sit at low clearances to the board substrate. The flux reflows and spreads under the component during reflow. The flux seals the underside of the component with a thick resinous material that is difficult to completely remove (Figure 19). Devices placed in tightly packed arrays further increase the cleaning difficulty, as there is very limited access for the cleaning fluid to reach the contaminant.

The design challenge requires both improved chemical and mechanical technology. The chemical driving forces can be improved by adding materials to increase the speed of cleaning and improve the wet-ability of the material to penetrate under low standoff components5. The mechanical driving forces require optimization in nozzle design, selection and positioning to address difficult cleaning challenges on the board (Figure 20). Even with improved chemical and mechanical forces, time is a critical factor. The time the board is exposed to the cleaning material (wash time) along with the time between reflow and cleaning (aging time) are very important variables.


Figure 19: Flux residue under a chip cap


Figure 20: Directional Sprays [Courtesy of Steolting Corporation]
 
The lesson learned points to the removal of flux residue under advanced packages and flush mounted chip caps as being very a difficult cleaning challenge. The data suggests the importance of time in the wash section and the directional spray from the wash manifolds.

Lesson 8: Batch cleaning equipment challenges.
With industry success adopting eutectic Sn/Pb low residue no-clean soldering materials, many assembly houses required limited cleaning on select boards. Batch cleaning machines were a nice fit to for low volume cleaning requirements.

Batch cleaning machines use flow, time, temperature, and advanced cleaning fluids as the critical drivers for delivering a clean part. The dishwasher style machine has become an industry standard and used by many assembly houses. Batch spray-in-air systems use higher flow nozzles that clean by flooding the part. The system limitation comes from the lack of direct impingement. This style of equipment works very well on eutectic Sn/Pb flux residues but for the harder to clean Pb-free materials, the batch spray-in-air requires new process development considerations. Figure 21 illustrates residue around the legs of a surface amount component solder with a hard residue Pb-free solder paste.


Figure 21: Batch cleaning illustration

To address this limitation of the batch spray-in-air cleaning machine design, advanced cleaning fluids are under development that improves the static cleaning rate. The challenge is to develop a cleaning fluid that improves the static cleaning rate and does attack and dull the solder joint with longer exposure time and temperature. Figure 22 illustrates the cleaning performance on selected Pb-free hard residue solder pastes with new cleaning fluids under development.


Figure 22: Batch cleaning fluid development

The lesson learned suggests the importance of process parameters and the cleaning fluid selection when cleaning hard residue Pb-free residues using batch spray-in-air cleaning machine designs.

Lesson 9: Solder joint appearance influenced by cleaning chemistry.
Engineered aqueous cleaning fluid designs use a combination of solvency, reactivity, and wetting. The reactive component used in aqueous cleaning fluids improves cleaning efficacy on many Pb-free solder flux residues. To improve cleaning on hard residues, higher reactivity is an option used by cleaning fluid formulators. High reactivity must be managed since these materials dull and etch the solder filet. Figure 23 illustrates a solder filet attacked or dulled by the cleaning chemistry. Figure 24 illustrates a solder filet that is not attacked by the cleaning chemistry while also improving the appearance of the solder filet.


Figure 23: Solder filet attacked by cleaning chemistry


Figure 24: Solder Filet of a Properly Designed Cleaning Chemistry

The lesson learned points to the importance of the cleaning fluid selection. Properly designed cleaning fluids improve cleaning while also protecting the solder filet from alkaline attack.

Lesson 10: Nitrogen atmosphere improves wetting, solder joint appearance and cleaning.
Soldering in a nitrogen-inerted environment improves wetting and reduces oxide formation of the solder flux and alloy6. The level and appearance of the flux residue is less for boards solder in a nitrogen-inerted environment. Test data conclusively correlates to the flux residues being less, transparent and easier to clean. Figure 25 illustrates the improved appearance and wetting from a board reflowed in a nitrogen­inerted environment.


Figure 25: Pb-free soldered in a nitrogen-interted atmosphere and cleaned

The lesson learned points to improved cleaning, solder joint appearance and wetting when soldering in a nitrogen-inerted environment.

Summary:
Numerous Pb-free cleaning studies and evaluations on customer applications provide process data for developing a series of lessons learned and best practices when cleaning Pb-free flux residues. The results from Pb-free cleaning studies point to flux residues as being more difficult to clean. The data from Pb-free cleaning studies indicate that most Pb-free flux residues are cleanable but up-stream process conditions and paste selection must be considered when cleaning electronic assemblies.
The lessons learned were developed from customer projects submitted to Kyzen’s Application Testing lab. The lessons point to the importance of process development before purchasing cleaning equipment. The lessons learned point to the importance of engineered cleaning fluid design. Advanced cleaning fluid designs improve the static cleaning rate, lower surface tension and protect the board surface from alkaline chemical attack. Consulting the lessons learned should assist the process engineer’s task of developing a Pb-free cleaning process.

Authors:
Mike Bixenman is the CTO of Kyzen Corporation. Kyzen is the world market leader for engineered cleaning fluids targeted at electronic assembly and advanced packaging process cleaning. Mr. Bixenman has over 16 years field experience in the design and processing of cleaning fluids within high technology environments. For questions or additional information, Mr. Bixenman may be contacted via email by clicking here .

Erik Miller is the Asia Pacific Manager for Kyzen Corporation. Mr. Miller works with Asian and Multinational organizations to assist their efforts in developing demanding cleaning processes. Mr. Miller has over 10 years experience in the design and implementation of electronic assembly and advanced packaging cleaning processes. Mr. Miller may be contacted via email by clicking here.

Fernando Rueda is the European Manager for Kyzen Corporation. Mr. Rueda is bilingual and instrumental in European market development. Mr. Rueda has over 5 years experience in the design and implementation of electronic assembly cleaning processes. Mr. Rueda may be contacted via email by clicking here.

References
1Biocca, P. (2006, May). Preventing Lead-free reflow defects and maintaining process yields.
International Conference on Lead-free Soldering.
2Bixenman, M. & Stach, S. (2004, September). Optimizing cleaning energy in batch and inline spray systems. SMTA Technical Conference. Donald Stephens Convention Center. Rosemont, IL.
3Bixenman, M. & Stach, S. (2005, September). Optimizing cleaning energy in electronic assembly spray in air systems: Phase II. SMTA Technical Conference. Donald Stephens Conference Center. Rosemont, IL.
4Ellis, D. & Bixenman, M. (2005, Feb.). Applied research for optimizing process parameter for cleaning Pb-free flux residue. IPC APEX Technical Conference. Anaheim Convention Center. Anaheim, CA.
5Bixenman, M. & Stach, S. (2006, September). Optimized static and dynamic driving forces for removing flux residue under flush mounted chip caps. SMTA Technical Conference. Donald Stephens Conference Center. Rosemont, IL.
6Bixenman, M. & Ellis, D. (2004, Feb). Lead-Free soldering: DOE study to understand its affect on electronic assembly defluxing. IPC APEX Technical Conference. Anaheim Convention Center. Anaheim, CA.

Acknowledgements
Much of the data generated for this paper was completed by Kevin Soucy, JoAnn Quitmeyer, Don McWright, Carolyn Leary and John Garvin at Kyzen’s Application and Research & Development Labs.

 

spacer
spacer spacer spacer
spacer spacer
spacer spacer spacer
spacer

Last Updated:
July 30, 2007


Copyright © 2007


Publisher: Megan Wendling
spacer spacer
spacer spacer spacer