Cleaning up after Pb-free: lacking field data for no-clean Pb-free solders and confidence in accelerated aging tests? Effective cleaning is the only way to guarantee an assembly won't fail.
At first, with a blast of something ranging between open skepticism and sheer hilarity, no doubt. Such a reaction would be perfectly understandable, given an economic climate that, over the past few years, has focused electronics manufacturers' efforts squarely on taking cost out, and on removing all processes outside the company's core activities.
But while the cleaning process is considered by many an unnecessary luxury in our no-clean world, it could do wonders for many a company's quality and product reliability, not to mention profits.
If process time for cleaning (30 min. maximum) is compared to reworking or producing a new assembly to replace a faulty one, cleaning will always be the quicker option. As the cleaning process is rigorous and, once set up, low maintenance, any resource freed by not having to rework or build new assemblies can be, in part, used to monitor cleaning. As the model calculation in Figure 1 shows, it is possible to improve yields, throughput and quality. The economics in the model will not work for all companies and market sectors, of course, but the essential question of justifying the cost for a semiaqueous batch cleaning process is: Are the total costs associated with establishing a cleaning process, depreciated over three years or more, less than the total costs in scrapped or returned units from service? If yes, then profits and product quality could be improved by investing in a batch process.
Every bare board, component and assembly undergoes numerous handling and production processes and each leaves a variety of chemical residues. These can be ionic or nonionic and may become reactive when the board is live in a humid or aggressive environment. This can cause dendritic growth, electromigration, corrosion, or coating delamination or breakdown.
The assembly residues which can cause coating and reliability issues include flux (correctly or incorrectly processed), metallic salts, mold release agents on components, hand contamination such as oily and salty fingerprints and hand creams, food residues, fungus, mold or bacteria, halides and certain hygroscopic glycols. If the board is coated, any of these residues can be the cause of aesthetic and functional conformal coating defects such as:
* Wet state: Poor coverage, dewetting, pinholes.
* Dry state: Craters, fish eyes, blisters, loss of adhesion.
Some residues can also inhibit the curing of some coating chemistries. Conformally coated, uncleaned assemblies have the potential to cover, but not render benign, chemical species which subsequently cause circuit failure or reduced efficiency. If circuits are cleaned prior to conformal coating, all uncertainties regarding reliability caused by residues are eliminated; this is also true of uncoated circuits. Residues can also reduce the efficiency and performance of solder pastes or fluxes, which in turn can lead to poor soldering or poor joint reliability.
Whether the assembly is conformally coated or sealed in a hermetic housing, maximum long-term reliability is achieved, particularly in aggressive environments. Misprinted boards are cleaned of solder paste or adhesives. Stencils are cleaned of solder pastes or SMT adhesives to guarantee precision printing and to eliminate cross-contamination. Final reflowed assemblies are cleaned and ionic cleanliness is tested.
By cleaning, manufacturers ensure optimized production yields, throughput and quality, and also guarantee reliability and longevity. This also works for manufacturers building Class 1 and 2 products.
Manufacturers that are unsure whether they should be cleaning can look at a few simple indicators. Any of these occurances may flag the need for further investigation:
* Conformal coating defects in the wet or dry state.
* Circuit failures during functional/ATE tests.
* Electrical failures, reduced SIR, corrosion, delamination or blistering of the conformal coating following environmental conditioning.
* Soldering defects.
A no-clean process requires validation. This would entail, for example, using test vehicles populated with appropriate dummy components and materials used in the process; using surface insulation resistance test methods; and being subjected to appropriate environmental conditioning (e.g., 65[degrees]C/95%RH, 85[degrees]C/85%RH, salt spray, corrosive gases). This technique can yield a high level of confidence in the process window, and ensure any residues are benign. However, once the process is full scale and commissioned, ask the following questions:
* Is monitoring and control of the soldering process and handling of boards strictly controlled?
* Are all process variables measured and recorded for every board?
* Are on-board thermal profiles and flux quantities available for every board?
* Do any primary measured data categorically prove every board has benign residues on its entire surface?
These questions become more pressing as component population density increases and finer pitches are used. Analysis to identify and quantify the chemical species on a board surface, particularly ionic ones, can be carried out with specialized techniques such as ion chromatography. But these techniques are not standard manufacturing quality assurance equipment.
Indeed, no economical instruments currently lend themselves for use as a QA rapid test to measure all uncleaned residues--visible, or not, to the naked eye. If ionic cleanliness measurements are carried out, the readings are true only if all residues are removed in the IPA/water blend used in the machine. Most no-clean paste or flux residues are not extractable (cleaned) with this solvent blend. Effective cleaning is the only way to guarantee an assembly will not fail in service due to circuit board residues.
Using Pb-free materials means reevaluation and retesting for reliability. Pb-free soldering introduces a flux chemistry that uses more powerful and aggressive acids to cope with higher processing temperatures while maintaining good solderability.
Even if materials classified as no-clean are incorrectly applied and processed, the residues left on the board can cause downstream processing issues, particularly during conformal coating, and can be extremely destructive to the assembly in service. Solder paste residues for Pb-free circuits are chemically different than Pb-based ones, undergo higher processing temperatures, polymerize to a greater degree and may not respond to cleaning using older types of cleaning products.
This introduces several challenges:
* Companies currently cleaning circuits need to reevaluate and revalidate existing cleaning processes, taking into account the new soldering materials. They will likely discover a new cleaning chemistry is needed to achieve desired results. Re-evaluation provides the opportunity to evaluate operator--and environmentally friendly solutions for semiaqueous processes.
* Manufacturers of high-reliability products that have to change to Pb-free require long-term solder joint reliability and may evaluate solder pastes for physical, chemical, thermal and electrical properties, versus ease of cleanliness.
* No-clean manufacturers may have to introduce a cleaning process to ensure optimum product reliability.
Field performance data over many years for no-clean Pb-free assemblies are lacking. Simulated accelerated aging tests leave less than 100% confidence for optimum reliability. Therefore, effective cleaning is the only way to guarantee an assembly will not fail in service due to circuit board residues.
Cleaning processes are generally batch or inline, with a range of equipment available to suit budgets, throughputs, chemistries and liquid agitation mechanism. Batch equipment can be classified as multi-stage tanks employing ultrasonic generators or spray, and under-immersion systems with or without automated board basket handling systems. There are also spray-in-air machines. Inline cleaning machines employ high-pressure liquid spray delivery onto a continuous conveyorized belt. In these processes, residue cleaning is achieved through a combination of thermal, mechanical and chemical energies, with each process choice having a different balance of these energies.
Cleaning processes can be generically classified as follows:
* Pure water (aqueous).
* Semiaqueous (ready to use or concentrate, with and without saponifiers).
* Nonflammable solvent used in liquid state.
* Nonflammable solvent in vapor state.
* Flammable solvents.
The selection of a suitable process may seem difficult, with so many suppliers of equipment and cleaning media. Start by considering existing and imminent legislation on chemicals and the environment, including Directive 2004/42/CE, Limitation of Emissions of Volatile Organic Compounds Due to the Use of Organic Solvents (reduction of VOCs to atmosphere). Other legislation to consider: the RoHS directive, REACH legislation and the WEEE directive.
Certain semiaqueous products for batch or inline processes that contain low or no VOCs are classified as nonhazardous, nonflammable, low odor and capable of use in all types of equipment design. They work on a combination of solvency, detergency and chemical reaction, and can remove all types of contamination whether ionic, non-ionic, organic or inorganic. Because they contain water and other harmless organic matter, they can dissolve water-soluble and organic-based residues. However, pure water- or solvent-based processes favor dissolution of either inorganic or organic residues, not both.
Semiaqueous processes are more economical than solvent-based processes, because solvent costs range from twice to 30 times more per liter. Equipment is often double or triple the cost of equivalent semiaqueous equipment, and can be even greater if specialist, completely non-emissive machines are specified.
Leigh Jansen is development manager, Humiseal Europe (Chase Specialty Coatings division) (concoat.co.uk); firstname.lastname@example.org.
Production Example Cost/board $50 Sale value/board $200 Volumes: 5,000 boards/year Total production costs over 3 years $750,000 Revenue generated over 3 years $3,000,000 Profit generated over 3 years $2,250,000 No-Clean Products returned under warranty due 5%. to contaminant failure: Cost of returns over 3 years $150,000 + $10,000 service costs Subtotal $160,000 Reviewed profit $2,250,000-$160,000 Profit over 3 years $2,090,000 Cleaning Capital equipment cost $25,000 Chemical costs $18,000/year Total cleaning process costs over 3 $304,000 years Cleaning cost/board $2.02 Total profit (no board returns for $2,250,000-$204,000 contaminants and no service costs) Subtotal $2,230,000 Additional profit through investment $140,000 in cleaning Additional profit 6.7% FIGURE 1: Economics of cleaning vs. no-clean over three years. Table 1. Overview of Possible Combinations of Generic Equipment and Material Technologies Water or Nonflammable Self-rinsing Solvent Nonflammable Evaporative Water Semiaqueous Solvent Drying Ultrasonics (specific Yes Yes Yes Yes frequencies) Spray under immersion Yes Yes Yes No Spray in air Yes Yes Yes No Inline Yes Yes Yes No
|Printer friendly Cite/link Email Feedback|
|Date:||Mar 1, 2006|
|Previous Article:||Will your product survive in the field? HALT and HASS testing are not just for aerospace electronics anymore.|
|Next Article:||Concurrent manufacturing in an EMS environment: changes required and precautions to take for successful implementation of Pb-free manufacturing.|