Toxic ,use Reduction Planning in the Electronics Industry Table of Contents

The environmental impacts of alternatives to ozone depleting solvent cleaners wereexaminedin this investigation. The focus of the program was to identify alternative cleanerswithout ozone depleting potential that would meet Raytheon’s needs for the cleaning of CircuitCard Assemblies (CCA’s). Technologies evaluated were saponified aqueous cleaning systemsand semi-aqueous cleaning systems. Rinsewater samples from the two referenced technologieswere sampledand analyzed for a variety of parameters including BOD,COD, pH, oil & grease,flashpoint,and heavy metals. Analytical testing of the rinsewaters determined that in bothtechnologies the rinsewaters require treatment prior to disp xal. INTRODUCTIONIn response to increasing evidence of the depletion of the ozone layer by solventcleaning chemicalsand the large volumes of these solvents that Raytheon uses, a task force wascreated to evaluate effective alternatives to ozone depleting solvent cleaners. The initiativeis called The Alternate Cleaning Technology Committee or ACT and was formed in May 1990by the Raytheon Executive Officeof Manufacturing and Environmental Quality. Thecommittee consists of representativesfrom 15 Raytheon facilities, thus forming a company--wide, multi-divisional initiative with thirty-five members. The objectiveof the ACT Initiativeis to eliminate ozone depleting solvents, Freon 113, and 1,1,1 Trichloroethane for the cleaningof Circuit Card Assemblies (CCA’s) by the end of calendar year 1992. Sixmain requirements were established by the ACT Committee and were used todetermine what alternate cleaners would be viable for considerationin the Act Initiative.The six requirements,aie listed below: 0 Must not degrade reliability0 Non ozone depleting0 Environmentally acceptable0 An effective cleaner0 Safe alternative0 Currently available Using thislist of requirements, the ACT group conducted an equipment and chemistry surveythroughout the industry to determine what types of alternative cleaners would be evaluated.The next genurt ionof solvent cleaners, or HCFC‘s, were ruled out because of their uncertainregulatoryfuture and ozone depleting potential. The alternate cleaner could not degrade the reliability of our sophisticated militaryelectronics product line. High levels of cleanliness are required to ensure long term reliability. The alternate cleaner had to be environmentally acceptable which meant we did notwant to solve one environmental problem and create another. Using that reasoning, rinsewaterswere sampled during all tests of alternate cleaners to determine the environmental impact thoserinsewaters would haveon our waste treatment facilities and the environment. The alternate cleaners had to be an effective cleaner. Here at Raytheon we work under a variety of military specifications and high levels Of cleanliness are mandatory in order toensure consistent high quality and reliability. The alternate cleaners had to be safe for our workers. Corporate Health and Safety wasa memberof the ACT Committee to evaluate the cleaning chemistries that were identified.These involved the evaluation of potential health impacts of compounds such as glycol ethersand some of the new citrus based compounds in the terpene family. Also considered were thepotential flammability issues of the alternatives such as the terpene based cleaners that arecombustible liquids. Fire detection and suppression systems were fully evaluated. The alternate cleaner obviously had to be currently availableon the marketplace insufficient quantities to meet Raytheon’s near-term manufacturing and cleaning needs. Once the cleaning equipment and cleaning chemistries had been selected by the ACTCommittee for consideration, the next phase involved the development ofa detailed cleaningprocess specification which would be used for the evaluation of all the cleaners andchemistries. Included in this evaluation was an environmental test plan and health and safetyevaluation of the combinations of cleaner and chemistries. The goal of the ACT Initiative was to identify an effective alternate cleaning procedurethat will meet the near and long-term needs of Raytheonin the cleaning of CCA’s to militaryspecifications. A solution was needed both for in-line and batch type applications. Based onthe findings in the ACT Initiative for CCA cleaning, i t is expected that these findings will helpto accelerate the phase-out of ozone depleting solvent cleaners in other cleaning applicationssuch as machining, bench top, etc. EXPERIMENTAL PROCEDUREThe first step of the ACT Initiative was to identify the cleaning equipmentmanufacturers that would be evaluated in the Phase I program. A total of ten cleaningequipment vendors were identified for the PhaseI evaluation. The semi-aqueous cleaningchemistries that were included in the evaluation were the terpene compound EC-7R which ismade by Petroferm, Inc., and Axarel-38 which is manufactured by Dupont. The saponifiedaqueous cleaners were made by Kester, Federated Fry, and Alpha Metals. The second task of the ACT Initiative was to design and fabricate a test board thatwould represent the Raytheon product line and provide a viable testing vehicle for thealternate cleaners. The final board design wasa mix of plated through hole (PTH)and surfacemount technology(SMT), measuring four inches by seven inches in size and consisting of twodif ferent substrates, EpoxyE10 and Polyimide. The ne%! task was to develop a cleaning process specification which would provide aformidable c l d n g challengefor the alternate cleaners. The challenge would be significantlygreater than m y that would be encountered during a production process, which would allowthe alternate cleaners to be ranked based on levels of cleanliness. The EPA/IPC/DOD PhaseI1 testing protocol was integrated into the specification for reasons of comparison and customeracceptance. The cleaning baseline against which the alternate cleaner would be referenced was anexisting cleaning process which used l ,I*l trichloroethane, followed by a deionized water rinse(2 megohm) and completed with an isopropyl alcohol rinse. The cleaning process specification involved hand soldering of the SMT devices on-siteat the cleaning vendor, followed by a one minute immersion in Kester 185 RMA flux, followed by wave soldering with preheat topside temperature of 220 degrees Fahrenheit and a wavesolder tempcrature of 500 degrees Fahrenheit. The boards were then allowed to dwell one hour after after which the!: were processedthrough the cleaning equipment. Thirty boards of each substrate, G-10 a.ld Polyimide, wereprocessed fo r each cleaning evaluation. Following the cleaning process, 15 boards were visually examined for flux residue andthe remaining boards were packaged and returned to Raytheon laboratories for surfaceinsulation resistance (SIR) testing. Each board contained 5 SIR patterns for testing purposes.Upon completion of the SIR testing, a total of 840 components were cut, lifted and the boardunder the part was visually inspected for residual rosin contamination. Ionic residue wasdetermined with rhe Omegameter and organic residues were identif ied using High PressureLiquid Chromatography (HPLC). A Phase I environmental test plan was developed to identify parameters that would be testedin the rinsewaters during the Phase I ACT testing program at the vendor sites. Theenvironmental plan called for taking a representative composite sample of the rinsewatersduring cleaning, and testing the rinsewater for the presence of five heavy metals, lead, tin,copper, nickel, and-zinc. In addition, rinsewa ter samples were tested for Biochemical OxygenDemand (BOD), Chemical Oxygen Demand (COD), total suspended solids, pH, flashpoint, andoil and grease. The samples were taken at the vendor sites throughout the country and expressmailed back in coolers to an analytical laboratory in Massachusetts for testing. The-environmental objective of sampling the rinsewaters was two-fold; f irst was to evaluate thequantity and quality of the rinsewaters that would be generated from the cleaning equipmentand second to evaluate those rinsewater streams for potential closed-loop processing duringinstallation in the Raytheon facilities. The wide variety of wastewater discharge outlets atRaytheon facilities nationwide necessitated the evaluation ofa number of treatmenttechnologies for effective treatment of the rinsewaters. Another key element of environmentalinformation that needed to be gathered during the ACT testing protocol was to performdetailed material balances of all the cleaners used in the alternate cleaning methods anddetermine their fa te in the form of fugitive air emissions, stackair emissions, wastewaterdischarges, or off-site shipment as hazardous waste. RESULTS AND DISCUSSIONOf the multiple combinations of cleaning equipment and cleaner chemistries evaluatedin the Phase I ACT Initiative, three of the processes using semi-aqueous cleaning solvents wereequal to or better than the 1,l.l Trichloroethane, DI water, IPA rinse baseline. In addition, onesaponified aqueous in-line system cleaned effectively in the aqueous mode. The ElectronicControls Design (ECD) 6307/6300 cleaning equipment and Accel's Microcel I1 are viable forbatch semi-aqueous cleaning processes. These two machines were used in combination with EC-7R citrus b u a d cleaner. The ECD system utilizes two dishwasher style units and a n auxiliarycirculating oven for drying. The Accel machine incorporates a technology which spins the partto be cleaned about its center of gravity and utilizes centrifical force to remove material frombeneath components. This cleaning process is known as spin under immersion. The Detrex Model SA-20 and Hollis Automation Hydro-Station 332 are viable for in-lineprocesses. The Detrex unit can be utilized either in the aqueous or semi-aqueous mode,however, the machine was evaluated in the semi-aqueous configuration only using Axarel-38.To date, our evaluation has not tested the Detrex machine in the saponified aqueous mode. TheHollis Automation Hydro-Station332 was used in conjunction with Federated Fry 3555 in thesaponified aqueous cleaning mode. The degree of cleaning was determined to be equipment dependent. This was moreevident for saponified aqueous cleaning systems than semi-aqueous systems. Although, all theaqueous chemistries were similar, only the Hollis Automation unit could clean well enough to fulfill the test requirements. The environmental findings determined that the rinsewaters from both saponifiedaqueous and semi-aqueous cleaning processes require treatment prior to discharge. Our resultsindicated that the BOD,COD, and oil and grease limits of the semi-aqueous cleaners inparticular, arc well above the allowable discharge limits in most parts of the country. The in-line rinsewater discharge quantities averaged 23 gallons per minute for running DI waterrinse, while the batch semi-aqueous cleaning processes involved the discharge of between 3 and15 gallons per batch. The in-line saponified aqueous system which averaged approximately 4 -5 gallons per minute had a BOD, COD content of 9,295 milligrams per liter (mg/l) and 8,830milligrams per liter (mg/l) respectively, with an oil and grease content of 419 milligrams perliter (mg/l). Generally, acceptable discharge values for BOD average 250 mg/l, with CODseldom listed as a discharge parameter. Oil and grease acceptable values range between 100 and150 mg/l. In addition, the pH of the rinsewater from the aqueous treatment system was 10.9which is slightly above allowable limits of 5.5 9.5 pH. There is a significant difference between the handling of the waste wash sumps foraqueous and semi-aqueous systems. The semi-aqueous cleaning baths are reported to be veryresilient and able to assimilate large quantities of rosin based fluxes prior to requiring disposal.A steady state equilibrium is apparently achievable, where drag-out of flux residues equals thevolume of flux drag-in. Thus only periodically, approximately once or twice a year, the semi-aqueous cleaning bath is discarded and sent off-siteas a hazardous waste fuel supplement. Inthe case of EC-7R, the purchase price of the material includes the cost of disposal as ahazardous waste. Ultimately, basedon volumes generated, the EC-7R may be recycled by ahazardous waste vendor. In the case of the saponified aqueous cleaning sumps, these sumps are generallydischarged every eight to sixteen hours of operation and generally average aboutlOQ gallonsinsize. These wash tank dumps from the saponified aqueous systems contain the majority ofheavy metals, flux residues, and extremely high pH and BOD/COD values. The need forfrequent batch dumpson the saponified aqueous cleaning systems results in the use of highquantities of heated water with saponifier and also a large requirement to treat these washtank dumps prior to discharge. Historically, the reason for short bath life has been that thesurfactants, which are volatile, evaporate over time which lowers cleaning effectiveness. Based on a matrix evaluation of semi-aqueous versus saponified aqueous cleaning, whichincluded health B safety impacts, environmental impacts, cleaning effectiveness and operatingcosts, the decision was made to pursue the useof the terpene cleaner EC-7R in the Phase I1Pilot Facility. The Phase II Pilot Facility, which includes 2 cleaners and 2 closed-loopprocessors, will be installed in a Raytheon Massachusetts facility for cleaning tests on Raytheonproducts. Although the Hollis machine cleaned to an acceptable level, i t was not incorporatedin the pilot facility due to the presence of glycol ethers in the saponifier, projected higherannual operating costs, as a result of frequent wash tank dumps, and a lower potential to allowclosed-loop rinsewater processing. In the area of closed-loop rinsewater processing, the semi-aqueous systems have beendesigned to readily separate the cleaner from water thus providing an advantage oversaponified aqueous systems. Due to the accelerated nature of Raytheon's Alternate CleaningTechnology Program, the decision to move forward witha semi-aqueous cleaner was made inDecember of 1990. At that time, the technology to close-loop the Axarel-38 product had not yetbeen defined by Dupont.On the other hand, the EC-7R cleaning compound had been re-formulated to enhance its separability from water thus providing an apparent advantage forclosed-loop rinsewater processing. The ACT Phase I1 Pilot Test Facility which was installed in May of 1991, in one ofRaytheon's Massachusetts facilities, included an ECD batch system and a Detrex SA-20 in-linecleaning system, bothof which will use the EC-7R cleaning chemistry. To evaluate the 7131n13a3313033rJaJJ-feasibility of closed-loop rinsewater processing, Separatioa Technolc ;ists and Simon-WTSsystems were installed in the Pilot Test Facilityto evaluate their capabi! Y in close-looping therinsewater streams from both the batch and in-line cleaning units. The i\ 3wledge gained fromevaluating the closed-loop rinsewater processors will also be used t : dentif y appropriatetechnology that can then be used to polish rinsewaters prior to dischar This Pretreatmentwill involve removing residuals terpene to low enough concentratic to be suitable fordischarge to an on-site industrial waste treatment Plant or a municipal ste treatment plant. Both closed-loop processors use phase separation by specific grav j,granular activatedcarbon for trace terpene removal, ion-exchange, particulate filtration a c 3 a heating boost backto rinsewater temperatureof approximately 100 degrees Fahrenheit. The quality of deionized water that is requiredfor rinsing of Raytheon CCA's to meetmilitary specifications, ranges between 0.5 and 2.0 megohm. Therefore, the closed-loopprocessors need to supply deionized water ona continuous basis within that acceptability range. During the pilot test running of the closed-loop equipment, four significant items ofinformation are needed: 1.Using the present technologies, costs are a function of the amount of consumable-exchange media (Granular Activated Carbon and Ion-Exchange Resins) that will beusedon an annual basis.The life expectancy of the exchange medias is directly proportional to the efficiencyof the phase separator unitsand inversely proportional to the temperature of therinsewaters. Initially, rinsewater temperatures will be 120 degrees Fahrenheit, howeveraneffort will be made to lower the temperature to 80 degrees Fahrenheit, which will loweroperating costs, reduce terpene air emissions and extend exchange media life.Definition of Annua 1 ODeration & Maintenance Costs istics of Closed-LooD Rinsewater Processiqge .2.A hydraulic balance needs to be attained between the cleaners and the closed-loopprocessor. The feasibility of this balance will be determined during the pilot test. Due to the lack of a large installed base of closed-loop processors on RMA FluxSvstem Cleaners, analytical workare accumulating.will be doneon the treated water to determine if any contaminants In addition, the system will be monitored closely for any signs of microbial growth thatnecessary, a disinfection/sterilization step may need to be added to themay develop. Ifsystem. T r e a t m t Effrciencvof the Phase SeDa ratora4.Depending on the type and quantity of cleaner installations and the facility's capabilityto discharge wastewater, i t may be most practical to treat and discharge rinsewaters. Anexample would be the installation of 2 batch units, which would not justify the need toclose-loop rinsewaters.For that application, a reasonable treatment scheme may be to phaseseparate the rinsewater and then polish with granular activated carbcil prior to discharge.Evaluating the treatment efficiency of the phase separators and granular activated carbonsteps will determine the quality of effluent that is achievable.. . SUMMARYResults from the Raytheon ACT Initiative Phase 1 indicated that the following 4-processes, 3 semi-aqueous and one saponified aqueous, cleanedtoor better than the TCA/DI water/IPA baseline: