-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueComing to Terms With AI
In this issue, we examine the profound effect artificial intelligence and machine learning are having on manufacturing and business processes. We follow technology, innovation, and money as automation becomes the new key indicator of growth in our industry.
Box Build
One trend is to add box build and final assembly to your product offering. In this issue, we explore the opportunities and risks of adding system assembly to your service portfolio.
IPC 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.
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Ionic Cleanliness Testing Research of PWBs for Purposes of Process Control, Part I
August 24, 2010 |Estimated reading time: 23 minutes
AbstractIonic cleanliness testing machines are designed to determine the total ionic content extractable from the printed wiring board (PWB) for purposes of process control. The conductivity of the extract solution is measured and the results are expressed as sodium chloride equivalence per unit area. The problem with this method is two-fold: 1) Many of today's low residue flux and lead-free flux residues are not soluble in the extract solution and 2) contamination of concern is with site specific components, from which contamination does not correlate to the area of concern. The purpose of this study is to research low residue and lead-free flux structures, identify solvent compositions that will dissolve these residue types, and offer options for performing both bulk and site specific ionic cleanliness testing methods. IntroductionAlthough the detection method may differ, all non-destructive cleanliness tests today require the remaining residue to be dissolved to measure it [1]. This is true for bulk or site specific testing. This being said, the solubility of the residue in the test solution becomes very important. The approach selected by the researchers in this study for determining the solubility, or insolubility, of current flux types focuses determining the Hansen Solubility Parameters (HSPs) for the current flux types. Once the solubility parameters for the flux types are established, a suitable solvent can be selected based upon known HSPs to fully dissolve the remaining residues for analysis. The solubility data generated in this study will be valuable in developing new and more accurate cleanliness test methods for process control and reliability prediction. Why is Cleanliness Testing Important?Many a company has learned the hard way that product reliability is directly related to the ionic cleanliness of a circuit board. The most tragic example of a suspected circuit board cleanliness failure occurred on January 27, 1967 [2] when a circuit assembly in an Apollo command module developed a short circuit in an on-board control system. The resulting fire consumed the entire module and all three astronauts on-board. There are doubtlessly many other examples of failed brake systems, pacemakers, weapons systems, and navigation satellites and other electronic systems we depend on for our everyday lives and well being. This being a well-established fact, much time and capital is spent every day on assembly lines worldwide to assure ionic residues are controlled. Assessing Cleanliness Today One would like to think an Apollo type failure could be avoided today by following the current IPC Electronic Industry Standards. In fact, just following the current standards can leave electronic systems vulnerable to a host of failure modes related to circuit residues. Non-destructive visual inspection often misses residues under capacitors, resistors and SMT array components. These "spot" residues are difficult analyze in a non-destructive way. Newer methods have been developed using steam or heated IPA/DI water extraction and measured by ion chromatography or electromigration methods both of which require a trained technician or chemist and a considerable amount of time and effort. Even though these newer analytical tests have the ability to detect these hidden residues, a simpler, easier to perform test has held the top position as the most used cleanliness test performed to assess the cleanliness of electronic circuits for many years. ROSE Definition The resistivity of solvent extract (ROSE) test has been the industry standard production line test for measuring the cleanliness of circuit assemblies for 37 years [3]. Originally established as a pre-conformal coat cleanliness requirement for military electronic hardware soldered with RMA and RA fluxes [4], the test was performed by flowing a 75%/25% mixture of isopropyl alcohol and deionized water across the surface of the circuit and measuring the drop in resistivity of the mixture. The ROSE test was incorporated into the military specifications in the 1970s as a requirement for building military hardware. In the late 1980s and early 1990s, the military adopted industry standards developed jointly through the IPC. This kept the ROSE test as the daily standard for class 3 (high-reliability electronics). This remains as the current standard for conducting daily production tests. New Fluxes and New Designs This all worked quite well for some time. Then things changed and there were new fluxes and the new designs. With the introduction of surface mount, spaces became tighter. Chip cap 1206s became 0201s. Then there were the new fluxes; the water soluble fluxes in the 1970s, the no-cleans of the 1990s and today's lead-free fluxes. These fluxes contain many new resins that are not necessarily soluble in the 75/25 IPA water mixture. Some have tried heating the mixture to improve solubility with limited success. Exacerbating the problem, the continual miniaturization of electronic packages is creating further problems dissolving residues in smaller and smaller spaces.Miniaturization: Perhaps miniaturization is the most obvious trend of electronic industry. With Moore's law serving as the engine, electronic products are more versatile, running faster, and rapidly shrinking from huge floor model to tiny pocket size.For state-of-the-art board level assembly, the mechanical hole size and pitch is 0.125 mm and 0.3 mm, respectively, while the non-mechanical via diameter and pitch is 0.05 mm and 0.175 mm, respectively. For components, among the finest are CSP with 0.3 mm pitch and 01005 [9]. For portable devices, the leading edge package standoff is 0.3 mm, and the pad diameter is 0.2 mm. In the mean time, the cost of board assembly conversion has been driven down to 0.2 cents per I/O, with a board assembly escape rate of 200 ppm [10]. Miniaturization is driving multiple changes in flux technology, which ultimately influence the ability to measure in real time the cleanliness of production circuit boards. Flux Consistency: As a result of miniaturization, the solder powder used in solder paste is shifting from type 3 (25 to 45 microns) to type 4 (20 to 37 microns), and the flux employed is more homogeneous and more thixotropic to achieve satisfactory consistency in printing and soldering. Also because of miniaturization, the assembly process is more vulnerable toward bridging; therefore the solder paste needs to be more slump resistant.Oxide: In general, the volume of soldering materials, including fluxes and solder reduces in proportion with decreasing pitch. However, when the solder materials are shrunk in proportion to the pitch, the thickness of metal oxide does not shrink in proportion, as shown in Figure 1. The metals here refer to PCB pads, component leads and solder powder. Consequently, the amount of oxide to be removed by unit volume of flux increases with decreasing pad dimension. To compensate for this increasing work load, the fluxing capacity per unit amount of flux needs to be increased.Figure 1: Relation between pad size and oxide thickness and oxygen penetration path.Oxygen Penetration Path: While the pitch and pad size decreases, the oxygen penetration path through flux or solder paste also decreases, as also shown in Figure 1. This inevitably results in a more rapid oxidation of both flux materials and metals covered by the flux if soldered under air. Hence, a flux with greater oxidation resistance as well as a greater oxygen barrier capability is needed for finer pitch applications. Symptoms such as graping (see Figure 2) and head-in-pillow become common if the flux is not upgraded in terms of those properties. In Figure 2, the reflow was adequate when the dot size was large. However, graping occurred when the dot size was small.Figure 2: Example of graping phenomenon observed for small dot sizes.Flux Burn Off: In general, the vaporization rate of liquid increases with increasing exposure area per unit volume. For the same token, the flux burn-off increases with decreasing flux quantity deposited [11]. To offset this unfavorable trend, the flux employed for finer pitch needs to be more nonvolatile, hence more resistant to flux burn-off.Wetting Speed: Although good wetting is a desired feature at soldering--and fast wetting is also desired at wave soldering or hand soldering--fast wetting is actually causing problems during reflow soldering. At SMT assembly, defects due to unbalanced wetting force, such as tombstoning or swimming, increases with decreasing component size, partly due to increasing sensitivity toward misregistration. Under this situation, fluxes with a slower wetting speed would allow more time for the wetting force to be balanced, therefore promise a lower defect rate [12]. Spattering: Miniaturization brings the solder joints closer to the gold fingers, hence is more vulnerable toward solder spattering. Spattering can be caused by moisture pickup of the solder paste. It can also be caused by fast solder coalescence action. To minimize solder spattering, fluxes with low moisture pickup and slow wetting speed should be employed [13].Soldering Under Air: As stated earlier, the cost of board assembly conversion has been driven down to 0.2 cents per I/O. Low cost driver is pushing the industry hard to have soldering done under air. Although miniaturization is causing solder fluxing more difficult due to more oxides and easier oxygen penetration, the assembly manufacturing houses are reluctant to use nitrogen and are pushing the challenge to solder and flux suppliers. As a result, the flux desired should exhibit the following properties: 1) A great oxidation resistance to prevent the flux from being oxidized and (2) a great oxygen barrier capability to protect parts and solder from being oxidized.Lead Free: The environmental consideration has driven electronic industry toward lead-free soldering operation, with July 1, 2006 being the milestone date for European conversion dictated by RoHS. This European environmental legislation actually initiated a global lead-free conversion. As of today, about 60% of the global solder paste shipments were lead-free already.Although some other alloys, such as eutectic BiSn, eutectic SnZn or their modification, are also in use, the main stream lead-free solder alloys adopted by electronic industry include Sn-Ag-Cu (SAC), Sn-Ag (SA), Sn-Cu (SC), and modification of those alloys [14]. As to PCB surface finish, the prevailing choices include OSP, HASL, immersion Ag, ENIG and immersion Sn.High Temperature: The main stream lead-free solders exhibit a melting temperature ranging from 217 to 227°C. Thus, the soldering temperature typically is 20-40°C higher than that of eutectic SnPb. Use of a higher soldering temperature generally causes more flux thermal decomposition, more flux burn-off, and more oxidation of fluxes and metals. To avoid problems caused by the higher soldering temperature, fluxes with the following features are desired: 1) a higher thermal stability, 2) a higher resistance against burn-off, 3) a higher oxidation resistance and 4) a higher oxygen barrier capability.Poor Wetting: The surface tension of lead-free alloys (0.55-0.57N/m for SAC) is about 20% higher than Sn63Pb37 (0.51N/m) [15]. This consequently results in a poorer spreading and wetting for lead-free alloys [13]. This deficiency in alloy wetting is mainly compensated with new development of fluxes with a better wetting. In general, lead-free fluxes with adequate wetting exhibit the following features: (1) a lower surface tension facilitating a better solder spread [4], and (2) a higher flux capacity and/or higher flux strength.Large Dendrite: Due to the high tin content, lead-free alloys typically display a joint with large dendrite formation, as exemplified by Figure 3 [16]. Although large dendrite not necessarily causes early failure [14], combining with anisotropic nature of tin crystal lattice and unfavorable grain orientation does pose reliability concern [17]. This reliability concern can be alleviated by forming solder joint with refined grain structure. Besides employing alloys with grain refining additives [18], use of proper flux chemistry can also achieve a similar result [16].Figure 3: Surface dendrite structure of Sn95.5Ag3.5AgCu1.0 solder joint [9].
Miniaturization and lead-free soldering increase the flux innovation demands. Changes in flux compositions and soldering profiles increase ionic cleanliness complexities. To understand the magnitude of these complexities, nine flux residues were evaluated in this study. Two of the fluxes were organic acid based (water-soluble), one designed for tin-lead and the other for lead-free. One flux was rosin. Six fluxes were no-clean, four designed for tin-lead and two designed for lead-free.
ROSE Problems Problems with the ROSE test [5] exist. As mentioned previously, the test method relies on dissolving the flux residue to measure the effect on the resistivity. The electronic assembly drivers of miniaturization and the new flux technologies needed to achieve high yields further complicates this issue. If the residue is not dissolved, then the remaining residue is not detected by the ROSE method. This limitation leaves companies producing high reliability military and medical hardware with significant exposure. The main problem with the ROSE test today is that it is limited to isopropyl alcohol and water, and under current processing conditions, these solvents do not adequately dissolve most of today's fluxes trapped under SMT components.HypothesesH1: The IPA 75%/H2O 25% will not adequately dissolve many of today's flux technologies.
H2: A new test solvent that dissolves all flux technologies (water soluble, no-clean, rosin and lead-free) is needed.Predicting Solvent ActionScientists' postulate that the solvency behavior of a pure solvent is proportional to the cohesive energy of the solvent and that this energy is proportional to heat of vaporization of the solvent and calculated from the equation below. The cohesive energy density is the amount of energy needed to completely dissolve one unit volume of molecules from their like molecules, which is equal to the heat of vaporization divided by molar volume [21].Where:
ΔH = heat of vaporizationR = gas constantT = temperatureVm= the molar volume of the solventHildebrand Solubility Parameters In 1936, the Hildebrand solubility parameter was introduced by Joel Hildebrand. He proposed that the solvent's behavior to affect solids could be predicted by looking at the square root of the cohesive energy of that solvent. For a material to dissolve, in this case flux residue, the cohesive energy of the molecules must be overcome by molecules that have similar solvency behavior [21]. This parameter (δ) can be calculated as shown in equation 2 below.Equation 2: Hildebrand Solubility Parameter.
The units of Hildebrand solubility were originally expressed as (δ/cal½ cm-3/2). With the advent of the metric system the units are now (δSI/MPa½). For conversion, one δSI equals roughly two δ (2.0455 to be exact).The primary use of the Hildebrand solubility parameter was to predict the affect of solvents on materials. From selecting solvents to strip paints, to removing machining oils, this approach shortened and improved the formulation process. Furthermore, this approach could be used to avoid damaging certain materials of construction, such as epoxy fiberglass under the paint.The Cleaning Universe A good analogy is to think of all the thousands of solvents as points in 3 dimensional spaces, like stars in the night sky. The stars in this universe are arranged such that the stars closest to one another have similar solubility properties. Now imagine we can plot the residues to be removed and the materials of construction the widgets are made from in the same space such that if the residue or a widget material close to a solvent point we could predict the dissolution or deterioration of the material. This would be quite useful in both selecting a solvent to clean the residue of interest and avoid solvent selections that would deteriorate the materials of construction.Along Comes Hansen
This is precisely the concept introduced by Dr. Charles Hansen in 1966. He proposed the Hildebrand solubility parameter could be broken into three parts. The Hildebrand approach is much like looking into the night sky and seeing two stars that appear to be close. They could be close, or one could be light years behind the other. Dr. Hansen added three parameters to establish true location in this three dimensional relationship.The Hansen parameters estimate three important forces that influence solubility and that the sum of the squares of these equaled the Hildebrand solubility parameter squared as given in the equation below.
Where:
δt2 = total Hildebrand solubility parameterδd2 = dispersion force component of solubilityδp2 = polar force component of solubilityδh2 = hydrogen bonding force componentThe Components of Hansen Space
The first is the dispersive/cohesive force (δp). In non-polar solvents this force predominates. It is a measure of the molecule to molecule interaction created by momentary differences in electron distribution. The energy of vaporization can be used to estimate the dispersive force. The second parameter in Hansen space is an estimate of the polar cohesive force (δp). It is the force created by the permanent differences electron densities created when electron rich and electron poor atoms are found in the same molecule. The polar force can be estimated from the dipole moment of the molecule. The third force of Hansen space is the hydrogen bonding parameter (δh). It is a measure of the ability to exchange electrons though hydrogen bonding. It can be estimated from the heat of mixing, or can be calculated as the sum of everything not included in the first two parameters. Table 1 provides an overview of HSPs for common solvents.Table 1: Hansen parameters for common solvents. (Source BFK Solutions newsletter [5].)Charting Solvents in Hansen Space Representing 3-D space on a sheet of paper can be a challenge. At best, you have to use your imagination to visualize the data. Software is available to allow "point of view" rotation to assess the 3-D data. Figure 4 shows a single point of view look at solvents plotted in Hansen space. Figure 4: 2-D representation of 3-D Hansen space.Looking at the data represented in Figure 4, it is apparent that hexane has a lower polar force than water. It would be difficult to judge if acetone was greater than n-propyl bromide. Traditional two dimensional (x,y) charts can be used to evaluate Hansen data, but this requires two charts to tell the full story.
Teas Charts Teas Charts were developed by J P Teas to allow a 2-D view of 3-D data. Figure 5 shows a generalized view of solvent classes plotted with a Teas graph [7]. The three axes of the chart represent the percentage of each of the three Hansen solubility parameters.Figure 5: Teas chart of Hansen solubility parameters for general solvent classes.Computer Generated "3-D" charts Figure 6 illustrates a chart that places solvent properties into a three dimensional sphere, which is the heart of HSP. The sphere positions solvents in relationship to their ability to dissolve the contaminant, in this case flux residue. The solvents inside the sphere dissolve the contaminant and none of those outside the sphere dissolve the contaminant [8]. The objective is to identify a group of relevant solvents that work best on the specific flux residue. The broader the range of dispersive, polarity and hydrogen bonding properties away from the sphere the lease amount of work those solvents have on the selective soil.Figure 6: The Interaction Sphere is at the heart of the HSP theory.
MethodologyTo determine the HSP's for a given residue (water soluble flux, rosin flux, no-clean flux, lead-free no-clean flux), a simple series of tests were performed. For tests performed in these experiments, samples of the various flux residues were heated to peak reflow temperatures and held at that temperature until low boiling solvents were removed (approximately five minutes) and cooled to room temperatures. An alternative method of preparing samples would be to cut up a soldered assembly and use these small samples for testing. Using the prepared residues the following test is performed. Each material of interest was exposed to a group of 20 solvents selected to represent specific regions of Hansen space. In this test, a "shaker table" and 24 position test rack was used to hold and agitate the samples for a fixed time. A one hour exposure at room temperature was used for determining the HSP of each flux residue presented in this study.The objective is to define the sphere which says "all or most of the solvents inside the sphere dissolve the flux and none or few of the solvents outside the sphere fully dissolve the flux residue [19]." The 20 solvents selected to characterize the flux residues exhibit a broad range of dispersion, polarity and hydrogen bonding parameters. The residues from the nine fluxes in the study were added individually to each of the 20 test solvents at roughly 3% by weight.
The samples were prepared using 12 ml sample vials. Visual results were used to grade the rate of dispersion (Figure 7). Table 2 documents the grading scale used [19]. A score of 1 is very close to the center of the sphere while a six is very far outside the sphere radius. The smaller radius that gives the closest fit to a 1.0 is optimal.Figure 7: Grading scale sample illustrations.Table 2: Grading scale.Upon completion of the test, the samples were graded and entered into the HSPiP software. Analyzing the results of all 20 solvents tested on the specific flux residue type gives the location of the "interaction zone" where dissolution occurs [8]. The software program performs a calculation to find the best fit of the data. Inside solvents (blue circles) are within the radius and all outside solvents are shown as (red squares) [19]. There are some solvents that will disperse the soil but are categorized as "wrong solvents" since their HSP parameters are outside the interaction zone. Such wrong solvents are marked with a "*" and represented as blue open circles. Red open circles represent poor solvents that are inside the sphere. This test procedure establishes the Hansen Solubility Parameter for the specific flux residue tested.Interaction Zone The interaction zone indicates the region where other solvents can be found that could effectively remove the residue. Even though we did not test with these solvents, we now know they too have a high probability of removing the residue. This area of positive interaction in the Hansen plot is interaction space for that material. Depending on the material, this area can be large or small and is usually spherical or oval in shape. The size of the interaction zone is indicated by the Interaction Radius (R), which is the average radius of the zone. The center of the material interaction zone is used to set the Hansen parameters for that solid material (Figure 7).Upon completion of the test, the samples were graded and entered into the HSPiP software. Analyzing the results of all 20 solvents tested on the specific flux residue type gives the location of the "interaction zone" where dissolution occurs [8]. The software program performs a calculation to find the best fit of the data. Inside solvents (blue circles) are within the radius and all outside solvents are shown as (red squares) [19]. There are some solvents that will disperse the soil but are categorized as "wrong solvents" since their HSP parameters are outside the interaction zone. Such wrong solvents are marked with a "*" and represented as blue open circles. Red open circles represent poor solvents that are inside the sphere. This test procedure establishes the Hansen Solubility Parameter for the specific flux residue tested.Interaction Zone The interaction zone indicates the region where other solvents can be found that could effectively remove the residue. Even though we did not test with these solvents, we now know they too have a high probability of removing the residue. This area of positive interaction in the Hansen plot is interaction space for that material. Depending on the material, this area can be large or small and is usually spherical or oval in shape. The size of the interaction zone is indicated by the Interaction Radius (R), which is the average radius of the zone. The center of the material interaction zone is used to set the Hansen parameters for that solid material (Figure 7).Figure 8: Interaction Radius of the Soil Tested.RED Number Relative Energy Distance (RED) is the Hansen Solubility Distance between the given solvent and the reference value, divided by the radius that defines the goodness of fit [19]. The equation for the RED number is the ratio of the distance Ra (distance between the given solvent to the reference value) divided by the Ro (interaction radius). The RED number reflects the relative energy difference [8]. A RED number of 0 has no energy difference. RED numbers less than 1.0 are considered inside solvents, which indicate high affinity. RED numbers equal to or close to 1.0 are considered a boundary condition; which indicates that the solvent will disperse the soil. Progressively higher RED numbers indicate progressively lower affinities that indicate that these materials are the wrong solvents.
RED = Ra/Ro.The distance between flux soils based on their respective partial solubility parameters indicates the boundary regions that those soils find in today's electronic assembly process [8]. What this means is that some flux soils will have a higher affinity for cleaning agents that are inside the interaction zone. Cleaning agents that exhibit a boundary condition or are the wrong solvent for the soil will not clean as well.Editor's Note: Be sure to check next Tuesday's newsletter for the conclusion of this article.AuthorsDr. Mike Bixenman, Chief Technology Officer of Kyzen Corporation, is responsible for R&D, Analytical, Application Testing, Tech Service and Engineering groups at Kyzen. He has twenty years experience in research, development and optimization of electronic assembly cleaning agents and processes. He has authored and/or joint authored greater than 50 research papers on the topic of electronics assembly and advanced packaging cleaning. He holds a Doctorate of Business Administration from University of Phoenix School of Advanced Studies. Dr. Ning-Cheng Lee, Vice President of Technology at Indium Corporation, has been responsible for development of fluxes and solder pastes since 1984. He received his Ph.D. in polymer science from University of Akron in 1981. He is the author of "Reflow Soldering Processes and Troubleshooting: SMT, BGA, CSP and Flip Chip Technologies" by Newnes, and co-author of "Electronics Manufacturing with Lead-Free, Halogen-Free and Conductive-Adhesive Materials" by McGraw-Hill.
Steve Stach, President and CEO of Austin American Technology Corporation, has been responsible for development of new cleaning new cleaning systems for the last 22 years. He also has 10 years of experience as a Process Engineering Manager for both Defense and Medical Electronics firms specializing in cleaning processes. He has authored or co-authored more than 50 research papers on cleaning as early as 1979. He has a BS in chemistry and graduate work in chemical engineering. He holds several patents in cleaning technology.References:
1. IPC TM-650. 2. Stach, S., "Combining Cleaning and Testing in One Machine," US Tech, February, 2009.3. Mil-P-28809, "Military Specification Circuit Card Assemblies, Rigid, Flexible and Rigid Flex," 1972. 4. Kenyon, W.G., "Why Test for Ionics Anymore?" SMT Magazine, August 2004.5. Stach, S., "Using Hansen Parameters to Optimize Solvent Cleaning," SMT Magazine Online, June 2009.6. "The Physics of Cleaning; Part 5," BFK Solutions Newsletter.7. Burke, J., AIC book, Volume 3, 1984.8. Hansen, C., "Hansen Solubility Parameters, A Users Handbook," CRC Press, Boca Raton, Florida, 2007. 9. IPC International Technology Roadmap for Electronic Interconnections 2006-2007. 10. iNEMI Roadmap, 2006. 11. Ning-Cheng Lee, "Combining Superior Anti-Oxidation and Superior Print - Is it Really Impossible?" EPP EUROPE, pp. 20-21, December, 2007. 12. Gregory Evans and Ning-Cheng Lee, "Solder Paste: Meeting the SMT Challenge," SITE Magazine, 1987. 13. Ning-Cheng Lee, "Reflow Soldering: Processing and Troubleshooting SMT, BGA, CSP and Flip Chip Technologies," Newnes, pp.288, 2001. 14. Ning-Cheng Lee, "Achieving high reliability lead-free soldering--materials consideration," ECTC, short course, San Diego, California, May 26-29, 2009. 15. Benlih Huang, Arnab Dasgupta and Ning-Cheng Lee, "Effect of SAC Composition on Soldering Performance," Semicon West, STS: IEMT, San Jose, California, July 13-16, 2004. 16. Wusheng Yin, Ning-Cheng Lee, Fred Dimock and Kristen Mattson, "Effect of Flux and Cooling Rate on Microstructure of Flip Chip SAC Bump," SMTA International, Chicago, Illinois, September, 2005. 17. T.R. Bieler, H. Jiang, L. P. Lehman, T. Kirkpatrick, and E. J. Cotts, "Influence of Sn Grain Size and Orientation on the Thermomechanical Response and Reliability of Pb-free Solder Joints," ECTC, p. 1462-1467, San Diego, California, May 30-June 2, 2006.18. Weiping Liu, Ning-Cheng Lee, Adriana Porras, Dr. Min Ding, Anthony Gallagher, Austin Huang, Scott Chen and Jeffrey ChangBing Lee, "Achieving High Reliability Low Cost Lead-Free SAC Solder Joints Via Mn Or Ce Doping," 59th ECTC, San Diego, California, May 26-29, 2009.19. Abbott, S., Hansen, C., Yamamoto, H. and Valpey III, R. "Hansen Solubility Parameters in Practice. Hansen-Solubility," ISBN: 978-0-9551220-2-6, 2008.20. Kenyon, W.G., "Ionic Testing, A Primer," SMT Magazine, April, 2007.21. Hildebrand Solubility Parameter, Wikipedia, 2009.As originally published in the proceedings of SMTA International, October 4-7, 2009.Access thousands of full-length technical articles at the SMTA Knowledge Base.Don't miss the August issue of SMT Digital! Find it here.