Solders for Thick Gold Plating Consideration of solder characteristics, joint shear strength, and soldering conditions determines the choice of a solder-flux system

BY DAVI D H. BREWER

ABSTRACT. A study was made to find a produce joints that appear "cold" and tie, acicular crystals of AuSn4 in the solder more suitable than 63Sn-37Pb for are extremely weak. Certain other solder. These large crystals give the use with thick gold plating. Conclusions high-temperature solders that are bet­ solidified solder a crystalline appear­ were based upon the results of wettabil- ter adapted for use with gold plating ance that can be confused with a 'cold ity and shear-strength tests. The test often cannot be used because of the solder joint.' methods used provided results that are required use of rosin fluxes or the "3. Except for very thin deposits of reproducible with equipment that is un­ approximately 0.01 mil and less, alloy complicated. presence of organic insulating materi­ Eutectic 63Sn-37Pb solder is satisfac­ als. golds containing small quantities of tory for use with thick gold plating only This investigation was made to find nickel, cobalt, or silver cause consid­ if soldering time and temperature are a solder more suitable for use with erably weaker solder joints than do closely controlled to prevent excessive thick gold plating than 63Sn-37Pb. pure golds of comparable thicknesses. alloying of tin and gold. Used under Thick gold plating was considered to The weak solder joints are due pri­ conditions that produce excessive alloy­ be a deposit of 0.3 mil or greater. The marily to a dewetting condition. The ing, 90In-10Ag has a higher shear solder initially wets the gold surfaces strength than that of 63Sn-37Pb. Al­ investigation was made in two phases: 1. A study of the wettability of gold and then draws back to leave ridges of though high in shear strength, 80Au- solder connecting the joint. 20Sn solder is restricted in use because plating by the various solders. of its high soldering temperature. 2. A determination of the shear "4. Pure golds and alloy golds con­ A study of the wettability and shear- strength of the solders after they were taining nickel, cobalt, or silver are strength test results presented in this alloyed with gold. significantly, although not markedly, paper will aid in the choice of a solder- All of the solders used in the inves­ more readily wetted by tin-lead sol­ flux system for almost any intended use tigation are commercially available. ders than is bare copper. involving thick gold plating. "5. Pure golds are only slightly Background more readily wetted than alloy golds. Introduction An earlier study of gold-plating for "6. Thinner golds of both pure and Technological advancement in the solderability by Harding and Pressly* alloy types are slightly more readily electronics industry has led to increas­ presents the following six conclusions: wetted than thicker golds." ing use of gold-plated components. "1. Thin, pure gold plating of ap­ In the same report, Harding and Gold plating may be used to minimize proximately 0.05 mil and less and very Pressly make the following recom­ the resistance of a conductor, de­ thin, alloy gold plating of approx­ mendations: crease contact resistance, prevent cor­ imately 0.01 mil and less dissolve "Thin gold plating (approximately rosion, or, in thin deposits, to increase readily in tin-lead solder and have no 10 millionths) is definitely a soldering solderability. Difficulty is often en­ measurable detrimental effects on the aid and has no detrimental effects on countered in soldering, however, when structure or strength of the solder the structure or strength of solder design considerations dictate that a joints. joints. thick gold plating be used. "2. Thick gold plating of approx­ "Thick gold plating (approximately The pins of a miniature connector imately 0.05 mil and greater dissolves 50 millionths and greater) should be block are a typical example of a thick readily in tin-lead solder at normal avoided when strength is important. gold-plating application. While it soldering temperatures. The gold-tin- "In those applications where thick would be desirable to have a thin lead alloy formed is significantly gold plating is required for other rea­ plating on the end of the pin that is weaker than eutectic tin-lead solder, sons, as thin a gold as can be tolerated soldered to the conductor, repeated its strength varying inversely with the should be used; pure gold deposits abrasion of the opposite end by the gold content. Thus, any factor that should be chosen; and the soldering mating connector demands that a will increase the amount of gold in the operation should be precisely con­ thick plating be used. The small size solder will produce weaker solder trolled to minimize soldering tempera­ of the pins makes it impractical to use joints; that is, increased gold thick­ ture and soldering time." two different plating thicknesses. ness, increased soldering temperature, Emphasis is placed on the fact these Excessive alloying of tin-lead solder and increased soldering time. The conclusions and recommendations are with gold has long been known to gold-tin-lead alloy formed is weaker for tin-lead solders, and the use of than the eutectic tin-lead solder be­ gold plating, to enhance solderability. DAVID H. BREWER is with the Kansas cause of the formation of large, brit- City Division, Bendix Aerospace-Electron­ Solder Selection ics Co., Kansas City, Mo. "Harding, W. B., and Pressly, H. B., The choice of solders to be included Paper presented at the AWS National Fall "Soldering to Gold Plating," Technical in this investigation was based upon Meeting held in Baltimore, Md., during Proceedings, 1963, American Electroplaters' Oct. 5-8, 1970. Society, pp. 90-106. several factors. Table 1 lists the solder

WELDING RESEARCH SUPPLEMENT | 465-s Wettability Study Table 1—Properties of Solders Considered for Investigation Method Eutectic (E) or liquidus (L) Wettability of the solders chosen melt for study was determined by melting a Solder temperature, specific quantity of each solder on a composition % °F Comments circular specimen punched from a 63Sn-37Pb 361 E Common, general-purpose solder for electrical gold-plated panel. After cooling, the connections. outline of the wetted area was traced on translucent paper cross-ruled in 80ln-15Pb-5Ag 315 E Indium-base solders are weaker without tin, but 0.1 in. squares (each unit having an 90ln-10Ag 446 L gold-indium intermetallics may not be as brittle. area of 0.01 sq in.). The included squares were counted, and fractional 96.5ln-3.5Bi» 304 L Recommended by manufacturer "for soldering parts were estimated to the nearest thin films of gold without appreciable gold tenth of a unit. scavenging." A supplementary method of ex­ 97.5Pb-1.5Ag-1.0Sn 588 E pressing solder wettability was used— Fig. 1. This method consists of deter­ 68Sn-32Cd 349 E Alloys containing cadmium are used for solder­ mining the angle between the surface 83Pb-17Cd 478 E ing in the jewelry industry. of the solder (at its leading edge) and the surface of the base metal on which 80Au-20Sn 536 E Gold-base alloys recommended by manufac­ turer for soldering "because of their excellent the gold is plated. For this purpose, a 88Au-12Ge 673 E wettability, good thermal conductivity, high representative cross section of the strength, resistance to many etchants, and specimen is photomicrographed at 94Au-6Si 698 E low melting point." Alloys with melting points 50X. A V2 in. chord is drawn on the higher than that of 80Au-20Sn were found in­ photomicrograph from the leading 27AU-73T1 268 E capable of being fluxed by rosin fluxes. Most edge of the solder to its top surface. of them are quite brittle. The 27Au-73Tl alloy The included angle is termed the was so brittle that it could not be rolled into "contact angle." preforms needed for the shear specimens. Results from the two methods of 56.5Bi-43.5Pb 255 E expressing solder wettability were cor­ related, and acceptable limits were a Composition determined by laboratory analysis. defined for each method.

compositions that were initially con­ and gold, despite their frosty appear­ Specimen Preparation sidered, together with the preliminary ance. Test panels for the wettability study data obtained in preparation for the High melting temperatures excluded consisted of two 10V2 by 12 in. study. Except for 63Sn-37Pb, solders 88Au-12Ge and 94Au-6Si solders sheets of 0.020 in., 70-30 half-hard containing a high percentage of tin from the investigation because of the brass. Preparation before plating, pri­ were avoided because of the brit- presence of organic insulating and marily done to remove tarnish, con­ potting materials in the components to tleness of the gold-tin intermetallic sisted of vapor degreasing, cleaning compounds that would be formed. be soldered, and the required use of anodically in an alkaline cleaner The 63Sn-32Cd composition was in­ rosin-alcohol and rosin-terpene fluxes. (Oakite 191), and rinsing. For econo­ cluded because of its reported use in Of the original twelve solders con­ my, and to prevent edge build-up of the jewelry industry. sidered, the following nine were cho­ plating, the back and edges of each sen for study: In general, gold- and indium-base panel were masked with platers' tape solders were considered to be the 1. 63Sn-37Pb. to leave an exposed area of 8V2 by most promising. It was reasoned that 2. 80In-15Pb-5Ag. 10 in. The panels were then dipped in the gold-base solders would pick up 3. 96.5In-3.5Bi a 2 lb/gal solution of ammonium per- additional gold from the plating, 4. 68Sn-32Cd. sulfate at room temperature for 2 thereby decreasing the percentage of 5. 90In-10Ag. min, rinsed, dipped in 10% sulfuric intermetallics, and probably increasing 6. 83Pb-17Cd. acid at room temperature for 20 sec, the ductility. It was also thought that 7. 80Au-20Sn. rinsed, and copper-plated in a sulfate indium-gold intermetallics would be 8. 97.5Pb-1.5Ag-1.0Sn. less brittle than those formed from tin bath for 4 min at 15 asf (amp/sq. ft.) 9. 56.5Bi-43.5Pb. to obtain a maximum deposit thick­ ness of 0.1 mil. Following the copper plating, the panels were rinsed, dipped again in 10% sulfuric acid, rinsed, dipped in sodium cyanide at room temperature for 30 sec, and gold-plated in a pro­ prietary pure-gold plating solution (Sel-Rex BDT 200) for 42 min at 3 asf. This was calculated to produce a gold deposit ranging in thickness from 0.3 to 0.4 mil. A formed anode was used in the plating tank in an attempt to obtain a BRASS SUBSTRATE uniform deposit. After the first panel Fig. 1—Method of determining contact angle on 50X photomicrograph was plated and checked for plating

466-s I OCTOBER 1970 •I" BORDER AND BACK MASKED WITH PLATERS Table 2—Beta Backscatter Thickness TAPE PRIOR TO PLATING Values for Gold Plating on Wettability Specimens Panel 1 Thick- Panel 2 Thick- specimen ness, specimen ness, No. microin. No. microin. l-l i i. 18 N / 1-19 I 1-20 \ 2 1-1 235 2-1 230 V 2" J I \ 18 ) \ 2-19 2-20 j 1-2 242 2-2 245 y / 1-3 230 2-3 235 225 2-4 215 "N N 1-4 \ ( 1-6 1-5 235 2-5 225 16 i r i-7 2 6 2 7 1-6 242 2-6 240 16 ) I 1: \ - v v - y ^ 1-7 305 2-7 295 10" 1-8 305 2-8 295 1-9 293 2-9 305 4 \ x 1-10 355 2-10 330 •\0 1 2-8 5 'a , 1-11 297 2-11 290 \ •' / 'v-/ \ v 1-12 290 2-12 300 1-13 280 2-13 265 \ / 1-2 N 1-14 230 2-14 225 I I-! 212 2-15 215 -3 ' 2"2 J 2-9 y 1-15 \ v ^ 1-16 230 2-16 225 1-17 265 2-17 260 1-18 280 2-18 265 I — I 0 > 2-19 D 2 1-19 300 285 2 ,2 y v 2-M ; V 2-10 y 285 2-20 330 V -'V \ " -Ml 1-20 Avg 267 Avg 264

1 hi/ *" Table 3—A Comparison of Beta - 8-1/2" Backscatter and Microscopic Measure­ -10-1/2" ments of Gold Plating Microderm Microscope Fig. 2—Location of wettability specimens on plated panels Specimen thickness, thickness,11 reference microin. microin. thickness and uniformity, a modifica­ measurements may be obtained from 1-1 235 233 tion was made in the shape of the Table 3. The marked increase in thick­ 1-20 285 496 anode before plating the second ness shown by the microscopic meas­ 2-1 230 248 panel. urements made near specimens 1-20 2-20 330 448 and 2-20 may be attributed to the Following removal of the masking a Measured near the indicated specimen. tape, each panel was scribe-marked location being near the edge of the 1 into 20 two-inch squares. A l /2 in. panels. Analysis of the data in Tables diameter specimen was punched from 2 and 3 indicates that the plating the investigation consisted of 0.062 in. the center of each square and iden­ thickness of Panel 2 averaged slightly diameter wire. All of the others were tified according to the plot shown in less, but was more uniform, than that obtained in the form of foil that meas­ Fig. 2. The specimens were then va­ of Panel 1. Although the thickness on ured 0.500 by 0.375 by 0.005 in. The por degreased and individually stored most of the specimens was less than amount of wire and of foil required to in tarnish-inhibiting paper. had been desired, it was considered provide a volume of 0.002 cu. in. The thickness of the gold plating adequate for the wettability study. (equivalent to a V8 in. cube) was was measured at the center of each The 63Sn-37Pb test solder used in calculated. Foils for each solder speci­ specimen with a nondestructive beta men were stacked and folded twice, backscatter thickness-measuring in­ and the wire solder was rolled into small spheres, to produce small, com­ strument. The values obtained are SUPPORT FUTURE shown in Table 2. The plating- pact masses. thickness measurements of Panel 1 LEVELING RODS (3) The apparatus with which the speci­ ranged from 212 to 355 microin., with mens were prepared is illustrated in TEFLON SUPPORT RING an average of 267; those from Panel Fig. 3. A solder bath of eutectic tin- 2, 215 to 330 microin., with an aver­ lead solder served as the heat source age of 264. for melting the test solder. A variable In addition, two microscopic meas­ autotransformer was used to control urements were made from the re­ the temperature. maining portion of each panel for Each specimen was secured in the comparison with the beta backscatter Teflon support ring, and three drops reading of the nearest specimen. The of an activated rosin flux in alcohol locations of the microscopic measure­ (Kester 1544) were spread over its ments corresponded on the two panels entire bottom surface with a small and were near specimens 1-1, 1-20, brush. The support ring was then 2-1, and 2-20. screwed onto the three leveling rods, A comparison of the beta backscat­ and leveling was accomplished by ad­ ter and microscopic plating-thickness Fig. 3—Wettability test apparatus justing the rods while observing a

WELDING RESEARCH SUPPLEMENT | 467-s small bubble level that had been placed on the gold-plated surface of Table 4—Spread Area and Contact Angle of Wettability Specimens the specimen. Specimens were selected Solder Bath Contact for use in the order of decreasing Solder-flux temperature,b Wetted area, angle, 3 plating-thickness. system ° F Panel no. sq in. deg The test solder was placed in the 63Sn-37Pb 410 ±5 1-8 0.195 16 center of each specimen in such a Kester 1544° 1-7 0.205 manner that a minimum area of the 1-15 0.175 gold-plated surface was contacted. Avg 0.191 The solder was then covered with two drops of the specified flux. Rosin 90ln-10Ag 500 ± 5 2-6 0.130 fluxes using alcohol as the solvent Kester 1015d 1-2 0.145 were used at temperatures below 500° 1-6 0.130 20 F. Because of its lower volatility, a flux using terpene as the solvent was Avg 0.135 used at temperatures of 500° F and 80Au-20Sn 586 ±5 1-3 0.115 above. Kester 1015 1-14 0.121 22 After the dross had been skimmed 2-1 0.121 from its top, the solder bath was raised by means of the laboratory jack Avg 0.119 to contact the bottom surface of the specimen. The bath was kept in con­ 63Sn-37Pb 410 ± 5 1-10 0.115 e 2-10 0.093 tact with the specimen for 5 sec after Alpha 100-30 2-20 0.108 18 melting the test solder. The test solder was allowed to solidify undisturbed, Avg 0.105 after which the specimen was re­ moved from the leveling rods and 97.5Pb-1.5Ag-1.0Sn 638 ±5 2-15 0.110 — support ring and immersed in trichlo- 2-5 Ruined roethylene to remove the excess flux. 56.5Bi-43.5Pb 370 + 5 2-14 0.100 — Results Kester 1544 -0 1-16 0.095 The wetted area of each specimen Avg 0.098 was measured, as previously de­ scribed, and the results recorded. The 96.5ln-3.5Bi 370+ 5 2-11 0.078 specimen nearest the mean value of Kester 1544 -0 1-12 0.088 31 1-9 0.107 each solder-flux group measured was then chosen for determination of the Avg 0.091 contact angle. The results obtained from both methods of denoting wetta­ 80ln-15Pb-5Ag 370 + 5 1-11 0.090 36 bility are shown in Table 4, in the Kester 1544 -0 1-19 0.090 order of decreasing wetted area. As 2-9 0,090 might be expected, an analysis of the data shows an inverse correlation be­ Avg 0.090 tween the wetted area and the contact (Continued on next page) angle of the test solder. Based upon an examination of the terns that are considered suitable for study. specimens and a study of the data general use with thick gold plating, obtained, the following guidelines based upon the average wetted areas Lap Shear-Strength Study were established as criteria for judging determined from the wettability Method solder wettability: Application of the criteria estab­ 1. Solder-flux systems producing lished for determining acceptable wet­ wetted areas smaller than 0.050 sq in. Table 5—Solder-Flux Systems Suitable tability characteristics eliminated for 0.002 cu. in. of solder, or contact for General Use With Thick Gold 83Pb-17Cd solder from the shear- angles of 65 deg or more, do not wet Plating, as Determined From Wettability Study strength investigation. Six lap shear- sufficiently. strength specimens were made from 2. Solder-flux systems producing Average each of the other solders used in the wetted areas larger than 0.100 sq in. Solder- Use wetted wettability study. Of each group, three for 0.002 cu. in. of solder, or contact flux temperature, area, specimens were made at 50° F above, angles of 25 deg or less wet sufficient­ system °F sq in. and three at 150° F above the eutec­ ly to flow. A solder meeting this defi­ 63Sn-37Pb 410 ± 5 0.191 tic or liquidus temperature. The object nition is termed, "suitable for general Kester 1544 of the higher-temperature study was use." to determine the effect upon shear 90ln-10Ag 500 ±5 0.135 3. Solder-flux systems producing Kester 1015 strength of additional gold being al­ wetted areas ranging between these loyed with the solder. two sets of conditions should be used 80Au-20Sn 586 ± 5 0.119 All six of the 80Au-20Sn solder only when the solder can be preplaced Kester 1015 specimens were made at the lower and will not be required to flow to temperature, since a temperature form a good joint. 63Sn-37Pb 410 ± 5 0.105 150° F above the eutectic tempera­ Alpha 100-30 Table 5 shows the solder-flux sys- ture was considered excessive. Prior to

468-s I OCTOBER 1970 creased mass of the lap shear speci­ Table 4—(Continued) mens (compared to that of the wetta­ Solder Bath Contact bility specimens), the animal fat was Solder-flux temperature,b Wetted area, angle, heated 5° F above the desired temper­ system" °F Panel no. sq in. deg ature for temperatures below 500° F, 56.5Bi-43.5Pb 370+ 5 1-4 0.070 — and 10° F above the desired tempera­ Alpha 100-30 - 0 2-16 0.082 ture for temperatures of 500° F and above. Under these conditions, 50 sec Avg 0.076 were required for each specimen to reach its designated temperature. 68Sn-32Cd 400 ± 5 2-18 0.060 With an additional 5 sec allowed for Kester 1544 1-17 0.068 44 1-18 0.070 formation of the joint, the total time in the bath was 55 sec. Avg 0.066 Results 80ln-15Pb-5Ag 370 + 5 2-12 0.063 Table 6 shows the results of the lap Alpha 100-30 -0 2-8 0.050 shear-strength tests. A comparison of 2-7 0.057 64 column 3 with column 2 illustrates the effect on shear strength of additional Avg 0.057 gold being alloyed with the solder. While the average strength of some of 96.5ln-3.5Bi 370 + 5 1-13 0.055 59 the solders increased with the pick-up Alpha 100-30 - 0 2-19 0.058 of additional gold, four of them 1-20 0.050 showed a decrease. This is an impor­ Avg 0.054 tant consideration in choosing a solder for a loosely controlled operation. A 83Pb-17Cd 530 ± 5 1-5 0.041 solder that gains strength as the gold Kester 1015 1-1 0.030 content increases is more desirable 2-3 0.037 86 than one that loses strength. 576 2-4 0.040 Microsections of the joint com­ Avg 0.037 ponents, made after the shear tests were completed, showed that the en­ 68Sn-32Cd 400 ± 5 2-2 0.025 tire thickness of the gold plating had Alpha 100-30 2-17 0.037 | ' I 69 been alloyed with solder at tempera­ 2-13 0.040 tures of about 500° F, and higher. This accounts for the similarity in the Avg 0.034 strength values obtained from the high-temperature solders. • All solder volumes were 0.002 cu in. b Temperature measurements made with ASTM glass-stem mercury thermometer Table 6 also shows the disastrous " Kester 1544 is an activated rosin flux in alcohol. d Kester 1015 is an activated rosin flux in terpene solvent. results that occur from the use of '•' Alpha 100-30 is a no nactivated rosin flux in alcohol. tin-lead solder when soldering condi­ tions are not controlled. By accident, the shear tests, three of the 80Au- out. Each assembly was held together three specimens were soldered with 20Sn specimens were subjected to with a steel binder clip. 63Sn-37Pb at 284° F above the eutec­ thermal shock in accordance with Solder joints were formed by dip­ tic temperature. A study of the data Military Standard No. 202, Method ping the assemblies in animal fat shows that the strength of the joints 107, Test Condition B. Because 80Au- which had been heated to the required averaged about one-third that of the 20Sn, as purchased, is quite brittle, it temperature. Flux was not used, as joints produced at 150° F above the was thought the joints might prove to the animal fat was considered suffi­ eutectic temperature, and one-sixth be sensitive to thermal shock. ciently active. that of the joints produced at 50° F Shear-strength tests were per­ To compensate for the slower heat above the eutectic temperature. formed at room temperature, at a rate transfer of animal fat (compared to Data in Table 6 further indicate of 0.050 ipm, in a tensile-testing ma­ that of liquid solder) and the in- that thermal shock has no effect upon chine with two self-aligning grips. A the strength of 80Au-20Sn solder. It grip-length of 0.625 in. was used on may be noted that the highest strength all specimens. was exhibited by a joint that had .00* COPPER WIRE undergone thermal shock, and the Specimen Preparation SPACER. 2 PLACES —V^Q SOLDER lowest, by a joint that had not been so PREFORM The dimensions and assembly of the subjected. lap shear-strength specimens are illus­ Table 7 shows the average shear trated in Fig. 4. The panels consisted strengths, in decreasing order, of the of annealed copper, gold-plated in an \ three solders that are considered suit­ acid citrate pure-gold bath to a thick­ able for general use with thick gold ness of from 0.3 to 0.4 mil. Pre­ plating. A study of Tables 5 and 7 will formed foils of the various solders, aid in choosing a solder-flux system .500 each measuring 0.500 by 0.375 by for almost any intended use. Good 0.005 in., were used between the results can be obtained from eutectic lapped areas of the panels, and 0.004 tin-lead solder if the soldering time in. diameter wire spacers were inserted Fig. 4—Dimensions and assembly of and temperature are held to a mini­ to prevent excessive solder squeeze- lap shear-strength specimens mum.

WELDING RESEARCH SUPPLEMENT! 469-s Table 6. Lap Shear Joint Strength Table 7—Average Shear Strength of Solders Suitable for General Use with Shear strength Shear strength Thick Gold Plating (joint 50° F above (joint 150° F above eutectic or liquidus), eutectic or liquidus) Average Average Solder psi psi shear shear strength strength 63Sn-37Pb 4270 510 1970 (joints 3550 750} 1730 (joints 50° F above 150° F above 3590 6501 1820 eutectic eutectic Avg 3800 Avg 640 Avg 1840 or liquidus), or liquidus), Solder psi psi 80ln-15Pb-5Ag 1820 1900 80Au-20Sn 6800 — 1674 1800 63Sn-37Pb 3800 1840 1460 1840 90ln-10Ag 1090 1010

Avg 1650 Avg 1850 be obtained only if soldering time and 96.5ln-3.5Bi 740 860 temperature are controlled at a mini­ 310 1230 mum to prevent excessive alloying 720 1150 with gold. Ordinarily, soldering Avg 590 Avg 1080 procedures do not provide such con­ trol. 68Sn-32Cd 2280 3070 2. The 80Au-20Sn solder has the 2460 2800 highest shear strength of all those 2610 3190 tested and ranks third in wettability. Avg 2450 Avg 3020 Its use is restricted because of its high soldering temperature (586° F). Al­ 90ln-10Ag 1020 1010 though it is somewhat brittle, its 1090 1040 strength is not adversely affected by 1160 980 thermal-shock cycling. 3. The 90In-10Ag solder, used with Avg 1090 Avg 1010 Kester 1015 flux, ranks second in 80Au-20Sn' 70401 wettability. Its application tempera­ 5490;No thermal shock ture (500° F) is moderate. The shear 7080) strength is lower than that of either 6980 63Sn-37Pb or 80Au-20Sn when joints 7770J-Thermal shock are made within 5 sec at 50° F above 6450] the eutectic or liquidus temperature. Used under conditions that produce Avg 6800 excessive alloying with gold, 90In- 97.5Pb-1.5Ag-1.0Sn 3070 2130 lOAg has a higher shear strength than 2880 4150 that of eutectic tin-lead solder. 3550 3120 The findings of the study described in this paper show that either 90In- Avg 3170 Avg 3130 lOAg or 80Au-20Sn may be more suitable for use with thick gold plating 56.5Bi-43.5Pb 1560 1200 1630 1520 than 63Sn-37Pb. The 63Sn-37Pb can 1990 1530 be used effectively only when solder­ ing time and temperature are con­ Avg 1730 Avg 1420 trolled to minimize alloying of gold with tin. "Joints formed at 645° F (284° F above eutectic temperature) by accident. Data included for comparison. b Lap distance was decreased from V* to 3/i« in. to cause break to occur in the joint. A cknowledgments The author gratefully acknowledges Conclusions 63Sn-37Pb solder has greater wetta­ the assistance of H. B. Pressly in Results of the wettability and lap bility than any of the other solders gathering preliminary information on shear-strength studies performed on tested. It also has the second-highest various solders, W. R. Van Luchene in various solders used with thick gold shear strength if joints are formed the preparation of plated panels, and plating indicate the following conclu­ within 5 sec at 410° F. Its primary K. Gentry, Jr. in the preparation of sions: advantage is its low melting tempera­ wettability and shear-strength speci­ 1. If used with Kester 1544 flux, ture (361° F). Satisfactory results can mens.

470-s OCTOBER 1970