Room Temperature Aging Properties of Some Solder Alloys

Sn-Pb and Sn-Pb-Sb alloys undergo a loss in shear strength and hardness during 1st 30 days of storage af ter soldering, and some alloys require 30-60 days of ag ing to stabilize their mechanical properties and microstructure although post-solder annealing at 200 F for 80 hours effectively stabilizes the mechanical properties of 50Sn-47Pb-3Sb

BY B. T. LAMPE

ABSTRACT. Tin-lead and tin-lead- ly inexpensive, have desirable elec softening" transformation at room antimony alloys were found to age- trical properties, and can be used to temperature. soften during storage at room tem join most metals. The mechanical The mechanical properties of many perature. Some alloys experienced strength of tin-lead solder, however, low melting alloys change gradually greater than a 20% loss in shear is extremely low compared with the with time at room temperature be strength and hardness during the first metals which they are generally used cause of minute changes in struc 30 days of storage after soldering. to join. When the solder joint must ture. This aging at room temperature The loss in strength which is accom carry a load, the joint must be de is sometimes referred to as room panied by structural changes in the signed to avoid dependence upon temperature recrystallization or alloys is attributed to the precipita the strength of the solder. annealing. Aging sometimes has the tion of (S-tin out the supersaturated The strength of a solder joint is effect of increasing the hardness and lead-rich phase. The presence of anti strongly dependent upon the joint gap strength of the alloy, but it can also mony reduces the precipitation rate or the clearance between the sol have the opposite effect. For exam and prolongs the age-softening pro dered interfaces. Joint strengths far ple, a chill-cast quaternary eutectic of cess in addition to refining the grain exceeding the strength of the bulk tin, bismuth, cadmium, and lead structure. solder can be obtained with very shows an increase in strength of Some alloys required 30 to 60 days small (less than 0.001 in. or 0.025 about 34% after 16 weeks of aging at of aging before their mechanical mm) joint gaps. Joint gaps exceed room temperature (Ref. 1). properties and microstructure stabil ing 0.005 in. (0.13 mm), however, gen Lead-antimony alloys exhibit the ized. The mechanical properties and erally produce joint strengths ap same age-hardening property. If aging characteristics were found to be proaching that of the bulk solder. Tin- heated above their solid solubility sensitive to prior thermal treatment. A lead solders also have characteris temperature and then quenched, they post-solder anneal was found effec tically low creep strengths, especially show a marked increase in hardness tive in stabilizing the properties of the at elevated temperatures. Alloying on aging, the extent of which is alloys. elements such as antimony, copper, dependent on the amount of anti and silver are commonly added to tin- mony present and the rate of cooling lead solders to improve their creep Introduction (Ref. 2). Tin-antimony-cadmium strength. alloys containing 7-14 wt-% antimony The term "solder" has become A production requirement called and 0-2 wt-% cadmium also age- synonymous with the tin-lead alloy for a low (less than 400 F or 204 C) harden when tempered at 100 C system. Tin-lead solders are relative- melting point solder with good shear (212 F) after water quenching from strength and creep strength. A tin- above 200 C (414 F) (Ref. 3). This lead-antimony solder (50Sn-47Pb- hardening is attributed to the disper B. T. LAMPE is Supervisor - Advanced 3Sb) was selected, but solder joints sion of the SnSb intermetallic in the Development Engineering, Bendix Corpo made with this alloy were found to matrix. Tin-antimony alloys contain ration, Kansas City, Missouri. ing less than 7% antimony age-soften Paper presented at the Fifth Interna lose considerable strength during when tempered at 100 C (212 F). tional AWS Soldering Conference held in storage. It appeared that the solder St. Louis, Missouri, May 11-13, 1976. was undergoing some type of "age- Softening occurs in two stages: first

330-s I OCTOBER 1976 ANI MONUL EUTECTICS

TIME AFTER CASTING (DAYS) 10 15 20 25 30 35 4 U5 50 55 60 (FROM BAKER - REFERENCE 4)

TIME AFTER SOLDERING (DAYS)

60Sn-40Pb (0.2St>)

SSb)

8Sb)

0 20 40 60 80 IOO 120 140 160

TIME AFTER SOLDERING (DAYS) 0 5 IO I5 20 25 30 35 40 45 50 55 60

(FROM DREFAHL - REFERENCE 5) TIME AFTER CASTING (DAYSl Fig. 1 — Age-softening characteristics of chill-cast eutectic Fig. 2 — Room temperature aging of 50Sn-47Pb-3Sb alloy solders

very rapid softening during the first temperatures as shown in the lower 2. 58.5S n-37Pb-4.5Sb. 20-100 hours (h), followed by a much plot of Fig. 1. The tensile strength of 3. 50Sn -49Pb-1Sb. longer period of very gradual sof these solders is reduced by up to one- 4. 50Sn -50Pb. tening. third as compared to their strength in 5. 50Sn -47Pb-3Sb. Other soft solders are subject to the homogenized poured condition. 6. 50Sn -49.6Pb-0.35Ag-0.05Cu age-softening at room temperature as The loss in strength is attributed to tin 7. 60Sn -37Pb-3Sb. illustrated in Fig. 1. The upper plot precipitating from the supersat 8. 63Sn -37Pb. shows the steady decrease in hard urated lead-rich solid solution («) dur 9. 60Sn -38.6Pb-1.0Sb-0.4Ag. ness with time at room temperature of ing storage at room temperature. The some chill-cast eutectic solders. The formation of a globular precipitate antimonial solders (55Sn-41.5Pb- during aging was observed using an Lap Shear Test 3.5Sb) exhibited slight age-hard electron microscope. Formation of ening during the first few days after the intermetallic SnSb was not de It was desired to measure the shear casting which was followed by a pro tected. strength of those soft solder alloys longed age-softening. The nonanti- and minimize the effects of joint gap monial eutectics (63Sn-37Pb) age- Experimental Procedure and alloying from the base metal. A softened continuously at room tem lap shear specimen was designed perature. These alloys appeared to Materials with a 0.008 in. (0.2 mm) gap (dis reach a stable hardness after aging The room temperature aging prop tance between soldered surfaces). for 100-150 days at room temper erties of nine soft solder alloys were The lap shear specimens were pre ature. The original hardness could be studied. The list included two of the pared from copper panels. Two 0.375 restored by annealing at high tem most widely used tin-lead solders and (9.5 mm) by 0.5 in. (13 mm) preforms peratures, followed by quenching or a commonly used tin-lead-antimony of 0.005 in. (0.13 mm) thick solder foil fast air-cooling. It was observed by solder. Other alloys were specially were placed between the panels Baker (Ref. 4) that the hardness after made to evaluate the effects of minor along with two lengths of 0.008 in. (0.2 annealing was directly related to the alloying elements of copper, silver, mm) diameter copper wire to control annealing temperature, making it and antimony. The alloys were pur the joint spacing. The overlap was likely that the hardness of the alloys chased in the form of 0.005 in. (0.13 maintained at 0.375 in. (9.5 mm). was determined by the amount of tin mm) thick ribbon which had been Several drops of rosin-base flux were retained in the lead-rich phase. The cold rolled to final form from vacuum swabbed on each panel before alloys were also found to age-soften cast, high purity materials. The alloys assembly. The assembled lap joint more rapidly at 176 F (80 C) than at studied are listed below.* was clamped together and then im room temperature. mersed for 25 seconds (s) in a hot 1. 47Sn-47Pb-6Sb. Eutectic tin-lead solders contain peanut oil bath maintained at 480 ±10 F (249 ±5.6 C). Time at solder ing 0.2-0.8 weight percent (wt-%) anti 'Alloy compositions are given in weight mony undergo age-softening at room percent (wt-%) ing temperature was 15 s.

WELDING RESEARCH SUPPLEMENT! 331-S Hardness Test face finish. Polishing was done in ature and then undergo a prolonged such a way to minimize any heating of age-softening as shown in Fig. 2. The The relative low hardness of these the specimen. Diamond pyramid shear strength increased during the solder alloys required the use of a hardness measurements were made first three days from about 5830 psi microhardness tester. Fifty-gram 6 using a 100 g load. (40.2 X 10 Pa) to a little over 6000 samples of the various alloys were psi (41.4 x 106 Pa) and then de melted in a stainless steel crucible Metallography creased rapidly until it leveled off after and chill-cast from a temperature about 30 days at 5200 psi (35.9 X 106 (unless otherwise specified) of 575 Metallographic specimens similar Pa). The hardness vs. aging time ±10 F (301.7 ±5.6 C) by placing the to the hardness test specimens were curve showed the same character crucible in a reservoir of running tap prepared for each of the nine alloys istic except it started to decrease at (60 F or 15.6 C) water. The cast by chill-casting from 575 F (301.7 C). about 24 h after casting. solder slugs were ground and pol These specimens were carefully pol Structural changes were observed ished to less than a 1 micrometer sur- ished and etched to reveal the micro- in the alloy during aging. These structure. The following etching changes are shown in the series of r** procedure was used. * * i photomicrographs in Fig. 3. The - 1. Etchant: 10 parts concentrated structure immediately after chill-cast nitric acid; 10 parts glacial acetic acid; ing from 575 F (301.7 C) shows large 80 parts glycerin. . dark crystals of the lead rich solid ' '%, % 2. Time: 10-30 s. solution (a) interspersed in a matrix of ;° 3. Application: Cotton swab. very fine grained eutectic. The eutec • #^ Structure changes occurring in the tic consists of a mixture of very fine 0 HOURS H HOURS samples during aging were recorded crystals of a and $ (tin-rich solid solu .by making X1000 photomicrographs tion). of the etched specimens at regular - J&,, Within 4 h after casting, the eutec intervals using a metallograph. After f tic starts to coarsen. After about 3 d, a the photomicrographs were taken, white precipitate begins to form be the chemical etch was removed by tween the a and p crystals in the light polishing with an aluminum eutectic and in the grain boundaries oxide-water slurry. At the end of each J •rf **v of the large primary cv-crystals. aging interval, the specimens were re- With further aging, the eutectic etched. continues to coarsen and the white precipitate grows in size. The nucle Aging Tests %fl? ation of the precipitate appears to be The lap shear joints, hardness spontaneous and the particles appear specimens, and metallographic spec to grow at the expense of other par imens were stored at room temper ticles rather than by a continuous nu ature (70 ±10 F, or 21.1 ±5.6 C) and cleation and growth process. Sphe- tested at intervals of 30 minutes (m), roidizalion appears to take place, i.e., 24 h, and then 3, 6, 10 16, 24 40, and the precipitate particles increase in 60 days (d). Some aging tests were size and decrease in number as their conducted at 200 i 10 F with the shape becomes more nearly spher 50Sn-47Pb-3Sb alloy. Each slotted ical. point on the hardness and shear strength curves of Figs. 2, 4, 7, 9, 12, 47Sn-47Pb-6Sb Alloy 14 and 16 represents an average of This high antimony solder was five values. found to have aging characteristics

30 DAYS similar to those of the 50Sn-47Pb-3Sb Results alloy. As shown in Fig. 4, the alloy under F/g. 3 — Structural changes in alloy 50Sn- 50Sn-47Pb-3Sb Alloy 47Pb-3Sb during aging at room temper goes initial age-hardening followed by ature. X1000 (reduced 48% on repro This alloy was found to initially age- prolonged softening at room temper duction) harden with time at room temper ature. The degree of softening is not

0 5 10 15 20 25 30 35 40 45 50 55 60 SO 55 60

TIME AFTER SOLDERING (DAYS) TIME AFTER CASTING (DATS)

Fig. 4 — Room temperature aging of 47Sn-47Pb-6Sb alloy

332-s I OCTOBER 1976 as great, however, as for the pre present since the antimony content the precipitate crystals growing at the vious alloy. The shear strength drops (7.7% of tin content) exceeds the sol expense of others during the aging from a maximum of about 5100 psi to ubility limit of antimony in tin. The process. around 4600 psi (35.2 X 106 to 31.7 X matrix is essentially a eutectic struc 6 ture with a few primary a lead-rich 10 Pa) during againg which is 60Sn-37Pb-3Sb Alloy about a 500 psi (3.4 x 106 Pa) de crystals present along with the large crease as compared to a total de cubic crystals of SnSb. The eutectic The aging curves for this alloy are crease of about 850 psi (5.9 X 106 structure coarsens with aging time shown in Fig. 7. The shear strength vs. Pa) for the 50Sn-47Pb-3Sb solder and a white precipitate becomes vis aging time curve does not exhibit alloy. The hardness increases a cou ible after about 3 d and grows with the initial age-hardening characteris ple of points during the first 24 h and further aging time. As in the previous tic of the previous tin-lead-antimony then drops off to a little below the in alloys, the nucleation of the precip alloys, although the hardness curve itial value. The shear strength1 and itate appears to be spontaneous with does. The shear strength of the alloy hardness curves stabilize after about 10 d at room temperature. Structural changes occurring in this alloy during aging are shown in Fig. 5. The antimony content of this alloy is 12.8% of the tin content which is well above the reported (Ref. 1) solubility limit of 7% antimony in tin. White cubic crystals of the inter metallic compound SnSb can be seen in all the photomicrographs of Fig. 5. Noticeable structural changes are ob served in this alloy during room tem perature aging. A white precipitate begins to appear after about 3 d. The precipitate grows with time and ap pears to reach a stable size after 30- 60 d of aging. In some of the photomicrographs of Fig. 5, clusters of small cubic inter metallic crystals can be seen. They appear to be in a process of co alescing into the larger cubic crys tals. The precipitate crystals can be distinguished from the crystals of SnSb because they have a spherical shape rather than cubic.

58.5Sn-37Pb-4.5Sb Alloy The aging characteristics for this alloy are very similar to those of the previous tin-lead-antimony alloys. The shear strength and hardness of 30 DAYS 60 DAYS the alloy appear to stabilize after 10- 30 DAYS 60 DAYS 15 d at room temperature. Fig. 5 — Structural changes in alloy 47Sn- Fig. 6 — Structural changes in alloy Structure changes occurring dur 47Pb-6Sb during aging at room temper 58.5Sn-37Pb-4.5Sb during aging at room ing aging are shown in Figure 6. Cubic ature. X1000 (reduced 48% on repro temperature. X1000 (reduced 49% during crystals of SnSb intermetallic are duction) reproduction)

7I00

6900 I

6700 \ » 6500

6300 .

6100

5900 •

i i , i i i i I

15 20 25 30 35 H5 50 55 60 15 20 25 30 35 4 15 50 55 60

TIME AFTER SOLDERING (DATS) TIME AFTER CASTING (DAYS)

Fig. 7 — Room temperature aging of 60Sn-37Pb-3Sb alloy

WELDING RESEARCH SUPPLEMENT! 333-s decreases more than 1000 psi (6.98 x dark lead-rich («) lamellae of the Structural changes occur in the 106 Pa) during the first 25 d after eutectic after 3 d. The precipitate 50Sn-50Pb alloy during room tem soldering while the diamond pyramid grows with time, with some of the pre perature aging as shown in Fig. 11. hardness (DPH) only drops about 4 cipitate crystals growing completely The microstructure consists of large points. The strength and hardness across the width of the lamellae. The dark a (lead-rich solid solution) crys appear to stabilize after about 25-30 lead-rich lamellae also coarsen dur tals in an eutectic matrix. The eutec d. ing aging. The structure appears to tic structure has a "salt and pepper" Structural changes during room stabilize after about 60 d at room tem appearance with small dark n-crys- temperature aging are shown in Fig. perature. tals dispersed in a light 0-tin matrix. 8. The structure is basically a laminar The small

5700

5500

5300 - \ • 5100 • —!-_____

4900

1700

1 1 IIII 1 1 1 1 1

15 20 25 30 35 45 50 55 60 J 25 30 35 L

IME AFTER SOLDERING (DAYS) TIME AFTER CASTING (DAYS)

Fig. 9 — Room temperature aging of 50Sn-49Pb-1Sb alloy

334-s I OCTOBER 1976 of any of the alloys tested. Its DPH are shown by the hardness vs. aging chill cast at the lower temperature hardness also decreased significant time curves of Fig. 18. The faster cool (480 F, 249 C) did not exhibit initial ly (5 points) during aging. Softening ing rate produced higher hardnesses age-hardening which is characteris occurs at a rapid rate and the shear at all aging times. tic of this alloy. strength of the alloy does not appear The cooling rates specified in Fig. Aging curves based on the shear to stabilize during the testing period 18 represent average cooling rates strength of lap joints soldered at this of 60 d. The hardness, however, ap from the casting temperatures. Chill- temperature and cooled at a some pears to level out after about 30 d. casting from 480 F (249 C) pro what slower rate (see Fig. 2) show a The structure of this alloy is essen duced much lower hardnesses than very significant age-hardening during tially a fine-grained eutectic structure chill-casting from 600 F(316 C). The the first 3 d of aging at room tem that coarsens during aging as shown hardness aging curve for specimens perature. in Fig. 15. The dark lead-rich lamel lae of the eutectic grow in size during aging and tend to become spherical in shape. Structural changes oc curred throughout the 60-day aging period. The structure of the alloy did not appear stabilized at the end of the 60-day aging period.

50Sn-49.6Pb-0.35Ag-0.05Cu Alloy This alloy exhibited a very high shear strength immediately after sol dering but age-softened rapidly dur ing the first 10 d of aging at room tem perature as shown in Fig. 16. The hardness also drops off rapidly during the first 5 days of aging. Both the shear strength and hardness appear i Xt- to stabilize in about 10-20 d. 3 DAYS Structural changes occurring in this alloy during aging are shown in Fig. 17. The basic microstructure con sists of large a lead-rich crystals in an eutectic matrix (light). Small white '• •,, ' '' precipitate crystals are observed to form within the large a grains after about 4 h at room temperature. With 16 DAYS 16 DAYS further aging, the precipitate grows in size, accompanied by a general •:-. • . ( coarsening of the overall structure. The structure appears to stabilize af ter about 30 d. I I Effect of Thermal History on Aging : Characteristics of 50Sn-47Pb-3Sb "' * v_.l 60 OAYS The aging characteristics of the 30 DAYS 60 DAYS 30 DAYS 50Sn-47Pb-3Sb soft solder were Fig. 10 — Structural changes in alloy Fig. 11 — Structural changes in alloy found to be dependent upon the ther 50Sn-49Pb-1Sb during aging at room tem 50Sn-50Pb during aging at room temper mal history of the alloy. The effects of perature. X1000 (reduced 48% on repro ature. X1000 (reduced 48% on repro casting temperature and cooling rate duction) duction)

20 25 30 35 20 25 30 35 I 45 60 56 60

TIME AFTER SOLDERING (DAYS) TIME AFTER CASTING (DAYS) Fig. 12 — Room temperature aging ot 63Sn-37Pb alloy

WELDING RESEARCH SUPPLEMENT! 335-s Effect of Aging room temperature, except that the ag creasing in number as they became at Elevated Temperatures ing process at the elevated temper more spherical in shape. No precip Figure 19 shows the effects of ele ature is much faster. The shear itate formation was observed in the vated temperature storage on the ag strength vs. annealing time curve 50Sn-49Pb-1Sb alloy. ing characteristics of the 50Sn-47Pb- levels out after about 80 h at 200 F The eutectic structure of the widely 3Sb solder alloy. The specimens were (93 C) as does the hardness curve. used solder alloy 63Sn-37Pb coars placed in an oven immediately after Discussion ened considerable during the first 6 d soldering and aged at 200 F (93 C) after chill-casting and approached a Aging Characteristics for the desired time. The aging curves stable structure after about 30 d. The for specimens stored at 200 F (93 C) The results of this investigation period of the greatest structural have the same general shape as the indicate that tin-lead soft solder un change corresponded to the period of curves for samples stored at room dergo a prolonged age-softening pro greatest decrease in the mechanical temperature (Fig. 2). This indicates cess at room temperature after properties (shear strength and hard that the aging process in this alloy is soldering. This softening process is ness). This was true for all the alloys. the same at 200 F (93 C) as it is at accompanied by structural changes In the 50Sn-50Pb alloy, formation of a in the alloys. Tin-lead alloys contain second phase was observed within :....._,..:_ ing more than 1% antimony undergo the large crystals of lead-rich o solid an initial age-hardening process dur solution immediately after casting and ing the first three days after solder grew in size during aging. The second ing, followed by prolonged age-soft phase was not identified but was as ening. This was also characteristic of sumed to be ftf-tin. In the noneutectic the eutectic tin-lead alloy containing tin-lead alloy containing small addi Qy 1% antimony and 0.4% silver. How tions of copper and silver, a similar ever, in this alloy, the onset of the soft second phase was observed in the ening began after only 24 h at room dark n crystals. However, it did not temperature. The aging characteris become visible until after 4 h of aging. tics of these tin-lead solders were The aging curves for 50Sn-50Pb identified by changes in the shear and 50Sn-49Pb-1Sb alloys are very 1 strength and hardness of the alloys similar. The shear strength and hard mm®* with time at room temperature. ness values at the start and at the end Structural changes were observed (60 d) of the aging cycle for both in all the alloys during aging. Consid alloys are about the same. However, erable coarsening of the overall the softening process occurs more microstructure was invariably ob rapidly in the tin-lead alloy without the served. For the tin-lead-antimony antimony. Apparently, the antimony alloys containing more than 1% anti slows down and prolongs the soft mony, a white precipitate formed in ening process. The antimony ap the grain boundaries of the lead-rich pears to have an even greater effect (o) solid solution. This precipitate was on the microstructure. The structure analyzed and found to be a solid solu of the 50Sn-49Pb-1Sb alloy is much tion containing 98 wt-% tin and 2 wt-% finer grained than that of the 50Sn- antimony. The precipitate became 50Pb alloy, with no large crystals of microscopically visible at X1000 primary «-phase evident. In addition, magnification after about 3 d aging at no second phase formation was ob room temperature, which corres served in the crystals of the anti ponded with the end of age-hard mony-containing alloy. ening and the onset of age-softening The noneutectic tin-lead alloy con in these alloys. The precipitate grew taining 0.35% silver and 0.05% cop and became spherical with further ag per exhibited a very high shear ing at room temperature. The nucle strength immediately after soldering Fig. 13 — Structural changes in alloy ation of the precipitate appeared to but age-softened rapidly during the 63Sn-37Pb during aging at room temper be spontaneous with the precipitate first few days at room temperature. ature. X1000 (reduced 48% on repro particles increasing in size and de duction) During the first 10 days of aging, the

0 5 10 15 20 25 30 35 40 45 50 55 60

TIME AFTER CASTIHG (DAYS) Fig. 14 — Room temperature aging of 60Sn-38.6Pb-1.0Sb-0.4Ag alloy

336-s I OCTOBER 1976 shear strength decreased more than tics. Both showed initial age-hard tals in a eutectic matrix. The alloy is 1200 psi. The hardness decreased ening followed by extended soft much harder in this state than in its rapidly also, but the magnitude of the ening. It is interesting to note that the equilibrium state. However, the alloy decrease was not as great. The small shear strength values for both the tends to revert to its equilibrium state copper and silver additions appar 47Sn-47Pb-6Sb and 58.5Sn-37Pb- at room temperature by the precip ently have a profound strengthening 4.5Sb alloys are lower than compar itation of iti-tin from the supersat effect on the tin-lead alloy, but the ef able values for the 50Sn-47Pb-3Sb urated a solid solution. Precipitation fect is short-lived and unstable. The alloy. The reverse is true for the hard of tin from the a-phase softens the shear strength of the alloy after 25 ness values. One would expect the alloy. The same process occurs in sol days aging time approaches that of alloys containing the higher anti dering, whether it be a chill-cast the 50Sn-50Pb solder. There is some mony additions to have greater solder slug or an air-cooled lap joint. permanent strengthening (400 psi at strength and hardness due to the The cooling rate is fast enough to 60 days) but the effect is not as great presence of the intermetallic SnSb. retain a significant amount of super- as first indicated immediately after The initial shape of the shear soldering. strength aging curve (Fig. 12) for the There was little difference between 63Sn-37Pb eutectic alloy is open to Esn the microstructure of this alloy and question because of the wide scatter that of the pure tin-lead alloy, indi in the data points. In drawing the cating that the minute additions of curve to fit the data, it was assumed silver and copper were not very effec that shear strength of this alloy would tive in refining the grain structure. The follow the same general trend during tin-lead solder with the antimony and aging as does the hardness. silver additions (60Sn-38.6Pb-1.0Sb- 0.4Ag) exhibited aging characteris tics similar to those of the high anti Mechanism of Age Softening mony alloys (i.e., initial age-hard Tin-Lead Alloys. Binary tin-lead ening followed by prolonged age-soft alloys solution-harden when ening); however, the rate of softening quenched or fast-cooled from near or is more rapid. above their eutectic temperature due When compared to the 63Sn-37Pb to the retention of supersaturated eutectic solder, the 60Sn-38.6Pb- lead-rich solid solution («) at room 21 HOUR 1.0Sb-0.4Ag alloy actually softens to a temperature. At a temperature just lower shear strength value although below the eutectic temperature its final (60-day) hardness is greater. (361 F, or 182.8 C), a 50 wt-% tin and If compared to the 60Sn-37Pb-3Sb 50 wt-% lead alloy will consist of pri alloy, the maximum shear strengths mary cv crystals suspended in a eutec and hardnesses are about the same tic matrix (a+fS). but this alloy softens at a somewhat As shown in the tin-lead equilib faster rate. The results suggest that rium phase diagram of Fig. 20 the pri 0.4 wt-% silver and 1.0 wt-% anti mary (v-phase consists of 81% lead mony added in combination are about and 19% tin. As the temperature de equivalent to 3.0 wt-% antimony with creases, the solubility of tin in lead respect to the strengthening effect on decreases until, at room temperature, eutectic tin-lead solder. However, the the equilibrium composition of « is strengthening effect from the silver about 97% lead and 3% tin. and antimony is less stable as ev If the 50Sn-50Pb alloy is quenched idenced by the more rapid rate of (fast cooled) from just below the age-softening. eutectic temperature, the elevated The two solder alloys containing temperature «-phase (81Pb-19Sn) is Fig. 15 — Structural changes in alloy antimony additions in excess of the retained at room temperature. The 60Sn-38.6Pb-1.0Sb-0.4Ag during aging at 7% solubility limit of antimony in the alloy exists in a metastable condition room temperature. X1000 (reduced 49% exhibited similar aging characteris consisting of supersaturated « crys on reproduction)

6700

6500

6300

6 100

5900

5700 • \

5500

• ' •• • I I I i I i i [111 0 5 10 15 20 25 30 35 10 15 50 55 60 I5 20 25 30 35 40 15 50 55 60

TIME AFTER SOLDERING (DAYS) TIME AFTER CASTING (DAYS) Fig. 16 — Room temperature aging of 50Sn-49.6Pb-0.35Ag-0.05Cu alloy

WELDING RESEARCH SUPPLEMENT! 337-s saturated a-phase at room temper The curve of hardness vs. aging microstructure of eutectic tin-lead ature. During storage, the solder soft time for the 50Sn-50Pb solder alloy solder is unique and consists of a ens as it transforms to its equilibrium illustrates this softening process in finely dispersed mixture of lead-rich a state. tin-lead solders. The alloy has its (dark) and tin-rich 0 (light) solid solu maximum hardness immediately after tions. As seen in Fig. 13, the a crys "1 chill casting. During the first 10 d of tals are very small initially but coars storage, the hardness decreases sig ened during the aging process. No 0- nificantly. This decrease in hardness tin globules can be observed within is accompanied by significant struc the a crystals. The d-tin which precip tural changes in the alloy. Small white itates from the supersaturateda- globules of d-tin form within the large phase apparently is indistinguishable dark crystals of the primary a-phase. from the existing d-phase. i_±:_L__] The globules of d-tin appears imme According to Turnbell ef al (Refs. 6- diately after casting and reach a 9), j8-tin precipitates from the super stable size at about the time the hard saturated lead-rich (a) solid solution ness curve starts to level off. in two stages. The first stage involves At this point in time (6-10 d), the a discontinuous precipitation process 50Sn-50Pb alloy appears to be close which oroceeds at a rapid rate and to equilibrium. drains about 60% of the excess tin No further significant changes in from solution. Only 10-30 min are re hardness or microstructure are noted. quired a complete 50% of the pre The shear strength of the alloy be cipitation process. The remaining haves in a similar fashion and ap 40% comes out by a much slower pears to level out after about 20 d ag reaction (about 2 orders of magni ing time at room temperature. If the tude slower than the fast reaction). alloy is reheated to a temperature Precipitation occurs by the nucle close to the eutectic temperature ation and growth of cells. These cells (361 F, or 183 C) and quenched, the of transformed material (/3) nucleate original high hardness is restored and at the beginning of the precipitation the softening process starts over process mostly in the vicinity of grain again. boundaries and grow into the grains. The eutectic tin-lead alloy (63Sn- The reaction rate of the first stage of 37Pb) undergoes the same type of precip tation is obviously too fast to softening process after chill-casting account for the relatively slow and from the soldering temperature. The prolonged age-softening charac aging curves for this solder indicate teristics observed in the high tin-lead that the softening process takes about alloys used in the investigation. How 10-20 days at room temperature. The ever, the first stage (fast reaction) only 30 DAYS

Fig. 17 — Structural changes in alloy 50Sn-49.6Pb-0.35Ag-0.05Cu during aging at room temperature. X1000 (reduced 49% on reproduction)

• COOLED AT 50°F/SEC. FROM 600°F ® COOLED AT 30°F/SEC. FROM 600°F * COOLED AT 50°F/SEC. FROM l 20 40 60

ANNEALING TIME AI 200°F (HRS.)

50 55 60 20 40 60 80 IOO 120 I40 I50 180 200 220 240

TIME AFTER CASTING (DAYS) ANNEALING TIME AT 200°F (HRS. ) Fig. 18 — Effect of cooling rate and casting temperature on aging Fig. 19 (right) — Effect of annealing on aging characteristics of characteristics of 50Sn-47Pb-3Sb alloy " 50Sn-47Pb-3Sb alloy

338-s I OCTOBER 1976 accounts for 60% of the excess tin in the supersaturated a-phase. The much slower second stage reaction accounts for the remaining 40% of the transformation. LIQUID It appears that the gradual changes in structure and mechanical proper ties of the tin-lead alloys studied in this investigation are a ressult of the second stage precipitation of d-tin from the supersaturated lead-rich « solid solution. Livingston (Ref. 6) in his studies of the second stage process, noted a gradual spheroi- dization and coarsening of the d-tin precipitate during aging. A similar ef fect was observed in this investiga tion with the 50Sn-50Pb alloy. Effects of Antimony and Other Minor Alloying Elements. Age-soft 0 10 20 30 40 50 60 70 80 90 100 ening in tin-lead alloys containing LEAD - WEIGHT PERCENT • antimony occurs at a reduced rate. The antimony appears to slow down Fig. 20 — Tin-lead phase diagram the second stage precipitation of d-tin from the supersaturated lead-rich alloys, the softening effect over which a white precipitate is observed solid solution. The addition of anti comes the hardening effect very to form after 3 days aging at room mony to tin-lead alloys also has other quickly due to the fast transformation temperature. The presence of anti effects. The composition of d-phase rate. In the antimony-containing mony and silver combined in the precipitate is altered from essentially alloys, however, the slower transfor eutectic solder destroyed the fine pure tin to a solid solution of tin and mation rate and the increased num laminar structure and coarsened it antimony. ber of precipitate particles of smaller considerably. No precipitate was ob The precipitate in the 50Sn-47Pb- size causes the age-hardening effect served in the 50Sn-47Pb-3Sb alloy 3Sb alloy was identified as a solid to be predominant. during aging at room temperature. solution containing 98 wt-% tin and 2 During the first three days of aging, The structural changes observed dur wt-% antimony. The precipitate par a classical age-hardening process ing aging were similar to those ob ticles are smaller and greater in num takes place, with the precipitate parti served in the binary eutectic (63Sn- ber. In the alloys containing 3% or cles too small to be observed by opti 37Pb) alloy — a general coarsening of more antimony, the precipitate did cal microscopy. Hardening continues the overall structure. not become visible optically until 3 d as the precipitate grows until a crit Adding 1% antimony to a 50Sn- after chill-casting from 575 F ical particle size is reached. At this 50Pb solder has a permanent hard (302 C). Antimony also serves to point the alloy "overages" and the ening effect and acts to prolong the refine the grain structure of tin-lead softening process becomes predom age-softening process. It also serves alloys. The antimony-containing inant. The critical particle size ap to refine the grain structure, making alloys were much finer grained than pears to correspond to the size which the lead-rich a-phase crystals much the binary tin-lead alloys. can be resolved with the optical smaller. The alloys containing 3% or'more microscope. This process might be Effects of Thermal Treatment. The antimony age-harden during the first compared to the typical age-hard aging characteristics of the 50Sn- 3 d of aging at room temperature and ening process for copper-beryllium 47Pb-3Sb alloy were found to be then proceed to soften upon further alloys carried out at elevated temper dependent upon its prior thermal aging. This can be explained in terms atures. This characteristic is not seen treatment. Samples chill-cast from of a classical precipitation hardening in the pure tin-lead alloys or the tin- 600 F (316 C) at an average cooling process. When the tin-lead-antimony lead alloys containing small addi rate of 50 F/s (27.8 C/s) had higher alloy is fast-cooled from above its tions of antimony, copper, or silver hardness values throughout the aging melting temperature, it exists in a because the precipitation rate is too cycle than did samples chill-cast at a metastable condition at room tem fast and the number of nucleating slower cooling rate of about 30 F/s perature due to the presence of the sites too few to cause effective age- (16.7 C/s). The rate of softening for supersaturated lead-rich phase. The hardening to occur. The alloys "over the two samples was about the same, alloy tends to approach its equilib age" almost immediately. and both aging curves had the same rium condition at room temperature The presence of 1.0% antimony characteristic shape. The faster cool by precipitation of d-tin from the and 0.4 percent silver in combination ing rate produced a finer grained supersaturated lead-rich phase. produce an aging effect very similar microstructure. During the precipitation process, to the effect produced by 3% anti Samples chill-cast from a lower there are two opposing effects on the mony in eutectic solder alloys. Both temperature of 480 F (249 C) had mechanical properties of the alloy. alloys undergo initial age-hardening significantly lower hardness values at One is a softening effect due to the followed by prolonged softening. Both all aging times. The rate of softening drainage of d-tin out of the supersat alloys have about the same max was reduced, and there was no initial urated a-phase. The other effect is a imum shear strength and hardness age-hardening effect. The casting hardening effect due to the formation values, but the high-antimony alloy temperature, cooling rate, and aging of precipitate particles in the grain softens at a faster rate. The micro- temperature also had a great effect boundaries and along dislocations in structure of the 3% antimony alloy is a upon the resulting structure of the the crystal lattice. In pure tin-lead typical laminar pearlitic structure in alloy.

WELDING RESEARCH SUPPLEMENT! 339-s Higher casting temperatures and analysis to be a solid solution of 98 wt- became "over-aged" and started to faster cooling rates produce a higher % tin and 2 wt-% antimony. This sup soften. The onset of softening cor degree of hardening and finer grain ports the work reported by Eyre responded to the time at which the structures in this alloy. This is due to a (Ref. 10) who reported the solid sol precipitate first becomes visible with a greater degree of supersaturation ubility of tin in antimony at room light microscope at X1000 magnifica (more retained d-phase) of the lead- temperature to be in the order of tion. The addition of 1% antimony and rich a-phase during cooling. This 0.6 wt-% rather than 7% as reported 0.4% silver in combination to tin-lead causes greater lattice strain, thereby by Manko (Ref. 1). solder had a similar but less pro creating more nucleating sites for Although the antimony content nounced effect on its aging charac precipitation. The precipitate par (3%) in this alloy represented 6% of teristics. ticles are smaller and greater in num the tin content, no evidence of the The mechanical properties and mi ber, producing a greater hardening intermetallic SnSb could be found us crostructure of the 50Sn-47Pb-3Sb effect on the alloy. Slow cooling ing optical microscopy. alloy were found to be dependent causes less of the d-phase (tin) to be upon the thermal history of the sam retained at room temperature pro Conclusions ple. Higher casting temperatures and ducing a softer alloy. Tin-lead alloys which are fast- faster cooling rates result in higher The same effects would also be ex cooled from the melt harden due to strengths and hardnesses at all aging pected for the other tin-lead alloys in the formation of a supersaturated times. These conditions also pro the study. lead-rich solid solution. Upon subse duce a finer grain structure with an in Effects of Aging Temperature. quent storage at room temperature, creased number of precipitate par Samples of the 50Sn-47Pb-3Sb alloy these alloys tend to approach equi ticles. Slow cooling from the melt pro aged at 200 F (93 C) produced the librium by precipitation of tin out of duces a typical laminar eutectic struc same characteristic aging curves as» the supersaturated lead-rich phase. ture which is inherently softer and similar samples aged at room tem The precipitation process softens the more stable. Reheating and quench perature. However, the age-hard alloys, causing a reduction in shear ing after prolonged age-softening re ening and age-softening processes strength and hardness of up to 20%. stores the high original hardness and occurred at a much faster rate. Aging The softening process takes from 6 starts the age-softening process over or annealing at 200 F (93 C) causes to 60 d at room temperature, de agin. the alloy to develop its maximum pending upon the tin-lead alloy. Annealing at 200 F (93 C) after hardness and shear strength after Binary tin-lead alloys age-soften rap soldering was found to be effective in about 10 h aging time as compared to idly after casting, requiring less than stabilizing the structure and strength 3 d at room temperature (70 F, or 10 d to reach a stable strength and of the 50Sn-47PB-3Sb alloy. The 21.1 C). The higher aging temper hardness. The addition of antimony to mechanical properties stabilized after ature also increases the softening tin-lead solders slows down the pre 80 h at this temperature as com rate, causing the aging curves to level cipitation rate and prolongs the soft pared to 25-30 d at room tem out after about 80 h. This compares to ening process. Small additions (less perature. 25 d at room temperature. than 0.5 wt-%) of silver and copper in An interesting note is that the alloy crease the as-cast strength and hard References stabilizes to a somewhat higher shear ness but have little effect on the age- 1. Manko, H. H., Solders and Solder strength and hardness when aged at softening process. ing, McGraw-Hill, (1964). 2. Holmes, J. F., "Lead and Its Alloys," 200 F (93 C). This is due primarily to The decrease in shear strength and Metal Industry, Vol. 8, (December 1961). the small amount of solution hard hardness of tin-lead alloys during 3. Pell Walpole, W. T., "The Effect of ening produced when cooling the storage (aging) at room temperature Quenching and of Subsequent Prolonged samples to room temperature for is accompanied by noticeable struc Tempering at 100 C or 140 C on Alpha- testing. The solubility of tin in lead at tural changes. In all of the alloys Base Antimony-Cadmium-Tin Alloys: I. 200 F (93 C) is about 8% compared tested, a general coarsening of the Changes in Hardness," Journal of the Insti to 3% at room temperature. Some overall structure was observed. In the tute of Metals, (1942). excess tin in the a-phase is retained 50Sn-50Pb alloy, globules of d-tin 4. Baker, W. A., "The Creep Properties causing a small amount of hard were observed to form within the of Soft Solders and Soft-Soldered Joints," Journal of the Institute of Metals, Vol. 65, ening. The alloy could be expected to large a-crystals immediately after soften further with time at room tem (1939). casting. These globules grew during 5. Drefahl, K., et al, "Technological perature even after prolonged aging aging and became spherical in shape. Examinations of Antimony-Containing and at 200 F (93 C). In tin-lead alloys containing 3 wt-% or Antimony-Poor Lead-Tin Soft Solders The results of the elevated tem more antimony, a white precipitate According to DIN 1707 and Their Solder perature aging tests indicate that a was observed after 3 d aging time. Bonds," Metali, (August 1969) (German post-soldering anneal at 200 F The precipitate particles increased in Translation). (93 C) would be an effective method size and became spherical in shape 6. Turnbull, D., and Treaftus, H. N., of stabilizing the strength and struc with further aging. The precipitate "Micrographic Investigation of Precip itation in Pb-Sn Alloys," Trans. Am. Inst. of ture chill-cast 50Sn-47Pb-3Sb alloy. It was identified as a solid solution con Mining Met. Petrol. Eng., Vol. 212, (1958). is safe to assume that the same would taining 98 wt-% tin and 2 wt-% anti 7. Turnbull, D., and Treaftis, H. N., mony. A precipitate was not ob hold true for the other alloys used in "Kinetics of Precipitation of Tin From this investigation. Annealing the alloy served in the eutectic tin-lead solder Lead-Tin Solid Solutions," Acfa Met., at 200 F (93 C) speeds up the pre or the eutectic alloy containing 1% Vol. 3 (1955). cipitation process, causing the alloy to antimony and 0.4% silver. 8. Livingston, J. D., "Precipitation and stabilize after about 80 h compared to In the tin-lead alloys containing 3 Super Conductivity in Lead-Tin and Lead- Cadmium Alloys," Journal of Applied about 25 d at room temperature. wt-% or more antimony, the size and Physics, Vol. 34, #10, (1963). distribution of the precipitate par 9. De Sorbo, W., and Turnbull, D., Solid Solubility of Antimony ticles were such that age-hardening in Tin at Room Temperature "Kinetics of Precipitation in Small Lead- occurred during the first 3 d of aging. Tin Spheres," Acta Met., Vol. 4, (1956). The precipitate that formed during After about 3 d, the particle size ex 10. Eyre, B. L., "The Solid Solubility of aging of the 50Sn-47Pb-3Sb alloy was ceeded the "critical" size for effective Antimony in Tin," Journal of the Institute of identified by electron microprobe precipitation hardening and the alloys Metals, Vol. 88 (1959-60).

340-s I OCTOBER 1976