technical note: effect of rare earths on short-circuiting arc
TRANSCRIPT
Technical Note: Effect of Rare Earths on Short-Circuiting Arc Welding
BY B. POLLARD
ABSTRACT. To examine the effect of rare earth additions on arc stability and weld bead appearance, short-circuiting (S-C) arc welds made in rare-earth treated vanadium (V) -aluminum (Al) -nitrogen (N), i.e., VAN-80, steel and in mild steel were compared. Wi th the low currents (less than 120 amp) usually employed for S-C arc welding of light gage steel, welds in the V-Al-N steel were comparable to those in mild steel, both w i th respect to the amount of spatter adhering to the plate surface and the frequency of overlapping at the weld toes. Similarly, at higher currents, when the usual edge preparation for heavy gages was used, V-Al-N welds were equal in appearance to those in mild steel. Compared to mi|d steel bead-on-plate welds, deterioration of V-Al-N weld bead appearance was evident only at higher currents where more weld metal dilution occurs. Since the latter practice repre-
B. POLLARD is a Senior Metallurgist, Graham Research Laboratory, Jones & Laughlin Steel Corporation, Pittsburgh, Pa.
sents poor operating conditions, it is concluded that w i th correct welding procedures, the presence of rare earths does not affect arc stability or weld bead appearance.
In t roduct ion
Rare earth elements are added to VAN-80 steel, a hot-rolled vanadium (V) -aluminum (Al) -nitrogen (N) steel w i th 80,000 psi' yield strength, to provide sulfide shape control, thereby improving transverse bend formability and notch toughness.1 The major constituents of the rare earth addition are cerium and lanthanum. These elements are readily oxidiz-able so that, when welding in an oxidizing atmosphere such as carbon dioxide, the slag formed wi l l contain cerium and lanthanum oxides.
The arc burns from the region of lowest thermionic work function. If this is a globule of oxide which wanders around in an irregular manner on the surface of the weld pool, the arc wil l l ikewise move in an irregular fashion, w i th an adverse effect upon the weld bead appearance. Slags normally formed in C0 2 arc
welding are primarily composed of MnO and Si02 and wi l l probably have electron thermionic work functions in the range 3.5 to 4.0 electron volts (eV), compared to a value of 4.5 eV for iron.2
The work functions of the oxides Ce02 and La 2 0 3 are 2.3 eV and 2.5 eV, respectively,2 so that slag globules rich in cerium and lanthanum oxides may be expected to have a greater effect on arc stabil ity than the slags normally formed in C0 2 arc welding. This effect may be magnif ied by the n o n u n i f o r m distribution of cerium and lanthanum, since much of the addition is present in the steel as discrete sulfide or oxide particles. The fol lowing investigation was undertaken in order to determine the significance of this effect when joining the V-AI -N steel by short-circuiting (S-C) arc welding.
Exper imenta l Procedure
The compositions of the 0.190-in. thick V-Al-N used in the investigation and the mild steel used for control specimens are given in Table 1. Weld-
Table 1 — Chemical Composition of Steel Test Specimens, %
Steel C Mn Si S P
V-Al-N .14 1.26 .45 .009 .013
Mild steel .21 .31 .056 .025 <.002
Al
.06
<.005
Cr
.04
.026
Ni
.03
.076
Cu
.04
.14
V
.12
< 0 0 2
Total N2
.021
.005
M D . N2
.014
N.A.(al
Ce
.013
N.A.
La
.010
N.A.
(a) N.A. = not analyzed.
W E L D I N G R E S E A R C H S U P P L E M E N T ! 533 -s
•i"„ :j& ^iMM^HktAMlAi . <m. Jk%jt^
[•gm*mgi'7¥i7Ma
;.
Q.-JSis:;:;* •"i * ** v : |
Fig. 1 — Weld beads made with a welding current of 120 amp: (5) mild steel with mill scale on; (6) V-Al-N with mill scale on; (7) mild steel with mill scale removed. Two-thirds actual size (reduced 25%)
Fig. 2 — Weld beads made with a welding current of 1 70-175 amp: (1) mild steel bead on-plate specimen with mill scale on; (2) V-Al-N bead-on-plate specimen with mill scale on; (4/ V-AI-N 60 deg vee butt weld. Two thirds actual size (reduced 25%)
Table 2 — Welding Conditions
Weld no.
5 la]
6 la)
7 (a)
1 1 8 1
2 l n .
4 la) gl..)
1 0 ta )
1 6 (b l
11<W 1 5 ' "
Steel
Mi ld V-Al-N Mild Mild V-Al-N V-Al-N Mild V-Al-N Mi ld V-Al-N V-Al-N
Mi l l scale
On On Off On On On Off Off Off Off Off
Current, amp
120 120 120 175 175 175 175 175 175 160 212
Voltage, V
19 19 19 21 21 21 21 21 21 22V2
20
Welding
speed, ipm
15 15 15 15 15 10 15 15 15 15 15
Electrode feed. ipm
200 200 200 300 300 300 300 300 300 300 300
Type of weld
Bead-on-plate Bead-on-plate Bead-on-plate Bead-on-plate Bead-on-plate 60°-V butt jo int Bead-on-plate Bead-on-plate Bead-on-plate Bead-on-plate Bead-on-plate
ia' Welded with .035 in. diam electrode and 100% CO: shielding gas. 1") Welded with .045 in. diam electrode and 100% CO , shielding gas. Ic: Welded with .045 in. diam electrode and 75% argon — 25% CO. shielding gas.
ing conditions are given in Table 2. Bead-on-plate specimens and one
60 deg vee butt weld were made wi th E70S-1B fi l ler metal and 100% C0 2 or 75% argon—25% C 0 2 shielding gas. Welds were made on plates both wi th a hot rolled surface and after removing the mill scale by shot blasting and grinding. The effects of electrode diameter and welding current on weld bead appearance were examined.
Results The appearance of VAN-80 welds
made using a 0.035 in. diam electrode, a welding current of 120 amp and w i th the mill scale intact was superior to mild steel welds, both wi th respect to the amount of spatter adhering to the plate surface and the frequency of overlapping at the weld toes [compare Figs. 1 (5) and 1 (6)]. The poor appearance of the weld bead on the mild steel plate may be attributed
to the nature of the heavier mil l scale, since when the scale was removed the appearance of the weld bead was similar to that on the V-AI-N plate [Fig. 1(7)1.
When the welding current was increased to 175 amp, the appearance of weld beads on plates of both mild steel and V-Al-N was worse [Figs. 2(1) and 2 (2)] although still suitable for many applications. By contrast, a 60 deg vee butt weld in the V-Al-N steel
534-s | N O V E M B E R 1 9 7 2
Fig. 3 — Weld beads made with a welding current of 175 amp; (9) mild steel with mill scale removed; (10) V-Al-N with mill scale removed. Two-thirds actual size (reduced 25%)
had an excellent appearance [Fig. 2(4)]. Removing the mill scale improved the appearance of the mild steel weld bead [Fig. 3(9)] but had negligible effect upon the appearance of the V-Al-N weld [Fig. 3(10)].
Increasing the fil ler metal diameter to 0.045 in. and using a welding current of 160—175 amp caused a slight deterioration in the appearance of beads on scale-free mild steel plates [Fig. 4(16)] and a considerable deterioration in the appearance of beads on scale-free V-Al-N plates [Fig. 4(11)]. The use of 75% argon — 25% C0 2 shielding gas in place of 100% C02 gave no improvement in weld bead appearance [Fig. 4(15)].
:r 7
.::.. v
llPi i§MI lilfl
Wmsmm mmmw77-- ••
Fig 4 — Weld beads made with 0.045 in. diam electrode; (16) mild steel, mill scale removed, welded with 100% C02 shielding at 175 amp; (11) V-Al-N mill scale removed, welded with 100% CO2 shielding at 160 amp; (15) V-Al-N mill scale removed, welded with 75% argon-25% C02 shielding at 212 amp. Two thirds actual size (reduced 25%)
Discussion and Conclusions The results of this investigation
show that, w i th welding conditions typical of those used for S-C arc welding of light gage material (0.035 in. diam electrodes wi th currents of less than 120 amp), welds in rare earth treated V-Al-N steel were equal to those in mild steel, w i th respect to the amount of spatter adhering to the plate surface and the frequency of overlapping at the weld toes. Similarly, when using higher currents and the usual edge preparation for heavy gages, V-Al-N welds were similar to those in mild steel.
Only w i th high current bead-on-plate welds, where more weld metal dilution occurs, was a poor weld bead appearance produced. In practice, however, high weld metal dilution is only obtained for square butt joints wi th no root opening. Since this type of joint is only practical for light gages, for which low welding currents and small diameter electrodes are normal practice, it is concluded that rare earths have no detrimental effects on weld bead appearance when welding under conditions that are typical of S-C arc welding.
References
1. Luyckx, L., et al, "Sulfide Shape Control in High Strength Low Alloy Steel," Met. Trans. AIME, Vol. 1, pp. 3341-50 1970.
2. Smithells, C. J., "Metals Reference Book," Vol. Ill, 4th Edition, New York, Plenum Press, 1967.
W E L D I N G R E S E A R C H S U P P L E M E N T ! 535 -s
Intermetallic Growth in Tin-Rich Solders
Study affords better understanding of the performance of tin-base
soldered joints exposed to temperatures of 100 to 200 C
BY LOUIS ZAKRAYSEK
Fig. 1 — Joint fracture that impairs conductivity of an electrical connection, as seen by scanning electron microscope (SEM) at X140 (reduced 50%)
ABSTRACT. For t in-r ich solder alloys, 200 C (392 F) is an extreme temperature. Intermetall ic growth in tin-copper systems is known to occur and is believed to bear a direct relationship to failure mechanisms. This study of morphological changes w i th t ime at elevated temperatures was made to determine growth rates of tin-copper intermetall ics. Preferred growth directions, rates of thickening, and notable changes in morphology were observed.
Each of four tin-base alloys was f lowed on copper and exposed to temperatures between 100 C and 200 C for t ime periods of up to 32 days. Metallographic sections were taken and the intermetall ics were examined. Intermetall ic layer thickening is characterized by several dist inct stages. The initial growth of side plates is extremely rapid and exaggerated. This is fol lowed by retrogression (spheroidization) of the elongated peaks and by general thick-
L. ZAKRAYSEK is Manager, Metallurgy and Welding, Electronics Laboratory. General Electric Co., Syracuse, N.Y.
ening. The dual-composition intermetallic thickens from 1 to 70 microns wi th some tendency to grow through voids in the solder matrix. The possibility was noted that mechanical properties can be enhanced by the preferred growth of the copper-rich phase. The general theory for the diffusion-controlled growth of plates appears to be applicable to this metallic system.
In t roduct ion
The successful use of solder alloys requires (1) getting the right kind of solder joint made, and (2) assuring the soldered joint wi l l wi thstand environmental stresses and strains. Each of these requirements is related to the materials being joined, to the fil ler metal alloy used, to the soldering procedure and to the environment encountered.
During the past ten years, much work has been done on improving soldering techniques. NASA's pioneering effort on contour soldering for highly reliable joints is probably the impetus behind most of the electronics industry's move in that direction.1 As a result, a great deal of information regarding a supposedly
well-established process has just recently come to light. More people now have a better understanding of how the soldering operation should be done.
During the same t ime period, the need for this understanding was re-emphasized by the trend toward automation in soldering. When large numbers of connections are attempted simultaneously, it is imperative that conditions for soldering be closely controlled,2 otherwise an intolerable proportion of joints wi l l be found unacceptable. Much industry effort is still devoted to this aspect of the automated soldering process.3
On the other hand, investigations into the useful life of wel l -made soldered joints, have been given sporadic attention. Detailed analyses are often made only after the inadvertent misapplication of a soldered component results in failure. Now the demand is for more reliability and longer life under more severe service conditions, including higher allowable temperature limits.
As a result, the type of failure shown in Fig. 1 is found to occur more and more often. A seemingly well made, and an apparently high quality connection becomes electr ically intermittent or open due to joint
536-s N O V E M B E R 1 9 7 2
fracture. Metallurgical analysis of a number of such examples indicates that microstructural changes wi th t ime at temperature are, in large part, controll ing factors in determining the useful l ifetime of soldered connections.4 Therefore, it seems advisable to learn more about the effect of the long-time exposure of common solder systems to high ambient temperatures.5
Despite the preponderance of technical information gathered on the subject of the elevated-temperature use of materials, much of the available data6 pertains to the use of materials at temperatures in excess of 350 C. Few investigators, w i th some notable exceptions,7*8 consider anything lower to fall w i th in the realm of a "h igh temperature" application. Yet, those who deal w i th t in-rich solder alloys f ind that, for them, 200 C is an extreme temperature.
Some of the frequently encountered problems, even though manifested by an extreme operating environment originate in the soldering process.9 Others are due to the effect of the environment itself. Some of the latter are considered in this paper.
200
150
T/C
100'
% K 1 Hours
2 4 Sn
8
200"
Experimental Procedure
On the intuitive assumption that problems of joint failure bear a direct relationship to the formation of intermetallic compounds, a series of experiments was designed for the evaluation of t in-r ich solder alloys and copper wett ing surfaces. An objective was to study microstructural changes after the prolonged exposure of soldered joints to temperatures as high as 200 C. As usual, the arbitrary l imitation and selection of test parameters directs attention pointedly toward those desirable experiments which are left undone.
150
r, °c
100
8 16 32 Days Sn
Fig. 2 — Tin solder morphology after exposure at temperatures shown for (a) VA to 8 hr, (b) 1 to 32 days. Joint cross section (X800, reduced 46%>)
Soldering and Sampling
There is a need, in work of this nature, to avoid becoming engrossed in the problems of soldering and solderability since these are arts and sciences wi th in themselves. Of all the factors which combine to facil itate the making of a soldered connection, it seemed most important, for the purpose of this experiment, to control those related to intermetall ic formation. Therefore, soldering temperatures were kept as low as possible, the volume of solder was kept nearly the same for each test, and the solder and terminal compositions were selected for high potential intermetallic content.
All samples were chemically prepared prior to the making of lap
joints, and the faying surfaces were tough-pitch copper. A number of soldered samples were soaked at temperatures between 100 C and 200 C for periods of t ime as short as 15 minutes and as long as 32 days. A representative specimen for each solder composition and each set of t ime-temperature condit ions w a s prepared for metallographic study, and another was used for an evaluation of mechanical properties.
Morphology
Through the use of the metallographic sections, changes in the morphology of the CuSn intermetall ic phases were observed. The extent of growth and growth rates were deter
mined for intermetall ic thickening. Other growth characteristics which indicate extreme complexity in these metallic systems were observed and studied. The bond interfaces and adjoining regions in the solder joints showed metallographic fea tu res characteristic of the CuSn metallic systems. The general morphology of the solder area was traced from that present in early (as-soldered) samples to that resulting from 32-day exposure. Selected samples were prepared for further analysis by scanning electron microscope (SEM) and x-ray probe methods.
Composition
In order to evaluate the effect of
W E L D I N G R E S E A R C H S U P P L E M E N T ! 537-s
2001
Fig. 3 — Tin-silver solder morphology after exposure (a) and (b) to the same conditions shown in Fig. 2. Joint cross section (X800. reduced 46%)
150 \:m •
! & • ' • • •
2001
% Hours
2 4 $n-5Ag
8
composition on thermal response, four solders were selected for study. At the start of this project, it seemed important that differences in component solubilities and in melt ing temperatures, among other things, be monitored. For this reason Sn, Sn-2.5Sb, Sn-5Ag and Sn-40Pb solders were prepared for evaluation. Some of the data10 pertinent to the objectives of the study are shown in Table 1. The major components (Cu and Sn) of each of these combinations are, for practical purposes, not soluble in one another. Tin is soluble in Sb, Ag and Pb to an appreciable extent, and in the SnPb alloy there is available a quantity of solvent sufficient to influence CuSn intermetallic growth rates if such an influence is to be of any importance.
150
IX
100
8 16 Sn-5Ag
32
Table 1
Solvent
Cu Cu Cu Cu
Sn Sn Sn Sn
Sb Sb
Ag Ag
Pb Pb
(a) RT = i (b) 5 0 % (c) neql
— Physical Properties foi
Solute
Sn Sb Ag Pb
Cu Sb Ag Pb
Cu Sn
Cu Sn
Cu Sn
oom temperature or more solvent = neglia>ble
Wt % in
200 C
1.3 1.0
neg l l c )
negl
negl 5 0
negl 1.0
negl 4.5
negl 10,0
negl 1 7 0
Var ious Cu
solution at
R T ) a 1
neg l l c l
negl negl negl
negl negl negl negl
negl 4.5
negl 8.0
negl 2 0
and Sn Alloy Sys tems
Intermetal l ic phases present at
200 C
Cu3Sn Cu 3Sb
C u r S n 5
SnSb Ag 3 Sn
Cu,Sb
Ag 3 Sn
—
R T ( a )
Cu3Sn Cu3Sb
Cu 6 Sn s
SnSb A g , S n
C u 2 S.
A g . S n
Lowest liquid
r 1 temp., c
415 526 779 326
?27 232 221 183
526 425
779 724
326 183
Exper imenta l Results
Copper-tin intermetall ic compound growth can be readily monitored in soldered joints. Actual growth modes are observed and measurements made by the use of metallographic techniques. Changes in the distr i bution of metallurgical phases as wel l as of chemical elements are plotted by combining metallography w i th scanning electron microscopy and x-ray fluorescence.
Intermetallic Growth
Sections were prepared by the use of standard metallographic techniques. Al l samples were etched using a 10-sec swab w i th 60% H 2 0 2 -40% NH4OH fol lowed by 10-sec swab w i th 2% H N 0 3 - 9 8 % alcohol. These solutions were developed for the purpose of increasing the contrast of the intermetall ics. In this way, morphological changes were monitored for all test conditions, and for each of the alloys selected for study.
Shown in Fig. 2 is a series of metallographic sections depicting the thermal response due to the Sn solder. A two-phase intermetall ic layer can be seen to form, to grow progressively thicker and eventually to bridge the joint cross section. The same series at temperatures lower than or higher than 1 50 C show s imilar morphologies w i th expected differences that can be explained by changes in growth rates which exhibit the normal exponential dependence on temperature.
538-s I N O V E M B E R 1 9 7 2
Fig. 4— Tin-antimony solder morphology after exposure (a) and (b) to the same conditions shown in Fig. 2. Joint cross section (X800. reduced 46%,)
2001
Figure 3 is a metallographic series showing the thermal response in Sn-5Ag solder. As is true of each of these materials, after a day of elevated temperature exposure, the intermetall ic layer is wel l - formed, continuous and generally free of angular side plates. The growth pattern is similar to that in Sn solder, except that complete bridging is delayed somewhat.
The metallographic series shown in Fig. 4 is that for Sn-2.5Sb solder. This system is characterized by slightly more irregularity in the intermetallic layer, otherwise the morphology for this alloy is not too different from that of the others. The two-phase intermetall ic grows into the solder matrix w i th occasional bridging after long time at the lower temperatures and complete bridging at 200 C.
Shown in Fig. 5 is the series of metallographic sections for Sn-40Pb solder. The intermetall ic layer is more rounded indicating that the spheroidization of side plates occurs rather early. Thickening also tends to proceed more rapidly, and the bridging which does not occur readily w i l l probably never be continuous.
In attempting the determination of growth rates, the investigator is presented wi th the problem of deciding which of a variety of morphological features are significant as wel l as w i th some special measurement difficulties. There seems to be little doubt that the growth phenomena encountered here can be described by established diffusion theory. Purdy and Kirkaldy11 show the mathematical procedures generally applicable to problems of this type. The work described here w i l l , hopefully, define those parameters that deserve more rigorous experimental work and mathematical treatment.
The measurement of allotrio-morphs or side-plates is probably not significant at exposures over 100 C because all traces of these morphological features are obliterated w i th in several hours. Other possible measurements include an average layer thickness, Cu3Sn thickness, Cu6Snb
thickness, change in joint gap, and decrease in solder matrix width.
Although the irregular surface at the intermetall ic-matrix interface makes layer thickness rather difficult to measure, the results of such measurements are listed in Table 2,
150
IX
100 wm*
% % 1 Hours
2 4 Sn-2.5Sb
8
: * ::* *•: t :
200-r
150
JrX
100 .-.'•'If •••;;> . , • •• p^ifWSf. . •. 7 ' ' : : • ' . • • ' •••:'::-:™:'fe:.--:: 77i~77' U**'C.;:*":•' '•:
Table 2 —Observed Intermetallic Layer Thickness, Microns Temper Time at temperature
ature, C
200 200 200 200
175 175 175 175
150 150 150 150
125 125 125 125
100 100 100 100
Hours Days Compositon
Sn-40Pb Sn-2.5Sb Sn-5Ag Sn
Sn-40Pb Sn-2.5Sb Sn-5Ag Sn
Sn-40Pb Sn-2 5Sb Sn-5Ag Sn
Sn 40Pb Sn-2.5Sb Sn-5Ag Sn
Sn-40Pb Sn-2 5Sb Sn-5Ag Sn
Vt v? i
3.8 4.5 5.3 1.4 1.6 2.1 1.5 17 2.4 17 14 2 4
5.3 2.4 2 6 2.7
3.1 2.9 3.1 3 6
7.6 3 6 4.1 4.3
0 9 1.0 1.5 1.8 2.6 3.5 1.1 1.5 1.7 17 1.8 2 5 1.1 1.1 12 1.5 2.0 22 2.1 1.4 1.3 19 1.7 2.8
1.0 1.1 1.4 1.4 1.2 1.5 2.1 2 4 1 5 1.0 2.7 2.1
1.2 16 0.5 1.6
1.2 16 0.8 1.6
1 9 2 0 1.8 2.4
1
7 8 8.1 5.2 62
9 7 4 3 3 9 6.8
16
12.5 12 0 18.7 8 5
9 9 5 8 4 6 7.0
14.4 12.5 18 1 148
10.1 6 8 5.6 7 3
1.1 1.0 1.0 1.1 2.0 1.0 1.5 1.4 2.1 1.9 2.4 1.5 1.1 18 1.2 2 9 1 4 1.2 19 19 2.2 2.4 2 0 2.1
All measurements about 1.0 micron or less
32
15 6 18 1 17 5 31 3 37 5 68 8 28 8 14 1 50 0
6 2 31 3 34 4
15 0 16 7 16 6 9 2 153 133 8 0 182 137 9 7 9 9 100
1.5 1 4 1.3 1.0
1.9 1 9 1.6 1.5
2 3 2 3 2 4 2.0
4 1 3.4 4.5 3.0
4 1 3.8 4 8 3 6
6 5 4 1 5 0 4 3
W E L D I N G R E S E A R C H S U P P L E M E N T ! 539 -s
200=••* *%*•'*.*>' • • • * * • * <&-***%
150 *• • > . . . m - ' •*" &.-.nu*4 *g& »- .\ JM»X~-W» Sl l
20
Sn
CuSn INTER METALLIC
(MICRONS)
/
• /
1
/ / JDO /
/ /
/
^-'^X-
/
^Pi
100!
F/'Si. 6 — Growth in thickness of CuSn intermetallic layer with time at temperatures shown, using tin solder on copper
% % 1 Hours
2 4 Sn-40Pb
8
200
150
Fig. 7 — Growth in thickness of CuSn intermetallic layer with time at temperatures shown, using tin-silver solder on copper
rfx
TOO
b ».:fc •
w - 4 , • > " • £ * - * * •
2 4 Days
8 16 Sn-40Pb
32
Fig. 5— Tin-lead solder morphology after exposure (a) and (b) to the same conditions shown in Fig. 2. Joint cross section (X800, reduced 46%,)
T h e s e d a t a a re s h o w n g r a p h i c a l l y i n F igs. 6, 7, 8 a n d 9.
T h e r e a p p e a r s t o be l i t t le d e p e n d ence of g r o w t h ra tes o n c o m p o s i t i o n . T h e d o m i n a t i n g fac to r is t h e m a x i m u m exposu re t e m p e r a t u r e .
Morpho logy
D u r i n g t h e i n i t i a l s t a g e s of so l de r j o i n t f o r m a t i o n , t h e m i c r o s t r u c t u r e is c h a r a c t e r i z e d by t h e p r e s e n c e o f a l l o -t r i o m o r p h s w h i c h are f o r m e d o n s o l i d i f i c a t i o n . T h e r e is no w a y to avo id t h e p r e s e n c e of t h i s t h i n ( PS 1 -2 m i c r o n ) layer of i n t e r m e t a l l i c d u e to n u c l e a t i o n f r o m t h e m e l t . I m m e d i a te ly a f te r s o l i d i f i c a t i o n a n d upon e l eva ted t e m p e r a t u r e e x p o s u r e , s i de p la te g r o w t h is e x t r e m e l y r a p i d . N e x t
5 4 0 - s I N O V E M B E R 1 9 7 2
t h e s ide p la tes t e n d t o s p h e r o i d i z e w h i l e t h e i n t e r m e t a l l i c layer t h i c k ens . In la ter s tages of g r o w t h , t h e t h i c k layer b e c o m e s a c o n t i n u u m of m a s s i v e i n t e r m e t a l l i c w i t h a r o u n d e d l ead ing edge. F ina l ly , t h e i n t e r m e t a l l ic t e n d s to b r idge t h e e n t i r e j o i n t gap , a l w a y s w i t h o u t a p a r t i n g l i ne w h e r e t h e i n t e r m e t a l l i c l aye rs e v e n tua l l y mee t . T h r o u g h o u t t h e e a r l y s tages of t h e g r o w t h p rocess , t h e C u 3 S n a n d t h e C u 6 S n 5 t h i c k n e s s e s are a p p r o x i m a t e l y e q u a l . A s t i n is de p le ted f r o m t h e so lde r m a t r i x , C u 6 S n 5 g r o w t h s l o w s w h i l e t h e C u 3 S n c o n t i n u e s t o g r o w a t t h e ex pense of t h e t i n - r i c h i n t e r m e t a l l i c . G i v e n s u f f i c i e n t t i m e at t e m p e r a t u r e , t h e c o p p e r - r i c h p h a s e c o n s u m e s a l l of t he ava i l ab l e t i n .
Sn-2.5Sb
CuSn IMTERMfTAIlIC
(MICRONS.
i
1 zffc
A / .
^ ^ jt—-••'Ti i D ° ° c ,
/
Fig. 8 — Growth in thickness of CuSn intermetallic layer with time at temperatures shown, using tin-antimony solder on copper
Sn-40Pb
CuSn INTER METALLIC
[MICRONS:
* .
* , ^
?00°C , ' 1 1 si
< •
<
y
/ l s c f c
o ^ ^ -
— — - ~ " | 100° C
1
.°
Fig. 9 — Growth in thickness of CuSn intermetallic layer with time at temperatures shown, using tin-lead solder on copper
* 8 « * 2 I I ( T yS:1898 HS: 58EV/CH
¥ S : i « « 9 HS: 5BEV;CH
10 Osy ISO'C
Jl Srt-40Pb
Fig. 10 — X-ray probe traces showing concentrations of five elements in three regions of the cross section of a joint made with tin-lead solder on copper. View by SEM taken after 16 days at 150 C (etchant given in text)
Composition Gradients
In a newly formed solder joint, the cross section consists of a th in Cu 6Sn 5 intermetall ic layer at both copper interfaces and a solder matrix of fairly uniform composit ion between. SEM and x-ray probe analyses indicate that some segregation occurs in the solder matrix w i th pr i mary tin as wel l as each intermetallic type dispersed throughout the joint area.
As intermetall ic g row th p ro gresses, the x-ray probe reveals that the Cu3Sn and Cu 6 Sn 5 intermetall ics grow into the solder matrix and, at a slower rate, into the copper terminal. Insoluble and impurity elements are concentrated ahead of the leading edge of the Cu 6 Sn 5 layer until growth is completed by the depletion of the available Sn. Diffusion continues until the intermetall ic conversion to Cu3Sn is ended. In the Sn-Ag and Sn-Sb systems, a complex ternary intermetall ic is the end result. In the Sn-Pb system, a Pb-rich zone remains between the intermetall ic faces.
Figure 10 shows the SEM and x-ray probe trace taken from the Sn-40Pb sample after 16 days exposure. These results indicate both the tendency toward equil ibrium in the intermetallic and the concentration of insolubles in the matrix.
Conclusions Copper-tin intermetall ic compound
growth occurs in soldered joints at low temperatures and in relatively short t imes. The growth that takes place can be significant in terms of joint characteristics.
The presence of insoluble alloying elements affects the intermetall ic growth rate to a minor degree. The t in content and the maximum exposure temperature have the greatest influence on growth rates.
Intermetall ic layer thickening is characterized by several distinct stages. The initial growth of side plates is extremely rapid and exaggerated. This is fol lowed by retrogression (spheroidization) of the elongated peaks and by general thickening. The intermetall ic thickens f rom 1 to 70 microns w i th some tendency to grow through voids in the solder matrix. In the high-t in alloys the Cu6Sn5 readily bridges the joint sect ion. In the f inal stages of growth, the dual intermetall ic is transformed to Cu3Sn by total consumption of the solder matrix and of the Cu 6 Sn 5
phase.
The possibility was noted that mechanical properties can be enhanced by the preferred growth of the copper-rich phase. The general theory for the diffusion-control led growth of plates appears to be applicable to this metallic system.
Acknowledgment
This project was sponsored by several departments of the General Electric Company. Sample preparation and metallography are due to J. E. Richardson, M.D. Distin and D. B. Blackwood. The SEM work was performed by J. A. DeVore.
References
1. NASA Qua! Publ. NCP 200-4, Aug. 1964.
2. Thwaites, C. W., Tin Research Institute, Publ. No. 382, 1968.
3. Bud, P., Handbook on Joining Techniques, Institute of Printed Circuits, 1972
4. Zakraysek, L, We/ding Journal, Research Suppl., Vol. 36, No. 12, Dec. 1971, P. 522-S.
5. Howes, M. A. H. and Saperstein, Z. P., Welding Journal, Research Suppl., Vol. 48, No. 2, Feb. 1969, p. 80-S.
6. Yen, T. C, Welding Research Council, Bulletin 72, Oct. 1962.
7. Lewis, W. R., Tin Research Institute, Publ. No. 42, Spring 1958.
8. Willhelm, A. C. and Hamilton, J. A., Southern Research Institute, MSFC Report TP 85-207 CPB 02-1064-61, Mar. 1962.
9. Beal, R. E., CDA Technical Report 804/9, Aug. 1969.
10. Hansen, M., Constitution of Binary Alloys. McGraw-Hill, New York 1958.
11. Purdy, G. R. and Kirkaldy, J. S„ Trans. TMS-ASM Vol. 2, No. 2, 1972, p. 371.
W E L D I N G R E S E A R C H S U P P L E M E N T ! 541-s