8/2/2019 Whisker Free Tin Jan Feb White Paper Download
1/6
The Elimination of Whiskersfrom Electroplated Tin
As the implementation of RoHScontinues to affect the indus-try, tin and tin alloys remain the firstchoice for replacing tin/lead. Lead-free soldering and LF solder, arepresently well implemented in theindustry. There is a good under-
standing of LF solders like: the fam-ily of SAC alloys and the tin/copperfor Lead-free hot air solder leveling(HASL). The understanding extendsto the type of intermetallic com-pound (IMC) formed, its propaga-tion, and the integrity and reliabilityof the solder joint that is formed.There is a continuous effort to comeup with even better products thatmay lower the reflow temperature orreduce the IMC propagation forgreater solder joint reliability and
for a wider assembly window. Someof these efforts involve smallamounts of dopants to presentlyused LF alloys.
On the surface finish side, replac-ing tin-lead has posed greater chal-lenges. Component leads and con-nector finishes were being convertedto tin as an obvious alternative. Tin iseasy to apply, is readily solderableand economical to use. Tin workswell as a soldering surface; howeverany part of the lead or the connec-tion surface that is not soldered to
has shown a propensity to form tinwhiskers over the life of the part.Internal stresses in the deposit, cou-pled with IMC formation along grainboundaries as well as external stress-es on the deposit, are known to initi-ate whisker formation.
A lot of time and effort has beendirected at tin whisker mitigation.There are a series of methods todetermine the propensity of a tin fin-
ish to whisker, as well as recommen-dations on how to evaluate the same.This article describes successfulefforts on how to eliminate whiskerformation.
This article describes the two
approaches that were successful ineliminating whisker formation. Bothapproaches dissipate the stress thatis formed from the interaction ofIMC formation and the inherentstructure of electroplated tin. The
first is to modify the substrate sur-face to control the growth in thick-ness and direction of propagation ofthe IMC, and the second is to modifythe large columnar tin deposit crys-tal structure to mimic the fineequiaxed structure of tin-lead solder.Controlling the IMC thickness wasachieved by micro-roughening thecopper substrate before tin deposi-tion. The modification of the crystalstructure was accomplished by theuse of specific organic additives that
TECHNICALLYspeaking
BY GEORGE MILAD, NATIONAL ACCOUNTS MANAGER FOR TECHNOLOGY,
UYEMURA USA, SOUTHINGTON, CONN.
Figure 2. Four different paths that lead to stress and whisker formation.
Figure 1. A schematic illustrating tin whisker formation.
10 I metalfinishing I January/February 2012 www.metalfinishing.com
Whisker free tin Jan Feb-WHITE PAPER DOWNLOAD:Metal Tech Speaking Template USE THIS.qxd 3/8/2012 10:30 AM Page 1
8/2/2019 Whisker Free Tin Jan Feb White Paper Download
2/6
www.metalfinishing.com January/February 2012 Imetalfinishing I 11
disrupt the columnar growth andgive rise to smaller equiaxed crystals.
Electroplated pure tin and tin-based alloys are being used as alter-natives to tin-lead in the majority ofelectronic components. These alter-natives are known to produce tinwhiskers, which may give rise toshort circuits on these components.
In the case of tin finish on copperand copper-based alloys, the majorcause of tin whisker formation is com-
pressive stress. The stress is mainlycaused by irregular growth of copper-tin IMC at ambient conditions [1].
It is known that tin whiskers arereadily formed on electroplated tindeposits on copper and are notobserved on electroplated tin-leaddeposits. The tin deposit and tin-lead deposit are different in thecrystal structure. Crystal structurehas a direct impact on tin whiskersformation [2] [3].
A tin deposit with modified crystal
structure (similar to tin-leaddeposits) is capable of preventingwhisker formation by dissipatingand delocalizing the stress that causewhiskers.
As shown in Figure 1, stress, chan-neled along the boundaries of thelarge grained columnar tin depositis responsible for the emergence oftin whiskers. Stress may be internalor external (see Figure 2). The pri-mary source of internal stress isattributed to the non-uniformincrease in the thickness of the IMC
layer over time at ambient condi-tions (30oC, 60%RH for 4000hours). Another condition that pro-duces internal stress is exposure tohigh temperature and high humidi-ty (55oC, 85%RH 4,000 hours) forextended periods of time whichgives rise to oxidation and/or corro-sion. Internal stress could also beinduced by thermal cycling (-55oCto 85oC 1,500 cycles) due to mis-
TECHNICALLYspeaking
Figure 3. Test Vehicle
Figure 4. SEM micrographs of different Ra values.
Figure 5. Illustration of the process used for this study (typical plating sequence).
0
20
40
60
80
100
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Maximumw
hiskerlength(m)
Surface roughness / Ra (m)
Figure 6. Maximum whisker length vs surface roughness (1000 Hrs at 30oC/60%RH).
0
10
20
30
40
50
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Whiskerdensity(/mm
2)
Surface roughness / Ra (m)
Figure 7. Whisker density vs surface roughness (1000 Hrs at 30oC/60%RH).
0.087 0.120 0.187 0.249
0.288 0.358 0.402 0.487
Whisker free tin Jan Feb-WHITE PAPER DOWNLOAD:Metal Tech Speaking Template USE THIS.qxd 3/8/2012 10:30 AM Page 2
8/2/2019 Whisker Free Tin Jan Feb White Paper Download
3/6
hours). The samples were examinedfor whisker formation at varioustime intervals.
matched CTE (coefficient of ther-mal expansion). The latter twoforms are commonly used to induce
internal stress in controlled experi-ments. External stress is also knownto initiate whisker growth. Anexample is the stress induced bypress fit connectors.
EXPERIMENTS AND RESULTSA. Copper Surface Modification
A study was conducted on the mor-phology of the copper substrateprior to plating. A series of sub-strates varying in roughness wereevaluated for whisker formationafter electroplated tin deposition.
The roughness was controlled bychemical etching procedures.
Average roughness (Ra) variedbetween 0.13 to 0.47 microns. Asshown in Figure 4, 0.47m Ra has amuch larger surface area comparedto 0.13m Ra. The propensity towhisker was evaluated as follows:
Test VehicleThe test vehicle - CDA19400 (Cu-2.3Fe-0.03P-0.12Zn) lead frame (seeFigure 3).
Tin platingThe plating bath was MSA-basedmatte tin. The plating was run at acurrent density of 10A/dm2. Platingtime was varied to produce a 3 micronand a 10 micron thick deposit. Theformer was for short term whiskerevaluation and the latter which is typ-ical of lead frame plating was used forlong term evaluations.
MethodologyThe test vehicles were subjected to
chemical micro-roughening to pro-duce a set of specific Ra values(Figure 4). The figure shows the SEMmicrographs of the different degreesof micro-roughening as measured inRa um. The samples were then runthrough a standard plating processas outlined in Figure 5. The sampleswere then stored under controlledambient conditions (30oC/60%RH)for extended periods of time (1000
Definition of a WhiskerA whisker is a protrusion >10m in
TECHNICALLYspeaking
Figure 8. SEM illustrating 3 m Tin after 1000 hours at 30 oC/60%RH.
Figure 9. Morphology of IMC surface after tin stripping.
Figure 10. Cross-section showing the IMC after tin strip.
0.00.20.40.60.81.0
Ra0.13um Ra0.47umZeroCrossTime(sec.)
SurfaceroughnessoncopperFigure 11. Comparison of zero cross time of 10 m tin deposit for two levels of roughness.
Ra 0.47, No whiskersRa 0.13, whiskers
Ra 0.13 m Ra 0.47 m
Ra 0.13 m Ra 0.47 m
12 I metalfinishing I January/February 2012 www.metalfinishing.com
Whisker free tin Jan Feb-WHITE PAPER DOWNLOAD:Metal Tech Speaking Template USE THIS.qxd 3/8/2012 10:30 AM Page 3
8/2/2019 Whisker Free Tin Jan Feb White Paper Download
4/6
www.metalfinishing.com January/February 2012 Imetalfinishing I 13
cross-sections of the same Ra val-ues. It is clear that the rougher Raof 0.47m produced a thinner,more uniform IMC, compared tothe smoother Ra of 0.13m, whichshowed increased IMC thickness inlocalized areas. A plausible explana-tion is that the IMC is spread over amuch larger area on the roughermorphology (Ra 0.47m) com-pared to the smaller area of thesmoother surface (Ra 0.13um). It
follows, then, that the stress result-ing from IMC formation would behighly reduced and dissipated withincreased surface roughness of theunderlying copper substrate.
The solderability and the ductilityof a 10m tin deposit on the twoextremes of surface morphologywere examined using WettingBalance Testing as well as the BendTest. There was virtually no differ-
length and that has an aspect ratio(length/diameter) >2.
Measurement of Whisker LengthThe measurement, according to
JEITA ET-7410, is the straight-linedistance from the point of emer-gence of the whisker to the most dis-tant point on the whisker.
Results and DiscussionWhiskers wereexamined, meas-ured and tabulat-ed after 1000hours of storageunder controlled
ambient condi-t i o n s(30oC/60%RH).The data gath-ered fromwhisker exami-nation on the
various mor-phologies ofroughening aregraphed inFigure 6 and 7.Figure 6 looks at maximum whisker
length as a function of roughness.Figure 7 looks at the whisker densityper mm2 as a function of roughness.
The data clearly indicates thatthere is clear correlation between sur-face roughness and whisker propen-sity. The rougher surface produceslower whisker length and also lowerdensity per mm2. Figure 8 showswhisker growth on 3m of tin platedon smoother copper (Ra 0.13) com-pared to no whiskers on the roughersurface (Ra 0.47)
Samples with a tin deposit thick-
ness of 10m were stored for 7000hours at 30oC/60% RH. The tin wasthen stripped by chemical means andthe IMC morphology was examined.In addition, cross-sections were pre-pared and examined to verify the topdown observation.
Figure 9 shows the top view of theIMC after tin stripping on twoextremes of Ra, namely Ra 0.13mand Ra 0.47m. Figure 10 shows
ence in performance (see Figure 11).
B. Modifying the Crystal Structureof the Tin Deposit
A close examination of the crystalstructure of both tin and tin-leadalloy shows a clear differencebetween the two deposits. The tin-lead which does not whisker has anequiaxed relatively fine-graineddeposit. The tin, on the other hand,shows larger columnar crystals.
Figure 12 shows the difference incrystal structure between tin andtin-lead alloy (10 wt%Pb).
It is believed that if the crystalstructure of the tin deposit can bemodified to the tin-lead crystalstructure, the stresses will be dissi-pated and whiskers will not form.
Tests were conducted using thesame test vehicle and the same plat-ing conditions as outlined earlier in
TECHNICALLYspeaking
Figure 12. SEM and schematics of the tin vs the tin-lead deposit structures.
Figure 13. SEM of a cross-section and surface morphology in the three types of tin deposits.
Type A Type B Type C
Whisker free tin Jan Feb-WHITE PAPER DOWNLOAD:Metal Tech Speaking Template USE THIS.qxd 3/8/2012 10:30 AM Page 4
8/2/2019 Whisker Free Tin Jan Feb White Paper Download
5/6
typical of lead frames) and wereplaced in an ambient environment(30oC/60%RH) for 4000 hours.
Figure 14 shows Type A tin depositwith relatively long whiskers devel-oped. Figure 15 shows Type B, withwhiskers that are shorter than theType A whiskers. Figure 16 shows nowhisker formation with a Type Ccrystal structure stored under thesame conditions.
Figure 18 is the result of the finegrained equiaxed crystal structure(Type C deposit) achieved by modify-ing the plating bath with specifictypes of additives.
The result is a very controlled,
evenly distributed, and relatively thinIMC producing minimum stress.The equiaxed crystal structure dissi-pates the stress resulting in nowhisker formation. In this study nowhiskers were observed with finegrained equiaxed tin deposits storedunder ambient conditions for up to22,000 hours.
CONCLUSION
the copper surface roughness study.Three types of tin deposits were
produced by the use of specific plat-ing additives to the bath: Type A isa standard tin deposit characterizedby large columnar crystals; Type Bis modified to produce smallercolumnar grain structure. Type Cwas further modified to produce astill smaller grain that is both colum-nar as well as equiaxed, almost mim-
icking the tin-lead structure (seeFigure 13). The level of additive inthe bath is maintained by continu-ous dosing. Dosing is based on
AmpHrs of plating and results onconsistent crystal structure through-out the life of the bath.
Results and DiscussionAll three types were plated to the typ-ical thickness of 10 m (thickness
In this study two distinct approacheswere attempted to restrain whiskergrowth in tin deposits over copper.
The first approach was to create auniform IMC, by mico-rougheningthe copper substrate before tin depo-sition. A uniform IMC would elimi-nate high stress in localized areas.The second approach was to modifythe grain, from a large columnarstructure to a fine grained equiaxedstructure, resembling the structureof tin-lead deposit. This was achievedby the use of commercially available,proprietary additives. Tin depositwhich had crystal structure similarto tin-lead deposit restrained tin
whisker formation effectively.Crystal structure modification of thetin deposit was demonstrated to be a
very effective way to restrain tinwhisker formation.
REFERENCES[1] P. Oberndorff, M. Dittes, P.
Crema, and S.Chopin ; WhiskerFormation on Matte SnInfluence of High Humidity,
TECHNICALLYspeaking
Figure 14. Type A, Whiskers (long).
Figure 15. Type B, Whiskers (short).
Figure 16. Type C, No Whiskers.
Figure 17. A graph depicting the length of whisker vs storage time of the three crystal types of tindeposits.
Figure 18. The actual SEM of a cross-section and a graphic presentation of the same.
14 I metalfinishing I January/February 2012 www.metalfinishing.com
Whisker free tin Jan Feb-WHITE PAPER DOWNLOAD:Metal Tech Speaking Template USE THIS.qxd 3/8/2012 10:30 AM Page 5
8/2/2019 Whisker Free Tin Jan Feb White Paper Download
6/6
www.metalfinishing.com January/February 2012 Imetalfinishing I 15
ACKNOWLEDGEMENTSThe development work was done atR & D department at C. Uyemura &Co., Ltd.Hirakata Japan. Theresearch team was headed by
Masanobu Tsujimoto with inputfrom: Shigeo Hashimoto, MasayukiKiso, Raihei Ikumoto,ToshikazuKano, and Genki Kanamori .
BIOGeorge Milad is the National AccountsManager for Technology at UyemuraUSA. he may be reached via [email protected].
55thECTC, 2005.[2]W.J. Boettinger, C.E. Johnson,
L.A. Bendersky, K.-W. Moon, M.E.
Williams, and G.R. Stafford:Whisker and Hillock Formationon Sn, Sn-Cu and Sn-PbElectrodeposits, Acta Mater.,
Vol.53, pp.5033-5050, 2005.[3]W. Zhan and F. Schwager:
Effects of Lead on Tin WhiskerElimination - efforts toward lead-free and whisker-free electrode-position of tin, Journal of TheElectrochemical Society, Vol. 153 (5),pp. C337-C343, 2006.
TECHNICALLYspeaking
Whisker free tin Jan Feb-WHITE PAPER DOWNLOAD:Metal Tech Speaking Template USE THIS.qxd 3/8/2012 10:30 AM Page 6