issn 1675-7009ir.uitm.edu.my/id/eprint/12940/1/aj_ramani mayappan srj... · 2016. 5. 30. ·...

21
Volume 7 No. 2 ISSN 1675-7009 DEC 2010

Upload: others

Post on 26-Feb-2021

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Volume 7 No. 2

ISSN 1675-7009

DEC 2010

Page 2: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

SCIENTIFIC RESEARCH JOURNAL

Chief Editor

Zaiki Awang Universiti Teknologi MARA, Malaysia

Managing Editor

Razidah Ismail

Universiti Teknologi MARA, Malaysia

Editorial Board

Abu Bakar Abdul Majeed, Universiti Teknologi MARA, Malaysia David Shallcross, University of Melbourne, Australia

Halila Jasmani, Universiti Teknologi MARA, Malaysia Hamidah Mohd. Saman, Universiti Teknologi MARA, Malaysia

Huang Siew Lai, Universiti Teknologi MARA, Malaysia Ichsan Setya Putra, Bandung Institue of Technology, Indonesia

Ideris Zakaria, Universiti Malaysia Pahang, Malaysia Ir. Suhaimi Abd. Talib, Universiti Teknologi MARA, Malaysia

Jamil Salleh, Universiti Teknologi MARA, Malaysia K. Ito, Chiba University, Japan

Kartini Kamaruddin, Universiti Teknologi MARA, Malaysia Luciano Boglione, University of Massachusetts Lowell, USA

Mohd Hanapiah Abidin, Universiti Teknologi MARA, Malaysia Mohd Rozi Ahmad, Universiti Teknologi MARA, Malaysia Mohd. Nasir Taib, Universiti Teknologi MARA, Malaysia

Muhammad Azmi Ayub, Universiti Teknologi MARA, Malaysia Norashikin Saim, Universiti Teknologi MARA, Malaysia

Nordin Abu Bakar, Universiti Teknologi MARA, Malaysia Robert Michael Savory, Universiti Teknologi MARA, Malaysia

Saadiah Yahya, Universiti Teknologi MARA, Malaysia Salmiah Kasolang, Universiti Teknologi MARA, Malaysia

Shah Rizam Mohd. Shah Baki, Universiti Teknologi MARA, Malaysia Titik Khawa Abd. Rahman, Universiti Teknologi MARA, Malaysia

Wahyu Kuntjoro, Universiti Teknologi MARA, Malaysia Yin Chun Yang, Universiti Teknologi MARA, Malaysia

Zahrah Ahmad, University of Malaya, Malaysia

Copyright © 2010 Universiti Teknologi MARA, Malaysia All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means; electronics, mechanical, photocopying, recording or otherwise; without prior permission in writing from the Publisher. Scientific Research Journal is jointly published by Research Management Institute (RMI) and University Publication Centre (UPENA), Universiti Teknologi MARA, 40450 Shah Alam, Selangor, Malaysia. The views and opinion expressed therein are those of the individual authors and the publication of these statements in the Scientific Research Journal do not imply endorsement by the publisher or the editorial staff. Copyright is vested in Universiti Teknologi MARA. Written permission is required to reproduce any part of this publication.

Page 3: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

SCIENTIFIC RESEARCH JOURNAL

Vol. 7 No. 2 December 2010 ISSN 1675-7009

1. Cu6Sn5 and Cu3Sn Intermetallics Study in the Sn-40Pb/Cu

System During Long-term Aging Ramani Mayappan Zainal Arifin Ahmad

1

2. The Properties of Agricultural Waste Particle Composite Reinforced with Woven Cotton Fabric Mohd Iqbal Misnon Shahril Anuar Bahari Mohd Rozi Ahmad Wan Yunus Wan Ahmad Jamil Salleh Muhammad Ismail Ab Kadir

19

3. Hyperelastic and Elastic-Plastic Approaches for Modelling Uniaxial Tensile Performance of Woven Fabrics Yahya, M. F. Chen, X

31

4.

Effects of Particle Sizes, Wood to Cement Ratio and Chemical Additives on the Properties of Sesendok (Endospermum Diadenum) Cement-bonded Particleboard Jamaludin Kasim Shaikh Abdul Karim Yamani Ahmad Firdaus Mat Hedzir Ahmad Syafiq Badrul Hisham Mohd Arif Fikri Mohamad Adnan

57

Page 4: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

5.

Synthesis, Characterization and Biological Activities of Nitrogen-Oxygen-Sulfur (NOS) Transition Metal Complexes Derived from Novel S-2, 4-dichlorobenzyldithiocarbazate with 5-fluoroisatin Mohd Abdul Fatah Abdul Manan Hadariah Bahron Karimah Kassim Mohd Asrul Hafiz Muhamad Syed Nazmi Sayed Mohamed

67

Page 5: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal Vol. 7 No.2, 1-17,2010

Cu6SnS and Cu3Sn IntermetallicsStudy in the Sn-40Pb/Cu System

during Long-term Aging

Ramani Mayappan' andZainalArifin Ahmad'

'Faculty ofApplied Sciences, Universiti Teknologi MARAArau Campus, 02600 Arau, Perlis, Malaysia

2School ofMaterials and Minerals Resources Engineering,Engineering Campus, Universiti Sains Malaysia,

14300 Nibong Tebal, Pulau Pinang, Malaysia'Esmail: [email protected]

ABSTRACT

Replacing Sn-Pb solder with lead-free solder is a great challenge in theelectronics industry. The presented lead-free solder is Sn based and formstwo intermetallic species upon reaction with the Cu substrate, namely CUr?n5and CuSn. The growth of Cu

6Sn5 and CuSn intermetallics have beeninvestigated with respect to Sn-40Pb1Cu solderjoints. The joints were agedunder long-term thermal exposure using single shear lap joints and theintermetallics were observed using scanning electron microscopy. As-solderedsolder joints exhibit a single CUr?n5 phase, however after aging a CuSnlayer below the CUr?n5 is observed to manifest. The CUr?n5 layer developswith a scalloped morphology, whereas the CujSn layer always develops anundulating planar shape in phase with the Cu

6Sn5• The C1I

6Sn5 layer beginsto transform from a scalloped- to a planar-shape as aging progresses inorder to minimize the interfacial energy. The intermetallic layers exhibit alinear dependence on the square root ofaging time, which corresponds todiffusion-controlled growth. The activation energy for the growth of theCuSn, intermetallic layer has been determined to be 56.16 kJlmol.

Keywords: Intermetallic, Lead-free Solder, Cu-Sn, Growth Kinetics, SolderI

ISSN 1675-7009© 2010 Universiti Teknologi MARA (UiTM), Malaysia.

Page 6: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

Introduction

When molten solder wets a base metal (substrate) metallurgical bondsform as the solder solidifies [1,2]. This metallurgical bonding occurs atthe boundary between the solder and the substrate, termed the interface.This interface consists ofa thin layer(s) ofalloy termed the intermetalliccompound (IMC). This intermetallic layer(s) is composed ofmetals fromboth the solder and the substrate [1]. Manko [3] defines an intermetalliccompound to be a distinguishable homogeneous phase with a relativelynarrow range of composition in simple stoichiometric proportions (i.e.atomic ratios). The intermetallic layer provides the necessary bondingstrength and interfacial continuity between the solder and the substrate.

The formation and subsequent growth ofthe IMC is a major issue insoldering. The formation of intermetallic compounds is desirable for agood solder joint and in soldering this formation is a consequence ofgoodwetting. The implementation of lead-free solders as new solderingmaterials enhances the importance of these IMCs, because althoughtheir formation is desirable in attaining good bondingbetween the substrateand the solder, there are some drawbacks. The intermetallics are quitebrittle and excessive thickness may degrade the interfacial strength andculminate in a mismatch in the physical properties: such as the thermalexpansion coefficient and the elastic modulus [4].

Reliability losses in many electronic systems have been identified tobe directly related to the failure of solder joints, as opposed to devicemalfunction, consequently there is greater focus on solder joint reliability[5]. In order to develop reliable lead-free solder joints, it is desirable tounderstand the kinetics governing the growth ofIMCs.

Lead-free solder is tin-based and the interaction between the solderand the Cu substrate produces Cu-Sn IMCs. In the case of liquid tin!copper and many liquid-tin alloy/copper interconnections; Cu6SnS

IMCsform almost immediately upon contact ofthe liquid solder with the solidcopper [6-9]. After solid state aging at high temperature a Cu

3Snlayer

forms below the Cu6SnS

phase [10]; a consequence of a lower growthactivation energy for Cu

3Snas opposed to that for Cu6SnS' Flanders et

al. [11] reported the activation energies for the growth of E-Cu3Sn

andCu

6SnSintermetallic phases in a Sn-3.5Ag/Cu system to be 77.21 and

107.10 kJ/mol, respectively. Yoon and lung [12] stated that solid stateisothermal aging in a Sn-5Bi/Cu system using the reflow method yieldedIMC growth activation energies of 90.50 and 98.35 kJ/mol for E-Cu3Sn

2

Page 7: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CuSn, and CuJSn Intermetallics Study in the Sn-40Pb/Cu System

and Cu6SnS

' respectively. Vianco et al. [13] studied the effect of solidstate aging on the intennetallic growth of Sn-3.9Ag-0.6Cu on a Cusubstrate and reported that the activation energy for the growth of theE-Cu

3Snphase is 50 and 44 kl/mol for the Cu

6SnSphase, which is

lower than those reported in the aforementioned lead-free solders.This study investigates the microstructural evolution and growth

kinetics of'Cu.Sn, and Cu3Sn

IMCs using conventional Sn-40Pb solderon a Cu substrate.

Measurements and Apparatus

The Sn-40Pb solder, ZnCI., flux and Cu substrate considered andimplemented in this study are all commercially available. Before soldering,the Cu surface was polished using SiC coated paper in order to removethe inherent copper oxide layer. The single shear lap tensile specimenswere constructed by soldering two dog-bone shaped Cu strips togetherusing approximately 0.2 g of Sn-40Pb solder sandwiched between thetwo copper substrates. The solder contact area was maintained at 4 x 8mm- and a thickness of0.5 mm, shown in Figure I.

Figure I: Single Lap Dog-bone Shape (All Dimensions in mm)

After applying ZnCI., tlux to the soldered area ofthe Cu substrates,the system was placed on ahotplate and heated to a maximum temperatureof 240°C as presented in the temperature profile in Figure 2.

After soldering, specimens were left on the hotplate to cool andremoved upon attaining 50°C. The samples were then cleaned using dishwashing soap to remove any impurities. After allocating some samplesfor as-soldered IMC reference analysis, the remainder underwent thermalaging. The soldered samples were placed in an oven and heated at 50,

3

Page 8: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

250

_ 200Ua, 150e::J 10010Q 50Q,~ 0 +-_-__~ -....__-__- __I- 0 2 4 6 8 10 12 14 16 18

Time (min)

Figure2: HotplateTemperature Profile

75, 100, 125 and 150°C with a temperature stability of± 3°Cand aged for100,250,375,500 and 700 hours.

Cross-sectional samples were taken using a low speed diamond cutterand mounted using epoxy resin. The mounted samples were mirrorpolished and etched. The Cu/solder/Cu interface was analyzed using aZeiss-Supra 35VP-24-58 Field Emission Variable Pressure ScanningElectron Microscope (FEVPSEM) operating at 15 kV in backscattermode. For each sample, 8-10 micrographs were taken and backscatterEDX has been used to identify the elemental composition ofthe phases.To determine the average thickness ofthe intermetallic layers, the totalarea of the layer was calculated using ImageJ and divided by themicrograph length to an accuracy of± 0.0 I mm.

Results and Discussions

mtermetalllc Formation

Figures 3-9 show the interface that exists between the Sn-40Pb solderand Cu substrate after 700 hours ofaging at different temperatures. TheIMC formed between the as-soldered Sn-40Pb solder joint and the Cusubstrate is presented in Figure 3. According to Arenas and Acoff [14]the nucleation and growth ofIMCs begins when the molten solder at theinterface becomes saturated with copper. The interface exhibits ascalloped shape with a 40.6:59.4 (Cu:Sn weight percentage) compositionand has been identified to be Cu

6SnSfrom the Cu-Sn phase diagram.

The Cu-Sn system has been studied extensively by many researchersand the formation ofthe Cu

6SnSphase has been well documented [7,15-

4

Page 9: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CU"SIl.l lim! CuJSn hnennetotlics StlU(l' in the SI1-40PhICu System

_...... - .... .- - ..... _..- -~ ....­--... . ---._- ._- ----Figure 3: As-soldered IMC Layer between the Sn-W Pb and Cu Substrate

19]. According to Tu [15] the formation o f the C U"Sn, pha se is aconseque nce of the interstit ial diffusion ofC u into Sn at tem peraturesbelow 60'C.

The growth of the sca lloped CU,Sn, phase is prom oted by exposureto heat over extended t ime periods. From the SE M image presented inFigure 4 it is evident that after aging at 50'C for 500 hours, the sca llopedstruct ures penet rate deep into the solder matrix.

The presence of a e-Cu.Sn layer below the CUr,Sn, phase as report edby Arenas and Acoff [14], Tu and T hompson [ 16], Li u et al. [20] andKim and Tu [2 1] is evident in Figure 5. Hwang et al. [4] used T EM toobserve the s-C u.Sn phase in Cu-Sn system soldered at 250'C andreported that Sn-3Ag-6Bi/Cu, Sn-Cu/C u and Sn-Bi/Cu systems exhibit ae-C u.S n phase approx imate ly 100 nm thick.

Figurc4: (MC Layer Fonncd between the Sn-4 0Pb andCu S ubstrate Isothcnnally Aged at 50 C for500 Hours

5

Page 10: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

In th is st udy. the format ion of a thin layer of E-C u,Sn below C u.Sn,and closer to the Cu subst rate has been identified with respect to EDXcomposit ional analysis; 62:38 (Cu:Sn weight percentage). At higher agingtcmperat ures and longer aging t imes this E-C u,Sn layer becomes thickerand more dist inct ive. Figu re 6.

No E-C u,Sn phase was observ ed usin g FEVPS EM for samplesage d at SO'C; however it was observed for th ose aged at 7S'C. Similarresul ts were rep orted by Tu [ IS, 22 ] and Tu and T hompson [16] , whodisco ve red using T EM ana lysis that th in 11 1m Cu/Sn di ffusion co uplesannea led bel ow 60'C conta ined no c-C u.Sn ph ase , w he re as thoseannea led above 60 'C did. .

Th e forma t ion of th e C U" Sn, phase at the interface depletes the Sn­rich phase nca r the C u subst rate and consequent ly Pb-rich phases begin

• :. (b) -TJ'·Cu sn,y . E· u.sn E·CU,9n ' .

I 'jCu CuM~ . 5 00Kll IpIl'l' ::.::.:":':.:':: :=:..'":_ ....

Figure S: The Formation of the IMC Layer belween the Sn-401'b and CuSubstrate forSamples Isothennally Agedat 7S'C 1<". (a) 100 and (b) 700 Hours

.. . ."

; .. .AS 15 c.N :0:3000 ~m ___

Figure6: T heManifestation ofthe E-Cu_\Sn Below the CuSn, (Mefor a SampleAged at 150 'C for 100 Hours

6

Page 11: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CII"Sn.{ lind CuJSu lntermetallics SII/((I' in the SI1-40PbICIl System

to acc umulate at the soldcr/C u.Sn, intermetallie layer interface. Thefon na tion o f thi s Pb-rich phase ncar the interface is evident in Figure7(a) and Figure 7(b) shows the penetration of the scalloped C u.Sn , phaseinto the Pb-rieh phase .

Sn depletedzone

.'..•. - _. '- _-_oH... I-t _ _ ._~ .-_.

Figure 7: The Reaction between Sn and Cu Creates (a) a Sn Depicted Zoneand Results in (b) the CUt>SnsInt crmetallic Penetration into the Pb-richZone.

Both Images arc forSamples Aged at 150'C for 100 Hours

The format ion of the C u.Sn, IMC red uces the Sn co nce nt rat ion ncarth e interface and results in the format ion of the C u-rieh E-C u,Sn IMC asexp lained by Choi et al. [23] and Islam et al. [24] . Th ey state th at thesupply ofSn thro ugh the C U"Sn, is more restricted than the supply o fC ufro m the subst rate. hence the C u.Sn, trans forms into th e C u-rieh E­C U,Sn phase.

After the fonnation of the E-C u,S n layer both the C u"Sn, and E­C U,S n IMC layers grow progressive ly. Figure 8 depicts the evo lut ion ofthe C U"Sn, phase chang es from a scalloped morp hology to a more layeredst ruc ture and Figure 9 shows the limit of the present ex perimen ta l work .whereby a flat C u.Sn, phase is achieved after aging at 150'C lor 700hours. T he findin gs arc in goo d agreement with the wo rks of Tu et al.[2]. Lee et a/. [25] and Yu et al. [26].

Th e transformation from a scalloped to a more pla nar morphology asaging progressesmay be attributed to changes in the interfacial energy.According to Tu et al. [2] scalloped morphologies have larger interfacialareas than flat inter faces. So the IMC converts to a layered morphologyin order to minimize the interfacial energy.

Furthermore, astheagingtime increases, neighbouring scallops growinto eac h other. and the overall thickness of the C U"Sn, layer increases.Lee et al. [25] exp lained that in so lid-state aging. the Intcrrncrallic gro wt hat the channels (va lleys) between two adjacent scallops is faster than at

7

Page 12: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

Cu

Figure R: Transformntion from Scallopto Planar LayerwithAging: (a) I ()(l"C for250 Ho urs. (b) 125'C for 100 Hours , (e) 100·C for 500 1I01irs. (d) 125'C for 375

Hours, (e) 125'C for 700 Hours and (I) 150'C for 375 110llrs

Figure9: IMCLayer Fanned bel ween Sn-40Pb andCu Subst rate Isothermally Aged at 150'C for 700 Hours

8

Page 13: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CUrPn5 and Cu Sn Intermetallics Study in the Sn-40Pb/Cu System

the peaks ofthe scallops. Consequently the Cu6SnSintennetallic layer is

transformed from a scallop-like shape to a layer-type shape. The twoCu-Sn phases described have been observed in other Sn-based lead-freesolders on Cu substrates including 100Sn [27], Sn-3.5Ag [28] and Sn­3.9Ag-0.6Cu [13].

Intermetallic Growth

The increase in temperature and time results in growth of the Cu6SnS

and e-Cu.Sn intennetallics as presented in Tables 1 and 2. At an agingtemperature of 50°C Cu

6SnSintennetallic growth is insignificant. The

thickness grew from 1.4 mm at room temperature to 2.3 mm after 700hours. For the same duration, but with aging at 75°C the intennetallicthickness increases to 2.9 mm. At this temperature, after 100 hours ofaging, the E-Cu

3Snlayer has a thickness of 0.44 mm, but the E-Cu3Sn

phase does not grow significantly, and within the experimental error ofaround ± 0.15 urn.

For an aging time of700 hours at 100 and 125°C, the intermetallicthicknesses have increased to 5.0 and 8.0 mm, respectively. However,the E-Cu

3Snphase still does not grow significantly at these aging

temperatures; only a slight increase from 0.43 mm to 0.57 mm at125°C.

Aging at 150°C significantly affects the intennetallic thickness. After250 hours ofaging, the scallop structure has become flat and the thicknesshas increased to 6.1 mm. As the aging time increases, the intermetallicthickness increases further to 10.0 mm and the E-Cu

3Snphase thickness

increases to 1.83 mm.

Table I: Cu6SnSIntennetallic Thickness

Temperature/C 50 75 100 125 150(1Jll1) (1Jll1) (1Jll1) (1Jll1) (1Jll1)

As soldered 1.4±0.1 1.4±0.1 1.4±0.1 1.4± 0.1 1.4±0.1100 hours 2.1 ±0.2 2.2±0.2 2.5±0.3 2.8±0.3 Nil250 hours 2.2±O.l 2.4±0.4 3.0±0.2 3.5±0.3 6.1±0.6375 hours 2.3±0.3 2.5±0.2 3.5±0.1 5.0±0.3 7.0±0.8500 hours 2.3±0.1 2.6±0.1 4.0±0.2 '6.O±0.6 8.0±O.l700 hours 2.3±0.3 2.9±0.4 5.0±0.5 8.0±0.3 10.0±0.6

9

Page 14: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

Table2: E-Cu3Sn

Intermetallic Thickness

Temperature/C 75 100 125 150(urn) (urn) (urn) (urn)

As soldered Nil Nil Nil Nil100hours 0.44 Nil 0.43 Nil250 hours 0.31 Nil 0.56 1.13375 hours 0.39 0.42 0.67 1.44500 hours 037 0.39 0.65 1.50700 hours 0.34 0.33 0.57 1.83

According to Choi et al. [23], the formation of the Cu6SnS

layerreduces the quantity of Sn available for reaction at the solder/substrateinterface. Furthermore, the Cu

6SnSlayer may inhibit Sn diffusion towards

the solder/substrate interface. This in turns increases the total diffusiondistance and thus delays the Sn supply required for IMC growth.

Intermetallic Growth Rate and Activation Energy

If the intermetallic growth process is a volume-diffusion-controlledphenomenon, then the isothermal growth ofthe intermetallic layer can bedescribed using a square root time law [21]. Generally, the thickness ofan IMC layer in diffusion couples can be expressed by a simple parabolicequation, Equation (I );

d = (k.t)" + do (1)

where d is the thickness ofthe intermetallic layer, do is the initial thickness,k is the growth rate constant (cmvs), n is the time exponent (0.5) and tisthe reaction time (in s). Equation (1) can be re-written as follows:

(2)

k is strongly related to the diffusion coefficient for the elements comprisingthe IMC and can be determined by linear regression. Plotting the averagemeasured intermetallic thickness, d - do , against the square root of theaging time, t1, yields lines ofgradient IJ ,Figure 10.

10

Page 15: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Cu6Sn5

and CU.lSn Intermetallics Study in the Sn-40Pb1Cu System

The growth rate constants for all the intennetallic compounds arepresented in Table 3, however kinetic analysis ofthe Cu

3Snlayer cannot

be presented in this study due to insufficient data.

• SOC

12 _7SC

E .l100C10 :t:a X12SC

(I)(I) 8 x1S0CCDe~ 6:cI- 4~iii 2~

E 0CI)

E 0 200 400 600 800 1000 1200 1400 1600(Aging Time)112 (s)112

Figure 10:The Relationshipbetween Cu6SnSIntennetallic

1Thickness and Aging Time, t 2

Table3: Growth Rate Constants for Cu6SnS

IMC Growth

Temperature(C)

5075100125150

kI k1

(cm/s ' ) (cmvs)

5 x 10-8 0.25 X 10-14

9 X 10-8 0.81 X 10-14

10 X 10-8 4.41 X 10-14

26 X 10-8 15.2X 10-14

42 X 10-8 27.0 X 10-14

The linearity of the data and the good correlation between the datapoints and the linear regression fits indicate that IMC layer growth is adiffusion-controlled process and hence the Arrhenius equation can beused to determine the activation energy for intennetallic growth:

11

(3)

Page 16: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

which can be rewritten as

Ink=-~ + In kRT 0

(4)

where Q is the activation energy for layer growth (J/mol), R is the idealgas constant (8.314 J/mo1.K) and T is the absolute temperature in Kelvin.The activation energy for Cu

6SnSlayer growth has been determined to

be 56.16 kllmol from the gradient of a plot ofIn k versus lIT, Figure 11.This value excludes the intermetallic thickness data for aging at 150°Cand 100 hours, because the layers are too irregular to measure the IMCgrowth rates accurately. Some literature values for the growth of theCu6SnS

intermetallic are presented in Table 4 for the purpose ofcomparison.

-25

-27

~E -29.2.Q -31

.5-33

Sn-40Pb

y = -6755.4x - 12.n8R2 = 0.9879

Q = 56.16 kJ/mol

0.003-35 of---------r-------.....,.---

0.0022

Figure 11:Arrhenius Plot ofthe Cu6SnS

Intennetallic Layer Growthfor the Sn-40Pb/Cu System

The data obtained by Abtew and Selvaduray [29] in which Sn-40Pbsolder paste reacted on a single crystal Cu substrate gives a significantlylower activation energy value than that determined in this study. Thismay be attributed to the form ofCu employed, i.e. a single Cu crystal asopposed to a Cu substrate.

12

Page 17: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CuSn, and CuSn lntermetallics Study in the Sn-40Pb/Cu System

Table4: Activation Energies forCu6SnS

Intennetallic Growth

Solder/substrate system Activation Temperat ure Aging time Referenceenergy (kJ/mol) range f'C) (hours)

Sn-40Pb solder 39.64-48.22 90-170 Nil [29]paste/single crystal Cu

Sn-Pb/Cu 45-66 Nil Nil [30]Sn-Pb/Cu III 50-150 4000 [23]

Sn-37Pb/Cu 77 110-160 1536 [I]Cu/Sn-40Pb/Cu 42.25 100 and 150 625 [31]

Sn-40Pb/Cu 56.16 50-150 700 This study

Conclusions

The morphology and kinetics of the growth of IMC layers at the Cu­solder interface in Sn-40Pb solder joints during solid state aging hasbeen determined. The IMC layers formed consist of'Cu.Sn, which formsadjacent to the Cu substrate and Cu

6SnS' which forms towards the

solder. The Cu6SnS

develops with a scalloped morphology, whereasCu

3Sngrows as an undulating planar layer in phase with the Cu6SnS

'

The Cu6SnS

layer transforms from a scalloped- to a planar-morphologyas aging progresses due to minimization of the interfacial energy. Thegrowth ofthe IMC is diffusion-controlled and consequently the activationenergy for growth ofthe Cu

6SnSintermetallic layer has been determined

to be 56.16 kJ/mol which is in good agreement with the literature values.This finding is ofvalue in that it provides kinetic insight into the formationand progressive growth ofIMCs in solderjoints, which may impinge onthe struct ural and mechanistic application of Sn-based solders in theelectronics industry.

Acknowledgement

The work described in this article was supported by AUN/SEED-netgrant.

13

Page 18: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

References

[1] D.R. Frear, H.S. Morgan, S.N. Burchett and J.H. Lau, 1994. TheMechanics of Solder Alloy Interconnections, New York:Chapman and Hall, International Thomson Publishing.

[2] K.N. Tu, T.Y. Lee, J.W. lang, L. Li, D.R. Frear, K. Zeng and J.K.Kivilahti, 2001. Wetting reaction versus solid state aging ofeutecticSnPb on Cu, Journal ofApplied Physics, vol. 89(9), pp. 4843­4849.

[3] H.H. Manko, 2001 Solders and soldering, materials, design,production, analysis for reliable bonding, New York: McGrawHill Book Company.

[4] C.-W. Hwang, K.-S. Kim and K. Suganuma, 2003. Interfaces inlead-free soldering, Journal ofElectronic Materials, vol. 32( 11),pp. 1249-1256.

[5] D.R. Frear, 1996. The mechanical behavior of interconnectmaterials for electronic packaging, JOM, 48(5), pp. 49-53.

[6] H.K. Kim, H.K. Liou and K.N. Tu, 1995. Morphology ofinstabilityofthe wetting tips SnBi, eutectic SnPb, and pure Sn on Cu, JournalofMaterial Research, vol. 10, pp. 497-504.

[7] R.A. Gagliano and M.E. Fine, 2003. Thickening kinetics ofinterfacialCu6SnS

and Cu3Sn

layers during reaction of liquid tin with solidcopper, Journal of Electronic Materials, vol. 32(12), pp. 1441­1447.

[8] R. Mayappan, A.B. Ismail, Z.A. Ahmad, T.Ariga and L.B. Hussain,2006. Effect ofsample perimeter and temperature on Sn-Zn basedlead-free solders, Materials Letters, vol. 60, pp. 2383-2389.

[9] R. Mayappan, A.B. Ismail, Z.A. Ahmad, T.Ariga and L.B. Hussain,2007. The effect of crosshead speed on the joint strength betweenSn-Zn-Bi lead-free solders and Cu substrate, Journal ofAlloysand Compounds, vol. 436, pp. 112-117.

14

Page 19: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CuSn, and Cu Sn Intermetallics Study in the Sn-40Pb/Cu System

[10] K. Jung and H. Conrad, 2001. Microstructure coarsening duringstatic annealingof60Sn40Pb solderjoints: III intennetallic compoundgrowth kinetics, Journal ofElectronic Materials, vol. 30, pp. 1308­1312.

[11] D.R. Flanders, E.G. Jacobs and R.F. Pinizzotto, 1997. Activationenergies ofintennetallic growth ofSn-Ag eutectic solder on coppersubstrates, Journal of Electronic Materials, vol. 26(7), pp. 883­887.

[12] J.-W. Yoon and S.-B. Jung, 2003. Investigation of interfacialreactions between Sn-5Bi solder and Cu substrate, Journal ofAlloys and Compounds, vol. 359, pp. 202-208.

[13] P.T. Vianco, J.A. Rejent and P.F. Hlava, 2004. Solid-stateintennetallic compound layer growth between Copper and 95.5Sn­3.9Ag-0.6Cu solder, Journal ofElectronic Materials, vol. 33(9),pp.991-1004.

[14] M.F. Arenas and V.L. Acoff, 2004. Contact angle measurementsof'Sn-Ag and Sn-Cu lead-free solders on copper substrates, JournalofElectronic Materials, vol. 33(12), pp. 1452-1458.

[15] K. N. Tu, 1973.lnterdiffusion and reaction in bimetallic Cu-Sn thinfilms, Acta Metallurgica, vol. 21(4), pp. 347-354.

[16] K.N. Tu and R.D. Thompson, 1982. Kinetics ofinterfacial reactionin bimetallic Cu-Sn thin films, Acta Metallurgica, vol. 30(5), pp.947-952.

[17] Z. Mei, A.J. Sunwoo and 1.W. Morris, 1992. Analysis of low­temperature intennetallic growth in copper-tin diffusion couple,Metallurgical Transactions A, vol. 23A(3), pp. 857-864.

[18] K.F. Dreyer, W.K. Neils, R.R. Chromik, D. Grosman and E.1.Cotts,1995. Calorimetric study of the energetic and kinetics ofinterdiffusion in CU/Cu6SnS

thin-film diffusion couples, AppliedPhysics Letters, vol. 67( 19), pp. 2795-2797.

15

Page 20: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

Scientific Research Journal

[19] H.K. Kim, H.K. Liou and K.N. Tu, 1995. Three-dimensionalmorphologyofa veryrough interfaceformedin the solderingreactionbetween eutectic SnPb and Cu, Applied Physics Letters, vol.66(18), pp. 2337-2339.

[20] A.A. Liu, H.K. Kim, K.N. Tu and P.A. Totta, 1996. Spalling ofCu

6SnSspheroids in the soldering reaction ofeutectic SnPb on Crl

Cu/Au thin films, Journal ofAppliedPhysics, vol. 80(5), pp. 2774­2780.

[21] H.K. Kim and K.N. Tu, 1996. Kinetic analysis of the solderingreaction between eutectic SnPb alloy and Cu accompanied byripening, Physics Review B, vol. 53(23), pp. 16027-16034.

[22] K.N. Tu, 1996. Cu/Sn interfacial reactions: thin-film ca~e versusbulk case, Materials Chemistry and Physics, vol. 46, pp. 217-223.

[23] S. Choi, T.R. Bieler, J.P. Lucas and K.N. Subramanian, 1999.Characterization ofthe growth ofintermetallic interfacial layers ofSn-Ag and Sn-Pb eutectic solders and their composite solders onCu substrate during isothermal long-term aging, Journal ofElectronic Materials, vol. 28(11), pp. 1209-1215.

[24] M.N. Islam, A.Sharif and Y.C. Chan, 2005. Effect of volume ininterfacial reaction between eutectic Sn-3.5%Ag-0.5%Cu solderand Cu metallization in microelectronics packaging, Journal ofElectronic Materials, vol. 34(2), pp. 143-149.

[25] T.Y. Lee, W.J. Choi, K.N. Tu, J.W. Jang, S.M. Kuo, J.K. Lin, D.R.Frear, K. Zeng and J.K. Kivilahti, 2002. Morphology, kinetics, andthermodynamics of solid-state aging ofeutectic SnPb and Pb-freesolders (Sn-3.5 Ag, Sn-3.8 Ag-Q.7 Cu and Sn-0.7 Cu) on Cu,Journal ofMaterial Research, vol. 17(2), pp. 291-301.

[26] D.Q. Yu, C.M.L. Wu, C.M.T. Law, L. Wang and J.K.L. Lai, 2005.Intermetallic compounds growth between Sn-3.5Ag lead-free solderand Cu substrate by dipping method, Journal of Alloys andCompounds, vol. 392, pp. 192-199.

16

Page 21: ISSN 1675-7009ir.uitm.edu.my/id/eprint/12940/1/AJ_RAMANI MAYAPPAN SRJ... · 2016. 5. 30. · Cross-sectionalsamples weretaken usinga lowspeed diamond cutter and mounted using epoxy

CuSn, and Cu Sn Intermetallics Study in the Sn-40Pb1Cu System

[27] Y.G. Lee and J.G. Duh, 1999. Interfacial morphology andconcentration profile in the unleaded solder/Cu joint reliabilityassembly, Journal of Materials Science: Materials inElectronics, vol. I0, pp. 33-43.

[28] S. Ahat, M. Sheng and L. Luo, 200 I. Microstructure and shearstrength evaluation of SnAg/Cu surface mount solder joint duringaging, Journal ofElectronic Materials, vol. 30(10), pp. 1317-1322.

[29] M. Abtew and G. Selvaduray, 2000. Lead free solders inmicroelectronics, Materials Science and Engineering R, vol. 27,pp.94-141.

[30] P.T. Vianco,A.C. Kilgo and R. Grant, 1995. Intermetallic compoundlayer growth by solid state reactions between 58Bi-42Sn solderand copper, Journal ofElectronics Material, vol. 24(10), pp. 1493­1505.

[31] H.-T. Lee and M.-H. Chen, 2002. Influence of intermetalliccompounds on the adhesive strength of solder joints, MaterialsScience and Engineering A, vol. 333, pp. 24-34.

17