Advertisement

Metallurgical and Materials Transactions B

, Volume 50, Issue 1, pp 502–516 | Cite as

Experimental Determination of the Sn-Cu-Ni Phase Diagram for Pb-Free Solder Applications

  • Hector M. HenaoEmail author
  • Ching-shun Chu
  • Juan P. Solis
  • Kazuhiro Nogita
Open Access
Article
  • 158 Downloads

Abstract

The rapid increase of scrapped electrical products disposed in landfill sites is significantly influencing the environment because of the composition of the associated components. This situation has been the stimulus for the international community to issue a directive to prohibit the use of Pb-containing solders in electronic devices and effectively ban such products on the market. The Sn-Cu-Ni has become increasingly significant as it is used in several solder alloys and more generally both Cu and Ni are common substrates that interact with Sn-based solders in microelectronic applications. However, there is currently an insufficient understanding of the phase equilibria and the associated phase diagrams of many Sn alloys. The present work implemented an experimental methodology to obtain high accuracy results for the phase equilibria between one liquid and one or two intermetallic compounds. The Sn-Cu-Ni and the associated Sn-Cu and Sn-Ni binary systems have been re-evaluated and compared with the literature. The liquidus and the invariant lines were determined for the Sn-Cu-Ni ternary at the temperatures ranges of between 240 °C and 270 °C. The result will contribute to filling phase equilibria data gaps in the Sn-rich corner of the Sn-Cu-Ni alloy system.

Introduction

On the 1st of July 2006, the European Union issued a directive prohibiting the use of Pb-containing solders in electronic devices and effectively banning such products from the market. Legislators in other jurisdictions, including Japan, China, and USA are considering regulations that would have a similar effect.[1, 2, 3] This is the reason that in recent decades, Pb-free solders have seen widespread use in the industry and there has been an increased stimulus for research. This research has included large multi-institution projects to develop thermodynamic databases for solder alloys; in the EU, this has been conducted within the COST project[4] and in the USA this has been led by NIST.[5] These efforts have resulted in the publication of an Atlas of Phase Diagrams for Lead-Free Soldering[6] and been the stimulus for commercial thermodynamic packages offering lead-free soldering databases including Thermo-Calc. (TCSLD2 database)[7] and MTDATA(MTSOLDERS database).[8]

The projects mentioned above have been successful with many key ternary and higher order systems, notably Sn-Ag-Cu and Au-Ni-Sn, being characterized but, importantly, key systems remain that have not been fully studied which is an impediment to solder alloy design. An important example is the ternary Sn-Cu-Ni system, which while present in the Thermo-Calc. and MTDATA databases, relies on databases that have not been optimized for the full range of temperatures and compositions relevant to the soldering of electronic materials.

Literature Review of the Associated Sn-Cu and Sn-Ni Binary Systems

The Sn-Cu binary system is fundamental of various soldering alloys, such as Sn-Cu-Ag, Sn-Cu-Ni, and Sn-Ni-Cu-Ag and Sn-Ag-Co.[9,10] A summary of the reported phase equilibrium for the Sn-Cu system between liquid and Cu6Sn5 is displayed in Figure 1. Harding et al.,[11] Yu et al.,[12] Vuorinen et al.,[13] and Thermo-Calc[7] are all in reasonable agreement. On the other hand, the experimental results from Hanson et al.[14] only agree with results near to the eutectic temperature but show a large deviation as temperature is increased. In addition, the experimental result of Gourlay et al.[15] (draw in the Figure 1 for concentrations of Ni < 0.05 wt pct) indicated a large range of variations from 0.7 to 1.5 wt pct of Cu.
Fig. 1

Sn-Cu phase diagrams reported in the literature

There are numerous of Sn-Ni phase equilibria data such as that reported by Liu et al.,[16] Schmetterer et al.,[17] Zemanova et al.,[18] Nash et al.,[19] Hanson et al.,[14] and Gourlay et al.[15,20]; Belyakov et al.[21] covered a temperature range from 200 °C to 300 °C, whereas Hanson et al.[14] covered 200 °C to 800 °C. A calculated phase diagram was provided by Ghosh[22]; however, only a few of them focus on low Ni concentrations and temperature ranges (e.g., 240 °C to 270 °C) useful for the present study. Figure 2 displays the current available liquid/Ni3Sn4 phase equilibrium at the temperatures of interest and the appreciable differences that exists in the literature. In particular, the experimental data of Hanson et al.[14] and Gourlay et al.[15] are quite similar. There are important differences between the experimental works and simulations.
Fig. 2

Sn-Ni phase diagrams reported in the literature

The published information relating to the eutectic points for the Sn-Cu and Sn-Ni binary systems is summarized in Table I. The ranges of compositions of Cu and Ni in the liquid phase in equilibrium with the Inter Metallic Compound (IMC) phases Cu6Sn5 and Ni3Sn4 (respectively) are relatively small while the difference of temperatures between eutectic points in the binary systems is 4 °C.
Table I

Summary of Composition and Temperature of Eutectics in Sn-Cu and Sn-Ni Binary Systems

System

Authors

Reference No.

Temperature (°C)

Weight Percent of Cu or Ni

Sn-Cu

Harding and Pell Walpole (1948)

11

227

0.754

Thermo-Calc

7

226.8

0.885

Hanson (1934)

14

226.91

0.746

Sn-Ni

Nash et al. (1985)

19

231.3

0.163

Hanson et al. (1934)

14

232

0.134

Ghosh (1999)

22

231.1

0.167

Belyakov and Gourlay (2012)

21

231.4

0.090

Literature Review of the Sn-Cu-Ni Ternary System

A review of the experimental results reported in the literature for the effect of various elements including Ni, Au, Sb, In, Co, Zn, and Bi on the behavior of Sn-Cu-based solder alloys has recently been published.[9]

In the case of the Sn-Cu-Ni system, there are a number of studies at high temperatures (> 400 °C); however, as is indicated in Table II, only a few publications provide thermodynamic data in the Sn-rich corner at conditions relevant to the soldering of electronic materials (that is temperatures between 220 °C and 270 °C and low concentrations of Cu and Ni).[12,15,23, 24, 25] Furthermore, each of them only covers a small portion of the data required and contain contradictory results with respect to the presence of thermodynamically stable phases and the position of the liquidus and the invariant line. As an example, the study on Snugovsky et al.[24] and Vourinen et al.[13] reported the presence of two quasi-peritectic reactions with a difference between both reactions of less than 1 °C, which related to the presence of a ternary intermetallic compound described as Cu33Ni23Sn44. The composition of the supposed ternary phase was assumed based on chemical analysis using electron probe microanalyzer (EPMA); however, the phase was not characterized by X-ray diffraction (XRD). Furthermore, there was no agreement regarding the composition of the ternary phase. A summary of reported liqiudus lines is displayed in Figure 3 including the experimental work reported by Gourlay et al.[15] and simulation reported by Vourinen et al.[13] and Yu et al.[12] The figure indicates small differences in terms of the position of the liquidus lines even between 240 °C and 268 °C.
Table II

Literature Summary for Ternary Sn-Cu-Ni System

Authors

Method

Results

Remarks

Schmetterer et al. (2009)

Sn-Cu-Ni

Ref. 26

equilibria/fast quenching/EPMA-XRD

crucible: vacuumed silica ampoule

isothermal at 220 °C, 400 °C, 500 °C, 700 °C

not much information about Sn-rich composition

ternary solubilities were characterized

no indication of achievement of a glassy phase during quenching

Lin et al. (2002)

Sn-Cu-Ni

Ref. 23

equilibria/fast quenching/EPMA-XRD

crucible: vacuumed silica ampoule

isothermal at 240 °C

only one reasonable result for equilibria Sn-Ni3Sn4-Cu6Sn5

at.pct Ni = 1.1, at.pct Cu = 0.6

no indication of achievement of a glassy phase during quenching

Snugovsky et al. (2006)

Sn-Cu-Ni

Ref. 24

equilibria/Fast quenching/EPMA of IMC.

crucible: BN (boron nitride)

invariant lines 235 °C to 227 °C

reported of a ternary compound Cu33Ni23Sn44

no reported XRD.

no indication of achievement of a glassy phase during quenching

Gourlay et al. (2010)

Sn-Cu-Ni

Ref. 15

equilibria/fast quenching/EPMA

crucible: stainless-steel cups coated with BN.

settling of the IMC on the bottom of the crucible

isothermal at 268 °C

no indication of achievement of a glassy phase during quenching

Yu et al. and Vuorinen et al. (2007)

Sn-Cu-Ni

Ref. 12

CALPHAD

isothermal at T = 240 °C and 250 °C

database constructed based on the experimental works of Wang et al., Lin et al., Oberndorff and Ho et al.

Fig. 3

Isoterms the Sn-Cu-Ni system reported in the literature

All the experimental works (and simulation) reported a high solubility of Ni and Cu in the phases Cu6Sn5 and Ni3Sn4, respectively. The solubility for both elements is temperature dependent. For instance, Ni has a solubility between 29 at. pct at 400 °C to 24 at. pct Ni at 220 °C in (Cu,Ni)6Sn5.[13] The solubility of Cu in (Ni,Cu)3Sn4 ranges from 4 at. pct at 700 °C to 6 at. pct at 220 °C.[26]

Critical Review of the Experimental Works Reported in the Literature

In the opinion of the authors, the discrepancies in the experimental position of the liquidus lines observed in the in Figures 1 through 3 are related mainly to the method of chemical analysis. For experimental works that used wet chemical analysis to determine the composition of the liquid phase (e.g., Hanson et al.), there was no detailed information regarding the procedure, number of duplicates or standard deviation of the measurements. Furthermore, there was no evidence that the selected sample for analysis was free of primary crystals phases. In addition, the modern experimental technique of Atomic Absorption (A. A) and Inductively Coupled Plasma (I.C.P) are more appropriate to analyze the relatively low concentration of Cu and Ni reported in the liquid phase.

On the other hand, the experimental work to obtain liquidus lines for the ternary Sn-Cu-Ni system relied upon the determination of chemical composition of the liquid phase in EPMA analysis as is indicated in Table II. There are two difficulties within the application of this technique for the system. Firstly, the accuracy of the analysis is not satisfactory for the relatively small concentrations of nickel and copper in the liquid phase. And, most importantly the analytical technique is only adequate for homogeneous samples, in this case requiring the formation of a glassy metal during the cooling of the sample. There was no evidence in the references mentioned in Table II that this condition was fulfilled. Indeed, evidence of non-glassy metals can be clearly observed e.g., in Figure 18 of the paper of work of Lin et al.[23] An example of a large discrepancy in the composition of the liquid phase is also observed in the work of Lin at temperature of 240 °C, with values of 0.6 and 1.1 wt pct for copper and Ni, respectively. It is also appreciable in Figures 1 and 3, where there is a large variation in composition within the experimental data reported by Gourlay et al [15] in the binaries and ternary systems.

Experimental Methodologies for Equilibrium Studies of Sn-Cu-Ni Alloy

The experimental methodologies used in the experimental works reported in the literature are indicated in Table II. In general, this corresponded to the equilibria of samples in silica ampoules or graphite, held isothermally then quenched in cold water.[23,24,26] The samples were analyzed with XRD and the reported compositions of solid and quenched liquid with EPMA. On the report of the invariant line, the technique used was based on the path of solidification close to equilibrium conditions. The methodology of EPMA analysis for the liquid phase required the formation of a glassy (amorphous) metallic material which is difficult to obtain by quenching in a liquid and there is no evidence in the published paper that the condition was achieved.

An alternative experimental approach to obtain equilibrium information is based on the differences of densities between the liquid and the IMC,[15,16] allowing settling of the IMC on the bottom of the crucible. Then, Scanning Electron Microscopy (SEM) with energy-dispersive X-ray spectroscopy (EDX) can be used to identify the IMC composition while the liquid phases compositions can be determined by Inductively Coupled Plasma-Atomic Emission Spectrometry (ICP-AES) or titration chemical analysis.

Objective of the Present Work

The present work was a starting point to address the incomplete and contradictory information of equilibrium data on the Sn-Cu-Ni alloy and establish an experimental protocol for a modified equilibria/quenching/Electron Probe Microscopy Analysis (EPMA)-Inductively Coupled Plasma (ICP-AES) analysis methodology. The aim of this project includes
  • To identify the underlying reasons for the disparity in the existing results in the literature.

  • To design a new experimental approach using state of the art phase equilibrium methodologies and analytical chemical analysis procedures.

  • To generate a new Sn-Cu-Ni phase diagram based on reliable experimental data including four liquidus lines at 240 °C, 250 °C, 260 °C, and 270 °C.

  • Characterize the invariant (or invariants) line for the equilibrium between a liquid phase with two IMCs

Corroboration of Attainment of Homogeneous Liquid During Fast Quenching

As indicated in Table II, most of the reported experiments relate to equilibrium of the Sn-Cu-Ni alloy obtained by fast quenching to form aglassy liquid phase. The present work explored the possibility of obtaining that glassy material. For that purpose, a Sn-0.5 wt pct Cu sample with dimensions 10 mm × 10 mm × 0.1 mm was set between two thin slides of quartz (0.2 mm thickness) fixed at a 0.1-mm aperture as shown in Figure 4(a). The sample was heated up to 250 °C for twenty hours and then was dropped directly into iced water. As indicated in Figure 4(b), even the most effectively quenched area in the sample was heterogeneous with clearly observable small crystal precipitates. EPMA point analysis of the indicated line at a distance between points of 2 μm indicated a standard deviation on the concentration of copper larger than the starting concentration of 0.5 wt pct Cu. Given the impossibility of obtaining a glassy material for a simple quenching technique, we ruled out this methodology and developed the experimental procedure described in the following section.
Fig. 4

Experiments of fast quenching/EPMA analysis of liquid phase. (a) Sample between two thin slides of quartz, (b) SEM image of quenched sample

Experimental Procedure

The experimental methodology adopted in the present work is based on the difference of densities between IMCs and the liquid phase. The densities of Cu6Sn5 and Ni3Sn4[27] are 8.1 and 8.64 g/cm3, respectively. The density of the liquid close to the eutectic point is 7.31 g/cm3 (Thermo-calc. calculation). Thus, it can be expected that a clear physical separation by sedimentation of the IMC will occur during liquid/IMC equilibration. Based on these considerations, the experiments were designed to:
  • Produce a few large crystals by selection of initial composition to allow an easy physical separation between crystals and the liquid phase.

  • Promote relatively fast quenching to distinguish the primary phase from what would appear during quenching.

  • Ensure the sample is large enough to easily manipulate and to obtain between 0.1 and 0.2g for the ICP-AES analysis. The target parameters and verification of the experimental procedure and a summary of parameters used during the experimental work are indicated in Table III. The starting material included 99.99985 pct purity Sn chips, 99.98 wt pct purity Ni chips, 99.99 wt pct purity Cu chips (provided by the sponsor of the present work). The electric resistance furnace, schematically described in Figure 5, consist of 6-mm-diameter quartz tubes, an input controller with two Resistance Temperature Detectors (RTD) 3200 series and a power input controller Modal 3216. The electronic scale has an accuracy of 4 decimal digits.

Table III

Targets for Experimental Methodology

Parameter

Technique

Target of Accuracy

Purity of metals

high purity materials

purities of materials are > 99.9 pct

Temperature

controller

vertical furnace assembled to allow fast quenching

hot zone > 0.5 mm

control of temp. ± 0.1 °C

Method of analysis

 IMC

 Liquid phase

electron probe micro analyze (EPMA) for IMC

inductively Coupled Plasma (ICP) for the liquid phase

crystal composition accuracy better an 1 wt pct for major composition (Sn) and 0.01 wt pct for minor compositions (Ni and Cu)

Achievement of equilibrium and settlement of large intermetallic phases

slow cooling from 700 °C to the target equilibrium temperature

long-time of equilibration (> 12 hours) at the target temperature ± 1 °C

equilibrium for 1 hour at the target temperature ± 0.1 °C

observation of large size (> 100 µm) intermetallic compounds

no intermetallic compounds observed on top of the sample

Fig. 5

Schematic representation of furnace and images of a sample. (a) Schematic representation of furnace, (b) a typical quenched sample, (c) SEM image of a section of the sample, (d) SEM images of upper and bottom side of a section of the sample

The RTDs were also calibrated in accordance with the National Australian Testing Association (NATA) standard. The accuracy of control of temperature was tested at the melting point of Sn. Samples of high purity (99.99985 pct) Sn chips were placed into the quartz tubes and sealed under vacuum. The sealed quartz tube was then placed into the tube furnace and positioned at the same level as the tip of the RTD in the hot zone. The sample was heated ~ 0.54 °C above the melting point (reported for Kahlbaum as 231.92 °C[28]) and a clear liquid phase was observed. Conversely, the sample at 231.39 °C was solid. Assuming that the activation energy of melting of Sn does not affect the outcome, this experiment indicated that the temperature can be controlled at a level better than ± 0.55 °C.

Cu and Ni were enveloped in a foil of Sn and placed in the bottom of a silica crucible and on top of the envelope; high purity Sn was added. The initial compositions were targeted to obtain a low proportion of IMC phases. The samples were positioned in the hot zone and the temperature in the tube furnace was slowly increased until it approached a preheating temperature of 700 °C. The preheat temperatures were maintained for at least 10 hours before slowly reducing the temperature at a rate of approximately − 2.5 × 10−3 °C/s until the target temperature was reached with an accuracy of ± 2 °C. The sample was left at the desired temperature for another 12 hours and finally was held for an one hour at the target temperature with an accuracy of ± 0.1 °C. The samples were then fast quenched in water. One-half of the samples were polished for microscopic observation and EPMA analysis of the IMC(s), while the other half of the sample was cut transversely into two sections and the top (and in selected instances the bottom) sections were analyzed by ICP-AES. Figure 5 includes a typical sample obtained after quenching and the respective microscopy image of the top and bottom sides.

The elemental analysis of the IMC was undertaken using a JEOL JXA 8200L (trademark of Japan Electron Optics Ltd., Tokyo) electron probe X-ray microanalyzer (EPMA) with wavelength dispersive detectors (WDD). In the setup of the EPMA device, an acceleration voltage of 15 kV and a probe current of 15 nA were used. The Duncumb-Philibert ZAF correction procedure supplied with the JEOL JXA 8200L probe was also applied. Standards of Cu6Sn5 were prepared for high-purity metals, and Ni (Charles M. Taylor, Stanford, CA) was used in the EPMA measurements. Eight points were analyzed in each sample; the standard deviation of measured compositions of the standards and samples indicated an accuracy of the elemental analysis of ± 0.2 wt pct.

It was confirmed using microscopy that the top of the samples were liquid before quenching. In order to analyze the sample by triplicate, the liquid phase (top of the samples) was cut in three pieces each one with a weight of approximate 0.1 g. The samples were dissolved in “aqua regia,” diluted with water in a volumetric of 25 mL and analyzed using ICP-AES. To corroborate the accuracy of results, a number of samples of selected composition to assure the formation of liquid phases (without intermetallic compounds) were melted at 400 °C, fast quenched and analyzed by ICP-AES. The results shows in Table IV indicate that the accuracy of the analysis is of the order of 0.05 wt pct for Cu and 0.01 wt pct for Ni. Furthermore, the triplicate analysis of the liquid phases in equilibrium with IMC indicated standard deviations of the order of 0.05 for experiments in the binary Sn-Cu and 0.01 for the experiments in the binary Sn-Ni.
Table IV

ICP Analysis of Selected Samples

Cu (Weight Percent)

Ni (Weight Percent)

Initial

ICP Analysis

Initial

ICP Analysis

0.44

0.36

0.08

0.07

0.48

0.40

0.10

0.09

0.50

0.52

0.12

0.11

0.55

0.53

0.15

0.14

0.80

0.78

0.00

0.00

0.00

0.001

0.00

0.00

1.80

1.83

0.00

0.00

0.00

0.001

1.00

1.02

Experimental Results and Discussion

A total of 30 experiments will be reported which fulfil the selection criteria based on the following observations:
  • A low proportion of IMCs in the sample

  • Good sedimentation with the IMCs of large size (> 100 μm) accumulating in the bottom of the sample

  • A large liquid area in the upper side of the sample free of IMCs. The experimental results for the Sn-Cu and Sn-Ni binaries are summarized in Tables V and VI, respectively. The tables include the initial composition, the equilibrium composition of liquid and the composition of the IMC phases and also the results of experimental points where only a liquid phase was observed. Figures 6 and 7 show the equilibrium results for the binaries and the discrepancies between these results and those of Hanson et al.,[14] Harding et al.,[11] and the Thermo-calc. simulation for the system Sn-Cu. The differences of the results of the present work with those of Hanson et al.[14] and the Thermo-calc. simulation for the Sn-Ni system are also apparent.

Table V

Composition for Liquid and IMC Phases for the Sn-Cu Binary System

No.

Tem. °C

Initial (Weight Percent)

Liquid (Weight Percent)

IMC (Normalized Weight Percent)

Cu

Cu

Sn

Cu

IMC

1

230

1.4

1.02

61.7

38.3

Cu6Sn5

2

240

1.8

1.20

60.1

39.9

Cu6Sn5

3

250

1.8

1.36

60.6

39.4

Cu6Sn5

4

260

1.8

1.48

59.8

40.2

Cu6Sn5

5

270

1.8

1.61

62.7

37.3

Cu6Sn5

6*

270

1.6

1.58

n/a

n/a

n/a

n/a, not applicable because only the liquid phase was observed

*Only the liquid phase was observed

Table VI

Composition for Liquid and IMC Phases for the Sn-Ni Binary System

No.

Tem. °C

Initial (Weight Percent)

Liquid (Weight Percent)

IMC (Normalized Weight Percent)

Ni

Ni

Sn

Ni

IMC

7

240

0.25

0.16

72.8

27.2

Ni3Sn4

8

250

0.25

0.17

74.5

25.5

Ni3Sn4

9*

250

0.15

0.16

n/a

n/a

n/a

10

260

0.25

0.19

73.9

26.1

Ni3Sn4

11*

260

0.15

0.16

n/a

n/a

n/a

12

270

0.5

0.2

72.8

27.2

Ni3Sn4

13*

270

0.15

0.16

n/a

n/a

n/a

14

280

0.3

0.22

74.2

25.8

Ni3Sn4

n/a, not applicable because only the liquid phase was observed

*Only the liquid phase was observed

Fig. 6

Section of the Sn-Cu binary system including the present experimental points and the proposed liquidus line for the equilibrium Cu6Sn5; the figure includes the liquidus lines of Thermo-Calc. simulation and that proposed by Harding et al

Fig. 7

Section of the Sn-Ni binary system including the present experimental points and the proposed liquidus line for the equilibrium Cu6Sn5; the figure includes the liquidus lines of Thermo-Calc. simulation and that proposed by Hanrson et al

A typical example of the observed coexistence of liquid/IMC is shown in Figure 8 and both the large size of the IMCs and the presence of two IMCs are apparent in Figure 8(c). The experimental results of liquidus lines for the ternaries are shown in Table VII. The isotherm at 240 °C, 250 °C, 260 °C, 270 °C is plotted in Figure 9 which includes the lines of separation between the primary phase fields of (Ni,Cu)3Sn4 and (Cu,Ni)6Sn5. These lines were obtained by extrapolating of the tie lines for the equilibrium liquid/(Ni,Cu)3Sn4/(Cu,Ni)6Sn5 at the respective temperature.
Fig. 8

SEM images of samples. (a) SEM image of the equilibrium liquid/(Ni,Cu)3Sn4, (b) SEM image of the equilibrium liquid/(Cu,Ni)6Sn5, (c) SEM image of the equilibrium liquid//(Ni,Cu)3Sn4/(Cu,Ni)6Sn5

Table VII

Composition for Liquid and IMC Phases for the Sn-Cu-Ni Ternary System

No.

Tem °C

Initial (Weight Percent)

Liquid (Weight Percent)

IMC (Normalized Weight Percent)

IMC

Cu

Ni

Cu

Ni

Sn

Cu

Ni

15

240

0.6

0.4

0.32

0.17

73.82

4.77

21.4

(Ni, Cu)3Sn4

62.47

23.3

14.2

(Cu, Ni)6Sn5

16*

240

0.44

0.08

0.36

0.074

n/a

n/a

n/a

n/a

17

250

0.4

0.4

0.33

0.18

73.42

4.39

22.2

(Ni, Cu)3Sn4

18*

250

0.48

0.1

0.4

0.09

n/a

n/a

n/a

n/a

19*

250

0.5

0.12

0.52

0.11

n/a

n/a

n/a

n/a

20

260

1.4

0.8

0.83

0.28

61.6

20.7

17.7

(Cu, Ni)6Sn5

21

260

1.2

0.2

1.07

0.13

73

2.1

24.9

(Ni, Cu)3Sn4

22

260

0.4

0.5

0.42

0.26

73.74

3.48

22.8

(Ni, Cu)3Sn4

23

260

0.7

0.6

0.47

0.28

73.98

3.77

22.3

(Ni, Cu)3Sn4

63.06

22.2

14.7

(Cu, Ni)6Sn5

24

260

1

0.2

0.85

0.16

62.92

25.6

11.5

(Cu, Ni)6Sn5

25

260

1.1

0.4

0.47

0.3

74.01

3.79

22.2

(Ni, Cu)3Sn4

26

270

1.2

0.2

1.07

0.14

61.93

25.68

12.4

(Cu, Ni)6Sn5

27

270

1

0.3

0.96

0.19

62.74

22.26

15.0

(Cu, Ni)6Sn5

28

270

0.4

0.5

0.4

0.3

73.05

2.3

24.7

(Ni, Cu)3Sn4

29

270

0.7

0.6

0.5

0.31

74.14

3.36

22.5

(Ni, Cu)3Sn4

62.79

22.08

15.3

(Cu, Ni)6Sn5

30*

270

0.55

0.15

0.53

0.14

n/a

n/a

n/a

n/a

n/a, not applicable because only the liquid phase was observed

*Only the liquid phase was observed

Fig. 9

Experimental points for the equilibrium liquid/IMCs. (a) Experimental points at 240 °C, (b) experimental points at 250 °C, (c) experimental points at 260 °C, (d) experimental points at 270 °C

The composition of the liquidus and the corresponding connection with the IMC are included in Figures 10 through 13 with projections at two different scales: one for the liquidus from 0 to 2 wt pct of Ni and Cu in order to appreciate the details of the liquidus lines and another scale from 0 to 40 wt pct of Ni and Cu to indicate the composition of the IMCs.
Fig. 10

Equilibrium liquid/IMCs at 250 °C

Fig. 11

Equilibrium liquid/IMCs at 260 °C

Fig. 12

Equilibrium liquid/IMCs at 270 °C

Fig. 13

Equilibrium liquid/IMCs at 280 °C

The compositions of the IMCs are summarized in Figure 14 in at. pct. At the experimental condition, the maximum solubilities of copper in (Ni,Cu)3Sn4 and nickel in (Cu,Ni)6Sn5 are around 8 at. pct Cu (4 wt pct) and 24 at. pct Ni (18 wt pct), respectively. These solubilities are in agreement with those reported by Schmetterer et al.[26]
Fig. 14

Composition of IMC indicating the limit of solubility of (Ni,Cu)3Sn4 and (Cu,Ni)6Sn5

The proposed ternary phase diagram is shown in Figure 15, including the liquidus lines reported by Vourinen et al.,[13] Yu et al.,[12] and Gourlay et al.[15] The position of the invariant line was clearly determined. The eutectics of the Sn-Cu and Sn-Ni binaries were extrapolated from the present experimental work and in agreement with the reported temperatures included in Table I. It is clear that the discrepancies with previous experimental works and simulations were the liquidus lines; within the present work it has been extended till 0.31 wt pct Ni at 270 °C. The position of the invariant line is also substantially different from that suggested by Vourinen et al.[13] and Yu et al.[12]
Fig. 15

Proposed phase diagram; the figure includes for comparison the liquidus lines proposed by Vourinen et al., Yu et al. and Gourlay et al

Conclusions

The main objective of the present work was generating a new Sn-Cu-Ni phase diagram based on reliable experimental data. This objective required the improvement of reported methodologies to overcome the experimental shortcomings observed in the literature review. The developed equilibration methodology allowed the formation of large sized intermetallic compounds, a clear solid and liquid phase separation and the analysis of crystals in the bottom of the sample. A careful separation of the sample then allowed the associated phases of the liquid and solids to be elementally analyzed by ICP-AES and EPMA, respectively. In addition, the separation of phases was corroborated using electronic microscopy to observe the microstructure of solids formed in the bottom of the sample.

A consistent dataset of experimental data was obtained. The phase diagram included the liquidus lines at temperatures of 240 °C, 250 °C, 260 °C, and 270 °C, the invariant line of the diagram, and the tie lines of connections of the liquid phase with the intermetallic compounds. Furthermore, the phase diagram was also plotted along with the data reported in the literature. It is clear that there are considerably discrepancies with the work-previous experiments and simulations.

References

  1. 1.
    J. W. Evans and W. Engelmaier: A Guide to Lead-free Solders: Physical Metallurgy and Reliability, Springer, London, UK, 2007, pp. 1-8.Google Scholar
  2. 2.
    K. N. Subramaniam: Materials reliability for electronics, Wiley Chichester, West Sussex, 2012, pp. 233-250.Google Scholar
  3. 3.
    O. A. Ogunseitan: J. Met, 2007, vol. 59, pp. 12-17.Google Scholar
  4. 4.
    A. Kroupa, A.T. Dinsdale, A. Watson, J. Vrestal, J. Vízdal and A. Zemanova (2007) J. Met., 59, 20-25.Google Scholar
  5. 5.
  6. 6.
    A.T. Dinsdale, A. Watson, A. Kroupa, J. Vrestal, A. Zemanova, J. Vizdal: Atlas of phase diagrams for Pb-free soldering. 2008, Action number: 531, vol. 1, The European Science Foundation, and MTSOLDERS. v.1.0 database.Google Scholar
  7. 7.
    Thermo-calc. software tcsld3-2_extended_info http://www.thermocalc.com/media/19852/tcsld3-2_extended_info.pdf. Accessed 22 Oct 2017.
  8. 8.
    MTDATA-Phase Diagram Software from the National Physical Laboratory, SOLDERS: Database for lead free solders http://resource.npl.co.uk/mtdata/soldersdatabase.htm. Accessed 22 Oct 2017.
  9. 9.
    T. Laurila, V. Vuorinen, and M. Paulasto-Kröckel: Mater. Sci. Eng. R, 2010, vol. 68, pp. 1-38.CrossRefGoogle Scholar
  10. 10.
    C. E. Ho, R. Y. Tsai, Y. L. Lin and C. R. Kao: J. Electron. Mater., 2002, vol. 31, pp. 584-590.CrossRefGoogle Scholar
  11. 11.
    J. V. Harding, W. T. P. Walpole (1948) ASM Int., 75, 115-130.Google Scholar
  12. 12.
    H. Yu, V. Vuorinen and J. K. Kivilahti: J. Electron. Mater., 2007, vol. 36, pp. 136-146.CrossRefGoogle Scholar
  13. 13.
    V. Vuorinen, H. Yu, T. Laurila and J. Kivilahti: J. Electron. Mater., 2008, vol. 37, pp. 792-805.CrossRefGoogle Scholar
  14. 14.
    D. Hanson, E. J. Sandford and H. Stevens: J. Inst. Met., 1934, vol. 55, pp. 115-133Google Scholar
  15. 15.
    C. M. Gourlay, K. Nogita, J. Read and A. K. Dahle: J. Electron. Mater., 2010, vol. 39, pp. 56-69.CrossRefGoogle Scholar
  16. 16.
    H. S. Liu, J. Wang and Z. P Jin: Calphad, 2004, vol. 28, pp. 363-370.CrossRefGoogle Scholar
  17. 17.
    C. Schmetterer, H. Flandorfer, K. W. Richter, U. Saeed, M. Kauffman, P. Roussel and H. Ipser: Intermetallics, 2007, vol. 15, pp. 869-884.CrossRefGoogle Scholar
  18. 18.
    A. Zemanova, A. Kroupa and A. Dinsdale: Monatshefte Für Chemie-Chemical Monthly, 2012, vol. 143, pp. 1255-1261.CrossRefGoogle Scholar
  19. 19.
    P. Nash and A. Nash: Bull. Alloy Phase Diagrams, 1985, vol. 6, pp. 350-359.CrossRefGoogle Scholar
  20. 20.
    C. M. Gourlay, K. Nogita, S. D. McDonald, T. Nishimura, K. Sweatman and A. K. Dahle: Scr. Mater., 2006, vol. 54, pp. 1557-1562.CrossRefGoogle Scholar
  21. 21.
    S. A. Belyakov and C. M. Gourlay: Intermetallics, 2012, vol. 25, pp. 48-59.CrossRefGoogle Scholar
  22. 22.
    G. Ghosh: Metall. Mater. Trans. A, 1999, vol. 30, pp. 1481-1494.CrossRefGoogle Scholar
  23. 23.
    C. R. Lin, S. W. Chen and C. H. Wang: J. Electron. Mater., 2002, vol. 31, pp. 907-915.CrossRefGoogle Scholar
  24. 24.
    L. Snugovsky, P. Snugovsky, D. Perovic and J. Rutter (2006) Mater. Sci. Technol., 22, 899-902.CrossRefGoogle Scholar
  25. 25.
    S. Chen, C. Wang, S. Lin and C. Chiu: J. Mater. Sci., 2007, vol. 18, pp. 19-37.Google Scholar
  26. 26.
    C. Schmetterer, H. Flandorfer, C. Luef, A. Kodentsov and H. Ipser: J. Electron. Mater., 2009, vol. 38, pp. 10-24.CrossRefGoogle Scholar
  27. 27.
    D. Mu: Critical properties of Cu6Sn5 for high-temperature and 3D Pb-free soldering applications, Ph.D. Thesis, The University of Queensland, Brisbane Australia, 2012.Google Scholar
  28. 28.
    Waidner & Burgess, 1910, Bur. Stand. (U. S.), Bull., vol. 6, pp. 149-230.CrossRefGoogle Scholar

Copyright information

© The Author(s) 2018

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors and Affiliations

  • Hector M. Henao
    • 3
    Email author
  • Ching-shun Chu
    • 2
  • Juan P. Solis
    • 3
  • Kazuhiro Nogita
    • 1
  1. 1.Nihon Superior Centre for the Manufacture of Electronic Materials (NS CMEM), School of Mechanical and Mining EngineeringThe University of QueenslandSt. LuciaAustralia
  2. 2.School of Mechanical and Mining EngineeringThe University of QueenslandSt. LuciaAustralia
  3. 3.Metallurgy and Materials Engineering DepartmentTechnical University Federico Santa MaríaValparaísoChile

Personalised recommendations