Warning: fopen(/home/virtual/kwjs/journal/upload/ip_log/ip_log_2024-03.txt): failed to open stream: Permission denied in /home/virtual/lib/view_data.php on line 88 Warning: fwrite() expects parameter 1 to be resource, boolean given in /home/virtual/lib/view_data.php on line 89 Laser Soldering Properties of MEMS Probe for Semiconductor Water Testing

J Weld Join > Volume 39(4); 2021 > Article
Hong, Kim, Kim, Yun, and Park: Laser Soldering Properties of MEMS Probe for Semiconductor Water Testing

Abstract

In this study, the laser soldering process of microelectromechanical system (MEMS) probes and multilayer ceramic (MLC) substrates was optimized with Type 4 (T4) and Type 7 (T7) Sn-3.0Ag-0.5Cu (SAC305), Type 4 Sn-0.3Ag-0.7Cu (SAC0307), and Type 4 Sn-5.0Sb (SnSb) soldering, and the probe solder joint properties were compared. SAC0307 and SnSb were used to confirm the bonding property of the low-Ag solder and the high-temperature durability. We conducted a thermal cycling test (TCT) with 500 cycles at temperatures of -25 ℃ to 105 ℃ with a 10-min dwell time at each temperature. Before and after the TCT, the microstructure of solder joint as well as the shear strength between the probe and the MLC were examined. The results revealed that, after the TCT, the degradation rate of T7 SAC305 solder was lower than that of T4 SAC305, and the SnSb solder did not exhibit bonding strength degradation. The fracture mode of the SAC solder joint was ductile-brittle. In the case of the SnSb joint, brittle fracture was the major fracture mode because of the Sn-Sb intermetallic compounds. The SnSb solder had an excellent bonding strength and degradation property after the TCT, but the lack of toughness caused brittle fracture. These results confirm the applicability of T4 SnSb and T7 SAC305 solders for high-temperature response and fine pitch bonding.

1. Introduction

A probe card is used as a signal transmission interface used in the wafer inspection process for electrical go/no- go testing of individual chips in the fabricated wafers during the semiconductor manufacturing process. The probe card is a product that is individually designed and fabricated depending on the specifications of the semiconductor device under test (DUT). Since the probe card needs to be newly fabricated for each time the specifications of the device changes, it is categorized as a high value-added product. Also, the probe card is one of the core components for testing fabricated wafers during the semiconductor manufacturing process1). In terms of the shape of a probe card, a contact pin or probe with the shape of a very fine line is attached to a Printed Circuit Board (PCB) of a set specification for electrical go/no-go testing of chips in the wafer, and the probe is electrically connected to the wafer, serving as the interface between the wafer and the test system electronics. In this way, 100% inspection is performed for screening defective semiconductor chips.
In recent trends of probe card technology, memory capacity density continues to increase with the miniaturization of semiconductor process nodes, along with the shrinking of chip size with the same capacity and increase in the number of net die-per-wafer1,2). Accordingly, the pad pitch of the IC that the probe card needs to measure decreases, and the number of pins rapidly increases1-5).
Also, for automotive electronics, with increasing proportion of semiconductor devices (ICs) requiring testing under high-temperature environment, the temperature in the wafer testing during the semiconductor manufacturing process has increased to more than 120 ℃. Consequently, there is a pressing need for high temperature durability in the joint between the probe and solder of the substrate.
As shown in Fig. 1 below, the general structure of a probe card consists of a MEMS (microelectromechanical systems) probe that applies electrical signals to the panel and semiconductor device patterns, multilayer ceramic (MLC), PCB, interposer, connector, jig and fixture. The probe card should have low contact resistance, as well as durability, service life for more than 100,000 times of cyclic fatigue tests (Fatigue test or Touch down (T/D)), and small thermal deformation under -40℃-100℃1).
Fig. 1
Schematic diagram of MEMS probe solder joint
jwj-39-4-368gf1.jpg
The MEMS probe card is joined on the MLC pad by laser (Light Amplification by the Stimulated Emission of Radiation) soldering using Pb-free solders. For methods of MEMS soldering, the conventional reflow method can be used, but in this case, thermal stress is applied to the MLC substrate, leading to high possibility of short circuit and substrate fracture. Therefore, selective soldering method is used in which the MEMS probe is bonded to the individual pad of MLC substrate using laser. In the laser soldering process, soldering is performed by individually applying heat through laser optics for each pad with a size of several tens to hundreds of ㎛. The advantage of this method is that thermal and mechanical stresses are not applied to the MLC substrate during the soldering process. In terms of solders used for probe soldering, there has been a recent trend of shift from Pb-Sn solder to eco-friendly Pb-free solder. The mainly used Pb-free solders include Sn-3.0Ag-0.5Cu (SAC305), and for improvement in the high temperature durability, application of solders such as Sn-0.7Cu or Sn-5Sb with high melting points is being investigated.
Therefore, in this study, a laser soldering process for probe-MLC soldering was developed taking into account the high temperature durability and responses to fine pitch requirements. In the developed laser soldering process, SAC305 Type 4 (T4) (Powder size: 20-38 ㎛) and Type 7 (T7) (Powder size: 2-11 ㎛) solder pastes were used. In addition, Sn-0.3Ag-0.7Cu (SAC0307) T4 solder was used for comparison of bonding properties of solders with low-Ag, and for implementation of high-temperature durability, Sn-5.0Sb (SnSb) T4 solder was used to perform the probe soldering process. Next, thermal shock tests were performed to compare the high temperature durability of the soldered joints, the shear strength of the probe solder joint was measured before and after the test, and the soldering properties were comparatively analyzed through the cross-sectional analysis of the microstructure of the as-soldered joints.

2. Methods

2.1 Probe, MLC and Pb-free solder paste

Fig. 2 presents the images of the probe and MLC substrate used in the experiment. For the MEMS probe, nickel alloys plated with gold (Au) at a thickness of 25-34 ㎛ was used. The dimension of the MLC substrate was 25 (L) × 25 (W) mm, that of the pad for bonding to the probe was 750 (L) × 120 (W) ㎛, and the pad pitch size was 250 ㎛. The thickness of the Cu pad was approximately 23 ㎛, and the surface of Cu pad was treated with Ni/Au electro-plating.
Fig. 2
SEM micrograph of (a) MEMS probe and optical stereo-microscope image of (b) MLC
jwj-39-4-368gf2.jpg
As for types of solders used in the probe soldering process, a total of 4 types of solder paste were used as outlined in Table 1 below. They include T4 SAC305 solder, which is the typical solder currently used, T7 SAC305 for fine pitch requirement, and T4 SAC0307 solder for comparison of improvement in the thermal fatigue. For investigation of high temperature durability, SnSb solder was used for the soldering process.
Table 1
Chemical composition, type, and powder size of Pb-free solder pastes
Solder alloy composition Solder paste type Powder size (㎛) Manufacturer
Sn-3.0Ag-0.5Cu (SAC305) T4 20-38 Ecojoin Co., Ltd.(Korea)
T7 2-11 Duksan Hi-Metal Co., Ltd.(Korea)
Sn-0.3Ag-0.7Cu (SAC0307) T4 20-38 Ecojoin Co., Ltd.(Korea)
Sn-5.0Sb (SnSb)

2.2 Laser Soldering Process and Laser Soldering Profile

A soldering process using a laser was performed for the probe-MLC bonding process. Fig. 3 is a photograph depicting the laser soldering process. The fiber laser source used in the experiment is Pearl™ of nLight with a wavelength of 808 nm, a wave tolerance range of ±3 ㎛, and a power output of 55 W.
Fig. 3
Photographs of (a) laser soldering process with (b) MEMS probe and MLC
jwj-39-4-368gf3.jpg
The laser soldering profile used in the laser soldering process is shown in Fig. 4. Fig. 4(a) shows the laser profile for SAC305 solder, and Fig. 4(b) shows the laser profile for SnSb solder. The soldering process was performed by irradiating the laser for a total of 2.5s for SAC305 solder and 3s for SnSb solder. In the case of SnSb solder with a melting point of 232-240 ℃, higher compared to SAC305 solder with a melting point of 217 ℃, the laser was additionally irradiated by 0.5s at 16 A. Due to the high melting point of the solder, additional irradiation of a high-power laser was required to induce complete melting of the solder.
Fig. 4
Laser soldering profiles for (a) SAC305 and (b) SnSb solders
jwj-39-4-368gf4.jpg

2.3 Thermal Cycling Test (TCT)

For analysis of the bonding strength degradation property of the probe joint, 500 cycles of TCT was performed under the conditions of -25-105 ℃ and 10 min holding time for each temperature condition5,6). Then, the shear strength of the probe joint was measured. In general, the shear strength standard after TCT is within about 40% of the shear strength of the as-soldered joints.

2.4 Measurement of Solder Joint Strength

For measurement of the bonding strength of the ProbeMLC solder joint, a shear strength test was performed as shown in Fig. 5 using Dage 4000 (Nordson Co. Ltd., USA) bondtester. The shear strength test was performed under the conditions of a test speed of 50 ㎛/sec and the height of the test fixture at 500 ㎛ from the substrate, and the bonding strength was comparatively analyzed with respect to different solder alloy composition and solder paste types.
Fig. 5
Photographs of shear strength measurement of probe solder joint, (a) overview of probe-MLC bonded coupon and
jwj-39-4-368gf5.jpg

3. Results

3.1 Bonding Strength Analysis of Probe Solder Joint

For the analysis of the degradation property in the bonding strength of the probe joint, after 500 thermal cycles, the shear strength of the probe solder joint was measured and the level of degradation was comparatively analyzed against the shear strength of the as-soldered joint. The result was presented in Fig. 6, and in all samples, degradation within 0%-17% was observed compared to the shear strength of the as-soldered joint, indicating good bonding state. The shear strength of T4 and T7 SAC305 solder joint was 54.9 gf and 48.7 gf, respectively, and after TCT, the shear strength was decreased to 45.5 gf and 46.6 gf, respectively. Although T4 SAC305 showed slightly higher shear strength of the as-soldered joint than T7 SAC305, the degradation rates after TCT were 17.1% and 5.5%, respectively, indicating that T7 solder showed a lower degradation rate. The shear strength of T4 SAC0307 decreased to 42.7 gf after TCT from 48.2 gf at as-soldered joint, indicating the degradation rate of shear strength at 11.4%. For T4 SAC0307 solder joint, the shear strength of the as-soldered joint was slightly smaller than that of T4 SAC305 but the degradation rate after TCT was smaller than that of T4 SAC305, and the as-soldered joint shear strength was similar level to that of T7 SAC305. In the case of T4 SnSb solder joint, the as-soldered joint shear strength was lower than that of T4 SAC305, but was similar level to that of T4 SAC0307 and T7 SAC305 joints. Furthermore, considering that there was no bonding strength degradation after 500 thermal cycles in TCT, it can be seen that T4 SnSb solder has excellent high temperature durability and degradation property.
Fig. 6
Shear strength comparison of laser solder joints before and after thermal cycles
jwj-39-4-368gf6.jpg

3.2 Cross-sectional Analysis of As-soldered Probe- MLC Joint

The results of cross-sectional analysis of each type of solder joint after performing optimization of the probe- MLC bonding process using 4 types of solder are pre- sented in Figs. 7 and 8 below. Fig. 8 is the magnified SEM images of the probe-solder joint and solder-MLC joint interface. In Fig. 7 (a-h) is the cross-sectional SEM images after probe-MLC bonding using T4 SAC305 (a, , T4 SAC0307 (b,c), T4 SnSb (e,f), and T7 SAC305 solder (g,h). Fig. 8 (a-h) is the SEM images of the probe-MLC joint using 4 types of solders, respectively, and Fig. 8 (i,j) represents the magnified SEM image and EDS component analysis result for T7 SAC 305 solder joint interface. It can be seen that fillets were well formed in the probe solder joints, and various types of intermetallic compounds (IMC) were formed on the interface of probe solder joints. Regardless of the type, in the Pb-free solder joint of SAC305 and SAC0307, Ni3Sn4 (Cu,Ni)6Sn5 and (Cu,Ni,Au)6Sn5 IMCs were formed on the interface of probe solder joint, and on the solder base metal, Ag3Sn, AuxSny, Cu6Sn5, and (Cu,Ni,Au)6Sn5 IMCs were formed7-15). Since the surface of the probe is plated with Ni/Au, various types of Au-Sn IMCs were observed inside the solder joint. There are four types of Au-Sn IMCs that can be formed inside the solder joint: AuSn, Au2Sn, AuSn2, and AuSn 8). In Fig. 7 (b,d,h), comparing the microstructure formed inside the SAC305 solder, it can be seen that less Ag3Sn IMCs are formed with SAC0307 solder than with SAC305 solder. This is because the amount of Ag3Sn formed in the solder base metal varies with the difference in the Ag content contained in the solder. According to Hong7) and Suh8), it was reported that the fraction of primary-Sn is up to 11% for Sn-3.0Ag-0.5Cu and up to 35% for Sn-1.0Ag-0.5Cu, depending on the content of Sn in the Sn-Ag-Cu alloy composition. With the decrease in the Ag content in the Sn-Ag-Cu alloy composition, the amount of primary-Sn formation is increased and the amount of Sn to form Ag3Sn decreases. Thus, the Ag3Sn fraction in the SAC0307 solder is smaller than that of SAC305 solder7). The formation of Ag3Sn IMCs positively contributes to the improvement of the bonding strength of as-soldered joints, but when exposed to high temperature for a long time, Ag3Sn IMCs grow and become coarse, causing a decrease in bonding strength7,10). For this reason, the result of bonding strength comparison after TCT shows that the bonding strength degraded more in the case of SAC305 solder than SAC0307 solder, and this is considered to be due to the coarsening of Ag3Sn IMCs formed on the base metal. On the interface of SnSb solder joint, formation of various IMCs including (Au,Ni,Cu)6Sn5, (Cu,Ni)3Sn, Ni3Sn4, Sb2Sn3, (Sb,Au)2Sn3, and AuSn4 were observed13-15). It can be seen that the formation of these IMCs in the as-soldered joints contributes to the improvement of bonding strength of as-soldered joint after soldering, and in SAC composition, Sn-Sb IMCs are considered to show positive contribution to the enhancement of the high temperature durability.
Fig. 7
Cross-sectional SEM micrographs of as-soldered probe-MLC joints, (a,b) T4 SAC305, (b,c) T4 SAC0307, (e,f) T4 SnSb, and (g,h) T7 SAC305
jwj-39-4-368gf7.jpg
Fig. 8
Magnified SEM images of as-soldered Probe-MLC joints, (a,b) T4 SAC305, (b,c) T4 SAC0307, (e,f) T4 SnSb, (g,h) T7 SAC305, and (i,j) magnified images of the probe and the MLC joint of T7 SAC305
jwj-39-4-368gf8.jpg
Fig. 9 presents SEM images of the fracture surfaces after measuring the initial shear strength of the as-soldered probe solder joint for 4 types of the solder paste. The solder joint fracture surface of all samples showed a pattern of the mixed presence of typical ductile fracture with multiple dimples and cleavage fracture, which is brittle fracture12). The fracture surface of T4 SAC305 solder joint showed ductile fracture with formation of dimples in the early stage of shear strength test, but from the middle of the test, brittle fracture patterns of cleavage fracture without ribs were shown. For T4 SAC0307 solder, ductile fracture with formation of dimples was observed in all fracture surfaces, which confirms that compared to SAC305 solder, less Ag3Sn IMCs were formed, thus the strength of ad-soldered joint is decreased but ductility has increased. For T4 SnSb solder, cleavage brittle fracture with almost no dimples was observed as the major fracture mode of the fracture surface, and some of the surfaces showed semi-cleavage fracture. SnSb solder is a solder with improved hightemperature durability due to the formation of Sn-Sb IMCs, and the image analysis showed that although bonding strength was increased, brittle fracture pattern was shown due to low ductility. In terms of bonding strength, SnSb solder has excellent bonding strength of the as-soldered joint and a small degradation rate after TCT, but brittle fracture was induced at surface fracture, due to lack of toughness of the solder compared to SAC composition14,15).
Fig. 9
SEM images of fracture surfaces of as-soldered probe joints after shear strength test, (a,b) T4 SAC305, (b,c) T4 SAC0307, (e,f) T4 SnSb, and (g,h) T7 SAC305
jwj-39-4-368gf9.jpg

4. Conclusion

In this study, for improving the high temperature durability of the probe solder joint and meeting the fine pitch requirements in the soldering process, T4 and T7 SAC305, T4 SAC0307, T4 SnSb solders were used to perform optimization of probe-MLC laser soldering process. With the developed laser soldering process, the bonding strength degradation rate was comparatively analyzed before and after the thermal cycling test (TCT), and the bonding property was analyzed with respect to changes in the solder alloy composition and powder size. The main findings of this study are outlined as follows
  • 1) T4 SAC305 showed slightly higher shear strength of as-soldered joint compared to that of T7 SAC305, but after TCT, the degradation rate of T7 SAC305 solder was lower than that of T4 SAC305. For SAC305 solder, the shear strength of the as-soldered joint was lower with T7 SAC305 than that of T4 SAC305, but the degradation rate of T7 SAC305 after TCT was lower than that of T4 SAC305. The shear strength of the as-soldered joint of T7 SAC305 was similar to that of T4 SAC0307. For T4 SnSb solder, the as-soldered shear strength was lower than that of T4 SAC305, but it was equivalent to that of T4 SAC0307 and T7 SAC305. Since no bonding strength degradation was observed after 500 thermal cycles of TCT, the excellent high temperature durability and degradation characteristics were confirmed with SnSb solders.

  • 2) In SAC305 and SAC0307 solder joints, (Cu,Ni,Au)6Sn5 IMCs were formed on the interface of the probe-MLC solder joint, and formation of Ag3Sn, Au3Sn, Cu6Sn5, and (Cu,Ni,Au)6Sn5 IMCs was observed inside the solder. When bonding strength was compared after TCT, higher rate of bonding strength degradation was observed with SAC305 compared to SAC0307. The reason for the difference is that since the amount of Ag3Sn formed in the base metal was larger with SAC305 solder than the amount with SAC0307 solder, the bonding strength decreased more with SAC305 after TCT due to the coarsening of Ag3Sn. On the interface of SnSb solder joint, various types of IMCs such as (Au,Ni,Cu)6Sn5, (Cu,Ni)3Sn, Ni3Sn4, Sb2Sn3, (Sb,Au)2Sn3, and AuSn4 were observed and it is considered that Sn-Sb IMCs have contributed to the improvement in the high temperature durability of the solder.

  • 3) In the fracture surface analysis of the SAC solder joints, mixed presence of ductile fracture and brittle fracture was observed. In the case of fracture surface of T4 SAC0307, ductile fracture was more dominantly observed compared to SAC305, and this is judged to be caused by increased ductility from less formation of Ag3Sn IMCs inside the solder. In the case of T4 SnSb joint, cleavage fracture, which is brittle fracture, was shown as the major fracture mode, and brittle fracture was observed due to low ductility although the hightemperature durability was improved due to the formation of Sn-Sb IMCs. SnSb solder is considered to show excellent properties in terms of bonding strength and degradation rate, but the result indicated that brittle fracture was caused due to lack of toughness of the solder.

Acknowledgments

This work was supported by the Technology Innovation Program Root(Ppuri) Technology Development Program for Quality Response of Global Core Industry (Grant No.: 20011705) by the Korean Ministry of Trade, Industry and Energy (MOTIE) in the Republic of Korea.

References

1. Korea Institute of Science and Technology Information (KISTI), Technical Tren of Semi-conductor Measurement and Inspection Equipment, The Ministry of Trade. Industry and Energy (MOTIE) and Materials &Component Technology Network (MCT-Net). Accessed on https://www.mctnet.org(2020)
2. G. T. Kim, G. H. Kang, and D. l Kwon, Effect of Material Property Uncertainty on Warpage during Fan Out Wafer-Level Packaging Process, J. Microelectron. Packag. Soc. 26(1) (2019) 29–33. https://doi.org/10.6117/kmeps.2019.26.1.0029
[CROSSREF] 
3. C. I. Chen, C. Y. Ni, C. M. Chang, S. C. Wu, and D. Liu, Bondability Study of Chip-on-Film (COF) Inner Lead Bonding (ILB) Using Conventional Gang Bonder, IEEE Trans. Electro. Packag. Manuf. 31(4) (2008) 285–290.
[CROSSREF] 
4. S. J. Kim, W. S. Hong, H. B. Nam, and N. H. Kang, Growth Behavior of Intermetallic Compounds in Various Solder Joints Induced by Electromigration, J. Weld. Join. 39(1) (2021) 89–102. https://doi.org/10.5781/JWJ.2021.39.1.11
[CROSSREF] 
5. W. S. Hong, W. S. Kim, B. S. Song, and K. B. Kim, Thermal Shock Cycles Optimization of Sn-3.0Ag-0.5 Cu/OSP Solder Joint with Bonding Strength Variation for Electronic Components, Korean J. Mater. Res. 17(3) (2007) 152–159.
[CROSSREF] 
6. W. S. Hong and C. M. Oh, PoF Based Accelerated Life Prediction with 3 Dimensional Packaging Technology Development, J. Korean Weld. Join. Soc. 27(3) (2009) 10–16.
[CROSSREF] 
7. W. S. Hong and C. M. Oh, Thermal Shock Reliability of Low Ag Composition Sn-0.3Ag-0.7Cu and Near Eutectic Sn-3.0Ag-0.5Cu Pb-free Solder Joints, J. Kor. Inst. Met. &Mater. 47(12) (2009) 842–851.
8. D. W. Suh, D. W. Kim, P. Liu, H. C. Kim, J. A. Weninger, C. M. Kumar, A. Prasad, B. W. Grimsley, and H. B. Tejada, Effects of Ag Content on Fracture Resistance of Sn-Ag-Cu Lead-Free Solders under High-Strain Rate Conditions, Mater. Sci. Eng. A. 460(461) (2007) 595–603.
[CROSSREF] 
9. JEITA, Lead Free Soldering Technologies, Corona Pub, Tokyo. (2003) 26–29.
10. W. S. Kim, W. S. Hong, S. H. Park, and K. B. Kim, Activation Energy and Interface Reaction of Sn-40Pb/Cu &Sn-3.0Ag-0.5Cu/Cu, Korean J. Mater. Res. 17(8)(402) (2007) 402–407. https://doi.org/10.3740/MRSK.2007.17.8.402
[CROSSREF] 
11. J. H. Back, B. S. Lee, S. H. Yoo, D. G. Han, S. B. Jung, and J. W. Yoon, Solderability of Thin ENEPIG Plating Layer for Fine Pitch Package Application, J. Microelectron. Packag. Soc. 24(1) (2017) 83–90. https://doi.org/10.6117/kmeps.2017.24.1.083
[CROSSREF] 
12. J. Y. Park, M. S. Kim, C. Oh, S. H. Do, J. D. Seo, D. K. Kim, and W. S. Hong, Solder Joint Fatigue Life of Flexible Impact Sensor Module for Automotive Electronics, Korean J. Met. Mater. 55(4) (2017) 232–239. https://doi.org/10.3365/KJMM.2017.55.4.232
[CROSSREF] 
13. W. Liu, Y. Tian, C. Wang, and L. Sun, Formation of AuSnx IMCs in Sn3.5Ag0.75Cu Micro-Solder Joints Fabricated by Laser and Hot Air Reflow Processes, J. Mater. Sci. Mater. Electron. 24(1) (2012) 1–7. https://doi.org/10.1007/s10854-012-0715-8
[CROSSREF] 
14. J. H. Son, D. Y. Yu, Y. H. Ko, and J. H. Bang, A Study on the Electrochemical Corrosion Property of Sn-xSb Solder Alloy, J. Weld. Join. 36(3) (2018) 78–82. https://doi.org/10.5781/JWJ.2018.36.3.12
[CROSSREF] 
15. J. H. Son, M. K. Kim, D. Y. Yu, Y. H. Ko, J. W. Yoon, C. W. Lee, Y. B. Park, and J. H. Bang, Thermal Aging Characteristics of Sn-xSb Solder for Automotive Power Module, J. Weld. Join. 35(5) (2017) 38–47. https://doi.org/10.5781/JWJ.2017.35.5.6
[CROSSREF] 


ABOUT
BROWSE ARTICLES
ARTICLE CATEGORY 
FOR CONTRIBUTORS
Editorial Office
#304, San-Jeong Building, 23, Gukhoe-daero 66-gil, Yeongdeungpo-gu, Seoul 07237, Korea
Tel: +82-2-538-6511    Fax: +82-2-538-6510    E-mail: koweld@kwjs.or.kr                

Copyright © 2024 by The Korean Welding and Joining Society.

Developed in M2PI