Effect of electromigration on mechanical shear behavior of flip chip solder joints Jae-Woong Naha) and Fei Ren Department of Materials Science and Engineering, University of California at Los Angeles, Los Angeles, California 90095-1595
Kyung-Wook Paik Department of Materials Science and Engineering, Korea Advanced Institute of Science and Technology, Taejon 305-701, Korea
K.N. Tu Department of Materials Science and Engineering, University of California at Los Angeles, Los Angeles, California 90095-1595 (Received 14 October 2005; accepted 2 December 2005)
Effect of electromigration on mechanical shear behavior of flip chip solder joints consisting of 97Pb3Sn and 37Pb63Sn composite solder joints was studied. The under bump metallurgy (UBM) on the chip side was TiW/Cu/electroplated Cu, and the bond pad on the board side was electroless Ni/Au. It was found that the mode of shear failure has changed after electromigration and the mode depends on the direction of electron flow during electromigration. The shear induced fracture occurs in the bulk of 97Pb3Sn solder without current stressing, however, after 10 h current stressing at 2.55 × 104 A/cm2 at 140 °C, it occurs alternately at the cathode interfaces between solder and intermetallic compounds (IMCs). In the downward electron flow, from the chip to substrate, the failure site was at the Cu–Sn IMC/solder interface near the Si chip. However, in the upward electron flow, from the substrate to chip, failure occurred at the Ni–Sn IMC/solder interface near the substrate. The failure mode has a strong correlation to microstructural change in the solder joint. During the electromigration, while Pb atoms moved to the anode side in the same direction as with the electron flow, Sn atoms diffused to the cathode side, opposite the electron flow. In addition, electromigration dissolves and drives Cu or Ni atoms from UBM or bond pad at the cathode side into the solder. These reactions resulted in the large growth of Sn-based IMC at the cathode sides. Therefore, mechanical shear failure occurs predominantly at the cathode interface.
I. INTRODUCTION
Reliability testing and failure analysis are important for estimating and predicting the life time of electronic products.1 Electromigration has become a serious reliability concern in flip chip solder joints because the dimensions of solder joints are expected to decrease and current density to increase.2 Many papers have been published about the electromigration behavior of flip chip solder joints.3–14 The general electromigration induced failure mode in flip chip solder joints is the loss of under bump metallurgy (UBM) and the interfacial void formation at the contact interface between the interconnect line
a)
Address all correspondence to this author. e-mail:
[email protected] DOI: 10.1557/JMR.2006.0086 698
J. Mater. Res., Vol. 21, No. 3, Mar 2006
and the solder bump. Furthermore, the effect of electromigration on mechanical behavior of flip chip solder joints has attracted much attention lately because mechanical reliability of solder joints can be affected by electromigration induced intermetallic compound (IMC) formation at the solder joint interfaces.15 However, the combination of electromigration and mechanical test in flip chip solder joints has not been reported. Flip chip solder joints in electronic products are experiencing electromigration by current flow and shear force by thermal stress simultaneously. The difference of thermal expansion coefficients between a Si chip and its organic substrate generates shear force on the flip chip solder joints between them because the package experiences thermal cycle from room temperature to the Si device operation temperature, about 100 °C. Actually many study of low cycle fatigue of flip chip solder joints © 2006 Materials Research Society
J-W. Nah et al.: Effect of electromigration on mechanical shear behavior of flip chip solder joints
II. EXPERIMENTAL
FIG. 1. Schematic diagram of the electromigration and mechanical shearing test sample.
have been reported,16–18 again there is no report on combining the effects of electromigration and thermal stress. We report here the effect of electromigration on the failure mode of shear of flip chip solder joints. For comparison, the shearing tests were conducted with and without electromigration, using samples of composite flip chip solder joints of 97Pb3Sn and 37Pb63Sn.19,20
Figure 1 is a schematic diagram of the cross-section of samples used for electromigration and die shear test. A test chip with 97Pb3Sn solder bumps was flipped and assembled on substrate with 37Pb63Sn solder. The 97Pb3Sn solder bumps on the test chip were electroplated and the 37Pb63Sn solder were stencil printed. The UBM on the chip side was sputtered TiW (0.2 m)/Cu (0.4 m)/electroplated Cu (5 m), and the bond pad on the substrate side was electroless Ni (5 m)/Au (0.1 m). The thickness of Al metal line on the chip side was 1 m, and that of Cu metal line on the substrate side was 18 m. For the flip chip assembly, the typical reflow condition of 37Pb63Sn solder was applied with a peak temperature of 220 °C and a dwell time of 90 s in nitrogen atmosphere. The composite solder joint was achieved by having the 37Pb63Sn solders wet the 97Pb3Sn solder bumps and flow around the entire surface of the 97Pb3Sn solder bump because only the former melts during the assembly process.
FIG. 2. Mechanical shear test setup for flip chip samples.
FIG. 3. SEM micrograph of the cross sectioned solder joints without current stressing.
FIG. 4. SEM micrograph of the cross-sectioned solder joints after mechanical shear testing without current stressing: (a) chip side and (b) substrate side.
J. Mater. Res., Vol. 21, No. 3, Mar 2006
699
J-W. Nah et al.: Effect of electromigration on mechanical shear behavior of flip chip solder joints
was 0.2 m/s. The Si chip size was 15 × 15 mm, and the substrate was 37.5 × 37.5 mm. To investigate the change of solder composition and IMC formation during electromigration testing as well as the failure mode after mechanical shearing testing, a set of solder joints was cross-sectioned after shear with or without current stressing and examined by field-emission scanning electron microscopy (FESEM). III. RESULTS AND DISCUSSION
FIG. 5. SEM micrograph of the cross sectioned solder joints after 20 h current stressing at 2.55 × 104 A/cm2 at 140 °C.
The circuit on the test samples was designed with daisy chain layout, and electrons flowed from a chip to a substrate and a substrate to a chip alternatively during electromigration test. Electromigration testing was performed with current density of 2.55 × 104 A/cm2. During current stressing, the sample was placed on a hotplate kept at 140 °C in the air. Mechanical shear test of assembled flip chip was performed before and after electromigration test. The set up of die shear test is shown in Fig. 2. The substrate was clamed and shear force was applied by a cross head at the edge of a chip. The speed of cross head during the test
A cross-sectional scanning electron microscopy (SEM) image of composite solder joins before current stressing is shown in Fig. 3. Because the composite solder joints were made by 97Pb3Sn solder on a chip surrounded by 37Pb63Sn solder on a substrate, a Pb-rich phase is found in the upper region and a Sn-rich phase in lower region. The bright color area represents the 97Pb3Sn solder and the dark area is 37Pb63Sn. Figure 4 shows a cross-sectional image of solder joints after mechanical shearing test without current stressing. It can be seen that fracture occurred not at solder/IMC interface or in the 37Pb63Sn region but in the 97Pb3Sn solder region. The failure mode was soft solder breaking in 97Pb3Sn solders and exhibited ductile dimple failure because the mechanical strength of 97Pb3Sn solder is lower than that of 37Pb-Sn solder.21 Figure 5 is a SEM image of the cross-sectioned solder bumps after current stressing of 20 h at 2.55 × 104 A/cm2
FIG. 6. Magnified cross-sectioned image of solder joints after current stressing: (a) bump 2 in Fig. 5, (b) chip side UBM of bump 2, (c) bump 3 in Fig. 5, and (d) substrate side of bump 3. 700
J. Mater. Res., Vol. 21, No. 3, Mar 2006
J-W. Nah et al.: Effect of electromigration on mechanical shear behavior of flip chip solder joints
FIG. 7. SEM micrograph of the cross sectioned solder joints after mechanical shear testing after 10 h current stressing at 2.55 × 104 A/cm2 at 140 °C: (a) chip side and (b) substrate side.
at 140 °C. The arrows marked by e− indicate the direction of electron flow. In Fig. 5, it is shown that failure was initiated in bumps at the cathode sides in downward electron flow (circled regions in bump 2 and bump 4) by the consumption of the Cu UBM. This is due to current crowding at the entrance of electron flow from chip to the solder bump. Electromigration at a high current density is fast because atomic diffusion is fast and local Joule heating is also higher at higher current density.5 There is a very large current density change at the contact window between the solder bump and Al line because the cross section of the Al line on the chip side is about two orders of magnitude smaller that that of the solder bump.22 This current crowing accelerated the IMC formation and void growth at the cathode side in bump 2 and bump 4. While there is also current crowding at the entrance of electron flow from substrate to the solder bump in upward electron bumps (lower regions in bump 1 and bump 3), the current crowding at these regions is smaller than the circled regions in bump 2 and bump 4 because the
thickness of Cu metal line (18 m) on substrate is much thicker than that of the Al metal line (1 m) on the chip. In addition to current crowding, the Pb atomic diffusion was driven from the cathode to the anode and the Sn in the opposite direction by electromigration. As shown in Fig. 6, after current stressing, the change of microstructures in the solder bumps and the thickened IMC at both the interfaces of the solder joint were observed. In the case of electron flow from the Si chip to the organic substrate [Figs. 6(a) and 6(b)], the solder joints showed one uniform color because the Pb had migrated to the substrate side and the Sn to the chip side. The average composition of the solder bump was measured 83Pb17Sn by energy dispersive x-ray (EDX) in area detection mode. Also, the increase of Sn concentration at the chip side caused an increase of formation of Cu–Sn IMCs, and finally a void was formed at the contact by the consumption of Cu UBM.6 However, the thickness of the Ni–Sn IMC layer at the substrate side did not change from the initial state before current stressing. In the opposite case of electron flow from the organic substrate to the Si chip [Figs. 6(c) and 6(d)], the Pb had migrated to the chip side, and much more Sn had accumulated in the substrate side. Therefore, the Ni–Sn IMC layer at the substrate side was thickened by consuming electroless Ni–P. However, the thickness of the Cu–Sn IMC layer at the chip side did not change from the initial state before current stressing. The above results show that the growth of IMC during electromigration in flip chip solder joints depends on the direction of electron flow. In a composite flip chip solder joints, IMC at the cathode side was thickened by the increase of Sn concentration and the migration of Cu or Ni into solder. To investigate the effect of electromigration on mechanical behavior of flip chip solder joint, mechanical shearing test was performed after electromigration. Figure 7 shows a cross-sectional image of solder joints after mechanical shearing test after current stressing of 10 h at 2.55 × 104 A/cm2 at 140 °C. After the current stressing, fracture occurred alternately at chip side and at substrate side. The direction of electron flow in the solder joints is indicated by the arrows marked by e−. In the upward electron flow from substrate to chip (bump 1 and bump 3), failure occurred at the Ni–Sn IMC/solder interface near the substrate; however, in the downward electron flow from chip to substrate (bump 2 and bump 4), the failure site was the Cu–Sn IMC/solder interface near the Si chip. During the electromigration, a large amount of intermetallic compound grew at the cathode side where current enters the solder, so failure occurred alternately at the chip side and at the substrate side in mechanical shearing test. On the basis of this result, it is reasonable to expect that electromigration weakens the cathode interface of solder joints and decreases its mechanical reliability.
J. Mater. Res., Vol. 21, No. 3, Mar 2006
701
J-W. Nah et al.: Effect of electromigration on mechanical shear behavior of flip chip solder joints
IV. SUMMARY
The effect of electromigration on mechanical shear reliability of composite solder joints consisting of 97Pb3Sn and 37Pb63Sn in a flip chip configuration was studied. The results of mechanical shear test before and after electromigration demonstrated that the mechanical properties of flip chip solder joints were seriously affected by electromigration. Mechanical shear failure mode changes from fracture in the bulk of solder before electromigration to the cathode interface after electromigration. The mode change is correlated to microstructural change. Electromigration drives the atomic diffusion of Sn from the anode to the cathode and leads to the growth of a large amount of IMC at the cathode. The large growth of IMC caused the brittle failure at IMC/solder interface in mechanical shear test. This electromigrationinduced mechanical properties change of flip chip solder joints would strongly impact the reliability of flip chip technology.
8.
9.
10.
11.
12.
13.
14.
15.
ACKNOWLEDGMENTS
The authors at UCLA would like to thank the support from Semiconductor Research Corporation (SRC) Contract No. NJ-1080 and Microelectronic Innovation and Computer Research Opportunities (MICRO)/Intel Contract No. 04-092.
16.
17.
REFERENCES 1. J.H. Lau and Y.H. Pao: Solder Joint Reliability of BGA, CSP, Flip Chip, and Fine Pitch SMT Assembles (McGraw-Hill, New York, 1997), p. 47. 2. K.N. Tu: Recent advances on electromigration in very-large-scaleintegration of interconnects. J. Appl. Phys. 94, 5451 (2003). 3. T.Y. Lee, K.N. Tu, S.M. Kuo, and D.R. Frear: Electromigration of eutectic SnPb solder interconnects for flip chip technology. J. Appl. Phys. 89, 3189 (2001). 4. T.Y. Lee, K.N. Tu, and D.R. Frear: Electromigration of eutectic SnPb and SnAg3.8Cu0.7 flip chip solder bumps and under-bump metallization. J. Appl. Phys. 90, 4502 (2001). 5. W.J. Choi, E.C.C. Yeh, and K.N. Tu: Mean-time-to-failure study of flip chip solder joints on Cu/Ni(V)/Al thin-film under-bumpmetallization. J. Appl. Phys. 94, 5665 (2003). 6. J.W. Nah, K.W. Paik, J.O. Suh, and K.N. Tu: Mechanism of electromigration-induced failure in the 97Pb–3Sn and 37Pb–63Sn composite solder joints. J. Appl. Phys. 94, 7560 (2003). 7. Y.C. Hsu, T.L. Shao, C.J. Yang, and C. Chen: Electromigration
702
18.
19.
20.
21.
22.
study in SnAg3.8Cu0.7 solder joints on Ti/Cr–Cu/Cu under-bump metallization. J. Electron. Mater. 32, 1222 (2003). T.L. Shao, K.C. Lin, and C. Chen: Electromigration studies of flip chip Sn95/Sb5 solder bumps Cr/Cr–Cu/Cu under-bump metallization. J. Electron. Mater. 32, 1278 (2003). T.L. Shao, S.H. Chiu, C. Chen, D.J. Yao, and C.Y. Hsu: Thermal gradient in solder joints under electrical-current stressing. J. Electron. Mater. 33, 1350 (2004). T.L. Shao, Y.H. Chen, S.H. Chiu, and C. Chen: Electromigration failure mechanisms for SnAg3.5 solder bumps on Ti/Cr–Cu/Cu under-bump metallization pads. J. Appl. Phys. 96, 4518 (2004). Y.H. Lin, C.M. Tsai, Y.C. Hu, Y.L. Lin, and C.R. Kao: Electromigration-induced failure in flip-chip solder joints. J. Electron. Mater. 34, 27 (2005). Y.H. Lin, Y.C. Hu, C.M. Tsai, C.R. Kao, and K.N. Tu: In situ observation of the void formation-and-propagation mechanism in solder joints under current-stressing. Acta Mater. 53, 2029 (2005). Y.H. Liu and K.L. Lin: Damages and microstructural variation of high-lead and eutectic SbPb composite flip chip solder bumps induced by electromigration. J. Mater. Res. 20, 2184 (2005). J.W. Nah, J.O. Suh, and K.N. Tu: Effect of current crowding and Joule heating on electromigration-induced failure in flip chip composite solder joints tested at room temperature. J. Appl. Phys. 98, 013715 (2005). F. Ren, J.W. Nah, H. Gan, J.O. Suh, K.N. Tu, B. Xiong, L. Xu, and J. Pang: Effect of electromigration on mechanical behavior of solder joints, in Materials, Technology and Reliability of Advanced Interconnects—2005, edited by P.R. Besser, A.J. McKerrow, F. Iacopi, C.P. Wong, and J. Vlassak (Mater. Res. Soc. Symp. Proc. 863, Warrendale, PA, 2005), B10.2. L.L. Mercado, V. Sarihan, Y. Guo, and A. Mawer: Impact of solder pad size on solder joint reliability in flip chip PBGA packages. IEEE Trans. Adv. Packag. 23, 416 (2000). J.H.L. Pang, D.Y.R. Chong, and T.H. Low: Thermal cycling analysis of flip-chip solder joint reliability. IEEE Trans. Compon. Packag. Technol. 24, 705 (2001). Y. Kariya, T. Hosoi, S. Terashima, M. Tanaka, and M. Otsuka: Effect of silver content on the shear fatigue properties of Sn–Ag–Cu flip-chip interconnects. J. Electron. Mater. 33, 321 (2004). V.K. Nagesh, R. Peddada, S. Ramalingam, B. Sur, and A. Tai: Challenges of flip chip on organic substrate assembly technology, in Proc. 49th Electronic Components and Technology Conference (IEEE, Piscataway, NJ, 1999), p. 975. R. Shukla, V. Murali, and A. Bhansali: Flip chip CPU package technology at Intel: A technology and manufacturing overview, in Proc. 49th Electronic Components and Technology Conference (IEEE, Piscataway, NJ, 1999), p. 945 Y. Guo, S.M. Kuo, and C. Zhang: Reliability evaluations of under bump metallurgy in two solder systems. IEEE Trans. Compon. Packag. Technol. 24, 655 (2001). E.C.C. Yeh, W.J. Choi, K.N. Tu, P. Elenius, and H. Balkan: Current-crowding-induced electromigration failure in flip chip solder joints. Appl. Phys. Lett. 80, 580 (2002).
J. Mater. Res., Vol. 21, No. 3, Mar 2006
