Electrical Performance of Micro-assembled Beads ...
Electrical Performance of Micro-assembled Beads under Different Temperatures and Loadings Yen Lin Tzeng, Kerwin Wang Department of Mechatronics Engineering, National Changhua University of Education, Changhua, Taiwan [email protected]
Abstract—Micro-assembly is an efficient tool to build electrical connections with metallic micro-beads. This process uses patterned photoresist AZ1512 as an adhesion for micro-bead arrangement. The assembled beads is immobilized with underfill embedment (ZYMET 2821). This method allows arbitrary geometric pattern designs. All of these processes can be completed below 150oC. This paper characterizes the electrical performance of these densely-arranged anisotropic conductive tunnels under different temperatures and stresses loading Experiment results suggest that using photoresist to assemble micro conductive beads with underfill immobilization can yield stable performance. Keywords-assembly; micro-assembly; micro beads; vertical conductive channel; underfill.
I.
INTRODUCTION
In recent years, 3D flip-chip packaging becomes an important technology to integrate miniaturized systems. Overall, the packaging scheme can involve high density electrical interconnections between chips and integral circuits or between vertical conductive channels and through silicon vias (TSV) [1, 2]. Through the vertical conductive path, the transmission and processing speed of the microelectronic device can be improved [3]. Anisotropic conductive film (ACF) bonding is one of the well knowing technologies for building vertical conductive paths. It is widely used in flat panel displays, chip packaging and ACF wafer-level packaging [4]. However, it is well known that the thermomechanical reliability of these vertical conductive paths is depended on the thermo-mechanical properties of ACF. Many researchers have conducted some useful research [5-6] in this area. Instead of using ACF, we presented a photolithography based micro-conductive-bead assembly method our previous research [7]. It can provide miniaturized and densely packed conductive paths. The assembled micro-conductive-beads have to be permanently immobilized and electrically connected to the binding sites for practical applications. In this paper, we arrange the micro-scaled conductive beads on the lithographic patterns and inject the unferfill into the gap between chips; it can immobilize the beads and help the conductive chip to against the mechanical stress and moisture. The effect of the mechanical stress and temperature variation of electrical conductivity was also investigated. In the following sections, we first discuss the materials, experimental procedures and methods. Then, we present the test results and conclusions.
II.
MATERIALS AND EXPERIMENT PROCEDURES
A. Matrials 3 µm micro-composite conductive beads are selected for building anisotropic conductive electrical channels. The beads consist of a 0.025 µm gold-nickel metal layer and an encapsulated benzoguanamine resin core. AZ1512 is used to define the require patterns for microbeads assembly by photolithography process. It is a widely used positive photoresist for micro-fabrication. ZYMET X2821 is an encapsulation underfill materials, it can be quickly cured under low temperature. In this paper, it is used to immobilize the conductive beads and to hold the assembled chip. It can also protect the connections from damages. TABLE I.
WL-CSP AND BGA UNDERFILL ENCAPSULANT
Type
ZYMET X2821(underfill)
Viscosity, 25°C Cure conditions CTE Tg, (TMA) Shear storage modulus
500 (cps) 150°C, 1minutes 38ppm/°C 135°C 3GPa
TABLE II.
BEADS COEFFICIENT OF THERMAL EXPANSION
Layer Au Ni
Micro conductive beads
Coefficient of Thermal Expansion 14.2 ppm /°C 13 ppm /°C 45 ppm /°C
B. Micro-assembly Process To achieve densely packed electrical connections between chip and substrate, the procedures for micro conductive beads assembly and underfill process are listed in Fig. 1. A similar process of micro conductive beads assembly has been published in our previous research [7]. This process is start from an oxidized and aluminum deposited 4” wafer. AZ1512 is spin coated on Si wafer to pattern the aluminum. Depends on the applications, after lithography, developing and etching process, one can create desired patterns on the photoresist and aluminum for micro conductive beads assembly. The beads are spread densely and evenly on the photoresist patterns.
The Hotplate heating makes the photoresist soften at 120OC. It can increase sticking force between beads and patterns. Micro-conductive beads are selectively immobilized on the photoresist patterns. After gently washed redundant beads away with deionized water, we injected underfill into the chip's gap and heated it up to 150OC for 1 minute to bond and to immobilize the combinations. UV
(a)
Si wafer
Mask Photoresist Al layer
F
Beads (b)
(c)
Deionized water
(f)
Underfill
(g)
Fig. 2 shows the immobilized micro-conductive bead on the photoresist. The diameter of the micro conductive beads is around 3µm. The micro conductive beads are selectively immobilized on the patterned region. This immobilization process does not require external force fields. It is a cost and time efficient process. The bead distributions on the patterns are random in nature; however, they are confined to predetermined area. To ensure a good encapsulation process, the designed pattern should have topologically opened cavity geometry to spread the underfill. This study is focus on the electrical performance of microassembled beads under different temperatures and loadings. The simplified bead assembly region is prepared to ensure the applicable array arrangement, which provides improved filling results on the curing, velocity and encapsulant distributions for capillary-driven by dispensing. Fig. 3 shows an assembled combination of top and bottom chip with sandwiched conductive beads, underfill and photoresist. Top view of the over lapped area 4.3mm
(d)
(E) (d)
2.7mm
(h)
Fig. 1: Process steps: (a) photolithography; (b) conductive-bead spreading; (c) photoresist soften and bead immobilization; (d) redundant bead cleaning; (e) patterned beads; (f) photoresist reflow and heating impression; (g) underfill injection; (h) underfill curing.
Cross-section view of the over lapped area
Underfill
3µ µm
2.6µm
Al layer
Fig. 3: The schematic top view and cross-section diagram shows the configuration of thickness and size of photoresist and micro conductive bead assembly. The overlap area of electrodes is 11.6mm2.
III. EXPERIMENTAL SETUP AND MEASUREMENT
Fig. 2: The testing result shows that this technique can assemble a single layer of microbeads on well-defined photoresist (AZ1512) patterns.
Unlike traditional soldering processes, the electrical connections are not built by chemical bonding or phase transition. There are only physical contacts between electric pads and micro-conductive beads. That means contact sliding gap deformation and stress variation may cause the resistance change. Measurements have been made of the electrical conductive performance of the assembled conductive beads, namely its resistance at different temperature, compressive stress and shear stress. The experiment setups of these tests are shown in Fig. 4.
cover (b) Weight meter PC
Multimeter
F
(a)
Hotplate
(c) Weight meter Fig. 4: Three kinds of experiment setups : That are measure and transmit data by mutimeter and computer (a) Use hotplate to Heat the products in the close cover. (b) Give the products downward stress on the weight meter. (c) Give the upright products shear stress on the weight meter and .
The test conditions and results of the resistances at different temperatures and loading stress are listed below: A. Electrical resistance of this connection at different temperatures The assembled chip is placed on a hot plate; the electrodes are connected to a digital multimeter (Agilent 34410A) to measure the resistance. The setup is installed in a sealed chamber which provides a closed environment and good temperature uniformity. The heating rate of this setup is around 29OC/min and the cooling rate is around 4OC/min. As shown in Fig. 5, the resistance measurement range is setup from 25 to 150 OC, the data are automatically collected by a computer. Contact resistance measurement results corresponding to the thermal cycle is shown in Fig. 6, the contact resistance between 3µm Ni/Au coated Benzoguanamine resin beads and bonding pads are increased after temperature increased from 27 to 148 OC. Photoresist above reflow temperature was thermally and mechanically less stable above 110oC; however, been embedded in the underfill, photoresist reflow did not show any significant impact to resistance.
B. Electrical resistance of this connection under different compressive loadings The sample was tested under compressive loadings (Fig. 7). The loadings are gradually increased from 0 to 800kN/m2. Then it is gradually decreased back to 0kN/m2 to form an enclosed loop. The assembled connection had low contact resistance at the initial stage. When the pressure force is gradually applied to the assembly, compressive loading pressure would promote contact stress redistribution which increases the resistance of the electric assembly. However, the data measured within the linear elastic deformation loop are steady and repeatable. C. Electrical resistance of this connection at different shear stresses To precede the resistance investigation under shear stress loading, an assembled chip is installed onto a testing platform with a force gauge. The shear stress was continuously loaded to the assembled chips until the catastrophic damage occurred. The value of contact resistance, measured correspond to the shear loading, is shown in Fig. 8. The experiment results suggest that shear stress may cause significant degradation on the contact resistance. It can cause physical detaching and delaminating which leads to mechanical and electrical failure. The composite metallic beads aging caused by high temperature can lead further degradation of the contacts. High bonding and reflow temperatures above 200 OC are not recommended for reliability issue.
Temperture curve
160.00 140.00 ) 120.00 C O e( 100.00 ru t re 80.00 p m 60.00 e T 40.00 20.00 0.00
Temperature (OC)
F
0.0 There are various explanations on what causes the resistance variation (Fig. 6) during the thermal cycle, the most possible explanation being Matthiessen’s rule and the coefficient of thermal expansion differences among the photoresist, micro metallic beads and the underfill.
10.0
20.0
30.0
Time(min)
Fig. 5: The temperature curve of the platform during heating and cooling process.
IV.
CONCLUSION AND SUMMARY
Micro-assembly is an efficient tool to build electrical connections with metallic microbeads. This process uses patterned photoresist AZ1512 as an adhesion for micro-bead adhesion. This method allows arbitrary geometric pattern designs. All of these processes can be completed below 150oC. This paper characterizes the electrical performance of this densely-arranged anisotropic conductive tunnels under different temperatures and stresses. Experiment results suggest that using photoresist to assemble micro conductive beads with underfill embedment (XYMET 2821) can yield stable performance with appropriate shear stress protection.
Fig. 6: The resistances versus temperature characteristics of the bonded chip.
The results indicate that the micro-bead assembly with the underfill encapsulation method developed in this study is adequate for micro conductive beads assembly in flip chip bonding process.
ACKNOWLEDGMENT The authors would also like to thank National Chiao Tung University, and the Nano Facility Center at National Changhua University of Education, for providing the fabrication facility. This work is supported by National Science Council, NSC 100-2221-E-018-015 and NSC-982221-E-018-012-MY2.
REFERENCES [1]
[2]
Fig. 7: The resistances versus compress stress characteristics of the bonded chip.
[3]
Resistance V.S. Shear stress
[4]
600
Resistane(Ohm)
500 400
[5]
Breakdown
300 200
[6]
100 0 0
10
20
30
40
50
60
Shear stress(kN/m2)
Fig. 8: The resistances versus shear stress characteristics of the bonded chip.
[7]
S. F. Al-Sarawi, D. Abbott, and P. Franzon, “A review of 3-Dpackaging technology,” IEEE Trans. Compon. Packag. Manuf. Technol, vol. 21,no. 1, pp. 2–14, 1998. Y. Yano, T. Sugiyama, S. Ishihara, Y. Fukui, H. Juso, K. Miyata,Y. Sota, and K. Fujita, “Three-dimensional thin stacked packaging technology for SiP,” in Proc. Electron. Compon. Technol. Conf, pp. 1329–1334, 2002. M, Motoyoshi, “Through-Silicon Via (TSV),” Proceedings of the IEEE, Vol. 97, No.2, 2009, p43-48H.-Y. Son, C.-K. Chung, “WaferLevel Flip Chip Packages Using Preapplied Anisotropic Conductive Films (ACFs),” IEEE Trans. on Electron. Packag. Manuf, 30, pp. 221227, 2007. K.-W Jang, and K.-W. Paik, “Effects of Heating Rate on Material Properties of Anisotropic Conductive Film (ACF) and Thermal Cycling Reliability of ACF Flip Chip Assembly,” IEEE Trans. on Components and Packaging Technologies, 32, pp. 339-346, 2009. S-T Lu, W-H Chen, “Reliability and Flexibility of Ultra-Thin Chip-onFlex (UTCOF) Interconnects With Anisotropic Conductive Adhesive (ACA) Joints,” IEEE Trans. on Advanced Packaging, 33, pp.702-712, 2010. T. Kubono, S. Teraoka, “Relationship between the contact resistance and the temperature near electrical contacts switching continually,” Electronics and Communications in Japan (Part II: Electronics), 72(5), pp. 89-96, 1989. K.-W Wang, C.-H Chen, “Binding Site Design For The Micro Conductive Bead Assembly In Flip Chip Bonding” in The 4th Asia Pacific Conference on Transducers and Micro/Nano Technologies, Tainan, Taiwan, June 22-25, 2008.
Abstract—Micro-assembly is an efficient tool to build electrical connections with metallic micro-beads. This process uses patterned photoresist AZ1512 as an adhesion for micro-bead arrangement. The assembled beads is immobilized with underfill embedment (ZYMET 2821). This method allows arbitrary geometric pattern designs. All of these processes can be completed below 150oC. This paper characterizes the electrical performance of these densely-arranged anisotropic conductive tunnels under different temperatures and stresses loading Experiment results suggest that using photoresist to assemble micro conductive beads with underfill immobilization can yield stable performance. Keywords-assembly; micro-assembly; micro beads; vertical conductive channel; underfill.
I.
INTRODUCTION
In recent years, 3D flip-chip packaging becomes an important technology to integrate miniaturized systems. Overall, the packaging scheme can involve high density electrical interconnections between chips and integral circuits or between vertical conductive channels and through silicon vias (TSV) [1, 2]. Through the vertical conductive path, the transmission and processing speed of the microelectronic device can be improved [3]. Anisotropic conductive film (ACF) bonding is one of the well knowing technologies for building vertical conductive paths. It is widely used in flat panel displays, chip packaging and ACF wafer-level packaging [4]. However, it is well known that the thermomechanical reliability of these vertical conductive paths is depended on the thermo-mechanical properties of ACF. Many researchers have conducted some useful research [5-6] in this area. Instead of using ACF, we presented a photolithography based micro-conductive-bead assembly method our previous research [7]. It can provide miniaturized and densely packed conductive paths. The assembled micro-conductive-beads have to be permanently immobilized and electrically connected to the binding sites for practical applications. In this paper, we arrange the micro-scaled conductive beads on the lithographic patterns and inject the unferfill into the gap between chips; it can immobilize the beads and help the conductive chip to against the mechanical stress and moisture. The effect of the mechanical stress and temperature variation of electrical conductivity was also investigated. In the following sections, we first discuss the materials, experimental procedures and methods. Then, we present the test results and conclusions.
II.
MATERIALS AND EXPERIMENT PROCEDURES
A. Matrials 3 µm micro-composite conductive beads are selected for building anisotropic conductive electrical channels. The beads consist of a 0.025 µm gold-nickel metal layer and an encapsulated benzoguanamine resin core. AZ1512 is used to define the require patterns for microbeads assembly by photolithography process. It is a widely used positive photoresist for micro-fabrication. ZYMET X2821 is an encapsulation underfill materials, it can be quickly cured under low temperature. In this paper, it is used to immobilize the conductive beads and to hold the assembled chip. It can also protect the connections from damages. TABLE I.
WL-CSP AND BGA UNDERFILL ENCAPSULANT
Type
ZYMET X2821(underfill)
Viscosity, 25°C Cure conditions CTE Tg, (TMA) Shear storage modulus
500 (cps) 150°C, 1minutes 38ppm/°C 135°C 3GPa
TABLE II.
BEADS COEFFICIENT OF THERMAL EXPANSION
Layer Au Ni
Micro conductive beads
Coefficient of Thermal Expansion 14.2 ppm /°C 13 ppm /°C 45 ppm /°C
B. Micro-assembly Process To achieve densely packed electrical connections between chip and substrate, the procedures for micro conductive beads assembly and underfill process are listed in Fig. 1. A similar process of micro conductive beads assembly has been published in our previous research [7]. This process is start from an oxidized and aluminum deposited 4” wafer. AZ1512 is spin coated on Si wafer to pattern the aluminum. Depends on the applications, after lithography, developing and etching process, one can create desired patterns on the photoresist and aluminum for micro conductive beads assembly. The beads are spread densely and evenly on the photoresist patterns.
The Hotplate heating makes the photoresist soften at 120OC. It can increase sticking force between beads and patterns. Micro-conductive beads are selectively immobilized on the photoresist patterns. After gently washed redundant beads away with deionized water, we injected underfill into the chip's gap and heated it up to 150OC for 1 minute to bond and to immobilize the combinations. UV
(a)
Si wafer
Mask Photoresist Al layer
F
Beads (b)
(c)
Deionized water
(f)
Underfill
(g)
Fig. 2 shows the immobilized micro-conductive bead on the photoresist. The diameter of the micro conductive beads is around 3µm. The micro conductive beads are selectively immobilized on the patterned region. This immobilization process does not require external force fields. It is a cost and time efficient process. The bead distributions on the patterns are random in nature; however, they are confined to predetermined area. To ensure a good encapsulation process, the designed pattern should have topologically opened cavity geometry to spread the underfill. This study is focus on the electrical performance of microassembled beads under different temperatures and loadings. The simplified bead assembly region is prepared to ensure the applicable array arrangement, which provides improved filling results on the curing, velocity and encapsulant distributions for capillary-driven by dispensing. Fig. 3 shows an assembled combination of top and bottom chip with sandwiched conductive beads, underfill and photoresist. Top view of the over lapped area 4.3mm
(d)
(E) (d)
2.7mm
(h)
Fig. 1: Process steps: (a) photolithography; (b) conductive-bead spreading; (c) photoresist soften and bead immobilization; (d) redundant bead cleaning; (e) patterned beads; (f) photoresist reflow and heating impression; (g) underfill injection; (h) underfill curing.
Cross-section view of the over lapped area
Underfill
3µ µm
2.6µm
Al layer
Fig. 3: The schematic top view and cross-section diagram shows the configuration of thickness and size of photoresist and micro conductive bead assembly. The overlap area of electrodes is 11.6mm2.
III. EXPERIMENTAL SETUP AND MEASUREMENT
Fig. 2: The testing result shows that this technique can assemble a single layer of microbeads on well-defined photoresist (AZ1512) patterns.
Unlike traditional soldering processes, the electrical connections are not built by chemical bonding or phase transition. There are only physical contacts between electric pads and micro-conductive beads. That means contact sliding gap deformation and stress variation may cause the resistance change. Measurements have been made of the electrical conductive performance of the assembled conductive beads, namely its resistance at different temperature, compressive stress and shear stress. The experiment setups of these tests are shown in Fig. 4.
cover (b) Weight meter PC
Multimeter
F
(a)
Hotplate
(c) Weight meter Fig. 4: Three kinds of experiment setups : That are measure and transmit data by mutimeter and computer (a) Use hotplate to Heat the products in the close cover. (b) Give the products downward stress on the weight meter. (c) Give the upright products shear stress on the weight meter and .
The test conditions and results of the resistances at different temperatures and loading stress are listed below: A. Electrical resistance of this connection at different temperatures The assembled chip is placed on a hot plate; the electrodes are connected to a digital multimeter (Agilent 34410A) to measure the resistance. The setup is installed in a sealed chamber which provides a closed environment and good temperature uniformity. The heating rate of this setup is around 29OC/min and the cooling rate is around 4OC/min. As shown in Fig. 5, the resistance measurement range is setup from 25 to 150 OC, the data are automatically collected by a computer. Contact resistance measurement results corresponding to the thermal cycle is shown in Fig. 6, the contact resistance between 3µm Ni/Au coated Benzoguanamine resin beads and bonding pads are increased after temperature increased from 27 to 148 OC. Photoresist above reflow temperature was thermally and mechanically less stable above 110oC; however, been embedded in the underfill, photoresist reflow did not show any significant impact to resistance.
B. Electrical resistance of this connection under different compressive loadings The sample was tested under compressive loadings (Fig. 7). The loadings are gradually increased from 0 to 800kN/m2. Then it is gradually decreased back to 0kN/m2 to form an enclosed loop. The assembled connection had low contact resistance at the initial stage. When the pressure force is gradually applied to the assembly, compressive loading pressure would promote contact stress redistribution which increases the resistance of the electric assembly. However, the data measured within the linear elastic deformation loop are steady and repeatable. C. Electrical resistance of this connection at different shear stresses To precede the resistance investigation under shear stress loading, an assembled chip is installed onto a testing platform with a force gauge. The shear stress was continuously loaded to the assembled chips until the catastrophic damage occurred. The value of contact resistance, measured correspond to the shear loading, is shown in Fig. 8. The experiment results suggest that shear stress may cause significant degradation on the contact resistance. It can cause physical detaching and delaminating which leads to mechanical and electrical failure. The composite metallic beads aging caused by high temperature can lead further degradation of the contacts. High bonding and reflow temperatures above 200 OC are not recommended for reliability issue.
Temperture curve
160.00 140.00 ) 120.00 C O e( 100.00 ru t re 80.00 p m 60.00 e T 40.00 20.00 0.00
Temperature (OC)
F
0.0 There are various explanations on what causes the resistance variation (Fig. 6) during the thermal cycle, the most possible explanation being Matthiessen’s rule and the coefficient of thermal expansion differences among the photoresist, micro metallic beads and the underfill.
10.0
20.0
30.0
Time(min)
Fig. 5: The temperature curve of the platform during heating and cooling process.
IV.
CONCLUSION AND SUMMARY
Micro-assembly is an efficient tool to build electrical connections with metallic microbeads. This process uses patterned photoresist AZ1512 as an adhesion for micro-bead adhesion. This method allows arbitrary geometric pattern designs. All of these processes can be completed below 150oC. This paper characterizes the electrical performance of this densely-arranged anisotropic conductive tunnels under different temperatures and stresses. Experiment results suggest that using photoresist to assemble micro conductive beads with underfill embedment (XYMET 2821) can yield stable performance with appropriate shear stress protection.
Fig. 6: The resistances versus temperature characteristics of the bonded chip.
The results indicate that the micro-bead assembly with the underfill encapsulation method developed in this study is adequate for micro conductive beads assembly in flip chip bonding process.
ACKNOWLEDGMENT The authors would also like to thank National Chiao Tung University, and the Nano Facility Center at National Changhua University of Education, for providing the fabrication facility. This work is supported by National Science Council, NSC 100-2221-E-018-015 and NSC-982221-E-018-012-MY2.
REFERENCES [1]
[2]
Fig. 7: The resistances versus compress stress characteristics of the bonded chip.
[3]
Resistance V.S. Shear stress
[4]
600
Resistane(Ohm)
500 400
[5]
Breakdown
300 200
[6]
100 0 0
10
20
30
40
50
60
Shear stress(kN/m2)
Fig. 8: The resistances versus shear stress characteristics of the bonded chip.
[7]
S. F. Al-Sarawi, D. Abbott, and P. Franzon, “A review of 3-Dpackaging technology,” IEEE Trans. Compon. Packag. Manuf. Technol, vol. 21,no. 1, pp. 2–14, 1998. Y. Yano, T. Sugiyama, S. Ishihara, Y. Fukui, H. Juso, K. Miyata,Y. Sota, and K. Fujita, “Three-dimensional thin stacked packaging technology for SiP,” in Proc. Electron. Compon. Technol. Conf, pp. 1329–1334, 2002. M, Motoyoshi, “Through-Silicon Via (TSV),” Proceedings of the IEEE, Vol. 97, No.2, 2009, p43-48H.-Y. Son, C.-K. Chung, “WaferLevel Flip Chip Packages Using Preapplied Anisotropic Conductive Films (ACFs),” IEEE Trans. on Electron. Packag. Manuf, 30, pp. 221227, 2007. K.-W Jang, and K.-W. Paik, “Effects of Heating Rate on Material Properties of Anisotropic Conductive Film (ACF) and Thermal Cycling Reliability of ACF Flip Chip Assembly,” IEEE Trans. on Components and Packaging Technologies, 32, pp. 339-346, 2009. S-T Lu, W-H Chen, “Reliability and Flexibility of Ultra-Thin Chip-onFlex (UTCOF) Interconnects With Anisotropic Conductive Adhesive (ACA) Joints,” IEEE Trans. on Advanced Packaging, 33, pp.702-712, 2010. T. Kubono, S. Teraoka, “Relationship between the contact resistance and the temperature near electrical contacts switching continually,” Electronics and Communications in Japan (Part II: Electronics), 72(5), pp. 89-96, 1989. K.-W Wang, C.-H Chen, “Binding Site Design For The Micro Conductive Bead Assembly In Flip Chip Bonding” in The 4th Asia Pacific Conference on Transducers and Micro/Nano Technologies, Tainan, Taiwan, June 22-25, 2008.
