Thermomechanical Response of Anisotropically Conductive Film

Transcription

1 Thermomechanical Response of Anisotropically Conductive Film Yung Neng Cheng, Shyong Lee and Fuang Yuan Huang Department of Mechanical Engineering National Central University, Chung-li, Taiwan Abstract Anisotropically conductive film (ACF) is a smart electronic packaging material that consumes minimal space for connecting an IC chip to a liquid crystal display (LCD) panel or a printed circuit board. It consists of an adhesive resin and fine conductive fillers such as metallic particles or metal-coated polymer balls. The fillers are compressed and maintained at a certain elastic capability enabling them to conduct between the electrodes. Thus the extent of contact area and the shape of conductive fillers are important factors for determining conductivity. The process of applying ACF is modeled into three consecutive steps, with the stress and deformation states studied by finite element analysis in each step. The first step of the manufacturing process involves compressing the conductive particles with an external load at a temperature of 190 o C, insuring that the matrix resin is at a melting fluid condition. In the second step, we maintain the previous temperature and allow the matrix resin to solidify. Then the load is released allowing the particles to spring back to create tension stress in the bonding resin matrix. The final step allows the bonded and conducting ACF to cool from 190 o C down to room temperature. The state of stress and deformation will be readjusted due to different contraction properties between the filler and matrix resin. The results presented in this paper show that the first step of the process is the key focus in determining the function of the ACF. Key words Anisotropically conductive film (ACF), liquid crystal display (LCD), finite element analysis (FEA), chip on glass (COG) 1. Introduction ACF has been very desirable for micro-pitch connection of IC circuits to LCD panels because of its reliability and compactness. This material consists of microscopic electrically conductive fillers 3-15 mm in diameter that is dispersed within a resin adhesive mm thick. Out of the various types of materials for conductive particles and adhesive film that have been proposed, thermosetting adhesive film with metal-coated plastic particles is the most popular type [1]. The reason for using metal-coated plastic particles over solid metal particles is not well The International Journal of Microcircuits and Electronic Packaging, Volume 24, Number 4, Fourth Quarter, 2001 (ISSN ) 379

2 described in the literature, so it will be studied from the point of the thermomechanical properties of the material. The bonding and conducting mechanism depends on the thermomechanical response of the conductive fillers and matrix adhesive resin. Initially, a load is applied to compress the ACF at a high temperature with respect to the matrix resin that is in a melting state. During this state, the conductive particles are compressed to yield some flattened, intimate contact area to the electrode pads. Intuitively, the amount of contact area should be an important factor in determining the electrical conductivity. The shape of the deformed conductive particle should also be influential as well. If the applied compression is too great, the coated metal layer may suffer a high bulging stress from the resin inside and crack. On the other hand, insufficient compression might not create enough contact area to facilitate current flow, producing insufficient elastic force for a reliable intimate contact. In the next step, an external load will be released after the matrix resin solidifies to sustain the elastic force induced from the reverting particles. As the bonded assembly cools down to room temperature, The state of stress and deformation will further develop due to the different thermal contraction properties of the conductive particles and the matrix resin. It would be worthwhile to understand the thermomechanical response of the ACF at the various stages described above during the bonding process. Although experimental studies in this respect are plentiful [2-5] and some associating finite element analyses were also performed [5-7], there has been no specific conclusion pointing out an optimal pressing condition for applying the ACF. In this paper, finite element analysis (FEA) is employed to study the stress and deformation states during the three steps of the process when considering 20-60% initial compression amount on conductive particles of 3 and 5 mm diameter. 2. Materials and Finite Element Analysis Modeling The material used for coating the conductive particle is assumed to be nickel since it is commonly referred to in literature [8]. The conducting pad of IC is assumed to be made of nickel as well, and the other end for contacting the particle is glass that is mostly likely to be a LCD panel. A schematic plot showing the arrangement of relevant components in connection with the ACF is shown in Figure 1. The materials for each component and their relevant properties involved in the study are listed in Table 1. For the 2-D finite element analysis (FEA) performed in this paper, ANSYS code [9] was used. The domain of the conductive particles, the corresponding contacting pads, and adhesive resin matrix is modeled as axissymmetrical and divided into a discrete grid as depicted in Figure 2. The FEA is primarily directed at the metal-coated resin structure for which the particle diameter is considered to be either 3 or 5 mm, and the nickel coating layer thickness is assumed to be mm. The initial condition prescribed to initiate the finite element analysis is the distance change between the two conduction pads, which is expressed in terms of percentage of the diameter of conducting ball. Incidentally, FEA was also performed on a solid monolithic particle and its solution was compared with that by elasticity theory to verify its validity. In addition, compliances of the monolithic and composite conductive particles can be compared. 3. Results and Discussion 3.1 Validity of the FEA as comparing with the solution by elasticity theory For a solid particle, with the dimensions and material properties as given in Table 1, 380

3 elasticity theory can solve the reverting force as a function of the linear deformation. Theoretical solution is an ideal tool to verify whether the finite element modeling is correct. If so, then FEA is valid for applying to the other cases to be discussed next. The results are shown in Figure 3 in which the x-axis represents the initial linear deformation applied at the amount of 20%, 30% and 40%. 3.2 Composite ball (resin interior with metal coating) vs. monolithic metal one It has been stated that a composite particle of resin and nickel coating is preferred over solid nickel as the conductive filler used in the ACF. A straightforward explanation would be that the compliances are very different, since the resin occupying most of the volume of the composite is around 100 times more compliant than the nickel metal (Table 1). To observe how the composite and solid ball (both diameters are 5 mm) deform, a 2 mm compressive displacement was applied to the damping pads. The results demonstrated that the composite ball was easily flattened on top and bottom as shown in Figure 4. However as for the solid ball, deformation was not obvious because the pads were indented to accommodate a big portion of the displacement. FEA shows that the contact area for the composite ball is around 40% larger than the solid ball, the contact force and stress are also lowered by approximately 5 to 10 times. 3.3 Thermomechanical response of the ACF by FEA First step The conductive particle considered is assumed to be of resin material 3 or 5 mm in diameter, with a nickel metal coating of mm and suffering a 20-60% initial linear compressive deformation. In this step, the ACF is applied at 190 o C and the resin matrix is in liquid state. The ANASYS code has the ability to let the discrete elements in this domain temporarily not work. The contact area between the reverting and flattened particle and the supporting pads are calculated and listed in Table 2. Maximum tensile stress in the nickel layer is also included, which is an important consideration since it can lead to breakage and significantly damage electric conductivity Second step In this step, the applied ACF is still at 190 o C, but the resin matrix solidifies, bonding to the pads and pulling them in a direction as to resist the pushing force caused by the reverting particle. The ANASYS code has the ability to revitalize the discrete elements in the resin matrix domain to sustain stresses due to the reverting force of the particle after release of the external load. The contact area and maximum tensile stress in the nickel layer are calculated and listed in Table Third (final) step Finally, the bonded package containing the conductive ball with its clamping pads and adhesive resin cools down from 190 o C to room temperature. The resin matrix, already in a state of tension stress in the previous step, contracts to further compress the ball which also contracts but in less amount because the nickel layer has smaller thermal contraction coefficient. Some calculated data are listed in Table 4. The above listed data are a small fraction of what have been calculated, and from these, various functional relations are investigated. An issue of most interest is to reveal what 381

4 Units: micrometers Conductive ball IC Chip IC Chip Bump Adhesive matrix or 5 LCD panel 20 Figure 1. Relevant Components in an Anisotropically Conductive Film (ACF) factors influence contact area since it directly determines the capability of electrical conductivity. The contact between conductive particle and clamping pads is not welded or glued, so the contact area can self-adjust in accordance to the thermomechanical response of the conductive particle and resin adhesive matrix. In Figure 5, the contact areas for various deformations from 20% to 60% at three different steps are plotted. It is seen that once the initial amount of deformation is selected in the first step, then contact area will not change significantly in the following steps. Another important information obtained from this plot is that contact area is pretty much linearly proportional to the applied compressive displacement. analysis results show that size effect on maximum tension stress in the nickel layer is not obvious as reviewing the data for 3 and 5 mm diameters considered in the work (Tables2-4). 382

5 Component/ Materials Table 1. Material Properties Adopted for Finite Element Analysis Constitutive Equation Young s Modulus Poisson s Ratio Coefficient of Thermal Expansion (ppm/ o C) IC chip/si Elastic 128, LCD panel/glass Elastic 70, IC chip bump/ Nickel Elastic-Plastic 20 o C 100 o C 20 o C 100 o C Adhesive matrix/ Resin Inner part of conductive ball/ Resin Coating layer of conductive ball/ Nickel Hyperelastic Hyperelastic Elastic-Plastic 190 o C 20 o C 100 o C 150 o C 190 o C 20 o C 100 o C 150 o C 190 o C 20 o C 100 o C 190 o C 190 o C 20 o C 100 o C 150 o C 190 o C 20 o C 100 o C 150 o C 190 o C 20 o C 100 o C 190 o C -40 o C 100 o C 190 o C -40 o C 130 o C 137 o C 190 o C -40 o C 130 o C 137 o C 190 o C -40 o C 100 o C 190 o C IC Bump IC Chip Resin Adhesive Resin Filling Nickel Coating LCD Panel Figure 2. Meshing of the domain to be analyzed. Finer meshes are constructed in the areas expecting higher stresses. The white area adjacent to the particle is a mesh-free region. The nickel-coated conductive sphere is enlarged as shown on the left. 383

6 -4 Force (N x 10 ) theo 3 um FEM 3 um theo 5 um FEM 5 um DEFORMATION (%) Figure 3. Comparison of the Reverting Force Calculated by FEM and Elasticity Theory Undeformed Deformed Figure 4. Deformed shapes of the monolithic nickel ball and the composite after the clamping pads (not shown) have been compressed toward each other by 2 mm. 384

7 Table 2. of area and maximum tensile stress in the nickel layer at various initial compressive displacement of the pads containing the conductive ball R = 3 R = 5 20% % % % % Table 3. of contact area and maximum tensile stress in the nickel layer R = 3 R = 5 20% % % % %

8 Table 4. of contact area and maximum tensile strength in the nickel layer R = 3 R = 5 20% % % % % % 30% 40% 50% 60% (um) (um) Step Step Figure 5. of contact area at three steps for various deformations left: ball diameter 3 microns right: ball diameter 5 microns 386

9 4. Summary Deformation and state of stress for an anisotropically conductive film (ACF) under supposedly realistic conditions are analyzed using the tool of finite element analysis. To be sure that the FEA modeling was on the right track, a verifying procedure was performed by comparing the FEA solution to the theoretical one for a commonly solvable case. FEA also shows that it is justified to adopt composite balls of resin and metal coating over monolithic metal ones for ACF. The major objective of this paper is to find the extent of variation of contact area between conductive ball and its clamping pads during different step in the process of applying ACF. It is concluded that initial amount of compressive deformation dictates the area of contact that is only slightly affected by the subsequent steps. References [1] H. Nishida et al, Micropitch connection using anisotropic conductive materials for driver IC attachment to a liquid crystal display, IBM Journal of Research and Development, Vol.42, NO.3/4, pp , MAY/JULY [2] F. G Shi et al, Electrical Conduction of Anisotropic Conductive Adhesives: Effect of Size Distribution of Conducting Filler Particles, Materials Science in Semiconductor Processing, vol.2, pp , [4] M. J. Yim and K. W. Paik, The Contact Resistance and Reliability of Anisotropically Conductive Film (ACF), IEEE Transactions on Advanced Packaging, Vol. 22, pp , May [5] M. J. Yim and K. W. Paik, Design and Understanding of Anisotropic Conductive Films (ACF s) for LCD Packaging, IEEE Transactions on Components, Packaging, and Manufacturing Technology-Part A, Vol.21, NO.2, pp , JUNE [6] R. Dudek et al, Flow Characterization and Thermo-Mechanical Response of Anisotropic Conductive Films, IEEE Transactions on Components and Packaging Technology, Vol.22, NO.2, pp , JUNE [7] A. Schubert et al, Proceedings of 2nd Reliability Investigation of Flip Chip in FCOB and FCOG Applications by FEA, IEEE/CPMT Electronics Packaging Technology Conference, 1998, pp.49-56, 1998 [8] M. Yamaguchi et al, Development of Novel Anisotropic Conductive Film (ACF), Electronic Component and Technology Conference, 1999 Proceedings 49th, pp , [9] Ansys, ANSYS User s Online Manuals Release 5.6, Structural Analysis Guide, ANSYS Inc., USA, [3] M. J. Yim et al, Anisotropic Conductive Film (ACF) Interconnection for Display Packaging Application, Electronic Component and Technology Conference, pp ,