|
Andy Extance, Member of SpecialChem Technical Expert Team
Introduction
Conducting Formulation Studies
Changing the Rules
Back to Basics
Introduction
As we approach the end of the first decade of the 21st century, the flat screen TV is arguably the most obvious advance that has been made beyond 20th century technology. It's appropriate then that in this characteristically modern product, we find one of the greatest contemporary challenges for adhesive technology.
Liquid-crystal and plasma display TVs, as well as other thin film electronic technologies, demand that electronics are bonded in very close proximity to glass, using the adhesive as a sophisticated addition to the circuit design. Specifically, the challenge is to allow the adhesive to conduct an electrical signal from elevated 'bumps' on an integrated circuit (IC) to the electrodes of the liquid crystal display on the other side of the bond. For this to happen, the adhesive film must be conductive along the bond between the driver IC bumps and the electrodes.
|
Figure 1: An LCD assembly
|
Each IC bump must transmit different signals, rather than simply passing an equal charge across the entire adhesive layer. Therefore, the adhesive must be insulating in between each of the driver-display contacts. Any horizontal conductivity at all will lead to shorting between the bumps. This direction-specific, or anisotropic, conductivity must be achieved with straightforward adhesive dispensing suitable for high-volume manufacturing. Clearly, this is an incredibly advanced adhesive formulation problem.
Conducting Formulation Studies
To make an adhesive into a conductor, it's usual to load small particle metal fillers into the formulation.1 In itself this requires fine scientific judgement, because as the proportion of metal increases, so does conductivity, but at the same time adhesive properties diminish. To make matters worse, at high concentrations the metal filler can precipitate out. This risk can be reduced by adding structurally reinforcing fillers to the formulation, but essentially to devise a conductive formulation demands compromise on adhesive characteristics.
Figure 2: Thermocompression bonding of an Anisotropically Conducting Film (ACF)
For anisotropically conductive films (ACFs,) metal contents are optimised in reactive, solvent-free, epoxy formulations. These must cure in seconds under high-temperature, high-pressure, thermocompressive bonding between the IC and electrodes on the glass surface of the LCD display. This process compresses adhesive between the IC bump and the glass substrate so that there is a greater concentration of metal particles in the compressed space than in the space between the bumps.2
The epoxy adhesives used for ACFs today might typically be loaded with around 20 percent by weight of 4 μm diameter metal particles. Adding up to ten percent of a non-conducting filler, such as 0.8 μm diameter silica, helps improve an adhesive's anisotropy. This is because the additional filler restricts the movement of conducting filler under compression, keeping more particles on top of the bump to pass current.3
Changing the Rules
One of the innovations that helps ensure that an adhesive remains anisotropically conducting are specifically developed insulator-covered conducting particles. These bizarre sounding particles consist of polymer coated in a mixture of nickel and gold, that is then coated with a further insulator layer. The insulator cover layer is fragile, and under the bonding pressure between the bump and substrate it breaks. This leaves the newly-uncovered conductor particles in the compression zone, while in the remaining spaces the filler still behaves as an insulator. Using such fillers has been shown to reduce the minimum pitch - or distance in-between bumps - that specific displays can go down to without shorting from 15 μm to 10 μm. Figure 3:Correlation between anisotropically conductive adhesive layer thickness and number of particles on bump in a 25 μm thick double-layer ACF
Another way of improving anisotropic conductivity is to dispense the adhesive in two separate parts, In such a double-layer arrangement the conducting adhesive is first placed onto the driver IC, and then a more conventional non-conductive epoxy is laid on top. The reasoning here is that when this assembly is cured under thermocompression, the conductive adhesive collects on the bumps as they are forced up through the dual adhesive layer. In the spaces between the insulating conductive filler is diluted, reducing the chances of any electrical shorting.
For such an arrangement the adhesive and associated dispensing equipment must be well tuned for the tiny dimensions involved. The gap between the IC and the glass of the LCD above would typically be only 25 μm in depth. Within this, carefully modifying the thickness of non-conductive and conductive adhesives dispensed can optimise the amount of conductive filler. Experiments suggest that a 7 μm conducting layer and an 18 μm thick non-conducting layer best balances ease of manufacture and best electrical performance.
Back to Basics
Outside of these electrical considerations the manufacturing issues surrounding LCDs and similar products will be familiar to most adhesive formulators. For example, the assemblies undergo thermal cycling that leads to expansion and contraction of the adhesive layers. Including thermally conductive fillers like silicon carbide in the formulation therefore helps improve the flow of heat in a way that reduces physical strain on the system, and lessens the likelihood of a heat buildup thermally decomposing the adhesive.4
|
Figure 4: Microscope image of an ACF in action. Credit: Epson.
|
Another important issue for LCD panels is the ability to deal with warping. For this purpose it is important that the adhesives are designed to have a low elastic modulus, reducing their resistance to warping and the resultant strain. Conveniently, this is achieved fairly well by the double-layer approach. However, with the existing adhesives a great deal of attention has been paid to optimizing curing conditions to keep crosslinking levels and hence elastic moduli low.5 Clearly there seems to be an opportunity here to optimise crosslinker content, so that critical elastic modulus levels are never reached.
Understandably, to date, it's electronics companies who hold many of the patents related to ACFs. Consequently the electrical performance of anistropically conductive adhesives has been well optimised, but there appears to be room for development in physical and adhesive characteristics. The market for flat screen displays is now well established and represents a huge opportunity for adhesive manufacturers. Thanks to this, not only are anisotropically conductive adhesives a high-volume manufacturing product, but they are primed for further contributions from adhesive formulators to help improve processes and lower costs.
References
- "http://www.andisil.com/oneptrtv.html" Petrie, E.M., "Improving the Electrical and Thermal Conductivity of Adhesives" SpecialChem4Adhesives.com, Jul 6, 2005.
- Fu, Y. ; Liu, J.; Willander, M. J. Electron. Manuf 1999, 9, 4, 275–281
- Yim, J. Y.; Hwang, J.; Paik, K. W.Int. J. Adhes. Adhes.2007, 27, 77–84
- Yim, J. Y.; Paik, K. W. Int. J. Adhes. Adhes., 2006, 26, 304–313
- "http://www.andisil.com/oneptrtv.html" Rizvi, M. Y.; Lu, H. ; Bailey, C.; Chan, Y. C.; Lee, M. Y.; Pang, C. H. Microelectronic Eng., 2008, 85, 238–244
 |