Self-filleting technology using smart die-attach paste
09/01/2009
Executive OVERVIEW
Strong consumer demand for increased stored content on cell phones, MP3 players, and digital cameras continues to drive the memory market to higher production volumes. This is an intensely competitive marketplace that pushed selling prices lower in 2007 vs. 2006 and even lower still in late 2008 and early 2009. In light of these market developments, semiconductor companies are seeking lower-cost die attach material solutions to replace the higher-cost, film-based adhesive materials used for certain applications. One such material is self-filleting die attach paste.
Debbie Forray and Ilya Furman, Henkel Corp., Irvine, CA USA
Recent product developments in self-filleting die attach technology have resulted in a low-cost, high-reliability, high-volume-compatible solution to the cost quandary for production die stacking. This latest material advance offers performance comparable to film without the adverse effects of potentially damaged wire bonds, warpage problems, or reduced machine throughput ??? issues that often accompany film-based processes. In addition, self-filleting’s competitive material price and required material investment in only known good die (KGD), compared to film, which covers the entire wafer including bad die, results in some packaging firms saving as much as 25% in die attach materials cost as compared to that of current film adhesives. Furthermore, if one factors in the equipment investment required to facilitate a film process, the savings ??? as high as 50% comparatively ??? are more profound, making self-filleting an increasingly appealing alternative.
Defining self-filleting die attach
As die become thinner and design rules for new package types become tighter, traditional die attach methods and processes reach a point of diminishing returns. For example, a standard paste die attach material process requires that material is dispensed and a bond force applied that is large enough to force the adhesive to form a fillet around the edge of the die. Typical package assembly specifications require 100% coverage of die attach paste under the die and the presence of a fillet surrounding the entire die. When the package requires very tight assembly tolerances (100??m or less) or very thin die (75??m or less), it is difficult to control fillet height and length consistently. In such a scenario, the traditional die attach process can result in adhesive contamination of the die top (Fig. 1) or wire bond pads. Because of this, device assemblers have historically turned to adhesive film technology but it too has challenges, discussed later in this article.
New developments in the form of self-filleting die attach adhesives address the limitations of traditional die attach pastes, while also providing a lower cost alternative to film. Self-filleting adhesive provides many of the same benefits of traditional die attach pastes without the problematic issue of inconsistent fillet control. The self-filleting dispense process follows that used for traditional paste: a dot or pattern is drawn on the substrate and the die is placed with some bond force (Fig. 2).
The die placement process for a self-filleting paste deviates from standard die attach methods: the bond force needs to be high enough to start the flow of the material but low enough so the maximum coverage under the die is ~80% to 90% upon release of the die from the pick-and-place tool. After the die is released, the capillary force of the material enables controlled flow to the edge of the die, where it stops and forms a fillet.
Self-filleting process parameters
To fullly realize the advantages of self filleting adhesives, the materials need to be processed slightly differently than traditional pastes. The initial goal of the dispense and die attach processes is to ensure that there is no overflow or fillet creation upon die placement, which allows the material to flow and fillet on its own and, in the process, minimize fillet length.
The most important process parameter is dispense volume, which will control paste overflow and set the bondline thickness. In addition, dispense pattern shape and centering are important, as pattern shape can determine the flow speeds and have an effect on the fillet length. Dispense pattern is somewhat dependent on die size but, in most cases, a double-cross pattern is recommended, with the material distributed towards the edges of the die. Bondforce should be set as low as possible to achieve full coverage, with bond forces of 100g or less sufficient for most applications.
While die attach temperature can be manipulated to decrease coverage time, this technique is not recommended as a too-high temperature may cause the material to cure before the fillet is formed by the self-filleting mechanisms. Actual time to achieve full coverage is primarily dependent on the coverage percentage immediately after die attach and the rheological properties of the given adhesive.
Controlling bondline and tilt with spacers
Bondline control in tradional die attach is accomplished using bondforce and time to distribute the material until the desired bondline is reached. In the case of self-filleting pastes, bondline control is established through volume alone. Although a bondforce is still required to start the flow of the material, it is significantly lower than that for standard die attach pastes. A traditional die attach may require a bondforce of 3???4 Newtons in comparison to approximately 1 Newton for a self-filleting material.
Die attach spacers ??? typically spherical particles with a very tight distribution ??? are often used to control the bondline. Organic or metallic in nature, spacers in such applications establish bondline thickness and help control die attach tilt. Generally, spacers are used at a low quantity in a die attach adhesive: ~200 spacers in a 10 ?? 10mm2 area, for example.
 Figure 3. Spacers are used to control bondline tilt in die attach pastes. |
The second advantage of spacers is their ability to control bondline tilt. Bondline tilt occurs when a silicon die is placed non-parallel to the substrate either by improper equipment set-up or inconsistent paste dispense. In self-filleting adhesives, the spacers can act as pillars that flow out with the paste to ensure that the silicon is set parallel to the substrate. Figure 3 shows a cross-section of a same die stack set up using a spacer to control bondline and tilt.
Self-filleting vs. film
As referenced previously, because of the inherent drawbacks of traditional die attach pastes, many packaging firms have migrated to film die attach materials to achieve the processability and performance required for today’s memory and stacked die applications. There are several benefits to film, one of which is conrolled material flow when attached.
With the advent of self-filleting paste, some of film’s advantages can be achieved with die attach paste at a fraction of the cost. Paste manufacturing costs are lower than that of film. In addition, there are process-related cost reductions that are realized through the use of self-filleting pastes:
- Reduced material waste: Film covers the entire wafer ??? both known good and known bad die ??? with adhesive. In a well-defined process, yield will range from 80% to 90%, which means that 10???20% of the film product is wasted. Paste is only used on the KGD, reducing the quantity needed and saving cost.
- Reduced capital equipment expenditure: The standard manufacturing process for dispensing paste is already defined, with required equipment in place in virtually every global packaging house. Film requires new equipment investment and adoption of different processes.
In addition to cost savings, reliaiblity advantages may be achieved when using self-filleting technology in place of film-based materials. One of the more predominant reliability issues associated with film technologies is defects caused by void entrapment during the lamination process. Undulations in substrates and/or die can challenge film materials, as they have difficulties filling or flowing into or around these obstacles. As pastes are liquid mediums, fill and flow are not limiting factors.
With the industry moving toward ever-thinner die, the pressure required in the film lamination process may also pose challenges. When placed under enough force, today’s very fragile wafers can crack. With self-filleting die attach paste, minimal bondforce is required to dispense and place the die; die cracking is significantly reduced or eliminated. In same size die stacking designs, film’s effectiveness is also somewhat limited. Although advances have been made using flow-over-wire films (FOW), few have been exceptionally successful in resolving problems such as trapping voids between the wires and/or bending the wires. A self-filleting paste may eliminate some of these film-related issues.
Conclusion
With the development of self-filleting technology, packaging specialists now have an effective film alternative for certain applications (Fig. 4). The lower cost and processability of self-filleting materials make them an attractive alternative for many manufacturers who may currently use film-based technologies.
Figure 4. Self-filleting paste has been used successfully on many applications as an alternative to die attach film adhesives.Click here to enlarge image To facilitate the stacking of die for more applications such as memory, traditional die attach pastes have steadily been replaced with film-type die attach, as die stacking designs make it difficult to use traditional paste materials. Now, however, self-filleting materials may mark the re-emergence of die attach paste as the material of choice for packaging firms seeking to lower costs and maximize the life of existing capital without sacrificing performance. Though self-filleting materials are not a film replacement for all die attach applications, they are indeed a groundbreaking advance for many.
Acknowledgments
The authors would like to thank Minghai Wang of Henkel Corporation and Michael Matthews, formerly of Henkel Corporation, for their contributions.
Debbie Forray received her BS in chemical physics from San Diego State U. and is a senior scientist at Henkel Corporation, 15350 Barranca Pkway, Irvine CA, 92618 USA; ph.: 858-536-4705; [email protected].
Ilya Furman received his BS in chemical engineering from Lehigh U. and is an applications engineer at Henkel Corporation.
More Solid State Technology Issue Articles