ASML – Advanced Adhesive Bonding Material Characterization

Background

Whether you are reading this on a computer, on a smart phone, on a tablet, or on any other device, you are using ASML’s lithography products right now. In lithography, an image on a reticle is projected onto a wafer to etch the design into the wafer. Figure 1 shows a simplified version of some of the key components in the process. The bottom section shows the round wafer in a square carrier, while the middle section shows the rectangular reticle. The top circular section is part of the optical system.

Figure 1 – Simplified Lithography Process

Improvement in the lithographic process is so important because as lithography improves, so does the speed and capability of the microchips that are created from the process. Following “Moore’s Law”, advances in lithography systems, such as the one shown in Figure 2, have allowed chip manufacturers to continue to shrink feature sizes, with the current state of the art possessing the ability to manufacture feature sizes in the range of 22 nanometers, or about the size of 100 silicon atoms.

Figure 2: Lithography System

With feature sizes in that range, tolerance budgets are pressed to the atomic level. All components and processes contributing to the precision of the equipment must be optimized. Key to the process are the numerous optical components in the system, some of which have extreme tolerances. For examples, the flatness of some of the precision glass mirrors are controlled so tightly that if one of the mirrors were the country of Germany, the biggest bump would be the thickness of a penny.

Glass is a brittle material, subject to cracks and fracture with minimal stress. Once cracked, the optical path is disturbed, leading to the overall system not serving its functions. Care must be taken to properly package the glass structure, and over constrained mounting systems can pre-stress the glass before its field use. Also, most applications will see changes in temperature. Thermal expansion and contraction of an optical structure in a rigid mount would also crack the glass and cause system failure.

The optical path is also subject to tight tolerances for the path transmission. When the light hits a prism to change direction, a minor error in angular alignment can lead to a large linear error at the target surface. “Abbe error” or “sine error” (most commonly, “Abbe error” refers to the effect, “sine error” refers to the error value) is defined as the total linear error from errors in angular positioning. As shown in Figure 3, this error is small across small linear distances, but large across large linear distances. At the nanometer level, a microradian angular misalignment can cause a beam to completely miss its target

Figure 3: Abbe Error

Common optical mounting systems often feature kinematic mounts with displacement relief flexures for thermal expansion and dynamic events. Adjustability with a focus (no pun intended[i]) on Abbe error reduction can be provided for accurate alignment, including a locking feature to keep the optic in place once the desired precision is obtained, although adjustability comes with a trade-off in assembly time and stability. Direct metal-to-glass contact is usually avoided unless the metal is flexured. Glass-to-adhesive-to-metal is common, although not “Design for Manfacturing and Assembly” (DFMA) friendly and subject to instability issues from creep, fatigue, humidity, and other effects. Adhesive joints are often over-simplified and under considered in the structural system, and common difficulties in DFMA (failure to cure, bond thickness / area variation, voids, etc.) typically lead to material properties much different than values published from the vendor. Adhesives also typically eliminate the potential for disassembly for service / rework, and will outgas, a limiting factor in many applications in ASML’s clean room environment. Do not underestimate the engineering complexity of a simple “glue joint.”

Problem Statement

Students will be challenged to research adhesive for optical bonds. This will involve design and optimization of manufacturing equipment, analysis of predicted material properties, and tests to validate analysis on various types of adhesive bonds used on optical components. These should include but are not limited to:

•  Design of experiments for finding ideal flow related geometry for epoxy injection. How big should the over flow gate be as a function of bond pad diameter and height, epoxy viscosity, injection hole size?

-  Adhesives: DP460, DP490, Araldite 2030

•  Design of experiments to characterize ultra-thin bond lines which occur as a byproduct of wicking.

-  Students to research wicking phenomenon and characterize properties

+  What are their material properties (tensile strength, peel strength, thermal stresses) compared to traditional bond lines?

•  Design of experiments to develop a prototype statistical number of “Glue wicking barriers” and judge their effectiveness

-  ASML Design vs. student design

•  Design of experiments to quantify residual bond preload. Time varying results will be difficult because they’re slow, but perhaps vary temperature (minor room temperature, elevated cure) time, humidity, dynamics (shaker table), etc. to see if bond preload relaxes. A prototype could be a simple as a thin glass plate bonded over deep epoxy wells and watch the glass deformation relax.

-  Strain gauges studying variations in parameters (applicable to clean room deltas)

+  Students to research testing standards

End Game

In the end, if all goes as planned, you will have an important contribution to the body of knowledge concerning adhesive bonds. Data such as this would be of tremendous benefit to ASML and its endeavors to keep Moore’s Law alive for the next generation, but should also be of general value to industry in general.

This is by no means an easy endeavor, but we believe your group is up to it and you have 9 months to show us what you got.

If you have any questions, please contact:

Andy Judge

ASML Group Lead – Mechanical Development, Reticle Positioning

1 203 761 4328

Chris Reed

ASML Group Lead – Mechanical Development, Optical Modules

1 203 761 6780

While they will help you, they will not do everything for you. Good luck (and skill)!

[i] Actually, pun may have been intended