Monolayer Glues

The goal of this research project is to join two pieces of solids under normal laboratory conditions (room temperature, normal atmosphere) by chemical activation of their surfaces. The ideal process would simply be the reverse of crystal cleavage, i.e. joining two solids with perfectly clean, flat surfaces by simply pressing them together, which - under exertion of cohesive forces between the surface atoms - should result in a single, interface-free piece of solid. In practice, this ideal has been closely approximated in a process known as wafer bonding, where two silicon wafers with native oxide layers can be bonded irreversibly via a condensation reaction between the surface hydroxyl groups

image.alternative

The goal of this research project is to join two pieces of solids under normal laboratory conditions (room temperature, normal atmosphere) by chemical activation of their surfaces. The ideal process would simply be the reverse of crystal cleavage, i.e. joining two solids with perfectly clean, flat surfaces by simply pressing them together, which - under exertion of cohesive forces between the surface atoms - should result in a single, interface-free piece of solid. In practice, this ideal has been closely approximated in a process known as wafer bonding, where two silicon wafers with native oxide layers can be bonded irreversibly via a condensation reaction between the surface hydroxyl groups

image.alternative

Despite the successful application of this bonding method for the fabrication of novel microelectronic and micromachined devices, two serious drawbacks limit the further exploitation of this technology: First, high temperature annealing up to 1200 °C is required to drive this condensation reaction to completion, which precludes the use of temperature-sensitive components (heteregeneous junctions, doped layer profiles, low-melting metal layers, etc.). And second, water is produced as a side product of the interface bonding reaction, which usually causes bonding defects such as interface voids or bubbles and reduces the bonding strength. Our approach for the development of a new low-temperature bonding process is based on the following concept:

image.alternative

Despite the successful application of this bonding method for the fabrication of novel microelectronic and micromachined devices, two serious drawbacks limit the further exploitation of this technology: First, high temperature annealing up to 1200 °C is required to drive this condensation reaction to completion, which precludes the use of temperature-sensitive components (heteregeneous junctions, doped layer profiles, low-melting metal layers, etc.). And second, water is produced as a side product of the interface bonding reaction, which usually causes bonding defects such as interface voids or bubbles and reduces the bonding strength. Our approach for the development of a new low-temperature bonding process is based on the following concept:

image.alternative

The two substrates are coated before bonding with a pair of monomolecular glue layers (M1-A and M2-B) terminated by appropriate surface functional groups A and B, which undergo spontaneous addition reactions (-A + -B ----> -C-) upon contacting the surfaces to form a strong, irreversible linkage between the substrates. The basic idea of this approach is therefore the replacement of the condensation reaction between surface hydroxyl groups in the conventional wafer bonding process by thermodynamically and kinetically more favorable addition reactions between suitable pairs of activated precursor surfaces, which, in addition to the desired quantitative conversion at room temperature, produce no troublesome side products. Examples for such reactions currently investigated in our group include Diels-Alder reactions, Michael additions and Epoxide Ring Opening Reactions.

image.alternative

The two substrates are coated before bonding with a pair of monomolecular glue layers (M1-A and M2-B) terminated by appropriate surface functional groups A and B, which undergo spontaneous addition reactions (-A + -B ----> -C-) upon contacting the surfaces to form a strong, irreversible linkage between the substrates. The basic idea of this approach is therefore the replacement of the condensation reaction between surface hydroxyl groups in the conventional wafer bonding process by thermodynamically and kinetically more favorable addition reactions between suitable pairs of activated precursor surfaces, which, in addition to the desired quantitative conversion at room temperature, produce no troublesome side products. Examples for such reactions currently investigated in our group include Diels-Alder reactions, Michael additions and Epoxide Ring Opening Reactions.

image.alternative

A novel spectroscopic technique (Infrared Internal Transmission Spectroscopy) is being developed for this project in our group.

image.alternative

A novel spectroscopic technique (Infrared Internal Transmission Spectroscopy) is being developed for this project in our group.

image.alternative

For this method, an IR beam is coupled into the bonded silicon sandwich via a silicon hemisphere at a large incidence angle, in which case the radiation tunnels through the bonding layers and selectivly probes the perpendicular vibrational modes of the interface with greatly enhanced sensitivity. This method is used to monitor the interface bonding reactions between the solid substrates and should allow for a specific selection and fine-tuning of the most suitable glue layer composition and contribute to a better understanding of the molecular basis behind the bonding process.