The basic idea of this study is based on a previously developed laser spallation technology in which a transient compressive stress pulse of 1-2 nanoseconds (ns) rise-time and 16-20 ns total duration is generated on the backside of a substrate by exfoliating a constrained metallic film by using a 3 ns-long Nd:YAG laser pulse (Fig. 1).

Figure 1. Laser spallation method

The stress pulse is made to propagate towards a test coating deposited on the substrate’s front surface, and whose fundamental interface tensile strength (adhesion) is to be measured. The compressive stress pulse reflects into a tensile wave from the free surface of the coating and pries off its interface. The free surface velocity of the coating during separation is measured using an optical interferometer to quantitatively determine the interface strength. Films with thickness greater than 1 micron could be separated.

Recently it was discovered that the profiles of these stress waves could be changed to one having a rarefaction shock (i.e, having finite rise times but with post-peak decay time of almost zero), after propagating the initial waves through glasses. This allows separation of films with thickness approaching 50 nm, and with interface strengths nearing 2.7 GPa! An interesting extension of this technique (Fig. 2), which is the subject of the present study, is to separate individual lines, squares, and circular thin film patches deposited on a Si or a glass or for that matter on any mother substrate using the laser-generated stress-waves, and catch the separated features or “legos” on a target substrate that is kept in close vicinity of the separating structures.

Figure 2. Pattern transfer idea

Combination with already matured alignment technology can allow development of MEMS structures and IC geometries in an efficient way. In some cases, the traditional wet-etching processing step, which frequently leads to problems of stiction, can be bypassed. However, appropriate adhesion-promoting molecular bonding sites must be chosen for bonding of individual “legos” on the target substrate. Thus, a first generation of circuits and MEMS structures can be constructed from transfer of basic building blocks from a mother substrate at essentially ambient temperature. Thus, structures and substructures can be created on any engineering substrate, including flexible plastics or tapes. Furthermore, films and membranes of any material can be transfer-deposited. Thus, remarkable advance can be made in construction of devices and ICs requiring heterogeneous material integration.

The following applications could be impacted in a major way, if the proposed work is successfully completed.

1.

The current application domain for MEMS and ICs will increase significantly as because of the high temperature processing (e.g., LPCVD polysilicon is deposited at ~600ºC, reflow temperature in most ICs can reach up to 220ºC), present MEMS and IC structures cannot be created on flexible polymeric substrates. Similarly, several sensors involve integration of films made from heterogeneous materials such as polymers and metals and now even biological cells. A polymeric impedance-based gas sensor that uses an SU-8 microwell structure is an example of such a structure [24]. In order to microfabricate arrays of sensors with unique polymers, the integration process must contend with the large volume of solvent that is typically present during polymer deposition. Furthermore, the microfabrication technique must not damage previously deposited polymers. A somewhat involved strategy is to use a permanent microwell structure to contain the polymer–solvent solution in a well-defined sub-millimeter area without disturbing previously deposited polymers. The technology stemming from this research will bypass all these inefficiencies by simply transferring the polymer film deposited on a separate mother substrate to the exact desired location by using the existing alignment technologies.

Preliminary Data

As a precursor to this proposal, a gold wire pattern with individual wires of 1 mm x 60 ?m x 30 nm were patterned on a Si wafer. An epoxy was spun on top of the lines. The laser-generated stress wave, without any glass modification, were used to separate the entire epoxy layer and caught on a regular tape. Figure 3 shows a micrograph of the underside of the separated epoxy layer, showing the separated gold line in its entirety. This provides confidence in the possible success of the proposed work, even though the gold adhesion is expected to be rather low. In a separate preliminary work, an IC package with solder bumps attached to underlying Si pads via intermetallic bonding (Fig. 4) was separated by focusing the laser-generated stress waves on the package’s backside. The separated bumps were caught on a regular tape, which was placed in contact with bumps from top, prior to any stress wave loading. The caught bumps are shown on the separated tape surface in Fig. 5 on the right. This shows the potential for success of the proposed work dealing with the transfer of IC features from Si to plastic substrates.