Background of the Invention As important as manufacturing an integrated circuit is the ability to apply electronic signals to and receive electronic signals from the integrated circuit. Ordinarily. an integrated circuit die is configured to have relatively large exposed metal areas, known as bonding pads. through which this electrical interface can be conducted. Often these metal areas are formed of aluminum or an aluminum alloy can be for example 5 mils, square.

Several well known techniques are commercially used to realize such electrical interface. One such well known technique is conventionally called wire bonding. The completed die is mounted to a lead frame which is integral to the package. such as a dual- in-line packatge (DIP). pin grid array package (PGA). or other pacages equally well known.

The package includes individual pins or other electrical contact devices that are configured for ready coupling to an external circuit or circuit board through soldering, socketing or other well known means. Electrical contact is made between the bonding pads and the lead frame by attaching a thin wire between these contacts. The wire is attached to the bonding pads and the lead frame by heating or ultrasonically welding the wire to the pads.

Unfortunately, the bonding wire introduces impedance into the path of an electric signal due to its thinness and length. The bonding wire acts as an inductor. This impedance operates to add noise to the signal, thereby decreasing the overall operating efficiency of a system including bonding wires. Further, because of the physical requirements of the wire and the

potential for undesired contact between adjacent wires, the spacing requirements are restrictive. The bonding pads are positioned near the edges of the die to allow a shortest path for the bonding wire. The bonding wire can also act as an antenna.

Another well known technique is commonly known as flip chip bonding. According to flip chip bonding, a substrate such as a printed circuit board includes electrically conductive lands that are formed into a mirror image of the bonding pads on an integrated circuit. Generally, the bonding pads are treated with a layer of solder. The integrated circuit and the substrate are mounted in a face-to-face relation to one another and the solder is melted with heat to join the bonding pads and to the corresponding land. In this way, the bonding pads and the lands necessarily provide electrical contact from the integrated circuit to the substrate. Unfortunately, electrical interconnection between two integrated circuits on a printed circuit board, each die using flip chip bonding, requires signal traces along the substrate. These traces, such as on a printed circuit board, introduce impedance into the path of the electric signals and operate to slow the transmission of the signals. thereby decreasing the overall operating efficiency of such a system.

Others have proposed chip to chip bonding techniques. For example, U. S. Patent 5,399,303 to Rostoker discloses a multi-chip semiconductor arrangement using flip chip dies.

Rostoker teaches arrangements using double sided flip chips which have raised bumps on both sides and single sided flip chips which have raised bumps on one side. The double sided flip chips are mounted on the substrate and the single sided flip chips bridge the gap between the double sided flip chips with minimal overlap. The Rostoker invention suffers from the several difficulties in implementation. For example, an upper integrated circuit must be mounted as a bridge between two integrated circuit dice. This requires such an assembly to consume considerable surface area on a substrate such as a printed circuit board, or requires an atypically large integrated circuit package to contain such an assembly. In addition, the upper integrated circuit must be manufactured using complex processing techniques that form electrically conductive vias through the body of that integrated circuit.

Such techniques are complex and thus expensive to perform. Electrical connection to this

assembly is made to the back of the upper integrated circuit using for example conventional wire bonding techniques.

Still others have proposed techniques for reducing the volume and thus surface area consumed by multiple integrated circuit devices. One such technique is taught in U. S.

Patent to Nicewarner, Jr. This technique is not concerned with the number of interconnections between integrated circuits but rather the volume of space consumed by plural integrated circuits. Nicewarner teaches a three dimensional modular assembly of integrated circuits. The chips are mounted back to back and are mounted on both sides of the primary substrate and between each of the two secondary substrates and the primary substrate. Because of the design, the array of chips between the primary substrate and the first secondary substrate must mirror the array of chips between the primary substrate and the second secondary substrate.

Yet others have proposed stacking integrated circuits one on top of another. Such techniques include forming interconnections along the sides of the stack. Heat dissipation from an integrated circuit within the stack can become a problem.

An emerging technology surrounds semiconductor micromachines that are used for forming displays. A device that can be used for such a display is disclosed in U. S. Patent 5.311.360 which is incorporated in its entirety herein by reference.

According to the teachings of the'360 patent. a diffraction grating is formed of a multiple mirrored-ribbon structure such as shown in Figure 1. A pattern of a plurality of deformable ribbon structures 100 are formed in a spaced relationship over a substrate 102.

Both the ribbons and the substrate between the ribbons are coated with a light reflective material 104 such as an aluminum film. The height difference that is designed between the surface of the reflective material 104 on the ribbons 100 and those on the substrate 102 is t/2 when the ribbons are in a relaxed, up state. If light at a wavelength X impinges on this structure perpendicularly to the surface of the substrate 102, the reflected light from the surface of the ribbons 100 will be in phase with the reflected light from the substrate 102.

This is because the light which strikes the substrate travels R/2 further than the light striking

the ribbons and then returns B/2, for a total of one complete wavelength? W. Thus, the structure appears as a flat mirror when a beam of light having a wavelength of A impinges thereon.

By applying appropriate voltages to the ribbons 100 and the substrate 102, the ribbons 100 can be made to bend toward and contact the substrate 102 as shown in Figure 2.

The thickness of the ribbons is designed to be k/4. If light at a wavelength X impinges on this structure perpendicularly to the surface of the substrate 102, the reflected light from the surface of the ribbons 100 will be completely out of phase with the reflected light from the substrate 102. This will cause interference between the light from the ribbons and light from the substrate and thus, the structure will diffract the light. Because of the diffraction. the reflected light will come from the surface of the structure at an angle (H) from perpendicular.

It will become apparent to one of ordinary skill in the art after studying the teachings of the'360 patent that the structure shown in Figure 1 can be used to represent a single pixel of a display. A typical display can contain 1024x1280 pixels arranged in an array of rows and columns. A semiconductor device using pixels such as shown in Figure 1 and having 1024x1280 pixels can have an array that is approximately 1 inch by 1. 3 in extent.

Such a device will likely require at least 2310 I/O structures (1024 + 1280 + 6). The additional 6 I/O pins are for specialized biasing. Using conventional wire bonding bond pad structures, the surface area necessary for making connection to such a device can exceed the active display area by several times. Because the cost to manufacture the device is primarily born by the active display area, such an expense of surface are is excessive and undesirable.

To solve this problem, it is undesirable to manufacture driver structures on the same semiconductor substrate as the active display area. This is due to two primary reasons.

First, the cost to manufacture the devices in the active display area is greater than to manufacture electronic circuit components for making drivers using conventional well known techniques. Second, the processing steps are not the same to manufacture the active displav area as for electronic circuit components. Thus, a processing specification for such a device

would be extremely complex. This would tend to further reduce the vield of such a structure and thereby further increase the cost. For these reasons, it is preferred that the drive electronics be placed in a separate integrated circuit.

Interconnections to such a device could not require the electrically conductive vias nor the bridging technique taught by Rostoker. Further, such an interconnecting structure could not function using a stacked arrangement because light would not be able to impinge onto the surface of the active display area.

What is needed is a method of and apparatus for providing electrical interconnections directly from one integrated circuit to another.

What is also needed is a method of and an apparatus for providing electrical interconnections to an integrated circuit with extremely high I/O requirements.

What is further needed is a method of and an apparatus for providing electrical interconnections to an integrated circuit without obscuring the surface of the integrated circuit.

Summary of the Invention A method of and an apparatus for electrically interconnecting two integrated circuits devices includes mounting the two devices face to face. A first device is mounted for example to a substrate or lead frame. The first device includes a plurality of electrical/physical mounting structures preferably positioned along one edge. The mounting structure provide both electrical interconnection and physical mounting. A second device includes a corresponding plurality of mounting structures configured as a mirror image of the mounting structures on the first device. The mounting structures on the second device are also positioned along one of its edges so that once the mounting structures are brought together in a face to face relationship, the second device cantilevers off the edge of the first device. Under certain circumstances, a dummy block can be mounted to the substrate adjacent to the first device to act as a strut or support for the second device. Under certain other circumstances, an epoxy potting compound can be used to provide structural support.

The mounting structures can be positioned sufficiently close to one another that surface area consumed for I/O is minimized. Another set of electrical interconnect structures can be formed on the surface of the second device on the edge opposite the mounting structures.

Electrical connection can be formed to these electrical interconnect structures using conventional techniques such as tape automated bonding (TAB) or flex connections.

Brief Description of the Drainas Figure I shows a pixel structure from a prior art semiconductor micromachine display device.

Figure 2 shows the structure of Figure 1 under a biased condition.

Figure 3 shows an exploded partial perspective view of the preferred embodiment of the present invention.

Figure 4 shows a side view of a first embodiment of the present invention.

Figure 5 shows a side view of a second embodiment of the present invention.

Figure 6 shows a side view of a third embodiment of the present invention.

Detailed Description of the Preferred Embodiment Figure 3 shows an exploded partial perspective view of the preferred embodiment of the present invention. A first semiconductor device 300 is substantially planar and includes a primary face 302 and a secondary face 304. An integrated circuit device (not shown) is formed in the primary face 302 of the first semiconductor device 300. The integrated circuit formed on the first semiconductor device 300 can be of any type of circuit but is preferably of a semiconductor micromachine display device. However, it will be apparent to one of ordinary skill in the art that the first integrated circuit device 300 could be another conventional circuit such as a microprocessor, controller, PAL, PLA, dynamic or non- volatile memory and the like.

The first semiconductor device 300 is mounted by its secondary surface 304 to a substrate 306 as shown by the dash-dot-dash line. The substrate can be of any convenient

form such as a printed circuit board, a ceramic or an IC package. A plurality of electrical/physical mounting structures 308 are formed near one edge on the primary surface 302 of the semiconductor device 300. The mounting structures 308 are preferably formed of a metal using conventional semiconductor processing techniques. The mounting structures 308 are electrically coupled to the integrated circuit (not shown) using conductive traces (either metallic or doped semiconductor). The mounting structure 308 can be conveniently formed of aluminum or an aluminum alloy. Other metals can also be used.

A second semiconductor device 310 is substantially planar and includes a primary face 312 and a secondary face 314. An integrated circuit device (not shown) is formed in the primary face 312 of the second semiconductor device 300. The integrated circuit formed on the second semiconductor device 300 can be of any type of circuit but is preferably of a driver circuit for a semiconductor micromachine display device. However. it will be apparent to one of ordinary skill in the art that the second integrated circuit device 310 could be another conventional circuit such as a microprocessor, controller. PAL, PLA, dynamic or non-volatile memory and the like.

The primary surface 312 of the second semiconductor device 310 is mounted to the primary surface 302 of the first semiconductor device 300 as shown by the dash-dot-dash line. As can be readily seen the second semiconductor device 310 cantilevers off the edge of the first semiconductor device 300 in an overhanging manner. This limits the amount of the primary surface 302 of the first semiconductor device 300 that is consumed with forming electrical connections to the integrated circuit on the semiconductor device 300. Because of the techniques described herein, the size of the mounting structure 308 can be as small as 50 microns on a side. The spacing between adjacent mounting structure 308 can be limited to 50 microns.

A plurality of electrical/physical mounting structures 318 are formed near one edge on the primary surface 312 of the semiconductor device 310. The mounting structures 318 and portions of the second semiconductor device 310 along with their respective lead lines are shown as ghost lines to indicate that those structures are hidden from view. The

mounting structures 318 are preferably formed of a metal using conventional semiconductor processing techniques. The mounting structures 318 are electrically coupled to the integrated circuit (not shown) using conductive traces (either metallic or doped semiconductor). The mounting structure 318 can be conveniently formed of gold or a solderable alloy. Other metals can also be used.

The mounting structures 318 formed in the second semiconductor device 310 are configured in a mirror image to the mounting structures 308 that are formed on the first semiconductor device 300. In this way, when the first semiconductor device 300 and the semiconductor device 310 are brought together in a face to face relation. the mounting structures 308 and 318 will join together electrically and physically.

Figure 4 shows a side view of the structure of Figure 3. Where appropriate, the same reference numerals will be used to identify identical structures to avoid obscuring the invention in extraneous detail. Note that a large portion of the first semiconductor device 300 is exposed. This is particularly ideal for the preferred embodiment of the present invention where the integrated circuit is a semiconductor micromachine display device.

Under such circumstances, it is essential that the integrated circuit be exposed to receive and reflect or diffract light.

In addition to the structural elements discussed above relative to Figure 3. Figure 4 also shows an electrical interconnection made from other systems to the second semiconductor device 310. A second plurality of mounting structures 320 are formed on the primary surface 312 of the second semiconductor device 310. Preferably, the mounting structures are located on an edge opposite those of the mounting structures 318. As shown, the size and spacing of the second plurality of mounting structures 320 need not be identical to that for the mounting structures 318. A flexible tape connector 322 includes a corresponding set of mounting structure 324 that are configured as a mirror image to those of the mounting structures 320. The flexible tape connector 322 is coupled to the second semiconductor device 310 in the conventional manner.

The process for forming the structure of Figures 3 and 4 follows. Separately, the first and the second semiconductor devices 300 and 310 are formed using semiconductor processing steps known and described in detail elsewhere. Either the mounting structures 308 or the mounting structures 318 or both include a mounting substance such as solder.

Once completed, the first semiconductor device 300 is mounted to the substrate 306.

Separately, a flexible tape connector 322 is mounted to the second semiconductor device 310. Once these two subassemblies are formed, the second semiconductor device 310 with the flexible tape connector 322 are brought to the first semiconductor device 300. The mounting substance is melted with heat or ultrasonically to join the mounting structures 308 and 318 electrically and physically.

Under certain circumstances, it may be determined that the strength of the joined mounting structures 308 and 318 will be insufficient to hold the second semiconductor device 310 in place without failing. Under those circumstances a support structure 330 is used as shown in Figure 5. The support structure 330 is mounted adjacent the edge of the first semiconductor device 300 to which the second semiconductor device 310 is mounted.

The support structure 330 can be formed of a dummy block of semiconductor material or can be formed of a defective/non-functioning device such as the first semiconductor device 300. It is desirable that the height of the support structure 330 approximate that of the first semiconductor device 300. An adhesive material 332 is placed on an upper surface of the support structure 330 to hold the second semiconductor device 310 in place. Preferably, the adhesive 332 is formed of a compressible material so that it will conform to an ideal height for supporting the second semiconductor device 310.

As an alternative to the structure shown in Figure 5, where it is determined that the strength of the joined mounting structures 308 and 318 will be insufficient to hold the second semiconductor device 310 in place without failing a potting compound 340 can be used as shown in Figure 6. According to the preferred embodiment of such a compound, an epoxy potting compound is used. Under those circumstances a support structure 330 is used as shown in Figure 5. The support structure 330 is mounted adjacent the edge of the first

semiconductor device 300 to which the second semiconductor device 310 is mounted.

Preferably, the electrical interconnection 322 is coupled to the second semiconductor device 310 before the potting compound is applied. The potting compound 340 can limited to the region under the second semiconductor device 310, can extend up and over the second semiconductor device 310 as shown at reference numeral 340'or in between. Because the preferred embodiment of the present invention is intended for applications wherein the first semiconductor device 300 is configured to reflect light, it is important in such applications that the potting compound 340 and/or 340'not cover the first semiconductor device 300.

The present invention has been described in terms of specific embodiments incorporating details to facilitate the understanding of the principles of construction and operation of the invention. Such reference herein to specific embodiments and details thereof is not intended to limit the scope of the claims appended hereto. It will be apparent to those skilled in the art that modifications may be made in the embodiments chosen for illustration without departing from the spirit and scope of the invention.

Specifically, it will be apparent to one of ordinary skill in the art that the device of the present invention could be implemented in several different ways and the apparatus disclosed above is only illustrative of the preferred embodiment of the invention and is in no way a limitation. For example. it would be within the scope of the invention to varv the materials and structures of the various components disclosed herein.