WO2014121300A2

WO2014121300A2 - Photonic data transfer assembly

Info

Publication number: WO2014121300A2
Application number: PCT/US2014/014740
Authority: WO
Inventors: Douglas R. Hackler, Sr.; Dale G. Wilson
Original assignee: American Semiconductor, Inc.
Priority date: 2013-02-04
Filing date: 2014-02-04
Publication date: 2014-08-07
Also published as: WO2014121300A3

Abstract

The described Photonic Data Transfer Assembly (100) is a flexible device capable of receiving, processing and transmitting photonic signals which can be activated, deactivated, tuned and controlled by included electronic circuitry which 5 may be controlled wirelessly using Radio Frequency (RF) communication. Flexible photonic waveguides and flexible electronic circuits are integrated with flexible interconnects into a smart card format only 0.25 mm thick. Waveguides (130) fabricated in semiconductor wafer form and converted to Semiconductor-on- Polymer (SOP) provide routing for light in three dimensions. Sub-micron sized 10 features offer flexibility for conformal mounting and maintain performance during dynamic deformations. Flexible SOP transponders (160) integrated with printed antennae (180) use RF signals to wirelessly transmit data from the flexible electronic-photonic circuit. Interconnections (190) use thin flexible material, such as polymer, which supports printed lines that connect pads and may contain vias. 15 Attachments use conductive and non-conductive epoxies to remain flexible while conforming to underlying topography.

Description

PHOTONIC DATA TRANSFER ASSEMBLY

TECHNICAL FIELD

The present invention relates generally to integrated circuits that are flexible. In particular, the described devices and methods use standard semiconductor processing techniques to develop assemblies that include optical as well as electrical circuits that operate in a flexible format such as a smart card.

BACKGROUND ART

In recent years, silicon photonics has attracted attention as an emerging technology for optical telecommunications and for optical interconnects in microelectronics. Based on sophisticated silicon semiconductor technology, silicon photonics can provide an inexpensive highly integrated electronic-photonic platform, in which ultra-compact photonic devices and electronic circuits converge.

Many functions and techniques must be called upon to produce any circuit assembly that is flexible. This is made more complicated when the convergence of interfaces between photonics and electronics is desired. To be truly flexible each of the components incorporated into the assembly must be individually flexible. In addition, the interconnections between the components must not impede flexibility.

To date any convergence of photonics and electronics has been based upon conventional construction techniques. Semiconductor die are mounted to conventional substrates using traditional methods of die attach such as eutectic, solder or epoxy bonding techniques. Subsequent formation of interconnects between the die and package, between one die and another, or from a die directly to a circuit board typically has been accomplished by wire bonds or bump bonds or solder. Individual bonds have been made independently, one at a time. These traditional mounting and interconnect methods are quite effective for rigid die, but fail to meet most requirements for flexible electronics. While epoxy die attach is well suited to flexible assembly, there is no support to be derived from these traditional interconnects.

A more recent method of providing flexible interconnects to flexible substrates uses flexible springs. Flexible semiconductor circuits are generally available and flexible "plastic" CMOS has been demonstrated, but a means of interconnecting them is not presently recognized. A primary requirement of an assembly such as that of present interest is some form of transponder to provide communications capability. A transponder is a device that emits an identifying signal in response to reception of an interrogating signal. Transponders, as used in applications such as smart cards, function as traditional transponders with contactless capability. They require no battery and are powered and read at short ranges via magnetic fields using electromagnetic induction. The wireless non-contact utilization of radio-frequency electromagnetic fields is also utilized to power logic and memory operations in the assembly and to transfer data from a card to an object such as a card reader.

A transponder in a smart card format is a type of data storage and/or computing device that is commonly used for contactless or hybrid smart cards. The device is a complex rigid assembly that includes one or more integrated circuit (IC), an antenna with a substrate, connection of the chip's bond pads to the substrate and a molded body to protect the chip. The ICs used in smart card transponders are very limited in die area due to reliability issues associated with the deformation of cards encountered during typical use. Rigid IC's fracture and break when bent. The larger the IC, the greater the failure rate. Transponder assemblies used in smart cards are typically 0.5mm (500um) in thickness and are individually inlayed in a complex cavity formed in a card body that is commonly made of PVC. For contactless smart cards the antenna is commonly a coil of copper wire. The antenna is integrated as an additional card inlay in another cavity on the same card and connected to an IC to provide wireless communication and enable RFID (RF Identification) capability. The requirement for a cavity limits the card thickness and increases the cost of manufacturing.

DISCLOSURE OF THE INVENTION

The Photonic Data Transfer Assembly (PDTA) described here is a flexible device capable of receiving, processing and transmitting photonic signals. The photonic signals can be activated, deactivated, tuned and controlled electronically by included electronic circuitry which in turn may be controlled wirelessly using radio frequency communication. The PDTA comprises flexible photonic waveguide circuits and flexible electronic circuits integrated with flexible interconnects. Present day photonic waveguides fabricated using semiconductor wafers are limited to routing light in two dimensions, that is, in x and y directions, within the plane of rigid crystalline silicon. By fabricating waveguides in a semiconductor wafer form and then converting those wafers to Semiconductor-on-Polymer (SOP), the waveguides can provide routing for light in x, y and z directions. The SOP format provides for sub-micron sized features that are flexible for conformal mounting or capable of maintaining performance while being deformed dynamically. The result is a fully flexible sub-micron feature-capable waveguide.

Electronic interface to the PDTA is accomplished by a flexible transponder which provides the device with capability to transmit data wirelessly using Radio Frequency (RF) signals from the flexible electronic-photonic circuit. Integrated circuits are necessary to achieve the complexity required for transponder

functionality. Ultra-thin flexible SOP Integrated Circuits (ICs) are integrated with a printed RF antenna and can be laminated as a layer without the use of cavities or cutouts. This reduces cost and simplifies manufacturing. One embodiment of the flexible transponder places the IC and the RF antenna in a flexible hybrid electronic system that is printed on a flexible substrate, including bonding of the IC on the flexible substrate in contact with the RF antenna. Since the flexible transponder is ultra-thin and flexible, it is not subject to the reliability failures associated with the deformation of conventional rigid transponder assemblies. This important feature eliminates limits on die size for reliability and enables the use of larger ICs and arrays of ICs for large scale memory and processing.

The devices described here are interconnected using flexible circuit interconnects. These are flexible overlays that can bridge between the devices that are to be interconnected, conforming to the underlying topography. They remain flexible and are capable of routing interconnect signal paths and providing low resistance electrical contacts. A basic interconnect includes a thin flexible material with at least one printed line having a connection pad at each end of the line to create a flexible interconnect. More complex interconnects may include multiple electrical conductors with the addition of electrical insulators serving as thermal conductors to a heat sink. BRIEF DESCRIPTION OF DRAWINGS

The particular features and advantages of the invention briefly described above as well as other objects will become apparent from the following description taken in conjunction with the accompanying drawings, in which:

FIG. 1 depicts an exploded view of a Photonic Data Transfer Assembly

(PDTA);

FIG. 2 shows the components within an assembled PDTA;

FIG. 3 shows a pair of 2-D Photonic Waveguides on a rigid substrate;

FIG. 4 shows two pair of 3-D Photonic Flexible Waveguides on polymer substrates, one as convex and one as concave;

FIG. 5 depicts traditional 2-D Photonic Waveguides fabricated in SOI (Semiconductor-On-lnsulator);

FIG. 6 illustrates a 3-D Flexible Waveguide fabricated in Semiconductor-on- Polymer (SOP) as a concave flexible structure;

FIG. 7 illustrates a 3-D Flexible Waveguide fabricated in SOP with a backside polymer;

FIG. 8 depicts a generic integrated circuit as an unmounted rigid die;

FIG. 9 illustrates attachment of the die of FIG. 8 to an antenna assembly;

FIG. 10 shows a typical smart card having a cavity for reception of a die and antenna assembly;

FIG. 1 1 shows the antenna assembly with an attached die mounted in the cavity of the smart card of FIG. 10;

FIG. 12 illustrates sealing of a top cover to FIG. 1 1 to produce a

conventional contactless smart card;

FIG. 13 depicts an unmounted rigid die without and with requisite bonding wires;

FIG. 14 illustrates placement of the rigid die with bonding wires into a cavity in a conventional smart card;

FIG. 15 depicts an unmounted ultra-thin die produced by a SOP process; FIG. 16 illustrates an adhering of the SOP die to a printed antenna assembly with subsequent lamination and sealing to produce the flexible smart card transponder of FIG. 17; FIG. 18 shows an ultra-thin die as depicted in FIG. 17 attached to a printed card body with contacts and vias, without wire bonds or molding, to produce a flexible smart card without a cavity;

FIG. 19 is a cross-section of a basic flexible interconnect showing two layers of metal with pads;

FIG. 20 illustrates the flexible interconnect of FIG. 19 when flexed;

FIG. 21 is a top view of a flexible interconnect including pads with vias;

FIG. 22 is a cross-section view of flexible interconnect of FIG. 21 ;

FIG. 23 shows the flexible interconnect of FIG. 22 when flexed;

FIG. 24 depicts in cross-section a flexible interconnect with vias

interconnecting pads of a flexible printed circuit board (PCB) with the pads of a semiconductor die;

FIG. 25 shows a top view of a flexible interconnect providing multiple interconnections; and

FIG. 26 illustrates multiple semiconductor die connected by a flexible interconnect to each other and to the underlying substrate.

BEST MODE FOR CARRYING OUT THE INVENTION

The Photonic Data Transfer Assembly (PDTA) demonstrates the

convergence of photonics and electronics. The described PDTA is a flexible device that can receive and send photonic signals. These can be activated, deactivated, tuned and controlled either by autonomous onboard electronic circuitry or remotely through the use of wireless radio frequency communication. The general utility of the PDTA becomes apparent when the assembly is constructed in a smart card format. This readily identifiable and accepted format provides considerable data storage and computation capability with extreme portability. As shown in FIG. 1 and FIG. 2 the described PDTA comprises flexible optical circuits, in the form of photonic waveguides (130, 150), and flexible microelectronic circuits (160, 165) which are integrated with the use of flexible interconnects 190.

A flexible substrate 170 supports a variety of circuit components. One of these is an ultra-thin flexible waveguide 130. A second instance of a waveguide 150 has been integrated with an active optical device 140 for application to the flexible substrate 170. A radio frequency (RF) circuit in the form of a

Semiconductor-on-Polymer (SOP) integrated circuit (IC) 160 is added to the assembly where it couples to flexible antenna 180 which has been printed onto the flexible substrate 170. All elements are coupled appropriately by flexible

interconnects 190. Where compatible processes are available, many, if not all, of these elements may be fabricated directly onto the flexible substrate 170. In other situations, each of the elements may be fabricated in wafer form after which the separate ultra-thin die may be bonded to the flexible substrate 170. The working elements of the assembly are then protected by lower and upper layers of thin flexible package material (110, 120).

The completely enclosed PDTA 100 shown in FIG. 2 is capable of receiving and transmitting electrical signals by means of flexible antenna 180 which is supported by processing elements in an SOP IC 160. Flexible antenna 180 also serves to inductively couple electrical power needed to drive the circuits of the PDTA 100. SOP IC 160 is also coupled to active optical device 140 which in turn is coupled to a waveguide 150 for transmission of optical signals. A second SOP IC 165 interfaces between waveguide 150 and waveguide 130 for additional optical communications.

The following discussion will elaborate on each of the subassemblies of the PDTA. These are the photonic waveguide, the radio frequency transponder and the interconnects between them.

Flexible 3-D Photonic Device

Ultra-small geometrical silicon photonic structures have been demonstrated as photonic waveguides. Integration of these waveguides with microelectronics provides a highly integrated platform for electronic-photonic convergence. The practical achievement of this platform requires reduction of such factors as the propagation and coupling losses in the interface to external fibers. State-of-the-art technologies specially tuned to the fabrication of nanometer structures, and the fundamental propagation performance has already become a practical standard. Some passive devices, such as branches and wavelength filters, and dynamic devices based on the thermo-optic effect or carrier plasma effect have been developed by using silicon photonic wire waveguides. These waveguides also offer an efficient media for nonlinear optical functions, such as wavelength conversion. Although polarization dependence remains a serious obstacle to the practical applications of these waveguides, waveguide-based polarization manipulation devices provide effective solutions, such as a polarization diversity system.

Bonding a wafer, individual die, or SoP device from a lll-V semiconductor, such as GaAs (Gallium Arsenide) or similar photonic material, onto the flexible silicon prior to demount can create an active region for lasers, amplifiers, modulators, and other photonic devices using standard processing techniques on the pre-demount flexible substrates. Following demount, additional photonic devices may be mounted to the opposite side of a flexible photonic waveguide structure for stacking of devices in three dimensions (3-D).

Waveguides are an essential component of photonic circuits. The presently described devices are flexible silicon strip photonic waveguides routed in silicon to create interconnects and couplers. These are photonic structures having sub- micron features that are integrated with CMOS (Complementary Metal Oxide Semiconductor). Single crystalline silicon structures are well known by those skilled in these arts to be effective for photonic waveguides.

Current structures utilize rigid wafer semiconductor substrates where the waveguides can be routed in the two-dimensional plane of the crystalline silicon, as depicted in waveguides 220 of FIG. 3. Implementing photonic devices using a Semiconductor-on-Polymer (SOP) process that has been developed for flexible CMOS results, as depicted in FIG. 4, in flexible single crystalline silicon structures that can be deformed into the third dimension. Such fabrication of photonic waveguides (230, 240) results in photonic devices that are flexible and provide for the routing of light in three dimensions.

In a Semiconductor-on-Polymer (SOP) process, such as that described in U.S. Patent No. 6,762,510 entitled "Flexible Integrated Monolithic Circuit" issued to Fock et al., a semiconductor wafer, such as one upon which CMOS circuitry has been fabricated, is coated with a polymer. The polymer conforms well to the CMOS circuitry and is cured to a solid. A carrier substrate is then temporarily bonded to the polymer. This carrier is used as support of the intermediate assembly while the original CMOS substrate, that is, the handle silicon, is removed by processes that may include grinding and etching in order to reduce the original substrate to less than about 12 μιτι. While still supported by the carrier substrate, additional devices may be fabricated, including complex integrated active devices, passive devices and interconnects. The ultra-thin semiconductor substrate with its CMOS devices intact is then released from the carrier substrate by breaking its temporary bond to the polymer. This results in a flexible integrated circuit in a SOP format.

Three-dimensional waveguides can be patterned in situ with silicon mesa isolation. Other photonic material can be integrated into the semiconductor wafer prior to the SOP processing. When used in conjunction with the SOP process, such waveguides can be fabricated simultaneously with flexible CMOS so that the resultant integrated circuits and waveguides are both flexible. Devices fabricated using lll-V materials, such as Gallium Arsenide (GaAs), and other photonic materials are bonded to the photonic circuits for the integration of lasers and diodes. These devices can be bonded to the waveguides while they are either still mounted on their original rigid carrier wafer or after they have been demounted from the wafer. The devices can either be thinned to the point of flexibility or be made small enough so that they can be surface-mounted to the SOP waveguides without impeding the flexibility of the SOP wafer. Integration with CMOS provides ready connectivity to electronic inputs and outputs. SOP processing results in flexible single crystalline photonic lll-V and Silicon materials.

Other embodiments of such integration include alternative semiconductor technologies such as GaN (gallium nitride), nanotubes, graphene, ferroelectric, magnetic memory, and chalcogenide. These and other late FEOL (Front End-of- Line) process options are applied as part of the photonic circuit, contributing to electronic or photonic functions.

The characteristic flexibility of the described devices enables these photonic waveguides to conform to a variety of radii. They are not affected adversely by deformation during storage, and they are physically robust, resisting damage from being dropped or other impact. The described design is adaptable to heatsinks, external device interconnects, high-temperature flexible materials other than SOP polymer, and claddings on the external surfaces of the semiconductor waveguides.

A traditional two-dimensional photonic waveguide fabricated in SOI

(Semiconductor-On-lnsulator) is depicted in FIG. 5. The photonic circuits and SOI waveguides 280 are fabricated using conventional methods on a buried oxide (BOX, 250) supported by handle silicon 260. This is protected by a passivation layer 270 resulting in a rigid two-dimensional device.

Unlike the traditional two-dimensional photonic waveguide depicted in FIG. 5, the photonic circuits described here may be mounted to conform to concave or convex surfaces depending upon a variety of applications and the environment in which they are to be stored and used. This capability to conform is enabled by removal of the rigid handle silicon 260 (FIG. 5).

The flexible photonic waveguide under present consideration is built upon a sub-micron single crystalline SOP layer supported by a flexible substrate with an intervening isolation layer. Passivation of the SOP provides an additional isolation layer effectively cladding the waveguide to confine light within the SOP. The isolation material may be any substance that supports confinement of the optical mode, such as silicon dioxide or silicon nitride.

A polymer layer 290 is applied to the passivated surface of the IC for support while the handle silicon 260 is removed (FIG. 6) and replaced by a flexible backside polymer 295 (FIG. 7). The result is a flexible photonic waveguide completely encapsulated in a polymer coating. These waveguide assemblies may be reshaped dynamically as part of a tuning mechanism, or they may be simply adjusted to conform to various environmental conditions. They may be controlled by electronic circuits through control elements such as MEMS (Micro-Electro- Mechanical System) mirrors, PIN diodes and other similar devices to route, amplify or delay photonic signals.

Since their flexibility is a consequence of having been constructed on a polymer substrate, wherein the handle wafer, whether silicon or another material, has been removed, these devices are inherently thin. This offers a benefit of reduced propagation losses. In general, single crystalline materials enable high mobility electronics and high performance photonics.

Their thinness also allows these devices to be stacked, enabling

construction of waveguides and other complex structures capable of routing signals in three dimensions, since photonic devices may be mounted to opposite sides of a flexible photonic waveguide structure. Alternatively, multiple flexible photonic waveguides may be stacked to increase functionality with light transfer occurring between the stacked waveguides. Waveguides within a stack may be optically coupled or they may be optically isolated. The flexible photonic circuit allows for integration of photonic devices such as low threshold lasers, tunable lasers, and other photonic integrated circuits with flexible Complementary Metal Oxide

Semiconductor (CMOS) integrated circuits. Multiple layers allow for highly complex systems with interconnects and devices, active or passive, fabricated within the layers prior to being stacked.

Significant benefits are gained by the merger of silicon and lll-V

semiconductor devices into flexible waveguides. Silicon is used to produce well understood waveguides, whereas lll-V devices are integrated to generate, amplify and modulate light for the waveguides. Other photonic materials may be integrated to affect the propagation or other characteristics of the light. CMOS devices convert the processed light into electronic signals and stimulate the lll-V materials. Construction of all of this on a SOP substrate provides a complete photonic circuit with the flexibility to route light in three dimensions.

It will be recognized by those skilled in these arts that many variations of the described embodiments are possible. Although silicon is the most likely material for constructing the photonic waveguides, any single crystalline wafer material is a feasible candidate. Furthermore, deposited materials such as TEOS (Tetra-Ethyl- Ortho-Silicate), polysilicon, amorphous silicon, silicon nitride, silicon carbide, gallium nitride or others may be used. Additional usable materials include graphene, nanotubes and non-crystalline materials. Each of these would still benefit from the flexibility afforded by sub-micron features. The essence of the presently described method is the fabrication of waveguides in wafer form and their subsequent conversion to flexible SOP.

Flexible Transponder

For a device such as a smart card to be useful it must have a means of communicating beyond itself. A radio frequency transponder as used in a smart card format consists of an integrated circuit (IC) computing device and an antenna. Conventional smart cards, described here in FIGs. 8-14, are constructed around, and constrained by, a rigid IC die 310 such as that shown in FIG. 8. A rigid IC is necessarily limited in size by the fact that larger ICs suffer a greater failure rate due to fracturing when they are subjected to bending. The computational and/or data storage capacity of the IC is to some extent limited by its size.

An antenna assembly 320 (FIG. 9) is conventionally formed from a coil of copper wire with some provision for connection with a bonding region 330 to which the IC 310 is attached. The foundation of a conventional rigid smart card 350 is formed, as shown in FIG. 10, with a complex, sometimes multilevel, recessed channel 370 into which the antenna assembly 320 with attached IC 310 is placed (FIG.1 1 ). A typical conventional smart card is formed from PVC and has a thickness of about 0.5 mm. The recesses necessary for mounting of the working components are either molded or milled into this foundation. After the antenna assembly has been inserted into the recess of the rigid smart card foundation and the computation circuitry of the IC has been connected to it (FIG. 1 1 ), a top cover 380 is placed over the foundation and sealed to produce the finished product (300) of FIG. 12.

For conventional smart card applications of a more general nature, a rigid IC 310 may be affixed to an exterior contact substrate 340 (FIG. 13). Bonding wires may be used to connect multiple ICs into an array. The exterior contact substrate with mounted circuitry is then placed (FIG. 14) into a cavity 360 in the foundation 355 of an alternate form of a conventional smart card 300.

The above described process is considerably simplified by the presently described method to produce a flexible smart card with an overall thickness of less than 0.25 mm. This method is based upon a flexible IC produced by a process such as Semiconductor-on-Polymer (SOP). By virtue of its being thin and flexible, the IC 410 of FIG. 15 may be larger and therefore more capable while also being more reliable than the rigid ICs used in previous smart cards.

The flexible IC 410 does not need to be mounted on a rigid foundation. For the assembly shown in FIG. 16, a variety of flexible substrates 450 may be used, including thin PVC, PET, or even paper; that is, any flexible material that can provide suitable dielectric isolation. The substrate material may be processed in sheet or roll-to-roll form to enable large volume production at low cost. The IC 410 may be placed directly on the substrate 450 with no need for a protective cavity.

A flexible antenna 420 may be constructed without wire merely by printing it directly onto the substrate with a conductive ink, forming vias and printed contacts at the same time. The antenna substrate may be a polymer or paper and may easily be laminated onto another substrate for a specific application. Such an antenna is ultra-thin and flexible. It may be single-sided, or double-sided to accommodate printed structures and circuitry on both sides. The antenna substrate may be produced with interconnects or multilayer circuits to accommodate multiple ICs. Furthermore, additional circuitry, such as support logic and memory, may be included on the flexible smart card. For a transponder, the antenna supports both send and receive capability. Low-cost resistors and capacitors that are not available on an IC may be printed directly to the card substrate. There is no need for bonding wires since chip-to-chip interconnects are easily made by conductive inks printed directly onto the substrate and the bonding pads of all thin ICs are connected directly to the flexible substrate 450 by using conductive epoxy to printed contacts and vias. Sealing of the assembly to produce the finished flexible smart card of FIG. 17 is a simple lamination process such as that used for protecting other important papers. It is to be noted, however, that the use of SOP ICs, with their inherently protective polymer coating, allows for the transponder layer to remain as an external card layer without additional lamination.

The general case of a flexible IC 410 applied to a flexible substrate by means of an exterior contact substrate 440 is shown in FIG. 18. Here, a flexible substrate 450 is pre-patterned to provide all necessary interconnects, including vias 460. This is generally accomplished by printing with a conductive ink. The ICs used in this method are thin and their bonding pads provide an opening through the polymer of the SOP that readily exposes them for contact. When attached to a thin substrate, the product is effectively planar, enabling direct adhesion between ICs and substrate with no need for a cavity to contain and protect the ICs. The thin IC 410 is simply placed onto an exterior contact substrate 440 for attachment to a flexible substrate 450 with electrical connections being made by a conductive epoxy or similar adhesive. Vias 460 through the flexible substrate 450 enable contact to the back side of the exterior contact substrate 440. An appropriate selection of materials for the contact pads and their mating connections allows them to naturally attach to each other when placed in contact. When an SOP IC is used, its own polymer substrate may advantageously assist in adhesion to the antenna and/or card substrate. The polymer coating of the SOP also provides environmental protection for the IC, during card construction as well as in the end product.

Depending upon the application, the laminated cover of the flexible smart card may be transparent or opaque. A transparent cover enables access to light- sensitive circuitry, including optics, where such access is useful, in which applications the cover may also serve as a filter such as for color or polarization. More commonly, such a smart card will use an opaque cover printed with various logos or other identifying information. In any instance, exterior contacts may be directly written into an outer layer of the card where a contacting option is desired instead of, or in addition to, a contactless card format.

In addition to the transponder described here, the flexible smart card may be used in many other applications. The described technology is also applicable to any flexible label whether for product, packaging or personnel, as a replacement for barcodes and magnetic strips. Other applications include a variety of identification systems such as passports and driver licenses where increased "smart" capability is desired, especially for secure documents where it is desirable to have a considerable capacity for updates.

Though the above process has been described using flexible ICs and flexible substrates, there is nothing described here that precludes application of these techniques to a rigid substrate. If a rigid substrate is used, the polymer of the SOP IC could be replaced by a variety of dielectric materials.

It will be recognized by those skilled in these arts that many variations of the described embodiments are possible. Although Semiconductor-on-Polymer has been described here as a means of acquiring thin ICs, other means of producing thin ICs would be useful. Also, though silicon is the most likely substrate for flexible ICs, other single crystalline wafer materials are also feasible candidates for the IC substrate. Additional usable materials include graphene, nanotubes and non- crystalline materials. The foundation substrate may also be selected from a variety of thin and flexible materials, not to be limited by the few described here. The benefits of the described smart card transponder are derived from its thinness and flexibility which simultaneously enable low-cost production, durability and reliability. The realization of a transponder as a single card layer provides feasibility for contactless smart cards using a variety of low cost card stocks that may include paper.

Flexible Interconnects

The flexible interconnect described here enables interconnections between various combinations of semiconductor die and printed circuit boards, such as those components used to build a smart card. A basic interconnect includes a thin flexible material with at least one printed line having a connection pad at each end of the line to create a flexible interconnect. As shown here beginning in FIG. 19, the flexible interconnect 500 is made from a flexible non-conductive material such as polymer 590. The large flexible surface area material provides a structure on which various features can be printed, patterned, deposited or etched. Conductive pads 510 and metal lines 520 may be formed on or in a flexible interconnect using low cost electronic printing capability. Such features, including sub-micron and multi-layer lines, may be printed on the flexible interconnect using wafer fabrication techniques known to those skilled in such art. FIG. 20 shows a basic flexible interconnect in a flexed state.

The flexible interconnect can be attached to the assembly using materials such as conductive and non-conductive epoxies. The conductive epoxies or similarly suitable material can be applied so as to directly connect the interconnect pad to the pad of the die being contacted with the two surfaces coming into contact when the flexible interconnect is applied.

A more sophisticated interconnection includes the patterning of a via

(through-hole) completely through the pads of the flexible interconnect. In addition to the features of the basic flexible interconnect, a top view of an enhanced version of a flexible interconnect is illustrated in FIG. 21 where vias 530 have been formed. Vias extend through the thickness of the flexible material as well as the metal surface of the pad. A side view of the same interconnect appears in FIG. 22, while FIG. 23 depicts a flexed version of the same device. It is to be noted from these figures that the interconnects (510 and 520) are entirely contained within the flexible interconnect material (polymer, 590) so as to provide electrical isolation.

The flexible interconnect can be applied with the flexible interconnect pad surface on the side that is not adjacent to the die pad being contacted. To accomplish this, the flexible interconnect is adhered to the substrate with non- conductive epoxy or with an adhesive. An example of using the flexible

interconnect with vias (FIG. 23) in this manner is shown in FIG. 24. Such interconnections may be made between one semiconductor die and another, from a semiconductor die to a printed circuit board (PCB), or between one PCB and another. Here, connection is made between a flexible PCB 540 at pad 545 and a semiconductor die 550 at its pad 560 using a conductive epoxy 570.

The connection is made by printing a fill of conductive material, such as conductive epoxy, into the vias 530. The conductive material serves as a short circuit to the die pad, fills each via and overlaps the top of the flexible interconnect pad to form an electrical path from the die pad to the flexible interconnect pad. The filled vias 530 complete the electrical connection with pads 510 at the opposite side of the flexible interconnect 500. The epoxy fill of the vias maintains the thinness and flexibility of the interconnect. Depending upon the application, the materials being connected, and the relative dimensions, it may be desirable to fill the space between the flexible interconnect and the connected devices with a non-conductive epoxy 580 fill material to provide additional support.

A more complex, two-dimensional, flexible interconnect is shown in FIG. 25. This flexible interconnect 500 is used in FIG. 26 to make connections between two semiconductor die 550 and a substrate 600 such as a flexible PCB. Contact between the bonding pads 560 of the semiconductor die 550 are made by filling the vias with a printed conductive epoxy 570 that overflows onto the surface of the interconnect pad. Depending upon the dimensions, a printable conductive ink may be used in place of the epoxy. The flexible interconnect 500 conforms to the topography of the underlying devices. Though the pads 510 of the flexible interconnect 500 have been shown as being recessed from the surrounding surface, they may be fabricated so as to reach the surface. Depending upon the relative topographies of the mating surfaces, a surface-to-surface connection may be made without epoxy by using pad materials that naturally attach to each other when placed in contact. In any case, the flexible material of the pad is open to accept electrical bonding to a die pad or substrate pad. The flexible interconnect may also be applied to a die by extending, or wrapping, over the edge of the die to a substrate where it is attached using a non-conductive adhesive.

The surface area of the flexible interconnect may be large or relatively larger than the die being connected. The flexible material is large enough, and durable enough, that it can be handled during assembly without undue concern for its fragility. This accommodates ease of positioning that is independent of the die and substrate materials.

At the same time, the interconnect metal may be extremely small. A flexible direct-write printing technology is one means of producing a tightly packed interconnect. Printing with a conductive ink may be used to establish contact between two stacked material layers.

Another means of producing a tightly packed interconnect is to use a

Semiconductor-on-Polymer (SOP) technology. Such technology is capable of integrating extremely small, dense devices into the flexible interconnect. Furthermore, the SOP approach allows for integration of in-line devices such as resistors and capacitors, and even active devices. By replacing the conductive metal lines with a transparent material such as silicon, the described flexible interconnect may be adapted for use with optical components through photonic waveguides, providing for a mix of electronic and non-electronic capability.

Metal interconnects may be used to conduct heat or to form heat sinks. Similarly, flexible interconnects may be formed from material that is an electrical insulator but thermally conductive in order to transport heat away from the attached circuitry. By replacing the polymer with an insulator material that conducts heat, the flexible interconnect becomes usable as a conformal heat sink. This is in addition to the fact that unused surface area on the flexible interconnect may be layered with metal lines for the purpose of conducting heat away from the interconnected devices.

Though the above process has been described using flexible semiconductor devices and flexible substrates, there is nothing described here that precludes application of these flexible interconnect techniques to rigid components and there are other advantages to be gained in so doing. In its simplest form the flexible interconnect described here can be used as a replacement for bonding wires, especially as they can span long distances while conforming to underlying topography.

As such, multiple interconnects may be applied simultaneously, each with its own inherent insulation to protect it from the other interconnects, even when deformed. This reduces assembly time and cost while improving reliability.

Additionally, the interconnects may comprise multi-layer metal. In some

applications it will be useful that individual bonding connections may extend beyond the edge of a die or package.

On the other hand, the described flexible interconnects could be written one at a time using a material such as a conductive epoxy to trace from one pad to another on top of a flexible polymer strip that had been constructed with an array of vias, selectively addressing those contacts necessary to configure a particular circuit. It will be recognized by those skilled in these arts that many combinations and variations of the above-described devices and techniques are possible.

Claims

What is claimed is:

1. An assembly comprising:

a flexible substrate;

a flexible optical circuit on the flexible substrate for reception and/or transmission of a photonic signal,

a flexible microelectronic circuit on the flexible substrate for wireless reception and/or transmission of an electrical signal; and

at least one flexible interconnect,

wherein the at least one flexible interconnect couples the flexible optical circuit and the flexible microelectronic circuit to enable the assembly for data transfer, and

wherein the assembly is capable of sustained data transfer during flexure or other

deformation into a non-planar configuration.

2. The assembly of claim 1, wherein the flexible substrate comprises a polymer.

3. The assembly of claim 1, wherein a total thickness of the assembly does not exceed 0.25 mm.

4. The assembly of claim 1, wherein the assembly is flexible to conform to a non-planar mounting surface.

5. The assembly of claim 1, further comprising:

a control element,

wherein the flexible microelectronic circuit operates on the control element to change a behavior of the flexible optical circuit,

wherein the control element is a device for manipulating a photonic signal, such as a MEMS

(Micro-Electro-Mechanical System) mirror or a PIN diode, and

wherein the behavior is one of routing, amplification, or delay of a photonic signal.

6. A flexible photonic waveguide comprising:

a flexible substrate;

a first layer of isolation material on the substrate; a sub-micron single crystalline Semiconductor-On-Polymer (SOP) layer on the first isolation layer; and

a second layer of isolation material on the SOP layer,

wherein the SOP layer comprises photonic circuitry, and

wherein an optical mode of light entering or generated within the SOP layer is confined to the SOP layer.

7. The flexible photonic waveguide of claim 6, wherein the flexible photonic waveguide is flexible to conform to a non-planar mounting surface.

8. The flexible photonic waveguide of claim 6, wherein the photonic circuitry is functional during flexure or other deformation into a non-planar configuration.

9. The flexible photonic waveguide of claim 6, wherein the isolation material is a substance that supports confinement of the optical mode, such as silicon dioxide or silicon nitride.

10. The flexible photonic waveguide of claim 6, wherein a total thickness of the flexible photonic waveguide does not exceed 50 μιη.

11. The flexible photonic waveguide of claim 6, wherein two or more of the flexible

photonic waveguide are placed one upon another to form a stack of flexible photonic waveguides.

12. The stack of claim 11, wherein the flexible photonic waveguides are optically coupled.

13. The stack of claim 11, wherein the flexible photonic waveguides are optically isolated.

14. A method of forming a flexible photonic waveguide, comprising:

preparing a photonic circuit on a semiconductor substrate;

removing the semiconductor substrate; and

transferring the photonic circuit to a flexible support.

The method of claim 14, wherein the flexible support is a polymer.

16. A flexible transponder comprising:

a flexible substrate;

a flexible microelectronic circuit constructed on the flexible substrate,

wherein the flexible microelectronic circuit is capable of radio frequency operation; and a flexible antenna coupled to the microelectronic circuit,

wherein the flexible antenna is congruent with or conformable to the flexible substrate.

17. The flexible transponder of claim 16, wherein the flexible microelectronic circuit is produced by a Semiconductor-On-Polymer (SOP) process.

18. The flexible transponder of claim 16, wherein the flexible transponder is capable of continuous operation during flexure or other deformation into a non-planar configuration.

19. The flexible transponder of claim 16, further comprising a foundation,

wherein the flexible microelectronic circuit is mounted on the foundation, and

wherein the flexible antenna is printed on the foundation.

20. The flexible transponder of claim 16, wherein the flexible microelectronic circuit

receives power through the flexible antenna by electromagnetic induction or by electromagnetic radiation.

21. An interconnect comprising:

a flexible non-conductive material; and

a pattern of flexible conductive material on the flexible non-conductive material,

wherein the pattern includes at least two connection pads coupled by a line, and

wherein the interconnect is flexible, and

wherein a total thickness of the interconnect does not exceed 50 μιη.

22. The interconnect of claim 21, wherein the pattern is formed using a Semiconductor-on- Polymer (SOP) process.

23. The interconnect of claim 21, wherein the pattern is formed using ink.

24. The interconnect of claim 21, wherein the flexible non-conductive material is paper.

25. The interconnect of claim 21, further comprising a via (through-hole) in a connection pad.

26. The interconnect of claim 21, wherein of the flexible conductive material and the flexible non-conductive material at least one material is thermally conductive, whereby the interconnect serves as a heat sink.

27. The interconnect of claim 21,

wherein two or more of the interconnect are placed one upon another to form a multi-layer interconnect,

whereby the flexible non-conductive material of a first layer serves to insulate the pattern of flexible conductive material of the first layer from the pattern of flexible conductive material of a layer adjacent to the first layer.

28. The multi-layer interconnect of claim 27, wherein a conductive ink establishes electrical contact from a pattern in one layer to a pattern in an adjacent layer.