Patent Application
20080122119
1. Field
The present invention relates generally to manufacturing of semiconductor devices, and more particularly, to a method and apparatus for creating RFID devices using masking techniques.
2. Background
Automatic identification of products has become commonplace. For example, the ubiquitous barcode label, placed on food, clothing, and other objects, is currently the most widespread automatic identification technology that is used to provide merchants, retailers and shippers with information associated with each object or item of merchandise.
Another technology used for automatic identification products is Radio Frequency Identification (RFID). RFID uses labels or “tags” that include electronic components that respond to radio frequency commands and signals to provide identification of each tag wirelessly. Generally, RFID tags and labels comprise an integrated circuit (IC, or chip) attached to an antenna that responds to a reader using radio waves to store and access the ID information in the chip. Specifically, RFID tags and labels have a combination of antennas and analog and/or digital electronics, which often includes communications electronics, data memory, and control logic.
One of the obstacles to more widespread adoption of RFID technology is that the cost of RFID devices such as tags or labels is still relatively high as lower cost manufacturing of RFID devices has not been achievable using current production methods. Additionally, as the demand for RFID devices has increased, the pressure has increased for manufacturers to reduce the cost of the devices, as well as to reduce the size of the electronics as much as possible so as to: (1) increase the yield of the number of dies (i.e., chips) that may be produced from a semiconductor wafer, (2) reduce the potential for damage, as the final device size is smaller, and (3) increase the amount of flexibility in deployment, as the reduced amount of space needed to provide the same functionality may be used to provide more capability.
However, as the chips become smaller, the process of interconnecting them with other device components, such as antennas, becomes more difficult. Thus, to interconnect the relatively small contact pads on the chips to the antennas in RFID inlays, intermediate structures variously referred to as “strap leads,” “interposers,” and “carriers” are sometimes used to facilitate inlay manufacture. The intermediate structure include conductive leads or pads that are electrically coupled to the contact pads of the chips for coupling the chips to the antennas. These leads provide a larger effective electrical contact area between the chips and the antenna than do the contact pads of the chip alone. With the use of intermediate structure, the alignment between the chip and the antenna does not have to be as precise during the direct placement of the chip on the antenna as without the use of such intermediate structure. The larger contact area provided by the intermediate structures reduces the accuracy required for placement of the chips during manufacture while still providing effective electrical connection between the chip and the antenna. However, the accurate placement and mounting of the chips on the intermediate structures still provide serious obstacles for high-speed manufacturing of RFID tags and labels.
Some challenges that currently face manufacturers or suppliers to component manufacturers include:
1) Wafer Processing: Transfer of chips from a wafer to a suitable substrate.
2) Chip Attachment: Accurately positioning of chips for attachment to intermediate structures is difficult to achieve at the speeds needed to achieve the economies of scale obtainable through high volume manufacturing.
3) Bonding: It is difficult to accurately bond, cure, and electrically connect the chips to intermediate structures at rates necessary to achieve high volume manufacturing.
Several possible high-speed intermediate structure assembly strategies have been proposed. The first approach, which uses the “pick-and-place” machines typically deployed in the manufacturing of circuit boards for picking up electronic components and placing them on circuit boards, is accurate, but requires expensive machines that ultimately do not deliver a sufficient throughput to justify the increased cost. That is, pick-and-place equipment may only be able to achieve 20-25,000 units per hour (UPH) whereas 100,000 UPH or more is needed for true high speed manufacturing. However, utilizing multiple pick-and-place machines in a line significantly increases the complexity of the manufacturing process and the possibility of error.
Another approach, referred to as a “self-assembly process,” is a method in which multiple chips are first dispersed in a liquid slurry, shaken and assembled into a substrate containing chip receiving recesses. Some current processes are described in U.S. Pat. No. 6,848,162, entitled “Method and Apparatus for High Volume Assembly of Radio Frequency Identification Tags,” issued to Arneson, et al. on Feb. 1, 2005; U.S. Pat. No. 6,566,744, entitled “Integrated Circuit Packages Assembled Utilizing Fluidic Self-Assembly,” issued to Gengel on May 20, 2003; and, U.S. Pat. No. 6,527,964, entitled “Methods and Apparatuses for Improved Flow in Performing Fluidic Self Assembly,” issued to Smith et al. on Mar. 4, 2003. Publications, patents and patent applications are referred to throughout this disclosure. All references cited herein are hereby incorporated by reference.
Accordingly, there is a long-felt, but as yet unsatisfied need in the RFID device manufacturing field to be able to produce RFID devices in high volume, and to assemble them at much higher speed per unit cost than is possible using current manufacturing processes.
The present invention for creating RFID devices provides high yield of transfer and bonding of RFID chips from chip carriers to device matrices with a near infrared (NIR) bonding process. The chips are directly bonded to the strap leads or antenna leads from the chip carriers. In one preferred embodiment, the process includes the steps of selectively removing and transferring of RFID chips to the chip carrier with predetermined pitches; and bonding the chip to strap leads or antenna leads using the NIR bonding process.
In one preferred embodiment, the techniques that are used to implement the present invention include:
1. Selective removal and transfer of chips to a chip carrier where the chips have a pitch that substantially matches the pitch of the strap leads or the antenna leads.
Using the selective chip transfer techniques as described herein, chips can be selectively removed and transferred from a wafer or a chip carrier to another chip carrier to result in a chip carrier with chips spaced at a different pitch. There is no size limitation of chip carrier in these selective chip peeling and transfer processes. UV-reactive adhesive tapes and functional adhesive carriers, as described herein, can be effectively used in these selective transfer processes.
2. Selective removal and transfer of chips to a chip carrier where the chip position is not changed during the peeling and transfer process.
Since there is no tape expansion or extension during the selective chip removal and transfer processes, the original position of each chip is not changed through these removal and transfer processes.
3. Bonding of the chips to the strap leads or the antenna leads through a NIR bonding process:
The chip can be bonded directly to the strap lead carrier or the antenna lead carrier without limitation on the size of the chip carrier. No other process is needed between the selective removal and transfer process and the chip bonding process.
Other features and advantages of the present invention will become apparent to those skilled in the art from the following detailed description. It is to be understood, however, that the detailed description of the various embodiments and specific examples, while indicating preferred and other embodiments of the present invention, are given by way of illustration and not limitation. Many changes and modifications within the scope of the present invention may be made without departing from the spirit thereof, and the invention includes all such modifications.
The invention may be more readily understood by referring to the accompanying drawings in which:
Like numerals refer to like parts throughout the several views of the drawings.
As used herein, the term “quasi-wafer” refers to a subset of chips transferred from an original wafer, wherein the position of each chip on the quasi-wafer relative to the other chips in the original wafer does not change. In one preferred embodiment, a quasi-wafer is created by removing every other row of chips and then every other column of chips from an original wafer, resulting in at least one quasi-wafer where each chip on the quasi-wafer is without any adjacent chips. In this preferred embodiment, four quasi-wafers may be created from an original wafer. In other preferred embodiments, fewer or more chips may be removed so that strap leads, which are created on a substrate with a pitch matched to the pitch of the chips in the quasi-wafer, may be attached to the chips. Thus, the term quasi-wafer may refer to any subset of chips removed from the original wafer, but where each chip maintains its original position.
With reference to
In one preferred embodiment, the different levels of adhesion are created using a radiation source, such as an Ultraviolet (UV), Infrared (IR), or Near IR (NIR) light source in conjunction with shadow masks to create a pattern of exposed and covered areas on a carrier (or substrate) coated with an adhesive sensitive to the radiation source. In this embodiment, the adhesiveness in the areas exposed to the radiation source are deactivated (or activated), resulting in lower (or higher) adhesion. Coherent light sources (such as lasers) can also be used as the radiation source without using any masks. In another preferred embodiment, the patterning of the substrates is done by the selective dispensing or printing of adhesives or other materials in order to selectively increase/decrease the adhesion in desired locations. For example, a tape patterned with UV-reactive adhesives may be used to selectively remove any number of chips from a die matrix. In this example, an additional activation or deactivation step may be necessary, which can be achieved by using UV radiation. Other types of radiation such as the aforementioned IR or NIR radiation may be used, based on the specific type of adhesives used.
In one preferred embodiment of the quasi-wafer creation process, the transfer is a two-dimensional to two-dimensional transfer that involves selective chip transfer from one wafer into multiple quasi-wafers in two steps. The first step involves transferring the entire wafer (e.g., plurality of chips 206) to a new support (e.g., tape 202) having an UV-reactive adhesive 204. In one preferred embodiment, a tape having UV-reactive adhesives coated or dispensed thereon, such as the 1027R tape from Ultron System of Moorpark, Calif., as currently available from MINITRON Elektronik GmbH, may be used. It is desirable that the tapes used have a stable, substantially unstretchable substrate such as polyethylene terephthalate (PET). It is also desirable to use frames to mount each tape. The frame may be rigid in shape but can be bent if necessary.
The second step involves the attachment of a second tape having UV-reactive adhesives coated or dispensed thereon to the chips transferred to the first tape. Thus, referencing
In one preferred embodiment, first tape 202 and second tape 212 are then pulled apart in a peel angle in the range of between about 40-50°, with an angle of approximately 45° being preferred, relative to the chip edges, to separate the wafer into alternate rows of chips on each of the supports. When separating the tapes, a peel front is generated. This peel front may impinge on a chip either orthogonally or non-orthogonally. As defined herein, orthogonal refers to the fact that the peel front impinges on an entire first edge of a chip simultaneously. Preferably, the peel angle would be non-orthogonal but would involve symmetrically separating the support tapes. In one preferred embodiment, the non-orthogonal angle would be approximately 45° relative to a first edge of a chip.
Once a first subset of the chips of the original wafer has been created, it is covered with another support, again sandwiching the chips between the supports. The chips are now masked in alternate columns in the same fashion as when masked row-wise. Now after UV irradiation, when first tape 202 and second tape 212 are again pulled apart at an approximate 45° angle, the respective set of chips (first portion of chips 502 and second portion of chips 512), are arranged on the support in a checkerboard or alternating chip/space pattern. Other patterns are possible, such as two chips/two spaces, one chip/two spaces and so on.
In another approach, instead of transferring the complete wafer of chips to a new support, only predetermined portions of plurality of chips 206 are transferred to new supports directly from wafer tape 202. Thus, for example, every other chip in each row and column can be removed.
In one preferred embodiment, adhesion patterned through the use of UV-reactive adhesive is used because: 1) UV detack tape material is readily available; 2) UV methods are optical methods, so it is possible to achieve resolution to a micrometer level (much smaller than the RFID or diode die size); and 3) it is also very easy to design a system to use UV to selectively pattern a surface, such as a scan laser system, a projection system, or such systems that are commonly used for lithography applications.
Other selective transfer methods that may be used in addition to or in place of the adhesion patterning method as described include: vacuum picking; pick-and-place; and freezing. For example, selective picking may be used to achieve a pitch appropriate for strap lead or antenna lead placement. Selective picking may also be used first to increase the space between the chips (e.g., the space between each of the chips can be increased by 100%) and then freeze/peel/free and selective picking/dispensing may then be used to ultimately achieve the pitch appropriate for attaching to strap leads or antenna leads.
Referring to
As discussed above, once the quasi-wafer creation step 102 has been performed, in one preferred embodiment four quasi-wafers will have been created. Two of these quasi-wafers have the chips in a “pads up” orientation and two have the chips in a “pads down” orientation. In one preferred embodiment, an optional orientation step 104 is used to reorient the chips on one set of quasi-wafers to have a desired pad orientation, either contact pads up or contact pads down. Thus, after the execution of chip orientation step 104, all sets of quasi-wafers would have the same pad orientation.
After step 104 has been completed, the chips on the quasi-wafers are being carried on tape and held in place by a functional adhesive (e.g., UV-reactive adhesive in this embodiment). In step 106, the chips are transferred to another substrate and held in place with a temperature-sensitive adhesive (TSA). In one preferred embodiment, the TSA is made by mixing a crystallized additive to a solvent based pressure sensitive adhesive (PSA). The PSA is available, for example, from KIWO, Inc. of Seabrook, Tex. The crystallized additive of turns the PSA into a temperature sensitive adhesive (TSA). Different weights of this additive result in different temperature functionalities of the TSA. In a preferred embodiment, a mixture of 33% of the additive is added to the aforementioned solvent-based PSA.
In one preferred embodiment, the plurality of chips on a quasi-wafer is transferred to a suitable carrier by sandwiching the chips between the UV-reactive carrier and a new carrier coated with TSA. Irradiation with UV light detackifies the UV-reactive adhesive and accomplishes the transfer upon separation of the two carriers. Similar to the separation of the tapes during the quasi-wafer manufacturing process, the separation is preferably performed at a 45° angle relative to the leading edge of the plurality of chips.
As noted above in step 104, half of the quasi-wafers produced have chips held in an orientation that is opposite of what is desired (i.e., either in a pads up or pads down orientation). These chips require one additional transfer in order to obtain quasi-wafers with pads all in the same orientation. This may be done prior to or after the transfer to a TSA coated support. For example, the chips can be transferred to a first TSA coated support, then transferred to a second TSA coated support. In one preferred embodiment, the first and second TSA coated supports have different formulations of TSA to allow the transfer of the chips.
As described herein, functional adhesives such as a TSA or UV-reactive adhesives can be used in the various embodiments. However, it should be noted that other functional adhesives can also be used. The adhesive properties of the functional adhesive may be changed (e.g., increased, reduced or eliminated) by exposure to a predetermined range of stimuli. The stimuli may include UV radiation, heat, light or other stimuli that effect changes in the adhesive properties. The change of the adhesive properties may be temporary or permanent. Further, the change in adhesive properties may be gradual or sudden. For example, the functional adhesive may be a TSA that loses its adhesiveness at higher temperatures, but once the temperature falls below a predetermined temperature, the adhesiveness of the TSA returns.
As discussed, there are many different methods to transfer and bond chips to strap leads and antenna leads. However, the process of transferring and bonding the chips from the die matrix or silicon wafer to strap leads and antenna leads are composed of many different process steps, such as: the expansion of the matrix of silicon wafer, the transfer of the chips from one carrier to another (perhaps multiple times), and using special instruments to pick up and bond the chips to strap leads or antenna leads. An increase in the number of these process steps both limits the production yield and increases the risk of chip contamination.
To reduce the number of steps, it is proposed herein that a process whereby selective chip transferring and NIR bonding steps are used to directly bond chips, contained on a chip carrier, to strap leads or antenna leads. As used herein, the term “strap” refers to an assembly where a chip is already bonded to a strap lead or lead frame, whereas the terms “strap lead” or “lead frame” refers to a chip to antenna connection structure. The bonding process can be performed for a plurality of chips and, thus, a large number of chips can be bonded onto strap leads or antenna leads through just one bonding operation. The chip carrier is, in one preferred embodiment, a carrier that is coated with adhesives. Functional adhesives are used on the substrates supporting the chips during the transferring and bonding of the chips. The advantages of using functional adhesives include the fact that the adhesives can be screen printed or printed with other printing methods and the adhesive printed on the carrier can be recyclably used for many cycles. Use of the reusable adhesives provides many ways to simplify and speed up the chip transfer process.
In one preferred embodiment, functional adhesives are recyclably used for many times in chip transfer and bonding. The chips are transferred from one substrate to the functional adhesive carrier at ambient conditions and then transferred again or bonded from the functional adhesive carrier to another substrate or device at a predetermined temperature, i.e., at a temperature that causes the functional adhesive to release the chips. The utilization of recyclable functional adhesives can simplify the chip transfer process, increase chip transfer and bonding yields, and reduce production costs.
1. Exemplary use of functional adhesives.
2. In one preferred embodiment, the functional adhesive is comprised of printable liquid materials and dried as a rubbery adhesive material for a surface. Examples include hot melt or solvent-based adhesives.
3. Examples of applications of the functional adhesive during the chip transfer and bonding process.
In one preferred embodiment, adhesive dots of approximately 1 mm in diameter are screen printed onto a PET substrate having a thickness of approximately 7 mils by syringe. The adhesive dots are dried at approximately 70° C. for approximately 10 minutes. These adhesive dots are able to peel the chip away from the UV detacked tape at room temperature. The PET substrate with the chips attached by adhesive dots is heated to 60° C., then a tape such as Ultron System No. 1020° R tape, which is a UV detackable adhesive coated on polyvinyl chloride (PVC), is pressed against and contacted with the chips and the PET substrate. The tape is peeled away from the PET substrate, with the chips being transferred onto 1020R tape.
In a first preferred embodiment for transferring chips to a TSA carrier, as illustrated in
If the adhesive pattern or dot size is smaller than the chip area, a little stretch of the wafer backing tape may be used to separate chips. Thus, there may not be a need to selectively transfer the wafer before this step. An alignment system may then be used to align a wafer tape 812 having a plurality of diced chips 806 (stretched or selectively transferred or in the original form) to the substrate of patterned adhesive. In one preferred embodiment, wafer tape 812 is an UV backing tape that is detackified by UV radiation. In one preferred embodiment, plurality of chips 806 have their pads facing toward wafer tape 812. The plurality of chips 806 is then aligned to adhesive dots 850 or patterns and the gap between the two surfaces (wafer tape 812 and roll substrate 802) are closed to get certain chips to contact the adhesive printed areas. Once the gap between those two surfaces is opened, the chips in contact with the adhesive dots or patterns will separate from the wafer and stay on the adhesive dots.
In another preferred embodiment, as illustrated in
In yet another preferred embodiment, as illustrated in
Once the step of transferring the chips to the TSA carrier has been completed, process 100 continues with step 108.
Referring to
In one preferred embodiment, NIR thermocompression unit 1280 has a web handling system capable of positioning two webs within a precision of better than +/−30 micrometers (•m). The carrier 802 is laminated to the ACP dispensed final substrate 1202 (e.g., strap lead web) before entering into the NIR thermocompression unit 1280, in which the chips 806 are bonded to the strap leads by the ACP 1204 being cured under pressure. In case the functional adhesive used is a TSA, it is detacked under heat and the carrier substrate (e.g., carrier 802) can be easily removed from the strap lead web (e.g., final substrate 1202) at the exit from the NIR thermocompression unit 1280. In cases where other functional adhesives are used (e.g., UV detackable adhesives), a UV radiation unit may be used inline after the NIR thermocompression unit 1280 to detack the adhesive and peel off the carrier substrate.
As shown in the figure, a subset of NIR rays 1310a of the plurality of NIR rays is allowed through pair of masks 1370, 1372 to heat up the subset of chips 1306a. Along with the pressure applied, the heat cures the plurality of dispensed ACP droplets 1350. In another preferred embodiment, the plurality of dispensed ACP droplets 1350 are heated and cured by its exposure to the subset of NIR rays 1310a. A subset of NIR rays 1310b are reflected by the pair of masks 1370, 1372 and does not reach the subset of chips 1306b. In one preferred embodiment, both subset of chips 1306a and subset of chips 1306b are affixed to the chip carrier 1302 with a TSA and the heat generated by the subset of NIR rays 1310a reduces the adhesiveness of the TSA under subset of chips 1306a. The reduction of adhesiveness of the TSA facilitates the removal of subset of chips 1306a from chip carrier 1302 when the subset of chips 1306 are bonded to the strap lead carrier 1312.
The pair of masks 1370, 1372 allows selective transfer and bonding of the plurality of chips because the pair of masks 1370, 1372 is used to selectively allow NIR/IR to reach the chips to be transferred. In one preferred embodiment, pair of masks 1370, 1372 are aluminum and have polished surfaces so that NIR radiation is reflected therefrom and does not heat up. Other reflective materials may also be used, the selection and configuration of which depending on the type of radiation to be reflected. In another preferred embodiment, an insulating material can be used to protect the TSA attaching the chips on the chip carrier from temperature changes. In yet another preferred embodiment, instead of protecting predetermined locations on a chip carrier from being exposed to temperature changes, radiation absorption elements or temperature changing elements (e.g., heated grid) may be used to localize the temperature changes needed to reduce the adhesiveness of the TSA and/or cure the adhesives attaching the chips to the strap leads. Further, targeted coherent light such as lasers may also be used to heat specific locations.
It should be noted that although quasi-wafers have been used to describe certain concepts of chip transfer and attachment contained herein, arrangements of chips other than in quasi-wafers can also be used. For example, in the chip transfer and bonding approach shown and described with respect to
The embodiments described above are exemplary embodiments of the present invention. Those skilled in the art may now make numerous uses of, and departures from, the above-described embodiments without departing from the inventive concepts disclosed herein. Accordingly, the present invention is to be defined solely by the scope of the following claims.
