The present disclosure relates to semiconductor devices and methods thereof. In particular, the present disclosure relates to wafer lamination systems and methods thereof for processing semiconductor wafers.
In semiconductor processing, devices are formed in parallel on a wafer. A singulation process is performed to separate the wafer into individual devices. Typically, a mechanical saw is employed to separate the wafer into individual dies. To increase performance, high-k dielectric materials are used in the formation of interconnects as part of the back end (BE) dielectric stack. However, the high-k BE dielectric stack is prone to damage due to the vibration of mechanical sawing
To avoid damaging the high-k BE dielectric, plasma dicing is employed instead of mechanical sawing to singulate the wafer. Plasma dicing includes laminating a wafer on a dicing tape attached to a wafer ring. The wafer ring with the wafer (wafer ring assembly) is positioned in a plasma dicing chamber to singulate the wafer.
To laminate a wafer for plasma dicing, two separate pieces of equipment or tools are needed. For example, to expedite the plasma dicing process, the wafer is thinned by back grinding (BG) the back (inactive) side of the wafer to reduce plasma dicing time. This means that one tool is needed to remove the BG tape and one tool is needed to laminate the dicing tape onto the wafer. Two separate tools make the plasma dicing of wafers more complex and time-consuming, increasing the overall production cost. In addition, two separate tools require more space, further adding to the overall production cost.
From the foregoing discussion, there is a desire to provide a cost-effective way to laminate wafers for plasma dicing.
Embodiments generally relate to lamination systems and methods for fabricating devices.
One embodiment relates to a lamination system for manufacturing semiconductor devices. The system includes an input sub-system, a processing sub-system and an output sub-system.
The input sub-system is configured with a robotic arm of a robotic module for picking up a wafer. The wafer includes first and second wafer surfaces. A backgrinding (BG) tape is attached to the first wafer surface. For example, the second wafer surface is a ground surface. The wafer may be thinned by grinding the second wafer surface until a desired thickness is reached. A wafer aligner module aligns the wafer on the robotic arm. In one embodiment, the robotic arm attaches to the BG tape on the wafer. For example, an end effector of the robotic arm attaches to the BG tape on the first wafer surface. The aligned wafer on the robotic arm is transferred to the processing sub-system.
The processing sub-system includes a wafer receiving module (WRM) and various processing modules configured to perform delamination of the BG tape from the first wafer surface and laminating a dicing tape after the BG has been delaminated. The dicing tape laminates the wafer to a wafer ring to form a wafer ring assembly. The WRM includes a positioning chuck. The robotic arm is configured to place the wafer aligned on thereon onto the positioning chuck. A second surface of the wafer is attached to and held into position by the position chuck, leaving the BG tape exposed. The WRM transfers the wafer to the various modules for processing to form the wafer ring assembly.
As for the output sub-system, it includes an output module for an output module. The output module is configured to pick up the wafer ring assembly from the WRM and outputs it from the lamination system. The wafer ring assembly may be further processed to dice the wafer thereon, separating it into individual dies or devices.
Another embodiment relates to a method of fabricating devices. The method includes providing a wafer having first and second major surfaces. The first major surface includes a backgrinding tape (BG) tape laminated thereto. A robotic module picks up the wafer and is aligned by a wafer aligner module. The robotic module transfers the aligned wafer to a wafer receiving module (WRM). The robotic module places the wafer on a positioning chuck of the WRM. The second major surface of the wafer is held by the positioning chuck.
The WRM transfers the wafer to various processing modules of a processing sub-system to delaminate the BG tape from the first wafer surface and to laminate a dicing tape on the first wafer surface. The dicing tape is laminated to a wafer ring to form a wafer ring assembly. The wafer ring assembly is transferred to the output module by the WRM. The output module outputs the wafer ring assembly for subsequent processing, including wafer singulation to separate the wafer into individual devices.
These and other advantages and features of the embodiments herein disclosed will become apparent through reference to the following description and the accompanying drawings. Furthermore, it is to be understood that the features of the various embodiments described herein are not mutually exclusive and can exist in various combinations and permutations.
In the drawings, like reference characters generally refer to the same parts throughout the different views. Also, the drawings are not necessarily to scale, emphasis instead generally being placed upon illustrating the principles of various embodiments. In the following description, various embodiments of the present disclosure are described with reference to the following, in which:
Embodiments described herein generally relate to semiconductor processing for producing devices. The embodiments, in particular, relate to cost-effectively laminating wafers for plasma dicing to form devices.
Referring to
As shown, the lamination system includes an input sub-system 111, a processing sub-system 141 and an output sub-system 191. The different sub-systems are mounted on a rectangular-shaped system platform or body 101. Other shaped system platforms may also be useful. In one embodiment, the input sub-system and the output sub-system are disposed on opposing sides of the processing sub-system along the y-direction. For example, incoming wafers are provided at the input sub-system, processed by the processing sub-system and output by the output sub-system. Other configurations of the sub-systems which efficiently process the incoming wafers with minimal system footprint may also be useful.
In one embodiment, the input sub-system includes a robotic module 110, a wafer input module 120 and a wafer aligner module 130. The processing sub-system includes a wafer receiving module (WRM) 140 and a WRM track 104 for moving the WRM within the lamination system. On the first opposing side of the WRM along the x direction is disposed a back grinding (BG) tape delamination module 150; on a second opposing side of the WRM along the x direction is disposed a dicing tape lamination module 170. A wafer ring pick and place module 180 is disposed proximate to the output sub-system. In one embodiment, the wafer ring pick and place module includes a pick and place arm 181 and a wafer ring track 187. The various processing modules may be supported on the system platform by a system frame 102. The WRM track is disposed on the system body and/or frame to facilitate the movement of the WRM to the different processing modules. As for the output module 190, as already discussed, it is disposed on an opposing side of the processing sub-system as the input sub-system. Other layout configurations may also be useful. The layout should promote efficient processing flow for delamination and lamination of a wafer ring.
As discussed, the input sub-system includes the wafer input module, the robotic module and the wafer aligner module. The input module is configured to receive a cassette of incoming wafers. An incoming wafer includes a BG tape on the front (active) side of the wafer. For example, the back (inactive) wafer side has been thinned to the desired thickness by back grinding and polishing. The input module is configured to feed incoming wafers for the robotic module. The incoming wafers are preferably configured such that the inactive surface is facing upwards. For example, the polished side of the wafer is facing upwards. This avoids the polished surface of the wafer from contacting any other surface to maintain cleanliness after polishing.
The robotic module is configured to pick up an incoming wafer from the input module and transfer it to the wafer aligner module. In one embodiment, the robotic module includes a multi-axis or degrees of freedom robotic arm. The robotic arm includes end effectors for grasping an incoming wafer from the input module using, for example, vacuum pressure. In one embodiment, the end effectors contact the BG tape on the wafer. For example, the end effectors grasp the tape side of the wafer. Other techniques or configurations for grasping the wafer may also be useful.
In one embodiment, prior to transferring it to the aligner module, the robotic module flips the incoming wafer 180° so that the BG tape is facing upwards. The wafer aligner module aligns the wafer on the robotic arm. The robot arm can be configured to rotate and laterally shift x-y directions during the alignment process. In one embodiment, the aligner module aligns the wafer using the wafer notch as a reference for wafer orientation and the edge of the wafer as a reference for alignment and centering. Other alignment techniques or configurations may also be useful. For example, in some embodiments,
After alignment, the robotic module transfers the aligned wafer to the WRM of the processing sub-system for processing. The robotic module, in one embodiment, is disposed on an x-y robot track system to facilitate movement to the input module, the wafer aligner module and the WRM module of the processing sub-system.
Regarding the WRM module, it receives the incoming wafer from the robotic module with the BG tape facing upwards. The WRM module includes a WRM table with a positioning chuck to hold the wafer in position, the polished surface is facing the positioning chuck. For example, the chuck may hold the wafer in position by vacuum pressure. In addition, the WRM may include alignment pins for aligning a wafer ring to the WRM in a subsequent process. Since the aligner module has already aligned the wafer to the robotic arm, the wafer is also aligned to the positioning chuck of the WRM. Once positioned onto the chuck, the wafer remains thereon until processing is completed. This advantageously avoids the need for realignment, simplifying the overall wafer lamination process.
The positioning chuck, for example, holds the wafer in position by vacuum pressure on the polished surface. The positioning chuck may be under the wafer. For example, the positioning chuck is located under the polished side of the wafer. In other embodiments, the positioning chuck can be configured to be on top of the wafer when the BG tape is facing down. This ensures that the polished side of the wafer is attached to the positioning chuck. The wafer can be flipped either before or after alignment. Preferably, flipping of the wafer is performed before alignment to prevent any change in alignment during flipping.
The WRM module proceeds to the delamination module for processing. In one embodiment, the delamination module includes a curing unit 155 and a tape removal unit 150. The curing unit cures the BG tape using UV radiation before it reaches the tape removal unit. The curing reduces the adhesivity of the BG tape, facilitating its removal. In one embodiment, the WRM moves through the curing unit as the BG tape is cured on its way to the tape removal unit. Preferably, sticky assist tape is applied to one edge of the BG tape, the sticky assist tape is pulled to peel the BG tape from the wafer surface. After peeling, the BG tape and the sticky assist tape are disposed of. For example, the peeled BG tape is placed in a disposal bin for peeled BG tapes.
After BG tape removal, the WRM moves to the wafer ring pick and place module. The ring pick and place module includes a pick and place arm which picks up a wafer ring, positions the ring and places it onto the WRM table with the delaminated wafer. The wafer ring is positioned using, for example, position pins on the WRM table. The pick and place module, for example, receives a wafer ring from a wafer ring repository, which places it on a ring track for pick up by the pick and place arm.
WRM with the ring and delaminated wafer proceeds to the lamination module for processing. The lamination module includes a cleaning unit 160 and a tape lamination unit 170. The cleaning module cleans the surface of the wafer. For example, the surface from which the BG tape has been removed is cleaned. In one embodiment, the cleaning module includes an ionizer and air cleaning unit. The cleaning module, for example, employs air and suction force to remove particles from the surface of the wafer. The ionizer is configured to remove static. In one embodiment, the WRM moves through the cleaning unit as it cleans.
The cleaned wafer progresses to the lamination unit. The lamination unit is configured to laminate the dicing tape on the wafer surface. For example, the dicing tape is laminated on the cleaned surface. In one embodiment, the dicing tape is laminated on the same surface from which the BG tape was removed. Since the inactive surface is maintained on the chuck, this minimizes the inactive surface from particle contamination by minimizing its exposure or handline.
The lamination unit includes a tape supply which dispenses the dicing tape from lamination over the wafer and wafer ring. In one embodiment, the dicing tape is sized appropriately to be about the same size or slightly smaller than the wafer ring. This advantageously avoids the need for cutting or trimming the dicing tape. For example, as shown in
The WRM moves the wafer ring assembly to the output module. As shown, the WRM track goes through the wafer pick and place module to the output module. Other configurations of the output module may also be useful. The output module is configured to output the wafer ring assembly from the lamination system, completing the lamination process.
In one embodiment, the output module includes an output robot for picking up the wafer ring assembly from the WRM. For example, the robot grasps the wafer ring assembly by the edge thereof, upon which the WRM chuck releases the wafer. This enables the wafer ring assembly to be moved by the robot. The robot transfers the wafer ring assembly to, for example, a conveyor which feeds a wafer ring assembly repository. The wafer ring assemblies, for example, a subsequently processed to singulate the wafer into individual devices by plasma dicing.
In one embodiment, a third arm member 676 is connected to the first wrist joint for rotating it about a fourth axis of rotation A4. An opposing end of the third arm member includes a second wrist joint 689. The second wrist joint is connected to a fourth arm member 686. The wrist joint rotates the fourth arm member about a fifth axis of rotation A5. An end 696 of the arm formed by an end of the fourth arm member is a third wrist joint. In one embodiment, an end effector (not shown) is coupled to the third wrist joint, rotating the end effector about a sixth axis of rotation A6. For example, this enables flipping of the wafer held by the end effector.
The end effector is configured to grasp an incoming wafer from the input module. For example, the robotic arm can be rotated and extended to pick up the incoming wafer from the input module. The robotic arm can then be retracted, rotated and extended to the wafer aligner module for alignment. After alignment, the robotic arm may be retracted, configured to rotate the end effector to flip the wafer and extended to the WRM of the processing sub-system. In some embodiments, the base of the robotic module may be mounted onto an x-y translatable platform to facilitate movement among the modules, such as the input module, wafer aligner module and the WRM.
At 727, the incoming wafer on the robotic arm is aligned by the wafer aligner module. For example, the robotic arm with the incoming wafer is moved to the wafer aligner. At the wafer aligner module, a sensor determines the location of the notch on the wafer for alignment. The robot arm adjusts the position of the wafer based on the location of the notch. This is done while the wafer is on the end effector. After the position of the notch is adjusted, the robot arm rotates the wafer 360 degrees. If the notch is misaligned after the rotation, the robot arm adjusts the position of the wafer to center the wafer. The process may be repeated until the wafer is aligned and centered. In one embodiment, the wafer notch is used to orient the wafer and the wafer edge is used for positioning, such as aligning and centering.
The robotic module moves to the processing module at 737. The robotic module moves between the modules via the robotic track. In one embodiment, the robotic module positions the incoming wafer on the chuck of the wafer receiving module (WRM) of the processing sub-system. The BG tape is facing up on the chuck.
The WRM, at 747 transports the wafer to the delamination module for removal of the BG tape. In one embodiment, the BG tape is cured by the curing unit along the way to the tape removal unit. For example, the WRM travels along the WRM track through the curing unit without stopping to cure the BG tape. When the WRM reaches the tape removal unit, it stops for the tape removal unit to remove the BG tape.
After tape removal, the WRM moves to the wafer ring pick and place module at 757. At the wafer ring pick and place module, a pick arm picks a wafer ring from the wafer ring repository and places on the chuck table. Subsequently, the WRM with the wafer and wafer ring moves to the lamination module at 767. In one embodiment, the wafer and wafer ring are cleaned by the cleaning unit. In one embodiment, the WRM moves through the cleaning unit, similar to the curing unit, at stops at the lamination unit. At the lamination unit, a dicing tape is laminated onto the wafer ring and inactive wafer surface, producing a wafer ring assembly.
After lamination, at 777, the WRM moves to the output sub-system. The output module then removes the wafer ring assembly from the WRM. The wafer ring assembly is loaded onto a conveyor of the output module. The conveyor feeds the wafer ring assembly to a wafer ring repository. When the wafer ring repository is full, it is removed. For example, the wafer ring repository is transported for subsequent processing by a plasma dicing chamber to singulate the wafers into individual dies.
As described, efficient and cost-effective dicing tape lamination is achieved using a single tool with a single chuck.
The inventive concept of the present disclosure may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The foregoing embodiments, therefore, are to be considered in all respects illustrative rather than limiting the invention described herein. The scope of the invention is thus indicated by the appended claims, rather than by the foregoing description, and all changes that come within the meaning and range of equivalency of the claims are intended to be embraced therein.
This application claims the benefit of U.S. Provisional Application No. 63/434,458, filed on Dec. 21, 2022, which is incorporated herein by reference in its entirety for all purposes.
Number | Date | Country | |
---|---|---|---|
63434458 | Dec 2022 | US |