MicroLED Vacancy Repair Using Parallel Donor Coupon Transfer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge lies in efficiently repairing vacancies on microLED display substrates resulting from manufacturing defects and chip failures during mass transfer, where donor wafers also contain non-functional chips, making it difficult to find an optimal sequence of transfers to fill all vacancies in as few steps as possible, as the problem is NP-hard and infeasible for large substrates.
Innovation Solution
A method involving the identification of target and donor substrates with vacancies, defining a bounding box to match chip positions, and using machine learning or algorithms to simulate and determine optimal translational offsets for parallel transfer of functional chips from donor coupons to fill vacancies on the target substrate, allowing for efficient repair in minimal transfers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional sequential methods are used to repair vacancies on large substrates, then completeness of vacancy filling can be achieved, but the number of transfer steps becomes excessively large and the process becomes infeasible
Solution Approach 1:
The patent divides the large substrate into multiple smaller sub-arrays, each containing a subset of the vacancies. By segmenting the repair task, the system can process multiple sub-arrays in parallel during each transfer step, dramatically reducing the total number of sequential transfers needed while ensuring all vacancies are filled
Solution Approach 2:
The patent introduces a new dimension of parallelism by enabling simultaneous transfers to multiple different sub-arrays from a single donor array. Instead of sequentially filling one sub-array at a time, the system performs multi-target parallel transfers, effectively adding a dimensional layer to the transfer process that reduces the overall transfer step count
2Quantity of substance
If donor wafers with non-functional chips are used, then available donor substrates increase, but the difficulty of finding optimal transfer sequences increases significantly
Solution Approach 1:
The patent performs preliminary identification and mapping of functional chip locations on donor wafers before the actual transfer process. By pre-characterizing the donor substrates and creating detailed maps of functional chip positions, the system simplifies the subsequent optimization process, making it feasible to handle multiple donor wafers with non-functional chips without exponentially increasing computational complexity
Solution Approach 2:
The patent creates digital representations or models of the donor and target arrays, including the locations of functional and non-functional chips. These virtual copies allow the system to simulate and evaluate multiple transfer sequences computationally, finding optimal solutions without physically attempting every possible sequence, thus managing the complexity of matching functional chips to vacancies across multiple donor substrates
3Reliability
If multiple transfer steps are used to fill vacancies, then all vacancies can be filled, but manufacturing yield decreases and waste increases
Solution Approach 1:
The patent merges multiple transfer operations into fewer consolidated steps by enabling parallel transfers to multiple sub-arrays simultaneously. By combining what would traditionally be sequential operations into concurrent processes, the system reduces the total number of transfer steps required, thereby improving manufacturing yield and reducing waste associated with repeated handling and potential chip failures during multiple transfers
Data Source
AI summary
Data representations are formed of a target substrate and a plurality of donor coupons that are incompletely filled with functional chips. The data representations are abstracted into a current state description of the target substrate and the donor coupons and input into a machine learning model that has been trained on previous mass transfer sequences. An optimal output of the machine learning model defines at least a selected one or more of the donor coupons and corresponding functional chips of the selected one or more of the donor coupons used to fill the vacancies. A parallel transfer of the corresponding functional chips is performed to fill the vacancies on the target substrate using the selected one or more of the donor coupons.