Photolithographic process simulation including efficient result computation for multiple process variation values

• US 7,921,383 B1
• Filed: 02/12/2007
• Issued: 04/05/2011
• Est. Priority Date: 01/11/2006
• Status: Expired due to Fees

• Current Assignee: Olambda Incorporated
• Original Assignee: Olambda Incorporated
• Inventors: Wei, Haiqing
• Primary Examiner(s): WHITMORE, STACY
• Application Number:
  US11/705,808
• Time in Patent Office: 1,513 Days
• Field of Search: 703/13, 716 1- 2, 716/4, 716 19- 21
• US Class Current: 716/54
• CPC Class Codes: G03F 7/705 Modelling or simulating from physical phenomena up to complete wafer processes or whole workflow in wafer productions

Abstract

A method, system, and related computer program products and computer-readable numerical arrays for computer simulation of a photolithographic process is described. In one preferred embodiment, simulation of a photolithographic process is provided in which a computation time for computing each subsequent result for each subsequent combination of process variation values and/or target depths is significantly less than a computation time for computing an initial result for an initial combination of the process variation values and/or target depths. Accordingly, where computation for the initial combination requires a first time interval T, results for (N−1) subsequent combinations can be achieved such that a total time interval for the N results is substantially less than NT. Computation of a process model used for the computer simulation is also described, as well as calibration of the process model to a physical photolithographic processing system.

Claims

1. A method for simulating a photolithographic process, the method being implemented by a computing device comprising at least one processor, comprising:

receiving a plurality of model kernels characterizing the photolithographic process, the model kernels not being dependent on a predetermined process variation associated with the photolithographic process;

receiving a plurality of parameters dependent on said process variation; and

computing with the computing device a plurality of resultant intensities for a corresponding plurality of process variation values using said parameters, a mask transmittance function, and said plurality of model kernels.

2. The method of claim 1, said computing a plurality of resultant intensities comprising:

computing a first resultant intensity for a first process variation value by processing said mask transmittance function with said model kernels to generate a plurality of intermediate results and processing the intermediate results with said parameters, said parameters being set according to said first process variation value; and

computing a second resultant intensity for a second process variation value by processing said intermediate results with said parameters, said parameters being set according to said second process variation value.

3. The method of claim 2, wherein said processing said mask transmittance function with said model kernels has a relatively high degree of computational complexity as compared to said processing the intermediate results with said parameters, whereby said computing said second resultant intensity is achieved in substantially less time than said computing said first resultant intensity.

4. The method of claim 3, wherein said processing said mask transmittance function with said model kernels comprises convolving said mask transmittance function with said model kernels, said intermediate results being convolution results.

5. The method of claim 4, wherein said processing the intermediate results with said parameters comprises one of (i) summing the squared magnitudes of the convolution results as weighted according to said parameters, and (ii) coherently summing the convolution results by group as weighted according to a first subset of said parameters and summing the squared magnitudes of the coherent sums as weighted by a second subset of said parameters.

6. A computer readable medium tangibly embodying one or more sequences of instructions wherein execution of the one or more sequences of instructions by a computing device comprising at least one processor causes the computing device to simulate a photolithographic process according to the method of claim 1.