1. Field of the Invention
The present invention relates to a semiconductor device fabrication process and, more particularly, to characterization of a photoresist process in a semiconductor coating process.
2. State of the Art
Semiconductor processing for forming integrated circuits requires a series of processing steps. These processing steps include the deposition and patterning of a variety of material layers. The material layers are typically patterned using a photolithographic process, which uses a patterned photoresist layer as an etch mask that is patterned over the material layer. The photoresist layer is formed by first depositing liquid photoresist onto the semiconductor wafer and then spin-coating the wafer to the desired thickness. The photoresist is dried or baked and subjected to light through a photomask or reticle, and then developed to form a photoresist etch mask.
As integrated circuit dimensions decrease, the uniformity of semiconductor processes becomes increasingly important. Photolithography processing equipment is used for various types of semiconductor wafers and processes are set up and taken down as semiconductor equipment is reused over various processes and for various specifications. Photolithography process set up currently is a tedious, time-consuming chore. The photoresist pump must be primed, a wafer must be coated with photoresist, and then the coated wafer must be baked. The wafer coating thickness is then measured at a random sampling of points across the wafer. Known measuring equipment requires a significant amount of time to measure each point.
Defective coatings may be identified when the average coating thickness measurement is beyond the range of process specifications, or when the standard deviation of thickness measurements around the wafer is larger than a specific tolerance. Once a process parameter is found to be outside of the process specifications, the coating process must be adjusted, another wafer must be coated and baked, and the coating must be manually rechecked until the photoresist thickness is within the process specifications. As a result, a substantial delay often occurs before production processing may begin.
Optimization of photoresist processes has conventionally been time-consuming and conducted on an ad hoc basis. A series of test wafers is coated at various spin rates and for various times. This series of test wafers is then measured and processes are adjusted accordingly. A series of spin curves is generated based on the spin rate vs. the thickness information. The operator of the process then makes several adjustments to obtain the best possible uniformity for the target thickness. Such a trial and error approach requires the running of several wafers and such processing can take anywhere from 1-6 hours per thickness and still not guarantee an optimal setup. For example, the best possible uniformity for a given photoresist thickness when the wafer is spun out for 5 seconds may be a variation of 25 Angstroms, but the optimal uniformity for the same thickness might be achieved at 4.2 second with a slightly lower spin time and yield a variation of 15 Angstroms.
Conventionally, each of the data points on a spin curve is derived from a separate wafer and then recorded for future reference. Due to inherent processing variations, when a subsequent process is set up, a spin curve is referenced for the best possible candidate and then a process wafer is run to identify small operator adjustments. It would be advantageous to obtain additional data points across a spin curve without processing specific wafers for each data point.
An in situ photoresist thickness characterization process and apparatus is provided. In one embodiment, a method is provided for characterizing a photoresist process used for processing a semiconductor wafer. Photoresist is dispensed on a spinning semiconductor wafer as part of the characterization process. The thickness of the photoresist is monitored at a plurality of locations on the spinning semiconductor wafer and at specific time intervals while the photoresist flows across the wafer. The thicknesses are recorded from the plurality of locations and for the specific time intervals.
In another embodiment of the present invention, a photoresist process characterization system for performing the characterization method is provided. A photoresist dispenser controllably dispenses photoresist on the semiconductor wafer while a spinning system rotates the wafer at a specified spin rate. A thickness measurement apparatus monitors the thicknesses of the photoresist on the wafer at a plurality of locations and at specific time intervals while the photoresist flows across the semiconductor wafer.
In yet another embodiment of the present invention, a process for coating a semiconductor wafer according to characteristics derived from the characterization process is provided. Photoresist is deposited on a semiconductor wafer and the wafer is spin-coated according to a recipe derived from a photoresist process characterization system. The photoresist process characterization system includes a photoresist dispenser which controllably dispenses photoresist on the semiconductor wafer while a spinning system rotates the wafer at a specified spin rate. A thickness measurement system monitors the thicknesses of the photoresist on the wafer at a plurality of locations and at specific time intervals while the photoresist flows across the semiconductor wafer.
In yet a further embodiment of the present invention, a computer-readable medium having computer-executable instructions thereon for performing the method of characterizing a photoresist process is provided.
In the drawings, which illustrate what is currently considered to be the best mode for carrying out the invention:
Photoresist coating processes include several variables, values for which may be maintained as “recipes” for referencing and reuse. Conventionally, recipes for photoresist coating of a semiconductor wafer were derived from only a very few data points, since the generation of each data point required the processing of a separate wafer and many photoresist and process environment parameters contribute to the variations in possible photoresist processes. For example, different photoresists have different viscosities that affect the spin-coating process. Also, the vapor pressure of the solvent that is in the photoresist to assist in the coating process presents variations to the overall process.
Reference to
As the photoresist flows across the semiconductor wafer 14, the thickness is monitored 22 at multiple locations across the semiconductor wafer by measurement system 42. Measurement of the photoresist thickness at multiple locations is indicative of the flow and thickness uniformity across the wafer. Measurement system 42 is configured to provide concurrent multiple readings across the radius or diameter of the semiconductor wafer at specific time intervals while the photoresist is flowing outwardly during the spinning process. Various measurement techniques for measuring film thickness are contemplated. One exemplary measurement system 42 includes one or more forms of sensors 44 which may assume various configurations, one of which is a multihead reflectometer as illustrated in
In
As indicated in
In order to more accurately calibrate the thickness data and uniformity data stored in database 26, one or more actual test semiconductor wafers corresponding to the data in the database may undergo further physical processing. The resulting semiconductor wafer is further measured to determine actual finished process thickness measurement data which may then be correlated 32 with the thickness measurement data stored in the database 26. Once semiconductor wafers are coated with photoresist, the next processing step includes a soft-bake step which accomplishes several important purposes, including driving off the solvent from the spun-on photoresist as well as providing adhesion and annealing benefits. Once the photoresist is soft-baked, characterization tests are performed on the photoresist thickness to determine actual soft-baked thickness measurement data 30 which is then correlated 32 to calibrate or improve the accuracy of thickness measurement data and uniformity data within database 26.
The present method further contemplates the generation of multiple spin curves at multiple spin rates. A query 34 determines whether further spin curves are desired and when such curves are desired, the spinning rate is changed 36 to another desired spin rate and processing returns with a new spin rate. When the data for the desired spin curves are derived, data from database 26 is output 38 for selection or utilization by either a manual operator or an automated operator for making the desired selection for the process setup. An output device 52 (
While the invention may be susceptible to various modifications and alternative forms, specific embodiments have been shown by way of example in the drawings and have been described in detail herein. However, it should be understood that the invention is not intended to be limited to the particular forms disclosed. Rather, the invention includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defmed by the following appended claims.
This application is a divisional of application Ser. No. 10/747,542, filed Dec. 29, 2003, pending.
Number | Date | Country | |
---|---|---|---|
Parent | 10747542 | Dec 2003 | US |
Child | 11599225 | Nov 2006 | US |