The invention relates to process control, and particularly to methodology that reduces CD (Critical Dimension) non-uniformity in resist processing for semiconductor mask or wafer manufacturing, and compensates interfield signature from other processes.
For commonly used commercialized chemically amplified resist, the PEB sensitivity is of the order of 10 nm/degree. That is, change of PEB temperature by 1 degree will result in about 10 nm change of resist CD. Thus, for application of, e.g., the 90-nm node and beyond, temperature uniformity of a hot plate is controlled to be less than 0.2°, being a big challenge for track manufacturers.
Tokyo Electronic Industry Co., Ltd (TEL) uses a RTD (Resistance Temperature Detector) wafer to calibrate their chilling precision hot plate (CPHP). Since the chamber accommodating CPHP is opened in order to place the RTD wafer, the temperature measured in such a way cannot reflect the original thermal environment that a wafer experiences during its PEB. Further, it takes a long time for the RTD wafer to complete the whole calibration procedure. Additionally, since there are 29 sensors on the RTD wafer, such characterization of a hot plate will not be detailed enough.
OnWafer Technologies proposes a wireless device that mimics a wafer with built-in thermal sensors. Thus, it can be sent into the track just like a production wafer and measures the whole thermal cycle at correct thermal environment. However, if the required accuracy is very high, matching of response between different sensors should be guaranteed. If it is uniformity of CD, rather than temperature, that is to be adjusted, then the relation between CD and the whole thermal cycle needs to be established. Also, similar to the RTD wafer, there are 42 sensors on the wireless sensor wafer, still too few to give a sufficient characterization of a hot plate. Moreover, such methodology is of high cost due to short lifetime of the wireless sensor wafer.
What is pursued here is a methodology directly utilizing resist as the thermal sensor for accurate, efficient, and economical measurement of the temperature of a hot plate, so realizing process control by adjusting the hot plate, therefore improving CD uniformity.
Accordingly, a heating apparatus/system and a process control method for use in the heating apparatus/system are provided. A baseline CD map W is first obtained by measuring CD at predetermined locations within a substrate processed with the heating apparatus at a baseline setting. A target CD map M is also obtained by assigning CD value at each of the predetermined locations within a substrate. An error CD map ΔW is then calculated by subtracting CD value of the target CD map from that of the baseline CD map at each of the predetermined locations within a substrate. An original CD map N is obtained by measuring CD at the predetermined locations within a substrate processed with the heating apparatus at an original setting. A perturbed CD map Wi for each heating zone i is also obtained by measuring CD at predetermined locations within a substrate processed with the heating apparatus at a setting where the temperature of the heating zone i is deviated from the original setting by a predetermined value ΔTi. A basis function Pi for each heating zone i is then calculated by subtracting CD value of the original CD map from that of the perturbed CD map at each of the predetermined locations within a substrate. Afterwards, the error CD map ΔW is expanded by a collection of the basis functions with the expansion coefficient ci for each basis function Pi, respectively. The baseline setting of each heating zone i of the heating apparatus is then adjusted according to the error CD map ΔW, the expansion coefficient ci, and the basis functions.
A substrate is then processed with the heating apparatus at the adjusted baseline setting.
The baseline setting of each heating zone i of the heating apparatus is adjusted by an amount −ciΔTi.
The expansion coefficient ci is obtained by using a least-square-fitting method.
The CD of the baseline CD map is based on an after-development or after-etching image.
The CD of the original CD map and the perturbed CD map is based on an after-development or after-etching image.
The interfield signature of prior processes, e.g., CVD (Chemical Vapor Deposition), CMP (Chemical Mechanical Polishing), can also be compensated as long as it is consistent from run to run.
A semiconductor fabrication method incorporating a heating apparatus is also provided. An exposed photoresist on a substrate is heated with the heating apparatus at a baseline setting, and is further processed to delineate a collection of patterns at predetermined locations. CD of the collection of patterns is measured to obtain a baseline CD map W. CD value at each of the predetermined locations is assigned to generate a target CD map M. Then, temperature distribution of the heating apparatus is adjusted according to an error CD map ΔW defined as the difference between the baseline CD map W and the target CD map M.
An exposed photoresist on a substrate is heated with the heating apparatus at an original setting, and is further processed to delineate a collection of patterns at the predetermined locations. CD at the predetermined locations on the substrate is measured to obtain an original CD map N. An exposed photoresist on a substrate is heated with the heating apparatus at a setting where the temperature of the heating zone i is deviated from the original setting by a predetermined value ΔTi, and is further processed to delineate a collection of patterns at the predetermined locations. CD at the predetermined locations on the substrate is measured to obtain a perturbed CD map Wi for the heating zone i of the heating apparatus. Then, a basis function Pi for the heating zone i is calculated by subtracting CD value of the original CD map from that of the perturbed CD map at each of the predetermined locations within a substrate. Repeat the above procedure until the basis function Pi for each heating zone i is obtained. The error CD map ΔW is then expanded by a collection of the basis functions with a corresponding expansion coefficient ci for each basis function Pi. The baseline setting of each heating zone i of the heating apparatus is adjusted according to the error CD map ΔW, the expansion coefficient ci, and the basis functions.
A substrate is then processed in semiconductor mask or wafer manufacturing with the heating apparatus at the adjusted baseline setting.
The baseline setting of each heating zone i of the heating apparatus is adjusted by an amount −ciΔTi.
The expansion coefficient ci is obtained by using a least-square-fitting method.
The CD is based on an after-development or after-etching image.
The interfield signature of prior processes, e.g., CVD, CMP, can also be compensated as long as it is consistent from run to run.
The aforementioned methodology may take the form of program code embodied in a tangible media. When the program code is loaded into and executed by a machine, the machine can become an apparatus for practicing the disclosed methodology.
The aforementioned features and advantages will become apparent by referring to the following detailed description with reference to the accompanying drawings, wherein:
Though one knows that CDU is not good, also that such non-uniformity is mainly from temperature non-uniformity of a heating apparatus, such as a hot plate, for PEB, one cannot adjust the temperature distribution of the hot plate at his will. There are, e.g., for TEL's track, 7 or 15 degrees of freedom, corresponding to 7 or 15 heating zones of the hot plate (210 and 220), as illustrated in
First, in step S410, an exposed photoresist on a substrate, such as a wafer, is heated on the heating apparatus 310 at a baseline setting. In step S420, the exposed and heated photoresist is then developed, or further processed, to delineate a collection of patterns at predetermined locations of the substrate. Then, in step S430, CD of the collection of patterns is measured to obtain a baseline CD map W. In step S440, CD value at each of the predetermined locations is assigned to generate a target CD map M. Thereafter, in step S450, temperature distribution of the heating apparatus 310 is adjusted according to an error CD map ΔW defined as the difference between the baseline CD map W and the target CD map M. The heating apparatus 310 with the adjusted baseline setting (temperature distribution) can be used for processing subsequent substrates in semiconductor mask or wafer manufacturing. Detailed descriptions of the adjustment follow.
In step S510, a baseline CD map W is first obtained by measuring CD at predetermined locations within a substrate, such as a wafer, processed with the heating apparatus 310 at a baseline setting. In step S520, a target CD map M is also obtained by assigning CD value at each of the predetermined locations within a substrate. In step S530, an error CD map ΔW is then calculated by subtracting CD value of the target CD map from that of the baseline CD map at each of the predetermined locations within a substrate, i.e., ΔW=W−M. More explicitly ΔWj=Wj−Mj at each location j.
In step S540, an original CD map N is obtained by measuring CD at the predetermined locations within a substrate processed with the heating apparatus at an original setting. In step S550, a perturbed CD map Wi for each heating zone i is obtained by measuring CD at the predetermined locations within a substrate processed with the heating apparatus at a setting where the temperature of the heating zone i is deviated from the original setting by a predetermined value ΔTi. It is understood that the original CD map corresponding to the original setting may be identical to the baseline CD map corresponding to the baseline setting. How worse the CD non-uniformity of the baseline CD map is determines how large ΔTi should be used. If ΔTi is too small, extrapolation will have to be employed in the subsequent fitting procedure, making it less precise, while if ΔTi is too large, nonlinear effect will occur. In step S560, a basis function Pi for each heating zone i is then calculated by subtracting CD value of the original CD map N from that of the perturbed CD map Wi at each of the predetermined locations within the substrate, i.e., Pi=Wi−N. More explicitly, Pji=Wji−Nj at each location j. The operation of step S560 is repeated for each heating zone of the heating apparatus 310 until all basis functions are obtained.
Thereafter, in step S570, the baseline setting (temperature distribution) of each heating zone i of the heating apparatus 310 is adjusted according to the error CD map ΔW, the expansion coefficient ci, and the basis functions. In step S570, the error CD map ΔW is expanded by a collection of the basis functions with a corresponding expansion coefficient ci for each basis function Pi, i.e.,
The coefficient ci can be found by minimizing the total error
by using a least-square-fitting method. The method can be found in standard textbooks on numerical analysis, and is thus omitted here. Therefore, if the baseline setting of each heating zone i of the heating apparatus 310 is adjusted by −ciΔTi, then the baseline CD map will theoretically improve
It is understood that W, M, ΔW, Wi, N, Pi, and Wc are all scalar fields, i.e., a collection of a scalar value at each of the predetermined locations on a substrate.
The process control describe above, initially intended to be used at the ADI stage, can be extended to improve the CD uniformity at the AEI stage. This corresponds to compensate the etching chamber characteristics by PEB by adjusting the temperature profile of the heating apparatus (hot plate). The change represented by the procedures is that CD is measured at the AEI stage. Further, any interfield signature of prior processes, e.g., CVD, CMP, is also corrected implicitly if the substrate already undergoes these prior processes.
Embodiments of the method and system, or certain aspects or portions thereof, may take the form of program code (i.e., executable instructions) embodied in tangible media, such as floppy diskettes, CD-ROMs, hard drives, or any other machine-readable storage medium, wherein, when the program code is loaded into and executed by a machine, such as a computer, the machine thereby becomes an apparatus for practicing the embodiment. Embodiments of the method and system may also be embodied in the form of program code transmitted over some transmission medium, such as electrical wiring or cabling, through fiber optics, or via any other form of transmission, wherein, when the program code is received and loaded into and executed by a machine, such as a computer, the machine becomes an apparatus for practicing the embodiments. When implemented on a general-purpose processor, the program code combines with the processor to provide a unique apparatus that operates analogously to application specific logic circuits.
Although the present invention has been described according to predefined embodiment, it is not intended to limit the invention thereto. Those skilled in this technology can still make various alterations and modifications without departing from the scope and spirit disclosed.
Number | Name | Date | Kind |
---|---|---|---|
6566016 | Ziger | May 2003 | B1 |
6590219 | Ziger | Jul 2003 | B1 |
7001697 | Park et al. | Feb 2006 | B2 |
20030157415 | Ziger | Aug 2003 | A1 |
20040182822 | Chen et al. | Sep 2004 | A1 |
Number | Date | Country | |
---|---|---|---|
20060196960 A1 | Sep 2006 | US |