Data processing: generic control systems or specific application – Specific application – apparatus or process – Product assembly or manufacturing
Reexamination Certificate
2002-07-03
2004-12-21
Picard, Leo (Department: 2125)
Data processing: generic control systems or specific application
Specific application, apparatus or process
Product assembly or manufacturing
C700S117000
Reexamination Certificate
active
06834212
ABSTRACT:
TECHNICAL FIELD
The present invention relates generally to semiconductor fabrication, and more particularly to systems and methods for advanced process controls for semiconductor fabrication.
BACKGROUND OF THE INVENTION
As dimensions of semiconductor devices decrease, the available process window size decreases and manufacturing design rules shrink requiring tighter control over the manufacturing process. Generally, improvements in semiconductor fabrication processes and/or improvements in structural fabrication are required in order to further decrease critical dimensions and, thereby semiconductor devices. However, tighter control over the manufacturing process can be difficult to achieve, especially as critical dimension decrease further.
Semiconductor fabrication is a manufacturing process employed to create semiconductor devices in and on a wafer surface. Polished, blank wafers come into semiconductor fabrication, and exit with the surface covered with large numbers of semiconductor devices. The semiconductor fabrication includes a large number of steps and/or processes that control and build the devices. The basic processes utilized are layering, patterning, doping and heat treatments. Layering is an operation that adds thin layers to the wafer surface. Layers can be, for example, insulators, semiconductors and/or conductors and are grown or deposited via a variety of processes. Some common deposition techniques are chemical vapor deposition (CVD), evaporation and sputtering. Patterning is a series of steps that results in the removal of selected portions of surface layers. After removal, a pattern of the layer is left on the wafer surface. The material removed can be, for example, in the form of a hole in the layer or a remaining island of the material. The patterning transfer process is also referred to as photomasking, masking, photolithography or microlithography. The actual subtractive patterning, e.g. removal of material from the surface film, is done by plasma etching. The goal of the patterning process is to create desired shapes in desired dimensions (e.g., feature size) as required by a circuit design and to locate them in their proper location on the wafer surface. Patterning is generally considered the most important of the four basic processes. Doping is the process that adds specific amounts of dopants to the wafer surface. The dopants can cause the properties of layers to be modified (e.g., change a semiconductor to a conductor). A number of techniques, such as thermal diffusion and ion implantation can be employed for doping. Heat treatments are another basic operation in which a wafer is heated and cooled to achieve specific results. Typically, in heat treatment operations, no additional material is added or removed from the wafer, although contaminates and vapors may evaporate from the wafer. One common heat treatment is called annealing, which repairs damage to crystal structure of a wafer/device generally caused by doping operations. Other heat treatments, such as alloying and driving of solvents, are also employed in semiconductor fabrication.
Processes employed in semiconductor fabrication typically employ a significant number of variables. A process can, for example, utilize one or more flow rates, composition ratios, temperature, pressure, spin rate, time and the like. Additionally, these variables are generally subject to processing constraints that are employed to reduce or prevent damage to the wafer and/or semiconductor device. Constraints may also be employed to ensure the creation and/or maintenance of desirable device characteristics (e.g., fast switching speeds, low leakage current, etc.) For example, an exemplary constraint could be that oxygen flow must be greater than 15 sccm and less than 20 sccm or that a CHF
3
/CF
4
ratio not to exceed 20. Because of the number of variables involved, the constraints on those variables, and the desire to create favorable device characteristics while suppressing undesirable device characteristics, it can be difficult to control the manufacturing process let alone maximize the benefits of the manufacturing process. It has been empirically observed that controlling the manufacturing process becomes even more challenging as smaller and smaller devices need to be fabricated. The control of fabrication processes can be improved via feed forward and feedback control techniques that use in situ (or in-line) data to improve the results of the process. These techniques are known as “Advanced Process Control” techniques and typically work by building a predictive model of the manufacturing fabrication process. The predictive model is able to “predict” the outcome (e.g., values for one or more device characteristics) of the manufacturing process given the state of process variables and applicable constraints. One can invert the usage of the predictive model by using a value for a desired outcome and then “solving” the predictive model to identify the required state of process variables.
Fabrication processes that are modeled with only one or two variables can generally be “solved” fairly easily. The “Solved” means finding and/or assigning values to these variables that produce desired results (e.g., dimensions, locations, doping concentrations . . . ). These one or two variable process models can typically be quickly solved as linear or quadratic problems. However, more complex process models that utilize a higher number of variables are more difficult to solve. Additionally, such process models are often nonlinear further increasing the difficulty of solving the models. Accordingly, these more complex process models require relatively large amounts of computational power to obtain a solution that substantially meets all goals and constraints. Additionally, these non-linear models can be unsolvable (e.g., a solution that satisfies the goal and all applicable constraints cannot be achieved in finite time using finite computational resources). Thus, obtaining solutions for complex process model (e.g., containing a large numbers of variables) are unobtainable or computationally expensive to solve and can preclude identifying solutions in real time.
SUMMARY OF THE INVENTION
The following is a summary of the invention in order to provide a basic understanding of some aspects of the invention. This summary is not intended to identify key/critical elements of the invention or to delineate the scope of the invention. Its sole purpose is to present some concepts of the invention in a simplified form as a prelude to the more detailed description that is presented later.
The present invention can facilitate improvements in semiconductor fabrication processes by permitting tighter control over the manufacturing process. Feed back and feed forward techniques can be employed to improve a fabrication process by reducing variations in the process output. Improved or tighter control of the fabrication process can result in the following benefits: a) better device characteristics; b) improved device performance; and/or c) improved device yield. Collectively these benefits improve device yield and/or produce devices with additional intrinsic value (erg., better performance, lower heat dissipation, lowered leakage current). Furthermore, the present invention serves to facilitate substantially automating all or a portion of semiconductor device fabrication. Fabrication processes can be initiated remotely or locally by sending directives. Additionally, the fabrication processes are dynamically adapted in real time to account for variations in expected inputs and/or outputs.
The invention adapts and controls a desired fabrication process by utilizing a control model. A suitable control model is obtained or derived from one or more other control models. This is said to be an “instantiated” model, viz. a model associated with a particular tool recipe (description of settings such as power, time, etc.), specified with sufficient detail so as to be capable of predicting outputs for a set of input values.
A formulaic description is then o
Freeland Mark
Patel Sukesh
Amin & Turocy LLP
Blue Control Technologies, Inc.
Cabrera Zoila E.
Picard Leo
LandOfFree
Method and apparatus for APC solver engine and heuristic does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for APC solver engine and heuristic, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for APC solver engine and heuristic will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3296136