The present invention relates to a configuration for a semiconductor wafer processing (e.g., atomic layer deposition, chemical vapor deposition, plasma vapor deposition, cleaning or etching, etc.) apparatus having multiple, single-wafer processing chambers (reactors).
In the field of thin film technology, larger production yields and higher productivity have been, and continue to be, driving forces behind the development of new wafer processing apparatus. For example, many atomic layer deposition (ALD) systems now being used commercially employ a batch processing approach wherein substrates to be coated are arranged in different planes, and wherein relatively large numbers of substrates are coated in a single reactor simultaneously. The popularity of such devices is due largely because ALD has an inherently lower deposition rate than competing processes. By processing several substrates at the same time (in parallel) in a batch reaction chamber, total wafer throughput can be increased.
Unfortunately, batch processing has some inherent disadvantages, and addressing the throughput limitations of ALD by batch processing seems to trade one set of problems for another. For example, in batch processor systems cross-contamination of substrates poses a significant problem. Batch processing also inhibits process control, process repeatability from substrate to substrate and batch to batch, and necessitates post-processing film-removal solutions for backside deposition. All of these factors severely affect overall system maintenance, yield, reliability, and therefore net throughput and productivity.
What is needed therefore is a high productivity ALD system architecture that allows multiple substrates to be processed while still providing attractive throughput and yield, and which at the same time conservatively uses expensive clean room and associated production floor space.
One prior solution that attempts to address this need is described in U.S. Pat. No. 5,855,681 of Maydan et al. In particular, the '681 patent describes a semiconductor wafer processing apparatus that includes multiple processing chambers, each with a pair of single wafer processing regions. The processing regions of each chamber are isolatable from each other but share a common gas supply and common exhaust pump. The processing chambers are configured to allow multiple, isolated processes to be performed concurrently in the different processing regions, so that a pair of wafers can be processed simultaneously in each chamber. Each processing region of each processing chamber is connected to a common transfer chamber, which includes a wafer handler adapted to transfer simultaneously two wafers from a load lock chamber to the twin processing regions of a processing chamber.
The '681 patent points out that the processing regions of each processing chamber are isolatable from one another inasmuch as the “processing regions have a confined plasma zone separate from the adjacent region which is selectively communicable with the adjacent region via an exhaust system.” However, the gas lines which provide gas into the gas distribution system within each processing region are connected to a single, common gas source line and are therefore shared or commonly controlled for the delivery of gas to each processing region of a chamber:
The '681 design appears to preclude more than 2 processing regions per processing module with respect to loading, and limits the assembly of more than 2×3 processing regions within a relatively small footprint. Thus, while the solution proposed in the '681 patent does provide for some of the benefits of single wafer processing in a batch-type environment, there are limits as to the number of wafers which can be processed at a time.
In one embodiment of the present invention, a wafer processing apparatus includes one or more processing modules, each processing module having multiple, distinct, single-wafer processing reactors configured for semi-independent ALD and/or CVD film deposition therein; a robotic central wafer handler configured to provide wafers to and accept wafers from each of said wafer processing modules; and a single-wafer loading and unloading mechanism that includes a loading and unloading port and a mini-environment coupling the loading and unloading port to the robotic central wafer handler. The wafer processing reactors of any or all the processing modules may be arranged for wafer processing (i) along axes of a Cartesian coordinate system, or (ii) in quadrants defined by said axes, one axis of said coordinate system being parallel to a wafer input plane of the at least one of the process modules to which the single-wafer processing reactors belong. Each processing module can include up to four single-wafer processing reactors and preferred arrangements include 3 or 4 such reactors per module. Each of the single-wafer processing reactors of each processing module includes an independent gas distribution module.
The wafer processing apparatus may further include a chemical source sub module stacked atop a processing chamber containing the single-wafer processing reactors, and an electrical controller sub module stacked atop the chemical source sub module. The electrical controller sub. module and the chemical source sub module may be vertically displaceable from each other and from the processing chamber along one or more guide posts.
A further embodiment of the present invention includes a wafer processing module having up to four (and, preferably 3 or 4) semi-independent process zones arranged (i) in quadrants of, or (ii) along axes of a Cartesian coordinate system, one axis of said coordinate system being parallel to a wafer input plane of the process module, said process zones being configured for wafer processing such that reactant leakage from a subject one of the process zones to adjacent process zones thereof occurs in an amount no more than 5×10−2 times the reactant deposition rate in the subject process zone. The process zones are preferably equally accessible by a wafer indexer configured to load/unload wafers to/from the semi-independent process zones. Each of the semi-independent process zones may include an independent gas distribution module and/or the semi-independent process zones may share a common gas exhaust system (e.g., arranged so as to provide azimuthally-symmetric exhaust from each of the semi-independent process zones).
A still further embodiment of the present invention provides a wafer process module having a stack of electrical controls and gas source modules with said gas source modules being coupled to a reactor lid, the stack being capable of vertical motion and guided separation from a reactor chamber thereunder, thereby providing for removal of the lid, the electrical controls and the gas source modules collectively, or individually.
Another embodiment of the present invention provides for wafer handling by moving single wafers into or out of multi-single wafer reaction chamber zones using individual wafer end effectors of an indexer to sequentially accept wafers from a central vacuum robotic wafer handler, and to place said wafers substantially simultaneously on reactor susceptors within each reaction chamber zone.
The present invention is illustrated by way of example, and not limitation, in the figures of the accompanying drawings, in which:
Described herein is a unique configuration for a semiconductor wafer processing (e.g., atomic layer deposition, chemical vapor deposition, plasma vapor deposition, cleaning or etching, etc.) apparatus having multiple, single-wafer processing chambers. In the present description there are a number of details set forth in order to provide readers with a thorough understanding of the present invention, however, it will be apparent to those of ordinary skill in the art that there are many alterations in detail and scale that may be made in the embodiments described herein without departing from the spirit and scope of the present invention. For example, there are many wafer sizes presently in use in integrated circuit manufacturing, and processing stations configured according to embodiments of the present invention may be constructed to accommodate individual wafer sizes or a range of wafer sizes. Furthermore, in addition to the features described in detail below, embodiments of the present invention may include some or all of the features of related wafer processing apparatus developed at least in part by some of the present inventors and described in the following patents and patent applications assigned to the assignee of the present invention, each of which is incorporated herein by reference:
As more fully described below with reference to the accompanying drawings, the present multi-single wafer architecture provides a throughput enhancement over conventional systems by a factor of 8 or 12 to 3 or 4 (i.e., 8/3 or 12/4, where the reference number of reactors per process module is 3 and 4, respectively), in the same or smaller footprint. Conventional systems typically have an “a real productivity metric” of about 3 wph/m2 (e.g., 30 wph in 10 m2) and use a standard robotic central handler using 3 or 4 single wafer process chambers. In contrast, the present wafer processing apparatus uses up to three process modules (described further below), each of which have up to (and preferably) four multiple single wafer (MSW) reactors. The invention is also applicable for use with dozens to hundreds of smaller piece parts that may placed on carriers whose sizes are substantially the same as 200 or 300 mm diameter. The reactor design for the present wafer processing apparatus may be optimized by using an array of four, semi-independent reactors within a given process module. The reactors may be laid out in an “on-axis” configuration within the process module, or, preferably, in an “in-quadrant” configuration, which provides for certain floor space and process control advantages. As illustrated in the accompanying drawings, the wafer processing apparatus may utilize a stacked supporting module configuration in which such modules are vertically movable for access to and service of chemical sources, electrical controls, and reactor lids. Further, the present wafer processing apparatus includes a unique indexing mechanism that allows for efficient loading and unloading of the reactor chambers.
Referring first to
The wafer processing apparatus (100) is illustrated with three conventional FOUP loading modules (112), a conventional mini-environment with an atmospheric robotic wafer transfer (120) to two 25 wafer capacity vacuum load locks (130) utilizing 2 to 25 wafers. If desired, a wafer aligner may be placed in the mini-environment but such a configuration is not shown here. The wafer processing apparatus (100) may be implemented in a 300 mm or 200-300 mm bridge configuration. The modules have shared pumping but may have or may not have independent precursor feed injection above the substrate surface. Independent precursor feed provides for some flexibility and control in matching film deposition characteristics. Included within each process module (112) of the wafer processing apparatus (100) is a unique wafer pick and place indexer mechanism configured to move wafers to each single wafer reactor of a respective process module (112). The indexer design is shown in detail in
The wafer processing apparatus (200) is illustrated with three conventional FOUP loading modules (212), a conventional mini-environment with an atmospheric robotic wafer transfer (220) to two vacuum load locks (230) of 2 to 25 wafer capacity. If desired, a wafer aligner may be placed in the mini-environment but such a configuration is not shown here. This configuration has a high system performance metric for smaller, limited production granularity and may be implemented in a 300 mm or 200-300 mm bridge configuration. The process module has shared pumping but independent precursor feed injection above the substrate surface.
Similarly, process modules housing more than 4 single wafer reactors may be used and are considered to be within this scope of the present invention. For example, modules housing 5, 8, or other numbers of reactors may be used. In such cases, the indexer apparatus described herein would need to be modified to accommodate the appropriate number of wafers. In some cases, this may mean departing from the central, circular indexer design discussed below and, instead, adopting an indexer that includes linear translation motion as well as rotational (e.g., one which resembles a race track around or between the periphery of the reactors housed within the process module; or a central, linear track arrangement between the reactors, which may be arranged on alternate sides thereof).
The quadrant layout configuration has a smaller module area (1911 sq. units vs. 2021 sq. units) than the axis layout, and also provides better packing density in the overall system architectures illustrated in
To load the wafers in each reactor, an indexer (described further below), having wafers loaded thereon, is rotated about a central axis of the process module. An entry load circle position is illustrated by a circle centered some distance (555) from the plane of the input port (530) in the drawing illustrating the quadrant layout in the upper right portion of
One of the benefits of the in-quadrant design is the sharing of perturbations caused by the effects of the wafer entrance slot valve. In the on-axis design, the perturbation is applied to a single wafer. Additionally, the effects of the slot valve may be offset by the use of a vertically movable susceptor as described in U.S. Pat. Nos. 5,855,675 and 6,174,377, both assigned to the assignee of the present invention and incorporated herein by reference.
Thus, in some embodiments, the present invention provides a process module having up to four independent process zones, said zones arranged for wafer processing in quadrants of a Cartesian coordinate system, the axes of said coordinate system being parallel and/or perpendicular to the wafer input plane of the process module. The in quadrant (or axis) reactor zones may be used in an apparatus for ALD and/or CVD film deposition or other single wafer processes such as plasma, cleaning or etching processes having an architecture consisting of one or more multiple, semi-independent, wafer process modules.
The stacked electrical, source module boxes and reaction chamber lid (645) may be moved vertically to elevate them relative to the process chamber (640) using parallel guiding support post(s) (680). This design provides for modular access to different sub modules and different service functions.
Thus, a wafer process module configured in accordance with embodiments of the present invention includes a stack of electrical controls and gas source modules with said source modules being coupled to a reactor lid. The entire stack is capable of vertical motion and guided separation from the reactor chamber, thereby providing the removal of the lid, the electrical controls and the source modules collectively, or individually.
Other features of the overall wafer processing apparatus are similar to those described above. The back module is a gas box (710). Stacked thereunder is a chemical source module (730), which is stacked on the process chamber (740) containing the reactors laid out in the quadrant design within. A wafer entrance slot (750) is included, and individual wafer reaction cylinders housings (760) that contain the susceptor-heater hardware are also shown. Ports (770) are provided for viewing, as required.
The first operation (shown in load view 1410) is wafer loading, whereby all four wafers are loaded onto the arms of the indexer in a chamber otherwise without wafers or precursor process gases. During this process, the four wafers are loaded sequentially on the indexer arms. An end effector (1412) is shown placing the last of 4 wafers on the indexer receiving arm at the south-east location. Once all four wafers have been loaded onto the indexer, the indexer is rotated by 45°, positioning the wafers over the centers of the susceptor-heaters (see view 1430). Two sets of circles are shown: one with wafers loaded onto the four arms of the indexer, and displaced 45° from the four quadrant wafer susceptor positions (for ease of drawing the indexer sequencing, these illustrations show the “axis layout” instead of the “quadrant layout,” but the operations described herein are equally applicable to both).
The second operation is placement. Placement view (1430) shows the four wafers being pin lifted above the plane of the indexer and in particular above the indexer's paddles or “grippers.” Once the wafers are above the plane of the indexer, the indexer is rotated by 45° so that its end effectors are positioned between the susceptor-heaters. The lift pins are retracted thus placing the wafers down on the susceptors with the indexer arms between the susceptors-heaters (see view 1450). A vertically translatable elevated susceptor-heater (pedestal) may be used to position the wafers in an optimal process zone with respect to gas distribution and annular pumping conduits as discussed in U.S. Pat. No. 6,387,185.
The third operation is process run. Process view (1450) shows the wafer processing apparatus configuration when precursors are to be exposed to the wafers surfaces. The indexer's end effectors remain out of the direct pathway of the precursors, allowing precursors to reactor with the wafer first. The indexer arms may be adapted to provide minimal impact on the gas flow during the deposition period. During depositions, the parasitic leakage depositions on the indexer are preferably less than 5×10−2 of that in a given reactor.
The fourth operation is placement of wafers from the suceptor to the end effectors of the indexer, as illustrated in the next process view (1470). Once film deposition on the wafers is complete, a vertically translatable retracted susceptor-heater (pedestal) may be used to achieve a lower position, suitable for the wafer pick operation. Lift pins elevate the wafers above the plane of the indexer, and the indexer is rotated under the wafers. The lift pins retract and the wafers are placed on the end effectors of the indexer, over the centers of the susceptor-heaters as shown.
The fifth operation is unloading as illustrated in the unload view (1490). The indexer is rotated 45°, providing a wafer with a film deposited on it to face the exit (entrance) slot, viewed in the southeast direction of the unload view (1490). The wafers are then removed from the indexer one at a time via the end effector (1412) of the central wafer robotic wafer handler.
Of course, other loading/processing sequences than that described above may be used.
The system throughput is a function of the rate at which wafers can be loaded from a front opening unified pod (FOUP) (112, 212, 312, 412) into the batch load locks (130, 240, 340, 430) and from there through the central vacuum robotic chamber to the process module, as well as the process time. For a process module with 50 wph throughput, the system throughput will be approximately 46 wph. For a two process module system with an intrinsic gross throughput of 100 wph, the system throughput is approximately 75 wph, but this may be improved by enhancements in wafer handling. Thus, embodiments of the present invention provide a wafer handling apparatus and process for moving wafers into or out of multi-single wafer reaction chamber zones for the purpose of ALD or CVD film deposition. As indicated above, an embodiment of such an apparatus includes four receiving wafer end effectors, said end effectors being used to sequentially accept wafers from a central vacuum robotic wafer handler, and configured to place said wafers (substantially simultaneously) on the reactor susceptors for film deposition.
The present system may be operated in a parallel mode, wherein all wafer are processed together and simultaneously after wafers are loaded onto the susceptors. Alternately, a process may be run in one semi-independent station that is followed by another process. In the case of ALD, an exposure may be taking place in one process module, while a different exposure or a purge may be taking place on another process module. The present wafer processing apparatus may also be compatible with plasma enhanced processes, where remote or direct plasma hardware is configured with each semi-independent reactor. Sources for each quadrant may be parallel or independently fed. Pump configurations may be shared or independent.
The present system may be operated in a parallel mode, wherein all wafer are processed together and simultaneously after wafers are loaded onto the susceptors. Alternately, a process may be run in one semi-independent station that is followed by another process. In the case of ALD, an exposure may be taking place in one process module, while a different exposure or a purge may be taking place on another process module. The present wafer processing apparatus may also be compatible with plasma enhanced processes, where remote or direct plasma hardware is configured with each semi-independent reactor. Sources for each quadrant may be parallel or independently fed. Pump configurations may be shared or independent. Alternatively, or in addition, serial and/or parallel processing can be carried out in the reactors in one or more of the process modules. For example, in the case where the deposition rates are balanced, 2 reactors can run one process (e.g., a film type) and 2 reactors a different process (e.g., a different film type). In still another case, where the deposition rates are not balanced, a larger number of reactors can be dedicated to the lower deposition rate process and a smaller number of reactors used for the higher deposition rate process.
The present application is a non-provisional of, claims priority to and incorporates by reference U.S. Provisional Patent Application 60/609,598, filed Sep. 13, 2004.
Number | Date | Country | |
---|---|---|---|
60609598 | Sep 2004 | US |