WO1999054694A1 - Method and apparatus for monitoring plasma processing operations - Google Patents
Method and apparatus for monitoring plasma processing operations Download PDFInfo
- Publication number
- WO1999054694A1 WO1999054694A1 PCT/US1999/008894 US9908894W WO9954694A1 WO 1999054694 A1 WO1999054694 A1 WO 1999054694A1 US 9908894 W US9908894 W US 9908894W WO 9954694 A1 WO9954694 A1 WO 9954694A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- plasma
- processing chamber
- optical emissions
- data
- calibration light
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 1749
- 238000012545 processing Methods 0.000 title claims abstract description 755
- 238000012544 monitoring process Methods 0.000 title claims abstract description 369
- 230000008569 process Effects 0.000 claims abstract description 1228
- 230000003287 optical effect Effects 0.000 claims abstract description 594
- 238000011156 evaluation Methods 0.000 claims abstract description 96
- 230000000694 effects Effects 0.000 claims abstract description 69
- 235000012431 wafers Nutrition 0.000 claims description 422
- 238000013479 data entry Methods 0.000 claims description 233
- 230000009183 running Effects 0.000 claims description 155
- 238000004519 manufacturing process Methods 0.000 claims description 127
- 239000000835 fiber Substances 0.000 claims description 105
- 230000008859 change Effects 0.000 claims description 103
- 238000003860 storage Methods 0.000 claims description 102
- 230000003750 conditioning effect Effects 0.000 claims description 93
- 230000003595 spectral effect Effects 0.000 claims description 89
- 238000004140 cleaning Methods 0.000 claims description 64
- 230000009471 action Effects 0.000 claims description 41
- 238000000429 assembly Methods 0.000 claims description 38
- 230000000712 assembly Effects 0.000 claims description 38
- 238000012797 qualification Methods 0.000 claims description 33
- 239000004065 semiconductor Substances 0.000 claims description 28
- 238000001914 filtration Methods 0.000 claims description 27
- 230000006870 function Effects 0.000 claims description 27
- 238000009826 distribution Methods 0.000 claims description 23
- 238000011068 loading method Methods 0.000 claims description 23
- 230000000670 limiting effect Effects 0.000 claims description 18
- 239000000463 material Substances 0.000 claims description 17
- 238000004886 process control Methods 0.000 claims description 17
- 238000012552 review Methods 0.000 claims description 16
- 230000000977 initiatory effect Effects 0.000 claims description 12
- 230000001186 cumulative effect Effects 0.000 claims description 11
- 230000002411 adverse Effects 0.000 claims description 10
- 230000004913 activation Effects 0.000 claims description 9
- 239000006117 anti-reflective coating Substances 0.000 claims description 9
- 238000012986 modification Methods 0.000 claims description 9
- 230000004048 modification Effects 0.000 claims description 9
- 230000001965 increasing effect Effects 0.000 claims description 8
- 230000002250 progressing effect Effects 0.000 claims description 8
- 239000011358 absorbing material Substances 0.000 claims description 7
- 238000000926 separation method Methods 0.000 claims description 6
- 239000007787 solid Substances 0.000 claims description 6
- 230000002123 temporal effect Effects 0.000 claims description 5
- 238000013500 data storage Methods 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 230000001154 acute effect Effects 0.000 claims description 3
- 230000004044 response Effects 0.000 claims description 3
- 230000001419 dependent effect Effects 0.000 claims description 2
- 238000007619 statistical method Methods 0.000 claims description 2
- 230000003116 impacting effect Effects 0.000 claims 4
- 238000007789 sealing Methods 0.000 claims 2
- 238000012795 verification Methods 0.000 claims 1
- 210000002381 plasma Anatomy 0.000 description 1636
- 238000001228 spectrum Methods 0.000 description 688
- 230000036541 health Effects 0.000 description 202
- 230000002159 abnormal effect Effects 0.000 description 89
- 238000003909 pattern recognition Methods 0.000 description 77
- 238000004458 analytical method Methods 0.000 description 49
- 239000000047 product Substances 0.000 description 42
- 230000004075 alteration Effects 0.000 description 36
- 238000001514 detection method Methods 0.000 description 35
- 239000007789 gas Substances 0.000 description 31
- 239000010410 layer Substances 0.000 description 11
- 238000005530 etching Methods 0.000 description 9
- 238000011160 research Methods 0.000 description 9
- 238000013480 data collection Methods 0.000 description 8
- 239000000203 mixture Substances 0.000 description 7
- NJPPVKZQTLUDBO-UHFFFAOYSA-N novaluron Chemical compound C1=C(Cl)C(OC(F)(F)C(OC(F)(F)F)F)=CC=C1NC(=O)NC(=O)C1=C(F)C=CC=C1F NJPPVKZQTLUDBO-UHFFFAOYSA-N 0.000 description 7
- 230000000007 visual effect Effects 0.000 description 6
- 238000013523 data management Methods 0.000 description 4
- 239000010453 quartz Substances 0.000 description 4
- 238000007670 refining Methods 0.000 description 4
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 238000012512 characterization method Methods 0.000 description 3
- 230000033001 locomotion Effects 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 229910052751 metal Inorganic materials 0.000 description 3
- 239000002184 metal Substances 0.000 description 3
- 150000002739 metals Chemical class 0.000 description 3
- 238000005192 partition Methods 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- 239000002210 silicon-based material Substances 0.000 description 3
- 239000002904 solvent Substances 0.000 description 3
- 238000004544 sputter deposition Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000012546 transfer Methods 0.000 description 3
- 238000012384 transportation and delivery Methods 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 239000013626 chemical specie Substances 0.000 description 2
- 238000005229 chemical vapour deposition Methods 0.000 description 2
- 238000010835 comparative analysis Methods 0.000 description 2
- 230000000052 comparative effect Effects 0.000 description 2
- 230000007812 deficiency Effects 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000014759 maintenance of location Effects 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 238000001020 plasma etching Methods 0.000 description 2
- 238000011112 process operation Methods 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 230000006641 stabilisation Effects 0.000 description 2
- 238000011105 stabilization Methods 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 239000000725 suspension Substances 0.000 description 2
- 230000036962 time dependent Effects 0.000 description 2
- TVEXGJYMHHTVKP-UHFFFAOYSA-N 6-oxabicyclo[3.2.1]oct-3-en-7-one Chemical compound C1C2C(=O)OC1C=CC2 TVEXGJYMHHTVKP-UHFFFAOYSA-N 0.000 description 1
- RYGMFSIKBFXOCR-UHFFFAOYSA-N Copper Chemical compound [Cu] RYGMFSIKBFXOCR-UHFFFAOYSA-N 0.000 description 1
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 1
- RTAQQCXQSZGOHL-UHFFFAOYSA-N Titanium Chemical compound [Ti] RTAQQCXQSZGOHL-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 239000003124 biologic agent Substances 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 239000006227 byproduct Substances 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 229910052802 copper Inorganic materials 0.000 description 1
- 239000010949 copper Substances 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000003111 delayed effect Effects 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000009658 destructive testing Methods 0.000 description 1
- 239000003989 dielectric material Substances 0.000 description 1
- 238000001312 dry etching Methods 0.000 description 1
- 229920001971 elastomer Polymers 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 1
- 229910052737 gold Inorganic materials 0.000 description 1
- 239000010931 gold Substances 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 150000002500 ions Chemical class 0.000 description 1
- 238000012886 linear function Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005459 micromachining Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000013618 particulate matter Substances 0.000 description 1
- 238000002161 passivation Methods 0.000 description 1
- 238000000059 patterning Methods 0.000 description 1
- 229920002120 photoresistant polymer Polymers 0.000 description 1
- 239000004033 plastic Substances 0.000 description 1
- 229920003023 plastic Polymers 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000004801 process automation Methods 0.000 description 1
- 238000005086 pumping Methods 0.000 description 1
- 238000007788 roughening Methods 0.000 description 1
- 239000005060 rubber Substances 0.000 description 1
- 229910052594 sapphire Inorganic materials 0.000 description 1
- 239000010980 sapphire Substances 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 229910052710 silicon Inorganic materials 0.000 description 1
- 239000010703 silicon Substances 0.000 description 1
- 238000010183 spectrum analysis Methods 0.000 description 1
- 229910052719 titanium Inorganic materials 0.000 description 1
- 239000010936 titanium Substances 0.000 description 1
- WFKWXMTUELFFGS-UHFFFAOYSA-N tungsten Chemical compound [W] WFKWXMTUELFFGS-UHFFFAOYSA-N 0.000 description 1
- 229910052721 tungsten Inorganic materials 0.000 description 1
- 239000010937 tungsten Substances 0.000 description 1
- 235000013311 vegetables Nutrition 0.000 description 1
- 239000001993 wax Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J3/443—Emission spectrometry
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01J—MEASUREMENT OF INTENSITY, VELOCITY, SPECTRAL CONTENT, POLARISATION, PHASE OR PULSE CHARACTERISTICS OF INFRARED, VISIBLE OR ULTRAVIOLET LIGHT; COLORIMETRY; RADIATION PYROMETRY
- G01J3/00—Spectrometry; Spectrophotometry; Monochromators; Measuring colours
- G01J3/28—Investigating the spectrum
- G01J2003/2866—Markers; Calibrating of scan
Definitions
- the present invention generally relates to the field of plasma processes and, more particularly, to monitoring/evaluating such plasma processes.
- Plasma is used in various types of industrial-type processes in the semiconductor and printed wiring board industries, as well as in various other industries such as in the medical equipment and automotive industries.
- One common use of plasma is for etching away materials in an isolated or controlled environment.
- Various types of materials may be etched by one or more plasma compositions, including glasses, silicon or other substrate materials, organics such as photoresist, waxes, plastics, rubbers, biological agents, and vegetable matter, and metals such as copper, aluminum, titanium, tungsten, and gold.
- Plasma is also utilized for depositing materials such as organics and metals onto an appropriate surface by various techniques, such as via chemical vapor deposition.
- Sputtering operations may also utilize plasmas to generate ions which sputter away material from a source (e.g., metals, organics) and deposit these materials onto a target such as a substrate.
- Surface modification operations also use plasmas, including operations such as surface cleaning, surface activation, surface passivation, surface roughening, surface smoothing, micromachining, hardening, and patterning.
- Plasma processing operations can have a significant effect on a company's profit margin. This is particularly true in the semiconductor and printed wiring board industries. Consider that a single semiconductor fabrication facility may have up to 200-300 processing chambers and that each processing chamber in commercial production may process at least about 15-20 wafers per hour.
- an eight inch wafer which is processed in one of these chambers in some cases may be used to produce up to 1 ,500 semiconductor chips which are each worth at least about $125, and that each of these semiconductor chips are in effect "pre-sold.” Therefore, a single wafer which has undergone an abnormal plasma process and which is scrapped will result in lost revenues of at least about $187,500.
- plasma recipe The particular plasma process which acts on the wafer such that a semiconductor device may be formed therefrom is commonly referred to as a plasma recipe.
- a plasma recipe as used in relation to the present invention means a plasma processing protocol which includes one or more different and distinct plasma steps (e.g., a certain combination of certain steps).
- "Different and distinct" means that each plasma step produces a different, predetermined result on the product being processed (e.g., a wafer).
- Differences between plasma steps may be realized by changing one or more process conditions, including without limitation the composition of the plasma, the temperature and pressure in the processing chamber, DC bias, pumping speeds, and power settings.
- the sequence of the plasma steps, as well as the result of each plasma step, also produces a desired overall or cumulative end result for the plasma recipe.
- Plasma processes may be run on wafers in a commercial production facility in the following manner.
- a cassette or boat which stores a plurality of wafers (e.g., 24) is provided to a location which may be accessed by a wafer handling system associated with one or more processing chambers.
- One wafer at a time is processed in the chamber, although some chambers may accommodate more than one wafer at a time for simultaneous plasma processing.
- One or more qualification wafers may be included in each cassette, and the rest are commonly referred to as production wafers. Both the qualification and production wafers are exposed to the same plasma process in the chamber.
- One common monitoring technique associated with plasma recipes run on wafers is endpoint detection
- Current endpoint detection systems attempt to identify when a single plasma step of a given plasma recipe is complete, or more specifically that point in time when the predetermined result associated with the plasma step has been produced on the product
- a representative "predetermined" result is when a layer of a multi- layered wafer has been completely removed in a manner defined by a mask or the like
- prior art systems exist for attempting to identify the endpoint of a single step of a multiple step plasma recipe, no known system is able to identify the endpoint of each step of a multiple step plasma recipe, or even any two steps of a multiple step recipe for that matter
- the amount of gases which are used to generate the plasma may be reduced by terminating a given plasma step when it has achieved its desired result More importantly, terminating a given plasma step at or very shortly after its endpoint has been reached prevents the wafer from being over-etched to an undesired degree Over-etching a wafer removes more material from the wafer than desired, such as by etching away portions of the layer immediately following that which was to be etched, and may also result in the undesirable sputtering of materials onto other portions of the wafer The resulting effect on the semiconductor dev ⁇ ce(s) formed from this wafer may reduce the quality of the semiconductor dev ⁇ ce(s), may go undetected until the semiconductor dev ⁇ ce(s) has been delivered to the customer which would not be desirable if the dev ⁇ ce(s) was defective or deficient in any way, or both Finally, a certain degree of over- etching of
- Endpoint detection is desirable in theory for plasma processes Certain deficiencies became evident as attempts were made to implement endpoint detection techniques in commercial fabrication facilities Initially, all known endpoint detection techniques were developed by first chemically analyzing the subject plasma operation to identify a wavelength to key in on as being indicative of endpoint Fabrication facilities typically run a multiplicity of plasma recipes As such, these known endpoint detection techniques increase costs due to the required retention of an experienced chemist Moreover, these techniques often do not produce the intended result - that is the wavelength which is selected by the chemist may in fact not be at all indicative of endpoint when the plasma step is actually run since it is only "theory" based A given endpoint detection technique may also be dependent upon the processing chamber on which the technique was developed Accurate results may not be realized when the endpoint detection technique is used on other processing chambers Therefore, it would be desirable to have a plasma monitoring system in which the amount of chemical "pre-analysis" is reduced and which would allow the plasma monitoring system to work to an acceptable degree on multiple processing chambers (i e , a generic plasma monitoring system which was
- Plasma processing of product (e g , wafers) within the processing chamber will likely have an effect on the interior of the processing chamber which in turn may have an adverse effect on subsequent plasma recipes which are run on product within the chamber
- Certain "byproducts" of a plasma process run on product in the chamber may be deposited on one or more interior surfaces of the chamber These deposits may have some type of adverse effect on one or more plasma recipes which are being run in the processing chamber (e g , a processing chamber may be used to run more than one type of plasma recipe)
- Deposits on the interior surfaces of the processing chamber may have the following exemplary effects on the performance of the chamber a longer period of time may be required to reach the endpoint of one or more plasma steps of the plasma recipe, endpoint of one or more plasma steps may never be reached; and a result which is different than expected of the current plasma step may be undesirably realized (i.e., an unexpected/undesirable result).
- Processing chambers are typically removed from the production line on a scheduled, periodic basis for a cleaning operation to address the above-noted conditions, regardless of whether the chamber is actually in condition for a cleaning and even if the chamber was ready for cleaning well before this time. It would be desirable to have a plasma monitoring system which would provide an indication of when a processing chamber should be removed from production for cleaning. Cleaning operations which are used to address the above-noted deposits include plasma cleans of the interior of the processing chamber, wet cleans of the interior of the processing chamber, and replacement of certain components of the processing chamber which may actually be consumed by the plasma processes conducted therein and are therefore commonly referred to as "consumables".
- a plasma clean addresses the above-noted deposits by running an appropriate plasma in the processing chamber typically without any product therein (e.g., no production wafers), and therefore with the chamber being in an "empty" condition.
- the plasma acts on these deposits in a plasma clean and reduces the thickness thereof by chemical action, mechanical action, or both. Resulting vapors and particulate matter are exhausted from the chamber during the plasma clean. It would be desirable to have a plasma monitoring system which would provide an accurate indication of both the health and endpoint of the plasma clean currently being conducted within the processing chamber.
- a plasma clean alone will not adequately address the condition of the interior of the processing chamber.
- Another cleaning technique which may be employed, alone or in combination with a plasma clean is commonly referred to as a "wet clean.”
- Various types of solvents or the like may be used in a wet clean and are manually applied by personnel. In this regard, the subject processing chamber is depressurized, the chamber is opened to gain appropriate access, and the interior surfaces of the chamber are manually wiped down such that the solvents may remove at least some of the deposits by chemical action, mechanical action, or both.
- Plasma cleaning operations address the solvent residuals from the wet clean, "prep" the new components of the chamber for plasma processing of product in the chamber, or both It would be desirable to have a plasma monitoring system which would provide an accurate indication of both the health and endpoint of the plasma cleaning operation in this type of case
- Conditioning wafers may be run through the processing chamber before running production wafers through the processing chamber after any type of cleaning of the processing chamber, after any components of the chamber have been replaced, or in the case of a new chamber which has never had any plasma processes conducted therein
- An entire plasma processes is typically run on one or more conditioning wafers disposed in the subject processing chamber in a conditioning wafer operation
- Conditioning wafers may simply be "blanks" or may have some semiconductor device components thereon, and the running of entire plasma processes thereon may do nothing to the conditioning wafers or portions of the conditioning wafer may be etched. Nonetheless, no semiconductor devices are ever formed from a conditioning wafer and no integrated circuit of any kind is ever etched onto a conditioning wafer while running the plasma recipe thereon.
- conditioning wafers of this type are either refurbished (e.g., material is redeposited back into those areas which were etched during the conditioning wafer operation) and re-used again as a conditioning wafer or they are scrapped.
- the processing of these conditioning wafers further "preps” or "seasons” the chamber and is done for the purpose of placing the chamber in a certain condition for production. No devices are currently being used to identify when the processing of the conditioning wafers has achieved its intended purpose. Therefore, it would be desirable to have a plasma monitoring system which would provide an accurate indication of when the conditioning wafer operation may be terminated, as well as the health of such an operation.
- the present invention generally relates to various aspects of a plasma process. These aspects may be grouped into four main categories. One category relates in at least some manner to a calibration or initialization procedure, associated components, or both. The first aspect through the fourth aspect presented below are within this category. Another category relates in at least some manner to various types of evaluations which may be undertaken of a plasma process which was run, and more typically one which is currently being run, within the processing chamber (e.g., plasma health evaluations, plasma process/plasma process step identification, plasma "on” determinations). The fifth aspect through the eighth aspect presented below are within this second category.
- Yet another category associated with the present invention relates in at least some manner to the endpoint of a plasma process (e.g., plasma clean, conditioning wafer operation) or discrete/discernible portion thereof (e.g., a plasma step of a multiple step plasma recipe).
- a plasma process e.g., plasma clean, conditioning wafer operation
- discrete/discernible portion thereof e.g., a plasma step of a multiple step plasma recipe.
- the ninth aspect through the thirteenth aspect presented below are within this third category
- the fourth category associated with the present invention relates to how one or more of the above-noted aspects may be implemented into a semiconductor fabrication facility
- the fourteenth aspect through the seventeenth aspect presented below are within this fourth category
- a first aspect of the present invention is embodied in a plasma processing system having calibration capabilities in relation to the monitoring of plasma processing operations
- the plasma processing system includes a processing chamber having a window with an inner surface which is exposed to plasma processes conducted within the chamber and an outer surface which is isolated from such processes
- a plasma generator is associated with the plasma processing system to provide the plasma for the plasma processes Any technique and corresponding structure for forming a plasma in the chamber is appropriate for this first aspect of the present invention
- a first spectrometer assembly e g , one or more spectrometers of any type, such as scanning-type spectrometers and solid state spectrometers
- a calibration light source is also located outside of the chamber and operatively interconnected with the window through a second fiber optic cable assembly (e g , one or more fiber optic cables)
- Ends of the first and second fiber optic cable assemblies may be disposed on, but are preferably spaced from
- calibration involves a comparison between data relating to the calibration light which is sent to the window by the calibration light source (e g , a pattern, intensity, or both, of the corresponding optical emissions) and data relating to a first portion of this same calibration light which is reflected by the inner surface of the window on the processing chamber (e g , a pattern, intensity, or both, of the first portion)
- the inner surface of the window of the processing chamber is that portion of the window which is typically affected by plasma processes conducted within the chamber Changes on the inner surface of the window may have an effect on any evaluation of a plasma process being conducted within the chamber if such an evaluation is based upon the transmission of optical emissions through the window
- Information on the window in relation to calibration in accordance with the first aspect of the present invention preferably includes information which is specific to the inner surface of the window on the processing chamber through which optical emissions are obtained That is, calibrations in accordance with this first aspect are preferably in relation to only the inner surface of the window and not the outer surface of the window Steps may be undertaken such that the portion of the calibration light which is reflected by the inner surface of the window is readily available for comparison with the calibration light in the form as it is being sent to the window in another embodiment of the subject first aspect This may be accomplished through appropriately configuring the window For instance, at least a portion of the window, which includes that area where the calibration light impacts the window, may have a generally wedge- shaped configuration (e g , variable window thickness) Another characterization of a window configuration in this embodiment is that at least a portion of the inner and outer surfaces of the window may be disposed in non-parallel relation These types of configurations are particularly useful when the relevant ends of the first and second fiber optic cable assemblies are coaxially disposed or are at least disposed in
- the end of the second fiber optic cable assembly would be disposed on a first side of this reference plane, displaced therefrom and directed toward the window and at least generally in the direction of the reference plane such that light leaving its end would impact the outer surface of the window at an angle other than perpendicular
- the end of the first fiber optic cable assembly would be disposed on a second side of this same reference plane (opposite the first side), displaced therefrom, and directed toward the window and at least generally in the direction of the reference plane such that at least a portion of the calibration light which is reflected by the inner surface of the window would be "collected" by the first fiber optic cable assembly for provision to the first spectrometer assembly
- the thickness of the window will define at least in part the amount by which that portion of the calibration light which is reflected by the inner surface of the window is offset from that portion of the calibration light which is reflected by the outer surface of the window, and thereby the "sensitivity" to the relative positionings between the ends of the first and second fiber optic cable assemblies to collect only light reflected by the inner surface
- Anti-reflective coatings may also be applied to the outer surface of the window to reduce the effects of that portion of the calibration light which is reflected by the outer surface of the window - that is such that a comparison may be made between the calibration light that is sent to the window with that portion of the calibration light which is reflected by the inner surface of the window
- a window with parallel inner and outer surfaces could be used with an arrangement whereby the ends of the first and second fiber optic cable assemblies were coaxially disposed and oriented such that reference axes projecting from their respective ends impacted both the inner and outer surfaces in at least substantially perpendicular fashion
- Application of an anti-reflective coating to the outer surface of the window would reduce the amount of light which is reflected from the outer surface of the window and directed back to the first fiber optic cable assembly for provision to the first spectrometer assembly in this instance
- Another embodiment of the first aspect of the present invention relates to the use of at least two different types of light by the calibration light source
- One of these calibration lights may include a plurality of discrete intensity peaks, while the other of these lights may be defined by a continuum of intensity or where there are no discernible peaks (e g , a constant intensity, a continually changing intensity, or a combination of both)
- one of these calibration lights may be used to identify one type of condition requiring calibration (e g , a wavelength shift associated with the optical emissions data obtained through the window) while the other may be used for another, different type of condition requiring calibration (e g , an intensity shift associated with the optical emissions data obtained through the window, a complete filtering of some part of the optical emissions transmitted through the window)
- One embodiment of the calibration assembly associated with the second aspect of the present invention calibrates the plasma monitoring assembly for one or more conditions
- One of these conditions is a wavelength shift which may be experienced in relation to the optical emissions data obtained on the subject plasma process
- Another of these conditions is an intensity shift which may be experienced in relation to the optical emissions data obtained on the subject plasma process
- Yet another of these conditions is where certain of the optical emissions, which should be available on the subject plasma process, are being at least substantially completely filtered (e g , blocked out) by the window
- one of these conditions is where the window is having different effects on different portions of the optical emissions This would be the case where there are differing intensity shifts or multiple dampening effects throughout the optical emissions data being obtained on the subject plasma process
- the calibration assembly discussed above in relation to the first aspect of the present invention may be used to identify and calibrate the subject plasma monitoring assembly for any of the above-noted types of conditions in relation to the second aspect of the present invention
- Wavelength shifts may be identified through using a calibration light having a plurality of discrete and displaced (at different wavelengths) intensity peaks Any shifting in the wavelengths at which these peaks appear in the calibration light which is sent to the window (calibration light) in relation to that portion of the calibration light which is reflected by the inner surface of the window (reflected light) would be indicative of a wavelength shift and which could be addressed and more preferably at least substantially alleviated by calibration Intensity shifts may also be identified with this type of light by noting how the intensity of the peaks vary between the calibration light and the reflected light Some peaks in the reflected light may be dampened in relation to the calibration light more than others, which would indicate the existence of multiple dampening effects Peaks which were present in the calibration light but which were absent in the reflected light would indicate that there is filtering
- intensity shifts, complete filtering, and different dampening effects are identified through using a type of calibration light having a continuum of intensity which provides a more complete picture than the case where a calibration light having discrete intensity peaks is used for any of these purposes That is, little or no information is provided on the "behavior" of the window in relation to those wavelengths which are located between the intensity peaks in the calibration light (i e the effect of the window on the intensity of these wavelengths), and therefore assumptions must be made There is no need for such assumptions in the case of using a calibration light with a continuum of intensity for the above-noted purposes
- a third aspect of the present invention is directed to monitoring a plasma process through initializing a plasma monitoring assembly
- the plasma monitoring assembly evaluates at least one aspect of a subject plasma process (e g , one currently being conducted within a processing chamber) by obtaining optical emissions data through a window on the processing chamber
- Optical emissions which are obtained on the subject plasma process include at least wavelengths from about 250 nanometers to about 1 ,000 nanometers which defines a first wavelength range, and at least at every 1 nanometer throughout this first wavelength range
- Initialization of the plasma monitoring assembly in a first embodiment of this third aspect includes directing a calibration light toward the window through which optical emissions are obtained reflecting a first portion of the calibration light from the window, and comparing the original calibration light which was sent with this first portion Consequently, any combination of the various features discussed above in relation to calibration in accordance with the first and second aspects may be implemented in this third aspect as well When the comparison of the calibration light with the first portion of the reflected light yields a first result
- At least one adjustment is made in relation to the plasma monitoring assembly
- Adjustments which may be made in relation to the plasma monitoring assembly in this first embodiment of this third aspect include physical adjustments to the plasma monitoring assembly
- the grating, one or more of its mirrors, or both may be moved (e g , pivoted) to calibrate the plasma monitoring assembly
- Any calibration of the plasma monitoring assembly involving a physical adjustment of the spectrometer assembly in this manner will typically be to address a wavelength shift which is typically due to "drifting" of the spectrometer assembly, although this type of physical adjustment may be used to address wavelength shifts from other sources
- Another type of adjustment which may be made in relation to the plasma monitoring assembly is a calibration of the optical emissions which are collected or obtained on the subject plasma process, or more typically data which is representative of these optical emissions
- the "adjustment" may include the implementation of a single calibration factor or multiple calibration factors in the plasma monitoring assembly A single calibration factor is typically utilized when there is a "uniform"
- Initialization of the plasma monitoring assembly in a second embodiment of this third aspect includes the steps of monitoring the window on the processing chamber through which optical emissions are obtained
- the second embodiment further includes the step of determining if the window is filtering out optical emissions within a first wavelength region which is contained within the first wavelength range of about 250 nanometers to about 1 ,000 nanometers, which again defines the optical emissions being obtained and made available for evaluation by the plasma monitoring assembly
- the various features discussed above in relation to the second aspect of the invention in relation to "filtering" may be included in this second embodiment of the third aspect as well
- the second embodiment of the third aspect includes the step of having the plasma monitoring assembly ignore any optical emissions within any first wavelength region or that reg ⁇ on(s) where filtering has been detected
- Notification may be provided that a filtering condition has been identified Moreover, a recommendation that the window be replaced may be issued in this situation
- the monitoring step of the second embodiment of the third aspect may include the step of directing a calibration light toward the window, reflecting a first portion of this calibration light from the window, and comparing the calibration light with this first portion
- any one or more of the features discussed above in relation to the first and second aspects of the present invention may be utilized by this second embodiment of the third aspect as well
- the second embodiment may also include the step of making at least one adjustment in relation to the plasma monitoring assembly when certain conditions are identified by the above-noted calibration procedure
- any one or more of the features discussed above in relation to the first embodiment of this third aspect may also be utilized by this second embodiment of the third aspect of the present invention
- Initialization of the plasma monitoring assembly in a third embodiment of the above-noted third aspect of the present invention includes the steps of monitoring the window on the processing chamber through which optical emissions on the subject plasma process are obtained
- the third embodiment further includes the step of determining if the window is having a first effect (e g , dampening) on a first wavelength region which is contained within the first wavelength range of about 250 nanometers to about 1 ,000 nanometers (which defines the optical emissions being obtained and made available for evaluation by the plasma monitoring assembly), as well as a second effect (e g dampening) on a second wavelength region which is also contained within the first wavelength range but outside of the first wavelength region associated with the first effect
- the various features discussed above in relation to the second aspect of the invention in relation to identifying different dampening effects may be included in this third embodiment of the third aspect as well
- the third embodiment of the third aspect includes the step of making at least one adjustment in relation to the plasma monitoring assembly if any of these first and second types of effects are identified As such, any one or more of the
- the monitoring step of the third embodiment of the third aspect may include the step of directing a calibration light toward the window reflecting a first portion of this calibration light from the window, and comparing the calibration light with this first portion
- any one or more of the features discussed above in relation to the first and second aspects of the present invention may be utilized by this third embodiment of the third aspect as well
- the third embodiment may also include the step of making at least one adjustment in relation to the plasma monitoring assembly when certain conditions are identified by the above-noted calibration procedure
- any one or more of the features discussed above in relation to the first embodiment of this third aspect may also be utilized by this third embodiment of the third aspect of the present invention
- a fourth aspect of the present invention relates to a method for monitoring a plasma process which includes monitoring a window on the processing chamber in which the plasma process is conducted
- a quantity of product is loaded into the processing chamber (e g , at least one wafer)
- the plasma process is thereafter conducted on this product (e g , a plasma recipe)
- data on the plasma process e g , optical emissions of the plasma in the chamber during the process
- the plasma process is evaluated based upon both the data which is obtained through the processing chamber window and the monitoring of the window
- the monitoring of the window more specifically includes the step of monitoring an actual condition of the window
- the condition of the window in the case of the subject second embodiment is monitored other than through data which is obtained on the plasma process That is, the data which is obtained on the plasma process being conducted within the processing chamber is not utilized in any manner by the step of monitoring the condition of the window in this first embodiment of the fourth aspect of the present invention
- a second embodiment of the subject fourth aspect characterizes the monitoring of the window in a different manner than as discussed above in relation to the first embodiment
- the monitoring step of this second embodiment includes the steps of directing a calibration light toward the window, reflecting a first portion of this calibration light from the inner surface of the window, and comparing the calibration light as it was sent to the window with that portion of the calibration light which was reflected by the inner surface of the window
- one or more of the features presented above in relation to the first and second aspects of the invention may be included in this second embodiment of the fourth aspect as well
- the types of conditions which may be identified through this monitoring of the processing chamber window are presented above in relation to the second aspect of the present invention and any one or more of these features may be included in this second embodiment of the fourth aspect as well
- a fifth aspect of the present invention relates to determining when plasma exists or is "on" within a processing chamber based upon machine-based optical analysis (i e , not by a human eye) More specifically, the fifth aspect relates to obtaining optical emissions from within the processing chamber evaluating these optical emissions, generating a plasma in the processing chamber and identifying when plasma exists within the processing chamber through a machine-based evaluation of the optical emissions from within the processing chamber
- the identification of when plasma exists within the chamber through optical analysis may implement various techniques
- the time at which the plasma comes on within the chamber may be identified by determining when the optical emissions from within the processing chamber exceeds a certain predetermined output (e g , when the intensity of the optical emissions or a certain portion thereof within the chamber exceeds a certain amount)
- the identification of when plasma exists through optical analysis may also be directed toward evaluating how the optical emissions change over time For instance, when no plasma exists within the chamber, there will be no corresponding optical emissions being emitted from the chamber Therefore, the identifying step may simply be directed toward noting any change from a "dark" condition to a "light” condition
- Another way to determine when plasma exists within the chamber through an optical analysis is to determine when the optical emissions from within the chamber include at least a certain number of discrete intensity peaks, each of which has at least a certain intensity
- the presence of plasma within the chamber may be identified by
- Another feature which may be incorporated in the subject fifth aspect relates to the processing of a product after the plasma exists within the chamber
- the window on the chamber may be monitored in accordance with the fourth aspect of the invention discussed above These monitoring operations may be automatically terminated at a time when plasma is first identified within the chamber through the noted optical analysis provided by this fifth aspect
- plasma processes conducted within the chamber may be monitored by a plasma monitoring assembly Calibration of this plasma monitoring assembly may be made available in accordance with the third aspect of the invention discussed above These calibration operations may be automatically terminated when plasma is identified within the chamber through the noted optical analysis provided by this fifth aspect
- a sixth aspect of the present invention relates to a plasma spectra directory which contains at least optical emissions data from plasma processes previously conducted within the processing chamber and which are used to evaluate plasma processes subsequently conducted in this very same processing chamber
- the plasma spectra directory is stored on a computer- readable storage medium and for ease of description includes a first data structure having a plurality of data entries Each of these data entries includes data representative of optical emissions from at least one time during the subject plasma process and this data is associated with one of a first category, a second category, and a third category
- the data entries associated with the first category are those plasma processes which have been run in the chamber and which define a "standard” against which subsequent plasma processes are judged Plasma processes which are run in the processing chamber are evaluated to determine if they "correspond" or "match” with at least one data entry associated with the first category These types of plasma processes associated with the first category thereby may be characterized as "normal" runs In this case, plasma processes which are associated with the first category are at least assumed to have proceeded without substantially any error, and may be tested in some manner to confirm that they did in fact proceed without any substantial error or aberration
- the optical emissions of the plasma in the processing chamber will typically reflect whether a given plasma process is proceeding in a "normal” fashion
- optical emissions associated with a data entry of the first category preferably include at least wavelengths from about 250 nanometers to about 1 ,000 nanometers at least at every 1 nanometer throughout this range and at least at every 1 second from the subject plasma process
- Virtually any type of plasma process may be included in data entries associated with the first category as long as its optical emissions data provides an indication that the plasma process is proceeding in a certain fashion
- One or more plasma recipes (run on production wafers, qualification wafers, or both), plasma cleanings (before or after a wet clean), and conditioning wafer operations may each be included in the plasma spectra directory and associated with the first category
- Multiple "species" of these types of plasma processes may also be included in the plasma spectra directory in association with the first category (e g , different types of plasma recipes)
- Multiple data entries of the same "spec ⁇ es" may also be included in the plasma spectra directory in association with the first category as well (e g , multiple entries of the same plasma recipe run on the same type of product)
- the data entries associated with the second category of the subject sixth aspect are those plasma processes (e g , plasma recipes, plasma cleans, conditioning wafer operations) which have been run in the processing chamber and which have encountered at least one error or aberration
- This error or aberration will typically be represented by a change in the optical emissions of the plasma in the processing chamber, and the cause may be identified by a review of these optical emissions Typically this review is after termination of the subject plasma process
- Obtaining optical emissions data within the above- noted wavelength range increases the likelihood that optical emissions data which is representative of the error or aberration will in fact be available for inclusion in a data entry which is associated with the second category
- the identification or cause of the error or aberration is included in some manner with the data entry which is associated with the second category Various actions may be initiated based upon this information
- An alert or the like (audio, visual, or both) may be issued to apprise personnel that an error has been encountered in the subject plasma process in the subject processing chamber
- Specific information on the error may also be made available as well as one or more ways to address or correct the error or aberration
- corrective actions may be automatically undertaken if desired
- the entire run in which the error occurred is not included in the data entry associated with the second category Instead only those optical emissions which reflect the existence of the subject error or aberration are typically included in such a data entry This may include optical emissions data from only a single point in time during the subject plasma process or from multiple times Optical emissions included in any data entry associated with the second category may also be of the above-noted wavelength range However, if the error or aberration is only reflected in a certain portion of the optical emissions which are obtained on the subject plasma process, only this portion need be included in the plasma spectra directory for the subject data entry associated with the second category
- the data entries associated with the third category in relation to the subject sixth aspect are those plasma processes which have been run in the processing chamber and which are "unknown" to the plasma spectra subdirectory That is, the optical emissions from the subject plasma process have failed to correspond with any data entry associated with the first category or with the second category Moreover the reason as to why this is the case has yet to be determined, or more accurately the cause has yet to be associated with a data entry on the computer-readable storage medium Two situations will typically encompass each case where a data entry is recorded in the plasma spectra directory and associated with the third category Plasma processes which have not yet been recorded in the plasma spectra directory and associated with the first category are one such situation In this case, the entirety of the subject plasma process may be recorded in the plasma spectra directory and associated with the third category Once this data entry is identified as being a new plasma process which did or was assumed to have proceeded without substantially any error or aberration, the data entry may be "transferred" from the third category to the first category Plasma processes which have encountered an error which has not been recorded in the plasma
- a seventh aspect of the present invention relates to various analytical techniques which may be used to evaluate a plasma process in at least some manner
- a computer-readable storage medium includes a plurality of data entries At least one of these data entries is associated with the type of first category discussed above in relation to the sixth aspect, while at least one of these data entries is associated with the type of second category also discussed above in relation to the sixth aspect
- the evaluation technique embodied by this first embodiment of the seventh aspect first determines if the subject plasma process corresponds with any data entry associated with the first category Any such correspondence may be used to characterize the subject plasma process as "normal" or the like If the subject plasma process at any time fails to correspond with at least one data entry under the first category, this first embodiment of the seventh aspect will then "search" those data entries under the second category to see if the subject plasma process has encountered a known error or aberration Therefore, data entries under the second category are not searched in each case
- Various actions may be initiated if the current plasma process corresponds with a data entry associated with the second category, either manually or automatically For instance, the subject plasma process may be terminated, an alert may be issued that an error has been encountered, further use of the processing chamber for processing product may be suspended, adjustment of the plasma process may be undertaken in an attempt to remedy the subject error(s), or any combination thereof
- a second embodiment of the subject seventh aspect utilizes a computer- readable storage medium which includes a first data entry which is associated with a first category of the type identified above in relation to the sixth aspect
- This data entry includes a plurality of first data segments from a plurality of different times during one plasma process previously conducted in the processing chamber
- Each data segment includes optical emissions of the plasma in the chamber for wavelengths of at least about 250 nanometers to about 1 ,000 nanometers which defines a first wavelength range, and at least at every 1 nanometer throughout this first wavelength range
- This second embodiment entails obtaining current optical emissions from another plasma process run in this same processing chamber which are also within the first wavelength range and at least at every 1 nanometer throughout this first wavelength range.
- a comparison is undertaken between the current optical emissions and those associated with at least one first data segment of the first data entry throughout the first wavelength range and at least at every 1 nanometer throughout the first wavelength range.
- a third embodiment of the subject seventh aspect also utilizes a computer-readable storage medium.
- a first plasma process is run in the processing chamber.
- Optical emissions of the plasma in the chamber are obtained for wavelengths of at least about 250 nanometers to about 1 ,000 nanometers which defines a first wavelength range, and at least at every 1 nanometer throughout this first wavelength range.
- This data is obtained at a plurality of times during this first plasma process and such is recorded in a first data entry on the computer-readable storage medium.
- a second plasma process is conducted after termination of the first and similar data is obtained.
- the second plasma process is evaluated based upon at least a portion of the optical emissions data from the second plasma process.
- this may not be practical, desirable, or necessary.
- the progress of the second plasma process in relation to the first plasma process recorded in the first data entry on the computer-readable storage medium may be based upon an evaluation of at least a 50 nanometer bandwidth and at least every 1 nanometer throughout this smaller bandwidth.
- a smaller wavelength region may be selected for evaluating the second plasma process in relation to the first plasma process in a variety of manners.
- the particular wavelength(s) at which error(s) have been previously encountered in running this same plasma process may be used to select that portion of the first wavelength range which should be used in the subject evaluation (e.g., ⁇ 25 nanometers of each wavelength which is indicative of an error or aberration).
- a wavelength region may be selected which includes each of the errors previously encountered on the same type of plasma process.
- the "width" of the region may be defined by the two extreme wavelengths, although it would be preferred to include a "buffer" of sorts on each of these ends (e.g., expand the range by 25 nanometers on each end).
- the particular wavelength(s) which is indicative of the endpoint of the subject plasma process or discrete/discernible portion thereof may be used to select that portion of the first wavelength range which should be used in the subject evaluation ( ⁇ 25 nanometers of each such wavelength).
- Individual endpoint indicator wavelengths are discussed in more detail below in relation to the ninth aspect of the present invention.
- An eighth aspect of the present invention relates to identifying the type of plasma process conducted within the processing chamber. This aspect may be used to identify whether a plasma process is a certain type of plasma recipe being run on a certain type of production wafer, a certain type of plasma recipe being run on a certain type of qualification wafer, a certain type of plasma recipe being run on a certain type of conditioning wafer, or a plasma clean being run in a chamber.
- a first embodiment of this eighth aspect is able to identify the particular type of a plasma recipe being run on product (e.g., production wafer, qualification wafer) in a processing chamber based upon the storage of at least two plasma recipes on a computer-readable storage medium.
- the computer-readable storage medium includes a plurality of data entries.
- a first of these data entries includes relevant data from a plurality of times during a first plasma recipe run on product in the processing chamber (and preferably of the entirety of this first plasma recipe at least after stabilization of the plasma).
- a second of these data entries includes relevant data from a plurality of times during a second plasma recipe (different from the first plasma recipe) run on product in the same processing chamber (and preferably of the entirety of this second plasma recipe at least after stabilization of the plasma).
- Data on a subject plasma recipe which is being run on product in the same processing chamber is obtained. This data is used to determine if the current plasma recipe is of the same type as the first or second plasma recipe stored on the computer- readable storage medium.
- this determination is completed prior to termination of the current plasma recipe and at least before the next product is loaded into the chamber
- This first embodiment of the eighth aspect may be used to determine not only the identity of the subject plasma process, but the type of product (e g , whether a production wafer or a qualification wafer) that is being processed by including relevant data from prior plasma processes on the computer-readable storage medium That is, by including a plasma recipe "A" run on a certain type of production wafer in one data entry and the same plasma recipe "A" on a certain type of qualification wafer in another data entry, the ability exists to determine if the current plasma recipe is being run on a production versus a qualification wafer
- the data obtained on the current plasma process may be optical emissions of the plasma in the processing chamber
- These optical emissions may include at least wavelengths from about 250 nanometers to about 1 ,000 nanometers (inclusive) which defines a first wavelength range, and optical emissions may be obtained at least at every 1 nanometer throughout this first wavelength range
- Optical emissions of the subject plasma process may be compared with one or both of the first and second plasma recipes stored on the computer-readable storage medium to see if there is sufficient correspondence therebetween
- the techniques discussed above in relation to the seventh aspect may be implemented in this first embodiment of the eighth aspect as well
- a second embodiment of the subject eighth aspect is directed toward inputting the plasma recipe to be run in the chamber and using the principles discussed above in relation to the first embodiment of the eighth aspect to verify that no errors were made when inputting the subject plasma recipe That is, the identify of the subject plasma process is determined in accordance
- a third embodiment of the subject eighth aspect is directed to identifying a subject plasma recipe based upon at least two plasma recipes which are stored on a computer-readable storage medium and which were previously run in the same processing chamber
- the first execution of the subject plasma recipe is initiated and is of the type associated with either the first or second plasma recipe
- At least one characteristic of the plasma is monitored during the execution of each subject plasma recipe
- Both the first and second plasma recipes are available for comparison against the first execution of the subject plasma recipe
- subsequent executions of the subject plasma recipes are evaluated at least initially only in relation to the identified plasma recipe on the computer-readable storage medium
- This embodiment is particularly pertinent to the case where the first wafer of a cassette or boat of wafers is evaluated in accordance with the foregoing since the same plasma recipe is typically run on the entire cassette Therefore, once the third embodiment of the eighth aspect determines the identify of the plasma recipe being run on the first wafer, all subsequent wafers in the cassette are at least initially evaluated against only
- a ninth aspect of the present invention relates to engaging in research to identify one or more indicators of a first endpoint which is when the plasma process (e g , plasma recipe, plasma clean, conditioning wafer operation) or portion thereof (e g , plasma step of a plasma recipe) has achieved a first predetermined result (e g , the etching away of a certain layer from a multi-layer structure such as a wafer)
- a first plasma process is run in the processing chamber
- Optical emissions of the plasma are obtained at a plurality of times during this first plasma process
- These optical emissions include at least wavelengths from about 250 nanometers to about 1 ,000 nanometers (inclusive) which defines a first wavelength range
- Optical emissions are preferably obtained at least at every 1 nanometer throughout this first wavelength range
- These optical emissions are evaluated or analyzed and at least one endpoint indicator is selected based upon this analysis
- the subject analysis may include generating a plot of intensity versus time for a plurality of individual wavelengths which are within the first wavelength range
- plots are generated for each wavelength which is available based upon the data collecting resolution of the relevant "collecting" structure (e g , spectrometer(s))
- These plots are analyzed after the conclusion of the running of the plasma process, preferably in view of information as to about what time the first endpoint should have occurred (e g , calculated based upon knowledge of process conditions and thickness of layer to be etched away)
- Any wavelength having a plot with a distinctive change in intensity around that time where the first endpoint should have occurred may be identified as a possible endpoint indicator candidate
- a comparison of the plots between two or more runs may identify a pattern which stays the same, but which undergoes some type of change
- This change may be a temporal shift, a shift in the intensity associated with the pattern, a uniform enlargement of the pattern, a uniform reduction in the pattern, or any combination thereof
- Patterns which undergo this type of change are an indicator that the corresponding wavelength is in fact indicative of the first endpoint
- One "controlled" way of inducing such a shift is to process two or more products having different thicknesses
- the analysis used to select at least one indicator of the first endpoint may also include examining the optical emissions to identify the existence of intensity peaks, and determining if any of these
- a tenth aspect of the present invention relates to monitoring at least two aspects of a plasma process, one of which may be the "health" of the plasma process and another of which may be at least one endpoint associated with the plasma process
- This tenth aspect is applicable to any plasma process, including plasma recipes which are run on product (e g , production wafers, qualification wafers) in a processing chamber, plasma cleanings (e g , with or without a wet clean), and conditioning wafer operations
- substantially the entirety of the plasma process may be evaluated in relation to its "health" except possibly the initial portion of the plasma process where the plasma is typically unstable
- the evaluation of the plasma process in relation to identifying an endpoint need not be initiated until closer to the time at which the subject endpoint should be reached
- the frequency at which the plasma health is evaluated need not be the same as the frequency at which the evaluation is undertaken to identify the subject endpoint
- the plasma health may be assessed less frequently than the evaluation relating to identifying the subject endpoint
- a plasma process is conducted within the processing chamber and at least a first endpoint is associated with this plasma process
- the plasma process is monitored to identify the occurrence of the first endpoint
- Any endpoint detection technique may be used for this first embodiment of the tenth aspect, including those addressed below in relation to the eleventh through the thirteenth aspects of the present invention
- the "condition" of the plasma is also evaluated during, and more preferably throughout the entirety of, the plasma process (again excluding possibly the initial portion when the plasma is typically unstable)
- One way of defining the "condition" associated with this first embodiment of the tenth aspect is equating the same with the cumulative result of all parameters having an effect on the plasma in the processing chamber This may be done by evaluating optical emissions of the plasma in the chamber which includes at least wavelengths from about 250 nanometers to about 1 ,000 nanometers which defines a first wavelength range, and at least at every 1 nanometer throughout this first wavelength range and at least at a plurality of different times during the subject plasma process
- a second embodiment of this tenth aspect involves generating a plasma in the processing chamber and running a first plasma step in the chamber Associated with this first plasma step is a first endpoint which is when the first plasma step has produced a first predetermined result At least one characteristic of the plasma in the chamber is evaluated during the first plasma step using a first time resolution. Although typically equal increments will be utilized in this evaluation, such is not required by this second embodiment of the tenth aspect. An evaluation is also undertaken to identify an occurrence of the first endpoint using a second time resolution which is different than the first.
- the above-noted "at least one characteristic" may be the condition of the plasma during the subject plasma process, although it need not be the case.
- An eleventh aspect of the present invention generally relates to monitoring a plasma process to identify an occurrence of a first endpoint associated with the plasma process. More specifically, at least two different techniques are used to evaluate the current plasma process to identify the first endpoint in this eleventh aspect. Endpoint may be called when only one of these techniques identifies the occurrence of the first endpoint, or may be called after each of these techniques identifies the occurrence of the first endpoint.
- This eleventh aspect of the present invention is applicable to any plasma process having at least one endpoint associated therewith (e.g., plasma recipes which are run on product in a processing chamber, plasma cleanings, and conditioning wafer operations).
- One of the techniques which may be used in the subject eleventh aspect involves a comparison of the current optical emissions of the plasma in the chamber with optical emissions of the plasma in the chamber from a previous time in the same process, preferably the immediately preceding time at which optical emissions were obtained.
- these optical emissions include at least wavelengths from about 250 nanometers to about 1 ,000 nanometers at least at about every 1 nanometer.
- endpoint may be deemed to have been reached. Stated another way, when there is no longer any substantial change in the optical emissions, endpoint may be deemed to have been reached.
- Another technique for identifying endpoint which may be used in the subject eleventh aspect involves a comparison of the current optical emissions of the plasma in the chamber with a standard.
- This "standard” may be optical emissions of the plasma in the chamber from a previous execution of this same plasma process in the same processing chamber at a time when endpoint was at least deemed to have been reached Moreover, this standard may be stored on a computer-readable storage medium In one embodiment, these optical emissions include at least wavelengths from about 250 nanometers to about
- Yet another technique which may be used in the subject eleventh aspect of the present invention includes determining if there is at least a first change in the impedance of the processing chamber which is reflected in the optical emissions of the plasma in the processing chamber
- a "modal" change in the plasma may be indicative of a change in impedance which in turn is indicative of endpoint
- This "modal" change may be a rather sudden and significant increase or decrease in the intensity of the entirety of the plasma or of a particular wavelength(s)
- Another technique which may be used to identify endpoint in relation to the subject eleventh aspect includes evaluating at least one individual wavelength of light forming the plasma of the subject plasma process This one wavelength of light may be evaluated to determine when a plot of intensity versus time deviates by more than a predetermined amount from a predetermined equation (e g , when there is no longer a "fit" between the current data and the subject equation) Therefore, the features discussed above in relation to the ninth aspect of the present invention are also relevant to this portion of the eleventh aspect as well Moreover, any one or more individual wavelengths of light may be evaluated to determine when the change in slope over time of the wavelength(s) changes by more than a predetermined amount Second order derivatives may be used as well
- a twelfth aspect of the present invention is directed toward a technique for identifying the occurrence of a first endpoint associated with a plasma process (e g , plasma recipe, plasma clean, conditioning wafer operation) or a discrete/discernible portion thereof (e g , a plasma step of a multiple step recipe or process)
- Optical emissions of the plasma in the chamber from the process are obtained These optical emissions include at least wavelengths from about 250 nanometers to about 1 ,000 nanometers which defines a first wavelength range
- the data resolution which is used in collecting the optical emissions is no more than about 1 nanometer This means that optical emissions are obtained at least at every 1 nanometer throughout the first wavelength range
- Identification of the first endpoint involves a comparison of the most current optical emissions of the plasma in the chamber with a first output
- This first output may be optical emissions of the plasma in the chamber from a previous time in the same plasma process, preferably the immediately preceding time at which the optical emissions were obtained in relation to the now current optical emissions
- This first output may also
- Confidence in the calling of the first endpoint by the above-noted technique may be enhanced by using a second technique and not calling the first endpoint until both of the techniques have "seen” the first endpoint
- a thirteenth aspect of the present invention relates to identifying the occurrence of multiple endpoints in a single plasma process
- Many plasma recipes will include a number of different plasma steps Each of these plasma steps typically has an identifiable endpoint associated therewith Therefore, the eleventh aspect of the present invention allows for the identification of at least two of these endpoints and including each endpoint associated with the subject plasma process
- Each of the techniques identified in eleventh aspect discussed above may be utilized in this thirteenth aspect
- Appropriate actions include terminating the current plasma process, issuing an alert, suspending execution of any further plasma processes in the chamber until it is appropriately cleaned, or any combination thereof
- Appropriate actions include terminating the current plasma process, issuing an alert, suspending execution of any further plasma processes in the chamber until it is appropriately cleaned, or any combination thereof
- Various features may be utilized by the fourteenth aspect of the present invention, and these features may be used alone in relation to this fourteenth aspect as well as in any combination
- the data which is obtained on the current plasma process may be optical emissions of the plasma in the chamber Wavelengths obtained may include at least from about 250 nanometers to about 1 ,000 nanometers which defines a first wavelength range Data may be obtained at least at every 1 nanometer throughout the noted first wavelength range
- Plasma cleaning operations are embodied within a fifteenth aspect of the present invention
- a plasma clean removes materials from the interior of the processing chamber by having plasma exist within an "empty" chamber
- No product e g , wafers
- Optical emissions of the plasma in the "empty" chamber are obtained at a plurality of times during the process in a first embodiment of this fifteenth aspect
- a pattern of at least a portion of the optical emissions is compared with a first standard pattern during at least a portion of the process (e g , a plot of intensity versus time)
- This first standard pattern may be recorded on a computer-readable storage medium
- this first standard pattern may be from optical emissions of plasma in a plasma process previously conducted within the same processing chamber at a time when the plasma clean had reached its endpoint
- the plasma process is terminated
- Predetermined amount contemplates using pattern recognition techniques, as well as taking a differential and noting when this differential is at least substantially free from any substantial intensity peaks
- the first embodiment of the fifteenth aspect of the present invention includes comparing the pattern of a specific wavelength(s) within the optical emissions of the plasma with the first standard pattern which will include the corresponding wavelength(s) Moreover, the first embodiment also includes comparing the pattern of the entirety of the optical emissions obtained on the current plasma process with the first standard pattern
- the first standard pattern may be part of a first standard optical emissions segment which includes a plurality of wavelengths
- the intensity peak associated with the subject wavelength of the first standard pattern may be identified in relation to its intensity (e g , it is the "largest" intensity peak around a certain wavelength), one or more other intensity peaks (e g , the subject wavelength is the "middle” peak in a certain wavelength region), or both Noting these characteristics of the wavelength for the first standard pattern in the first standard optical emissions segment may then be used to identify the subject wavelength in the current optical emissions segment
- the first embodiment of the subject fifteenth aspect may also be terminated if the subject plasma process has reached a predetermined maximum time limit before the current pattern and first standard pattern are within a predetermined amount of each other Typically
- the differential between the optical emissions at a current time in the process and the optical emissions from a previous time in the same plasma process is determined
- this differential is no more than a first amount, the current plasma process is terminated Therefore, this second embodiment equates the time at which the plasma clean should be terminated with a situation where the current plasma process is no longer changing the condition of the interior of the processing chamber at a desired rate All or a portion of those
- Conditioning wafer operations are addressed in a sixteenth aspect of the present invention
- At least one conditioning wafer is loaded in a processing chamber and a plasma process is run thereon
- the plasma process will etch a pattern on the conditioning wafer which is something other than an integrated circuit or a pattern which would not be associated with a semiconductor device
- Plasma processing of the conditioning wafer is monitored through obtaining optical emissions of the plasma in the chamber
- a number of conditioning wafers are processed in this manner until the conditioning wafer operation is terminated based upon the results of the monitoring of one of the plasma processes conducted on a conditioning wafer
- a production wafer operation is initiated whereby at least one production wafer is loaded in the chamber and a plasma recipe (e g , one or more plasma steps) is run thereon
- a plasma recipe e g , one or more plasma steps
- Termination of the conditioning wafer operation may also be based upon when consecutive runnings of the plasma process on conditioning wafers are within a certain amount of each other as determined through the data obtained on the process That is, the termination of the conditioning wafer operation may be equated with the conditioning wafer operation having reached a steady state (e g , the processing of one conditioning wafer looks at least effectively the same as the processing of the next conditioning wafer) determined in accordance with an evaluation of optical emissions data Termination of the conditioning wafer operation may also be based solely on the data obtained on the conditioning wafer operation That is, no wafer need be analyzed before the production wafer operation is initiated
- One or more of a plasma cleaning operation, a wet cleaning operation, or a replacement of consumables may also be initiated before the initiation of the conditioning wafer operation as well
- At least two chambers are involved in the plasma processing of wafers disposed therein Each plasma process conducted within these chambers is monitored in at least some respect Wafers will continue to be sequentially processed in these chambers unless the monitoring of the current plasma process on the wafer(s) in one of these chambers detects the existence of one or more conditions
- These conditions include the existence of a "dirty chamber", a known error condition, an unknown condition, or a combination thereof as these terms have been used in relation to the sixth and fourteenth aspects discussed above
- the distribution of wafers to this particular chamber may be suspended immediately after this type of condition is identified, or suspension may be delayed until a certain number of these types of conditions are encountered in multiple plasma processes That is, a given chamber may not be taken "off line” until this same condition (or another of the conditions) have been identified in multiple runs
- the chamber may be cleaned in some manner Plasma cleans, wet cleans, replacement of consumables, or any combination thereof are contemplated as an appropriate "cleaning' in the context of this first embodiment of the seventeenth aspect
- the distribution of wafers for running plasma processes thereon may be reinitiated Encountering a known error during the plasma processing of a wafer(s) in one of the chambers may result in the modification of one or more process control parameters to address this error
- the first embodiment contemplates analyzing the plasma process after termination thereof in an attempt to identify the corresponding cause
- a second embodiment of the seventeenth aspect relates to the running of plasma processes on product in at least three chambers
- the wafers are distributed to these chambers using a first sequence Modification of this sequence is initiated if the monitoring of the plasma process in one of the chambers identifies the existence of a certain condition Any of those identified above in relation to the first embodiment would be applicable to this second embodiment as well In this regard the corresponding features from the first embodiment may be implemented in this second embodiment as well
- a third embodiment of this seventeenth aspect involves the distribution of wafers to at least two processing chambers for the running of a plasma process thereon The time required to complete each plasma process is monitored The distribution sequence which is utilized is based upon this monitoring of time For instance, the distribution sequence may involve maximizing the use of the "fastest" processing chamber
- An eighteenth aspect of the present invention relates to a virtual optical filter of sorts for use in monitoring plasma processing operations
- Optical emissions data throughout a first wavelength region e g , a range of wavelengths extending from a first wavelength to a second wavelength, the distance between which defines a bandwidth
- a second wavelength region is selected for monitoring at least one aspect of the first plasma process
- This second wavelength region is a subset of the first wavelength region (i e has a smaller bandwidth) That is the second wavelength region is totally contained within, but is smaller than, the first wavelength region Therefore, only a portion of the optical emissions data which is being collected on a given plasma process is used to evaluate this process in at least some manner in accordance with the subject eighteenth aspect
- Monitoring one portion of a plasma process may require optical emissions data within one wavelength region or at one specific wavelength, while monitoring another portion of the same plasma process may require optical emissions data within a different wavelength region or at a different wavelength Similarly, monitoring one type of plasma process may require optical emissions data within a certain wavelength region or at a certain wavelength while monitoring a different type of plasma process may require optical emissions data within a different wavelength region or at a different wavelength
- a significant benefit of the subject eighteenth aspect is that so long as the desired optical emissions for monitoring a given plasma process or a portion thereof are within the first wavelength region of the optical emissions data which is being collected no physical adjustments will be needed to accommodate any of these scenarios
- the eighteenth aspect avoids the situation where one bandpass filter is required to monitor one type of plasma process and where another bandpass filter is required to monitor another type of plasma process
- the optical emissions data over the first wavelength region may be stored on a computer-readable storage medium in a database or otherwise cataloged such that a plasma monitor (e g , plasma health module, endpoint detection module) may seiectably retrieve which particular subset of the first wavelength region (e g , a specific wavelength or a wavelength region) is desired for use by the plasma monitor That is, each wavelength may be assigned some type of identifier such that all that is required to retrieve the optical emissions data on this wavelength is to input the corresponding identifier to the requisite plasma monitoring module In order to retrieve optical emissions data on a plasma monitor (e g , plasma health module, endpoint detection module) may seiectably retrieve which particular subset of the first wavelength region (e g , a specific wavelength or a wavelength region) is desired for use by the plasma monitor That is, each wavelength may be assigned some type of
- One aspect of any plasma process which may be monitored through the subject eighteenth aspect is to identify an occurrence of at least one endpoint which is associated with a plasma process being run in a processing chamber "Endpoint" is defined as when the plasma process has realized or affected a certain, predetermined result (e g , the removal of a certain layer)
- the occurrence of an endpoint may be monitored by monitoring one or more individual wavelengths which are each contained in the first wavelength region, by monitoring one or more wavelength regions which are each contained within the first wavelength region but which need not have the same bandwidth, or any combination thereof
- Another aspect of a plasma process which may be monitored through the eighteenth aspect is to determine if a plasma process currently being run in a processing chamber is proceeding in accordance with at least one plasma process which was previously conducted within the very same processing chamber Although this may be done by comparing optical emissions data over the entire first wavelength region, this may be done through the eighteenth aspect and its use of a smaller wavelength region for monitoring purposes
- the second wavelength region may have a bandwidth of at least
- wavelengths or wavelength regions may be used to monitor the plasma process in accordance with the subject eighteenth aspect, and again without making any physical change to the plasma monitoring system so long as each of these wavelengths or wavelength regions are in the first wavelength region where optical emissions data is being collected on the current plasma process
- one wavelength region within the first wavelength region may be monitored for a first endpoint of a first step of a certain plasma process
- another/different wavelength region within the first wavelength region may be monitored for a second endpoint of a second step of the same plasma process This may be done for each step of the plasma process
- Another possibility is to monitor for at least one endpoint associated with the current plasma process and at the same time to also monitor the health of the plasma process by comparing optical emissions data from a wavelength region having at least a 50 nanometer bandwidth, with optical emissions data of this same wavelength region from a plasma process previously conducted in the same processing chamber
- a nineteenth aspect of the present invention relates to identifying a wavelength region which would be appropriate for monitoring a plasma process for the occurrence of a certain endpoint of the plasma
- the nineteenth aspect will be described in relation to selecting a first wavelength region which has a first bandwidth, and which may be used to identify a first endpoint associated with a first plasma process
- Optical emissions data throughout a second wavelength region having a second bandwidth are obtained on the first plasma process
- the second wavelength region includes at least wavelengths within the range from about 250 nanometers to about 1 ,000 nanometers and at least at every 1 nanometer throughout this wavelength range
- these optical emissions are obtained on the plasma process at least at every 1 second during most if not all of the first plasma process
- a third wavelength bandwidth is selected which is less than the second bandwidth of the second wavelength region which again defines the particular optical emissions data which are to be collected on the first plasma process
- Plots are generated for a plurality of wavelength regions which have the third wavelength bandwidth and which are subsets of the second wavelength region over which optical emissions data are being collected on the first plasma process
- Each plot will then be specific to one of these wavelength region subsets, and will illustrate the change in area of the subject wavelength region over time
- the area of the subject wavelength region is reflective of or related to the intensity of the optical emissions in this particular wavelength region
- at least wavelength region may be selected as the first wavelength region for identifying the occurrence of the first endpoint during the first plasma process
- Optical emissions data are collected on the first plasma process over a second wavelength which includes at least wavelengths from about 250 nanometers to about 1 ,000 nanometers
- the third bandwidth may be selected as 5 nanometers
- Each wavelength region having this 5 nanometer bandwidth and plotted for purposes of identifying an appropriate endpoint indicator wavelength region may be referred to as an endpoint evaluation wavelength region
- the number and relationship between the various endpoint evaluation wavelength regions may be selected to cover at least most, and more preferably the entirety of, the second wavelength region over which optical emissions data are being collected on the first plasma process
- the endpoint evaluation wavelength regions may be disposed in overlapping relation or may be disposed in end-to-end fashion
- the first endpoint evaluation wavelength region may be from 250-
- the second endpoint evaluation wavelength region may be from 255-260 nanometers
- the third endpoint evaluation wavelength region may be from 260-265 nanometers, and so forth up to the 995-1 ,000 endpoint evaluation wavelength region in the subject example
- the first wavelength region for calling the first endpoint may be selected as the first wavelength region for calling the first endpoint.
- the first wavelength region for calling the first endpoint may be selected to encompass each of these endpoint evaluation wavelength regions For instance, if the plot of the 275-280 wavelength region, the plot of the 285-290 wavelength region, and the plot of the 300-305 wavelength region had the requisite identifiable or significant event, the first wavelength region could be defined as the 275-305 nanometer wavelength region
- no two adjacentmost endpoint evaluation wavelength regions are combined to define a particular first wavelength region if there is more than about a 15 nano
- the twentieth aspect may be used to conduct various experiments of sorts as to how making a certain change or combination of changes in the plasma monitoring system a particular plasma monitoring technique, or both would effect the results achieved by or the performance associated with the monitoring of plasma processes
- the subject "monitoring” may be for plasma health, endpoint, or both
- the twentieth aspect is implemented on a remote system which is interconnected with the plasma monitoring system in at least some way and which replicates or mimics at least portions of the plasma monitoring system so that these experiments of sorts may be conducted "off-line” so as to not affect production
- a twenty-first aspect of the present invention is a plasma monitoring network of sorts
- a first embodiment of this twenty-first aspect is a plasma processing system which includes a plurality of chamber clusters Each chamber cluster includes at these one plasma processing chamber and at least one plasma monitoring system which is interconnected with at least one of the processing chambers of the particular chamber cluster There may be a single plasma monitoring system for the entire chamber cluster, each chamber may have its own plasma monitoring system, or a single plasma monitoring system may service a plurality of processing chambers within a given chamber cluster but not all of such processing chambers
- the first embodiment of the twenty-first aspect further includes a clean room system
- One or more clean rooms may define this clean room system
- Each chamber cluster is contained within the clean room system
- a plurality of chamber clusters will be located in the same clean room
- the first embodiment of this twenty-first aspect covers a situation where one or more chamber clusters are located in a plurality of different clean rooms
- a master remote station is operatively interconnected with the plasma monitoring system(s) of each chamber cluster
- This master remote station is disposed outside of the clean room system and includes a display (e g , computer monitor) and data entry device (e g , keyboard)
- each chamber cluster remote station is disposed outside of the clean room system and includes a display (e g , computer monitor) and data entry device (e g , keyboard)
- a given chamber cluster remote station is operatively interconnected only with the plasma monitoring system(s) of its associated chamber cluster, and thereby not the plasma monitoring system(s) of any other chamber cluster
- the master remote station may have greater access rights to a given plasma monitoring system of a given chamber cluster than the chamber cluster remote station for this same given chamber cluster
- multiple modules may be associated with each of the plasma monitoring systems
- the master remote station may have access to a greater number of modules on a per plasma monitoring system basis than the corresponding chamber cluster remote station (i e , the chamber cluster remote station which interfaces with the plasma monitoring system(s) of a single chamber cluster)
- Examples of the above-noted modules include a data player module, a statistical analyzer module, a control module, and a data review module
- the data player module may have the characteristics discussed above in relation to the twentieth aspect of the present invention
- the statistical analyzer module may be configured to undertake various types of statistical analysis For instance the performance of a single processing chamber could be statistically analyzed against itself (e g , performance variations over time) Moreover the performance of one processing chamber could
- the data review module graphically illustrates the current plasma process being run on a certain processing chamber in at least some manner
- the data review module may be configured to simultaneously display multiple plasma processes being run on multiple chambers Representative examples of how the plasma process may be reviewed through the data review module would be to display the plot of change in area over time for the particular wavelength region being used to call a certain endpoint, to display a plot of a particular wavelength for purposes of calling endpoint, or to display one or more plots relating to how the current plasma process in proceeding in relation to a previous execution of this same plasma process
- the present invention will now be described in relation to the accompanying drawings which assist in illustrating its various pertinent features
- One application of the present invention is for processes which utilize plasma to provide at least one function or to achieve at least one predetermined result, and the present invention will hereafter be described in this context More specifically, the present invention will be described in relation to the running of plasma processes on wafers or the like from which semiconductor devices are formed (e g , etching where the "predetermined result” may be the removal of one or more layers, chemical vapor deposition where the predetermined result may be the buildup of one or more films, sputtering where the predetermined result may be the addition or removal of material)
- Figure 1 is a schematic view of a wafer production system
- Figure 2 is a perspective view of one embodiment of the wafer cassette incorporated in the wafer production system of Figure 1
- Figures 3A-B are top and side views, respectively, of one embodiment of the wafer handling assembly incorporated in the wafer production system of Figure 1 ,
- Figure 4 is a cross-sectional view of one embodiment of a plasma processing chamber which may be incorporated in the wafer production system of Figure 1 namely a dry etching chamber,
- Figure 5 is a schematic view of one embodiment of a gas delivery system for the processing chamber of Figure 4
- Figure 6 is a schematic view of one embodiment of a plasma monitoring assembly which may be incorporated in the wafer production system of Figure
- FIG. 7 is a flowchart of one embodiment of the plasma monitoring module used by the plasma monitoring assembly of Figure 6,
- Figure 8 is a spectral pattern of one embodiment of a plasma recipe which may be run on the system of Figure 1 ,
- FIG. 9 is a flowchart of one embodiment of a plasma spectra directory and its various subdirectories which may be used in plasma monitoring operations,
- Figure 10 is a flowchart of one embodiment of a general data management structure which may be utilized for the various subdirectories of the plasma spectra directory of Figure 9,
- Figure 11 is a flowchart of one embodiment of how data within the general data management structure of Figure 10 may be condensed/consolidated
- Figure 12A is one embodiment of a data management structure which may be used for the normal spectra subdirectory of Figure 9,
- Figure 12B is one embodiment of a data management structure which may be used for the abnormal spectra and unknown spectra subdirectories of Figure 9,
- Figure 13 is a flowchart of one embodiment of a pattern recognition module which may be used by the current plasma process module of Figures 7 and 32 in the evaluation of a plasma process being run in the processing chamber of Figure 1
- Figure 14 is a flowchart of one embodiment of a process alert module which may be used by the current plasma process module of Figures 7 and 32 in the evaluation of a plasma process being run in the processing chamber of Figure 1
- a pattern recognition module which may be used by the current plasma process module of Figures 7 and 32 in the evaluation of a plasma process being run in the processing chamber of Figure 1
- Figure 14 is a flowchart of one embodiment of a process alert module which may be used by the current plasma process module of Figures 7 and 32 in the evaluation of a plasma process being run in the processing chamber of Figure 1
- Figure 15 is a flowchart of one embodiment of a startup module to access the current plasma process module of Figures 7 and 32 for the evaluation of a plasma process being run in the processing chamber of Figure 1
- Figure 16 is a flowchart of one embodiment of a startup subroutine which may be accessed by the startup module of Figure 15,
- Figures 17A-C are exemplary spectra of one type of plasma process that may be run in any of the processing chambers of Figure 1 and monitored by the current plasma process module, namely a three-step plasma recipe,
- Figures 18A-C are exemplary spectra of another type of plasma process that may be run in any of the processing chambers of Figure 1 and monitored by the current plasma process module, namely a plasma cleaning operation without first wet cleaning the chamber at the start, at an intermediate time, and end of such a plasma cleaning operation, respectively,
- Figures 19A-C are exemplary spectra of another type of plasma process that may be run in any of the processing chambers of Figure 1 and monitored by the current plasma process module, namely a plasma cleaning operation conducting after a wet clean of the chamber at the start, an intermediate time, and end of such a plasma conditioning operation, respectively,
- Figures 20A-C are exemplary spectra of another type of plasma process that may be run in any of the processing chambers of Figure 1 and monitored by the current plasma process module, namely a conditioning wafer operation at the start, an intermediate time, and end of such a conditioning wafer operation, respectively,
- FIG 21 is a flowchart of one embodiment of a plasma health subroutine which may be used by the plasma health module of Figures 7 and 32,
- Figure 22 is a flowchart of another embodiment of a plasma health subroutine which may be used by the plasma health module of Figures 7 and 32
- Figure 23 is a flowchart of one embodiment of a plasma health/process recognition subroutine which may be used by the plasma health module of
- Figure 24 is a flowchart of another embodiment of a plasma health/process recognition subroutine which may be used by the plasma health module of Figures 7 and 32
- Figure 25 is a flowchart of one embodiment of a plasma health/process step recognition subroutine which may be used by the plasma health module of Figures 7 and 32
- Figures 26A-C are exemplary spectra from a "clean” processing chamber, from an “aging” processing chamber and from a “dirty” processing chamber, respectively,
- Figure 27 is a flowchart of one embodiment of a chamber condition subroutine which may be incorporated in the chamber condition module of Figures 7 and 32
- Figure 28 is a flowchart of another embodiment of a chamber condition subroutine which may be incorporated in the chamber condition module of Figures 7 and 32
- Figure 29 is a flowchart of another embodiment of a chamber condition subroutine which may be incorporated in the chamber condition module of Figures 7 and 32,
- Figures 30A-D are exemplary spectra from the processing chamber in a "dirty chamber" condition, at the end of a wet clean, at the end of a plasma clean, and at the end of a conditioning wafer operation, respectively,
- FIG 31 is a schematic view of another embodiment of a plasma monitoring assembly which may be incorporated in the wafer production system of Figure 1 , which includes the current plasma process module from Figure 7 above, and which also includes a calibration module,
- FIG 32 is a flowchart of one embodiment of a plasma monitoring module which may be used by the plasma monitoring assembly of Figure 31 , as well as by the plasma monitoring assembly of Figure 37 below,
- Figure 33 is a schematic view of one embodiment of the spectrometer assembly of Figure 31 .
- Figure 34 is one embodiment of the fiber optic cable assembly which operatively interfaces the window and plasma monitoring assembly in Figure 31
- Figure 35 is a schematic view of the axes of the calibration light sent and reflected by the inner and outer surfaces of the window of Figure 31
- Figure 36 is one embodiment a fixture assembly for interconnecting the fiber optic cable assembly of Figure 34 with the window on the processing chamber presented in Figure 31 ,
- FIG 37 is a schematic view of another embodiment of a plasma monitoring assembly which may be incorporated in the wafer production system of Figure 1 , which includes the current plasma process module from Figure 7 above, and which also includes a calibration module,
- Figure 38 is a schematic view of the axes of the calibration light sent and reflected by the inner and outer surfaces of the window of Figure 37
- Figure 39 is one embodiment a fixture assembly for interconnecting the fiber optic cables with the window on the processing chamber in the configuration presented in Figure 37,
- Figure 40 is a flowchart of one embodiment of the calibration module from Figure 32
- Figure 41 is a flowchart of one embodiment of a calibration subroutine which may be used by the calibration module of Figure 40,
- Figure 42 is one embodiment of a spectra of a calibration light which may be used by the calibration module of Figure 40,
- Figure 43 is one embodiment of a spectra of that portion of the calibration light of Figure 42 which is reflected by the inner surface of the processing chamber window when there are at least substantially no deposits formed thereon,
- Figure 44 is a cutaway view of another embodiment of a spectrometer which may be used by the spectrometer assembly of Figure 31 and which is operatively interfaced with the calibration module of Figure 40,
- Figure 45 is a flowchart of another embodiment of a calibration subroutine which may be used by the calibration module of Figure 40,
- Figure 46A is one embodiment of a spectra of a calibration light which may be used by the calibration module of Figure 40 to identify an intensity shift condition
- Figure 46B is another embodiment of a spectra of a calibration light which may be used by the calibration module of Figure 40 to identify an intensity shift condition
- Figure 47A is one embodiment of a spectra of that portion of the calibration light of Figure 46A which is reflected by the inner surface of the processing chamber window when in a degraded or aged condition,
- Figure 47B is one embodiment of a spectra of that portion of the calibration light of Figure 46B which is reflected by the inner surface of the processing chamber window when in a degraded or aged condition
- Figure 48 is a flowchart of another embodiment of a calibration subroutine which may be used by the calibration module of Figure 40,
- Figure 49 is a flowchart of one embodiment of a research subroutine which may be used by the research module of Figures 7 and 32,
- Figures 50A-C are exemplary plots of intensity versus time for 3 wavelengths generated by the research subroutine of Figure 49 from one running of a plasma recipe on product in one of the chambers from Figure 1 ,
- Figures 51A-C are exemplary plots of intensity versus time for same 3 wavelengths presented in Figures 50A-C, but from another running of the same plasma recipe on product in the same processing chamber
- Figure 52 is a flowchart of one embodiment of an endpoint detection subroutine which may be used by the endpoint detection module of Figures 7 and 32,
- Figure 53 is a flowchart of another embodiment of an endpoint detection subroutine which may be used by the endpoint detection module of Figures 7 and 32,
- Figure 54A is an exemplary spectra from a processing chamber at the start of a plasma process step
- Figure 54B is an exemplary spectra which has been selected as being indicative of endpoint of the plasma process step from Figure 54A for use as a reference by the endpoint detection subroutine of Figure 53,
- Figure 54C is the difference between the spectra of Figures 54A and 54B in accordance with the endpoint detection subroutine of Figure 53
- Figure 55A is an exemplary spectra from a processing chamber at an intermediate time of the plasma process step presented in Figure 54A
- Figure 55B is the same spectra from Figure 54B,
- Figure 55C is the difference between the spectra of Figures 55A and 55B in accordance with the endpoint detection subroutine of Figure 53,
- Figure 56A is an exemplary spectra from a processing chamber at the endpoint of the plasma process step presented in Figure 54A,
- Figure 56B is the same spectra from Figure 54B,
- Figure 56C is the difference between the spectra of Figures 55A and 55B in accordance with the endpoint detection subroutine of Figure 53,
- Figure 57 is a flowchart of another embodiment of an endpoint detection subroutine which may be used by the endpoint detection module of Figures 7 and 32,
- Figure 58 is an exemplary optical emissions output which would be indicative of the endpoint of a plasma process step in accordance with the endpoint detection subroutine of Figure 57,
- Figure 59 is a flowchart of one embodiment of a wafer distribution subroutine which may be incorporated in the wafer distribution module of Figures 7 and 32
- Figure 60 is a flowchart of another embodiment of a wafer distribution subroutine which may be incorporated in the wafer distribution module of Figures 7 and 19
- Figure 61 is another embodiment of a fixture for interconnecting an optical fiber with a processing chamber
- Figure 62 is another embodiment of a research subroutine for identifying a wavelength region has an indicator for endpoint
- Figures 63A-H are plots of change in area versus time for various wavelength regions
- Figure 64 is another embodiment of an endpoint detection subroutine
- Figure 65 is an embodiment of a plasma monitoring network
- Figure 66 illustrates various modules which may be accessed over the plasma monitoring network of Figure 65 Dp-tailed Description
- a wafer production system 2 is illustrated in Figure 1 and is generally for executing one or more plasma-based processes (single or multiple step) on wafers 18
- Semiconductor devices may be formed from wafers 18 processed by the system 2, including integrated circuit chips
- the system 2 generally includes a wafer cassette 6 which stores a plurality of wafers 18 and allows these wafers 18 to be readily transported to and from the system 2
- One wafer cassette 6 is disposed in each of the two load lock chambers 28 of the wafer production system 2
- a wafer handling assembly 44 is advanced into the respective load lock chamber 28, removes at least one of the wafers 18 from the wafer cassette 6, and transfers the wafer(s) 18 to one of the plurality of processing chambers 36 of the wafer production system 2 (four chambers 36a-d being illustrated)
- Other arrangements may be utilized for purposes of the present invention
- a main control unit 58 (hereafter "MCU 58")
- the MCU 58 is a computer having at least one computer-readable storage medium and at least one processor, such as a desktop PC or a main-frame having satellite terminals
- the MCU 58 is a computer having at least one computer-readable storage medium and at least one processor, such as a desktop PC or a main-frame having satellite terminals
- the MCU 58 is a computer having at least one computer-readable storage medium and at least one processor, such as a desktop PC or a main-frame having satellite terminals
- the MCU 58 is a computer having at least one computer-readable storage medium and at least one processor, such as a desktop PC or a main-frame having satellite terminals
- the MCU 58 is a computer having at least one computer-readable storage medium and at least one processor, such as a desktop PC or a main-frame having satellite terminals
- the MCU 58 is a computer having at least one computer-readable storage medium and at least one processor, such
- Wafer Cassette 6 - Figure 2 More details regarding the embodiment of the wafer cassette 6 which is incorporated in the wafer production system 2 of Figure 1 are presented in
- the wafer cassette 6 includes a frame 10 defined by a pair of laterally spaced sidewalls 22 which are interconnected by a back panel 26, as well as a pair of end panels 8
- the front of the frame 10 is substantially open such that the wafer handling assembly 44 may be advanced within and retracted from the wafer cassette 6 to remove wafers 18 from and provide wafers 18 to the associated wafer cassette 6
- a plurality of longitudinally spaced and laterally disposed partitions 16 (e g , each partition 16 being disposed at least generally perpendicular to the longitudinal axis of the cassette 6) are provided within the frame 10 for purposes of maintaining separation of adjacent wafers 18
- Each pair of adjacent partitions 16 defines a pocket 14 in which a single wafer 18 may be placed
- Loading of wafers 18 within the wafer cassette 6 which are to be plasma processed may be accomplished by disposing the one of the end panels 8 of the cassette 6 on an appropriate supporting surface and manually loading wafers 18 into the cassette 6, with only one wafer 18 being disposed in any
- Wafer Handling Assembly 44 Figures 3A-3B Additional details regarding the wafer handling assembly 44 which is incorporated in the wafer production system 2 of Figure 1 are illustrated in Figures 3A-B Other types of wafer handling assemblies may be utilized by the wafer production system 2, such as the types disclosed in U S Patent Nos 5,280,983 to Maydan et al , issued January 25, 1994, and entitled “SEMICONDUCTOR PROCESSING SYSTEM WITH ROBOTIC AUTOLOADER AND LOCK” and 5,656,902 to Lowrance, issued August 12, 1997, and entitled “TWO-AXIS MAGNETICALLY COUPLED ROBOT", both of which patents are incorporated by reference in their entirety herein
- the wafer handling assembly 44 of Figures 3A-B generally includes a robotic wafer handler 48 which is disposed within a central chamber 70 of the wafer production system 2 Load lock chambers 28 and processing chambers 36 are thereby disposed about the wafer handling assembly 44 Movement of the robotic wafer handler 48 is realized through a wafer handler
- Processing Chamber 72 - Figures 4 and 5 One embodiment of a processing chamber which may be incorporated in the wafer production system of Figure 1 as one of the chambers 36 is presented in more detail in Figure 4 Other types/configurations of processing chambers may be utilized by the wafer production system 2 for purposes of the present invention as well, including those disclosed in U S Patent Nos 5,614,055 to Fairbairn et al , issued March 25, 1997, and entitled "HIGH DENSITY PLASMA CVD AND ETCHING REACTOR", and 5,641 ,375 to Nitescu et al , issued June
- the processing chamber 74 of Figure 4 is specifically adapted for performing a plasma etching operation on a wafer(s) 18 when disposed therein
- the processing chamber 74 includes chamber sidewalls 78 which are disposed about a central, longitudinal axis 76 of the chamber 74 Access to the processing chamber 74 may be provided by a chamber cover 82 which is interconnected with the chamber sidewalls 78 in such a manner that at least in certain instances, at least a portion of the chamber cover 82 may be moved away from the chamber sidewalls 78 In the illustrated embodiment the chamber cover 82 is removed only to gain access to the interior of the processing chamber 74 for maintenance, cleaning, or both
- a window port 124 extends through a portion of the chamber sidewall 78 and is aligned with a transparent window 112
- the window 112 includes an inner surface 116 and an outer surface 120, and provides a way for the plasma to be viewed exteriorly of the processing chamber 74 and further to provide a mechanism for obtaining optical emissions data on the plasma recipe being run on the wafer(s) 18 within the chamber 74
- a bell jar 90 and a bell roof 86 which are each formed from transparent, dielectric materials (e g , quartz, sapphire)
- the bell jar 90 is spaced radially inward (e g , in the direction of the central, longitudinal axis 76 of the chamber 74) from the inner surface of the chamber sidewalls 78
- the bell roof 86 is disposed above the bell jar 90 and is axially movable in a direction which is at least substantially parallel with the central, longitudinal axis 76 of the chamber 74 through interconnection with an elevator 98 Movement of the elevator 98 may be desirable for one or more purposes For instance, this movement may be used to change the spacing between a showerhead 94 and a wafer pedestal 106/wafer platform 102 which in one embodiment are the electrodes or "plasma generator" for the chamber 74
- the wafer pedestal 106 is disposed radially inwardly of the bell jar 90 in spaced relation therewith, and the wafer platform 104 is disposed on top of the wafer pedestal 106
- both the wafer pedestal 104 and wafer platform 106 are formed from silicon-based materials since the wafers 18 are also commonly formed from silicon-based materials
- the wafer 18 is introduced into the processing chamber 74 through a wafer access 80 which extends through the chamber sidewall 78, and is disposed in a flat orientation on the upper surface of the wafer platform 104
- Various mechanisms may be used to retain the wafer 18 on the wafer platform 104 during the running of the plasma process on the wafer 18 in the chamber 74, such as by drawing a vacuum through a vacuum port 108 which is formed on the wafer platform 104 or by using electrostatic charges (not shown) Transport of the wafer 18 into the processing chamber 74 is again provided by the wafer blade 66 of the wafer handling assembly 44 ( Figures 1 and 3A-B) After the wafer blade
- the showerhead 94 is interconnected with the elevator 98 such that it is axially movable therewith, and in one embodiment is also formed from a silicon- based material for the above-noted reasoning
- the showerhead 94 includes one or more apertures (not shown) for the purpose of dispersing feed gases within the vacuum chamber 84 in a manner to define a desired gas flow pattern for the plasma Gases are provided to the showerhead 94 through a gas inlet port 100 formed in the quartz bell roof 86
- an appropriate voltage may be applied to one or more of the wafer pedestal 106 and the showerhead 94 to create the plasma within the chamber 74 above the wafer platform 104
- the wafer pedestal 106 and wafer platform 104, as well as the showerhead 94 thereby aiso function as electrodes in the illustrated embodiment as noted
- the electrical field generated by these electrodes also functions to effectively confine the plasma to the space between the electrodes
- FIG 5 illustrates one embodiment of a gas delivery system 150 which may be used to provide gases to the processing chamber 74 of Figure 4 for a given plasma process operation
- the gas delivery system 150 includes a plurality of storage tanks 154, each of which is fiuidly interconnected with the processing chamber 74 either directly or indirectly
- Storage tanks 154b-d are available for containing one or more types of feed gases which will define the gas composition of the plasma within the vacuum chamber 84
- Each storage tank 154b-d is fiuidly interconnected with a mixer 166 by gas lines 158b-d where the feed gases may be appropriately mixed prior to being provided to the processing chamber 74 through the showerhead
- One embodiment of an assembly for monitoring/evaluating plasma processes and which may be incorporated in the wafer production system 2 of
- FIG. 1 is illustrated in Figure 6
- the plasma monitoring assembly 174 operatively interfaces with the window 38 of the processing chamber 36 by receiving optical emissions of the plasma which pass out of the processing chamber 36 through the window 38
- These optical emissions are "collected" by an appropriate fiber optic cable 178 which is positioned at or near the outer surface 42 of the window 38
- Fixtures which illustrate ways of maintaining a fiber optic cable and a window of a processing chamber in a fixed positional relationship are presented in Figures 36 and 39
- Optical emissions of the plasma within the processing chamber 36 during processing of a wafer 18 enter the fiber optic cable 178 and are directed to a spectrometer assembly 182
- Both scanning-type and solid state spectrometers may be used as the spectrometer assembly 182
- the assembly 182 may also include one or more appropriately interconnected spectrometers, each of which obtains optical emissions data from a different region
- the spectrometer assembly 182 separates these optical emissions into a plurality of individual wavelengths and provides these separate optical components to
- a computer-readable signal is provided by the CCD array 186 to a plasma monitor control unit 128 (hereafter "PMCU 128") which is the primary control mechanism of the plasma monitoring assembly 174
- PMCU 128 plasma monitor control unit 128
- PMCU 128 is a computer which may be configured to include, but not limited to, at least one motherboard, at least one analog-to-digital conversion board, at least one central processing unit (CPU) for each motherboard, and one or more types of computer-readable storage mediums such as at least one floppy disk drive, at least one hard disk drive, and at least one CD ROM drive
- Other hardware may be operatively interconnected with the PMCU 128, such as a display 130 for providing visual/audio-based information to operations personnel
- One PMCU 128 may be provided for each chamber 36, or the PMCU 128 may be configured to service multiple chambers 36
- the PMCU 128 is also operatively interfaced or interconnected with the MCU 58 of the wafer production system 2 such that the PMCU and MCU 58 may communicate with each other
- the PMCU 128 includes a plasma monitoring module 200 and each of its sub-modules may be stored on a computer-readable storage medium associated with the PMCU 128 (e g , on a portable computer d ⁇ skette(s), on a hard drive, on a CD(s))
- the plasma monitoring module 200 and these sub-modules are illustrated in Figure 7
- One sub-module is a startup module 202 which provides a way of accessing other sub-modules through a current plasma process module 250
- the current plasma process module 250 of the plasma monitoring module 200 facilitates the monitoring or evaluation of the various types of plasma processes which may be conducted within the chamber 36 through the evaluation of optical emissions data of the plasma in the chamber 36
- optical emissions data are collected and delivered by the fiber optic cable 178 to the spectrometer assembly 182 which divides the light up into its individual optical components Data representative of these optical emission components are then made available to the current plasma process module 250 through the CCD array 186 as described above
- optical emissions of the plasma in the processing chamber 36 which are obtained and available for evaluation (e g , by the current plasma process module 250 manually by the appropriate personnel) include at least those wavelengths from about 250 nanometers to about 1 ,000 nanometers (inclusive), and more preferably at least those wavelengths from about 150 nanometers to about 1 ,200 nanometers (inclusive)
- the above-noted desired range or bandwidth of optical emissions data which are obtained/collected of the plasma in the chamber 36, and which includes each of the above-noted ranges or bandwidths, will be referred to as the "Preferred Optical Bandwidth " Optical or wavelength resolutions within and throughout the Preferred
- Optical Bandwidth are preferably no more than about 1 nanometer, and even more preferably no more than about 0 5 nanometers (presently contemplating a wavelength resolution of 04)
- the term "wavelength resolution” in this context means the amount of separation between adjacent wavelengths in the subject optical emissions data which is collected Therefore, if the wavelength resolution being used to collect optical emissions data from the plasma in the chamber 36 is 1 nanometer, no more than a 1 nanometer spacing will exist between any two data points within and throughout the Preferred Optical Bandwidth
- equal spacings will typically be utilized in relation to the wavelength resolution within and throughout the Preferred Optical Bandwidth, this need not be the case such that “wavelength resolution” encompasses equal spacings, unequal spacings, and combinations thereof
- the above-noted magnitude for the optical or wavelength resolution will be referred to as the "Preferred Data Resolution"
- Another factor relating to the effectiveness of the current plasma process module 250 in relation to the amount of optical emissions data of the plasma in the chamber 36 is the times at which this data is taken during the subject
- the spectrometer assembly 182 illustrated in Figure 6 should be capable of meeting the above-noted criteria, and a number of implementations may be utilized
- the spectrometer assembly 182 may be of the scanning type in which the spectrometer assembly 182 would include structure to scan the spectrum to obtain data encompassing the Preferred Optical Bandwidth using the Preferred Data Resolution and at the Preferred Data Collection Time Resolution (e g , scan a first optical emissions segment or region of the 250-550 nanometer wavelengths, scan a second segment of the 500-750 nanometer wavelengths, and scan a third segment of the 700-950 nanometer wavelengths, each of which overlaps so that the possibility of losing data is reduced and further to facilitate alignment of spectral segments)
- the spectrometer assembly 182 may also be a solid state device Multiple subunits or processing cards may be connected in parallel relation to function similar to the scanning type noted above That is, each subunit or processing card of the solid state device would then provide information on a specific optical emissions segment or region within the Preferred Optical Bandwidth
- Exemplary Plasma Recipe Spectra - Figure 8 A representative or exemplary spectra in computer-readable form which may be made available to the current plasma process module 200 of Figure 7 for analysis is presented in Figure 8 Only a portion of the Preferred Optical Bandwidth is represented by the spectra 246 However, it serves to illustrate certain principles associated with the present invention since the evaluation of the current plasma process need not be of the entire Preferred Optical Bandwidth in each case
- the spectra 246 contains data within the wavelength range of 400 nanometers to about 700 nanometers and at a certain fixed point in time in a plasma process being conducted within the processing chamber 36 of Figure 1 (e g , at a current time t n )
- Various characteristics of the spectra 246 of Figure 8 may be used in the analysis undertaken by the current plasma process module 250 of Figure 7 These characteristics include the overall pattern of the spectra 246, one or more of the location and intensity of one or more of intensity peaks 248 in the spectra 246, and one or more of the relative location and relative intensity
- the plasma spectra directory 284 of Figure 9 includes a number of subdirectories or subsets of categorically similar data is typically specific to a single processing chamber 36 (although the same directory 284 could be used for multiple chambers 36 if the data within the directory 284 was indexed in some way to the specific chamber 36), is accessed by one or more of the submodules of the current plasma process module 250, and is preferably stored in a computer-readable medium of or associated with the PMCU (e g , one or more computer diskettes, hard drive, one or more CDS)
- Relevant data which is included in each data entry in each of these subdirectories of the plasma spectra directory 284 is a spectra (optical emissions data), and in all but one case (spectra of calibration light subdirectory 310) is a spectra of the plasma in the processing chamber 36 at the relevant time and typically within the Preferred Optical Bandwidth and at the
- Spectral data is used by the current plasma process module 250 to determine if subsequent plasma processes conducted in this very same processing chamber 36 are proceeding in accordance with at least one of the plasma processes stored in the normal spectra subdirectory 288 Entries in the normal spectra subdirectory 288 are thereby used as a "model" or "standard” for the evaluation of plasma processes conducted in this very same processing chamber 36 at some future time How data is actually entered in the normal spectra subdirectory 288 will be discussed in more detail below in relation to the startup module 202 and Figures 13-14 Suffice it to say for present purposes that entries in the normal spectra subdirectory 288 are from actual plasma processes conducted in the subject chamber 36 These plasma processes are either confirmed (e g , by post-plasma processing testing) or assumed (and typically later confirmed) to have proceeded in a desired or predetermined manner, or more specifically without any substantial/significant errors or aberrations No pre-analysis or knowledge of any plasma process is required to use the current plasma process module 250 and the normal spectra subdirectory
- Spectral data from a plasma process ABC conducted in a given chamber 36 may be recorded in the normal spectra subdirectory 288 one day simply for purposes of determining if any subsequent running of this same plasma process ABC in this same chamber 36 has proceeded in accordance with the spectral data from the plasma process ABC previously recorded in the normal spectra subdirectory 288
- Spectra of "Abnormal” Plasma Processes and identified by reference numeral 292 (hereafter “abnormal spectra subdirectory 292")
- Data relating to any of the plasma processes referenced above in relation to the normal spectra subdirectory 288 may also be stored in the abnormal spectra subdirectory 292, and the above-noted organizational techniques may be utilized here as well
- Entries to the abnormal spectra subdirectory 292 are made when a given plasma process conducted in the processing chamber 36 does not proceed in the desired or predetermined manner (e g , when the process has not proceeded according to the relevant plasma process(es) of the normal spectra subdirectory 288), and further when the cause or causes of the error or aberration has been identified to the plasma spectra directory 284
- the module 250 may then compare this "deviating spectral data" on the current plasma process with spectral data in the abnormal spectra subdirectory 292 Any number of actions may be initiated if the current plasma process module 250 identifies a "match" between the spectral data from the plasma process currently being conducted in the processing chamber 36 and spectral data in the abnormal spectra subdirectory 292 These actions may include issuing an appropriate alert(s) of the error condition, addressing one or more aspects of or relating to the control of the chamber 36, or both as will be discussed in more detail below in relation to the process alert module 428 of
- Spectral data from a plasma process which is currently being conducted in the subject processing chamber 36 which does not “match” any plasma process stored within the normal spectra subdirectory 288, and which further does not “match” with corresponding spectral data in the abnormal spectra subdirectory 292, is recorded in a subdirectory of the plasma spectra directory 284 which is entitled “Spectra of "Unknown” Plasma Processes” and identified by reference numeral 296 (hereafter "unknown spectra subdirectory 296")
- Any error or aberration which is "new" to the current plasma process module 250 i e , spectral data, indicative of an error or aberration, which has not been previously recorded in the abnormal spectra subdirectory 292
- relevant data from the current plasma process being recorded in the unknown spectra subdirectory 296
- Spectral data recorded in the unknown spectra subdirectory 296 from prior plasma processes will typically be analyzed by personnel at some point in time after the process has been terminated If the spectral data from a plasma process recorded in the unknown spectra subdirectory 296 is identified as being a new plasma process, and if a determination is made to use this spectral data as a standard for evaluating further runnings of this same plasma process on this same processing chamber 36 this spectral data may be transferred to the normal spectra subdirectory 292 Entries may also be made to the abnormal spectra subdirectory 292 from the unknown spectra subdirectory 296 Analysis of the spectral data from a particular plasma process which is recorded in the unknown spectra subdirectory 296 may lead to the conclusion that the spectral data is associated with one or more particular errors/aberrations which is identifiable by its spectral data The relevant spectral data from the unknown spectra subdirectory 296 may then be transferred to the abnormal spectra subdirectory 292
- the plasma spectra directory 284 of Figure 9 also contains data which is indicative of when the endpoint has been reached of an entire plasma process or a discernible portion thereof such as a plasma step of a single or multi-step plasma recipe or other plasma process "Endpoint" in the context of a plasma process or a discernible portion thereof is that time in the plasma process when the plasma within the processing chamber 36 has achieved a certain predetermined result
- Each plasma step in a plasma recipe typically has one or more characteristics in its corresponding spectra which will indicate that the desired predetermined result has been achieved, as typically does the end of a plasma clean which was initiated without first wet cleaning the chamber 36, a plasma clean which was initiated after wet cleaning the chamber 36, and a conditioning wafer operation
- Spectral data of a plasma process conducted in a chamber 36 may be analyzed after the plasma process is terminated to identify one or more spectra (or portions thereof such as one or more individual wavelengths) which are indicative that endpoint of the subject plasma process or plasma process step has been reached Spectral data which is indicative
- a final subdirectory of the plasma spectra directory 284 of Figure 9 is a calibration light spectra subdirectory 310 which does not contain spectra of plasma from the chamber 36 Instead, one or more spectra of one or more calibration lights are stored in the subdirectory 310 Generally, a calibration light, whose spectra is in the calibration light subdirectory 310, is directed at the window 38 of the processing chamber 36 A comparision is made between the spectral pattern of the calibration light from the subdirectory 310 and the spectral pattern of that portion of the calibration light which is reflected by the inner surface 40 of the window 38 on the processing chamber 36 The results of the comparison are used to determine the type and amount of calibration which should be implemented in relation to the operation of the current plasma process module 250 as will be discussed in more detail below in relation to the calibration module 562 and Figures 40-48
- the plasma spectra directory 284 may have the same subdirectories as presented in Figure 9, although only the normal spectra subdirectory 288 and abnormal spectra subdirectory 292 are illustrated for convenience and only data on one category of plasma process which may be stored in the normal subdirectory 288 is illustrated (plasma recipes).
- Review of the normal spectra subdirectory 288 of Figure 10 indicates that spectra on multiple plasma recipes, which have been previously run on product in the processing chamber 36 which is associated with the plasma spectra directory 284, are each stored in their own main data entry 350. This would also be the case with other categories of plasma processes stored in the normal subdirectory 288. That is, each main data entry 350 is reserved for storing information which is used to evaluate plasma processes which are to be conducted in this same processing chamber 36.
- Each main data entry 350 for a given plasma process has a plurality of data segments 354 associated therewith, and each of these data segments 354 will include at least a spectra (e.g., Figure 8) of the plasma in the processing chamber 36 at a certain point in time and preferably within the Preferred Optical Bandwidth at the Preferred Data Resolution.
- the spectra associated with each data segment 354 may be stored as a single spectra which covers the Preferred Optical Bandwidth, or may be stored as multiple spectra which collectively cover the Preferred Optical Bandwidth.
- Spectra for the data segments 354 are taken periodically throughout the running of a plasma process within the processing chamber 36 through the window 38 on the chamber 36 (e.g., by the plasma monitoring assembly 174 of Figure 6 or any of the embodiments illustrated in Figures 31 and 37 below) using the Preferred Data Collection Time Resolution.
- the entirety of the plasma process may be recorded in the normal spectra subdirectory 288 in this manner, sometimes the plasma is rather unstable when it first comes on in the chamber 36. Therefore, it may be desirable to not retain optical emissions data in the normal spectra subdirectory 288 from this unstable time period.
- Entries of plasma processes in the normal spectra subdirectory 288 may consist of a plurality of totally different types or species of plasma processes within a given category or genus as also illustrated in Figure 10.
- Plasma recipe A is stored under main data entry 350a, which is different from a plasma recipe B which is stored under main data entry 350b, which is different from a plasma recipe "X" which is stored under main data entry 350c.
- Multiple runnings of the same plasma recipe or process may also be recorded in the normal spectra subdirectory 288 as well if desired (not shown). For instance, spectral data from two separate runnings of plasma recipe A on the same type of product in the associated processing chamber 36 may actually be included in the normal spectra subdirectory 288. Evaluation of a current plasma recipe being run on product in the subject processing chamber 36 would then potentially involve the comparison of optical emissions data on the current process in relation to both of these main data entries 350.
- optical emissions data within the normal spectra subdirectory 288 of Figure 10 may be consolidated or condensed to eliminate the storage of redundant data, to increase the speed of the search of the normal spectra subdirectory 288 by the current plasma process module 250, or both.
- Figure 11 illustrates one way in which this may be accomplished in the case of the normal spectra subdirectory 288a for one example where Plasma Recipes A-D are stored in the directory 288a. The same principles would apply to any type of plasma process which is stored in the normal spectra subdirectory 288.
- Plasma Recipe A under main data entry 358a and Plasma Recipe B under main data entry 358b each have the same spectra for purposes of the current plasma process module 250 from time t., (the first time data is recorded in the subdirectory 288a for the subject plasma recipe) to time t n (the "nth" time data is recorded in the subdirectory 288a for the subject plasma recipe).
- a under mam data entry 358a and Plasma Recipe B under mam data entry 358b each then include their own individual data segments 366a-c and 366d-f respectively, over the time period from t n+1 to t n+x (the "xth" time data is recorded in the normal spectra subdirectory 288a after the time t n )
- Each of the Plasma Recipes A and B end at the same time for purposes of the example of Figure 11
- the normal spectra subdirectory 288b of Figure 11 also has mam data entries 358c and 358d for Plasma Recipes C and D, respectively
- the same data storage concept used in relation to Plasma Recipes A and B is likewise employed for Plasma Recipes C and D in the normal spectra subdirectory 288a
- the spectra of Plasma Recipe C under data main entry 358c and Plasma Recipe D under mam data entry 358d are the same for purposes of the current plasma process module 250 from time t., to time t n
- the multiple spectra from this time range are stored only once in common data segments 362d-h in the normal spectra subdirectory 288a
- Common data segments 362d-h are thereby associated with both Plasma Recipe C of mam data entry 358c and Plasma Recipe D of mam data entry 358d
- the spectra of Plasma Recipe C under mam data entry 358c and the spectra of Plasma Recipe D under main data entry 358d differ for purposes of the current plasma process module 250 As such Plasma Recipe C under mam data entry 358c and the spectra of Plasma Recipe D under main data entry 358d differ for purposes of the current plasma process module 250 As such Plasma Recipe C under mam data entry 358c and the spectra of Plasma Recipe D under main data entry 358d differ for purposes of the current plasma process module 250 As such Plasma Recipe C under
- Each data segment 354 of each plasma process stored under a mam data entry 350 in the normal spectra subdirectory 288 of Figure 10 may contain a multiplicity of data types relevant to the monitoring of the current plasma process with the current plasma process module 250
- a representative example is presented in Figure 12A where these various data types of data are presented in data fields 322 which are associated with each data segment 354
- Spectral patterns of the plasma in the processing chamber 36 is a significant data type for comparing the current plasma process with the plasma spectra directory 284, and these spectra are stored in a spectra field 322d in the normal spectra subdirectory 288 of Figure 12A
- Each data segment 354 in the normal spectra subdirectory 288 also includes a time field 322a where the time associated with the spectra in the spectra field 322d is recorded (e g , the time into the plasma process when the spectra is taken) Data in the time field 322a may be used in various ways by the current plasma process module 250 as will be discussed in more detail
- each mam data entry 350 in the normal spectra subdirectory 288 "Associated" in this context means that this information may be provided once for each main data entry 350 for a given plasma process or for only a few times which is less than the total number of data segments 354 under a particular mam data entry 350, but it also encompasses a situation where this information is actually provided for each data segment 354 of the subject mam data entry 350 which is not as desirable because of redundancies
- Fields for these types of information include a plasma process "genus" field 322h (e g , to identify whether the main data entry
- a plasma recipe, plasma clean, or conditioning wafer operation is a plasma recipe, plasma clean, or conditioning wafer operation
- a wafer identifier field 322b e g , for information which corresponds with an identifier, such as a number or code, which appears on the wafer 18 which is to have a plasma recipe run thereon and which is used for tracking purposes
- a plasma process "species" field 322c a subset of a plasma process "genus” such as different types of plasma recipes (e g , plasma recipe A and plasma recipe B)
- a plasma process step field 322e e g , to identify the step of a particular plasma recipe or any other plasma process which provides a different function or achieves a different result than other portions of the process
- a maximum total plasma process step time field 322f e g , the maximum amount of time allowed to complete a given plasma step of a multi-step plasma recipe or other process
- a maximum total plasma process time field 322g e g , the maximum
- the category or genus of the subject plasma process may be identified in a plasma process genus field 338e (e g , plasma recipe, plasma clean, conditioning wafer operation), a particular type or species of a given category or genus of plasma process may be identified in a plasma process species field 338f (e g , a particular type of plasma recipe), and the type of plasma step may be identified in a plasma process step field 338g
- a plasma process genus field 338e e g , plasma recipe, plasma clean, conditioning wafer operation
- a particular type or species of a given category or genus of plasma process may be identified in a plasma process species field 338f (e g , a particular type of plasma recipe)
- the type of plasma step may be identified in a plasma process step field 338g
- Data in addition to the above-described spectra may be associated with each of the data segments 354 under each mam data entry 346 for known error/aberrations within the abnormal spectra subdirectory 2
- the current plasma process module 250 includes error identification capabilities as will be discussed in more detail below Once the current plasma process module 250 identifies a match between the current optical emissions data and a relevant spectra or portion thereof in the abnormal spectra subdirectory 292, information on the corresponding error/aberration may be issued based upon information in the error field 338c Moreover, corrective actions may be undertaken based upon the contents of this same error field
- each data segment 354 under each mam data entry 346 for known error/aberrations within the abnormal spectra subdirectory 292 may also include a protocol field 338d Information contained within the protocol field 338d will somehow relate to how the subject error or aberration may or should be addressed, or more specifically what action or actions may or should be undertaken to address the error
- Single or multiple protocols may be stored in any one protocol field 338d (e g , more than one protocol may be appropriate to address a certain condition)
- the current plasma process module 250 identifies a match between the current spectra and a relevant spectra in the abnormal spectra subdirectory 292, how the corresponding error/aberration is addressed may be based upon information contained in the subject protocol field 338d
- Data regarding process control parameters or conditions associated with the processing chamber 36 may also be included in each of the above-described subdirectories of the plasma spectra directory 284 for a specific data entry, particularly in the case of the abnormal spectra subdirectory 292 and the unknown conditions subdirectory 2
- Certain key principles of the present invention are simply based on whether one spectral pattern matches another spectral pattern (e g , does the spectral pattern of the plasma in the chamber 36 "match" the pattern of the relevant spectra in the relevant subdirectory of the plasma spectra directory 288) In many cases this determination may be made through the pattern recognition module 370 which is presented in Figure 13 Various "pattern recognition techniques" may be employed by the pattern recognition module 370 to provide the above-noted function
- One such pattern recognition technique is embodied by the flowchart depicted in Figure 13 and may be generally characterized as a point-by-point pattern recognition technique
- the pomt-by- pomt pattern recognition technique which is embodied by the pattern recognition subroutine 374 of Figure 13 is contained within step 378
- the intensity at a first wavelength in the current spectra at the current time t c (a fixed point in time) is compared with the intensity at this same first wavelength of the relevant spectra from a Target Directory Whichever sub-module of the current
- “matching spectra” in relation to one sub-module of the current plasma process module 250 may not be a “matching spectra” in relation to another of its sub-modules If the intensities of the two subject spectra are within a "match limit” of each other at this first wavelength in the subject optical emissions, the patterns of the two subject spectras are initially considered a “match” and the analysis is repeated at a second wavelength which is displaced and different from the first wavelength and which defines a second "point” on which the above-noted pomt- by-pomt analysis is repeated
- the particular "match limit" which is used by the pattern recognition subroutine 374 may be specific to which sub-module of the current plasma process module 250 calls the pattern recognition subroutine 374 as noted The above-described point-by-point analysis is repeated by
- step 378 "advancing along" the entirety of the current spectra at typically pre-seiected wavelength increments (e g , every nanometer) Typically, a fixed wavelength increment will be utilized by step 378, such that the comparison between the two subject spectra will be made at every "x" nanometers throughout the entirety of the entire "bandwidth” of the spectra However, there need not be equal spacings between each of the "points" examined by the pattern recognition subroutine 374
- Another match criterion which may be used in combination with the above-noted point-by-point comparison relates to how many of the examined points must be within the "match limit" in order for the two subject spectra to be considered a match
- the sub-module which calls the pattern recognition subroutine 374 may require each "point" examined in the subject point-by-point analysis to be within the selected match limit in order for the two spectra to be considered a match Alternatively, something less than 100% may also be utilized For instance, having at least 95% of the points being within the match limit may equate with the two subject spectras being considered a match
- the average variation of the multiplicity of points examined by the pattern recognition subroutine 374 may be calculated and compared with an average associated with the relevant spectra from the Target Directory to determine if it is within a predefined tolerance Any combination of the foregoing may be implemented for determining what is a "match " A number of factors will affect the accuracy which may be attributed to the results achieved by the execution of step 378 in the pattern recognition
- Analytical Wavelength Resolution in this context means the wavelength increments at which the above-described point-by-point analysis will be performed throughout the subject spectra
- a pattern is used such as a fixed wavelength increment
- the Analytical Wavelength Resolution used by the pattern recognition subroutine 374 is no more than about 2 nanometers, and even more preferably no more than about 0 5 nanometers
- this will be referred to as the "Preferred Analytical Wavelength Resolution
- match limit input to the pattern recognition subroutine 374, whether using a raw difference theory, a percentage difference theory, or a combination of a raw difference theory and a percentage difference theory (e g , the difference in intensity must be no more than a certain, input number of intensity units and must also be within a certain, input number of percentage points of each other in the case of a combination), the 200 nanometer wavelength "point" of the current spectra and the relevant spectra from the Target Directory will be characterized as a "match" at the current time t c The point-by-point analysis will then continue at the 201 nanometer wavelength in the above-described manner, and will be repeated at each 1 nanometer increment until reaching the last 900 nanometer wavelength Results of the point-by-point comparison of step 378 for the current spectra at the current time t c will then be provided to step 380 of the pattern recognition subroutine 374 of Figure 13 for use by the sub-module of the current plasma process module 250 which called the pattern recognition module 370 Control of the plasma monitoring
- spectral data should be taken at least every 1 second and the analysis of this data should be completed by the pattern recognition subroutine 374 as fast as possible
- the identification of the current plasma recipe and the analysis of the performance of the processing chamber 36 e g , plasma health
- the pattern recognition subroutine 374 of Figure 13 is able to meet the demands through simplifying the analysis of the spectra of the plasma in the processing chamber 36
- the sum total of the analysis provided by the pattern recognition subroutine 374 is simply whether the pattern of the current spectra "matches" the pattern of the relevant spectra from the Target Directory There is no need to locate or
- Process Alert Module 428 - Figure 14 Various conditions which may be encountered by the current plasma process module 250 may result in the transfer of control to or the sharing of control with the process alert module 428 of Figure 14
- One or more subroutines may be included under the process alert module 428
- Each of these subroutines may present various options in relation to how the relevant condition or situation is addressed which resulted in the activation of the process alert module 428
- actions are made available - the issuing of one or more alerts and addressing the control of the subject plasma process in some manner
- One or more alarms, alerts, or the like may be activated if the alarm alert function of the process alert subroutine 432 of Figure 14 is enabled at its step 454 in relation to the subject condition or situation At least one visual alarm may be activated in step 458 of the process alert subroutine 432
- Exemplary visual alarms include displaying a general indication of the existence of the relevant condition (e g , a flashing light), a more specific indication of the subject condition (e g , providing a textual description of the identified condition or situation), or both Appropriate locations where information relating to the subject condition may be presented include the display 130 associated with the particular processing chamber 36 where the subject condition was encountered, the display 59 associated with the wafer production system 2 which may be characterized as a master control panel of sorts for the wafer production system 2, any master control panel for the entire fabrication facility incorporating the wafer production system 2, any computer network on which the wafer production system 2 is included, or on any combination of the foregoing Other visual indications may be employed alone or in combination with any of
- Step 450a presents a protocol category which is the modification of one or more process control parameters
- One or more spectra from the abnormal spectra subdirectory 292, one or more conditions, or both, may be included in step 448a of the process alert subroutine 432 which will access step 450a
- Step 450a is directed toward a protocol category which is to attempt to "address" (e g , correct/remedy) the subject condition in the current plasma
- any one or more spectra or one or more conditions associated with step 448a may have one or more process control protocols associated therewith
- step 450a is not directly integrated with the relevant process controller(s)
- Step 450b contains one or more protocols which are directed toward terminating the subject plasma process, although typically termination of the current plasma process will simply entail terminating the gas flow to the chamber 36 and the electrical componentry which are responsible for creating the plasma Termination of the current plasma process may be automatically undertaken if desired by the facility incorporating the wafer production system 2 through operatively interfacing the process alert subroutine 432 with the appropriate process controller(s) (e g , by the PMCU 128 sending an appropriate signal to the MCU 58) Manual termination of the current plasma process is also contemplated by step 450b Execution of step 450b in this case may then simply entail apprising the appropriate personnel that a condition has
- One or more spectra/conditions associated with step 448e may be associated with different protocols in step 450e
- one protocol of step 450e corresponding with one or more spectra/conditions associated with step 448e may relate to a plasma cleaning operation which may be initiated in accordance with the foregoing Other spectra or conditions associated with step
- step 450e may access a protocol of step 450e which relates to a wet clean which may be initiated in accordance with the foregoing Spectra or conditions within the chamber 36 which are of a nature such that the wafer distribution sequence should be affected in some manner by their existence may be included in or associated with step 448c
- the protocol set forth in step 450c thereby addresses the manner in which wafers 18 are distributed to the various processing chambers 36 of the wafer production system 2 through the wafer distribution module 1384 which will be discussed in more detail below in relation to Figures 59-60 Addressing the sequence of distribution of wafers 18 to the processing chambers 36 of the wafer production system 2 may be automatically undertaken if desired by the facility incorporating the wafer production system 2 through operatively interfacing the process alert subroutine 432 with the appropriate process controller(s) (e g , wafer distribution module 1384, MCU 58) Manual techniques are also contemplated by step 450c in that the execution of step 450c of the process alert subroutine 432 may simply
- plasma process/plasma process step endpoint may be addressed through the process alert subroutine 432 of Figure 14
- one or more spectra indicative of the endpoint of the subject plasma process or discrete portion thereof e g , plasma process step
- the identification of the plasma process/process step itself may be included in or associated with step 448d
- the protocol set forth in step 450d addresses how the identification of the occurrence of the particular endpoint should be addressed This may include terminating the subject plasma process/process step, initiating the next plasma process/step (e g , if the subject plasma step is not the last step of a given plasma recipe or other process), or both depending upon the nature of the plasma process Automation and manual techniques are contemplated by step 450d as in the above-noted cases
- control of the plasma monitoring operations is relinquished by the process alert subroutine 432 of Figure 14 through steps 440 (if only the alarm alert function is enabled at step 454) or 462 (if the process control feature is enabled at step 436) Depending upon the circumstances, control may be returned to the particular sub-module of the current plasma process module 250 which called the process alert module 370.
- Another option which may be employed is to return control of the plasma monitoring operations in a particular case or in ail cases to the startup module 202 of Figure 7 through execution of steps 440 or 462.
- Startup Module 202 - Figures 15-16 Information on what is happening in the processing chamber 36 (e.g., spectra of the plasma in the chamber 36) is made available to the current plasma process module 250 for evaluation of the current plasma process operation through its various "sub-modules" as generally discussed above and as will be addressed in more detail below in relation to the following relevant figures. Access to the various "sub-modules" of the current plasma process module 250 may be controlled through the startup module 202 of Figure 15. As such, the startup module 202 may be viewed as a main menu of sorts for the various options that are available through the current plasma process module
- the startup routine 203 basically allows personnel to "enter” in some manner the type of action to be undertaken such that control of the plasma monitoring operations may be transferred to the appropriate sub-module of the current plasma process module 250. "Entry” may be accomplished by providing a listing on the display 130 associated with the PMCU 128 (e.g., Figure 6) of all of the actions which may be undertaken and allowing personnel to select which option should be pursued with the data entry device 132. Another option would be to allow personnel to input the action to be initiated using the data entry device 132.
- startup routine 203 may just immediately begin comparing the current plasma process to the plasma spectra directory 284 (e.g., using an appropriate order for searching the various subdirectories).
- step 136 of the startup routine 203 accesses a calibration module 562 through execution of step 140
- the calibration module 562 will be discussed in more detail below in relation to Figures 40-48
- Research in relation to the current plasma process to be run in the chamber 36 may be initiated through step 144
- research may be undertaken to identify one or more characteristics which are indicative of the endpoint of a particular plasma process or plasma process step This is accomplished through execution step 148 of the startup routine 203 which calls a research module 1300 which will be discussed in more detail below in relation to Figures 49-51 C
- a final option available through the startup routine 203 of Figure 15 relates to current plasma processes (i e , any plasma process run in the chamber 36 which is not recorded in the plasma spectra directory 284)
- Plasma processes such as plasma processing qualification/production wafers (step 230), plasma cleaning operations without first doing a wet clean of the chamber 36 (step 234), plasma cleaning operations conducted after the chamber 36 has been wet cleaned (step 238), and conditioning wafer operations (step 242) each may be accessed through the startup routine 203
- the endpoint of these types of plasma processes, a specific portion thereof, or both may be determined through an endpoint detection module 1200 which will be discussed in more detail below in relation to Figures 52-58 and which is called by step 240 of the startup routine 203
- the "health" of these types of plasma processes may also be evaluated through the plasma health module 252 which will be discussed in more detail below in relation to Figures 21-25 and which is called through execution of step 236 of the startup routine 203
- Step 236 of the startup routine 203 of Figure 15 relates to a plasma health evaluation and calls the startup subroutine 204 of Figure 16
- Two mam options may be pursued in relation to "plasma health" through the startup subroutine 204 of Figure 16
- Either a current plasma process may be recorded in the normal spectra subdirectory 288 to be used as a standard for evaluating plasma processes subsequently run in the chamber 36, or the current plasma process may be evaluated against the normal spectra subdirectory 288
- step 208 of the startup subroutine 204 of Figure 16 inquires as to whether the plasma process to be conducted in the subject processing chamber 36 should be recorded in the normal spectra subdirectory 288 associated with this chamber 36 If the "response" to the inquiry of step 208 is a "yes" the startup subroutine
- step 224 a determination is made as to the status of the plasma in the processing chamber 36 - specifically whether the plasma is "on” through optical analysis by the current plasma process module 250
- One way to determine if the plasma is "on” in the chamber 36 is to determine when the spectra obtained from the processing chamber 36 "matches” any spectra stored in the plasma spectra directory 284 or any of its subdirectories, such as through the pattern recognition module 370 of Figure 13
- Another way in which this may be done is to determine when any of the spectra from the interior of the processing chamber 36 have at least a certain number of discrete peaks of a least a certain intensity Using the same principles discussed above in relation to the pattern recognition module 370 of Figure 15 may also identify this type of spectra through the current plasma process module 250 Determining when there is at least a certain change in the optical emissions from within the chamber 36 may also be indicative that the plasma is "on” (e g , going from a "dark” condition to a "light” condition) Regardless of how the
- the startup subroutine 204 proceeds to step 228 where at least spectral data of the current plasma process is recorded in the normal spectra subdirectory 288 Preferably this encompasses the Preferred Optical Bandwidth at the Preferred Data Resolution and using the Preferred Data Collection Time Resolution After the plasma process is terminated, the subroutine 204 returns to the "mam menu-like" startup routine 203 of Figure 15 via step 226
- the other alternative available through the startup subroutine 204 of Figure 16 is to evaluate the current plasma process to be run in the subject processing chamber 36 against spectral data already recorded in the normal spectra subdirectory 288 In the example presented in Figure 16, this is accomplished by exiting step 208 of the startup subroutine 204 under a "no" logic condition, which directs the startup subroutine 204 to proceed to step 212
- Step 212 inquires as to whether the plasma process to be run in the processing chamber 36 should be evaluated against spectral data in the normal spectra subdirectory 288 Since the two mam options available under the startup subroutine 204 are to either record data or evaluate spectral data of the current plasma process against spectral data already recorded in the normal spectra subdirectory 288, and further since the subroutine 204 reached step 212 because a "decision" was made at step 208 of the subroutine 204 to not record data in the normal spectra subdirectory 288, a response of "no" at step 212 merely redirects the startup
- Plasma Health Evaluations The plasma health module 252 of Figures 7 is also included in the embodiment of Figure 32 and evaluates the overall health of the plasma in the subject chamber 36 or the "plasma health " "Plasma health” as used herein means the state or the condition of the plasma process as it relates to plasma performance when compared to typical "normal” plasma behavior resulting in usable product
- the "condition” of the plasma in turn may be characterized as the cumulative result of all parameters having an effect on the plasma in the processing chamber Stated another way, "plasma health” may be equated with a condition where a current plasma process is proceeding in accordance with one or more plasma processes stored in the normal spectra subdirectory 288
- the plasma health module 252 is able to determine if the current plasma process being conducted within the processing chamber 36 is progressing "normally” through a comparison of at least a portion of the optical emissions from the processing chamber 36 during the plasma process with the relevant spectra or portion thereof in the normal spectra subdirectory 288
- Spectral patterns of the plasma in the chamber 36 will change as the plasma process progresses Moreover, the spectral patterns of the plasma in the chamber 36 differ in relation to the category of plasma process being run This is evidenced by a review of exemplary spectra from a plasma recipe, a plasma clean conducted without first wet cleaning the chamber 36, a plasma clean executed after a wet clean of the chamber 36, and a conditioning wafer operation presented below In each case, “intensity” is plotted along the "y” axis and expressed in “counts” which is reflective of the intensity level, while “wavelength” is plotted along the "x" axis in nanometers Exemplary Plasma Recipe Spectra - Figures 17A-C
- FIG. 17A-C An example of a multiple step plasma recipe run on a wafer 18 in the chamber 36 is illustrated in Figures 17A-C in which the spectra of the plasma in the processing chamber 36 varies with a change in the current plasma step
- Figures 17A-C present a spectra 744 of an exemplary first plasma step of an exemplary plasma recipe A, a spectra 752 of an exemplary second plasma step of this same plasma recipe A, and a spectra 760 of an exemplary third plasma step of this same plasma recipe A, respectively
- Each of these spectra 744, 752, and 760 are characterized by a number of peaks 748, 756, and 764, respectively, of varying intensities at various wavelengths
- a comparison of the spectra 744, 752, and 760 reveals that their associated patterns differ, including without limitation as follows 1 ) at about the 425 nanometer wavelength region, peak 748a in the spectra 744 of Figure 17A has an intensity of about 3,300, peak 756a in the
- Figure 17A has an intensity of about 3,200
- peak 756a in the spectra 752 of Figure 17B has an intensity of about 3,900
- there is no peak in the spectra 760 of Figure 17C but the corresponding intensity (noise) is about 500, 3) at about the 525 nanometer wavelength region
- peak 748c in the spectra 744 of Figure 17A has an intensity in excess of 4,000
- peak 756c in the spectra 752 of Figure 17B has an intensity of about 3,400
- peak 764c in the spectra 760 of Figure 17C has an intensity of about 2,750, 4) at about the 587 nanometer wavelength region
- there is no peak in the spectra 744 of Figure 17A but the intensity is about 500 (noise)
- peak 764d in the spectra 760 of Figure 17C has an intensity of about 3,000
- Figures 18A-C Representative spectra are presented in Figures 18A-C to illustrate how the optical emissions of the plasma within the processing chamber 36 changes over time during a plasma clean conducted without first doing a wet clean of the chamber 36
- Figure 18A presents a spectra 770 of an exemplary plasma when the processing chamber 36 is in a dirty chamber condition and while plasma is present in the process chamber 36 without any product therein
- Figure 18B presents a spectra 774 of this same exemplary plasma at an intermediate time of the plasma clean in which the dirty chamber condition has begun to be addressed by the plasma clean
- Figure 18C presents a spectra 778 of this same exemplary plasma at the end of the plasma clean at which time the interior of the processing chamber 36 is deemed to be in condition to return to commercial production (e g , to etch integrated circuit designs on a production wafer 18)
- This spectra 778 may be selected by the operator of the facility implementing the wafer production system 2 as being indicative of the chamber 36 being in proper condition for resumption of production
- [02 778 as being indicative of a 'clean chamber condition" may be somewhat arbitrary
- Each of the spectra 770, 774, and 778 are characterized by a number of peaks 772, 776, and 780, respectively, of varying intensities at various wavelengths
- a comparison of the spectra 770, 774, and 778 reveals that their associated patterns are in fact different, including without limitation as follows 1 ) at about the 625 nanometer wavelength region, peak 772e in the spectra 770 of Figure 18A has an intensity of about 500, peak 776e in the spectra 774 of Figure 18B has an intensity of about 300, and there is no substantial peak in the spectra 778 of Figure 18C, 2) at about the 675 nanometer wavelength region, peak 772f in the spectra 770 of Figure 18A has an intensity of about 4,000, peak 776f in the spectra 774 of Figure 18B has an intensity of about 1 ,000, and there is no substantial peak in the spectra 778 of Figure 18C, and 3) at about the 685 nanometer wavelength region, peak 772g in the spectra
- Figure 19A presents a spectra 1328 of the exemplary plasma in the processing chamber at the start of such a plasma cleaning of the chamber 36
- Figure 19B presents a spectra 1336 of the exemplary plasma at an intermediate point in such a plasma cleaning of the chamber 36
- Figure 19C presents a spectra 1344 of the exemplary plasma at the end of such a plasma cleaning of the chamber 36
- Each of the spectra 1328, 1336, and 1344 are characterized by a number of peaks 1332, 1340, and 1348, respectively, of varying intensities at various wavelengths
- a comparison of the spectra 1328, 1336, and 1344 reveals that their respective patterns are different, including without limitation as follows 1 ) at about the 625 nanometer wavelength region
- peak 1332e in the spectra 1328 of Figure 19A has an intensity of about 600
- More than one entry of a plasma clean may be required in the normal spectra subdirectory 288 depending upon a variety of factors
- the spectral data of a plasma clean run on a chamber 36 after a wet clean may look different than a plasma clean that is run on a new chamber 36 which has not been wet cleaned
- the spectral data of a plasma clean which is run after the chamber 36 has been running a first type of plasma recipe may look different than a plasma clean which is run after the chamber 36 has been running a second type of plasma recipe which is different from the first type of plasma recipe
- Figure 20A presents a spectra 1288 of an exemplary plasma in the processing chamber 36 at the start of a conditioning wafer operation
- Figure 20B presents a spectra 1292 of an exemplary plasma at an intermediate point in the conditioning wafer operation
- Figure 20C presents a spectra 1296 of an exemplary plasma at the end of the conditioning wafer operation.
- Each of the spectra 1288, 1292 and 1296 are characterized by a number of peaks 1290, 1294, and 1298, respectively, of varying intensities at various wavelengths.
- a comparison of the spectra 1288, 1292, and 1296 reveals that there are certain differences in their respective patterns, including without limitation as follows: 1 ) at about the 440 nanometer wavelength region, peak 1290a in the spectra 1288 of Figure 20A has an intensity of about 3,550, peak
- peak 1294a in the spectra 1292 of Figure 20B has an intensity of about 3,750, and peak 1298a in the spectra 1296 of Figure 20C has an intensity of about 4,000; 2) at about the 525 nanometer wavelength region, peak 1290b in the spectra 1288 of Figure 20A has an intensity of about 2,800, peak 1294b in the spectra 1292 of Figure 20B has an intensity of about 2,900, and peak 1298b in the spectra 1296 of Figure 20C has an intensity of about 2,800; 3) at about the 595 nanometer wavelength region, peak 1290d in the spectra 1288 of Figure 20A has an intensity of about 2,100, peak 1294d in the spectra 1292 of Figure 20B has an intensity of about 2,150, and peak 1298d in the spectra 1296 of Figure 20C has an intensity of about 2,125; 4) at about the 675 nanometer wavelength region, peak 1290e in the spectra 1288 of Figure 20A has an intensity of about 600, peak 1294e in the spectra 12
- spectra 1288, 1292, and 1296 show that the progression of a conditioning wafer operation is evident in the spectral pattern of the plasma in the chamber 36 during the operation. More than one entry of a conditioning wafer operation may be required in the normal spectra subdirectory 288 depending upon a variety of factors. For instance, the spectral data of a conditioning wafer operation which is run after the chamber 36 has only been plasma cleaned may look different than the spectral data of a conditioning wafer operation which is run in the chamber 36 after it has been both plasma cleaned, wet cleaned, and then again plasma cleaned.
- the spectral data of a conditioning wafer operation run after the chamber 36 has been running a first type of plasma recipe may look different than a conditioning wafer operation which is run after the chamber 36 has been running a second type of plasma recipe which is different from the first type of plasma recipe.
- Plasma Health Module 252 - Figures 21-25 The current plasma process module 250 of Figures 7 and 32 is available for monitoring the health of any plasma process which is conducted within the processing chamber 36 first through a comparison of at least a portion of its spectral data with at least a portion of the spectral data stored in the normal spectra subdirectory 288 ( Figure 9). Plasma recipes (whether run on production wafers 18 or qualification wafers 18), plasma cleans (with or without wet cleans), and conditioning wafer operations, as well as the health of any other plasma process, may each be evaluated through the plasma health module 252.
- the plasma health module 252 deals with the presence of having different categories of plasma processes stored in the normal spectra subdirectory 288, as well as in the abnormal spectra subdirectory 292 and the unknown spectra subdirectory 296 which are also used in the plasma health evaluation of a current plasma process being run in the chamber 36, is really a matter of preference. Some ways of dealing with the existence of multiple categories of plasma processes may affect the speed of the evaluation by the plasma health module 252 more than others. For instance, the plasma health module 252 may limit its comparison of the current plasma process to the same category or genus of plasma processes stored in the normal spectra directory 288 and abnormal spectra directory 292.
- identifying information may be input into the plasma process genus field 322h ( Figure 12A) associated with each plasma process stored in the normal spectra subdirectory 288 and the plasma process genus field 333e ( Figure 12B) associated with each plasma process (or portion thereof) stored in the abnormal spectra subdirectory 292.
- the current plasma process to be conducted in the chamber 36 may be identified to the plasma health module 252 in some manner. This may be accomplished through the startup module 202 of Figure 15 (e.g., through including appropriate process category or genus identifying information in steps 230, 234, 238, and 242 of the startup subroutine 203, which is passed onto step 236 of the subroutine 203, and which may be then passed on to the plasma health module 252).
- Plasma health is also preferably evaluated by comparing optical emissions from the current plasma process in the chamber 36 with the plasma spectra directory 284 over at least those wavelengths within the Preferred Optical Bandwidth based upon the Preferred Data Resolution, and using the Preferred Analytical Wavelength Resolution.
- some subset of the optical emissions data of the plasma in the chamber 36 may be used to monitor the plasma health.
- One such circumstance is when processing speed is or potentially is an issue.
- the data within the abnormal spectra subdirectory 292 of Figure 9 may be used to generate the subset of data which may be reviewed for purposes of monitoring the plasma health.
- the plasma health evaluation may be conducted over optical emissions segments which include those wavelengths which are indicative of errors which occurred in processes previously conducted within the chamber 36.
- One alternative is to define an optical emissions segment ⁇ 25 nanometers on each side of each wavelength of a spectra in the abnormal spectra subdirectory 292 which is indicative of an error from a previous plasma process.
- the plasma health may be evaluated by looking at each of the 300-350, 400-450, and 550-600 nanometer region.
- a smaller optical emissions segment for monitoring plasma health may also be selected by defining a range which includes each of those wavelengths which are indicative of errors from the abnormal spectra subdirectory 292. For instance, if errors from previous runs are reflected at the 325, 425, and 575 wavelengths, the plasma health may be evaluated by looking at the wavelength region from about 325 nanometers to about 575 nanometers.
- each of the endpoints of this range may be desirable to include a "buffer" on each of the endpoints of this range as well (e.g., extend by about 25 nanometers on each end of the range).
- the above may be further limited by limiting the plasma health evaluation to those optical emissions segments which include only errors from the same type of plasma process which is to be run in the chamber 36 (e.g., same plasma recipe).
- information on endpoint of the plasma process or discrete portion thereof may be used to define the wavelengths to be evaluated in relation to plasma health.
- endpoint may be called based upon a change at one or more specific wavelengths.
- Plasma health may be evaluated by looking at a ⁇ 25 nanometer region around each wavelength which is used to call endpoint.
- Plasma Health Subroutine 253 - Figure 21 One embodiment of a subroutine is illustrated in Figure 21 which may be used by the plasma health module 252 to evaluate whether a current plasma process is proceeding in accordance with at least one plasma process stored in the normal spectra subdirectory 288 of Figure 9 (e.g., indicative of a "healthy" plasma). Summarily, spectral data is taken during and more preferably throughout the entirety of the execution of the current plasma process which is being run within the processing chamber 36. Consideration should be given to the first part of a plasma process being somewhat unstable.
- Spectral data from the current plasma process is first compared against the normal spectra subdirectory 288 to determine if the current plasma process "matches” any plasma process stored within the normal spectra subdirectory 288. As long as the current plasma process "matches” at least one plasma process stored in the normal spectra subdirectory 288, the current plasma process is characterized as being "normal” or “healthy” and the plasma health subroutine 253 will continue to limit its search for "matching" spectra to the normal spectra subdirectory 288. However, oftentimes there is an error or aberration during a plasma process which may have some type of adverse effect on the desired end result of the plasma process, and this should be identifiable from the spectra of the plasma in the chamber 36.
- the spectra of the plasma in the chamber 36 should no longer "match” any plasma process stored in the normal spectra subdirectory 288.
- the plasma health subroutine 253 will then discontinue its search of the normal spectra subdirectory 288 for evaluating the current plasma process and start comparing the current plasma process with the abnormal spectra subdirectory 292 of Figure 9. Errors or aberrations in plasma processes which have been encountered before by the plasma health subroutine 253 on this same chamber 36, and which have had their corresponding cause or causes identified, are recorded in the abnormal spectra subdirectory 292.
- Actions which may be initiated if spectral data of the current plasma process "matches" at least one spectra in the abnormal spectra subdirectory 292 range from issuing an appropriate alert to addressing one or more process control features of the wafer production system 2 as discussed above in relation to the process alert subroutine 432 of Figure 14. All of the data in the normal spectra subdirectory 288 and the abnormal spectra subdirectory 292 is obtained from the processing chamber 36 on which the plasma health module 252 is being used to evaluate any plasma process currently being run or which was run in this very same chamber 36.
- Plasma processes, whether are run on qualification or production wafers 18, in the processing chamber 36 are evaluated by the plasma health module 252.
- the plasma health module 252 is able to complete its evaluation of a plasma recipe which was run on a production wafer 18 before the plasma recipe is initiated on the next production wafer 18 since the plasma health module 252 effectively relies on pure pattern recognition techniques, and not chemical analysis or chemical species identification techniques.
- the plasma health module 252 is able to not only determine the identify of the plasma process, but to determine that the plasma process is being run on a qualification wafer 18 versus a production wafer 18.
- step 254 is referenced in Figure 21 in relation to merely a "spectra” or optical emissions data, as noted above, other types of data may be taken at/associated with this time as well (e.g., the time into the plasma recipe at which the associated spectra was obtained from the chamber 36).
- a comparison is then made at step 258 of the plasma health subroutine 253 between the spectra of the current plasma process obtained at step 254 (current plasma process) and the relevant spectra from the normal spectra subdirectory 288 (stored plasma process).
- the normal spectra directory 288 will hereafter be described as having only a single plasma process stored therein ("Recipe A").
- Step 258 of the plasma health subroutine 253 calls the pattern recognition module 370 of Figure 13 to undertake a comparative analysis between the
- step 258 of the subroutine 253 setting the Target Directory used by the pattern recognition module 370 to the normal spectra subdirectory 288. Only the normal spectra subdirectory 288 is then searched by the pattern recognition module 370 at this time, through execution of step 258 of the plasma health subroutine 253, to determine if there is a "match" between the current spectra at the current time t c and Recipe A as stored in the normal spectra subdirectory 288. Which spectra of Recipe A is actually compared with the current spectra in this instance is addressed below after the discussion of the loop 190 of the subroutine 253 is completed.
- the pattern recognition module 370 of Figure 13 returns control of the plasma monitoring operation back to the plasma health subroutine 253 of Figure 21 after the pattern recognition module 370 has determined whether there is a "match" between the spectra at the current time t c (from step 254 of the plasma health subroutine 253) and the relevant spectra of Recipe A from the normal spectra subdirectory 288 in the subject example.
- the result ("match” or "no match") of the analysis by the pattern recognition module 370 is actually provided to step 260 of the plasma health subroutine 253 of Figure 21. If the current spectra at the current time t c was a "match" with the relevant spectra of
- the evaluation by the plasma health subroutine 253 will continue in relation to the normal subdirectory 288.
- the plasma health subroutine 253 inquires at step 261 as to whether the current plasma process being conducted in the processing chamber 36 has been terminated, or more accurately if there are any more spectra from the subject current plasma process being run in the chamber 36 to be evaluated by the subroutine 253.
- Other information on the subject current plasma process may be provided through execution of step 194 of the plasma health subroutine 253 which calls other sub- modules of the current plasma process module 250 as will be discussed in more detail below (e.g., to access a chamber condition evaluation function, to access an endpoint determination function).
- Step 278 of the subroutine 253 more specifically provides for adjustment of the "clock” by a predetermined increment “n” to thereby increase the current time t c by an increment of "n "
- the magnitude of "n” defines that portion of the collected data which will be analyzed All of the data may be analyzed, or only a portion thereof (e g , only every other "piece” of optical emissions data may actually be analyzed)
- this concept will be referred to as the Analytical Time Resolution
- the Analytical Time Resolution in relation to plasma health is at least at every 1 second, and more preferably at least at every 300 milliseconds
- the plasma health subroutine 253 then returns to step 254 where the next spectra of the plasma in the processing chamber 36 from the execution of the current plasma process is obtained for the subroutine 253 at the new current time t
- a time dependency requirement is an acceptable way to evaluate whether a current plasma process is proceeding in accordance with any one or more plasma processes stored in the normal spectra subdirectory 288. From a practical standpoint this is not necessarily the case. Variations throughout the wafers 18 in a given wafer cassette 6 on which the same plasma recipe is typically run may affect the amount of time required to complete one or more of the plasma steps of the current plasma recipe being run in the chamber
- the thickness of a certain layer to be etched away by a certain step of the plasma recipe may vary from wafer 18 to wafer 18 within an acceptable tolerance.
- Conditions within the chamber 36 may also have an effect on the amount of time required to reach the endpoint of one or more plasma steps of a given plasma recipe or any other plasma process for that matter.
- the performance of the chamber 36 may change. Changing the performance of the chamber 36 may change the amount of time required to reach the endpoint of one or more steps of a given plasma recipe. Other factors may affect timing issues associated with other types of plasma processes run in the chamber 36.
- the plasma health subroutine 253 issuing false alarms, or more specifically an indication that a current plasma process does not conform with at least one plasma process stored in the normal spectra subdirectory 288 when such is not the case.
- Relevant in the context of the plasma health subroutine 253, and in fact for each sub- module of the current plasma process module 250, may simply be whether the current plasma process being run in the chamber 36 is progressing in a manner consistent with at least one of the plasma processes stored in the normal spectra subdirectory 288, although not necessarily at the same speed and therefore not being time dependent
- this will be referred to as a "progression dependency requirement” and is exemplified by the following
- the first spectra obtained for the plasma health subroutine 253 at the current time t is compared with one or more spectra of Recipe A in the normal spectra subdirectory 288 in the subject example
- the spectra of Recipe A which matched the spectra of the plasma in the chamber 36 at the current time t,, and which has the earliest time associated therewith, is identified as the current status spectra of Recipe A This accounts for the possible, although likely improbable, situation where a spectra at time t, in a given plasma process is substantially the same as
- the loop 190 defined by steps 254, 258, 260, 261 , and 278 of the plasma health subroutine 253 will continue to be re-executed until one of two events occurs.
- One event which will cause the plasma health subroutine 253 to exit the loop 190 is when all of the spectral data on the current plasma process has been evaluated by the subroutine 253 in accordance with the foregoing. That is, the entirety of the current plasma process run in the processing chamber 36 proceeded in accordance with at least one of the plasma processes recorded in the normal spectra subdirectory 288 ("at least one" referring to the fact that more than one entry of a given plasma process may be included in the normal spectra subdirectory 288), or in the subject example Recipe A.
- Control of the plasma monitoring operations is then transferred from the plasma health subroutine 253 back to, for instance, the startup module 202 of Figure 15 through execution of step 279 of the plasma health subroutine 253.
- the results of "normal" runs may be recorded in a "normal run” log file. Data such as that presented in Figure 12A may be included in this "normal run” log file and will provide a historical record of the subject plasma process. If data storage space is an issue, the spectral data may be omitted from the historical record although it is desirable to retain this data.
- this historical data need not be stored for access by the current plasma process module 250. For instance, the historical data may be stored on a network associated with the wafer production system 2 or any other data storage area.
- the plasma health subroutine 253 may also exit the loop 190 (the evaluation of the current spectra at the current time t c in relation to the normal spectra subdirectory 288 of Figure 9) when this current spectra is no longer a "match" with any plasma process stored in the normal spectra subdirectory 288. This would be the case when at some point in time the current plasma recipe being run on product within the processing chamber 36 was not a "match" with
- Recipe A of the normal spectra subdirectory 288 in the subject example The results of the pattern recognition module 370 of Figure 13 provided back to step 260 of the plasma health subroutine 253 in this case would cause the subroutine 253 to proceed from step 260 to step 266.
- Step 266 of the plasma health subroutine 253 calls the pattern recognition module 370 of Figure 13 to undertake a comparative analysis between the spectra of the plasma in the chamber 36 at the current time t c and the relevant spectra of the abnormal spectra subdirectory 292. This is accomplished by step 266 of the subroutine 253 setting the Target Directory used by the pattern recognition module 370 to the abnormal spectra subdirectory 292. Only the abnormal spectra subdirectory 292 is then searched by the pattern recognition module 370 through execution of step 266 to determine if there is a "match" between the current spectra of the plasma in the chamber 36 at the current time t c and the relevant spectra stored in the abnormal spectra subdirectory 292.
- Each spectra stored within the abnormal spectra subdirectory 292 may and preferably does have a time associated therewith, which is the time into the plasma process in which the spectra was obtained from within the chamber 36 (i.e., its corresponding t c ).
- the search of the abnormal spectra subdirectory 292 for "matches" by the pattern recognition module 370 may be limited to those spectra which were recorded at the same current time t c or within a predetermined amount of time on each side of the subject current time t c (e.g., ⁇ "x" seconds of the subject current time t c ).
- the point-by-point analysis embodied by step 386 of the pattern recognition subroutine 374 of Figure 13 may be performed in relation to only those spectra within the abnormal spectra subdirectory 292 which were also recorded at the same 20 second time period or within ⁇ 10 seconds (or any other desired amount) of this time period.
- Another subset of the abnormal spectra subdirectory 292 which may be used as a refining search criteria is the plasma process category/genus, or even the plasma process type or species within a plasma process category/genus having multiple types/species of plasma processes (e.g., a specific type of plasma recipe). That is, only those spectra in the abnormal spectra subdirectory 292 which are associated with a plasma process which is at least possibly the same as that currently being run on product within the processing chamber 36 will be analyzed by the pattern recognition module 370 using a plasma process criterion.
- Plasma health module 252 may not have narrowed down the identification of the current plasma process to a single plasma process within the normal spectra subdirectory 288. How the plasma health module 252 may identify a current plasma process being run on product in the processing chamber 36 is addressed below in relation to the plasma health/process recognition subroutines 790, 852, and 924 of Figures 22- 24.
- the plasma step of a plasma process may also be used as a refining search criterion for which spectra of the abnormal spectra subdirectory 292 are analyzed by the pattern recognition module 370.
- the plasma health module 252 may identify a current plasma step of a current plasma process being run in the processing chamber 36 is through the plasma health/process step recognition subroutine 972 which will be discussed below in relation to Figure 25. Any combination of the foregoing may be used as initial search criteria to initially refine the search of the abnormal spectra subdirectory 292. Finally, no refining search criteria need be used. That is, the search of the abnormal spectra subdirectory 292 for "matches" may compare the current spectra at the current time t c with each spectra within the abnormal spectra subdirectory 292, thereby removing both the time element and plasma process category/plasma process type within a given plasma process category/plasma step element as required "initial match" criteria.
- the pattern recognition module 370 returns control of the plasma monitoring operation back to the plasma health subroutine 253 of Figure 21 after the pattern recognition module 370 has determined whether there is a "match” between the spectra of the plasma in the chamber 36 at the current time t c (from step 254 of the plasma health subroutine 253) and the relevant spectra from the abnormal spectra subdirectory 292.
- the result (“match” or “no match”) of the analysis by the pattern recognition module 370 is provided to step 276 of the plasma health subroutine 253 of Figure 21.
- the plasma health subroutine 253 effectively takes two actions and these actions may be undertaken in any order, including simultaneously.
- One of these actions is that the plasma health subroutine 253 will proceed to step 274 which calls the process alert module 428 which was discussed above in relation to Figure 14.
- alerts may be issued as to the identification of the abnormal condition, control of the wafer production system 2 may be addressed, or both through the process alert module 428.
- Another action which is taken by the plasma health subroutine 253 in this type of case is to record data of the remainder of the plasma process in an "abnormal run" log file for historical purposes.
- spectral data is recorded in an "abnormal run” log file through execution of step 264 for the current time t c (the first spectra which did not "match” any relevant plasma process stored in the subdirectory 288, but which did "match” at least one entry in the abnormal spectra subdirectory 292). Proceeding from step 264 to step 265 of the plasma health subroutine 253 of Figure 21 , a determination is made as to the status of the current plasma process.
- any termination of the plasma process, or more accurately recordation of data on the remainder of the process, will cause the subroutine 253 to proceed to step 267 where control of the plasma monitoring operations may be returned to, for instance, the startup module 202 of Figure 15 Continuation of the plasma process after the error is identified will cause the subroutine 253 to proceed from step 265 to step 268 where the current time t c is increased by a factor of "n" such that another spectra of the plasma in the chamber 36 can be obtained for the subroutine 253 at this new current time t c through step 272 for recordation in the "abnormal run” log file
- the magnitude of "n” may be the Preferred Analytical Time Resolution Steps 264, 265, 268, and 272 will continue to be repeated in the described manner to continue to record data in the "abnormal run” log file until the current plasma process is terminated, at which time the subroutine 253 will exit at step 267 as described
- the plasma health subroutine 253 of Figure 21 handles this type of situation by exiting step 276 to where the subroutine 253 effectively takes two actions, and these actions may be undertaken in any order and including simultaneously One of these actions is that the plasma health subroutine 253 will execute step
- alerts may be issued as to the existence of the unknown condition, control of the wafer production system 2 may be addressed, or both through the process alert module 428
- Another "action" which is taken by the plasma health subroutine 253 in this type of case is to record data of the remainder of the plasma process in the unknown spectra subdirectory 296
- spectral data is recorded in the unknown spectra subdirectory 296 through execution of step 270 for the current time t e (the first spectra which did not "match” any relevant plasma process stored in the normal spectra subdirectory 288, and which also did not “match” any relevant entry in the abnormal spectra subdirectory 292) Proceeding from step 270 to step 283 of the plasma health subroutine 253, a determination is made if the current
- I20 plasma process has been terminated (e g , is the plasma "off' in the chamber 36) Any termination of the plasma process will cause the subroutine 253 to proceed to step 281 where control of the plasma monitoring operations may be returned to, for instance, the startup module 202 of Figure 15 Continuation of the plasma process after the unknown condition is encountered will cause the subroutine 253 to proceed from step 283 to step 280 where the current time t c is increased by a factor of "n" (e g , Preferred Data Collection Time Resolution) such that another spectra of the plasma in the chamber 36 can be obtained for the subroutine 253 at this new current time t c through step 282 for recordation in the unknown spectra subdirectory 296 Steps 270, 283, 280, and 282 will continue to be repeated in the described manner until the current plasma process is terminated or until data on the remainder of the process has been recorded in the subdirectory 296, at which time the subroutine 253 will exit at step 281 as described Plasma process runs which are recorded in
- Plasma Health/Process Recognition Subroutine 790 - Figure 22 Another embodiment of a subroutine which may be used by the plasma health module 252 is presented in Figure 22 Not only is the health or condition of the plasma assessed by the subroutine 790 in Figure 22, but the particular plasma process which is being run in the chamber 36 is also identified. That is, the subroutine 790 is able to determine the identify of the plasma process (e.g., to distinguish between different types of plasma recipes, to distinguish between the same plasma recipe run on a production wafer 18 and a qualification wafer 18, etc). Consequently, the subroutine 790 is characterized as a plasma health/process recognition subroutine 790.
- the plasma health/process recognition subroutine 790 also presents one way in which a current plasma process being run in the subject chamber 36 may be evaluated against multiple plasma processes stored in the normal spectra subdirectory 288 of Figure 9. These very same principles may be implemented in the plasma health subroutine 253 of Figure 21.
- a number of prerequisites are addressed before the plasma health/process recognition subroutine 790 actually initiates its analysis of the current plasma process being run in the processing chamber 36.
- the order in which these steps are executed is not important to the present invention.
- Initialization of the plasma health/process recognition subroutine 790 includes setting the Target Directory associated with the pattern recognition module 370 of Figure 13 to the normal spectra directory 288 of Figure 9 at step 796 of the subroutine 790.
- the pattern recognition module 370 is used by the subroutine 790 to compare the pattern of a "run spectra" (i.e., a spectra of the plasma from the processing chamber 36 during a plasma process being run in the processing chamber 36) with the relevant spectra of the plasma processes stored in the normal spectra subdirectory 288.
- Preparation for the analysis of the current plasma process being run in the chamber 36 by the plasma health/process recognition subroutine 790 also requires execution of step 816 which "calls up” or “flags" the first plasma process in the normal spectra subdirectory 288 which is to be compared with the current plasma process by the subroutine 790.
- the logic of the subroutine 790 is to compare the current plasma process being run in the processing chamber 36 with only one plasma process stored in the normal spectra subdirectory 288 at a time. That is, the subroutine 790 will first compare the current plasma process with Process A in the normal spectra subdirectory 288.
- the subroutine 790 will compare the entirety of the subject current plasma process with Process B in the normal spectra subdirectory 288. Only if the current plasma process deviates from Process B will other plasma processes stored in the normal spectra subdirectory 288 be compared one at a time with the current plasma process by the plasma health/process recognition subroutine 790.
- the plasma health/process recognition subroutine 790 may be configured to make all plasma processes stored in the normal spectra subdirectory 292 available for comparison with the current plasma process, or the above-noted refining criterion/criteria may be used.
- the first spectra of the plasma in the processing chamber 36 obtained for the plasma health/process recognition subroutine 790 is through execution of step 794 and which is also part of the initialization of the subroutine 790.
- This spectra is associated with the time t 0 (hereafter "start time t 0 "), and is stored along with each spectra obtained for the subroutine 790 until its analysis is completed. Any failure to retain the spectra of the current plasma process would not allow the subroutine 790 to use its "one process at a time" comparative logic.
- step 798 a current time t c variable is introduced, and further where this current time t c is set equal to the starting time t 0 .
- Comparison of the current plasma process being run in the processing chamber 36 with the data from step 816 is undertaken at step 800 where the plasma health/process recognition subroutine 790 is directed to proceed to the pattern recognition module 370 of Figure 13.
- An analysis of the spectra of the plasma from the processing chamber 36 at the current time t c is undertaken at step 800 to determine if the pattern of this current spectra is a "match" with the relevant spectra of Process A of the normal spectra subdirectory 288.
- the "match determination” is effectively a comparison of the patterns of the two noted spectra to determine if the pattern of the current spectra is sufficiently similar to the pattern of the relevant spectra of Process A from the normal spectra subdirectory 288 to be considered a "match” therewith.
- “Relevance” in terms of which spectra of a given plasma process from the normal spectra subdirectory 288 is compared with the spectra of the plasma in the chamber 36 at the current time t c by the subroutine 790 may be determined in accordance with either the time dependency requirement or the progression dependency requirement discussed above in relation to the plasma health subroutine 253 of Figure 21.
- the results of the analysis from step 800 are evaluated at step 812 of the plasma health/process recognition subroutine 790. If the spectra of the plasma in the processing chamber 36 associated with the current plasma process at the current time t c is a "match" with the relevant spectra of Process A in the normal spectra subdirectory 288, the subroutine 790 proceeds to step 802 where the results are displayed.
- an indication may be provided to operations personnel on the display 130 ( Figure 6), or by any of the other methods described above, that the plasma health/process recognition subroutine 790 has determined that the current plasma process being run in the processing chamber 36 corresponds, through the current time t c , with Process A. It may be inaccurate and therefore inadvisable at this point in time to indicate that the plasma process currently being run in the processing chamber 36 is definitively Process A.
- the comparison of the current plasma process with the normal spectra subdirectory 288 up to this time has been limited to Process A.
- the spectra of the plasma in the chamber 36 up through the current time t c could in fact also "match" the relevant spectra of one or more other plasma processes stored in the normal spectra subdirectory 288.
- step 806 of the plasma health/process recognition subroutine 790 determines if all of the data from the current plasma process has been evaluated by the subroutine 790 (e.g., has all the data obtained up until the plasma goes "off been evaluated).
- any continuation of the current plasma process or a failure to have examined all of its optical emissions data will cause the plasma health/process recognition subroutine 790 to proceed to step 804 which causes the current time t c to be adjusted by an increment of "n.”
- the magnitude of "n” defines the Analytical Time Resolution, and preferably the Preferred Analytical Time Resolution is implemented. For instance, if the start time to was set at one second (where the initial spectra reading was obtained for the subroutine 790 at step 794) and the variable "n" was set at two seconds, the current time t c upon exiting step 804 would be 3 seconds.
- the spectra at this new current time t c from the processing chamber 36 is then obtained for the subroutine 790 at step 808, and the subroutine 790 returns to step 800 where the pattern of this new spectra is compared with the pattern of the relevant spectra of Recipe A to determine if they "match" in accordance with the foregoing.
- Steps 800, 812, 802, 806, 804, and 808 define a loop 818 which continues to be executed to compare the current plasma process being run in the processing chamber 36 with one of the plasma processes stored in the normal spectra subdirectory 288 (Process A in the subject example) until one of two conditions exists. One of these conditions is where the current plasma process has been completed and "matched" an entire plasma process stored in the normal spectra subdirectory 288. In this case, the subroutine will exit from step 806 to step 810. Control of the plasma monitoring operations may be returned by step 810 to, for instance, the startup module 202 of Figure 15.
- Step 818 Another condition where the subroutine 790 will exit the loop 818 is when the spectra of the plasma in the processing chamber 36 at the then current time t c does not "match" with the relevant spectra of the plasma process stored in the normal spectra subdirectory 288 currently being used by the plasma health/process recognition subroutine 790 (Process A in the subject example).
- the subroutine 790 will exit from step 812 to step 814.
- Step 814 basically inquires as to whether each plasma process stored in the normal spectra subdirectory 288 has been compared with the current plasma process by the subroutine 790 through the loop 818.
- the plasma health/process recognition subroutine 790 will proceed from step 814 to step 822 where data on the next plasma process stored in the normal spectra subdirectory 288 is recalled in some manner for use by the subroutine 790.
- This data on a plasma process stored in the normal spectra subdirectory 288 is recalled for evaluation by the subroutine 790 against the current plasma process from the time ⁇ through the latest current time t c (i.e. , from the very beginning of this plasma process).
- the subroutine 790 returns to step 798 from step 822 where the current time t c is returned to the start time t 0 , and the loop 818 of the subroutine 790 is entered to evaluate the current plasma process against the next plasma process stored in the normal spectra subdirectory 288 in the above-described manner.
- step 814 the plasma health/process recognition subroutine 790 will exit step 814 and proceed to step 820.
- the protocol of step 820 generally directs the plasma health/process recognition subroutine 790 to determine if the current plasma recipe has encountered a known error/aberration which is stored in the abnormal spectra subdirectory 292.
- the plasma health/process recognition subroutine 790 may include the portion of the plasma health subroutine 253 of Figure 21 which pertains to the abnormal spectra subdirectory 292 and the unknown spectra subdirectory 296 for that matter (i.e., starting with step 266 of the subroutine 253 and including everything thereafter).
- the spectra of the current plasma process which is compared with the abnormal spectra subdirectory 292 in the manner discussed in relation to the plasma health subroutine 253 of Figure 21 would be the spectra following (in time) the spectra of the last current time t c which matched any of the plasma processes stored in the normal spectra subdirectory 288.
- t 0 is 1 second and "n" is 2 seconds
- the current plasma process "matched” with Process A up until time t 39 the current plasma process "matched” with Process B until time t 61
- the current plasma process "matched” with Process C until only time t 3 .
- the spectra at time t 61 would be that which is compared with the abnormal spectra subdirectory 292 in the manner described above in relation to the plasma health subroutine 253 of Figure 21. If the plasma process identified by the plasma health/process recognition subroutine 790 was a plasma recipe stored in the normal spectra subdirectory 288, a variation of the subroutine 790 may be implemented which may enhance the speed of the plasma health evaluation.
- the logic of the subroutine 790 may be modified such that the subroutine 790 would thereafter at least start its analysis of each subsequent plasma process run in the chamber 36 with that plasma recipe from the normal spectra subdirectory 288 which was previously identified by the subroutine 790.
- each wafer 18 subsequently processed could first be checked against the plasma recipe run for a production wafer 18 from the normal spectra subdirectory 288, and then against the plasma recipe run for a qualification wafer 18 from the normal spectra subdirectory 288.
- Plasma Health/Process Recognition Subroutine 852 - Figure 23 The plasma health/process recognition subroutine 790 of Figure 22 may be described as incorporating a "series" logic. That is, the comparison of the current spectra of the plasma in the processing chamber 36 at the current time t c is made in relation to only one plasma process stored in the normal spectra subdirectory 288 at a time.
- a plasma health/process recognition subroutine which may be used by the process health module 252 and which proceeds with a "parallel" logic is presented in Figure 23.
- the plasma health/process recognition subroutine 852 of Figure 23 begins at step 854 where the Target Directory for the pattern recognition module 370 of Figure 13 is set to the normal spectra subdirectory 288 (i.e., the search for "matching" spectra will initiate in the normal spectra subdirectory 288).
- Another preliminary step of the plasma health/process recognition subroutine 852 is at step 856 where a logic operator Flag 2 is set to "T" for each of the plasma processes stored in the normal spectra subdirectory 288 to be evaluated through the subroutine 852.
- the order in which steps 854 and 856 are executed is not particularly important in relation to the present invention.
- Data relating to the current plasma process being run on product in the processing chamber 36 is obtained for the plasma health/process recognition subroutine 852 at step 860. Included in this data is at least a spectra of the plasma within the processing chamber 36 during the execution of a plasma process within the processing chamber 36 at the current time t c which was obtained from the chamber 36 over the Preferred Optical Bandwidth and at the Preferred Data Resolution.
- a comparison is thereafter made of the pattern of this current spectra at the current time t c with the relevant spectra of each plasma process stored in the normal spectra subdirectory 288 which has matched the current plasma process being run in the processing chamber 36 up until the now current time t c , and this comparison is made before spectra associated with a later current time t c is compared with these plasma processes stored in the normal spectra subdirectory 288
- the current plasma process is effectively concurrently compared with each plasma process stored in the normal spectra subdirectory 288 which has "matched" the current plasma process being run in the processing chamber 474 up until the current point in time If the current spectra at a current time t c does not match a particular plasma process stored in the normal spectra subdirectory 288, this plasma process is dropped from the list of possible plasma processes and spectra at new, later in time, current times t c are no longer compared with this plasma process "Relevance" in terms of which spectra of
- the spectra at the current time t c from step 860 of the plasma health/process recognition subroutine 852 is effectively concurrently compared with each of the relevant plasma process stored in the normal spectra subdirectory 288 the first time through the mam body of the plasma health/process recognition subroutine 852
- the logic operator "Flag 2 " associated with each such plasma process has been set to "T" at the previous step 856, so the subroutine 852 will proceed through steps 864 (Process A), 880 (Process B), and 892 (Process "X") to steps 868 (Process A), 884 (Process B), and 892 (Process "X") to where the subroutine 852 is directed to proceed to the pattern recognition module 370 of Figure 13
- the pattern recognition module 370 determines if the pattern of the current spectra at the current time t c is a "match" with the relevant spectra of the subject plasma process stored in the normal spectra subdirectory 288 (Process A in the
- Some plasma processes stored in the normal spectra subdirectory 288 being used by the subroutine 852 sooner or later will fail to "match” the current plasma process being run in the processing chamber 36 That is, the pattern recognition module 370 will determine that the pattern of the spectra at the current time t c does not match the relevant spectra of the subject plasma process stored in the normal spectra subdirectory 288 One or more of steps 868, 884, and 896 will then exit in such a manner that the logic operator "Flag 2 "of their respective plasma process will be set to "F” (at step 876 for Process A, at step 888 for Process B, at step 900 for Process "X”) Any plasma process in the normal spectra subdirectory 288 having its logic operator “Flag 2 " set to "F” will no longer be compared to the current plasma process being run in the processing chamber 36 by the pattern recognition module 370 through the subroutine 852 Step 868 associated with Process A will be bypassed through step 864 when the logic operator "Flag 2 " for Process A is
- Process B is set to "F"
- step 896 associated with Process "X" will be bypassed through step 892 when the logic operator "Flag 2 " for Recipe “X" is set to "F"
- the plasma health/process recognition subroutine 852 will continue via steps 904 and 912 However, if this is not the case, the plasma health/process recognition subroutine 852 will exit from step 904 to step 908
- the protocol of step 908 is directed to determining if the current plasma process being run in the processing chamber 36 has encountered a known error/aberration that is stored in the abnormal spectra subdirectory 292 Therefore, the plasma health/process recognition subroutine 852 may include the portion of the plasma health subroutine 253 of Figure 21 which pertains to the abnormal spectra subdirectory 292 and the unknown spectra subdirectory 296 for that matter (i e , starting with step 266 of the subroutine 253 and including everything thereafter, but not shown) The spectra compared with the abnormal spectra subdirectory 292 in step 266 of the plasma health subroutine 253
- step 916 would indicate that the current plasma process was now only potentially
- the spectra for the new time t 12 is obtained for the subroutine 852 at step 860 in the subject example, and the subroutine 852 proceeds to the logic operator check for each of the plasma processes stored in the normal spectra subdirectory 288
- the subroutine 852 would continue the comparison of the current plasma process being run in the processing chamber 36 with Processes A and B in the subject example, through steps 864 and 868 for Process A and through steps 880 and 884 for Process B However Process C would no longer be compared with the current plasma process at the current time t 12 since step 892 associated with Process C would bypass its associated comparison step
- step 904 The subroutine 852 would proceed with the comparison of the current plasma process being run in the processing chamber 36 with the normal spectra subdirectory 288 since the logic operator "Flag 2 " for each of Process A and B at step 904 was still "T" at the now current time t 12 based upon the logic from step 904
- the spectra for the new current time t 13 is obtained for the subroutine 852 at step 860 in the subject example, and the subroutine 852 proceeds to the logic operator check for each of the plasma processes stored in the normal spectra subdirectory 288 which are available to the subroutine 852. The subroutine 852 would continue the comparison of the current plasma process being run in the processing chamber
- Plasma Processes B and C would no longer be compared with the current plasma process since step 880 associated with Process B would bypass its comparison step 884 and direct the subroutine 852 to instead proceed to step 904, and since step 892 associated with Process C would bypass its comparison step 896 and direct the subroutine 852 to proceed to step 904.
- the subroutine 852 would proceed with the comparison of the current plasma process being run in the processing chamber 36 with the normal spectra subdirectory 288 via step 904 since the logic operator "Flag 2 " for Process A is still "T" at the now current time t 13 .
- Completion of an entire plasma process which was run in the processing chamber 36 while matching at least one plasma process stored in the normal spectra subdirectory 288 will cause the subroutine 852 to exit from step 918 and proceed to step 920.
- Control of the plasma monitoring operations may then be returned by step 920 of the plasma health/process recognition subroutine 852 to, for instance, the startup module 202 of Figures 13.
- Plasma Health/Process Recognition Subroutine 924 - Figure 24 Yet another embodiment of a plasma health subroutine which may be used by the plasma health module 252 is illustrated in Figure 24.
- the plasma health/process recognition subroutine 924 is generally directed to achieving an increase in the speed of the comparison between the current plasma process being run in the processing chamber 36 and the normal spectra subdirectory 288 by at least initially limiting the search within the subdirectory 288 to a single plasma process of the subdirectory 288
- personnel are allowed to indicate which plasma process is to be run in the processing chamber 36
- the data entry device 60 for the mam control unit 58 ( Figure 1 ) may be used to select a plasma recipe to be run from a list of plasma recipes on the display 130
- the startup module 202 could prompt personnel to input the recipe if desired through execution of step 230 of the startup routine 203 of Figure 16 More typically, the recipe to be run on a certain lot of wafers 18 will be input somewhere in the fabrication facility (e g , main control panel), and when
- the plasma health/process recognition subroutine 924 proceeds to step 932 to confirm that this plasma process is in fact stored in the normal spectra subdirectory 288
- step 932 of the plasma health/process recognition subroutine 924 of Figure 24 A spectra of the plasma in the processing chamber 36 at the current time t c is obtained for the subroutine 924 at step 940 if the process selected or input at step 928 was located in the normal spectra subdirectory 288 through execution of step 936 This current spectra is compared with the relevant spectra of the selected plasma process in the normal spectra subdirectory 288 The comparison at step 944 of the recipe recognition subroutine 924 determines if the pattern of the current spectra at the current time t c (from the current plasma process being run in the processing chamber 36) is a "match" with the relevant spectra of the selected plasma process stored in the normal spectra subdirectory 288 "Matches" in accordance with
- Notification of the deviation of the current plasma process from the process selected in step 928 of the plasma health/process recognition subroutine 924 may be provided through execution of step 956 which calls the process alert module 428 discussed above in relation to Figure 14 and which may also offer one or more protocols in relation to this condition if the process control feature is enabled at step 436 of the process subroutine 432
- Other options such as allowing the present plasma process to be terminated (even though it may be a valid plasma process) may also be provided (not shown)
- a variation of the subroutine 924 relates to the fact that the same plasma recipe is typically run on an entire cassette 6, and that the cassette 6 may have one or more qualification wafers 18 in with the production wafers 18 Even though the same plasma recipe is run on these wafers 18, certain differences between the production wafers 18 and the qualification wafer(s) 18 may produce differences in their respective spectral patterns
- the logic of the subroutine 924 may be to first compare the current plasma process against
- Plasma Health/Process Step Recognition Subroutine 972 - Figure 25 Another embodiment of a subroutine which may be used by the plasma health module 252 is presented in Figure 25 Not only does the subroutine 972 of Figure 25 monitor or evaluate the health of the plasma from a plasma process being run in the processing chamber 36, but the subroutine 972 is also able to identify the current plasma step of the current plasma process being run in the processing chamber 36 As such, the subroutine 972 is characterized as a plasma health/process step recognition subroutine 972 Two factors are key to providing this plasma step identification function One is that the steps of the subject plasma process actually differ sufficiently in relation to their subject spectra such that they can be distinguished as is the case of the multi-step recipe illustrated in Figures 17A-C above Another is that the identify of the plasma step be associated in some manner with its corresponding spectra, such as through inputting information to the plasma step field 322e discussed above in relation to Figure 12A
- the plasma health/process step recognition subroutine 972 proceeds with a "parallel" logic and in the same manner as the plasma health/process recognition subroutine 852 of Figure 23
- the plasma health/process step recognition subroutine 972 of Figure 25 begins at step 976 where the Target Directory for the pattern recognition module 370 of Figure 13 is set to the normal spectra subdirectory 288 (i e , the search for "matching" spectra will initiate in the normal spectra subdirectory 288)
- Another preliminary step of the plasma health/process step recognition subroutine 972 is at step 980 where a logic operator Flag 3 is set to "T" for each of the plasma processes stored in the normal spectra subdirectory 288
- the order in which steps 976 and 980 are executed is not particularly important to the present invention
- Data relating to the current plasma process being run in the processing chamber 36 is obtained for the subroutine 972 at step 984 Included in this data is a spectra of the plasma within the processing chamber 36 during execution of a plasma process being run
- the current plasma process is effectively concurrently compared with each plasma process stored in the normal spectra subdirectory 288 which has "matched" the current plasma process up until the current point in time and which is made available to the subroutine 972 If at any time the spectra at the current time t c from the current plasma process does not match a particular process in the normal spectra subdirectory 288, this plasma process is dropped from the list of possible plasma processes and spectra at new, later in time, current times t c are no longer compared with this particular plasma process "Relevance" in terms of which spectra of the selected plasma process is compared with the spectra of the plasma in the chamber 36 at the current time t c by the subroutine 972 may be determined in accordance with either the time dependency requirement or the progression dependency requirement discussed above in relation to the plasma health subroutine 253 of Figure 21 Moreover, as in the case of the plasma health subroutine 253 discussed above in relation to Figure 21 , the plasma health/process step recognition subroutine 9
- the pattern recognition module 370 determines if the pattern of the current spectra at the current time t c is a "match" with the pattern of the relevant spectra of the subject plasma process stored in the normal spectra subdirectory 288 (Process
- step 1012 the clock of the subroutine 972 is adjusted by increasing the current time t c by a factor of "n"
- the magnitude of "n”def ⁇ nes the Analytical Time Resolution i e , what portion of the collected data is actually analyzed
- the subroutine 972 then proceeds from step 1012 to step 1016 where all of the plasma processes from the normal spectra subdirectory 288 which are still a potential "match” for the current plasma process being run in the processing chamber 36 are displayed to the appropriate personnel (e g , on display 130 in Figure 6) Moreover, the specific process step, if any, of each of these potential plasma processes is also displayed at step 1016 Another spectra at the new current time t
- Step 992 associated with Process A will be bypassed through step 988 when the logic operator "Flag 3 " for Process A is set to “F”
- step 1000 associated with Process B will be bypassed through step 996 when the logic operator “Flag 3 " for Process B is set to “F”
- step 1008 associated with Process “X” will be bypassed through step 1004 when the logic operator "Flag 3 " for Process “X” is set to "F”
- the plasma health/process step recognition subroutine 972 will continue via steps 1032 and 1036 However, if this is not the case, the subroutine 972 will proceed from step 1032 to step 1040
- the protocol of step 1040 is directed to determining if the current plasma process being run in the processing chamber 36 has encountered a known error/aberration that is stored in the abnormal spectra subdirectory 292 Therefore, the plasma health/process recognition subroutine 972
- the primary purpose of the plasma health module 252 is to monitor the health of the plasma used to execute a plasma process within the chamber 36 As noted above, the running of plasma processes on product in the chamber 36 will eventually start to adversely impact its performance This "aging" chamber condition is often, if not always, reflected by the spectra of the plasma in the chamber 36 during a plasma process How the pattern of the spectra of the plasma may change over time as the chamber 36 "ages” is illustrated in Figures 26A-C
- Figure 26A presents a spectra 1052 of an exemplary plasma when the processing chamber 36 is in a clean condition and while running a corresponding plasma recipe on product (e g , a "healthy" plasma)
- Figure 26B presents a spectra 1060 of this same exemplary plasma after a number of plasma recipes have been conducted within the processing chamber 36 and while this same plasma recipe is actually being run on product in the chamber 36
- the plasma recipes run in the chamber 36 between the times associated with Figure 26A and 26B have started to age the chamber 36, the interior of the chamber 36 has not sufficiently degraded the health of plasma to the point where the chamber 36 needs to be cleaned
- Figure 26C presents a spectra 1068 of this same exemplary plasma, during the running of the same plasma recipe on product in the same chamber 36 as presented in Figures 26A-B, and where the running of the previous plasma recipes on product in the processing chamber 36 has further deteriorated the condition of the interior of the processing chamber 36
- This spectra 1068 may be selected by the operator of the facility implementing
- Each of the spectra 1052, 1060, and 1068 are characterized by a number of peaks 1056, 1064, and 1072, respectively, of varying intensities (plotted along the "y" axis and expressed in "counts” which is reflective of the intensity level) at various wavelengths (plotted along the "x" axis in nanometers)
- a comparison of the spectra 1052, 1060, and 1068 reveals that their associated patterns are in fact different, including without limitation as follows 1 ) at about the 440 nanometer wavelength region
- peak 1056a in the spectra 1052 of Figure 26A has an intensity of about 3,300
- peak 1064a in the spectra 1060 of Figure 26B has an intensity of about 3,300
- peak 1072a in the spectra 1068 of Figure 26C has an intensity of about 2,700, 2) at about the 525 nanometer wavelength region
- peak 1056b in the spectra 1052 of Figure 26A has an intensity of about
- peak 1064b in the spectra 1060 of Figure 26B has an intensity of about 2,900
- peak 1072b in the spectra 1068 of Figure 26C has an intensity of about 2, 100, 3) at about the 560 nanometer wavelength region
- peak 1056d in the spectra 1052 of Figure 26A has an intensity of about 400
- peak 1064d in the spectra 1060 of Figure 26B has intensity of about 700
- peak 1072d in the spectra 1068 of Figure 26C has an intensity of about 1 ,200, 4) at about the 595 nanometer wavelength region
- peak 1056e in the spectra 1052 of Figure 26A has an intensity of about 2,100
- peak 1064e in the spectra 1060 of Figure 26B has intensity of about 2,000
- peak 1072e in the spectra 1068 of Figure 26C has an intensity of about 2,000, and 5) at about the 625 nanometer wavelength region
- the distmctiveness of the patterns between the spectra 1052, 1060, and 1068 may be utilized to apprise the appropriate personnel of the condition of the processing chamber 36 in relation to cleaning schedules
- At least two options exist for implementing the spectra in the plasma spectra directory 284 of Figure 9 which are at least deemed indicative of a chamber 36 which is in need of cleaning Spectra of plasma in a dirty chamber condition may be included in the abnormal spectra subdirectory 292 of Figure 9
- the plasma health subroutine 253 of Figure 21 , the plasma health/process recognition subroutine 790 of Figure 22, the plasma health/process recognition subroutine 852 of Figure 23, the plasma health/process recognition subroutine 924 of Figure 24, and the plasma health/process step recognition subroutine 972 of Figure 25 would each then include "chamber condition monitoring" capabilities in the manner presented above How one or more spectra which is indicative of a dirty chamber condition may be obtained is as follows Consider the situation where a plasma recipe being run on product in the processing chamber 36 does not "match" any plasma recipe stored in the normal spectra subdirectory 288, and further does not "match” any of the known errors/aberrations stored in the abnormal spectra subdirectory 292 such as discussed above in relation
- Chamber Condition Module 1084 - Figures 27-29 Another way of implementing spectra indicative of a dirty chamber condition is to include this data in a chamber condition subdirectory 300 and to utilize a chamber condition module 1084 which is separate from the plasma health module 252
- a subroutine which may be used to monitor the condition of the chamber 36 through a comparative analysis with the chamber condition subdirectory 300 is illustrated in Figure 27
- the loop 190 of the plasma health subroutine 253 of Figure 16 may include protocol in its step 194 to call the chamber condition subroutine 406 of Figure 27. That is, the subroutine 406 may be called by the plasma health subroutine
- step 253 for each execution of step 254 of the subroutine 253 where a current spectra at the current time t c is obtained for the subroutine 253.
- This same spectra may then be made available to the chamber condition subroutine 406 through execution of step 410 of the subroutine 406.
- the pattern of this spectra may then compared with the pattern of spectra in the chamber condition subdirectory 300 at step 412 (step 408 of the subroutine 406 sets the Target Directory for the pattern recognition module 370 of Figure 13 to the chamber condition subdirectory 300). More specifically, step 412 of the chamber condition subroutine 406 directs the subroutine 406 to go to the pattern recognition module 370 of Figure 13.
- the chamber condition subroutine 406 will proceed from step 414 to step 416. Control is then returned to the loop 190 of the plasma health subroutine 253 of Figure 21 where the "clock" is adjusted by an increment of "n” at its step 278. Another spectra is then obtained for the chamber condition subroutine 406 at this new current time t c when step 194 of the loop 190 from the plasma health subroutine 253 of Figure 21 is again encountered, which calls the chamber condition subroutine 406 of Figure 27 for a repetition of the above-described analysis. Any of the subroutines 790, 852, 924, and 972 for providing plasma health evaluations may include this type of feature (not shown).
- the chamber condition subroutine 406 of Figure 27 will continue to execute in the above-noted manner until one of two conditions is met.
- the first is when plasma monitoring operations are terminated through the plasma health subroutine 253 of Figure 21 , such as when all of the data on the current plasma process has been evaluated and the plasma processing operations have been terminated.
- the second condition which will cause the termination of the chamber condition subroutine 406 in the above-described manner is when the pattern recognition module 370 indicates that there is a match between the spectra of the plasma in the processing chamber 36 at the then current time t c and at least one spectra in the chamber condition subdirectory 300.
- the subroutine 406 will proceed from step 414 to step 418 where the chamber condition subroutine 406 of Figure 44 transfers control to the process alert module 428 where any of the actions described above in relation to a dirty chamber condition may be initiated.
- the chamber condition subroutine 406 may be run in parallel with the plasma health module 252 each time the plasma health module 252 is accessed
- step 236 of the startup routine 203 also include protocol to call the chamber condition subroutine 406.
- step 236 of the startup routine 203 also include protocol to call the chamber condition subroutine 406.
- an additional step of adjusting the clock would have to be included, as well as a loop defined by this step as well as steps 410, 412, and 414 in a manner similar to the other subroutines presented herein.
- the premise employed by the chamber condition subroutine 1088 is that the interior of the processing chamber 36 has degraded to the point where a cleaning operation should be employed when the time required to complete any plasma step of a multiple step plasma process takes longer than a time limit previously established for the plasma step.
- the time required to complete a plasma step may increase as the condition of the interior of the processing chamber 36 degrades.
- a given plasma step may take 30 seconds to achieve its desired/predetermined end result in a "clean" processing chamber 36, may take 50 seconds in a chamber 36 which is at an intermediate time in relation to a "cleaning" cycle, and may take in excess of 70 seconds under dirty chamber conditions.
- the chamber condition subroutine 1088 assumes that when a given plasma step takes longer than its associated time limit, the associated cause is the existence of a dirty chamber condition
- the chamber condition subroutine 1088 may be integrated in some manner with the operation of any one or more of the plasma health/process recognition subroutine 790 of Figure 22, the plasma health/process recognition subroutine 852 of Figure 23, or the plasma health/process recognition subroutine 924 of Figure 24 (e g , by being incorporated therein, by being called simultaneously therewith) In each of these cases, once the identity of the plasma process is determined, one will know each of the particular plasma steps which are included in this plasma process.
- the chamber condition subroutine 1088 may also be integrated in some manner with the plasma health/process step recognition subroutine 972 of Figure 25 which identifies the current plasma step being executed in the chamber 36
- a maximum time limit for each plasma step, if any, of the plasma process to be run in the chamber 36 should be obtained by the chamber condition subroutine 1088 at its step 1092
- Personnel may manually input the maximum time limit for the subject plasma step(s) of the plasma process with the data entry device 132 for purposes of step 1092 of the chamber condition subroutine 1088
- a more preferred approach is to include these time limits in the maximum total process step time field 322f for the mam data entry 350 of the plasma process as stored in the normal spectra subdirectory 288 ( Figure 12A)
- the maximum time limit may be empirically determined and input to the subject maximum total process step time field 322f
- the limits referred to in step 1092 may simply coincide with a time in which the operator of the fabrication facility employing the wafer production system 2 has determined is necessary to maintain a desired production rate, which would then be input to the subject maximum total process step time field 322f
- Information for step 1092 of the chamber condition subroutine 1088 may then be automatically retrieved from the corresponding maximum
- step 1096 of the chamber condition subroutine 1088 A process step clock (not shown) is started once the subject plasma step is initiated and will not stop until the termination of this plasma step
- Step 1096 may utilize the endpoint detection module 1200 to be discussed below in relation to Figures 52-58 to identify the endpoint of the current plasma step
- a comparison is made at step 1 100 of the chamber condition subroutine 1088 between the time spent on the current plasma step from step 1096 and its associated maximum time limit from step 1092 So long as this limit has not yet been exceeded, the chamber condition subroutine 1088 will proceed to step 1 108 where a determination is made as to whether the current plasma process has been terminated Any continuation of the plasma process will allow the chamber condition subroutine 1088 to continue its analysis through execution of steps 1096 and 1 100 as described When the plasma process is terminated, however, the subroutine 1088 will proceed from step 1 108 to step 1 112 where control of plasma monitoring operations may be returned to, for instance,
- the chamber condition subroutine 1 120 presented in Figure 29 and which may be implemented in the same way as the chamber condition subroutine 1088 described above in relation to Figure 28
- the premise employed by the chamber condition subroutine 1120 is that the interior of the processing chamber 36 has degraded to the point where a cleaning operation should be employed when the time required to complete an entire plasma process (all of the plasma steps) takes longer than a time limit previously established for completing the plasma process
- the time required to complete an entire plasma process may increase as the condition of the interior of the processing chamber 36 degrades, such as by the formation of deposits on its interior surfaces
- a given plasma process may take 180 seconds to achieve its desired/predetermined result in a "clean" processing chamber 36, may take 220 seconds in a chamber 36 which is at an intermediate time in relation to a "cleaning" cycle, and may take in excess of 300 seconds when in a dirty chamber condition
- the processing chamber 36 may actually degrade to the point where the desired end result of the plasma process may never be realized Therefore
- the chamber condition subroutine 1120 Since the chamber condition subroutine 1120 needs to "know" the identify of plasma process to execute its analysis, the chamber condition subroutine
- the chamber condition subroutine 1088 may also be integrated in some manner with the plasma health/process step recognition subroutine 972 of Figure 25 which also identifies the current plasma process being executed in the chamber 36, as well as the individual process step Referring to Figure 29, a maximum time limit for the plasma process to be run in the chamber 36 is input to the chamber condition subroutine 1120 at its step 1124 Personnel may manually input the maximum time limit for the subject plasma process with the data entry device 132 for purposes of step 1124 of the chamber condition subroutine 1120 A more preferred approach is to include this time limit in the maximum total process time field 322g for the mam data entry 350 of the subject plasma process as stored in the normal spectra subdirectory 288 The maximum
- Step 1128 The amount of time required to complete the current plasma process being run in the processing chamber 36 is monitored at step 1128 of the chamber condition subroutine 1120
- a recipe clock (not shown) is started once the plasma process is initiated (e g , when the plasma comes "on” in the chamber 36) and will not stop until the plasma process is terminated (e g , when the plasma goes "off' in the chamber 36)
- Step 1128 may also utilize the endpoint detection module 1200 to be discussed below in relation to Figures 52-
- step 1132 of the chamber condition subroutine 1120 A comparison is made at step 1132 of the chamber condition subroutine 1120 between the time spent on the current plasma process from step 1128 and its associated maximum time limit from step 1124 So long as this limit has not yet been exceeded, the chamber condition subroutine 1120 will proceed to step 1140 where a determination is made as to whether the current plasma process has been terminated Any continuation of the plasma process will allow the chamber condition subroutine 1120 to continue its analysis through execution of steps 1128 and 1132 in the described manner When the plasma process is terminated, however, the subroutine 1120 will proceed from step 1140 to step 1144 where control of plasma monitoring operations may be returned to, for instance, the startup module 202 of Figure 15 The chamber condition subroutine 1120 will continue to execute in the above-noted manner unless the time spent on the current plasma process exceeds its corresponding maximum time limit In this case the chamber condition subroutine 1120 will proceed from step 1132 to step 1136 where the process alert module 428 of Figure 14 is called based upon a dirty chamber condition,
- Figures 30A-D illustrate the differences in the spectral patterns of the plasma in the chamber 36 at various stages to illustrate how the current plasma process module 250 not only can evaluate the health of the subject plasma process, but how it can distinguish between different types of plasma processes through spectral analysis
- Figure 30A is an exemplary spectra of a plasma recipe being conducted in a processing chamber 36 which has been determined to be in need of a cleaning operation Compare this spectra 1440 of Figure 30A with the spectra 1450 of Figure 30D which is this same plasma recipe in this same processing chamber 36, but when in a clean condition and after a conditioning wafer operation has been completed to prepare the chamber 36 for processing production wafers 18 Note the different intensities of the peaks in these two spectras 1440 and 1450 at the various wavelengths There are two strong peaks around the 550 nanometer
- the cleaning protocol illustrated in Figure 30 starts with a wet clean where the chamber 36 is vented and opened such that interior surfaces of the chamber 36 may be wiped with one or more appropriate solvents.
- a plasma cleaning operation may have been previously run in the chamber 36, but may have been unable to adequately address the condition of the interior of the chamber 36 to the desired degree.
- the chamber 36 is resealed and a plasma is introduced into the chamber 36 with no product therein.
- the spectra 1444 of Figure 30B is of an exemplary plasma within the processing chamber 36 with no product therein at the start of such a plasma cleaning operation.
- the spectra 1448 of Figure 30C is of an exemplary plasma within the processing chamber 36 at the end of a plasma clean of the chamber 36 which has appropriately addressed the residuals of the wet clean. Compare this spectra 1448 from Figure 30C with the spectra 1450 of Figure 30D which is this same plasma in this same processing chamber 36, but after a number of conditioning wafers have been processed in the chamber 36. Note the different intensities of the peaks in these two spectras
- Plasma Monitoring Assembly 500 Degradation of the interior of the processing chamber 36 from plasma processes run therein also degrades the inner surface 40 of the window 38 on the chamber 36 which is exposed to the plasma (the outer surface 42 being isolated from the plasma, and thereby not affected by the plasma) Recall that data for the current plasma process module 250 in the nature of optical emissions of the plasma in the chamber 38 is obtained through the window 38 Therefore, as the window 38 degrades, so to may the reliability of the results of the current plasma process module 250 An embodiment which addresses this condition, as well as other conditions which may adversely impact the reliability of the results provided by the current plasma process module 250, is presented in Figure 31 in the nature of the plasma monitoring assembly 500
- the plasma monitoring assembly 500 of Figure 31 includes all of the features discussed above in relation to the plasma monitoring assembly 174 of
- the plasma monitoring assembly 500 includes a plasma monitoring module 560 ( Figures 31-32), which includes all of the same modules as identified above in relation to the plasma monitoring module 200 of Figure 7 and which is part of the PMCU 128' (e g , the same current plasma process module 250 and all of its sub-modules) Moreover, the spectrometer assembly
- the plasma monitoring assembly 500 and more specifically its plasma monitoring module 560, includes a window monitoring or calibration assembly 552 which includes a window monitoring or calibration module 562 which is also part of the plasma monitoring module 560 as illustrated in Figure 32
- the PMCU 128' is therefore the same as presented above in relation to the plasma monitoring assembly 174 of Figure 6, except that it includes this additional feature in relation to the calibration module 562 As such, the "prime" designation is utilized in relation to the PMCU 128 of Figure 31
- the calibration assembly 552 provides certain advantages for the plasma monitoring assembly 500 of Figure 31 over the plasma monitoring assembly 174 of Figure 6
- the calibration assembly 552 of Figure 31 provides two main functions One of these functions is to calibrate the current plasma process module 250 (Figures 7 and 32) for any wavelength shift of the spectra of the plasma within the processing chamber 36. Wavelength shifts may be attributed to the spectrometer assembly 506, but may also exist for other reasons.
- Another function provided by the calibration assembly 552 of Figure 31 is to calibrate the current plasma process module 250 for any intensity shift of the spectra of the plasma within the processing chamber 36. Intensity shifts may be due to an "aging" of the window 38 on the chamber 36 through which optical emissions are obtained, but may also be due to other conditions.
- the window 478 in Figure 31 has a different configuration than the window 38 presented in Figure 6, a different reference numeral is used not only for the window, but for the processing chamber as well. Therefore, the processing chamber of Figure 31 identified by the reference number 474 may be used in place of one or more of the chambers 36 presented in Figure 1.
- the calibration assembly 552 of Figure 31 includes a calibration light source 556 which is operatively interfaced with the window 478 of the processing chamber 474 by a fiber optic cable assembly 504.
- this calibration light 556 source actually has a first calibration light source 556a and a second calibration light source 556b.
- the first calibration light source 556a uses a first type of light to identify wavelength shifts associated with the plasma monitoring assembly 500. Intensity shifts are identified through the second calibration light source 556b which uses a second type of light which is different than the first type of light. Benefits associated with the use of two different types of light for identifying wavelength and intensity shifts, respectively, will be addressed in relation to the calibration module 562 and Figures 40-48. Although not preferred, the same type of light could be used to identify both intensity and wavelength shifts (e.g., light having a plurality of discrete intensity peaks).
- the fiber optic cable assembly 504 also operatively interfaces the window 478 with the spectrometer assembly 506 such that the window monitoring assembly 552 is able to directly monitor the condition of the window 478.
- a calibration light is directed to the window 478 from the calibration light source 556 through the fiber optic cable assembly 504, preferably when there is no plasma within the processing chamber 474.
- operation of the calibration assembly 552 is not dependent in any manner on the plasma within the chamber and its data is in fact independent of any data relating to the plasma
- a portion of the calibration light is reflected by the window 478 and is directed by the fiber optic cable assembly 504 to the spectrometer assembly 506
- the spectrometer assembly 506 of Figure 31 preferably is of the solid- state type, and in one embodiment includes three individual solid state spectrometers 516a-c, each of which analyzes a different wavelength region and as illustrated in Figure 33
- the spectrometer 516a may analyze the 246 nanometer to 570 nanometer region
- the spectrometer 516b may analyze the 535 nanometer to 815 nanometer region
- the spectrometer 516c may analyze the 785 nanometer to 1014 nanometer region
- the overlap between these individual spectrometers 516 once again facilitates alignment of the three spectral segments and reduces the potential for losing optical emissions data at the transition zones
- Each of these spectrometers 516 are interconnected in effectively parallel relation by the fiber optic cable assembly 504 which is illustrated in more detail in Figure 34
- the fiber optic cable assembly 504 of Figure 34 includes three inner cables 508 surrounded by six outer cables 512
- Light from the calibration light source 556 is directed to the window 478 through the outer cables 512 during calibration operations to be discussed below in relation to the calibration module 562 of Figures 40-48, while light reflected by the window 478 (as well as light from within the chamber 474 during the running of plasma processes within the chamber 474 for that matter, at which time the calibration assembly 552 is not activated or running) is directed to the spectrometer assembly 506 through the inner cables 508 of the fiber optic cable assembly 504
- One inner cable 508 of the fiber optic cable assembly 504 is directed to each of the spectrometers 516 of the spectrometer assembly 506 as illustrated in Figure 33
- This data received by the spectrometer assembly 506 from the calibration light source 556 is directly from the inner surface 482 of the window 474, and therefore it is proper to characterize the calibration assembly 552 as directly monitoring the condition of the inner surface of the window 478
- the calibration assembly 552 also addresses the presence of the outer surface 486 of the window 478.
- the outer surface 486 and inner surface 482 of the window 478 are disposed in a non-parallel relationship.
- the inner surface 482 of the window 478 is disposed at least substantially perpendicular to a reference axis 490 which coincides with the primary axis of the light from the calibration light source 556 as it is directed toward the window 478, whereas the outer surface 486 of the window 478 is disposed in non- perpendicular relation to this reference axis 490.
- the angle between the reference axis 490 and the outer surface 486 of the window 478 is within the range of about 2° to about 45°, and in another embodiment this angle is less than the critical angle.
- This relative positioning of the inner surface 482 and outer surface 486 of the window 478 has the effect of having that portion of the calibration light, which is reflected by the outer surface 486 of the window 478, be directed away from the axis 490 and thereby away from the inner cables 508 of the fiber optic cable assembly 504 which lead to the spectrometer assembly 506.
- the only significant portion of light which the inner cables 508 of the fiber optic cable assembly 504 "sees” is the light which is reflected by the inner surface 482 of the window 478 - not from the outer surface 486 of the window 478.
- the inner surface 482 of the window 478 is that which is affected by conducting plasma processes within the processing chamber 474, and which thereby affects the light which is emitted from the processing chamber 474 through the window 478. Therefore, the light received by the spectrometer assembly 506 during calibration operations by the window monitoring assembly
- 552 presents a more accurate depiction of the condition of the inner surface 482 of the window 486.
- Further enhancement may be realized by incorporating a broad band anti-reflection coating (e.g., of multiple-layer or laminated construction) on the outer surface 486 of the window 478 at least in that region where the calibration light impacts the outer surface 486.
- a broad band anti-reflection coating e.g., of multiple-layer or laminated construction
- Maintaining a proper relative positioning between the fiber optic cable assembly 504 and the window 478 on the chamber 474 is important to the operation of not only the calibration module 562, but the current plasma process module 250 as well.
- One way of interconnecting the fiber optic cable assembly is important to the operation of not only the calibration module 562, but the current plasma process module 250 as well.
- the fixture assembly 1534 includes a window fixture 1538 which securely retains the window 478 and allows for detachable interconnection of the same with the processing chamber 474 (e.g, via one or more threaded fasteners).
- a cavity or recess 1542 exists within an interior portion of the window fixture 1538 and interfaces with the outer surface 486 of the window 478.
- Surfaces of the window fixture 1538 which define the recess 1542 are black anodized so as to absorb any portion of the calibration light from the calibration light source 506 which is reflected by the outer surface 486 of the window 478 during calibration operations. Light absorbing coatings could also be utilized to provide this function.
- the fixture assembly 1534 also includes a fiber fixture 1546 which is appropriately interconnected (e.g., detachably) with the window fixture 1538 (e.g., via one or more threaded fasteners).
- the recess 1542 in the window fixture 1538 is thereby a closed space in the assembled condition via the outer surface 486 of the window 478 and a portion of the back side of the fiber fixture 1546.
- Appropriate treatment of the portion of the fiber fixture 1546 which closes the recess 1542 may also be implemented to reduce the potential for that portion of the calibration light which is reflected by the outer surface 486 of the window 478 interfering with that portion of the calibration light which is reflected by the inner surface 482 of the window 478 and provided to the spectrometer assembly 506 via the fiber optic cable assembly 504.
- the fiber optic cable assembly 504 is removably or detachably interconnected with the fiber fixture 1546 by a fiber fixture coupling 1554 on the fiber fixture 1546 and a cable coupling 1558 on an end of the fiber optic cable assembly 504 which houses both the inner cables 508 and the outer cables 512 ( Figure 34)
- the ends of the inner cables 508 and outer cables 512 project toward the outer surface 486 of the window 478 in axial alignment with a port 1550 which extends through the fiber fixture 1546 to intersect the recess 1542 in the window fixture 1538 Therefore, calibration light from the calibration light source 556 is directed through the outer cables 512, through the port 1550 in the fiber fixture 1546, through the recess 1542 in the window fixture 1538, and to the outer surface 486 of the window 478 Calibration light which is reflected by the inner surface 482 of the window 478 travels through the recess 1542 in the window fixture 1538, through the port 1550 in the fiber fixture 1546, into the inner cables 508 of the fiber optic cable assembly
- Plasma Monitoring Assembly 700 - Figures 37-39 Another embodiment of a plasma monitoring assembly which also reduces the potential for that portion of the calibration light which is reflected by the outer surface of the window interfering with that portion of the calibration light which is reflected by the inner surface of the window and provided to the calibration module 562 ( Figure 32) is illustrated in Figure 37 This embodiment may be used in place of the plasma monitoring assembly 500 of Figure 31 although the configuration presented in Figure 31 is more preferred
- the plasma monitoring assembly 700 of Figure 37 generally includes a calibration assembly 726, the calibration module 562 ( Figure 32) which is part of the PMCU 128' ( Figure 1 and 32), and the plasma monitoring module 560 ( Figures 32 and 37)
- the spectrometer assembly 712, CCD array 716, and calibration light source 728 are the same in relation to the similarly identified components of the Figure
- the plasma monitoring assembly 700 of Figure 37 differs from the plasma monitoring assembly 500 of Figure 31 principally in relation to the optical arrangement of the calibrating componentry
- the calibration assembly 726 uses a fiber optic cable 704 to operatively interface the calibration light source 728 and the window 38, and another fiber optic cable 708 to operatively interface the window 38 and the spectrometer assembly 712
- the fiber optic cable 704 is disposed to have light from the calibration light source 728 impact the outer surface 42 of the window 38 at an angle other than perpendicular, and the fiber optic cable 708 is disposed to receive only that portion of the calibration light which is reflected by the inner surface 40 of the window 38 and not any light which
- the outer surface 42 of the window 38, the fiber optic cable 704, and the fiber optic cable 708 has the effect of having that portion of the calibration light, which is reflected by the outer surface 42 of the window 38, be reflected in a manner so as to not enter the fiber optic cable 708 Therefore, when a calibration light is sent from the calibration light source 728 to the window 38 through the fiber optic cable 704, the only significant portion of light which the fiber optic cable 708 "sees" is the light which is reflected by the inner surface 40 of the window 38 - not light from the outer surface 42 of the window 38
- the inner surface 40 of the window 38 is that which is affected by conducting plasma processes within the processing chamber 36, and which thereby affects the light which is emitted from the processing chamber 36 through the window 38 Therefore, the light received by the spectrometer assembly 712 during calibration operations by the window monitoring assembly 700 presents a more accurate depiction of the condition of the inner surface 40 of the window 36
- FIG. 37 Further enhancement of the arrangement presented in Figure 37 may be realized by incorporating a broad band anti-reflection coating (e g , of multiple- layer or laminated construction) on the outer surface 42 of the window 38 at least in that region where the calibration light impacts the outer surface 42
- a broad band anti-reflection coating e g , of multiple- layer or laminated construction
- These types of coatings increase the amount of the calibration light which passes through the outer surface 42 of the window 38 to the inner surface 40 by reducing the amount of the calibration light which is reflected by the outer surface 42
- the embodiment presented in Figure 31 could be used with the type of window presented in the Figure 37 embodiment where the above- noted coating is included on the window 38 and with the ends of the fiber optic cable assembly 504 projecting toward the window 38 to form at least a substantially perpendicular angle relative to both the outer surface 42 and the inner surface 40
- This arrangement is less preferable in that notwithstanding the presence of an anti-reflection coating on the outer surface 42 of the window 38, some portion of the calibration
- the fixture assembly 1564 includes a window fixture 1568 which securely retains the window 38 and allows for interconnection of the same with the processing chamber 36 (e.g, via one or more threaded fasteners).
- a cavity or recess 1572 exists within an interior portion of the window fixture 1568 and interfaces with the outer surface 42 of the window 38.
- the fixture assembly 1564 also includes a fiber fixture 1576 which is appropriately interconnected with the window fixture 1568 (e.g., via one or more threaded fasteners).
- the recess 1572 in the window fixture 1568 is thereby a closed space in the assembled condition via the outer surface 42 of the window 38 and a portion of the back side of the fiber fixture 1576.
- Each of the fiber optic cables 704 and 708 are removably interconnected with the fiber fixture 1576.
- a fiber fixture coupling 1580b on the fiber fixture 1576 is disposed in the proper orientation for establishing an appropriate interconnection with the fiber optic cable 704 from the calibration light source 704, while a fiber fixture coupling 1580a on the fiber fixture 1576 is disposed in the proper orientation for establishing an appropriate interconnection with the fiber optic cable 708 leading to the spectrometer assembly 712.
- the ends of the fiber optic cables 704 and 708 project toward the outer surface 42 of the window 38 at the proper angle and in axial alignment with a port 1584a and 1584b, respectively.
- the ports 1584a and 1584b each extend through the fiber fixture 1576 to intersect the recess
- calibration light from the calibration light source 728 is directed through the fiber optic cable 704, through the port 1584b in the fiber fixture 1576, through the recess 1572 in the window fixture 1568, and to the outer surface 42 of the window 478.
- Calibration light which is reflected by the inner surface 40 of the window 38 travels through the recess
- FIG. 61 Another embodiment of a device for interconnecting a fiber optic cable end 1587 (e g , of fiber optic cable assembly 504, of fiber optic cable 704, of fiber optic cable 708, or an adapter for interfacing with the same) is presented in Figure 61 in the form of a fixture assembly 1586
- the fixture assembly 1586 generally includes a housing 1588 which may be appropriately interconnected with the processing chamber 36 so that an aperture 1598 of the housing 1588 is aligned with the window 38 on the chamber 36 to at least a degree (i e , to allow for the passage of light to the fiber optic cable end 1587
- An appropriate seal ring 1592 may be disposed within and become part of this interconnection
- a cable mount 1590 is disposed within at least a portion of the aperture 1598 of the housing 1588 to which the fiber optic end 1587 is actually attached
- An appropriate interfacing relationship exists between the cable mount 1590 and the housing 1588 such that the cable mount 1590, and thereby the fiber optic cable end 1587, may be moved relative to the housing 1588
- drift over the life of the spectrometer assembly 506 due to various factors such as temperature Any drifting of the output from the spectrometer assembly 506 will cause a wavelength shift in the spectra which is evaluated by the current plasma process module 250 An example of drifting would be that a peak in a spectra from the processing chamber 474 which is actually at the 490 nanometer wavelength, may appear at the 491 nanometer wavelength from the output of the spectrometer assembly 506 due to this drifting Moreover, the window 478 may have an effect on the spectra of the plasma from the current plasma process which is passing through the window 474 to the spectrometer assembly 506, namely by providing an intensity shift in one or more regions of the spectra representative of the plasma in the processing chamber 474 Failure to address either of these conditions may adversely affect the performance of the current plasma process module 250
- the calibration module 562 includes a calibration routine 564 which is run without any plasma in the processing chamber 474 and which is run at times determined by the facility using the wafer production system 2 (e g , once a day, on every shift change) Instructions may be included in the calibration routine 564 to detect the existence of plasma in the processing chamber 474 in any of the above-noted manners, and to exit the calibration routine 564 if any such plasma is detected (not shown)
- Step 568 of the calibration routine 564 directs the calibration light source 556 to send the calibration light to the window 478 Thereafter, step 572 directs the calibration routine 564 to proceed to an appropriate calibration subroutine
- the calibration subroutine referred to in step 572 of Figure 40 may include the calibration subroutine 576 which is illustrated in Figure 41 Generally, the O 99/54694
- step 580 of the calibration subroutine 576 is directed toward making at least one adjustment in relation to the spectrometer assembly 506 to compensate for a wavelength shift associated with the spectral data obtained through the window 478
- a comparison is undertaken by step 580 of the calibration subroutine 576 between the spectra of the calibration light from the calibration light source 556 which is directed to the window 478 and the spectra of the calibration light which is reflected from the inner surface 482 of the window 478 and provided to the spectrometer assembly 506 Since no plasma exists in the processing chamber 474 during calibration operations, the light received by the spectrometer assembly 506 should be limited to that portion of the calibration light which is reflected by the inner surface 482 of the window 478 which is therefore a direct monitoring of a condition of the window 478, specifically its inner surface 482 Any wavelength shift from the comparison of step 580 will be noted in step 584 of the calibration subroutine 576, an adjustment will be made in relation to the plasma monitoring assembly 500 to account for this shift at step 588 of the subroutine
- the comparison of the subject spectra at step 580 of the calibration subroutine 576 and the identification of any wavelength shift in step 584 of the subroutine 576 may be implemented in the following manner
- the spectra of the calibration light which is sent to the window 478 may be obtained from the calibration light subdirectory 310 of Figure 9
- This spectra is analyzed to identify the location of one or more of the intensity peaks in this spectra and possibly the relative wavelength positioning of these intensity peaks
- a "peak" may be equated as any portion of the spectra with an intensity greater than a predetermined amount (e g , at least about 100 intensity units) which exists over a predetermined wavelength range (e g , no more than about 2 nanometers) Therefore, the above-referenced analysis of the spectra of calibration light sent to the window 478 may simply entail noting the intensity over at least a portion of this spectra using an appropriate wavelength resolution For example, one or more peaks in this spectra may be identified by noting the intensity of the
- a spectra 676 which is output by the spectrometer assembly 506, due to a reflection of the calibration light represented in Figure 42 from the inner surface 482 of a "clean" window 478 on the processing chamber 474, is presented in Figure 43 (when the window 478 has not yet been exposed to any plasma processes).
- the spectra 676 is characterized by a plurality of peaks 678 of varying intensity, with "intensity” again being plotted along the "y” axis and expressed in “counts” which is reflective of the intensity level, and with “wavelength” again being plotted along the "x" axis in nanometers.
- the spectra 676 of Figure 43 may be compared with the spectra 672 of Figure 42 in the above-described manner to determine if the output from the spectrometer assembly 506 needs to be calibrated for a wavelength shift
- the peaks 678 of the spectra 676 of Figure 43 should appear at the same wavelengths as the corresponding peaks 674 of the spectra 672 of Figure 43
- the peak 678a from Figure 43 and the peak 674a from Figure 42 should be at the same wavelength
- the adjustment made in relation to the plasma monitoring assembly 500 at step 588 will be correspondingly "small "
- the amount of wavelength shift which will initiate an adjustment in relation to the plasma monitoring assembly 500 need not be zero That is, a certain amount of wavelength shift may be tolerated before an adjustment is made in relation to the plasma monitoring assembly 500 at step 588 For instance, in one embodiment no adjustments are made in relation to the plasma monitoring assembly 500 unless a wavelength shift of at least a certain amount is identified by steps 580 and 584 of the subroutine 576 (e g , a wavelength shift of at least about 025 nanometers is required before any adjustment is made)
- step 588 of the subroutine 576 provides for the adjustment in relation to the plasma monitoring assembly 500 At least two options exist for this "making at least adjustment in relation to the plasma monitoring assembly 500 "
- One alternative is to physically adjust the spectrometer assembly 506 if such is of the scanning type Figure 44 presents one embodiment of the spectrometer 516a' which is of the scanning- type
- the spectrometer 516a' includes an aperture 520 through which light from the inner cable 508a of the fiber optic cable 504 assembly enters the spectrometer 516a' Light passing through the aperture 520 is reflected by a mirror 524 onto a grating 532 Both the mirror 524 and grating 532 may be mounted for pivotal movement through a mirror pivot 528 and a grating pivot
- the calibration assembly 552 of Figure 31 may also be used to calibrate the output from the spectrometer assembly 506 when there is an intensity shift in the spectra of the plasma in the processing chamber 474. Intensity shifts in the spectra will typically be due to "aging" of the window 478. "Aging" of the window 478 as used herein means that the plasma processes which have been conducted within the processing chamber 474 have affected the inner surface 482 of the window 478 in some manner (e.g., by forming deposits on the inner surface 482, by etching the inner surface 482, by a combination of forming deposits on and etching the inner surface 482).
- the pattern recognition subroutine 374 of Figure 13 with its point-by-point analysis is employed by the pattern recognition module 370 to compare the pattern of the current spectra with the spectra in the plasma spectra directory 284 ( Figure 9), and where the intensity "match limit” is set on 10% (a percentage difference basis)
- the intensity "match limit” is set on 10% (a percentage difference basis)
- certain deposits have formed on the inner surface 482 of the window 478 such that the intensity of light emitted through the window 478 is reduced by 30%
- the pattern recognition subroutine 374 will therefore indicate that the current spectra is not a "match” with any spectra in the relevant subdirectory of the plasma spectra directory 284 (because its "match limit” is a 10% variation in the intensity of corresponding wavelengths in the subject spectra, and because the window deposits have reduced the intensity of light from the chamber 474 by 30%), even though the plasma is
- the calibration module 574 of Figure 40 may also include a calibration subroutine which is generally directed toward making an adjustment to account for the above-noted type of intensity shift in the spectra of the plasma in the chamber 474 as emitted through its window 478
- a calibration subroutine which is generally directed toward making an adjustment to account for the above-noted type of intensity shift in the spectra of the plasma in the chamber 474 as emitted through its window 478
- One embodiment of such a subroutine is illustrated in Figure 45 in the nature of the calibration subroutine 600 No plasma exists in the chamber 474 during the execution of the calibration subroutine 600 or at least when obtaining data therefore Moreover, detection of plasma in accordance with the foregoing may automatically terminate calibration operations through the subroutine 600
- the subroutine 600 may be executed on a periodic basis which is established by the operator of the facility using the wafer production system 2 (e g , daily)
- the light provided to the spectrometer assembly 506 should be limited to that portion of the calibration light which is reflected by the inner surface 482 of the window 478 As such, the calibration subroutine 600 actually monitors/evaluates the condition of that portion of the window 478 which may be adversely affected by plasma processes conducted in the chamber 474
- the comparison of the subject spectra at step 604 of the calibration subroutine 600 and the identification of any intensity shift in step 608 of the subroutine 600 may be implemented in the following manner
- the spectra of the calibration light which is sent to the window 478 may be obtained from the calibration light subdirectory 310 of Figure 9
- Its intensity peaks are identified in the manner discussed above in relation to the calibration subroutine 576 of Figure 41 , as well as possibly their relative wavelength positionmgs These same peaks should appear at the same wavelength (assuming that there is no wavelength shift) and "same" intensity level (taking into consideration the above- noted principles of optics) in that portion of the calibration light which is reflected by the inner surface 482 of the window 478
- the amount of intensity shift may be identified simply by finding these same peaks in that portion of the calibration light which is reflected by the inner surface 482 of the window 478 and noting any corresponding intensity shift
- the intensity of the peaks alone may be
- a spectra 682 of a calibration light is illustrated in Figure 46A (e g , a mercury lamp or other similar calibration light source with light defined by wavelengths from about 200 nanometers to about 1 ,000 nanometers), and this calibration light may be used by the calibration light source 556 of the calibration assembly 552 ( Figure 31 ) and by the calibration subroutine 600 ( Figure 45) to identify an intensity shift
- the spectra 682 of Figure 46A is characterized by a plurality of discrete intensity peaks 686 of varying intensity.
- FIG. 46A depicts the actual pattern of light which is sent to the window 478 in the subject example and when the window 478 has been exposed to a plurality of plasma processes (e.g., an "aged” window 478).
- a spectra 690 which is output by the spectrometer assembly 506 is presented in Figure 47A after the above-noted principles of optics has been accounted for and which is representative of that portion of the calibration light which is reflected from the inner surface 482 of the aged window 478 on the processing chamber 474.
- the spectra 690 is characterized by a plurality of discrete peaks 694 of varying intensity, with “intensity” again being plotted along the "y” axis and expressed in “counts” which is reflective of the intensity level, and with “wavelength” being plotted along the "x” axis in nanometers.
- the spectra 690 of Figure 47A may be compared with the spectra 682 of Figure 46A to determine if the output from the spectrometer assembly 506 needs to be calibrated for an intensity shift, such as due to an aging window 478. This comparison is again undertaken at step 604 of the calibration subroutine 600 of Figure 45.
- the peaks 694 of the spectra 690 of Figure 47A should appear not only at the same wavelengths as the corresponding peaks 686 of the spectra 682 of Figure 46A (as discussed above in relation to calibrating the output of the spectrometer assembly 506 for a wavelength shift), but they should also be at the same level of intensity. For example, the peak 694a from Figure 47A and the peak 686a from Figure 46A should be at the same intensity level, the peak
- Peak 694a of the spectra 690 from Figure 47A is at substantially the same intensity (about 9300) as its corresponding peak in the spectra 682 of Figure 46A, namely peak 686a.
- peak 694b of the spectra 690 from Figure 47A has an intensity of about 5,100
- the intensity of its corresponding peak in the spectra 682 from Figure 46A namely peak 686b
- peak 694c of the spectra 690 from Figure 47A has an intensity of about 9,600
- the intensity of its corresponding peak in the spectra 682 from Figure 46A namely peak 686c
- peak 478 is not having the same dampening effect on the entire spectra being obtained through the window 478
- the window 478 is having at least a first dampening effect on one part of the spectra (e g , the 450 nm region)
- the adjustment referred to in step 612 of the calibration subroutine 572 may be generally viewed as normalizing the spectra 690 of Figure 47A (calibration light reflected by the inner surface 482 of the window 478) to the spectra 682 of Figure 46A (calibration light sent by the calibration light source 556 to the window 478)
- One way of "normalizing” the data is by "regression fitting " This option may be used regardless of what type of spectrometer is implemented for the spectrometer assembly 506, such that this option may be employed for both scanning-type spectrometers and solid state spectrometers
- Another way of characterizing the subject adjustment is through the concept of calibration factors or gam If there is a "uniform" intensity shift, one calibration factor or gain may be applied to the optical emissions data which is collected on the current plasma process Multiple dampening effects identified in accordance with the foregoing may then be addressed through multiple calibration factors or gains One or more parts of the spectra of reflected light may require the application of one calibration factor or gain thereto, while one or more parts of the subject spectra may require the application of another calibration factor or gain thereto, and so forth. Finally, the spectrometer assembly 506 may be manipulated in some manner to obtain more light, although such is not as preferred as the foregoing.
- the comparison between the spectra 690 of Figure 47A and the spectra 682 of Figure 46A indicates the existence of an intensity shift.
- the output from the spectrometer assembly 506 may be regression fit to account for the intensity shift which was identified.
- information on the presence of this intensity shift may be input to the current plasma process module 250, more specifically the pattern recognition module 370 of Figure 13, to account for this intensity shift when comparing spectra from a current plasma process with spectra in the plasma spectra directory 284.
- Limits may be utilized for adjustments made in relation to the plasma monitoring assembly 500 by any of the calibration subroutines noted herein. For instance, in the event that the amount of calibration or gain required to address an intensity shift exceeds a first limit (or a wavelength shift for that matter), a message may be displayed to the appropriate personnel that the window 478 has degraded or aged to the degree where the accuracy of the results provided by the plasma monitoring assembly 500 may be affected to an undesirable degree. Alternatively, exceeding this first limit may in fact disable the plasma monitoring assembly 500 and a corresponding indication may be provided to operations personnel. Imposing a higher limit than the noted first limit may also be used to trigger disabling of the plasma monitoring assembly 500 (i.e., warn when a first limit is exceeded, and disable when a second, higher limit is exceeded).
- the calibration light illustrated in Figure 46A and discussed above has a plurality of discrete intensity peaks. Therefore, only 31 data points may be evaluated by the subroutine 600 of Figure 45 to identify the effect that the window 478 is having on the optical emissions passing therethrough (i.e., there are only 31 intensity peaks, and the remainder is merely noise). Assumptions would have to be made as to the effect that the window 478 is having on optical emissions at those wavelengths between these peaks.
- An embodiment of a calibration light which may be used by the calibration light source 556 ( Figure 31 ) and the calibration subroutine 600 (Figure 45) to identify an intensity shift, which alleviates the need for these types of assumptions, is presented in Figure 46B.
- the calibration light of Figure 46B is of a different type than that illustrated in Figure 46A in that the calibration light of Figure 46B presents a continuum of intensity, while the calibration light of Figure 46A has a plurality of discrete intensity peaks.
- This light may be the second type used by the second calibration light source 556b noted above (the first type being the calibration light of Figure 42/46A to identify a wavelength shift).
- the calibration light for identifying an intensity shift is a white light source defined by wavelengths which include at least the Preferred Optical Bandwidth. Therefore, the calibration light source 556 may actually include one type of light source for identifying a wavelength shift (e.g., Figure 42/46A), and may use a different light source for identifying an intensity shift (e.g., Figure 46B).
- a comparison of the spectra 666 of Figure 46B and the spectra 670 of Figure 47B indicates how the window 478 is having a different effect on different portions of the spectrum. Note how the shapes of the spectra 666 and the spectra 670 are generally the same between about the 200 nm and
- the profile of the spectra 670 of Figure 47B between about the 575 nm and 950 nm wavelengths is "flatter" than the corresponding portion of the spectra 666 of Figure 46B. Therefore, the window
- 478 is having one type of effect on optical emissions generally between about the 200 nm and 575 nm wavelength region, and another different effect on optical emissions generally between about the 575 nm and 900 nm wavelength region
- FIG. 48 Another embodiment of a calibration subroutine which may be utilized by the calibration module 574 of Figure 40 is illustrated in Figure 48
- the calibration subroutine 616 is able to identify when the window 478 is having different dampening effects on different portions of the spectral data being emitted through the window 478 Moreover, the subroutine 616 is also able to identify when the window 478 is completely filtering yet different portions of the spectral data being emitted through the window 478
- the calibration subroutine 616 uses a light source which has a continuum of intensity to provide these functions, such as the light source depicted in Figure 46B
- the calibration subroutine 616 starts with step 620 which initiates a comparison between the spectra of the calibration light which is sent by the calibration light source 556 to the window 478 (hereafter "reference spectra” for purposes of the subroutine 616) (e g , Figure 46B), and the spectra of that portion of the calibration light which is reflected by the inner surface 482 of the window 478 (hereafter “reflected spectra” for purposes of the subroutine 616) (e g , Figure 47B)
- Any "change” in intensity between the reference spectra and the reflected spectra is noted at step 624 of the calibration subroutine 616
- Steps 620 and 624 may employ the same type of logic as presented by steps 580 and 584 of the calibration subroutine 576 of Figure 41 and steps 604 and 608 of the calibration subroutine 600 of Figure 45
- the "change" in intensity referenced in step 624 of the calibration subroutine 616 of Figure 48 may be a predefined tolerance or a "
- Step 628 indicates that no calibration of the plasma monitoring assembly 500 is required, and that the control is passed to the startup module 202 of Figure 15
- Step 632 analyzes the manner in which the reflected spectra differs from the reference spectra in relation to the subject intensity criteria If there was a "uniform" intensity shift of the reflected spectra in relation to the reference spectra, the calibration subroutine 616 proceeds to step 636 where a single calibration factor or uniform gam is applied to the entirety of the reflected spectra (and thereby to optical emissions data obtained during a current plasma process) to "normalize” the same to the reference spectra Control is then transferred to the startup module 202 of Figure 15 via step 640 of the calibration subroutine 616 "Uniform" in relation to step 632 of the calibration subroutine
- step 616 need not be limited to a fixed number of intensity units (e g , the entirety of the reflected spectra need not differ from the entirety of the reference spectra by the same amount), but instead may utilize a raw difference, a percentage difference, or a combination thereof as the "match limit" as those terms are used in relation to the pattern recognition module 370 of Figure 13
- any change in intensity between each wavelength in the reflected spectra which is within ⁇ 5% of the intensity of the corresponding wavelength in the reference spectra may be considered a "uniform" intensity shift for purposes of step 632 of the calibration subroutine 616
- the subroutine 616 proceeds from step 632 to step 644 Generally, step 644 is directed to determining if there is any evidence of complete filtering of data over any wavelength range within the reflected
- any "complete filtering" of any wavelength range in the reflected spectra as discussed herein will cause the calibration subroutine 616 to proceed from step 644 to step 648.
- Data within the completely filtered region is ignored in any analysis subsequently provided by the current plasma process module 250 of Figure 32 as set forth in step 648 of the calibration subroutine 616.
- at least two different calibration factors are applied to different portions of the remainder of the data in the reflected spectra as set forth in this same step 648, or the reflected spectra is normalized in relation to the reference spectra over that portion which is not being filtered.
- the analysis of any plasma process thereafter performed which is analyzed by the current plasma process module 250 will be limited to only part of the desired optical emissions data. Data within the wavelength region which is being completely filtered by the window 478 is ignored.
- the calibration subroutine 616 may be programmed to proceed from step 648 to the process alert module 428 of Figure 14 as well.
- One or more alerts may be issued through the process alert subroutine 432. For instance, it may be appropriate to apprise operations personnel at this time that the window 478 is aging and should be replaced, that the results provided by any further executions of the current plasma process module 250 may provide inaccurate results since certain data from the current plasma process will be ignored in the comparative analysis, or both.
- any exiting of step 644 of the calibration subroutine 616 as a "yes" condition may call upon the process alert module 428 to terminate all further plasma processing operations until the window 478 is replaced (not shown).
- Control of the calibration subroutine 616 of Figure 48 may pass from step 644 to step 656 if there was no complete filtering of data in the reflected spectra
- Calibration of the reflected spectra pursuant to step 656 would then entail the application of at least two different calibration factors or a plurality of gams throughout the reflected spectra, or alternatively the normalization as discussed above
- one calibration factor may be applied to the reflected spectra over the 200 nanometer to 500 nanometer wavelength region, while a different calibration factor may be applied to reflected spectra over the 501-900 nanometer wavelength region
- Calibration subroutine 616 then exits step 656 to step 660 where control may be transferred to, for instance, the startup module 202 of Figure 15
- the calibration subroutine 616 of Figure 48 may be characterized as monitoring the window 478, and based upon these results specifying how subsequent plasma processes should be evaluated If the monitoring of the condition of the window 478 through the subroutine 616 determines that there is no significant dampening of optical emissions from the chamber 474 through the window 478 or there is only dampening and not any complete filtering, the calibration subroutine 616 provides that plasma monitoring operations proceed normally by comparing the "normalized" output from the spectrometer assembly 506 with spectra in the plasma spectra directory 284 Situations where some dampening of the optical emissions from the chamber 474 through the window
- the current plasma process module 250 includes a research module 1300 which is a submodule thereof and which is presented in Figure 49
- the research module 1300 includes a research subroutine 1478 which is run to identify which character ⁇ st ⁇ c(s) of those optical emissions of the plasma in the processing chamber 36 may be indicative of the endpoint of the subject plasma step, and which may then be used by the endpoint detection module 1200 of Figures 7 and 32 to identify the occurrence of the endpoint of the subject plasma step
- the research module 1300 may be accessed through the startup module 202 of Figure 15 through execution of steps 144 and 148
- the research subroutine 1478 of Figure 49 typically is set to evaluate multiple executions of the same plasma step and identifies the optical emissions data of the plasma in the chamber 36 which may be used to call endpoint Some information about endpoint may be obtained by looking at the optical emissions data from only a single run
- the research subroutine 1478 of Figure 49 utilizes, but does not necessarily require, some knowledge of the plasma step, such as a time estimate of the length of time required to reach the endpoint of the plasma step In one embodiment, this a priori knowledge may be used such that optical emissions data is obtained on the plasma step at a point in time which should include this endpoint
- a time estimate t e for completing the subject plasma step is input to the subroutine 1478 by execution of step 1480 This time estimate for reaching endpoint may be calculated based upon, for instance, knowing the etch rate of the subject process and the thickness of the layer to be etched away
- Step 1490 of the subroutine 1478 directs that product be loaded into the processing chamber 36 and step 1482 directs that the plasma step thereafter be initiated by the introduction of plasma into the processing chamber 36 under the appropriate conditions to initiate the desired plasma step
- Optical emissions of the plasma in the processing chamber 36 during the plasma step are obtained for the subroutine 1478 at a current time t c for the subroutine 1478 through execution of step 1484, preferably using the Preferred Optical Bandwidth at the
- step 1486 Preferred Data Resolution Adjustment of the "clock" of the subroutine 1478 occurs at step 1486 where the current time t c is increased by an increment of "n "
- " ⁇ " is set at the Preferred Data Collection Time Resolution If the new current time t c is less than a preset value relating to the time estimate t e , step 1488 of the subroutine 1478 causes the subroutine 1478 to exit step 1488 and return to step 1484 for repetition in accordance with the foregoing
- the "preset value" referred to in step 1488 may be the time estimate t e from step 1480, it may be desirable to use a larger value to ensure that optical emissions data is obtained for the time when endpoint of the plasma step is actually reached
- Another way of characterizing the foregoing is that optical emissions data should be recorded until the plasma goes off in the chamber 36 to be able to identify one or more wavelengths which may be indicative of endpoint
- Steps 1484, 1486, and 1488 of the research subroutine 1478 are basically directed to obtaining optical emissions data of the entirety of the subject plasma step within the Preferred Data Collection Bandwidth, at the Preferred Data Resolution, and at the Preferred Data Collection Time Resolution
- This data could be analyzed at this time to identify characteristics in the optical emissions data which are candidates for being an indicator of endpoint
- data is obtained in accordance with the foregoing on multiple executions of this same plasma step in the same processing chamber 36
- step 1492 of the subroutine 1478 which directs that the foregoing be repeated for the desired amount of executions of the plasma step (at least 2 runs)
- the desired amount of data is obtained (the desired number of runs)
- the data is analyzed through execution of step 1496 of the research subroutine 1478 to identify that optical emissions data or a portion thereof which may be used as an indicator of endpoint This analysis is currently done manually
- the analysis referred to in step 1496 of the research subroutine 1478 may be generally directed to identifying those optical emission lines (e g , individual wavelengths) which undergo some type of a recognizable or discernible change around the time that the endpoint of the plasma step is supposed to have occurred
- optical emission lines e g , individual wavelengths
- FIGS 50A-C Each of these figures represents a plot of intensity versus time through a time which should include the endpoint of the subject plasma step for three specific wavelengths ⁇ ⁇ 2 , and ⁇ 3
- the plots were generated from optical emissions data which was obtained from one execution of the subject plasma step on a wafer 18 in the processing chamber 36
- a plot of this type is obtained for each wavelength of light which is obtained based upon the optical resolution being used in collecting the optical emissions data
- An examination of Figure 50A for the wavelength h ⁇ from this first running of the plasma step reveals that there is no real discernible change at any time during the plasma step That
- both the wavelengths ⁇ 2 and ⁇ 3 remain candidates for being endpoint indicators because they each have at least one distinct change in their respective emission lines
- Figures 51A-C present the optical emissions data for the same wavelengths ⁇ 1? ⁇ 2 , and ⁇ 3 that are presented in Figures 50A-C, but from another running of the same plasma step in the processing chamber 36
- the plot of the wavelength ⁇ ., in Figure 51 A still indicates that nothing about endpoint can be derived from this wavelength
- the plot for the wavelength ⁇ 2 in Figure 51 B is at least substantially the same as presented in Figure 50B
- the two distinct changes in its emissions line may be due, for instance, to certain changes in the process such as the opening/closing of a valve(s)
- the plot for the wavelength ⁇ 3 in Figure 51 C has the same general pattern, the two distinct changes occur about 5 seconds later than they did in the run depicted in Figure 50C This may be an indication that the wavelength ⁇ 3 is reflective of the endpoint of the subject plasma step where endpoint may vary in time by some acceptable time differential
- wavelengths which may undergo a change which corresponds with endpoint of the subject step may be identified by comparing the plots of the various wavelengths from run to run and identifying those patterns which are generally the same from run to run, but which have some type of change For instance, this change may be one or more of a temporal shift, an intensity shift, an expansion of the pattern, and a reduction in the pattern
- this change may be one or more of a temporal shift, an intensity shift, an expansion of the pattern, and a reduction in the pattern
- these types of changes may be due to factors which vary from run to run, such as the thickness of the layer which is being addressed by the plasma step which may cause a temporal shift Therefore, one way to to identify a particular wavelength which is indicative of endpoint is to run the same plasma step on multiple products, each having a different thickness of the layer to be etched away such that the time at which endpoint will occur will also vary Wavelengths which have changes which may indicative of endpoint are those where the subject change ( "around" endpoint) also shifts in time
- an endpoint detection technique is still required to use this information to identify endpoint
- One such techniques entails defining the pattern of that portion of the plot of intensity versus time for the subject wavelength(s) up to where the change occurred which was selected as being indicative of endpoint Definition of this portion of the plot may be through an equation or a function (e g , linear function, first order polynomial, second order polynomial) Endpoint may then be deemed to have been reached for subsequent executions of this same plasma step when the corresponding wavelength(s) no longer fits the equation or function
- Another option is to take the first or second derivative of this function to identify the slope of the resulting line
- the change in slope over time of the current optical emissions data may be plotted Endpoint may be called when this plot deviates by more than a predetermined amount from that identified by the first or second derivative of the subject function.
- wavelength(s) may provide some type of indication that endpoint has been reached from the optical emissions data collected on the current plasma process.
- the optical emissions data obtained for the research subroutine 1478 may be reviewed to identify a peak which changes throughout the process and then reaches a steady state at about the time that endpoint is to have occurred.
- these plots may be reviewed to identify a peak which remains at a steady state for the process and then begins to change at about the time that endpoint is to have occurred.
- a small subset of this behavior includes behavior in which a suspect wavelength or wavelengths diminish to background or a new wavelength appears at about a time where endpoint was suspected to occur.
- the endpoint of a plasma process or a discrete portion thereof may also be evident in one or more wavelength regions. Research typically must be undertaken to identify which particular wavelength region is indicative in some manner of the subject endpoint. An embodiment of an appropriate subroutine of the research module 1300 for affecting this particular function is illustrated in Figure 62.
- the research subroutine 1650 initiates with a step 1652 in which one or more endpoint evaluation wavelength regions is selected. That is, one or more wavelength regions are selected for review to determine if the same are possibly indicative of the subject endpoint.
- One way to select an endpoint evaluation wavelength region is to identify a research bandwidth to be used to review a plurality of endpoint evaluation wavelength regions over a larger wavelength region (i e , to look at a plurality of subsets of the larger wavelength region or one having a larger bandwidth)
- the research bandwidth may be selected as five nanometers
- a endpoint evaluation wavelength regions may be evaluated throughout the entire Preferred Optical Bandwidth which is again the data which is preferably collected from any given plasma process
- the Preferred Optical Bandwidth includes at least wavelengths from about 250 nanometers to about 1 ,000 nanometers Therefore a first endpoint evaluation wavelength region could be from 150-155 nanometers, a second endpoint evaluation wavelength region could be from 155-160 nanometers, a third endpoint evaluation wavelength region could be from 160-165 nanometers, and so forth until the upper extreme was reached
- the selected endpoint evaluation wavelength regions need not overlap but may do so in the above-noted manner, or they may be disposed in end-to-end fashion
- step 1654 in which product (e g , one or more wafers) is loaded into the processing chamber 36. Since endpoint may be called on processes which are run in the processing chamber 36 in the absence of product, step 1654 may be alleviated when research is being done to identify an endpoint wavelength region for these types of processes
- the plasma process is nonetheless executed as indicated by step 1656 of the research subroutine 1650
- Optical emissions data are obtained as indicated by step 1658, preferably throughout the entirety of the plasma process, and more preferably using the Preferred Optical Bandwidth, the Preferred Data
- Step 1658 indicates that a plot of an endpoint evaluation wavelength region area versus time is generated for each endpoint evaluation wavelength region, preferably over the entirety of the subject plasma process or portion thereof (i e , determine/calculate the area under the spectral pattern of the subject endpoint evaluation wavelength region at time l
- Endpoint may be indicated by the presence of some identifiable event, change, or condition in one or more of the plot(s) which are the subject of step 1662.
- step 1664 of the research subroutine 1650 indicates that if a given plot from step 1662 has only insignificant changes over the entirety thereof, the endpoint evaluation wavelength region corresponding with this particular plot may be discarded from consideration as an endpoint wavelength region candidate (e.g., the plot is effectively a horizontal line).
- step 1666 indicates that if the plot of step 1662 for any endpoint evaluation wavelength region does has some identifiable event, change, or condition of at least a certain significance, the corresponding endpoint evaluation wavelength region may be indicative of some endpoint associated with the subject plasma process.
- Each research wavelength region in Figures 63A-H has a 5 nanometer bandwidth and is within the Preferred Optical Bandwidth.
- Endpoint evaluation wavelength region 1690 of Figure 63A, endpoint evaluation wavelength region 1693 of Figure 63D, endpoint evaluation wavelength region 1695 of Figure 63F, and endpoint evaluation wavelength region 1697 of Figure 63H each have a plot 1698 of change in area versus time which is substantially horizontal or which does not deviate from a horizontal reference line by more than a predetermined amount (e.g., using a raw difference, a percentage difference basis, or both as described herein).
- step 1664 of the research subroutine 1650 This may be used to define "insignificant” for purposes of step 1664 of the research subroutine 1650, and as such the endpoint evaluation wavelength regions 1690, 1693, 1695, and 1697 may be discarded or not considered as endpoint indicators through step 1664 of the research subroutine 1650 of Figure 62.
- endpoint evaluation wavelength regions 1691 , 1692, 1694, and 1696 each have a plot 1698 which exceeds this tolerance and therefore may be indicative of some endpoint associated with the subject plasma process through step 1666 of the research subroutine 1650 of Figure 62 (e g , an endpoint indicator candidate)
- Endpoint evaluation wavelength region 1691 of Figure 63B and endpoint evaluation wavelength region 1694 of Figure 63E each have a plot 1698 of the change in area over time which define a bell curve of sorts that is centered at time t
- Endpoint evaluation wavelength region 1692 of Figure 63C has a plot 1698 of change in area versus time which initially proceeds at least substantially horizontally, spikes up rather abruptly at time t,, and then proceeds again in at least substantially horizontal fashion
- the wavelength region which is used to call endpoint may be defined to include both of these endpoint evaluation wavelength regions (e g , if one endpoint evaluation wavelength region was 200- 205 nanometers and another endpoint evaluation wavelength region was 210- 215 nanometers, each had a plot of change in area over time which had the requisite "significant event" at about the relevant time period into the plasma process, the endpoint wavelength region could be selected as 200-215 nanometers)
- Endpoint Detection Module 1200 Figures 52-58 and 64
- the actual detection of the endpoint of a plasma process e g , plasma clean, conditioning wafer operation
- discrete portion thereof e g , one or more steps of a plasma recipe, all steps of a multi-step plasma recipe
- the endpoint detection module 1200 may be used to identify the endpoint of a plasma process having only a single endpoint Multi-step plasma processes having corresponding multiple endpoints may also be evaluated through the endpoint detection module 1200
- Two or more plasma steps of a multiple step plasma process may each have their respective endpoints identified through the endpoint detection module 1200 All steps of a multiple step plasma process having endpoints also may each be identified through the endpoint detection module 1200 as well
- Endpoint detection module 1200 may interface with the process alert module 428 of Figure 14 For instance, when endpoint is identified through the endpoint detection module 1200, information on the endpoint condition may be provided to the appropriate personnel through the "alert" function of the process alert subroutine 432 of Figure 14 Control of the plasma process in relation to the identification
- the current plasma process is evaluated both in relation to its health (i e , the plasma health in accordance with the plasma health module 252 discussed above) and in relation to endpoint through the endpoint detection module 1200
- the endpoint detection module 1200 need not be initiated at the start of the plasma process
- the endpoint module 1200 preferably begins its evaluation of the current plasma process to identify its endpo ⁇ nt(s) at an intermediate time in the process (e g , at least ⁇ A way into the subject process or portion thereof)
- an intermediate time in the process e g , at least ⁇ A way into the subject process or portion thereof
- the plasma health evaluation and endpoint evaluation need not be executed at the same frequency or Analytical Time Resolution.
- the amount of time between which the health of the plasma process is checked may be greater than the time between which the plasma process is checked to identify endpoint.
- the optical emissions of the plasma may be checked at least at every 1 second for a plasma health evaluation, and may be checked at least at every 300 milliseconds to identify endpoint.
- each of the endpoint detection subroutines may be addressed below which also relates to plasma health evaluations by the plasma health module 252. If the plasma health module 252 identifies either an error or unknown condition in relation to the current plasma process, this may have an effect on the operation of the endpoint detection module 1200. For instance, the endpoint detection module 1200 may be configured to respond to such a condition by displaying information to the appropriate personnel that endpoint will not be called because of the identification of the error or unknown condition by the plasma health module 252.
- identification of an error or unknown condition through the plasma health module 252 may terminate the ability of the endpoint detection module 1200 to affect any changes in the plasma process, to terminate a current plasma process after identifying its endpoint, to initiate the next plasma process, or both. That is, the endpoint detection module 1200 may be "shut off' or disabled if the plasma health is found to be unacceptable by the plasma health module 252 since endpoint information obtained under these circumstances may be unreliable.
- endpoint detection subroutine 1456 is used to determine when a given plasma step of a plasma recipe has affected its intended purpose or achieved the desired result A current spectra of the plasma in the processing chamber
- step 1464 An assessment of this spectra from the processing chamber 36 with the endpoint subdirectory 316 is undertaken at step 1464 to determine if endpoint has been reached.
- One technique contemplated by step 1464 is to determine if this spectra from the processing chamber 36, or at least a portion thereof, is a "match" with at least one spectra in the endpoint subdirectory 316 of Figure 9
- Optical emissions data over a wavelength range e g , the Preferred Optical Bandwidth using the Preferred Data Resolution
- the pattern of the spectra of the plasma in the processing chamber 36 will typically remain substantially constant for a time after endpoint has been reached Analyzing the data from a previous execution of this plasma step in the same processing chamber 36 may therefore allow for identification of a
- One or more wavelengths which were identified by the research module 1300 of Figure 49 may have their respective plots of intensity versus time from a previous execution of the same plasma step in the same processing chamber 36 defined by an equation or function
- the assessment through step 1464 may be to plot the optical emissions data of this wavelength in the subsequent execution of the same plasma process in the same chamber 36 and to determine if this data indeed fits with this equation or function If so, endpoint may be called when the current optical data no longer "fits" the equation or function (e g , employing linear fit techniques, polynomial fit technique, etc) which should be at about the time where the "change" discussed above in relation to the research module 1300 occurred
- Other techniques which may be used to assess this type of current data would be to utilize a first derivative or a second derivative of this equation or function to define a linear function, such that deviations from this function may be more readily identified in some cases as noted above
- the wavelengths which are indicative of endpoint are identified through the research module 1300 of Figure 49, certain relational characteristics of these particular wavelengths may be noted as well
- the intensity of the peak of this wavelength may be noted to identify the same in any subsequent executions of the same plasma process
- the subject wavelength may represent the largest peak within a certain wavelength region Therefore, the subject wavelength may be found in these subsequent executions by looking for the largest peak in a certain area of the spectrum
- the "position" of the peak of the subject wavelength may be identified in relation to one or more other peaks
- the subject wavelength may be represented by a peak which is located between two larger peaks in a certain wavelength region Therefore, the subject wavelength may be found in the subsequent executions by looking for this pattern in the optical emissions of the plasma
- Step 1476 resets the "clock" of the subroutine 1456 by increasing the current time t c by a factor of "n "
- the magnitude of "n” defines the Analytical Time Resolution Step 1470 makes a determination as to whether the current plasma process has been terminated.
- the subroutine 1456 returns to step 1460 where another spectra is obtained for the subroutine 1456 at this new current time t c for a repetition of the above-described analysis If the plasma process has been terminated, the subroutine 1456 proceeds from step 1468 to steps 1476 and 1470 which may be executed in any order Step 1476 resets the "clock" of the subroutine 1456 by increasing the current time t c by a factor of "n " The magnitude of "n" defines the Analytical Time Resolution Step 1470 makes a determination as to whether the current plasma process has been terminated The same types of techniques discussed above in relation to determining when the plasma is "on" in the chamber 36 may be used in step 1470 So long as the current plasma process has not been terminated
- step 1466 control of the plasma monitoring operations may be returned to, for instance, the start-up module 202 of Figure 15 If no endpoint has been detected by the time at which the plasma process is terminated, information may be provided to personnel that endpoint was not detected and that there may have been some error or aberration in the process, even though the plasma health module 252 did not necessarily identify this condition
- step 1468 identifies endpoint through steps 1464 and 1468
- step 1474 the process alert module 428 of Figure 14 is called
- actions may be taken to apprise personnel that the endpoint of the subject plasma step has been reached (through execution of steps 454 and 458 of the process alert subroutine 432), actions may be taken in relation to control of the plasma process (e g , initiating the next plasma step or terminating the plasma recipe if the subject step is the last step of the recipe), or both Whether through the process alert module 428 or through step 1472 of the endpoint detection subroutine 1456, control of plasma monitoring operations may then be returned to, for instance, the startup module 202 of Figure 15
- FIG. 53 Another embodiment of an endpoint detection subroutine which may be accessed through the endpoint module 1200 is presented in Figure 53
- the endpoint detection subroutine 1506 of Figure 53 initiates at step 1510 where a spectra of the plasma in the processing chamber 36 at the current time t c is obtained for the subroutine 1506
- This spectra and a "reference" spectra are subtracted from each other at step 1514 Only the differential is important in relation to step 1514 That is, it is not of particular importance whether the current spectra is subtracted from the "reference" spectra or vice versa
- both the reference spectra and the spectra on the current plasma process which is obtained are defined by the Preferred Optical Bandwidth and
- the "reference" spectra for step 1514 may be retrieved from the endpoint subdirectory 316 of Figure 9
- “reference" spectra for step 1514 of the endpoint detection subroutine 1506 is a spectra which is associated with the same plasma process That is, the spectra in the endpoint subdirectory 316 which is involved in the "subtraction” of step 1514 would be a spectra from a previous execution of the same plasma process in the same processing chamber 36 at a time in the plasma process when endpoint is at least assumed to have occurred How this spectra is selected for inclusion in the endpoint subdirectory 316 may be in accordance with the discussion presented above Moreover, information provided by the operator, by direct communications between the wafer production system 2 and the current plasma process module 250, or by the pattern recognition module
- step 1514 Another option for the "reference" spectra for purposes of step 1514 is a previous-in-time spectra from the same plasma process in which the current spectra is obtained at step 1510 (e g , the immediately preceding current time t c ) For instance, the spectra at the current time t c and the spectra at the current time t c.n may be subtracted from each other at step 1514
- the subtraction operation involving the current spectra of the plasma in the processing chamber 36 at the current time t c and the "reference" spectra through execution of step 1514 of the endpoint detection subroutine 1514 generates an output which is indicative of the differential If the subroutine 1506 determines that this differential is within a certain predetermined tolerance at step 1518, endpoint
- Step 1522 resets the "clock" of the subroutine 1506 by increasing the current time t c by a factor of "n "
- the magnitude of "n” defines the Analytical Time Resolution Step 1524 makes a determination as to whether the current plasma process has been terminated
- the subroutine 1506 returns to step 1510 where another spectra is obtained for the subroutine 1506 at this new current time t c for a repetition of the above-described analysis If the plasma process has been terminated, the subroutine 1506 proceeds from step 1524 to step 1528 where control of the plasma monitoring operations may be returned to, for instance, the start-up module 202 of Figure 15
- the "differential" referred to in step 1518 of the endpoint detection subroutine 1506 of Figure 53 may be compared with a predetermined tolerance, such as baseline intensity A raw difference basis, a percentage difference basis, or both may be used for establishing this tolerance
- a predetermined tolerance such as baseline intensity A raw difference basis, a percentage difference basis, or both may be used for establishing this tolerance
- the subroutine 1506 may be directed to proceed from step 1518 to step 1530
- Another way of saying this is that endpoint is deemed to have been reached for purposes of the endpoint detection subroutine 1506 when there are no longer any substantial peaks in the differential defined by step 1514
- Figures 54A-C, 55A-C, and 56A-C The spectra 1496a of Figure 54A is representative of plasma in the processing chamber 36 in the initial part (e g , for a current time t 0 ) of a plasma step being run on product within the processing chamber 36 (e g , the spectra of Figure 54A is
- the spectra 1496b of Figure 55A is representative of plasma in the processing chamber 36 at an intermediate time (e g , for a current time t 30 ) in the same plasma step presented in Figure 54A (i e , the spectra of Figure 55A is the spectra at the current time t c from step 1510 of the endpoint detection subroutine 1506 of Figure 53)
- the spectra 1500 of Figure 55B is again the ' reference" spectra from step 1514 of the endpoint detection subroutine 1506, and
- Figure 55C is an illustration of the "difference" between the spectra 1496b of Figure 55A and the spectra 1500 of Figure 55B which is presented in the nature of an output 1504b
- the output 1504b in Figure 55C still has substantial peaks therein which is indicative that endpoint has not yet been reached based upon the logic of the endpoint detection subroutine 1506
- the plasma step has progressed in that the size of the peaks in the output 1504b of Figure 55C is less than
- Optical emissions data are obtained for the endpoint detection subroutine 1204 of Figure 57 at its step 1208 These optical emissions data are more specifically of the plasma in the processing chamber 36 at a time t c1
- Two other time-related variables are introduced at step 1212
- a time t c1+n is set equal to a time t c2 at step 1212
- the time t c2 is greater than the time t cl by an increment of "n "
- the magnitude of "n” for the subroutine defines the Analytical Time Resolution
- Optical emissions of the plasma in the processing chamber 36 are obtained for this time t c2 through execution of step 1216 of the endpoint detection subroutine 1204
- the intensities of the optical emissions from the current values for the times t c1 and t c2 are compared with each other at step 1220 of the endpoint detection subroutine 1204 If the differential between the optical emissions at these two times is less than a predetermined amount (e g , an increase or a
- the differential between the optical emissions associated with the times t c1 and t c2 is ever more than the predetermined amount associated with step 1224 of the endpoint detection subroutine 1204, this may be an indication of the above-noted "modal" change in the plasma which is in turn indicative of endpoint It is not definitively indicative of the type of "modal” change at this point in time however Only those "modal" changes associated with the plasma which occur at about the time that endpoint is estimated to occur, which appear quickly or abruptly, and which are persistently observed in subsequent executions of the same plasma process are indicative of a change in impedance which occurs at endpoint Therefore, it may be desirable to execute a plurality of runs for each particular plasma process in which the subroutine 1204 is used before relying upon the subroutine 1204 to call endpoint by execution of the process alert module 428 through execution of step 1232
- the subroutine 1204 may also be implemented to confirm or increase the confidence level that endpoint has been reached when another endpoint detection technique is being used to call endpoint as well
- the "modal" change associated with the endpoint detection subroutine 1204 of Figure 57 may be monitored in relation to one or more specific wavelengths, but by definition impacts plasma performance and is detectable throughout the optical emissions range (although some wavelengths may demonstrate a stronger effect of the modal change than others)
- the optical emissions associated with the differential referred to in step 1224 may be in relation to a single wavelength
- the optical emissions associated with the differential referred to in step 1224 may be from a plurality of wavelengths, and the "differential" which is potentially indicative of a change in impedance may be when one or more of these wavelengths demonstrates the above-noted "modal" change.
- Another option is to monitor for the "modal" change over a certain bandwidth.
- Figure 58 illustrates the change in intensity of the plasma in the chamber 36 over time throughout the plasma step. More specifically, each point of the output 1260 illustrates the differential between the optical emissions of the plasma in the chamber 36 over the Preferred Optical Bandwidth and at the Preferred Data Resolution of a current time (e.g., t c ) and a preceding time (e.g., t c.n ). Preferably, each data point which defines the output 1260 is the differential between two adjacent times where optical emissions data is obtained from the chamber 36. As can be seen in the output 1260 from Figure 58, there is a significant change in the intensity of the plasma from about the 35 second mark and the 40 second mark. This is the type of "modal" change which may be associated with a change in impedance associated with the processing chamber 36, which in turn the endpoint detection subroutine 1204 of Figure 57 associates with endpoint.
- modal which may be associated with a change in impedance associated with the processing chamber 36, which in turn the endpoint detection subroutine
- FIG. 64 is illustrated in Figure 64 in the form of an endpoint detection subroutine 1670.
- the subroutine 1670 includes preparation or initialization steps 1672 and
- Step 1672 directs that the endpoint wavelength region(s) be selected.
- One or more endpoint wavelength regions may be used to call a specific endpoint within the subject plasma process.
- Endpoint "A” may be called when indicated by relevant plots of either of the endpoint wavelength regions or only when indicated by the relevant plots of all relevant endpoint wavelength regions.
- One endpoint wavelength region may be designated as the primary indicator for endpoint "A", and any other endpoint wavelength region associated with the same endpoint may be used merely to increase the confidence level that the call on endpoint "A" by the primary endpoint wavelength region was correct.
- Step 1674 is another initialization step of sorts in that it is directed toward the selection of how a certain pattern associated with the endpoint wavelength region selected in step 1672 will be identified, and which has been determined to be indicative of a specific endpoint (e g through the research module 1300)
- a specific endpoint e g through the research module 1300
- Various alternatives have already been discussed in relation to how this event may be identified, and such may also be used by the endpoint detection subroutine 1670 of Figure 64 Whether a specific endpoint is being called based upon the behavior of a specific wavelength or a wavelength region, one way of identifying a "pattern" of this specific wavelength or wavelength region which is indicative of the specific endpoint is through algorithms as indicated by step 1674
- a number of algorithms may be utilized to identify the relevant pattern and equate the same with endpoint, and therefore these algorithms will be characterized as endpoint algorithms
- One type of endpoint algorithm which is appropriate in certain situations is one that identifies when the negative slope of the plot associated with the endpoint wavelength/wavelength region exceed
- optical emissions of the plasma are obtained as indicated by step 1678
- these optical emissions encompass the Preferred Optical Bandwidth, and further are obtained using the Preferred Data Resolution and Preferred Data Collection Time Resolution
- This optical emissions data is preferably stored on a computer- readable storage medium, such as in a log file and/or in the plasma spectra directory 284
- the various wavelengths comprising the optical emissions data are cataloged in some manner so as to be independently identifiable
- the endpo ⁇ nt(s) of the subject plasma process may then be called through execution of step 1680 and in any of the above-described manners (e g , through any of the endpoint algorithms A1 , A2, B1 , B2)
- certain designated data may be retrieved or isolated for use by the endpoint detection module 1200
- multiple endpoint wavelength regions may be selected at step 1672 to call multiple endpoints in the plasma process of step 1676 (e g , the multiple steps of a plasma recipe)
- step 1676 e g , the multiple steps of a plasma recipe
- no hardware changes are required for any plasma monitoring system using the endpoint detection subroutine 1670 to change the wavelength region which is desired for use in calling a specific endpoint This may then be characterized as a virtual bandpass filter or virtual filtering technique
- optical emissions are preferably collected on any plasma process at the Preferred
- any specific portion thereof may be called upon or retrieved by the endpoint module 1200 for use in calling endpoint (e g , any portion of the collected data may be used by any of endpoint algorithms A1 , A2, B1 , B2) Since this is done without adding a filter to obtain the desired optical emissions, the endpoint module 1200 has virtual bandpass filter capabilities (i e , collecting a certain amount of data on a plasma process, and then only selecting only a portion of this data to monitor plasma health and/or to call one or more endpoints, provides a filtering function which does not require changes to the optical system, and may then be described as being done “virtually”) This makes the endpoint module 1200 very “generic" in that it may be used to
- Wafer Distribution Module 1384 - Figures 59-60 Various of the above-noted evaluations provided by the current plasma process module 250 may provide information to the wafer distribution module 1384 of Figures 59-60 to have some type of effect on the distribution of wafers to the various process chambers 36 of the wafer production system 2
- One embodiment of a subroutine which may be used by the wafer distribution module 1384 is illustrated in Figure 59
- the subroutine 1388 of Figure 55 includes steps 1392, 1396, 1400, and 1402 in which the protocol is for the wafer distribution subroutine 1388 to proceed to the plasma process product module 252 for each of chambers 36a-d ( Figure 1 )
- the subroutine 1388 may of course accommodate wafer production systems having different numbers of processing chambers 36 Monitoring of the current plasma recipes being run on product in each of these processing chambers 38a-d is included in the protocol of 1404 Any deviation from the normal spectra subdirectory 288 of the current plasma recipe being run on product in any of the chambers 36a-d is noted
- Plasma Monitoring Network 1610 - Figures 65-66 Wafer fabrication facilities will typically utilize a plurality of wafer production systems 2 (e g , a plurality of clusters of chambers 36, with a cluster of processing chambers 36 being served by a common wafer handling assembly 44 and typically controlled by a common MCU 58)
- a plurality of wafer production systems 2 is illustrated in the form of a plasma monitoring network 1610 which is presented in Figure 65
- Four wafer production systems 2a-d are illustrated in Figure 65, although any number of wafer production systems 2 may be networked utilizing principles of the plasma monitoring network 1610
- Each wafer production system 2 on the network 1610 is illustrated as being contained within a single clean room 1646 defined by an appropriate enclosing barrier 1648 However, the plasma monitoring network
- 1610 may be used to interconnect wafer production systems 2 which are located in any number of clean rooms 1646 (not shown)
- Each wafer production system 2 is illustrated as having its own mam control unit 58 and plasma monitor control unit 128 which as noted may be appropriately operatively interconnected to control operations in at least some manner based upon the monitoring of plasma processing operations (not shown)
- the plasma monitor control unit 128 for a given wafer production system 2 may be configured so as to be able to simultaneously monitor plasma processes being conducted in each of the chambers 36 within its system 2, and further a given plasma monitoring control unit 128 for a certain wafer production system 2 may be controlled on a chamber-by-chamber basis (i e , even though a single plasma monitoring control unit 128 may be used for a plurality of chambers 36, the plasma monitoring control unit 128 may be configured so that plasma monitoring operations in each of these chambers 36 will be independently controllable).
- the main control unit 58 and plasma monitoring control unit 128 of each given wafer production system 2 are both operatively interconnected with a remote chamber cluster station 1614 which is located outside of the clean room
- a remote master station 1630 is also located outside of the clean room 1646 and is operatively interconnected with the main control unit 58 and plasma monitor control unit 128 of each wafer production system 2 on the plasma monitoring network 1610.
- Each remote chamber cluster station 1614 and the remote master station 1630 may have an appropriate display (e.g., computer monitor) and data entry device (e.g., keyboard, mouse).
- Access to the main control unit 58 of a particular wafer production system 2 and the plasma monitoring control unit(s) 128 of a given wafer production system 2 is controlled through an access module 1682.
- the access module 1682 in turn has a plurality of modules or accessible items. Flexibility exists as to what may be done on each of the remote chamber cluster stations 1614 and the remote master station 1630. That is, the plasma monitoring network 1610 may be "programmed" to tailor which of the modules illustrated in Figure 66 that each of the remote chamber cluster stations 1614 and remote master station 1630 may access.
- the access to the modules of Figure 66 may be different for the remote master station 1630 and each of the remote chamber cluster stations 1614, the access to the modules of Figure 66 may be different between one or more of the remote chamber cluster stations 1614, or any combination thereof.
- the modules illustrated in Figure 66 also may be made available to or incorporated in any one or more of the plasma monitoring control units 128 within the clean room 1646, whether included on a plasma monitoring network 1510 or in a non-networked environment. In many cases, neither any of the remote chamber cluster stations 1614 nor the remote master station 1630 will have access to any main control unit 58 to control the operation of the same.
- this type of access may be controlled by its associated access module 1682 through a process control module 1681
- a process control module 1681 When a given process control module 1681 is "on”, access to the corresponding MCU 58 will be made available Conversely, when a given process control module 1681 is "off,” access to the corresponding MCU 58 will be denied
- This type of access is also controlled through their corresponding access module 1682
- One of the modules which may be made available in the above-noted manner on the plasma monitoring network 1610 is a data player module 1684
- the data player module 1684 basically allows the optical emissions data that was recorded in the plasma spectra directory 284 or otherwise in a data log file from a plasma process previously executed on a certain chamber 36 to be "rerun” in a sense on the remote master station 1630 and/or the corresponding remote chamber cluster station 1614 That is, this same data will be reused by the data player module
- Operational control of a certain plasma monitoring control unit 128 at the chamber 36 may be affected through a remote control module 1688 which is also illustrated in Figure 66
- a certain remote chamber cluster station 1614, the remote master station 1630, or any combination thereof may be allowed to make certain changes to any one or more of the sub-modules of any one or more of the plasma monitoring modules 200 of the subject plasma monitoring control unit 128 at the chamber(s) 36 to affect how the same will operate on the current plasma process or the next current plasma process run in the subject chamber(s) 36
- the remote chamber cluster stations 1614 will not have access to their respective plasma monitoring control unit 128 for any of their respective chambers 36 through the remote control module 1688 (i e , having a given
- a process review module 1689 may be made available to any one or more of the remote chamber cluster stations 1614, the remote master station 1630, or any combination thereof to allow the same to view the data on the plasma process currently being run on designated chambers 36 within a designated wafer production system 2 to which access has been provided through the plasma monitoring network 1610 via the process review module 1690
- a process review module 1690 may allow the remote master station 1630 to view a plasma process being conducted within any of the processing chambers 36 in real-time (e g , data reflective of the plasma process or any portion thereof, and including the plot of the endpoint wavelength(s)/wavelength reg ⁇ on(s)), including reviewing two or more processes simultaneously
- the remote chamber cluster stations 1614 will typically only have access through the process review module 1690 to view plasma processes being conducted within chambers 36 of its own wafer production system 2, although process review capabilities on any one or more of the remote chamber cluster stations 1614 could be expanded to review plasma processes being conducted in different wafer production systems 2 as well.
- the plasma monitoring assemblies 174, 500, and 700 as described above offers numerous advantages.
- One advantage is that the logic of the current plasma process module 250 and its various sub-modules can be used with any type of plasma chamber and further can be used to evaluate any type of plasma process. No significant adaptation of the logic of the module 250 or any of its sub-modules is required to use the module 250 on one chamber type/design, and then to turn around and use this same module 250 on another chamber type/design. Similarly, no adaptation of the module 250 or any of its sub- modules is required to use the module 250 on one type of plasma process, and then to turn around and use this same module 250 on a plasma process which is somehow different. In this sense the module 250 is a generic plasma monitoring tool.
- Another advantage of the plasma monitoring assemblies 174, 500, 700 is that the data which is used to evaluate the plasma processes conducted within a processing chamber is taken from this very same processing chamber when running the very same type of plasma process. In this sense when the assemblies 174, 500, and 700 are installed on a particular processing chamber, the assemblies 174, 500, 700 become specific to the chamber. The fact that a given plasma recipe will behave in one way on one chamber and another way on another chamber is of no significance to the plasma monitoring assemblies 174, 500, 700. Each assembly 174, 500, 700 adapts to the idiosyncrasies of its associated chamber and learns about the chamber through the running of actual plasma processes on the chamber to build up the plasma spectra directory 284 for that particular chamber. Again, each chamber will have its own plasma spectra directory 284 or the like in that all information used to evaluation a given plasma process on a given plasma chamber will have been obtained from this very same processing chamber.
- the module 250 has the ability to: determine if the current plasma process is proceeding like one or more previous runnings of this same plasma process on this same chamber, determine if the endpoint of a given plasma step has been reached, determine if the endpoint of each plasma step in a plasma recipe has been reached, identify the current plasma recipe to operations personnel, identify the current plasma step to operations personnel, determine when a chamber should be cleaned, and to determine when each of a plasma cleaning, and conditioning wafer operation may be terminated How this information is used will typically be left up to the discretion of the operator of the facility incorporating the wafer production system 2 For instance, the current plasma process module 250 may simply be used a source of information on the current plasma process The current plasma process module 250 may also take a more active roil in the operation of the wafer production system 2 by integrating the module 250 with process controls for the system 2, such as to automate control of plasma process operations based upon information provided by the module 250 Various
Abstract
Description
Claims
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
AU36630/99A AU3663099A (en) | 1998-04-23 | 1999-04-23 | Method and apparatus for monitoring plasma processing operations |
KR1020007011793A KR20010042965A (en) | 1998-04-23 | 1999-04-23 | Method and apparatus for monitoring plasma processing operations |
KR10-2004-7006151A KR20040053203A (en) | 1998-04-23 | 1999-04-23 | Method and apparatus for monitoring plasma processing operations |
EP99918803A EP1105703A4 (en) | 1998-04-23 | 1999-04-23 | Method and apparatus for monitoring plasma processing operations |
JP2000544994A JP2003524753A (en) | 1998-04-23 | 1999-04-23 | Method and apparatus for monitoring a plasma processing operation |
Applications Claiming Priority (40)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US6497298A | 1998-04-23 | 1998-04-23 | |
US6524598A | 1998-04-23 | 1998-04-23 | |
US09/065,247 | 1998-04-23 | ||
US09/064,793 | 1998-04-23 | ||
US09/065,362 | 1998-04-23 | ||
US09/065,274 | 1998-04-23 | ||
US09/065,307 US6261470B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,274 US6275740B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,991 | 1998-04-23 | ||
US09/065,680 US6077386A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,957 | 1998-04-23 | ||
US09/065,247 US6169933B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,965 | 1998-04-23 | ||
US09/065,358 | 1998-04-23 | ||
US09/065,359 | 1998-04-23 | ||
US09/065,257 US6123983A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,793 US6134005A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,680 | 1998-04-23 | ||
US09/065,359 US6157447A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,195 | 1998-04-23 | ||
US09/065,358 US6419801B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,957 US6221679B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,991 US6246473B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,006 US6090302A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,006 | 1998-04-23 | ||
US09/064,970 | 1998-04-23 | ||
US09/064,966 US6165312A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,245 | 1998-04-23 | ||
US09/064,965 US6254717B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,972 | 1998-04-23 | ||
US09/065,203 US6192826B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,362 US6132577A (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,970 US6269278B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/065,203 | 1998-04-23 | ||
US09/065,195 US6223755B1 (en) | 1998-04-23 | 1998-04-23 | Method and apparatus for monitoring plasma processing operations |
US09/064,966 | 1998-04-23 | ||
US09/065,257 | 1998-04-23 | ||
US09/065,307 | 1998-04-23 | ||
US09/290,903 | 1999-04-12 | ||
US09/290,903 US6383402B1 (en) | 1998-04-23 | 1999-04-12 | Method and apparatus for monitoring plasma processing operations |
Publications (2)
Publication Number | Publication Date |
---|---|
WO1999054694A1 true WO1999054694A1 (en) | 1999-10-28 |
WO1999054694A9 WO1999054694A9 (en) | 2001-08-09 |
Family
ID=27586430
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/US1999/008894 WO1999054694A1 (en) | 1998-04-23 | 1999-04-23 | Method and apparatus for monitoring plasma processing operations |
Country Status (4)
Country | Link |
---|---|
EP (1) | EP1105703A4 (en) |
JP (1) | JP2003524753A (en) |
KR (2) | KR20040053203A (en) |
WO (1) | WO1999054694A1 (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001040535A2 (en) * | 1999-12-03 | 2001-06-07 | The Regents Of The University Of California | System and method relating to vapor deposition |
WO2008009165A1 (en) * | 2006-07-03 | 2008-01-24 | He Jian Technology(Suzhou)Co.Ltd. | AN OPTICAL INSPECTING METHOD OF A PLASMA PROCESSING DEGREE OF A SiON FILM |
CN110246743A (en) * | 2014-10-20 | 2019-09-17 | 朗姆研究公司 | The system and method for detection processing point in multi-mode pulse processing |
WO2020219208A1 (en) * | 2019-04-26 | 2020-10-29 | Applied Materials, Inc. | Methods for calibrating an optical emission spectrometer |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7672747B2 (en) | 2000-03-30 | 2010-03-02 | Lam Research Corporation | Recipe-and-component control module and methods thereof |
AU2003262768A1 (en) * | 2002-09-30 | 2004-04-23 | Tokyo Electron Limited | Apparatus and method for use of optical system with plasma proc essing system |
US7565220B2 (en) | 2006-09-28 | 2009-07-21 | Lam Research Corporation | Targeted data collection architecture |
US7814046B2 (en) | 2006-09-29 | 2010-10-12 | Lam Research Corporation | Dynamic component-tracking system and methods therefor |
TWI379074B (en) * | 2007-05-07 | 2012-12-11 | Verity Instr Inc | Calibration method of a radiometric optical monitoring system used for fault detection and process monitoring |
CN104266324A (en) * | 2014-10-15 | 2015-01-07 | 北京雷克利达机电股份有限公司 | System and method capable of reminding user to replace fresh air haze removal machine HEPA (High Efficiency Particulate Air) filter |
CN107466420B (en) * | 2015-04-23 | 2019-10-25 | 应用材料公司 | In-situ etch rate for chamber clean terminal determines |
KR102543349B1 (en) * | 2016-07-11 | 2023-06-30 | 삼성전자주식회사 | Plasma monitoring apparatus |
CN107782447B (en) * | 2017-09-14 | 2019-11-15 | 中国科学院长春光学精密机械与物理研究所 | Space dimension automatic identifying method and system in imaging spectrometer spectral calibration |
JP6762927B2 (en) * | 2017-12-19 | 2020-09-30 | 株式会社日立ハイテク | Signal processing device and signal processing method |
CN112458440B (en) | 2020-11-18 | 2022-11-25 | 北京北方华创微电子装备有限公司 | Semiconductor process equipment, reaction chamber thereof and film deposition method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5014217A (en) * | 1989-02-09 | 1991-05-07 | S C Technology, Inc. | Apparatus and method for automatically identifying chemical species within a plasma reactor environment |
US5871658A (en) * | 1997-01-13 | 1999-02-16 | Taiwan Semiconductor Manufacturing Company, Ltd. | Optical emisson spectroscopy (OES) method for monitoring and controlling plasma etch process when forming patterned layers |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5347460A (en) * | 1992-08-25 | 1994-09-13 | International Business Machines Corporation | Method and system employing optical emission spectroscopy for monitoring and controlling semiconductor fabrication |
JPH08232087A (en) * | 1994-12-08 | 1996-09-10 | Sumitomo Metal Ind Ltd | Method for detecting end point of etching and etching device |
-
1999
- 1999-04-23 WO PCT/US1999/008894 patent/WO1999054694A1/en active IP Right Grant
- 1999-04-23 KR KR10-2004-7006151A patent/KR20040053203A/en not_active Application Discontinuation
- 1999-04-23 KR KR1020007011793A patent/KR20010042965A/en active IP Right Grant
- 1999-04-23 EP EP99918803A patent/EP1105703A4/en not_active Withdrawn
- 1999-04-23 JP JP2000544994A patent/JP2003524753A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5014217A (en) * | 1989-02-09 | 1991-05-07 | S C Technology, Inc. | Apparatus and method for automatically identifying chemical species within a plasma reactor environment |
US5871658A (en) * | 1997-01-13 | 1999-02-16 | Taiwan Semiconductor Manufacturing Company, Ltd. | Optical emisson spectroscopy (OES) method for monitoring and controlling plasma etch process when forming patterned layers |
Non-Patent Citations (1)
Title |
---|
See also references of EP1105703A4 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2001040535A2 (en) * | 1999-12-03 | 2001-06-07 | The Regents Of The University Of California | System and method relating to vapor deposition |
WO2001040535A3 (en) * | 1999-12-03 | 2001-12-13 | Univ California | System and method relating to vapor deposition |
WO2008009165A1 (en) * | 2006-07-03 | 2008-01-24 | He Jian Technology(Suzhou)Co.Ltd. | AN OPTICAL INSPECTING METHOD OF A PLASMA PROCESSING DEGREE OF A SiON FILM |
CN110246743A (en) * | 2014-10-20 | 2019-09-17 | 朗姆研究公司 | The system and method for detection processing point in multi-mode pulse processing |
WO2020219208A1 (en) * | 2019-04-26 | 2020-10-29 | Applied Materials, Inc. | Methods for calibrating an optical emission spectrometer |
CN113748322A (en) * | 2019-04-26 | 2021-12-03 | 应用材料公司 | Method for calibrating an optical emission spectrometer |
US11927482B2 (en) | 2019-04-26 | 2024-03-12 | Applied Materials, Inc. | Methods for calibrating an optical emission spectrometer |
Also Published As
Publication number | Publication date |
---|---|
KR20040053203A (en) | 2004-06-23 |
WO1999054694A9 (en) | 2001-08-09 |
JP2003524753A (en) | 2003-08-19 |
EP1105703A1 (en) | 2001-06-13 |
KR20010042965A (en) | 2001-05-25 |
EP1105703A4 (en) | 2005-08-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6077386A (en) | Method and apparatus for monitoring plasma processing operations | |
US6805810B2 (en) | Method & apparatus for monitoring plasma processing operations | |
US6090302A (en) | Method and apparatus for monitoring plasma processing operations | |
US6246473B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6419801B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6123983A (en) | Method and apparatus for monitoring plasma processing operations | |
US6132577A (en) | Method and apparatus for monitoring plasma processing operations | |
EP1105703A1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6275740B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6134005A (en) | Method and apparatus for monitoring plasma processing operations | |
US6261470B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6165312A (en) | Method and apparatus for monitoring plasma processing operations | |
US7695987B2 (en) | Method for automatic determination of semiconductor plasma chamber matching and source of fault by comprehensive plasma monitoring | |
KR0152355B1 (en) | Plasma processing method and its device | |
US6223755B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US5910011A (en) | Method and apparatus for monitoring processes using multiple parameters of a semiconductor wafer processing system | |
US7230720B2 (en) | Film thickness measuring method of member to be processed using emission spectroscopy and processing method of the member using the measuring method | |
US20100297783A1 (en) | Plasma Processing Method | |
US6269278B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6192826B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6157447A (en) | Method and apparatus for monitoring plasma processing operations | |
US7312857B2 (en) | Method and system for monitoring plasma using optical emission spectrometry | |
US6254717B1 (en) | Method and apparatus for monitoring plasma processing operations | |
US6221679B1 (en) | Method and apparatus for monitoring plasma processing operations | |
JP2007515804A (en) | Plasma equipment seasoning method and equipment therefor |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): AE AL AM AT AU AZ BA BB BG BR BY CA CH CN CU CZ DE DK EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MD MG MK MN MW MX NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT UA UG UZ VN YU ZA ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): GH GM KE LS MW SD SL SZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
REG | Reference to national code |
Ref country code: DE Ref legal event code: 8642 |
|
ENP | Entry into the national phase |
Ref country code: JP Ref document number: 2000 544994 Kind code of ref document: A Format of ref document f/p: F |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1020007011793 Country of ref document: KR |
|
WWE | Wipo information: entry into national phase |
Ref document number: 1999918803 Country of ref document: EP |
|
DFPE | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101) | ||
WWP | Wipo information: published in national office |
Ref document number: 1020007011793 Country of ref document: KR |
|
WWP | Wipo information: published in national office |
Ref document number: 1999918803 Country of ref document: EP |
|
AK | Designated states |
Kind code of ref document: C2 Designated state(s): AE AL AM AT AU AZ BA BB BG BR BY CA CH CN CU CZ DE DK EE ES FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MD MG MK MN MW MX NO NZ PL PT RO RU SD SE SG SI SK SL TJ TM TR TT UA UG UZ VN YU ZA ZW |
|
AL | Designated countries for regional patents |
Kind code of ref document: C2 Designated state(s): GH GM KE LS MW SD SL SZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG |
|
COP | Corrected version of pamphlet |
Free format text: PAGES 1/61-61/61, DRAWINGS, REPLACED BY NEW PAGES 1/63-63/63; DUE TO LATE TRANSMITTAL BY THE RECEIVING OFFICE |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: 1999918803 Country of ref document: EP |
|
WWG | Wipo information: grant in national office |
Ref document number: 1020007011793 Country of ref document: KR |