Feed-forward process control in the presence of inconsistent underlayer

Abstract

Methods and systems for generating a first sequence of signal values for a substrate; converting the first sequence of signal values into a first effective thickness trace; applying at least a portion of the first effective thickness trace to a first neural network to generate an adjusted first effective thickness trace; storing the adjusted first effective thickness trace on a networked computer; generating a second sequence of signal values for the scan of the second sensor across the substrate; converting the second sequence of signal values into a second effective thickness trace; applying at least a portion of the second effective thickness trace to a second neural network to generate an adjusted second effective thickness trace; retrieving the adjusted first effective thickness trace; and generating a thickness trace for the second conductive layer to compensate for a contribution of the first conductive layer underlying the second conductive layer.

Description

[0001] FEED-FORWARD PROCESS CONTROL IN THE PRESENCE OF INCONSISTENT UNDERLAYER

[0002] TECHNICAL FIELD

[0003] The present disclosure relates to chemical mechanical polishing, and more specifically to feeding measurements of a conductive layer to downstream machines during chemical mechanical polishing.

[0004] BACKGROUND

[0005] An integrated circuit is typically formed on a substrate by the sequential deposition of conductive, semi conductive, or insulative layers on a silicon wafer. A variety of fabrication processes require planarization of a layer on the substrate. For example, one fabrication step involves depositing a filler layer over a non-planar surface and planarizing the filler layer. For certain applications, the filler layer is planarized until the top surface of a patterned layer is exposed. For example, a metal layer can be deposited on a patterned insulative layer to fill the trenches and holes in the insulative layer. After planarization, the remaining portions of the metal in the trenches and holes of the patterned layer form vias, plugs, and lines to provide conductive paths between thin film circuits on the substrate.

[0006] Chemical mechanical polishing (CMP) is one accepted method of planarization. This planarization method typically requires that the substrate be mounted on a carrier head. The exposed surface of the substrate is typically placed against a rotating polishing pad. The carrier head provides a controllable load on the substrate to push it against the polishing pad. Polishing slurry' with abrasive particles is typically supplied to the surface of the polishing pad.

[0007] One problem in CMP is determining whether the polishing process is complete, i.e., whether a substrate layer has been planarized to a, desired flatness or thickness, or when a desired amount of material has been removed. V ariations in the slurry’ composition, the polishing pad condition, the relative speed between the polishing pad and the substrate, the initial thickness of the substrate layer, and the load on the substrate can cause variations in the material removal rate. These variations cause variations in the time needed to reach the polishing endpoint. Therefore, determining the polishing endpoint merely as a function of polishing time can lead to non-uniformity within a wafer or from wafer to wafer. In some systems, a substrate is monitored in-situ during polishing, e.g., through the polishing pad. One monitoring technique is to induce an eddy current in the conductive layer and detect the change in the eddy current as the conductive layer is removed.

[0008] SUMMARY

[0009] In general, an aspect disclosed herein is a method of semiconductor device fabrication, including forming a first patterned dielectric layer on a substrate: depositing a first conductive layer over the first patterned dielectric layer to fill trenches in the first patterned dielectric layer; chemical mechanical polishing the first conductive layer of the substrate until a first polishing pad until a top surface of the first patterned dielectric layer is exposed; in a first post-polish operation, maintaining the substrate in contact with the first polishing pad while substantially no polishing of the substrate is performed and scanning a first sensor of a first in-situ eddy current monitoring system across the substrate so as to generate a first sequence of signal values for a scan of the first sensor across the substrate; converting the first sequence of signal values for the scan of the first sensor across the substrate into a first effective thickness trace that includes a sequence of first effective thickness values for plurality of different positions across the substrate; applying at least a portion of the first effective thickness trace to a first neural network to generate an adjusted first effective thickness trace; storing the adjusted first effective thickness trace on a networked computer; forming a second patterned dielectric layer on the substrate over the first patterned dielectric layer; depositing a second conductive layer over the second patterned dielectric layer to fill trenches in the second patterned dielectric layer; chemical mechanical polishing the second conductive layer of the substrate with a second polishing pad; during polishing of second conductive layer, scanning a second sensor of a second in-situ eddy current monitoring system across the substrate so as to generate a second sequence of signal values for the scan of the second sensor across the substrate; converting the second sequence of signal values for the scan of the second sensor across the substrate into a second effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate; applying at least a portion of the second effective thickness trace to a second neural network to generate an adjusted second effective thickness trace; retrieving the adjusted first effective thickness trace from the networked computer; generating a thickness trace for the second conductive layer to compensate for a contribution of the first conductive layer underlying the second conductive layer; and at least one of halting polishing or adjusting a polishing parameter that affects the polishing of second conductive layer based on the thickness trace..

[0010] Examples may include one or more of the following features. The first polishing pad and the second polishing pad may be different polishing pads. The first neural network and the second neural network may be different neural networks. The first sensor and the second sensor may be different sensors, and the first in-situ eddy current monitoring system and the second in-situ eddy current monitoring system may be different in-situ eddy current monitoring systems. Storing the adjusted first effective thickness trace may include communicating the adjusted first effective thickness trace to the networked computer and storing the adjusted first effective thickness trace on a computer readable medium of the networked computer. Generating the thickness trace may include subtracting the adjusted first effective thickness trace from the adjusted second effective thickness trace. Maintaining the substrate in contact with the first polishing pad while substantially no polishing of the substrate is performed may include applying no positive pressure to the substrate. Applying zero positive pressure may include venting one or more chamber of a carrier head holding the substrate to atmospheric pressure. The method may include rinsing a polishing liquid from the first polishing pad before the first post-polish operation. The method may include repeatedly scanning the first sensor of the first in-situ eddy current monitoring system across the substrate so as to generate multiple sequences of signal values and converting the multiple sequences of signal values into the first effective thickness trace.

[0011] In general, an aspect disclosed herein is a computer program product for controlling a fabrication system. The computer program product includes forming a first patterned dielectric layer on a substrate; depositing a first conductive layer over the first patterned dielectric layer to fill trenches in the first patterned dielectric layer; chemical mechanical polishing the first conductive layer of the substrate with a first polishing pad until a top surface of the first patterned dielectric layer is exposed; in a first post-polish operation, maintaining the substrate in contact with a polishing pad while substantially no polishing of the substrate is performed and scanning a sensor of an in-situ eddy current monitoring system across the substrate so as to generate a first sequence of signal values for a scan of the sensor across the substrate; converting the first sequence of signal values for the scan of the sensor across the substrate into a first effective thickness trace that includes a sequence of first effective thickness values for plurality of different positions across the substrate; applying at least a portion of the first effective thickness trace to a neural network to generate an adjusted first effective thickness trace; storing the adjusted first effective thickness trace on a networked computer; forming a second paterned dielectric layer on the substrate over the first patterned dielectric layer; depositing a second conductive layer over the second patterned dielectric layer to fill trenches in the second patterned dielectric layer; chemical mechanical polishing the second conductive layer of the substrate with a polishing pad; during polishing of second conductive layer, scanning a sensor of an in-situ eddy current monitoring system across the substrate so as to generate a second sequence of signal values for the scan of the sensor across the substrate; converting the second sequence of signal values for the scan of the sensor across the substrate into a second effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate; applying at least a portion of the second effective thickness trace to a neural network to generate an adjusted second effective thickness trace; retrieving the adjusted first effective thickness trace from the networked computer; generating a thickness trace for the second conductive layer to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values, including subtracting the adjusted first effective thickness trace from the adjusted second effective thickness trace; and at least one of halting polishing or adjusting a polishing parameter that affects the polishing of second conductive layer based on the thickness trace.

[0012] In general, an aspect disclosed herein is a method of polishing a substrate, polishing a first conductive layer of the substrate with a polishing pad; in a post-polish operation, maintaining the substrate in contact with the first polishing pad while substantially no polishing of the substrate is performed and scanning a sensor of an in-situ eddy current monitoring system across the substrate so as to generate a sequence of signal values; converting the sequence of signal values into an effective thickness trace that includes a sequence of effective thickness values for plurality' of different positions across the substrate; applying at least a portion of the effective thickness trace to a neural network to generate an adjusted effective thickness trace; retrieving an adjusted second effective thickness trace from a networked computer, the second effective thickness trace representative of a second sequence of adjusted thickness values for a scan of the sensor across a second conductive layer beneath the first conductive layer on the substrate; generating a thickness trace for the first conductive layer to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values, including subtracting the adjusted second effective thickness trace from the adjusted first effective thickness trace; and at least one of halting polishing or adjusting a polishing parameter that affects the polishing of first conductive layer based on the thickness trace. Examples may include one or more of the following features. The first polishing pad and the second polishing pad may be different polishing pads. The first neural network and the second neural network may be different neural networks. The first sensor and the second sensor may be different sensors, and the first in-situ eddy current monitoring system and the second in-situ eddy current monitoring system may be different in-situ eddy current monitoring systems. Storing the adjusted first effective thickness trace may include communicating the adjusted first effective thickness trace to the networked computer and storing the adjusted first effective thickness trace on a computer readable medium of the networked computer.

[0013] In general, an aspect disclosed herein is a chemical mechanical polishing system. The chemical mechanical polishing system includes a networked computer; and multiple chemical mechanical polishing stations in communication with the networked computer, each chemical mechanical polishing station may include: a platen supporting a polishing pad; a carrier head to hold a surface of a substrate against the polishing pad; a motor to generate relative motion between the platen and the carrier head so as to polish an overly ing layer on the substrate; an in-situ eddy current monitoring system may include a sensor arranged within a recess of the platen; And a controller configured to cause the respective chemical mechanical polishing station to: polish a first conductive layer of the substrate with the polishing pad; in a postpolish operation, maintain the substrate in contact with the polishing pad while substantially no polishing of the substrate is performed and scan a sensor of the in-situ eddy current monitoring system across the substrate so as to generate a sequence of signal values; convert the sequence of signal values into an effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate; apply at least a portion of the effective thickness trace to a neural network to generate an adjusted effective thickness trace; retrieve an adjusted second effective thickness trace from the networked computer, the second effective thickness trace representative of a second sequence of adjusted thickness values for a scan of the sensor across a second conductive layer beneath the first conductive layer on the substrate; generate a thickness trace for the first conductive layer to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values, including subtracting the adjusted second effective thickness trace from the adjusted first effective thickness trace; and at least one of halt polishing or adjust a polishing parameter that affects the polishing of first conductive layer based on the thickness trace. Examples may include one or more of the following features. The method where the second adjusted effective thickness trace may be generated by a different chemical mechanical polishing station than the chemical mechanical polishing station polishing the first conductive layer. The second adjusted effective thickness trace may be generated by a different neural network than the first adjusted effective thickness trace.

[0014] Implementations may include one or more of the following advantages. In calculation of the thickness of the layer being polished based on a monitored signal, the technique can compensate for contributions to the signal by conductive underlay ers, e.g., a doped substrate or metal layers, even when the underlayer contribution is inconsistent on a wafer-to-wafer basis. Thus, the thickness of the layer being polished can be calculated with higher accuracy or reliability. The calculated thickness can be used for determining control parameters during a polishing process and / or determining an endpoint for the polishing process. Reliability' of the control parameter determination and endpoint detection can be improved, wafer underpolish can be avoided, and wafer-to-wafer non-uniformity (WTWNU) and within-wafer nonuniformity (WIWNU) can be reduced. The thickness of the conductive layer can be determined following deposition of the layer and used as a starting thickness profile for the same substrate in downstream polishing processes. Using the starting thickness profile of the same substrate and distributing the profile to any downstream system provides single-wafer control according to desired polishing final thicknesses.

[0015] The details of one or more implementations are set forth in the accompanying drawings and the description below. Other aspects, features, and advantages will be apparent from the description and drawings, and from the claims.

[0016] BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 illustrates a schematic cross-sectional view of an example of a polishing station including an electromagnetic induction monitoring system.

[0017] FIG. 2 illustrates a schematic top view of an example chemical mechanical polishing station showing a path of a sensor scan across a substrate.

[0018] FIGS. 3A-3C are schematic cross-sectional views of a substrate illustrating a polishing process.

[0019] FIG. 4 is a schematic cross-sectional view illustrating an example magnetic field generated by an electromagnetic induction sensor.

[0020] FIG. 5 is a schematic top view of a substrate being scanned by a sensor head of a polishing apparatus.

[0021] FIG. 6 is a schematic graph of measured signals obtained while monitoring locations on a substrate.

[0022] FIG. 7 is an example neural network.

[0023] FIG. 8 is a schematic illustration of the process of compensating for substrate doping while also applying edge reconstruction.

[0024] FIG. 9 illustrates a graph of an example signal from an eddy current sensor as a function of conductive layer thickness.

[0025] FIG. 10 is a schematic illustration of generation of an adjusted trace for a substrate being polished.

[0026] FIGS. 11A-11C illustrate a flow chart of a method for performing endpoint and profile control in chemical mechanical polishing.

[0027] FIG. 12 is a schematic plan view' of an example of networked polishing apparatuses.

[0028] DETAILED DESCRIPTION

[0029] One monitoring technique for a polishing operation is to induce eddy currents in a conductive layer on a substrate, e.g., using an alternating current (AC) drive signal. The induced eddy currents can be measured by an eddy current sensor in-situ during polishing to generate a signal. Assuming the outermost layer undergoing polishing is a conductive layer, then the signal from the sensor should be dependent on the thickness of the conductive layer. Based on the monitoring, control parameters for the polishing operation, such as applied pressure, can be adjusted in-situ to control the polishing rate. In addition, the polishing operation can terminate based on an indication that the monitored thickness has reached a desired endpoint thickness.

[0030] In practice, the magnetic field generated by the eddy current sensor does not stop within the outermost conductive layer but can extend into underlying conductive layers or the underlying substrate. As a result, the signal generated by the eddy current sensor can depend on the conductivity of the semiconductor wafer and the presence of the underlying conductive layers. If the semiconductor wafer is not doped, e.g., as typically used in “blank” wafers used for system calibration and basic substrate wafers, the electrical resistance of the wafer can be sufficiently high that the presence of the wafer does not have detectable influence on the eddy current signal.

[0031] However, for actual device fabrication the wafers will typically be doped, e.g., highly doped, for various purposes. In this situation, the signal generated by the eddy current sensor can have significant contribution from the substrate, depending on the conductivity of the semiconductor wafer. As such, thickness measurement based on signals captured by the eddy current sensor can be inaccurate.

[0032] If the contribution of the substrate due to doping is known, then calculation of the thickness of the layer being polished can take into account the contribution to the signal from the semiconductor substrate. However, often doping is poorly controlled during semiconductor device fabrication, both across a given wafer and, on a wafer, -to-wafer basis within a given production lot.

[0033] The magnetic field generated by the eddy current sensor does not stop within the conductive layer being polished and can extend into underlying layers. Without being limited to any particular theory, the skin depth in magnetic permeable materials for the electromagnetic frequency employed in eddy current sensor can be larger than thickness of any one conductive layer and the underlying conductive and dielectric layers. As a result, the signal generated by the eddy current sensor can depend on the conductivity of the layers beneath the conductive layer being polished.

[0034] If one assumes that the underlying layers do not vary from wafer-to-wafer or product-to-product, an eddy current measurement of the underlying layer can be performed prior to deposition of conductive layers to generate traces of the thickness of the combined layers. These traces of the underlying layer can then be stored as a reference to be subtracted from the conductive layer traces in downstream wafer production steps.

[0035] However, these assumptions do not account for wafer-to-wafer variation in the thickness of the underlying conductive layers or wafer-to-wafer variation in substrate doping, either of which can be substantial. One technique to overcome the wafer-to-wafer variation is to feed-forward eddy current thickness profiles to downstream wafer production systems. Feeding the thickness profiles forward enables individualized wafer profile control based on the unique underlying layer thickness profile and the conductive layer thickness profile.

[0036] The present disclosure concerns a technique to perform endpoint and polishing profile control when the underlayer, e.g., the doped semiconductor wafer or underlying deposited layers, has an inconsistent wafer-to-wafer contribution to the signal from the eddy current sensor.

[0037] An effective thickness profile of a first deposited conductive layer is calculated based on a series of sequence values received from the in-situ monitoring system. The effective thickness can then be adjusted by using a neural network which corrects for measurement effects such as edge correction. The adjusted effective thickness profile is attributed to the specific wafer having the first conductive layer. The adjusted effective thickness profile is stored and can be retrieved by additional wafer production systems to which the substrate is distributed. Thus, as the substrate continues to be polished on the additional wafer production systems, the adjusted effective thickness profile is wafer-specific and available to each system. Each system may perform a production step on the wafer enables wafer-specific polishing thus increasing within- wafer and wafer-to- wafer uniformity.

[0038] A subsequent deposited layer can be scanned to generate a second effective thickness profile representing the thickness of the subsequent deposited film. The adjusted thickness profile associated with the substrate can be subtracted from the second effective thickness profile to provide actual thickness traces for the second deposited conductive layer, e.g., a contribution of the underlayers in the form of an effective thickness.

[0039] FIG. 1 illustrates an example of a polishing station 20 of a chemical mechanical polishing system. The polishing station 20 includes a rotatable disk-shaped platen 24 on which a polishing pad 30 is situated. The platen 24 is operable to rotate about an axis 25. For example, a motor 22 can turn a drive shaft 28 to rotate the platen 24. The polishing pad 30 can have two layers, including an outer polishing layer 32 and a softer backing layer 34.

[0040] The polishing station 20 can include a supply port or a combined supply-rinse arm 39 to dispense a polishing liquid 38, such as an abrasive slurry, onto the polishing pad 30. The polishing station 20 can include a pad conditioner apparatus with a conditioning disk to maintain the surface roughness of the polishing pad 30.

[0041] A carrier head 70 is operable to hold a substrate 10 against the polishing pad 30. The carrier head 70 is suspended from a support structure 72, e.g., a carousel or a track, and is connected by a drive shaft 74 to a carrier head rotation motor 76 so that the carrier head 70 can rotate about an axis 71. Optionally, the carrier head 70 can oscillate laterally, e.g., on sliders on the carousel, by movement along the support structure 72, or by rotational oscillation of the carousel itself.

[0042] The carrier head 70 can include a flexible membrane 80 having a substrate mounting surface to contact the back side of the substrate 10. The membrane 80 can form a plurality of pressurizable chambers 82 to apply different pressures to different zones, e.g., different radial zones, on the substrate 10. The carrier head can also include a retaining ring 84 to hold the substrate below- the membrane 80.

[0043] In operation, the platen 24 is rotated about its central axis 25. The carrier head 70 is rotated about the central axis 71 and translated laterally across the top surface of the polishing pad 30. In examples having multiple carrier heads, each carrier head 70 can have independent control of its polishing parameters, for example each carrier head 70 can independently control the pressure applied to each respective substrate 10.

[0044] The polishing station 20 includes an in-situ eddy current monitoring system 100 which can be coupled to or be considered to include a controller 90. In operation, the polishing station 20 can use the in-situ eddy current monitoring system 100 to determine when the conductive layer has reached a target thickness, e.g., a target thickness for a metal layer overlying a dielectric layer, and then halt polishing. Alternatively or in addition, the polishing system can use the in-situ eddy current monitoring system 100 to determine differences in thickness of the conductive layer 16 across the substrate 10 and use this information to adjust the pressure in one or more chambers 82 in the carrier head 70 during polishing in order to reduce polishing non-unifonnity. During polishing, the measurements from the sensor 102 can be displayed on an output device to permit an operator of the polishing station to visually monitor the progress of the polishing operation, although this is not required.

[0045] The sensor 102 of the in-situ monitoring system 100 can be installed in a recess 26 in the platen 24. The sensor 102 can include a magnetic core 104 positioned at least partially in the recess 26, and at least one coil 106 wound around a portion of the core 104. Drive and sense circuitry 108 is electrically connected to the coil 106. The drive and sense circuitry 108 generates a signal that can be sent to the controller 90. The circuitry 108 can include a capacitor connected in parallel with the coil 106. Together the coil 106 and the capacitor can form an LC resonant tank. Although illustrated as outside the platen 24, some or all of the drive and sense circuitry 108 can be installed in the platen 24. A rotary coupler 29 can be used to electrically connect components in the rotatable platen 24, e.g., a sensor 102 of the eddy current monitoring system 100, to components outside the platen, e.g., drive and sense circuitry 108 or the controller 90. Alternatively, the sensor 102 can be in wireless communication with the controller 90.

[0046] The drive and sense circuitry 108 can include a marginal oscillator coupled to a combined drive / sense coil 106, and the output signal can be a current required to maintain the peak-to-peak amplitude of the sinusoidal oscillation at a constant value, e.g., as described in U. S, Patent No. 7,112,960. Other configurations are possible for the drive and sense circuitry’ 108. For example, separate drive and sense coils could be wound around the core. The drive and sense circuitry 108 can apply current at a fixed frequency', and the signal from the drive and sense circuitry 108 can be the phase shift of the current in the sense coil relative to the dri ve coil, or an amplitude of the sensed current, e.g., as described in U. S. Patent No.

[0047] 6,975,107.

[0048] In some implementations, the polishing station 20 includes a temperature sensor 64 to monitor a temperature in the polishing station 20 or a component of the polishing station 20.

[0049] A controller 90, e.g., a general purpose programmable digital computer, receives the signals from sensor 102 of the in-situ monitoring system 100. Since the sensor 102 sweeps beneath the substrate 10 with each rotation of the platen 24, information on the depth of the conductive layer, e.g., the bulk layer or conductive material in the trenches, is accumulated in-situ (once per platen rotation). The controller 90 can be programmed to sample signals from the in-situ monitoring system 100 when the substrate 10 generally overlies the sensor 102.

[0050] The controller 90 can be in electronic communication with one or more polishing systems or one or more polishing stations of the same polishing system. The controller 90 can communicate data between the one or more polishing systems or one or more polishing stations, such as sequences of values generated by the sensor 102, or estimated thickness profiles. In examples having more than one polishing station 20, the polishing system can include a first polishing station, such as polishing station 20, with a first electromagnetic induction monitoring system 100 and a second polishing station with a second electromagnetic induction current monitoring system.

[0051] In some implementations, a polishing system includes additional polishing stations. For example, a polishing system can include two or three polishing stations 20. For example, in operation, bulk polishing of the conductive layer on the substrate can be performed at the first polishing station, and polishing can be halted when a target thickness of the conductive layer remains on the substrate. The substrate is then transferred to the second polishing station, and the substrate can be polished until an underlying layer is exposed, e.g., a patterned dielectric layer.

[0052] Referring to FIG. 2, as the platen 24 rotates, the sensor 102 sweeps along a path 120 that passes below the substrate 10. By sampling the signal from the circuitry 108 at a particular frequency, the circuitry 108 generates measurements at a sequence of sampling zones 94 across the substrate 10. For each sweep, measurements at one or more of the sampling zones 94 can be selected or combined. Thus, over multiple sweeps, the selected or combined measurements provide a time-varying sequence of values.

[0053] The polishing station 20 can also include a position sensor 96, such as an optical interrupter, to sense when the sensor 102 is underneath the substrate 10 and when the sensor location opposite the carrier head 70. A flag 98 can be attached to the periphery of the platen 24. The point of attachment and length of the flag 98 is selected so that it can signal the position sensor 96 when the sensor 102 sweeps underneath the substrate 10. Alternately or in addition, the polishing station 20 can include an encoder to determine the angular position of the platen 24.

[0054] In addition, the controller 90 can be programmed to calculate the radial position of each measurement, and to sort the measurements into radial ranges, e.g., as discussed in U. S. Pat. No. 6,399,501.

[0055] Since the sensor 102 sweeps underneath the substrate 10 with each rotation of the platen 24, information on the conductive layer thickness is being accumulated in-situ and on a continuous real-time basis for each of multiple different radial zones on the substrate.

[0056] Referring to FIG. 3 A, the polishing station 20 can be used to polish a substrate 10 that includes a conductive layer 16 of a conductive material overlying and / or inlaid in a patterned dielectric layer 14. For example, the substrate 10 can include a layer of conductive layer 16, e.g., a metal, e.g., copper, aluminum, cobalt or titanium, that overlies and fills trenches in a dielectric layer 14, e.g., silicon oxide or a high-k dielectric. Optionally a barrier layer 18, e.g., tantalum or tantalum nitride, can line the trenches and separate the conductive layer 16 from the dielectric layer 14. The conductive layer 16 in the trenches can provide vias, pads and / or interconnects in a completed integrated circuit. Although the dielectric layer 14 is illustrated as deposited directly on a semiconductor wafer 12, one or more other layers can be interposed between the dielectric layer 14 and the wafer 12.

[0057] The semiconductor wafer 12 can be a silicon wafer, e.g., single crystalline silicon, although other semiconductor materials are possible, e.g., gallium arsenide or gallium nitride. In addition, the semiconductor wafer 12. can be doped, e.g., with p-type or n-type doping. The doping can be uniform laterally across the wafer, or the wafer can be selectively doped, e.g., as appropriate for fabrication of transistors in integrated circuits using the semiconductor wafer.

[0058] Initially, the conductive layer 16 overlies the entire dielectric layer 14 and the barrier layer 18 (see FIG. 3 A). As polishing progresses, the bulk of the conductive layer 16 is removed, exposing the barrier layer 18 (see FIG. 3B). Continued polishing then exposes the patterned top surface of the dielectric layer 14 (see FIG. 3C). Additional polishing can then be used to control the depth of the trenches that contain the conductive layer 16. For simplicity of the drawing, a barrier layer lining the sides and bottom of the trench is not illustrated in FIG. 3C.

[0059] The in-situ eddy current monitoring system 100 is configured to generate a signal that depends on a thickness of the conductive material, e.g., the metal, of the layer 16 being polished. Referring to FIG. 4, a substrate 10’ having multiple conductive layers is shown being polished on the polishing layer 32. The drive and sense circuitry 108 applies an AC current to the coil 106 which generates a magnetic field 110 between two poles 112a and 112b of the core 104. In operation, a portion of the magnetic field 110 extends into the substrate 10’ when the substrate 10’ intermittently overlies the sensor 102.

[0060] The magnetic field 110 can create an eddy-current in the conductive layer 16 thus modifying the effective impedance of an LC circuit of the monitoring system 100. The effective impedance of the LC circuit changes as the thickness of the conductive layer 16 changes. This enables monitoring of the thickness of the conductive layer 16, and changes therein.

[0061] The magnetic field 110 can also penetrate into various different layers other than the material 16 being polished. For instance, conductive layer 19 is “below” the conductive layer 16 and separated from the conductive layer 16 by a dielectric layer 15. The magnetic field can also extend into the semiconductor wafer 12. As such, the effective impedance of the LC circuit, and thus the signal from the drive and sense circuitry 108, can also depend on the conductivity of the underlying conductive layer 19 and on the doping and resultant conductivity’ of the semiconductor wafer 12. Patterning of the layers is not shown in FIG. 4 for ease of illustration.

[0062] Some variation in the signal intensity from the sensor 102 can be caused by the measurement region of the sensor 102 overlapping the substrate edge, rather than an intrinsic variation in the thickness or conductivity’ of the layer being monitored. Consequently, this distortion in the signal can cause errors in the calculating of a thickness value for the substrate, particularly near the substrate edge. To address this problem, the controller 90 can feed each effective thickness trace into a neural network that is configured to generate an adjusted thickness trace that compensates for compensate such signal distortions. Such a process is described in U. S. Patent Publication No. 2021 -0379723. The resulting thickness measurements can be sorted into the radial ranges.

[0063] Referring to FIGS. 5 and 6, the signal 220 includes a first portion 222 that corresponds to locations in an edge region 203 of the substrate 10 when the sensor 102 crosses a leading edge of the substrate 10, a second portion 224 that corresponds to locations in a central region 201 of the substrate 10, and a third portion 226 that corresponds to locations in edge region 203 when the sensor 102 crosses a trailing edge of the substrate 10. The signal can also include portions 228 that correspond to off-substrate measurements, i.e., signals generated when the sensor head scans areas beyond the edge 204 of the substrate 10 in FIG. 6.

[0064] The edge region 203 can correspond to a portion of the substrate where measurement spots 211 of the sensor 102 overlap the substrate edge 204. The central region 201 can include an annular anchor region 202 that is adjacent the edge region 203, and an inner region 205 that is surrounded by the anchor region 202. The sensor 102 may scan these regions on its path 210 and generate a sequence of measurements that correspond to a sequence of locations along the path 210.

[0065] In the first portion 222, the signal intensity ramps up from an initial intensity (typically the signal resulting when no substrate and no carrier head is present) to a higher intensity. This is caused by the transition of the monitoring location from initially only slightly overlapping the substrate 10 at the edge 204 (generating the initial lower values) to the monitoring location nearly entirely overlapping the substrate 10 (generating the higher values). Similarly, in the third portion 226, the signal intensity ramps down when the monitoring location transitions to the edge 204 of the substrate.

[0066] Although the second portion 224 is illustrated as flat, this is for simplicity, and a real signal in the second portion 224 would likely include fluctuations due both to noise and to variations in the layer thickness. The second portion 2.34 corresponds to the monitoring location scanning the central region 201. The second portion 224 includes two sub-portions 230 and 232 that are caused by the monitoring location scanning the anchor region 202 of the central region 201, and sub-portion 234 that is caused by the monitoring location scanning the inner region 205 of the central region 2.01.

[0067] As noted above, the variation in the signal intensity in the portions 222, 226 is caused in part by measurement region of the sensor 102 overlapping the substrate edge, rather than an intrinsic variation in the thickness or conductivity of the layer being monitored.

[0068] Consequently, this distortion in the signal 220 can cause errors in the calculating of a characterizing value for the substrate, e.g., the thickness of the layer, near the substrate edge. To address this problem, the controller 90 can include a neural network, e.g., neural network 300 of FIG. 8, to generate an adjusted signal corresponding to one or more locations of the substrate 10 based on the measured signals corresponding to those locations. Referring now to FIG. 7, the neural network 300 is configured to, when trained appropriately, generate adjusted signals that reduce and / or remove the distortion of computed signal values near the substrate edge. ’The neural network 300 receives a group of inputs 304 and processes the inputs 304 through one or more neural network layers to generate a group of outputs 350. The layers of the neural network 300 include an input layer 310, an output layer 330, and one or more hidden layers 320.

[0069] Each layer of the neural network 300 includes one or more neural network nodes. Each neural network node in a neural network layer receives one or more node input values (from the inputs 304 to the neural network 300 or from the output of one or more nodes of a preceding neural network layer), processes the node input values in accordance with one or more parameter values to generate an activation value, and optionally applies a non-linear transformation function (e.g., a sigmoid or tanh function) to the activation value to generate an output for the neural network node.

[0070] Each node in the input layer 310 receives as a node input value one of the inputs 304 to the neural network 300.

[0071] The inputs 304 to the neural network include measured signal values from the in-situ monitoring system 100 for multiple different spots 211 on the substrate 10, such as a first measured signal value 301, a second measured signal value 302, through an nth measured signal value 303. The measured signal values can be individual values of the sequence of values in the signal 220.

[0072] In general, the multiple different locations include locations in the edge region 2.03 and, optionally, the anchor region 202 of the substrate 10. In some implementations, the multiple different locations are only in the edge region 203 and the anchor region 202, In other implementations, the multiple different locations span all regions of the substrate.

[0073] These measured signal values are received at signal input nodes 344. Optionally, the input nodes 304 of the neural network 300 can also include one or more state input nodes 316 that receive one or more process state signals 304, e.g., a measure of wear of the pad 30 of the polishing station 20.

[0074] The nodes of the hidden layers 320 and output layer 330 are illustrated as receiving inputs from every node of a preceding layer. This is the case in a fully connected, feedforward neural network. However, the neural network 300 may be a non-fully connected feedforward neural network or a non-feedforward neural network. Moreover, the neural network 300 may include at least one of one or more fully connected, feedforward layers; one or more non-fully connected feedforward layers; and one or more non-feedforward layers. The neural network generates a group of adjusted signal values, e.g., outputs 350, at the nodes of the output layer 330, i.e., “output nodes” 360. In some implementations, there is an output node 360 for each measured signal from the in-si tu monitoring system that is fed to the neural network 300. In this case, the number of output nodes 360 can correspond to the number of signal input nodes 304 of the input layer 310.

[0075] For example, the number of signal input nodes 344 can equal the number of measurements in the edge region 203 and the anchor region 202, and there can be an equal number of output nodes 360. Thus, each output node 360 generates an adjusted signal that corresponds to a respective measured signal supplied as an input to a signal input node 344, e.g., the first adjusted signal 351 for the first measured signal 301, the second adjusted signal 352 for the second measured signal 302, and the nth adjusted signal 353 for the nth measured signal 303.

[0076] In some implementations, the number of output nodes 360 is smaller than the number of input nodes 304. In some implementations, the number of output nodes 360 is smaller than the number of signal input nodes 344. For example, the number of signal input nodes 344 can equal the number of measurements in the edge region 203, or equal to the number of measurements in the edge region 203 and anchor region 202, Again, each output node 360 of the output layer 330 generates an adjusted signal that corresponds to a respective measured signal supplied as a signal input node 304, e.g., the first adjusted signal 351 for the first measured signal 301, but only for the signal input nodes 354 that receive signals from the edge region 203.

[0077] The polishing station 20 can use the neural network 300 to generate adjusted signals. The adjusted signals can then be used to determine a thickness for each location in a first group of locations of a substrate, e.g., the locations in the edge region (and possibly the anchor region). For example, referring back to FIG. 5, the adjusted signal values for the edge region can provide an adjusted portion 236 of the signal 220.

[0078] In some implementations, for an adjusted signal value that corresponds to a given measurement location, the neural network 500 can be configured such that only input signal values from measurement locations within a predetermined distance of that given location are used in determining the adjusted signal value.

[0079] To train the neural network, the sensor 102 of the in-situ monitoring system 100 can be used to generate a profiles of reference substrates. In addition, ground truth measures of thickness of the reference substrates can be obtained; these measurements can be performed for locations that are to be processed by the neural network. The system can generate the ground truth measures of thickness using an electrical impedance measuring method, such as a four-points probe method. The signal values from the reference substrate are applied to the inputs 304 while the ground truth measurements are applied to the output nodes 360 and the system is run in a training mode, e.g., gradient descent with backpropagation.

[0080] The reference substrates can include wafers on which a uniform thickness of a conductive material is deposited. The amount of conductive material can be selected to simulate the presence of a doped wafer.

[0081] The reference substrates can also include sample device substrates at an equivalent stage of processing as the device substrate for which the in-situ monitoring system is to be used for controlling of polishing, e.g., substrates with layers having different edge profiles.

[0082] The signal generated by the eddy current sensor 102 also includes the contribution from underlying conductive layers and the doped semiconductor wafer. If not handled properly, attempts to compensate for the contribution to the signal from the doped wafer can introduce additional errors, e.g., at the substrate edge when edge reconstruction techniques are utilized. The eddy currents excited in the doped semiconductor wafer 12, the conductive layer 16, and the conductive layer 19 are independent and separated by insulating dielectric layer 15. Thus, the power dissipations of the silicon wafer 12 and the separated conductive film layers 16 and 19 are additive and should generally follow the superposition principle of linear systems, e.g., the signal contributions of the layers and wafer are additive.

[0083] In many situations the doping of the semiconductor wafer 12 is not tightly controlled, and thus can be highly variable both across a wafer, e.g., within-wafer, and on a wafer-to-wafer basis, even for wafers within a given production lot.

[0084] Similarly, in many situations the thickness of the underlying conductive layers, e.g., the conductive layer 19, is not tightly controlled, and thus can be variable both across a wafer, e.g., within-wafer, and on a wafer-to-wafer basis, even for wafers within a given production lot.

[0085] Hypothetically, a radial profile of the starting thickness of a conductive layer to be polished could be known with high precision and reliability by performing a measurement at an in-line or stand-alone metrology station. For example, the conductive layer starting thickness profile of the substrate could be measured with a four-point probe at an in-line or stand-alone station and fed-forward to the polishing station. The problem with such approaches is that such accurate in-line or stand-alone metrology systems are typically slow, particularly if many points are being measured to provide a radial profile. As such. measurement of each substrate by an in-line or stand-along metrology systems may not be practical from a cost or throughput perspective.

[0086] Referring to FIG. 8, a reference trace 420 across wafer is generated. This reference trace 420 is generated prior to polishing of the substrate or generated prior to polishing of a subsequent layer. The wafer has the same doping profile as the wafers to be used in the device substrate to be polished. In some implementations, the reference trace 420 is generated by scanning a sample wafer, with the sensor 102 of the in-situ monitoring system 100. For example, the reference trace 420 could be generated by fab operator. Alternatively, the system manufacturer could generate reference traces for wafers having a variety of different dopings (e.g., concentrations and / or doping materials), and these traces can be stored in a library1. The operator can then select one of the references traces from the library, e.g., from a drop-down menu or similar user interface, that corresponds most closely to the doping of the wafer in the device substrates to be polished.

[0087] The raw signal values in the reference trace 420 from the sensor 102 can be converted to effective thickness values (represented by effective thickness trace 420') using a correlation curve.

[0088] FIG. 9 shows a correlation curve 510, for a given resistivity, between the thickness of a conductive layer of the given resistivity and the signal from the electromagnetic induction monitoring system 100. DSTART represents the initial thickness of the conductive layer, SSTART is the desired signal value corresponding to the initial thickness DSTART; DHNAL represents the final thickness of the conductive layer, and SPINAL IS the desired signal value correspond to the final thickness; and K is a constant representing a value of the signal for zero conductive layer thickness.

[0089] The correlation curve 510 can be represented in the controller 90 by a function, e.g., a polynomial function, e.g., a second order function, a third order function, or a higher order function. The correlation between the signal X(x) and the thickness D(x) can be represented by the equation:

[0090] X(x) = W₁D(x)2+ W₂D(x) + Wa (Equation 1) where W₁, W₂, and W₃ are real number coefficients. Thus, the controller can store the values of the coefficients of the function, e.g., W₁, W₂, and W₃, as well as the resistivity po for which the correlation curve 510 applies. In addition, the relationship could be represented with a linear function, a Bezier curve, or a non-polynomial function, e.g., exponential or logarithmic. The relationship curve 510 can be used to convert the signal values in the raw signal of the reference trace 420 from a reference wafer to “equivalent” thickness measurements. That is, although there is no conductive layer on top of the reference wafer, the measurement can be represented as a thickness values. These are “equivalent” thickness values because each is a thickness of an equivalent conductive layer on an undoped wafer that would generate the same signal as the reference wafer.

[0091] Returning to FIG. 8, the reference trace 420' is then processed by the neural network as if it were a normal signal to perform the edge reconstruction algorithm on the reference trace. This generates an adjusted effective thickness trace 450 with a portion having adjusted signal values 430.

[0092] In some implementations, the conversion to thickness is performed in advance, and what is stored in the library (and selected by the operator) is the reference trace 420' with thickness values. In some implementations, the thickness conversion and edge reconstruction are performed in advance, and what is stored in the library (and selected by the operator) is an adjusted effective thickness trace 450.

[0093] During the polishing operation, the substrate 10 is monitored by the in-situ monitoring system, and the measured trace 220 for the substrate 10 is generated for each sweep of the sensor 102 across the substrate 10. This measured trace 220 can also be termed a “total” trace or signal, as it includes contributions from the conductive layer being polished, conductive layers beneath the conductive layer being polished, and the underlying wafer.

[0094] The correlation curve 510 (see FIG. 9) can be used to convert the signal values in the signal 220 from the substrate being polished to thickness measurements (represented by measured trace 220').

[0095] Each measured trace 220’ is processed by the neural network, as discussed above, to generate an adjusted measured trace 250 with a portion 2.36 having adjusted values. In some implementations, the conversion from raw signal to thickness can be performed for both the reference wafer and the substrate being polished after the edge reconstruction is performed.

[0096] To compensate for the wafer doping and contributions from underlying conductive layers, the controller 90 can generate an adjusted trace 480. Generating the adjusted trace includes subtracting the adjusted effective thickness trace 450 from the adjusted measured trace 250. Assuming the adjusted effective thickness trace 450 is represented by S(x), and the adjusted measured trace 250 is represented by T(x), with x being a radial position, then T(x)-S(x) provides an apparent thickness trace. For some configurations of the sensor 102, the contribution from the doped wafer and the substrate to the trace are not a simple superposition. Rather, the apparent thickness of the conductive layer can be somewhat smaller than the actual thickness. This problem can become more pronounced at higher driving frequencies.

[0097] However, any particular sensor configuration (e.g., driving frequency, shape and dimensions of core, location and number of windings of coil, etc.) can have a generally linear relationship between the actual thickness and the apparent thickness. This relationship is illustrated in FIG. 10. A function 520 that relates the apparent thickness to the actual thickness can be expressed as a linear function with a slope of k and a y-intercept (where the thickness should be zero) of b. These values k and b can be determined empirically by testing, and will vary between different sensor configurations. The value of k tends to be less than or equal to 1, e.g., a value from 0.7 to 1.

[0098] Thus, an adjusted film thickness profile, A(x), of the conductive layer on the substrate can be calculated according to A(x) = (T(x)-S(x)-b) / k.

[0099] Endpoint can be called when the adjusted thickness value A(x) reaches a target thickness value DTARGET. Similarly, the adjusted thickness values A'(x) can be used for control of the polishing parameters, e.g., for calculation of polishing pressures to reduce nonuniformity.

[0100] In some cases, the relationship between the apparent thickness and the actual thickness for a particular sensor configuration may not be linear. In such a case, a more complex equation, e.g., a polynomial, may be used to calculate the actual thickness.

[0101] In some implementations, the raw signal is normalized before conversion to thickness values. This technique is applicable to both the reference trace 420 and the substrate trace 220. For example, a calibrated signal X'(x) can be generated using a wafer having a conductive layer of known thickness and conductivity. The calibrated signal X'(x) is then used for the correlation curve, e.g., in place of X(x) in Equation 1 above, to determine the thickness values.

[0102] In addition, during conversion of the raw signal values to thickness values, the resistivity of the layer can be taken into account. For example, the thickness value calculated using the correlation curve, e.g., Equation 1 above, can be adjusted based on the resistivity of the layer to provide a corrected thickness value. This technique can be used for both the reference trace 420 and the substrate trace 220. The edge reconstruction algorithm can be applied to the corrected thickness values instead of initial thickness values D(x). In addition to the substrate-to-substrate variations in resistivity, changes in temperature of the layer can result in a change in the resistance of the conductive layer. For example, the conductive layer may become hotter as polishing progresses, and thus more conductive (lower resistivity). In particular, the controller carrying out the process can also calculate a resistivity profile of the conductive layer at the real time temperature T(t). The real time temperature T(t) can be determined from the temperature sensor 64. ’The an adjusted resistivity pr is then used in place of the resistivity px, e.g., in Equation 3 above (or in calculation of the gain and offset in Equation 2).

[0103] FIGS. 11A-11C illustrate a flow chart, showing a process 1100 for fabricating a substrate and detecting an endpoint or adjusting a polishing parameter during the polishing operation using a fed-forward thickness profile received from an upstream station.

[0104] A first patterned dielectric layer is formed on a substrate (step 1102). In some examples, forming the first patterned dielectric layer includes depositing the first patterned dielectric layer on the substrate and polishing the first patterned dielectric layer. Polishing the first patterned dielectric layer may planarize the first patterned dielectric layer and remove asperities before depositing conductive layers.

[0105] A first conductive layer is deposited over the first patterned dielectric layer to fill trenches in the first patterned dielectric layer (step 1104). The layer of a conductive material overlies the patterned dielectric layer. Optionally a barrier layer is deposited between the conductive layer from the dielectric layer.

[0106] The first conductive layer of the substrate is polished with a first polishing pad until a top surface of the first patterned dielectric layer is exposed (step 1106). The polished conductive layer provides vias, pads and / or interconnects in an integrated circuit. The polishing of the first conductive layer is monitored using an m-situ monitoring system, such as an in-situ eddy current monitoring system. The in-situ monitoring system generates a sequence of signal values representative of a thickness of a conductive layer deposited on the substrate.

[0107] In a first post-polish operation, the substrate is not removed from the polishing pad. Instead, the substrate is maintained in contact with the polishing pad. In some examples, the reference trace is acquired in which a sensor of an in-situ eddy current monitoring system is scanned across the substrate, e.g., by rotating the platen, so as to generate a first sequence of signal values for a scan of the sensor across the substrate (step 1108). The reference trace can be acquired while substantially no polishing of the substrate is performed. To perform substantially no polishing of the substrate, the carrier head pressure can be reduced to zero, or near zero, pressure, rotation and oscillation of the carrier head can be halted, polishing liquid, e.g., abrasive slurry, can be washed from the polishing surface, or a combination of two or more these techniques can be used. Typically the polishing liquid would be removed from the polishing pad by a high-pressure rinse, and the chambers in the carrier head would be vented to atmosphere so that there is no positive pressure applied to the substrate.

[0108] The first sequence of signal values for the scan of the sensor across the substrate is converted into a first effective thickness trace that includes a sequence of first effective thickness values for plurality of different positions across the substrate (step 1110). The signal values in the trace from the sensor can be converted to thickness values using a correlation curve. The term “effective thickness” is used because the values of the thickness trace are calculated as if the signal is entirely due to the conductivity of the layer being polished. However, the signal includes contributions from underlying layers and the wafer, such as a doped wafer. The underlying layers and the wafer can have different conductivities than the layer being polished but are incorporated into the measurement as an ‘additional’ thickness of the material being polished.

[0109] [[Need to mention that ou can do multiple scans to and then added all measured pointes together to form a single radial trace]]

[0110] At least a portion of the first effective thickness trace is applied to a neural network to generate an adjusted first effective thickness trace (e.g., effective thickness trace 420’) (step 1112). Some variation in the signal intensity from the sensor can be caused by the sensor overlapping the substrate edge. The controller feeds at least a portion, e.g., all, of each thickness trace into the neural network configured to generate an adjusted thickness trace (e.g., adjusted thickness trace 450) that compensates for such signal distortions.

[0111] The adjusted first effective thickness trace is stored on a networked computer (step 1114). Storing the first effective thickness trace on a networked computer allows the first effective thickness trace to be used in downstream polishing process as a baseline. Thus, if the substrate is returned to the same polishing system, or distributed to a different polishing system, in downstream production steps, the first effective thickness trace can be used by either system.

[0112] The first patterned dielectric layer includes the conductive materials in the trenches and holes. A second patterned dielectric layer is formed on the substrate over the first patterned dielectric layer (step 1116). The second patterned dielectric layer is deposited using any deposition process suitable for dielectric layer deposition. A second conductive layer is deposited over the second patterned dielectric layer to fill trenches in the second patterned dielectric layer (step 1118). The second conductive layer is a conductive material as described herein. The conductive material of the second conductive layer can be the same conductive material as the first conductive layer, or a different conductive material.

[0113] The second conductive layer of the substrate is polished with a polishing pad (step 1120). In some examples, the substrate is returned to the polishing station at which the first conductive layer was polished. As such, the second conductive layer can be polished on the polishing pad that polished the first conductive layer. In some examples, the substrate is delivered to a different polishing station. As such, the second conductive layer can be polished on a different polishing pad than the polishing pad that polished the first conductive layer.

[0114] During polishing of the second conductive layer, a sensor of an in-situ eddy current monitoring system is scanned across the substrate so as to generate a second sequence of signal values for the scan of the sensor across the substrate (step 1122). In examples in which the substrate was returned to the polishing station at which the first conductive layer was polished, the sensor is the same sensor which generated the first sequence of signal values. In examples in which the substrate was distributed to a new polishing station, the sensor of the new polishing station is a different sensor, e.g., a second sensor, than that which generated the first sequence of signal values. The sensor can be repeatedly scanned across the substrate, e.g., more than once, to generate additional sequences of signal values. The sensor can be repeatedly scanned until a sufficient number of sequences of signal values are available to determine an accurate thickness profile of the conductive layer being scanned.

[0115] The second sequence of signal values for the scan of the sensor across the substrate are converted into a second effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate (step 1124).

[0116] At least a portion of the second effective thickness trace are applied to a neural network to generate an adjusted second effective thickness trace (step 1126). In some examples, the second effective thickness trace is applied to the same neural network that generated the adjusted first effective thickness trace. In some examples, the second effective thickness trace is applied to a different neural network trained to generate adjusted effective thickness traces.

[0117] The adjusted first effective thickness trace are retrieved from the networked computer (step 1128). The adjusted first effective thickness trace for the first conductive layer and the substrate can be received from a networked location, e.g., an upstream fab, deposition, or polishing system. Receiving the adjusted first effective thickness trace allows for highly accurate determination of amount of thickness to be removed in subsequent processing steps.

[0118] A thickness trace for the second conductive layer is generated to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values (step 1130). The generating can include subtracting the adjusted first effective thickness trace from the adjusted second effective thickness trace. Subtracting the adjusted first effective thickness trace from the adjusted second effective thickness trace removes signal contributions from the first conductive layer from the adjusted second effective thickness trace. Removing the contributions compensates for the adjusted second effective thickness trace to generate the thickness trace for the second conductive layer.

[0119] Polishing is halted or a polishing parameter that affects the polishing of second conductive layer is adjusted based on the thickness trace (step 1132). One of these steps, or both steps, can be completed. For example as polishing progresses, the controller can extrapolate forward to determine a time at which the conductive layer thickness will equal a target thickness value. This time provides the endpoint time (ET). As another example, by substituting the adjusted target thickness values for the target thickness values in the profile control algorithm, the algorithm will drive the polishing process toward providing those adjusted target thickness values.

[0120] Steps 1108-1132 of the method 1100 can be performed iteratively, e.g., each time with the second conductive layer providing the first conductive layer of the subsequent iteration of the method. For example, for a given conductive layer, rather than immediately removing the substrate from the polishing pad after a polishing endpoint of the given conductive layer has been detected (e.g., after step 1132), the given conductive layer can be subject to a post-polish operation in which a reference trace 420 is acquired (see step 1108) by scanning a sensor of the in-situ eddy current monitoring system across the substrate. Thus, for each given deposited conductive layer, an adjusted effective thickness trace 450 can be determined which includes contributions for all previous layers on the substrate. The adjusted effective thickness trace 450 can then be stored and used as the reference trace 420 when generating the thickness trace 480 for the next deposited conductive layer. This iterative technique can be performed for more than one of the conductive layers to be polished, e.g., for each conductive layer to be polished that is deposited on a substrate, e.g., for three or more layers, e.g., eight layers, or twelve layers. FIG. 12 is a plan view of a chemical mechanical polishing system 1200 having multiple networked polishing stations 20 for processing one or more substrates. The polishing system 1200 includes a plurality of polishing stations 20, e.g., three polishing stations 20a, 20b, and 20c. Each of the polishing stations 20 includes at least one carrier head 70, e.g., three carrier heads 70a, 70b, 70c. Each of the polishing stations 20 includes at least one controller 90, e.g., three controllers 90a, 90b, 90c.

[0121] The polishing system 1200 includes a computing device 1210, e.g., a server, which is in electronic communication with the controllers 90 of the polishing stations 20. The computing device 1210 receives the adjusted effective thickness traces 450 from the respective controllers 90 following the post-polish operations. The computing device 1210 stores the received adjusted effective thickness traces 450 for distribution to the polishing stations 20. In some examples, one of the polishing stations 20 receives a substrate for which an adjusted effective thickness trace for a deposited conductive layer is available on the computing device 1210. The computing device 1210 provides the adjusted effective thickness trace associated with the substrate to the polishing station 20 on which the substrate is to be polished. Thus, the polishing stations 20 can use adjusted effective thickness traces generated by different production machines, e.g., polishing stations 20.

[0122] The above-described polishing apparatus and methods can be applied in a variety of polishing systems. Either the polishing pad, or the carrier heads, or both can move to provide relative motion between the polishing surface and the substrate. For example, the platen may orbit rather than rotate. The polishing pad can be a circular (or some other shape) pad secured to the platen. Some aspects of the endpoint detection system may be applicable to linear polishing systems, e.g., where the polishing pad is a continuous or a reel-to-reel belt that moves linearly. The polishing layer can be a standard (for example, polyurethane with or without fillers) polishing material, a soft material, or a fixed-abrasive material. Terms of relative positioning are used to refer to relative positioning within the system or substrate; it should be understood that the polishing surface and substrate can be held in a vertical orientation or some other orientation during the polishing operation.

[0123] Functional operations of the controller 90 can be implemented using one or more computer program products, i.e., one or more computer programs tangibly embodied in a non-transitory computer readable storage media, for execution by, or to control the operation of, data processing apparatus, e.g., a programmable processor, a computer, or multiple processors or computers. The above-described polishing apparatus and methods can be applied in a variety of polishing systems. Either the polishing pad, or the carrier heads, or both can move to provide relative motion between the polishing surface and the substrate. For example, the platen may orbit rather than rotate. The polishing pad can be a circular (or some other shape) pad secured to the platen. Some aspects of the endpoint detection system may be applicable to linear polishing systems, e.g., where the polishing pad is a continuous or a reel-to-reel belt that moves linearly. The polishing layer can be a standard (for example, polyurethane with or without fillers) polishing material, a soft material, or a fixed-abrasive material. Terms of relative positioning are used to refer to relative positioning within the system or substrate; it should be understood that the polishing surface and substrate can be held in a vertical orientation or some other orientation during the polishing operation.

[0124] Functional operations of the controller 90 can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware, and structural equivalents thereof, or in combinations of them. The computer software can be implemented as one or more computer program products, i.e., one or more computer programs tangibly embodied in a non- transitory computer readable storage media, for execution by, or to control the operation of, data processing apparatus, e.g., a programmable processor, a computer, or multiple processors or computers. A computer program (also known as a program, software, software application, or code) can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a standalone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file. A program can be stored in a portion of a file that holds other programs or data, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub programs, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers at one site or distributed across multiple sites and interconnected by a communication network.

[0125] The processes and logic flows described in this specification can be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows can also be performed by, and apparatus can also be implemented as, special purpose logic circuitry-’, e.g., an FPGA (field programmable gate array) or an ASIC (application specific integrated circuit). A number of embodiments of the invention have been described. Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims.

Claims

CLAIMSWhat is claimed is:

1. A method of semiconductor device fabrication, comprising:forming a first patterned dielectric layer on a substrate;depositing a first conductive layer over the first patterned dielectric layer to fill trenches in the first patterned dielectric layer;chemical mechanical polishing the first conductive layer of the substrate with a first polishing pad until a top surface of the first patterned dielectric layer is exposed;in a first post-polish operation, maintaining the substrate in contact with the first polishing pad while substantially no polishing of the substrate is performed and scanning a first sensor of a first in-situ eddy current monitoring system across the substrate so as to generate a first sequence of signal values for a scan of the first sensor across the substrate,converting the first sequence of signal values for the scan of the first sensor across the substrate into a first effective thickness trace that includes a sequence of first effective thickness values for plurality of different positions across the substrate;applying at least a portion of the first effective thickness trace to a first neural network to generate an adjusted first effective thickness trace;storing the adjusted first effective thickness trace on a networked computer; forming a second patterned dielectric layer on the substrate over the first patterned dielectric layer;depositing a second conductive layer over the second patterned dielectric layer to fill trenches in the second patterned dielectric layer;chemical mechanical polishing the second conductive layer of the substrate with a second polishing pad;during polishing of second conductive layer, scanning a second sensor of a second in-situ eddy current monitoring system across the substrate so as to generate a second sequence of signal values for the scan of the second sensor across the substrate;converting the second sequence of signal values for the scan of the second sensor across the substrate into a second effective thickness trace that includes a sequence of effectivethickness values for plurality of different positions across the substrate;applying at least a portion of the second effective thickness trace to a second neural network to generate an adjusted second effective thickness trace;retrieving the adjusted first effective thickness trace from the networked computer; generating a thickness trace for the second conductive layer to compensate for a contribution of the first conductive layer underlying the second conductive layer; andat least one of halting polishing or adjusting a polishing parameter that affects the polishing of second conductive layer based on the thickness trace.

2. The method of claim 1, wherein the first polishing pad and the second polishing pad are different polishing pads.

3. The method of claim 1, wherein the first neural network and the second neural network are different neural networks.

4. The method of claim 1, wherein the first sensor and the second sensor are different sensors, and the first in-situ eddy current monitoring system and the second in-situ eddy current monitoring system are different in-situ eddy current monitoring systems.

5. The method of claim 1, wherein storing the adjusted first effective thickness trace comprises communicating the adjusted first effective thickness trace to the networked computer and storing the adjusted first effective thickness trace on a computer readable medium of the networked computer.

6. The method of claim 1, wherein generating the thickness trace includes subtracting the adjusted first effective thickness trace from the adjusted second effective thickness trace.

7. The method of claim 1, wherein maintaining the substrate in contact with the first polishing pad while substantially no polishing of the substrate is performed comprises applying no positive pressure to the substrate.

8. The method of claim 7, wherein applying zero positive pressure comprises venting one or more chamber of a carrier head holding the substrate to atmospheric pressure.

9. The method of claim 1, comprising rinsing a polishing liquid from the first polishing pad before the first post-polish operation.

10. The method of claim 1, comprising repeatedly scanning the first sensor of the first in-situ eddy current monitoring system across the substrate so as to generate multiple sequences of signal values and converting the multiple sequences of signal values into the first effective thickness trace.

11. A computer program product for controlling a fabrication system, the computer program product residing on a non-transitory computer readable medium and comprising instructions for causing one or more computers to:forming a first patterned dielectric layer on a substrate;depositing a first conductive layer over the first patterned dielectric layer to fill trenches in the first patterned dielectric layer;chemical mechanical polishing the first conductive layer of the substrate with a first polishing pad until a top surface of the first patterned dielectric layer is exposed; in a first post-polish operation, maintaining the substrate in contact with a polishing pad while substantially no polishing of the substrate is performed and scanning a sensor of an in-situ eddy current monitoring system across the substrate so as to generate a first sequence of signal values for a scan of the sensor across the substrate; converting the first sequence of signal values for the scan of the sensor across the substrate into a first effective thickness trace that includes a sequence of first effective thickness values for plurality of different positions across the substrate;applying at least a portion of the first effective thickness trace to a neural network to generate an adjusted first effective thickness trace;storing the adjusted first effective thickness trace on a networked computer; forming a second patterned dielectric layer on the substrate over the first patterned dielectric layer;depositing a second conductive layer over the second patterned dielectric layer to fill trenches in the second patterned dielectric layer;chemical mechanical polishing the second conductive layer of the substrate with a polishing pad;during polishing of second conductive layer, scanning a sensor of an in-situ eddy current monitoring system across the substrate so as to generate a second sequence of signal values for the scan of the sensor across the substrate;converting the second sequence of signal values for the scan of the sensor across the substrate into a second effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate;applying at least a portion of the second effective thickness trace to a neural network to generate an adjusted second effective thickness trace;retrieving the adjusted first effective thickness trace from the networked computer;generating a thickness trace for the second conductive layer to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values, including subtracting the adjusted first effective thickness trace from the adjusted second effective thickness trace; andat least one of halting polishing or adjusting a polishing parameter that affects the polishing of second conductive layer based on the thickness trace.

12. A method of polishing a substrate, comprising:polishing a first conductive layer of the substrate with a polishing pad;in a post-polish operation, maintaining the substrate in contact with the first polishing pad while substantially no polishing of the substrate is performed and scanning a sensor of an in-situ eddy current monitoring system across the substrate so as to generate a sequence of signal values;converting the sequence of signal values into an effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate; applying at least a portion of the effective thickness trace to a neural network to generate an adjusted effective thickness trace;retrieving an adjusted second effective thickness trace from a networked computer, the second effective thickness trace representative of a second sequence of adjusted thickness values for a scan of the sensor across a second conductive layer beneath the first conductive layer on the substrate;generating a thickness trace for the first, conductive layer to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values, including subtracting the adjusted second effective thickness trace from the adjusted first effective thickness trace; andat least one of halting polishing or adjusting a polishing parameter that affects the polishing of first conductive layer based on the thickness trace.

13. The method of claim 12, wherein the first polishing pad and the second polishing pad are different polishing pads.

14. The method of claim 12, wherein the first neural network and the second neural network are different neural networks.

15. The method of claim 12, wherein the first sensor and the second sensor are different sensors, and the first in-situ eddy current monitoring system and the second in-situ eddy current monitoring system are different in-situ eddy current monitoring systems.

16. The method of claim 12, wherein storing the adjusted first effective thickness trace comprises communicating the adjusted first effective thickness trace to the networked computer and storing the adjusted first effective thickness trace on a computer readable medium of the networked computer.

17. A chemical mechanical polishing system, comprising:a networked computer; andmultiple chemical mechanical polishing stations in communication with the networked computer, each chemical mechanical polishing station comprising:a platen supporting a polishing pad;a carrier head to hold a surface of a substrate against the polishing pad;a motor to generate relative motion between the platen and the carrier head so as to polish an overlying layer on the substrate;an in-situ eddy current monitoring system comprising a sensor arranged within a recess of the platen; anda controller configured to cause the respective chemical mechanical polishing station to:polish a first conductive layer of the substrate with the polishing pad; in a post-polish operation, maintain the substrate in contact with the polishing pad while substantially no polishing of the substrate is performed and scan a sensor of the in-situ eddy current monitoring system across the substrate so as to generate a sequence of signal values;convert the sequence of signal values into an effective thickness trace that includes a sequence of effective thickness values for plurality of different positions across the substrate;apply at least a portion of the effective thickness trace to a neural network to generate an adjusted effective thickness trace;retrieve an adjusted second effective thickness trace from the networked computer, the second effective thickness trace representative of a second sequence of adjusted thickness values for a scan of the sensor across a second conductive layer beneath the first conductive layer on the substrate;generate a thickness trace for the first conductive layer to compensate for a contribution of layers of the substrate underlying the second conductive layer to second sequence of signal values, including subtracting the adjusted second effective thickness trace from the adjusted first effective thickness trace; and at least one of halt polishing or adjust a polishing parameter that affects the polishing of first conductive layer based on the thickness trace.

18. The method of claim 17, wherein the second adjusted effective thickness trace is generated by a different chemical mechanical polishing station than the chemical mechanical polishing station polishing the first conductive layer.

19. The method of claim 17, wherein the second adjusted effective thickness trace is generated by a different neural network than the first adjusted effective thickness trace.

Images