JP2012256911A - Device and method for monitoring chemical and mechanical polishing according to spectrum - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a step of determining when a terminal point of a polishing step is achieved.SOLUTION: A device and a method include detection of a terminal point according to the spectrum, adjustment of a polishing speed according to the spectrum, injection of an optical head top face, or a pad 30 with a window. In detection of the terminal point according to the spectrum, a reference spectrum, in which a terminal point determination theory based on a concrete spectrum is empirically selected, is used to achieve target thickness when the terminal point is called by applying the terminal point theory based on the spectrum. A polishing terminal point is determined using a different trace or a series of sequences. An injection system creates a layer-like gas flow to the top face of an optical head 53. A vacuum nozzle and a vacuum source are constituted so that the gas flow is formed into a gas form. The window includes a flexible plastic portion, and a crystalline or glass portion. Adjustment of the polishing speed according to the spectrum includes a step of obtaining a spectrum in a different zone on a substrate.

Description

[0001] The present invention relates generally to chemical mechanical polishing of substrates.

[0002] Integrated circuits are typically formed on silicon wafers by successively depositing conductive, semiconductive, and insulating layers on a substrate. One fabrication step includes depositing a filler layer on a non-planar surface and planarizing the filler layer. In a specific application, the filler layer is planarized until the top surface of the patterned layer is exposed. For example, a trench or hole in the insulating layer is filled by depositing a conductive filler layer on the patterned insulating layer. After planarization, the portions of the conductive layer remaining between the raised patterns of the insulating layer form vias, plugs, and lines, thereby providing a conductive path between thin film circuits on the substrate. In other applications such as oxide polishing, the filler layer is planarized until a predetermined thickness of filler layer remains on the non-planar surface. In addition to this, planarization of the substrate surface is usually required for photolithography.

[0003] Chemical mechanical polishing (CMP) is one accepted planarization method. This planarization method typically requires that the substrate be mounted on a carrier or polishing head. Typically, the exposed surface of the substrate is placed against a rotating abrasive disc pad or belt pad. The polishing pad may be either a standard pad or a fixed polishing pad. Standard pads have a durable rough surface, while stationary polishing pads have abrasive particles held in a contaminated medium. The carrier head provides a controllable load on the substrate and presses it against the polishing pad. Typically, a polishing slurry is supplied to the surface of the polishing pad. The polishing slurry contains at least one chemically reactive agent and, when used with a standard polishing pad, contains abrasive particles.

[0004] One problem with CMP is determining whether the polishing process is complete, that is, whether the substrate layer has been planarized to the desired flatness or thickness, or whether the desired amount of material is It is to determine when it has been removed. Overpolishing (removing too much) of the conductive layer or film leads to increased circuit resistance. On the other hand, insufficient polishing (too little removal) of the conductive layer leads to an electrical short circuit. Changes in the initial thickness of the substrate layer, the slurry composition, the polishing pad condition, the relative speed between the polishing pad and the substrate, and the load on the substrate can cause changes in the material removal rate. Due to these changes, the time required to reach the polishing end point also changes. Therefore, the polishing end point cannot simply be determined as a function of the polishing time.

[0005] In one general aspect, the invention features a computer-implemented method that includes selecting a reference spectrum. The reference spectrum is a spectrum of white light that is located on the first substrate and reflected by the target film that is thicker than the target thickness. The reference spectrum is empirically selected for the endpoint determination logic based on the specific spectrum such that the target thickness is achieved when the endpoint is called by applying the endpoint logic based on the specific spectrum. The present invention includes obtaining a current spectrum. The current spectrum is the spectrum of white light reflected on the target film located on the second substrate and having a current thickness greater than the target thickness. A polishing step is performed on the target film on the second substrate. The method includes determining for the second substrate when the end point of the polishing step has been achieved. This determination is based on the reference spectrum and the current spectrum.

[0006] In another general aspect, the invention features a computer-implemented method that includes selecting two or more reference spectra. Each reference spectrum is a white spectrum that is located on the first substrate and reflected by a target film that is thicker than the target thickness. The reference spectrum is empirically selected for the endpoint determination logic based on the specific spectrum such that the target thickness is achieved when the endpoint is invoked by applying the endpoint logic based on the specific spectrum. The method includes obtaining two or more current spectra. Each current spectrum is a spectrum of white light reflected by the target film, which is located on the second substrate and has a current thickness that is thicker than the target thickness. A polishing step is performed on the target film on the second substrate. The method includes determining, for a second substrate, whether an end point of the polishing step has been reached, the determination being based on a reference spectrum and a current spectrum.

[0007] In another general aspect, the invention features a computer program product that includes instructions that operate to cause a processor to select a reference spectrum. This reference spectrum is a spectrum of white light that is located on the first substrate and reflected by the target film that is thicker than the target thickness. The reference spectrum is empirically selected for the endpoint determination logic based on the specific spectrum such that the target thickness is achieved when the endpoint is invoked by applying the endpoint logic based on the specific spectrum. This product includes instructions that cause the processor to obtain the current spectrum. This current spectrum is the spectrum of white light reflected on the target film located on the second substrate and having a current thickness that is greater than the target thickness. A polishing step is performed on the target film on the second substrate. The product includes instructions that cause the processor to determine when the end of the polishing step has been reached for the second substrate. This determination is based on the reference spectrum and the current spectrum. This product is actually stored on a medium that can be read by the machine.

[0008] In yet another general aspect, the invention features a computer program product that is actually stored on a machine-readable medium. The product includes instructions that cause the processor to select more than one reference spectrum. The reference spectrum is a spectrum of white light that is located on the first substrate and reflected by the target film that is thicker than the target thickness. The reference spectrum is empirically selected for the endpoint determination logic based on the specific spectrum so that the target thickness is achieved when the endpoint is called by applying the endpoint logic based on the specific spectrum. The product further comprises instructions for obtaining two or more current spectra. Each current spectrum is a spectrum of white light reflected on a target film located on the second substrate and having a current thickness greater than the target thickness. A polishing step is performed on the target film on the second substrate. The product further comprises instructions for determining whether the end of the polishing step has been reached for the second substrate, the determination being made based on the reference spectrum and the current spectrum.

[0009] In one general aspect, the invention features an ejection system that ejects the top surface of an optical head. The system includes a gas source configured to provide a gas flow, a transfer nozzle, a transfer line connecting the gas source to the transfer nozzle, a vacuum source configured to provide a vacuum, a vacuum nozzle, and a vacuum source. Includes a vacuum line connected to the nozzle. The gas source and the transport nozzle are configured to guide the gas flow toward the top surface of the optical head. The vacuum nozzle and the vacuum source are configured so that the gas flow is layered.

[0010] In another general aspect, the invention features an injection system that injects the bottom surface of a polishing pad window. The system includes a gas source configured to provide a gas flow, a transfer nozzle, a transfer line connecting the gas source to the transfer nozzle, a vacuum source configured to provide a vacuum, a vacuum nozzle, and a vacuum source. Includes a vacuum line connected to the nozzle. The gas source and the transfer nozzle are configured to guide the gas flow at the bottom of the polishing pad window, thereby preventing the formation of condensation on the bottom surface of the polishing pad window.

[0011] In one general aspect, the invention features an assembly that performs chemical mechanical polishing. The assembly includes a polishing pad having a polishing surface. The assembly also includes a solid window installed in the polishing pad, which provides optical access through the polishing pad. The solid window includes a first portion made from polyurethane and a second portion made from quartz. The first portion has a surface that is coplanar with the polishing surface of the polishing pad.

[0012] In another general aspect, the invention features a polishing pad that includes a polishing layer having a top surface and a bottom surface. The pad includes an aperture having a first opening provided on the top surface and a second opening provided on the bottom surface. The top surface is a polished surface. The pad includes a window that includes a first portion made from a flexible plastic and a second portion that is crystalline or glassy. This window allows white light to pass through. The window is installed in the aperture so that the first part is inserted into the aperture and the second part is located on the bottom side of the first part, and the first part functions as a slurry barrier. To do.

[0013] In another general aspect, the invention features a polishing pad making method. In this method, a lump of crystalline or glassy material that transmits white light is placed in a mold of a polishing pad window. The method includes discharging a liquid precursor of a flexible plastic material that transmits white light into a mold. The method further includes curing the liquid precursor to form a window including a first portion made from a flexible plastic material and a second portion that is crystalline or glassy. It also includes the step of placing the window in a mold of the polishing pad. The method further includes discharging a liquid precursor of the polishing pad material into the mold of the polishing pad. The method also includes the step of producing a polishing pad by curing a liquid precursor of the polishing pad material. In this case, the window is installed in the polishing pad when the polishing pad is manufactured, and the first portion is anti-slurry. It is installed in the mold of the polishing pad so as to function as a barrier.

[0014] In another general aspect, the invention features a method of making a polishing pad. The method includes placing a mass of crystalline and glassy material that transmits white light into a mold of a polishing pad window. The method further includes discharging a liquid precursor of a flexible plastic material that transmits white light into the mold. The method also includes curing the liquid precursor to form a window including a first portion made from a flexible plastic material and a second portion that is crystalline or glassy. The method includes the step of forming a polishing layer including an aperture, the polishing layer having a top surface and a bottom surface, the aperture having a first opening provided on the top surface and a second opening provided on the bottom surface. The top surface is a polished surface. The method includes the step of inserting a window into the aperture, wherein the window is located on the bottom surface of the first portion where the first portion is inserted into the aperture and the second portion functions as a slurry-resistant barrier. So that it is installed in the aperture.

[0015] In another general aspect, the invention features a method of making a polishing pad. The method includes forming a first portion of a polishing pad window, the first portion having a recess and allowing white light to pass. The method further includes inserting a mass of crystalline or glassy material that transmits white light into the recess. The method further includes the step of forming a polishing layer including an aperture, the polishing layer having a top surface and a bottom surface, and the aperture having a first opening provided on the top surface and a second opening provided on the bottom surface. The top surface is a polished surface. The method also includes the step of inserting a window into the aperture so that the window is inserted into the aperture and the second portion is on the bottom side of the first portion that functions as a slurry barrier. It is installed in the aperture so as to be located at

[0016] In another general aspect, the invention features a computer-implemented method. During the polishing sequence, a first spectrum of reflected light is obtained in a first zone on the substrate and a second spectrum is obtained in a second zone on the substrate. In order to determine a first index for the first spectrum and a second index for the second spectrum, the first spectrum and the second spectrum are compared with a spectrum library. At different times during the polishing sequence, a third spectrum of reflected light is obtained from the first zone and a fourth spectrum is obtained from the second zone. The third spectrum and the fourth spectrum are compared with the library to determine a third index for the first zone and a fourth index for the second zone. The polishing rate in the first zone is determined from the first index and the third index, and the polishing rate in the second zone is determined from the second index and the fourth index. Based on the first polishing rate, the second polishing rate, the first target relative thickness of the first zone, and the second target relative thickness of the second zone, an adjusted polishing rate is determined for the second zone, thereby At the same time as the first zone is polished to the first target relative thickness, the second zone is polished to the second target relative thickness.

[0017] Implementations of the invention include one or more of the following features. The first zone may be an inner zone and the second zone may be an outer annular zone. Determining the adjusted polishing rate for the second zone may include determining when the first target relative thickness is within a predetermined threshold range from the second target relative thickness. Determining the adjusted polishing rate for the second zone may include determining an estimated end point time for the polishing sequence. Obtaining the first spectrum and the second spectrum may include obtaining a white light spectrum. The method further includes adjusting the polishing system parameters to polish the second zone at an adjusted polishing rate. The step of determining the adjusted polishing rate can be performed on the setting substrate, the step of adjusting the parameters of the polishing system can be performed on the product substrate, or both of the two steps can be performed on the product substrate. Can be run on. Adjusting the parameters of the polishing system may include adjusting the pressure. Determining the adjusted polishing rate can include determining the polishing rate so that, upon completion of the polishing sequence, the cross-section along the diameter of the substrate is a flat or bowl-shaped profile. . Obtaining the first spectrum and the second spectrum can include sampling the substrate at different rotational locations. Obtaining the first spectrum and the second spectrum includes measuring the spectrum reflected by the oxide film. The method further includes polishing the set substrate until it is over-polished, obtaining multiple spectra from a zone of the test substrate during polishing, and creating a spectral library, Storing in combination with the time the spectrum was acquired. The method may also include creating an index for the spectral library, where the index represents the spectrum obtained from the configuration substrate at a particular time.

[0018] In another general aspect, the invention features a method of monitoring a chemical mechanical polishing process. A multi-wavelength light beam is guided onto the substrate being polished and the spectrum of the light reflected from the substrate is measured. The light beam is moved in a path across the substrate surface. A series of spectral measurements is extracted from the signal and a radial position on the substrate is determined for each spectral measurement. Spectral measurements are stored in a plurality of radial ranges according to radial position. A polishing endpoint for the substrate is determined from at least one spectral measurement of the plurality of radial ranges. The method further includes determining an adjusted polishing rate for one of the plurality of radial ranges and applying the adjusted polishing rate to one of the plurality of radial ranges.

[0019] As used herein, the term "substrate" includes, for example, a product substrate (eg, including a plurality of memory or processor dies), a test substrate, a bare substrate, a gate substrate. it can. The substrate may be at various stages of integrated circuit fabrication, for example, the substrate may be a bare wafer and may include one or more deposited and / or patterned layers. The term “substrate” includes circular disks and rectangular sheets.

[0020] Feasible benefits of implementation of the present invention include one or more of the following. The end point can be determined substantially without considering the change in polishing rate. In general, there is no need to consider factors that affect the polishing rate, such as consumables. Using multiple reference and / or target spectra (relative to the use of a single reference spectrum and / or a single target spectrum) is generally simpler than traces generated using single reference spectrum technology Difference or endpoint traces can be provided, which improves the accuracy in endpoint determination. The spray system can reduce the likelihood that the slurry will dry on the substrate surface being polished. The polishing pad window can extend the accuracy and / or precision of endpoint determination.

[0021] The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

[0042] In many drawings, similar elements are denoted by similar reference numerals.

FIG. 1 shows a polishing apparatus 20 that can operate to polish a substrate 10. The polishing apparatus 20 includes a rotatable disk-shaped platen 24 on which a polishing pad 30 is installed. The platen is operable to rotate about the axis 25. For example, the platen 24 can rotate when the motor rotates the drive shaft 22. The polishing pad 30 can be removably secured to the platen 24 by, for example, an adhesive layer. When the polishing pad 30 is fatigued, it can be removed and replaced. The polishing pad 30 may be a two-layer polishing pad provided with an external polishing layer 32 and a flexible backing layer 34.

[0044] Including an aperture (ie, a hole through the pad) or a solid window provides optical access 36 through the polishing pad. Although the solid window can be secured to the polishing pad, in some implementations the solid window can be supported on the platen 24 and protrude into the aperture of the polishing pad. Usually, the polishing pad 30 is placed on the platen 24 so that an aperture or window installed in the recess 26 of the platen 24 overlaps the optical head 53. As a result, the optical head 53 has optical access to the substrate being polished through the aperture or window. Next, the optical head will be further described.

[0045] The window may be a hard crystalline or glassy material such as quartz or glass, a flexible plastic material such as silicone or polyurethane or a halogenated polymer (eg a fluoropolymer), or a combination of these materials. May be. The window may be transparent to white light. If the top surface of the solid window is a hard crystalline or glassy material, the top surface must be sufficiently retracted from the polished surface to prevent scratching. If the top surface is close to or in contact with the polishing surface, the top surface of the window must be made of a flexible plastic material. In some implementations, the solid window is a polyurethane window affixed within the polishing pad, or a window having a combination of quartz and polyurethane. The window may have a high transmittance of, for example, about 80% with respect to a single color light of a specific color such as blue light or red light. Further, by sealing the window with respect to the polishing pad 30, it is possible to prevent liquid from leaking from the boundary between the window and the polishing pad 30.

[0046] In one implementation, the window comprises a hard crystalline or glassy material covered with an outer layer of a flexible plastic material. The top surface of the flexible material may be coplanar with the polishing surface. The bottom surface of the hard material may be in the same plane as the bottom surface of the polishing pad or may recede from this surface. In particular, if the polishing pad includes two layers, the solid window can be integrated with the polishing layer so that the bottom layer has an aperture aligned with the solid window.

[0047] Assuming that the window includes a combination of a hard crystalline or glassy material and a flexible plastic material, it is not necessary to use an adhesive to secure the two parts. For example, in one implementation, no adhesive is used to bond the polyurethane portion of the window to the quartz portion. Alternatively, an adhesive that is transparent to white light can be used, or an adhesive can be added so that light passing through the window does not pass through the adhesive. As an example, the adhesive can be applied only around the boundary between the polyurethane portion and the quartz portion. A refractive index gel may be added to the bottom of the window.

[0048] Optionally, one or more recesses can be included in the bottom surface of the window. The shape of the recess may be capable of accommodating, for example, the end of an optical fiber cable or the end of an eddy current sensor. This recess allows the end of the optical fiber cable or the end of the eddy current sensor to be installed at a location away from the surface of the substrate being polished which is thinner than the window thickness. In implementations where the window includes a hard crystalline or glassy portion and such a portion is recessed by machining, the recess is polished to remove scratches caused by the machining. . Alternatively, a solvent and / or liquid polymer can be added to the surface of the recess to remove scratches generated during machining. By removing scratches that are usually caused by machining, scattering is suppressed and light transmission through the window is improved.

[0049] FIGS. 2A-2H show various window implementations. As shown in FIG. 2A, the window can have two portions, a polyurethane portion 202 and a quartz portion 204. These portions are layered, and a polyurethane portion 202 is installed on the quartz portion 204. A window may be installed in the polishing pad so that the top surface 206 of the polyurethane layer is flush with the polishing surface 208 of the polishing pad.

[0050] As shown in FIG. 2B, the polyurethane portion 202 may have a recess in which the quartz portion is installed. The bottom surface 210 of the quartz portion is exposed.

[0051] As shown in FIG. 2C, the polyurethane portion 202 may include protrusions, which may be protrusions 212 that protrude into the quartz portion 204, for example. This protrusion serves to reduce the tendency of the polyurethane portion 202 to be pulled away from the quartz portion 204 due to friction from the substrate or retaining ring.

[0052] As shown in FIG. 2D, the boundary between the polyurethane portion 202 and the quartz portion 204 may be a rough surface. Such a surface enhances the bonding of the two portions of the window and can further reduce the tendency of the polyurethane portion 202 to be pulled away from the quartz portion 204 due to friction from the substrate or retaining ring.

[0053] As shown in FIG. 2E, the thickness of the polyurethane portion 202 may be non-uniform. The thickness of the place that enters the light beam path 214 is smaller than the thickness of the place that does not enter the light beam path 214. According to the illustrated method, the thickness t ₁ is less than the thickness t ₂ . Or the thickness of the edge of a window can also be made thin.

[0054] An adhesive 216 can be used to attach the polyurethane portion 202 to the quartz portion 204, as shown in FIG. 2F. The adhesive is applied so that it does not enter the light beam path 214.

[0055] As shown in FIG. 2G, the polishing pad may include a polishing layer and a backing layer. The polyurethane portion 202 passes through the polishing layer and extends at least partially into the backing layer. The hole provided in the backing layer may be larger than the hole in the polishing layer, and the section of polyurethane corresponding to the backing layer may be wider than the section of polyurethane corresponding to the polishing layer. Thus, the polishing layer provides a lip 218 that extends to the window and functions to resist the force pulling the polyurethane portion 202 from the quartz portion 204. The polyurethane portion 202 coincides with a hole in the layer of the polishing pad.

[0056] As shown in FIG. 2H, a refractive index gel 220 can be added to the bottom surface 210 of the quartz portion 204 to provide a medium for moving light from the fiber cable 222 to the window. The refractive index gel 220 fills the capacitance between the fiber cable 222 and the quartz portion 204, and further has a refractive index that is equal to or intermediate between the refractive index of the fiber cable 222 and that of the quartz portion 204. be able to.

[0057] In implementations where the window includes both a quartz portion and a polyurethane portion, the polyurethane portion should have a thickness that does not cause the polyurethane to fatigue and expose the quartz portion during the life of the polishing pad. is there. Quartz can be retracted from the bottom surface of the polishing pad, or a portion of the fiber cable 222 can extend into the polishing pad.

[0058] The windows and polishing pads described above can be manufactured using a variety of techniques. A polishing pad backing layer 34 may be attached to the polishing outer layer 32 with, for example, an adhesive. For example, by cutting or casting the pad 30 to include an aperture, an aperture can be formed in the pad 30 that provides optical access 36, and a window can be inserted into the aperture, such as by adhesive. It can be fixed to. Alternatively, the window can also be formed by discharging a liquid precursor to be the window into the aperture of the pad 30 and curing it. Alternatively, a solid transparent element, such as the crystalline or glassy portion described above, may be positioned within the liquid pad material and the liquid pad material may be cured to form a pad 30 around the transparent element. In either of these two cases, a block of pad material can be formed and the layer of polishing pad from which the window has been cast can be removed from the block.

[0059] In an implementation in which the window includes a crystalline or glassy first portion and a second portion made of a flexible plastic material, the second portion is padded by applying the liquid precursor technique described above. It can be formed in 30 apertures. Thereafter, the first portion can be inserted. If the first part is inserted before the liquid precursor of the second part solidifies, the first part and the second part are bonded by curing. When the first part is inserted after the liquid precursor is cured, the first part and the second part can be fixed with an adhesive.

[0060] The polishing apparatus 20 may include an injection system for improving light transmission through the optical access 36. There are various ways to implement this injection system. In an implementation of the polishing apparatus 20 in which the polishing pad 30 includes an aperture instead of a solid window, the jet system can be implemented to provide a laminar flow of fluid, such as a gas or liquid, over the top surface of the optical head 53. Can do. (This top surface may be the top surface of a lens built into the optical head 53.) The laminar flow of fluid flowing through the top surface of the optical head 53 clears the opaque slurry from the optical access, and / or Alternatively, it is possible to prevent the slurry from drying on the top surface, and as a result, transmission is improved through optical access. In implementations where the polishing pad 30 includes a solid window rather than an aperture, an injection system is implemented to guide the gas flow at the bottom of the window. The gas flow prevents condensation from forming on the bottom surface of the solid window that can interfere with optical access.

[0061] FIG. 3 shows an implementation of a laminar flow injection system. The injection system includes a gas source 302, a delivery line 304, a transport nozzle 306, a suction nozzle 308, a vacuum line 310, and a vacuum source 312. The gas source 302 and the vacuum source can be configured such that the same or similar volume of gas can be introduced or aspirated. The transfer nozzle 306 can be installed so that a laminar flow of gas can be guided over the transparent top surface 314 in the in-situ monitoring module rather than on the substrate surface being polished. As a result, there is no situation where the slurry is dried by the laminar flow of gas on the surface of the substrate being polished, which has an undesirable effect on polishing.

[0062] FIG. 4 shows an implementation of an injection system to prevent the formation of condensation on the bottom surface of the solid window. This system reduces or prevents the formation of condensation at the bottom of the polishing pad window. The system includes a gas source 402, a delivery line 404, a transfer nozzle 406, a suction nozzle 408, a vacuum line 410, and a vacuum source 412. The gas source 402 and the vacuum source can be configured to introduce and draw the same or similar volumes of gas. The transfer nozzle 406 can be installed so that the gas flow is guided on the bottom surface in the polishing pad 30.

[0063] In one implementation that is an alternative to the implementation of FIG. 4, the injection system does not include a vacuum source or vacuum line. As an alternative to these components, the injection system has a vent formed in the platen, which allows gas introduced into the space under the solid window to the side of the platen or to the polishing device. It can be discharged to other places having moisture resistance.

[0064] The gas source and vacuum source described above can be placed away from the platen so that they do not rotate with the platen. In this case, each of the supply line and the vacuum line includes a rotary coupler for propagating the gas.

Returning to FIG. 1, the polishing apparatus 20 includes a combined slurry / rinse arm 39. During polishing, the arm 39 operates to discharge a slurry 38 containing a liquid and a pH adjuster. Alternatively, the polishing apparatus includes a slurry port that operates to discharge slurry onto the polishing pad 30.

[0066] The polishing apparatus 20 includes a carrier head 70 that operates to hold the substrate 10 against the polishing pad 30. The carrier head 70 hangs down from a support structure 72 such as a carousel, and is connected to the carrier head rotation motor 76 by a carrier drive shaft 74, so that the carrier head can rotate around the shaft 71. In addition, the carrier head 70 can vibrate laterally within a radial slot formed in the support structure 72. During operation, the platen rotates about the central axis 25, and the carrier head rotates about the central axis 71 and translates laterally over the top surface of the polishing pad.

[0067] The polishing apparatus also includes an optical monitoring system that can be used to determine the polishing endpoint described below. This optical monitoring system includes a light source 51 and a photodetector 52. Light passing from the light source 51 through the optical access 36 provided on the polishing pad 30 collides with the substrate 10, is reflected from the substrate 10, passes through the optical access 36 again, and moves to the photodetector 52.

[0068] Using the branching optical cable 54, light can be transmitted from the light source 51 to the optical access 36 and then returned from the optical access 36 to the photodetector 52. The branching optical cable 54 may include a “trunk” 55 and two “branching portions” 56 and 58.

[0069] As described above, the platen 24 includes the recess 26 in which the optical head 53 is installed. The optical head 53 holds one end of a trunk 55 of a branch fiber cable 54 configured to propagate light to and from the substrate surface being polished. The optical head 53 may include one or more lenses or windows overlying the ends of the branching fiber cable 54 (eg, as shown in FIG. 3). Alternatively, the optical head 53 may simply hold the end of the trunk 55 near the solid window provided on the polishing pad. The optical head 53 can hold the nozzle of the ejection system described above. Furthermore, the optical head 53 can be removed from the recess 26 as necessary, for example, to perform preventive or corrective maintenance.

[0070] The platen includes a removable in-situ monitoring module 50. The in-situ monitoring module 50 can include one or more of the following: a light source 51, a photodetector 52, and circuit components that transmit and receive signals between the light source 51 and the photodetector 52. For example, the output of detector 52 may be a digital electronic signal that passes through a rotary coupler, such as a slip ring, in the drive shaft and reaches the controller of the optical monitoring system. Similarly, the light source can be switched on and off in response to a control command in a digital electronic signal that reaches the module 50 from the controller through the rotary coupler.

[0071] The in-situ monitoring module can also hold corresponding ends of the branches 56, 58 of the branch optical fiber 54. The light source is transmitted through the branching portion 56 and emitted from the end portion of the trunk 55 installed in the optical head 53, and further operates to transmit light impinging on the substrate being polished. The light reflected from the substrate is received at the end of the trunk 55 installed in the optical head 53 and propagates to the photodetector 52 through the branching portion 58.

[0072] In one implementation, the branch fiber cable 54 is a bundle of optical fibers. The bundle includes a first group of optical fibers and a second group of optical fibers. One optical fiber in the first group is connected to propagate light from the light source 51 to the surface of the substrate being polished. One optical fiber in the second group is connected so as to receive light reflected from the surface of the substrate being polished and to propagate the received light to the photodetector. The optical fibers are such that the optical fibers in the second group form an X-shape centered on the longitudinal axis of the branched optical fiber 54 (when viewed in cross section of the branched fiber cable 54). Can be arranged. Alternatively, other arrangements are possible. For example, the optical fibers in the second group can form a V-shape that is a mirror image of each other. A suitable branching optical fiber is available from Verity Instruments, Inc., Carrollton, Texas.

[0073] Typically, there is an optimum distance between the polishing pad window and the end of the trunk 55 of the branched fiber cable 54 in the vicinity of the polishing pad window. This distance can be determined empirically and is affected, for example, by the reflectivity of the window, the shape of the light beam emitted from the branch fiber cable, and the distance to the substrate being monitored. In some embodiments, the branching fiber cable is installed so that the end near the window is as close as possible to the bottom of the window without actually contacting the window. In this implementation, the polishing apparatus 20 is a mechanism that is part of the optical head 53, for example, a mechanism that can operate to adjust the distance between the end of the branch fiber cable 54 and the bottom surface of the polishing pad window. Can be included. Alternatively, the vicinity of the branch type fiber cable is embedded in the window.

[0074] The light source 51 operates to emit white light. In one implementation, the emitted white light includes light having a wavelength of 200 to 800 nanometers. A suitable light source is a xenon lamp or a xenon mercury lamp.

[0075] The photodetector 52 may be a spectrometer. Basically, a spectrometer is an optical instrument for measuring the nature of light over a portion of the electromagnetic spectrum, such as intensity. A suitable spectrometer is a diffraction grating spectrometer. A typical output of a spectrometer is the intensity of light as a function of wavelength.

[0076] Optionally, the in-situ monitoring module 50 can include other sensor elements. The in-situ monitoring module 50 may include, for example, an eddy current sensor, a laser, a light emitting diode, and a photodetector. In implementations where the in-situ monitoring module 50 includes an eddy current sensor, the module 50 is typically installed so that the substrate being polished is within the working range of the eddy current sensor.

[0077] The light source 51 and photodetector 52 are connected to a computing device operable to control these operations and receive signals from them. The computing device can include a microprocessor, such as a personal computer, installed near the polishing apparatus. For control, the computing device can, for example, synchronize the activation of the light source 51 with the rotation of the platen 24. As shown in FIG. 5, the computer can cause the light source 51 to emit a series of flashes that start immediately before the substrate 10 passes through the in-situ monitoring module and end immediately after this pass. (Each point 501 to 511 in the figure represents a location where light from the in-situ monitoring module collides and is reflected.) Alternatively, the computer may cause the light source 51 immediately before the substrate 10 passes the in-situ monitoring module. Light that starts and ends immediately after this passage can be emitted continuously.

[0078] For signal reception, the computing device may receive a signal carrying information describing the spectrum of light received by the photodetector 52, for example. FIG. 6A shows an example of a spectrum measured from light emitted by one flash of the light source and reflected by the substrate. This spectrum shows a raw spectrum, that is, a spectrum before normalization. The spectrum 602 is measured from the light reflected from the product substrate. The spectrum 604 is measured from light reflected from a base silicon substrate (a wafer having only a silicon layer). The spectrum 606 is from light received by the optical head 53 when the substrate is not installed on the optical head. In this state, referred to herein as the dark state, the light received is typically scattered light from the pad window light.

[0079] The computing device may process the signal described above to determine the end point of the polishing step. Without being limited to any specific theory, the spectrum of light reflected by the substrate 10 develops as polishing proceeds. FIG. 6B provides an example of this development that proceeds with the progress of polishing of the target film. The spectrum of FIG. 6B is normalized. Different spectral lines represent different polishing times. As can be seen from the figure, the spectral properties of the reflected light change with changes in the thickness of the film, and the specific spectrum is presented at the specific thickness of the film. The computing device can perform logic based on one or more spectra to determine when the end point has been reached. The one or more spectra that are the basis for endpoint determination can include a target spectrum, a reference spectrum, or both.

[0080] As used herein, a target spectrum refers to a spectrum exhibited by white light reflected from a target film having a target thickness. As an example, the target thickness may be 1, 2, 3 microns. Alternatively, the target thickness may be zero, for example, when the target film is excluded and the underlying film is exposed.

[0081] A plurality of target spectra may exist, and usually exist, for a specific target thickness. This is because polishing occurs at a finite value to allow the target film to maintain the target thickness during the period when multiple spectra are collected. In addition, different spectra are typically produced from different areas of the patterned substrate (even if multiple spectra were obtained at the same time point during polishing). For example, the spectrum of light reflected from scribe lines in the substrate is different (eg, has a different shape) from the spectrum of light reflected from the same array of substrates. In this specification, such a phenomenon is called a pattern effect. Thus, multiple spectra may exist for a specific target thickness, and the multiple spectra may include different spectra due to pattern effects.

[0082] FIG. 7A shows a method 700 for obtaining one or more target spectra. The property of the substrate having the same pattern as the product substrate is measured (step 702). In this specification, a substrate to be measured is called a “setting” substrate. The setting substrate may simply be a substrate similar to or the same as the product substrate, and the setting substrate may be a single substrate in a batch. This property includes the thickness of the pre-polished target film at a specific target location on the substrate. Typically, thickness is measured at multiple locations. Typically, these locations are selected so that the same type of die characteristics is measured for each location. Measurements are performed at the metrology station.

[0083] The setting substrate is polished according to the polishing step of interest, and a spectrum of white light reflected from the substrate surface being polished is collected during polishing (step 704). Polishing and spectrum collection can be performed in the polishing apparatus described above. The spectrum is collected by an in situ monitoring system during polishing. Multiple spectra can be collected for each platen rotation. Overpolishing the substrate, i.e., polishing beyond the estimated end point, can provide a spectrum of light reflected from the substrate upon achieving the target thickness.

[0084] The properties of the overpolished substrate are measured (step 706). This property includes the thickness of the target film after polishing at one or more specific locations used for pre-polishing measurements.

[0085] Using this measured thickness and the collected spectrum, one or the spectrum determined to be exhibited by the substrate when the target thickness is achieved is selected from the collected spectra (step 708). In particular, linear interpolation can be performed using the pre-polishing film thickness and the post-polishing substrate thickness to determine the spectrum when the target film thickness is achieved. The spectrum determined as the spectrum when the target thickness is achieved is designated as the target spectrum for the batch of the multiple substrates. Typically, three of the collected spectra are designated as target spectra. Alternatively, 5, 7, and 9 spectra are designated as target spectra.

[0086] Optionally, the collected spectrum is processed to extend accuracy and / or precision. The processing of the spectrum is performed, for example, to normalize the spectrum to a common reference, to average it, and / or to remove noise from the spectrum. Specific implementation of these processing operations will be described below.

[0087] As used herein, a reference spectrum refers to a spectrum related to the target film thickness. Typically, one, two, or more reference spectra for a specific spectrum are empirically selected based on endpoint determination logic based on the specific spectrum. Thus, the target thickness is achieved when the computing device adds endpoint determination logic based on a specific spectrum and calls the endpoint. As described below with reference to FIG. 7B, it is possible to repeatedly select a reference spectrum or multiple reference spectra. Usually, the reference spectrum is not the target spectrum. Rather, the reference spectrum is usually the spectrum of light reflected from the substrate when the film of interest is thicker than the target thickness.

[0088] FIG. 7B illustrates a method 701 for selecting a reference spectrum for a specific target thickness and endpoint determination logic based on the specific spectrum. In some embodiments, more than one spectrum can be selected instead of just one. As described in steps 702 and 706 above, the setting substrate is measured and polished (step 703). In particular, the collected spectra and the measurement time of each collected spectrum are recorded. Multiple spectra can be collected for each platen rotation during polishing.

[0089] The polishing value of the polishing apparatus for a specific set substrate is calculated (step 705). Using the thickness T ₁ before polishing, the thickness T ₂ after polishing, and the actual polishing time TP, the average polishing value PR can be calculated from, for example, PR = (T ₂ = T ₁ ) / PT.

[0090] As described below, an endpoint time is calculated for a specific setup substrate to provide a calibration point for testing the reference spectrum (step 707). The end point time can be calculated based on the calculated polishing value PR, the starting thickness ST of the target film before polishing, and the target thickness TT of the target film. The end point time can be calculated as a simple linear interpolation, for example, as follows, assuming that the polishing value is constant throughout the polishing process: ET = (ST−TT) / PR.

[0091] In some cases, the calculated endpoint time is polishing another substrate in the batch of patterned substrates, stopping polishing at the calculated endpoint time, measuring the thickness of the target film, Can be evaluated. If this thickness is within the acceptable thickness range, the calculated end point time is acceptable. Otherwise, the calculated end point time can be calculated again.

[0092] One of the collected spectra is selected and designated as a reference spectrum (step 709). The selected spectrum is the spectrum of light reflected by the substrate when the thickness of the target film is greater than or approximately equal to the target thickness. Alternatively, two or more spectra are designated as reference spectra. Typically, three collected spectra are designated as reference spectra. Alternatively, five, seven, and nine spectra are designated as reference spectra. In the case of the target spectrum, since the polishing value is finite, there may be a plurality of reference spectra.

[0093] In one implementation, the specific platen rotation corresponding to the end point time calculated in step 707 is identified, and the spectrum collected during the specific platen rotation is selected and designated as the reference spectrum. As an example, the collected spectrum may be from the central region of the substrate. The platen rotation corresponding to the calculated end time is a platen rotation in which a time corresponding to the calculated end time has occurred. As an example, when the calculated end time is 25.5 seconds, the specific platen rotation related to the calculated end time is a platen rotation in which polishing for 25.5 seconds is occurring during the polishing process. It is.

[0094] Specific endpoint determination logic is performed in a simulation using one or more spectra selected to be assigned to one or more reference spectra, using the spectra collected for the configuration board. (Step 711). By executing this logic, an empirically derived and simulated endpoint time determined by this logic to be an endpoint is produced.

[0095] This empirically derived and simulated endpoint time is compared to the calculated endpoint time (step 713). If the empirically derived endpoint time falls within the calculated endpoint time threshold range, the currently selected one or more reference spectra are known to produce a result that matches the calibration point. . Thus, if the endpoint logic is performed using one or more reference spectra within the runtime environment, the system should reliably detect the endpoint at the target thickness. Accordingly, one or more reference spectra can be maintained as a reference spectrum for the polishing time of other substrates in the same batch. (Step 718) If not, steps 709 and 711 are repeated as appropriate.

[0096] Optionally, for each iteration (ie, each time steps 709, 711 are performed), one or more variables other than the selected spectrum can be changed. For example, the threshold range from the above-described spectrum (eg, filter parameter) processing and / or minimal difference trace can be changed. The difference trace and the minimum difference trace threshold range will be described below.

[0097] FIG. 8A shows a method 800 that uses spectrum-based endpoint determination logic to determine the endpoint of a polishing step. Using the polishing apparatus described above, another substrate in the patterned batch of substrates is polished (step 802). The following steps are performed during each rotation of the platen.

[0098] One or more current spectra are obtained for the current platen rotation by measuring one or more spectra of white light reflected at the substrate surface being polished (step 804). Spectral measurements at points 501-511 (FIG. 5) are examples of spectra measured during the current platen rotation. The spectrum measured during the current platen rotation is optional to extend accuracy and / or precision as described above with reference to FIG. 7A and described below with reference to FIG. It is processed.

[0099] In some implementations, if there is only one spectrum to measure, this one spectrum is used as the current spectrum. When measuring two or more spectra per platen rotation, these spectra are grouped, averaged within each group, and this average is designated as the current spectrum. Spectral grouping can be done by radial distance from the center of the substrate. As an example, the first current spectrum is obtained from the spectrum measured at points 502 and 510 (FIG. 5), the second current spectrum is obtained from the spectrum measured at points 503 and 509, and the third current spectrum is obtained. This spectrum is obtained from the spectrum measured at points 504 and 508, and the same can be done thereafter. The spectra measured at points 502 and 510 are averaged to obtain a first current spectrum for the current platen rotation. To obtain the second current spectrum of the current platen rotation, the spectrum measured at points 503 and 509 is averaged. To obtain the third current spectrum of the current platen rotation, the spectrum measured at points 504, 508 is averaged.

[00100] In some implementations, two or more spectra measured during the current platen rotation are selected as the current spectrum of this current platen rotation. In one implementation, the spectrum selected as the current spectrum is measured at a sample location near the center of the substrate (eg, at points 505, 506, 507 shown in FIG. 5). The selected spectra are not averaged, and each selected spectrum is designated as the current spectrum of the current platen rotation.

[0101] A difference between each current spectrum or a plurality of current spectra and each reference spectrum is calculated (step 806). One or more reference spectra can be obtained in the manner described with reference to FIG. 7B. In one implementation, this difference is the sum of the intensity differences over a range of wavelengths. That is, the following equation is obtained:

Where a and b are the lower and upper limits of the spectral wavelength range, respectively, and I _current (λ) and I _reference (λ) are the current spectral intensity and target spectral intensity for a given wavelength, respectively. is there.

[0102] One way to calculate the difference between each current spectrum and each reference spectrum is to select each current spectrum. For each selected current spectrum, the difference from each reference spectrum is calculated. For example, assuming a current spectrum e, f, g and a reference spectrum E, F, G, the difference is calculated for each of the following combinations of current spectrum and reference spectrum: e and E, e and F , E and G, f and E, f and F, f and G, g and E, g and F, g and G.

[0103] In some implementations, each calculated difference is added to the difference trace (step 808). In general, a difference trace is a plot of calculated differences. The difference trace is updated at least once per platen rotation. (When multiple current spectra are obtained for each platen rotation, the difference trace is updated multiple times per platen rotation.

[0104] Generally, the difference trace is a plot of one of the calculated differences (in this case, the plot of the smallest of the differences calculated for the current platen rotation). Instead of the minimum difference, another difference, for example, an intermediate value of a plurality of differences or a difference next to the minimum difference can be added to the trace.

[0105] Taking the smallest difference improves the accuracy of the endpoint determination process. Current spectra may include spectra from light reflected at different locations on the substrate (eg, scribe lines or arrays), and the pattern effects described above can greatly increase the difference between these spectra. can do. Similarly, the reference spectrum can include spectra from light reflected at different locations on the substrate. The comparison of these different spectra is flawed and can make the end point determination incorrect. For example, comparing the current spectrum of light reflected by a scribe line on a patterned substrate with a reference spectrum of light reflected by an array of patterned substrates results in an error in the endpoint determination calculation. there is a possibility. Figuratively speaking, such a comparison is like comparing apples and oranges. If only the smallest difference is considered, these types of comparisons are excluded from the calculation (even if they are done). Therefore, by using a plurality of reference spectra and a plurality of current spectra, and taking into consideration the minimum difference between these spectra, errors caused by the erroneous comparison described above can be avoided.

In some cases, for example, the difference trace can be simplified by removing a calculated difference that deviates beyond a threshold value from one or more preceding calculated differences and simplifying the difference trace.

[0107] It is determined whether the difference trace falls within a minimum threshold (step 810). When the minimum value is detected, the endpoint is called when the differential trace starts to rise above a specific minimum threshold. Alternatively, the end point can be called based on the slope of the differential trace. Specifically, the slope of the differential trace approaches zero and becomes zero when the minimum value is within the slope threshold range near zero.

[0108] Optionally, window logic may be applied to facilitate decision step 808. Suitable window logic is that described in co-assigned US Pat. Nos. 5,893,796, 6,296,548, the entirety of which is incorporated herein by reference. It is.

[0109] If it is not determined that the differential trace has reached the minimum threshold range, polishing can continue and steps 804, 806, 808, 810 are repeated as appropriate. Otherwise, the end point is called and polishing is stopped (step 812).

[0110] FIG. 8B illustrates a method of determining the end point described above. Trace 801 is a raw difference trace. Trace 803 is a simplified differential trace. When this simplified difference trace 803 reaches a threshold 805 above the minimum value 807, the end point is called.

[0111] As an alternative to the use of one or more reference spectra, one or more target spectra may be used in method 800. The difference calculation is between one or more current spectra and one or more target spectra, and the end point is determined when the difference trace reaches a minimum.

[0112] FIG. 9A illustrates an alternative method 900 for determining the endpoint of a polishing step using spectrum-based endpoint determination logic. The setting substrate is polished to obtain one or more target spectra and one or more reference spectra (step 902). These spectra can be obtained as described above with reference to FIGS. 7A and 7B.

[0113] A target difference is calculated (step 904). When using one reference spectrum, the target difference is the difference between the reference spectrum and the target spectrum, which can be calculated using the difference equation described above. If more than one reference spectrum is used, the target difference is the minimum difference between the reference spectrum and the target spectrum, which is calculated using the difference equation and difference calculation method described above (ie, step 808). .

[0114] Initiate polishing of another substrate in the batch of substrates (step 906). During polishing, the following steps are performed for each platen rotation. One or more spectra of white light reflected by the substrate surface being polished are measured to obtain one or more current spectra for the current platen rotation (step 908). A difference between one or more current spectra and a reference spectrum is calculated (step 910). One or more calculated differences (if there are two or more current spectra) are added to the difference trace (step 912). (Steps 908, 910, and 912 are similar to steps 804, 806, and 808, respectively.) Determine whether the difference trace is within the threshold range of the target difference (step 914). If it is not determined that the difference trace has reached the target difference threshold range, polishing can continue and steps 908, 910, 912, 914 are repeated as appropriate. Otherwise, the end point is called and polishing is stopped (step 916).

[0115] FIG. 9B illustrates the endpoint determination method described above. Trace 901 is a raw differential trace. Trace 903 is a simplified differential trace. When this simplified difference trace 903 falls within the threshold range 905 of the target difference 907, the end point is called.

[0116] FIG. 10A illustrates another method 1000 for determining the end point of a polishing step. One or more reference spectra are obtained (step 1002). The one or more reference spectra are obtained as described above with reference to FIG. 7B.

[0117] The spectra collected from the process of obtaining a reference spectrum are stored in a library (step 1004). Alternatively, the library can contain spectra that have not been collected but were theoretically generated. The plurality of spectra including the reference spectrum are indexed so that each spectrum has a unique index value. This indexing is implemented so that the index values are continuous in the order in which the spectra were measured. As such, the index correlates with time and / or platen rotation. In one implementation, a first spectrum collected at a first point during time has a smaller index value than a second spectrum collected at a later point in time. The library can be implemented in the memory of the computing device of the polishing apparatus.

[0118] One substrate in the substrate batch is polished and the following steps are performed for each platen rotation. One or more spectra are measured to obtain a current spectrum for the current platen rotation (step 1006). A spectrum is obtained as described above. Each current spectrum is compared with the spectrum stored in the library to determine the library spectrum that most closely matches any current spectrum (step 1008). The index of the library spectrum determined to best match any of the current spectra is added to the endpoint index trace (step 1010). When the end point trace reaches the index of any one or more reference spectra, the end point is requested (step 1012).

[0119] FIG. 10B illustrates the above-described method for determining an endpoint. Trace 1004 is a raw index trace. Trace 1003 is a simplified differential trace. Line 1005 represents the index value of the reference spectrum. Multiple current spectra can be obtained each time the optical head is swept under the substrate, for example, the spectrum of each radial zone on the substrate can be tracked and an index trace can be generated for each radial zone .

[0120] FIG. 11 shows an implementation for determining the endpoint during the polishing step. The following steps are performed for each platen rotation. A plurality of raw spectra of white light reflected from the substrate surface being polished is measured (step 1102).

[0121] Each measured raw spectrum is normalized to remove light reflections contributed from media other than the one or more target films (step 1104). Normalizing the spectra facilitates comparison of the spectra. The light reflection contributed from the medium other than the one or more target films includes light reflection from the polishing pad window and light reflection contributed from the base silicon layer of the substrate. The contribution from the window can be estimated by measuring the spectrum of the received light when the in-situ monitoring system is dark (ie, when no substrate is placed over the in-situ monitoring system). The contribution from the silicon layer can be estimated by measuring the light reflection spectrum of the bare silicon substrate. The contribution from the bare silicon substrate can be obtained before the start of the polishing step. However, the contribution from the window, the so-called dark contribution, is obtained dynamically, ie for each platen rotation as point 511 in FIG.

[0122] The measured raw spectrum is normalized as follows.

Normalized spectrum = (A-Dark) / (Si-Dark)
Here, A is a raw spectrum, Dark is a spectrum obtained under dark conditions, and Si is a spectrum obtained from a bare silicon substrate.

[0123] In some cases, the collected spectrum can be stored based on the pattern region that generated the spectrum, and spectra from several regions can be excluded. In particular, the spectrum from the light reflected from the scribe line can be removed from consideration (step 1106). Usually, different regions of the patterned substrate produce different spectra (even if the spectra were acquired at any time point during polishing). For example, the spectrum of light reflected by a scribe line in the substrate is different from the spectrum of light reflected by the array of substrates. Because the spectra differ in shape, the use of spectra from both these areas of the pattern usually leads to errors in determining the endpoint. However, spectra can be stored for each group of scribe lines and groups of arrays based on these shapes. In the case of a scribe line, the change in the spectrum is larger, so this spectrum can usually be excluded from consideration and the precision can be extended.

[0124] Step 1106 may be an alternative to a technique that uses multiple reference spectra (as described above in step 808 of method 800) to correct the above-described errors caused by inappropriate comparisons. Step 1106 can be performed as an alternative to or in addition to step 808.

[0125] So far, a subset of the processed spectrum has been selected and in some cases averaged (step 1108). The subset is made up of spectra obtained from light reflected from the substrate at a plurality of points within a region on the substrate. This region may be, for example, a region 503 or a region 507 (FIG. 5).

[0126] Optionally, a high pass filter is added to the measured raw spectrum (step 1110). Typically, the addition of a high pass filter removes the average low frequency distortion of a subset of the spectrum. The high pass filter can be added to the raw spectrum, this average, or both the raw spectrum and this average.

[0127] The average value is normalized so that its amplitude is the same as or similar to the amplitude of the reference spectrum (step 1112). The amplitude of the spectrum is the peak of the spectrum trough value. Alternatively, the average value is normalized so that its reference spectrum is the same as or similar to the reference amplitude. In this case, this reference spectrum is normalized to this reference amplitude. In some implementations, each spectrum in the subset of spectra is normalized so that their amplitudes are the same as or similar to those of the reference spectrum, or the reference spectrum is also normalized. Normalized to be the same as or similar to a certain reference amplitude.

[0128] The difference between the normalized average or spectrum and the reference spectrum is calculated (step 1114). As described with reference to FIG. 7B, one or more reference spectra are obtained. The difference is calculated using the above-described equation for calculating the difference between the spectra.

[0129] The difference trace is updated with the current difference or the minimum calculated difference (step 1116). The difference trace presents the calculated difference between the normalized average or spectrum and one or more reference spectra (or platen rotations) as a function of time.

[0130] An intermediate filter and a low-pass filter are added to the updated difference trace (step 1118). Typically, adding these filters simplifies the trace (by reducing or eliminating spikes in the trace).

[0131] An endpoint determination is performed based on the updated and filtered difference trace (step 1120). This determination is made based on when the differential trace reaches a minimum value. The window logic described above is used to make this determination.

[0132] More generally, the signal processing steps of steps 1104-1112 can be used to improve the endpoint determination procedure. For example, instead of generating a difference trace, a normalized average spectrum is used to select a spectrum from which to generate an index trace from the library, as described above with reference to FIG. 10A.

FIG. 12 shows the normalization of step 1112. As can be seen, only a part of the spectrum (or spectrum average) is considered for normalization. In this specification, this portion considered for normalization is called a normalization range, and can be selected by the user. Normalization is performed such that the highest and lowest points within the normalization range are normalized to 1 to 0, respectively. Normalization can be calculated as follows:
g = (1-0) / (r _max −r _min )
h = 1-r _max ^* g
N = Rg + h
Here, g is a gain, h is an offset, r _max is a maximum value within the normalization range, r _min is a minimum value within the normalization range, N is a spectrum after normalization, and R is a spectrum before normalization. .

[0134] FIG. 13 shows that multiple criteria are used to provide a smooth effect using a spectrum. Trace 1302 was generated using a single reference spectrum (average value). Trace 1304 was generated using three reference spectra (as described above with reference to FIG. 8). Trace 1306 was generated using nine reference spectra. As can be seen, the number of spikes included in trace 1304 is less than the number of spikes included in trace 1302. That is, the trace 1304 is simpler than the trace 1302. In addition, trace 1304 has a clearer drop than trace 1306, which is important because it allows the endpoint determination logic to call the endpoint. A clearer descent will facilitate end-point determination.

[0135] FIG. 14 illustrates a method 1200 that uses spectra to achieve a desired substrate profile. The end point time predicted for polishing the product substrate is determined (step 1210). In some implementations, this expected endpoint time is determined by polishing the setting substrate according to predetermined processing parameters, determining the time for the setting substrate to reach a desired thickness (eg, using conventional off-line method measurements), and setting This is determined by using the polishing time for the substrate to reach the desired thickness as the expected end point time.

[0136] Product substrate polishing is started (step 1218). Spectra are obtained at two or more radial positions on the two substrates (step 1226). For each spectrum measurement, the radial position on the substrate can be determined and the spectrum measurements can be placed in the zone based on these radial points. The substrate can have a plurality of zones, such as a central zone, an intermediate zone, and an edge zone. On a 300 mm wafer, the central zone may span a radius of 50 mm from the center, the intermediate zone may have a radius of 50 to about 100 mm, and the marginal zone may range from about 100 to 150 mm. In some implementations, the substrate has more or fewer zones than these three. The location where the spectrum is obtained is the method described in US application 10 / 922,110 filed August 18, 2004, “Determination of Position of Sensor Measurements Polishing”, or US Pat. No. 7,018,271. Can be determined using These are hereby incorporated by reference for all purposes.

[0137] As described above in connection with FIG. 10A, the spectrum from each zone (or the spectral average within a zone obtained with a single sweep of the sensor across the substrate for each zone) is defined as the spectrum in the spectral library. Compare. A corresponding index number is determined for each zone from the comparison in the spectral library (step 1238).

[0138] Polishing stops when the zone index matches one or more endpoint criteria. For example, polishing can be stopped when a preselected zone reaches the desired index, or when any zone first reaches the desired index, or when the desired index is achieved in each zone. (Step 1244). The desired index for each zone is ultimately determined by the desired substrate geometry. When the substrate after polishing is flattened or provided with a uniform oxide film, the spectra obtained in all zones must be the same or nearly the same, and all zones are the same or Has a similar desired index number.

[0139] The polishing rate in the zone can be adjusted using a feedback loop so that the final index number in all zones is equal to the desired final index number. FIG. 15 illustrates one method 1400 for adjusting the polishing process to obtain a desired substrate profile at an expected end point time. The desired number of indices at the expected end time is determined for each zone on the substrate (step 1402). Polishing begins (step 1404) and the substrate is optically monitored as described above to determine index traces for all zones on the substrate (step 1406). After the initial delay time that stabilizes the polishing process, the rate of change of the index according to time is calculated (the platen speed can be used to represent time) (step 1408). The index change rate can simply be calculated as the difference between the indexes at two different times divided by the number of platen revolutions that passed between the spectral measurements that generated the index at different times. The change rate of the index number represents the polishing rate. Typically, if there is no change in any polishing parameter, it can be assumed that the polishing rate is constant.

[0140] By extrapolating the index trace using the rate of change of the index for each zone, the number of indexes achieved at the expected endpoint time for the associated zone can be determined (step 1412). If the desired index number is exceeded or not yet reached at the expected end time, the polishing rate can be adjusted up or down as needed (step 1420). If the desired number of indexes is reached at the expected end time, no adjustment is necessary. The determination of whether two or more extrapolations and adjustments should be made can occur over the polishing sequence. The determination of whether or not the polishing rate needs to be adjusted includes determining whether the desired index number is achieved when the polishing endpoint occurs and that the final index is within an acceptable range from the desired final index number. And whether or not to enter.

[0141] In some implementations, an expected endpoint time is determined for a zone, such as a central zone. Then, by adjusting the polishing rate in this other zone as necessary, the desired endpoints of these zones are achieved simultaneously with the expected endpoint time for the selected zone, such as the central zone. The polishing rate can be adjusted, for example, by increasing or decreasing the pressure in the corresponding zone inside the carrier head. In some carrier heads, such as the carrier head described in US Publication No. 2005-0211377, for example, the carrier head has an adjustable pressure zone. It can be assumed that the change in polishing rate is directly proportional to the change in pressure, for example, a simple Prestonian model. In addition to this, the effect of the platen or the head rotation speed, the secondary effect of the combination of different head pressures, the polishing temperature, the slurry flow, and other parameters that affect the polishing speed are considered for polishing the substrate. It is possible to develop a control model.

[0142] Further, the endpoint determination logic based on the spectrum described above in method 800 can be used to determine the polishing endpoint, and this endpoint determination logic can be used in conjunction with the polishing process to provide a desired substrate profile. It can also be obtained. The relative thickness of each zone is determined from the equations provided above in connection with step 806 using the differences between the zones. A spectrum is obtained when the substrate is polished, and this spectrum is placed in the zone. Optionally, signal processing and filtering is added to the spectrum. A square error sum calculation is applied to the spectrum collected for each zone and the predetermined reference spectrum. The predetermined reference spectrum is a spectrum obtained when the polishing end point is reached.

[0143] When the sum of squared errors with a reference spectrum approaches the minimum value in one zone, the polishing pressures of the other zones are determined to determine whether to change for the polishing rate in any zone. Inspection is performed. The polishing rate in the zone where the sum of squared errors is approaching the minimum value can be reduced and the polishing rate in the other zones can be increased. It is also possible to analyze the sum of square errors over the entire polishing, and thereby change the adjustment of the polishing rate at the beginning of the polishing sequence. Unlike the method described in method 1400, this method does not require a correlation between the polished spectrum from the spectral library and the index number.

[0144] Referring to FIG. 16, if a specific contour is desired, such as having a uniform thickness over the entire substrate surface, the slope of the polishing rate represented by the change in index number over time is shown. It is possible to monitor and further adjust the polishing rate. After the polishing stabilization period 1505, spectra are obtained in the central zone 1510, the marginal zone 1515, and the intermediate zone 1520 between them. Here, the zone is a circular or annular zone. Each spectrum is correlated with a corresponding index. This process is repeated for the number of platen rotations or over time, and the polishing rate in each of the center zone 1510, the intermediate zone 1520, and the edge zone 1515 is determined. The polishing rate is represented by the slope of a line obtained by plotting the index 1530 (y axis) according to the number of revolutions 1535 (x axis). If either speed is faster or slower than the others, the speed in this zone can be adjusted. Here, this adjustment is based on the end point _CE of the central zone 1510. After sufficient data points or platen times have been collected for the current substrate, an approximate polishing endpoint (“EDP”) or estimated endpoint time (“EET”) is determined in the central zone. The EET is recalculated after each platen rotation. The first polishing time T ₁ of the during the polishing process, the polishing speed in the intermediate zone 1520 is decelerated, the polishing rate at the edge zones is increased. To be adjusted polishing rate in the middle zone 1520, the intermediate zone faster than other zones of the substrate, it will be polished with an excess polishing speed M _A speed. To be adjusted polishing rate _{T 1} of the edge zone 1515, the polishing of the edge zone 1515 becomes insufficient polishing at a rate _{E u.}

[0145] Even at a subsequent time (T ₂ ) during the polishing process, the speed can be adjusted if necessary. The goal in this polishing process is to finish polishing when the surface of the substrate becomes flat or when it is an oxide layer over the entire relatively flat surface. One method of determining the amount of adjustment of the polishing speed is to adjust the speed so that the indices of the center, middle, and edge zones are substantially equal at the polishing end point EDP. Therefore, the polishing rate in the edge zone it is necessary to adjust the center and intermediate zones are polished at the same speed as the T ₂ previously. The EDP is determined when the approximate straight line of the central zone reaches the desired ASL level. All other zones are controlled so that these approximate lines are as close as possible to simultaneously cross the ASL level.

[0146] Another method of using observation based on the polishing rate spectrum to achieve a specific profile is to polish the first substrate, monitor the polishing rate, and then against the subsequently polished substrate. There is one that supplies polishing speed information. Referring to FIG. 17, the first setting substrate is polished, and a spectrum that determines the polishing rate and relative oxide thickness of the center zone 1610, the intermediate zone 1620, and the edge zone 1630 is obtained. The start indexes of the intermediate zone 1620, the center zone 1610, and the edge zone 1630 are represented by M _O , C _O , and E _O , respectively. The central zone 1610 has an endpoint _CE spectrum that is selected for the target spectrum. At the end of polishing the current wafer, if the other two zones have an index number that falls within the threshold distance 1640 from the index of the center endpoint _CE , then the next wafer edge zone 1630 or intermediate zone 1620 There is no need to adjust the polishing rate. Similarly, even when the polishing rate and the number of indexes during polishing are within the allowable range 1650, it is not necessary to adjust the edge zone 1630 or the intermediate zone 1620. Here, the end of the polishing end point of the intermediate zone ME indicates that this intermediate zone is excessively polished, ending the marginal zone E _E is a show that this edge zone it is insufficient polishing. As a result, in subsequent wafers, the polishing rate parameter is adjusted so that the thicknesses at the end points of all zones E, M, and C are within the allowable range 1640. FIG. 18 represents the expected signal generation in zones E, M, C of the next wafer.

[0147] Referring to FIG. 18, during the polishing process when in-situ adjustment is used, the polishing rate is only changed several times, such as four times, three times, two times or only once. Is preferred. The adjustment can be performed near the start of the polishing process, in the middle, and near the end. In relation to the indexed spectrum, a linear comparison of the polishing in each zone is created and the calculations necessary to determine how to control the polishing process and eliminate the need for complex software or processing steps. Simplified.

[0148] Although the spectrum obtained from the difference zone of the substrate can yield a substrate profile, it is not necessary to provide the absolute thickness of the oxide layer. As such, the polishing rate adjustment method based on several spectra described herein can be used to monitor the relative oxide thickness across the substrate. Since these spectral-based methods can be used to determine and adjust the polishing rate within the zone of the substrate, the spectral-based method induces subsequent substrate thickness changes and wafer non-uniformity. The polished polishing can be corrected.

[0149] As described herein, relative thickness can be used to achieve a desired substrate profile. In some of the examples described above, the desired substrate profile after polishing is a flat profile. However, it is also possible to achieve contours other than flat. In many cases, the substrate is polished on two or more platens. Several polishing processes are known that inherently polish one zone at a higher speed than the other zones. In order to correct this non-uniform polishing, the polishing on the first platen can be controlled, for example, to leave one zone that is polished faster on the next platen thicker than the other zones. This difference in thickness can be achieved by selecting a difference in the number of target indexes or a ratio between the number of end indexes in one zone and the number of end indexes in another zone.

[0150] Embodiments of the present invention, and all functional operations described herein, have been disclosed within digital electronic circuit elements, or in the present specification and its structural equivalents, or combinations thereof. It can be implemented in computer software, firmware, hardware, including structural means. Embodiments of the present invention are implemented as one or more computer program products, i.e., for example, by a programmable processor, computer, multiple processors or a data processing device such as a computer, or to control the operation thereof. It can be realized as one or more computer programs that are actually embodied in an information carrier or propagation signal, such as a machine readable storage device. Computer programs (also known as programs, software, software applications, and codes) can be written in any programming language format, including compiled and translated languages, plus components, subroutines, and other computations It can be deployed in any form including units suitable for use in the environment as stand-alone programs or modules. A computer program is not necessarily associated with a file. A program is devoted to a program within another program or part of a file that holds data, a question or coordinated files (eg, a file that stores one or more modules, subprograms, parts of code) Can be stored in a single file. The computer program can be expanded so as to be executed on one computer, or on several computers installed in one place or distributed in a plurality of places and connected via a communication network.

[0151] The processing flow and logic flow described herein are one or more programs that execute one or more computer programs that perform functions by inputting data and generating output. It can be realized by a possible processor. The process flow and logic flow can be implemented by special purpose logic circuits such as, for example, FPGA (Field Programmable Gate Array) or ASIC (Application Specific Integrated Circuit), and the device can be further configured with such special purpose logic circuits. It is also possible to realize.

[0152] The polishing apparatus and method described above can be applied to various polishing systems. The polishing pad and / or carrier head can be moved to provide relative movement between the polishing surface and the substrate. For example, the platen can orbit instead of rotating. The polishing pad may be a circular (or other shape) pad fixed to the platen. Some aspects of the endpoint detection system can be applied, for example, to a linear polishing system in which the polishing pad is continuous or the reel-to-reel belt moves in a linear direction. The abrasive layer may be a standard (eg, polyurethane with or without filler) abrasive material, a flexible material, a fixed abrasive material. It should be understood that a relative positioning scheme can be used: the polishing surface and the substrate can be held in a vertical or other orientation.

[0153] Specific embodiments of the invention have been described. Other embodiments are within the scope of the claims. For example, it is possible to achieve a desired result even if the operations cited in the claims are performed in a different order.

[0022]
1 shows a chemical mechanical polishing apparatus. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Figure 5 illustrates a polishing pad window implementation. Fig. 2 shows an implementation of an injection system. Fig. 4 shows an alternative implementation of an injection system. FIG. 3 is an overhead view of a polishing pad, showing where in-situ measurements are taken. The spectrum obtained from in situ measurements is shown. Figure 8 illustrates the rotation of the spectrum obtained from in situ measurements as the polishing progresses. A method for obtaining a target spectrum will be described. A method for obtaining a reference spectrum is shown. The end point determination method is shown. The end point determination method is shown. An alternative method of end point determination is shown. An alternative method of end point determination is shown. Fig. 4 shows another alternative method of determining the end point. Fig. 4 shows another alternative method of determining the end point. Indicates the implementation of end point determination. Figure 8 illustrates the maximum / minimum normalization of the spectrum. Fig. 6 illustrates a simplification effect using provision of multiple reference spectra. A method for obtaining a spectrum within a zone during polishing is shown. A method of adjusting the polishing rate in the zone to achieve the desired profile is shown. The graph which represented the process which adjusts a grinding | polishing speed in grinding | polishing progress versus time is shown. The graph which represented the process which does not adjust a grinding | polishing speed in grinding | polishing progress versus time is shown. Fig. 6 shows a graph representing the process using the advance method of polishing control in polishing progress versus time.

DESCRIPTION OF SYMBOLS 10 ... Substrate, 20 ... Polishing device, 24 ... Platen, 25 ... Shaft, 30 ... Polishing pad, 32 ... External polishing layer, 34 ... Flexible backing layer, 38 ... Slurry, 51 ... Light source, 52 ... Photo detector 53 ... Optical head, 70 ... Carrier head, 72 ... Support structure, 202 ... Polyurethane part, 204 ... Quartz part, 206 ... Top surface, 208 ... Polishing surface, 210 ... Bottom surface, 302 ... Gas source, 304 ... Conveyance line, 306 ... Conveying nozzle, 308 ... Suction nozzle, 310 ... Vacuum line, 312 ... Vacuum source, 314 ... Transparent top surface, 402 ... Gas source, 404 ... Conveying line, 406 ... Conveying nozzle, 408 ... Suction nozzle, 410 ... Vacuum Line, 412 ... Vacuum source, 501, 502, 503, 504, 505, 506, 507, 508, 509, 510, 511 ... Point, 602, 604, 606 ... Spare Toll, 801, 803 ... Trace, 805 ... Threshold, 807 ... Minimum value, 901, 903 ... Trace, 905 ... Threshold range, 907 ... Target difference, 1003 ... Trace, 1005 ... Line, 1302, 1304, 1306 ... Trace, 1505 ... Stable polishing period, 1510 ... Center zone, 1515 ... Edge zone, 1520 ... Intermediate zone, 1530 ... Index, 1610 ... Center zone, 1520 ... Intermediate zone, 1630 ... Edge zone, 1640 ... Allowable range, 1650 ... Tolerance

Claims (100)

A computer-implemented method comprising:
Selecting a reference spectrum, wherein the reference spectrum is a spectrum of white light on a first substrate and reflected by a target film thicker than a target thickness, and the reference spectrum is based on a specific spectrum For endpoint determination logic, empirically selected such that the target thickness is achieved when the endpoint is invoked by applying endpoint logic based on the specific spectrum;
Obtaining a current spectrum, wherein the current spectrum is a spectrum of white light on a second substrate and reflected by a target film having a current thickness greater than the target thickness; A step of polishing the target film on two substrates;
Determining, for the second substrate, whether an end point of the polishing step has been reached, wherein the determination comprises a step based on the reference spectrum and the current spectrum. Method. Determining whether an end point of the polishing step has been reached,
Calculating a difference between the reference spectrum and the current spectrum;
Calling the end point when the difference reaches a threshold;
The method of claim 1 comprising: A current spectrum is obtained for each rotation of the platen used in the polishing step,
Determining whether an end point of the polishing step has been reached includes generating a differential trace that includes a plurality of points, each of the plurality of points being obtained per rotation of the platen. The method of claim 2, wherein the method represents a calculated difference between the current spectrum and the current spectrum. 4. The method of claim 3, wherein determining whether an end point of the polishing step has been reached comprises determining whether the difference trace has reached a minimum value when a polishing step has been achieved. The method of claim 4, wherein determining whether the endpoint of the polishing step has been reached comprises determining whether the differential trace has risen to a threshold above the minimum value. The method of claim 4, wherein determining whether the difference trace has reached a minimum includes calculating a slope of the difference trace. The method of claim 3, further comprising adding a filter to simplify the difference trace. Each of the current spectrum and the reference spectrum exhibits maximum / minimum amplification;
The method of claim 1, wherein the current spectrum is normalized such that the maximum / minimum amplification of the current spectrum is the same or similar to the maximum / minimum amplification of the reference spectrum. The method of claim 1, wherein each of the current spectrum and the reference spectrum is normalized to remove light reflections contributed by media other than the target film. The method of claim 1, wherein obtaining a current spectrum comprises obtaining a current spectrum that is not from light reflected by a scribe line of the second substrate. A computer program product stored on a machine-readable medium, the product comprising instructions that cause a processor to perform the following operations:
An operation of selecting a reference spectrum is included, wherein the reference spectrum is a spectrum of white light on a first substrate and reflected by a target film thicker than a target thickness, and the reference spectrum is a specific spectrum. For the endpoint determination logic based, it is chosen empirically such that the target thickness is achieved when the endpoint is invoked by applying the specific spectrum based endpoint logic,
An operation of obtaining a current spectrum is further included, wherein the current spectrum is a spectrum of white light on a second substrate and reflected by a target film having a current thickness greater than a target thickness; The target film on the two substrates is subjected to a polishing step,
A computer product, further comprising, for the second substrate, determining whether the end of the polishing step has been reached, the determining step being based on the reference spectrum and the current spectrum. An instruction to determine whether the end point of the polishing step has been reached,
Instructions to calculate the difference between the reference spectrum and the current spectrum;
The computer program product of claim 11, comprising: an instruction to call the end point when the difference reaches a threshold value. Instructions for obtaining a current spectrum include instructions for obtaining a current spectrum for each rotation of the platen used in the polishing step;
The instruction to determine whether the end point of the polishing step has been reached includes an instruction to generate a differential trace that includes a plurality of points, each of the plurality of points being each revolution of the reference spectrum and the platen. The computer program product of claim 12, wherein the computer program product represents a calculated difference between the current spectrum obtained from The computer program product of claim 13, wherein instructions for determining whether an end point of the polishing step has been reached include instructions for determining whether the differential trace has reached a minimum value. The computer program product of claim 14, wherein the instructions for determining whether an end point of the polishing step has been reached comprises instructions for determining whether the differential trace has risen to a threshold above the minimum value. . The computer program product of claim 14, wherein the instructions for determining whether the difference trace has reached a minimum include instructions for calculating a slope of the difference trace. The computer program product of claim 13, further comprising instructions for adding a filter to simplify the trace. Each of the current spectrum and the reference spectrum exhibits maximum / minimum amplification;
The product further comprises instructions for normalizing the current spectrum such that maximum / minimum amplification of the current spectrum is the same as or similar to maximum / minimum amplification of the reference spectrum. The computer program product described. The computer program product of claim 11, further comprising instructions for normalizing each of the current spectrum and the reference spectrum to remove light reflections contributed by media other than the target film. 12. The computer program product of claim 11, wherein the instructions for obtaining a current spectrum include instructions for obtaining a current spectrum that is not from light reflected by a scribe line of the second substrate. An ejection system for ejecting a top surface of an optical head,
A gas source configured to provide a gas flow;
A transfer nozzle;
A transport line connecting the gas source to the transport nozzle;
A vacuum source configured to provide a vacuum;
A vacuum nozzle;
A vacuum line connecting the vacuum source to the vacuum nozzle, and the gas source and the transport nozzle are configured to guide a gas flow over the top surface of the optical head, and the vacuum nozzle An injection system, wherein the vacuum source is configured so that the gas flow is stratified. The jetting system according to claim 21, wherein slurry is prevented from drying on the top surface of the optical head. The optical head guides light in a first direction;
The injection system according to claim 21, wherein the transfer nozzle is installed to guide a gas flow in a plane substantially perpendicular to the first direction. The injection system of claim 21, wherein the gas flow is configured to provide a specific amount of gas and the vacuum source is configured to draw the same or similar amount of gas. The injection system of claim 21, wherein the gas is clean and dry air or nitrogen gas. A first rotating coupler for propagating gas, wherein the first rotating coupler is installed in the transfer line;
The injection system according to claim 21, wherein the injection system is a second rotary coupler for propagating gas, wherein the second rotary coupler is installed in the vacuum line. An injection system for injecting the bottom surface of a polishing pad window, the system comprising:
A gas source configured to provide a gas flow;
A transfer nozzle;
A transfer line connecting the gas source to the transfer nozzle;
The gas source and the transfer nozzle are configured to guide a gas flow at the bottom of the polishing pad window, and prevent the formation of condensation on the bottom surface of the polishing pad window. system. A vacuum source configured to provide a vacuum;
A vacuum nozzle;
28. The ejection system of claim 27, further comprising a vacuum line connecting the vacuum source with the vacuum nozzle. 30. The injection system of claim 28, wherein the gas source is configured to provide a specific amount of gas and the vacuum source is configured to draw the same or similar amount of gas. 28. The injection system of claim 27, further comprising a vent configured to allow gas introduced by the gas source to escape. 32. The injection system of claim 30, wherein the vent includes an outlet located at a moisture resistant location. 28. A jetting system according to claim 27, wherein the gas is clean and dry air or nitrogen.
gas. 28. An injection system according to claim 27, further comprising a rotary coupler for transporting gas, wherein the coupler is installed in the transport line. A method of injecting the top surface of an optical head installed on the bottom of a recess in a platen, the method comprising:
Guiding a gas flow introduced at a specific speed over the top surface of the optical head,
The method further comprises the step of removing the gas from the recess at a specific rate, wherein the gas is stratified by a specific rate at which the gas is introduced and a specific rate at which the gas is removed. . 35. The method of claim 34, wherein the recess includes an opening and the gas flow is not guided in the opening. 35. The method of claim 34, wherein guiding the gas stream comprises guiding a flow of either clean, dry air or nitrogen gas. The platen rotates relative to the gas source;
35. The method of claim 34, wherein the step of directing the gas flow comprises transporting gas through a gas transport rotary coupler. A method of spraying a bottom surface of a polishing pad window installed in a platen space, the method comprising:
Introducing a gas into the space, the gas being guided at a bottom surface of the polishing pad window;
The method further comprising venting gas from the space, wherein condensation is prevented from forming on the bottom surface of the polishing pad window. 40. The method of claim 38, wherein the venting step comprises venting to a moisture resistant location. The platen rotates relative to the gas source;
39. The method of claim 38, wherein introducing the gas comprises delivering the gas through a gas delivery rotary coupler. An assembly for chemical mechanical polishing comprising:
A polishing pad having a polishing surface;
A solid window installed in the polishing pad to provide optical access through the polishing pad, the solid window being made from quartz and a first part made from polyurethane. An assembly including a second portion, the first portion having a surface that is coplanar with a polishing surface of the polishing pad. The polishing pad has an expected lifetime,
42. The first portion of the window of claim 41, wherein the first portion of the window has a thickness that does not fatigue the first portion and expose the second portion during an expected lifetime of the polishing pad. assembly. 42. The first portion is a first layer of the solid window, the second portion is a second layer of the solid window, and each layer has a side surface attached to the polishing pad with an adhesive. The assembly described in. 42. The assembly of claim 41, wherein the first portion and the second portion abut but are not attached to each other using an adhesive. 42. The assembly of claim 41, wherein the first portion and the second portion are attached using an adhesive that is transparent to white light. A fiber cable installed to guide the light beam along a path through an optical access provided by the solid window;
The first portion has a non-uniform thickness, so that the thickness of the first portion at a location within the light beam path is the first portion at a location not within the light beam path. 42. The assembly of claim 41, wherein the assembly is thinner than the thickness of the portion. 42. The assembly of claim 41, wherein the first portion of the solid window is transparent to white light. 42. The assembly of claim 41, wherein the solid window is about 80% transparent to a specific color of monochromatic light. A fiber cable installed to guide a light beam along a path through the optical access provided by the solid window;
42. The assembly of claim 41, wherein the first portion and the second portion are attached using an adhesive such that the adhesive does not enter the path of the light beam. 50. The assembly of claim 49, wherein the adhesive is impermeable to light from the fiber cable. The solid window includes a bottom surface, and the assembly further comprises:
50. The assembly of claim 49, comprising a refractive index gel added to a bottom surface of the window, the gel providing a medium for light to travel from the fiber cable to the window. 42. The assembly of claim 41, wherein the polishing pad includes a hole and a first portion that is inserted into the hole, the first portion having a recess in which the second portion is installed. A polishing pad for chemical mechanical polishing,
A polishing layer having a top surface and a bottom surface, wherein the polishing pad includes an aperture having a first opening provided in the top surface and a second opening provided in the bottom surface, and the top surface is a polishing surface ,
A window including a first portion made of a flexible plastic and a second portion made of crystalline or glass; wherein the window is transparent to white light; the window is in the aperture. A polishing pad inserted and installed in the aperture such that the second part is located on the bottom side of the first part, the first part functioning as a slurry-resistant barrier. The polishing pad has an expected lifetime,
54. The first portion of the window has a thickness such that the first portion does not fatigue and the second portion is not exposed during an expected lifetime of the polishing pad. Polishing pad. 54. The first portion is a first layer of the solid window, the second portion is a second layer of the solid window, and each layer has a side surface attached to the polishing pad by an adhesive. The polishing pad described in 1. 54. The polishing pad of claim 53, wherein the first and second portions are in contact but are not attached to each other using an adhesive. 54. The polishing pad of claim 53, wherein the first and second portions are attached using an adhesive that allows white light to pass through. A method of creating a polishing pad,
Placing a mass of crystalline or glassy material in a mold of a polishing pad window, said mass passing white light;
Discharging a flexible plastic material liquid precursor into the mold, the flexible plastic material passing white light; and
Curing the liquid precursor to form a window including a first portion made of a flexible plastic material and a second portion of crystalline or vitreous;
Installing the window in a mold of a polishing pad;
Discharging a liquid precursor of a polishing pad material into a mold of the polishing pad;
Curing the liquid precursor of the polishing pad material to produce the polishing pad;
The window is installed in the polishing pad mold so that when the polishing pad is manufactured, the window is in the polishing pad such that the first portion functions as a slurry barrier. A method that is installed. A method of creating a polishing pad,
Placing a crystalline or glassy material in a mold of a polishing pad window, the mass passing white light;
Discharging a liquid precursor of a flexible plastic material into the mold, the flexible plastic material passing white light;
Curing the liquid precursor to form a window including a first portion made of a flexible plastic material and a second portion of crystalline or vitreous;
Forming a polishing layer including an aperture, the polishing layer having a top surface and a bottom surface, the aperture having a first opening provided in the top surface and a second opening provided in the bottom surface; And wherein the top surface is a polished surface;
Inserting the window into the aperture, wherein the window is inserted into the aperture such that a first portion is inserted into the aperture and the second portion is located on a bottom side of the first portion. The first portion functioning as a slurry barrier. A method of creating a polishing pad,
Forming a first portion of a polishing pad window, wherein the first portion has a recess and allows white light to pass through;
Inserting a mass of crystalline or glassy material into the recess, the mass passing white light;
An aperture, a polishing layer having a top surface and a bottom surface, a step in which the aperture has a first opening provided in the top surface, and a second opening provided in the bottom surface, wherein the top surface is a polishing surface;
Inserting the window into the aperture such that the window is in the aperture, the first portion is inserted into the aperture, and the second portion is on a bottom side of the first portion. And the first part functions as a slurry barrier. A computer-implemented method comprising:
Obtaining a series of current spectra from the substrate being polished;
Calculating a difference between a predetermined spectrum and the current spectrum for each in the series of current spectra to generate a difference trace;
Determining a polishing endpoint using the differential trace;
With a method. 62. The method of claim 61, wherein determining the polishing endpoint comprises determining whether the difference trace has reached a minimum value. 64. The method of claim 62, wherein determining whether the difference trace has reached a minimum includes calculating a slope of the difference trace. A computer program product actually stored on a machine-readable medium, the product comprising instructions for causing a processor to perform the following operations, wherein the operations generate a series of current spectra from a substrate being polished: Receive operation,
Calculating a difference between a predetermined spectrum and the current spectrum for each of the series of current spectra to generate a difference trace;
A product, the operation of determining a polishing end point using the differential trace. The product of claim 64, wherein determining the polishing endpoint comprises determining whether the differential trace has reached a minimum value. 66. The product of claim 65, wherein determining whether the difference trace has reached a minimum comprises calculating a slope of the difference trace. A computer-implemented method comprising:
Obtaining a series of current spectra from the substrate being polished;
Storing a plurality of predetermined spectra and a plurality of index values, wherein each spectrum of the plurality of predetermined spectra is associated with a unique index value representing a polishing power;
Creating a series of index values for each current spectrum in the sequence by selecting a specific spectrum from the plurality of predetermined spectra and determining an index value for the specific spectrum;
Determining a polishing endpoint using the series of index values;
A method comprising: 68. Selecting a specific spectrum from the plurality of predetermined spectra comprises determining which spectrum in the plurality of predetermined spectra most closely matches the current spectrum. The method described. 68. The method of claim 67, wherein the index value is an integer. 68. The method of claim 67, wherein the index value represents a number of platen rotations. A computer program product actually stored on a machine readable medium, the product comprising instructions that cause a processor to perform the following operations:
Obtaining a series of current spectra from the substrate being polished;
An operation for storing a plurality of predetermined spectra and a plurality of index values, each of the predetermined spectra being associated with a unique index value representing a polishing power;
For each current spectrum in the sequence, selecting a specific spectrum from the plurality of predetermined spectra and determining an index value for the specific spectrum to create a series of index values;
A product that is an action that uses a series of index values to determine the polishing endpoint. The act of selecting a specific spectrum from a plurality of predetermined spectra includes determining which spectrum of the plurality of predetermined spectra is closest to the current spectrum. 72. The product of claim 71, wherein the index value is an integer. 72. The product of claim 71, wherein the index value represents a number of platen rotations. A computer-implemented method comprising:
Storing at least one normalized predetermined spectrum;
Obtaining a series of current spectra from the substrate being polished;
Normalizing each of the series of current spectra to generate a series of normalized current spectra;
Using the at least one normalized predetermined spectrum and the normalized current spectrum to determine a polishing endpoint. A computer program product actually stored on a machine readable medium, the product comprising instructions that cause a processor to perform the following operations:
Storing at least one normalized predetermined spectrum;
Obtaining a series of current spectra from the substrate being polished;
Normalizing each current spectrum in the series of current spectra to generate a series of normalized current spectra;
A method that is an act of determining a polishing endpoint using the at least one normalized predetermined spectrum and the normalized current spectrum. A computer-implemented method comprising:
Obtaining a series of current spectra from the substrate being polished;
Determining, for each of the series of current spectra, based on the current spectrum whether the current spectrum corresponds to a region of the substrate;
Determining a polishing endpoint using the current spectrum that was not determined to correspond to the first region;
A method comprising: 78. The method of claim 77, wherein the region is a scribe line. A computer program product that is actually stored on a machine-readable medium, the product comprising instructions that cause a processor to perform the following operations:
Instructions to obtain a series of current spectra from the substrate being polished;
Instructions for determining, based on the current spectrum, for each of the series of current spectra, whether the current spectrum is associated with a region of the substrate;
A product that is an instruction to determine a polishing endpoint using the current spectrum that was not determined to be associated with the first region. 80. The product of claim 79, wherein the region is a scribe line. A computer-implemented method comprising:
Obtaining a first spectrum from light reflected by a first zone on a substrate and a second spectrum from a second zone on the substrate during a polishing sequence;
Comparing the first spectrum and the second spectrum with a spectrum library to determine a first index for the first spectrum and a second index for the second spectrum;
Obtaining a third spectrum of light reflected at the first zone and obtaining a fourth spectrum from the second zone at different times during a polishing sequence;
Comparing the third spectrum and the fourth spectrum with the library to determine a third index for the first zone and a fourth index for the second zone;
Determining a polishing rate in the first zone from the first index and a third index, and determining a polishing rate in the second zone from the second index and the fourth index;
Adjusted polishing of the second zone based on the first polishing rate, the second polishing rate, the first target relative thickness for the first zone, and the second target relative thickness for the second zone Polishing the second zone to the second target relative thickness at substantially the same time that the first zone is polished to the first target relative thickness by determining a speed;
A method comprising: 82. The method of claim 81, wherein the first zone is an inner zone and the second zone is an outer annular zone. 82. The step of determining a polishing rate adjusted to the second zone comprises determining when the first target relative thickness is within a predetermined threshold from the second target relative thickness. The method described. 82. The method of claim 81, wherein determining an adjusted polishing rate for the second spectrum comprises determining an estimated endpoint time for the polishing sequence. 82. The method of claim 81, wherein obtaining the first spectrum and the second spectrum comprises obtaining a white light spectrum. 82. The method of claim 81, further comprising adjusting a parameter of the polishing system to cause the second zone to polish at the adjusted polishing time. 87. The method of claim 86, wherein determining the adjusted polishing rate is performed on a setting substrate, and adjusting the parameters of the polishing system is performed on a product substrate. 87. The method of claim 86, wherein determining the adjusted polishing rate is performed on a product substrate, and adjusting the parameters of the polishing system is performed on the product substrate. 90. The method of claim 86, wherein adjusting the polishing system parameters comprises adjusting pressure. 82. The step of determining an adjusted polishing rate comprises determining a polishing rate that causes a cross-section along the diameter of the substrate to have a flat outer shape when the polishing sequence is completed. Method. 82. The method of claim 81, wherein determining an adjusted polishing rate comprises determining a polishing rate that causes a cross-section along the diameter of the substrate to be bowl-shaped when the polishing sequence is completed. . 84. The method of claim 81, wherein obtaining the first spectrum and the second spectrum comprises sampling the substrate at different rotational locations. 82. The method of claim 81, wherein obtaining the first spectrum and the second spectrum comprises measuring a spectrum reflected by an oxide film. Polishing the setting substrate until it is excessively polished;
Obtaining a plurality of spectra from a zone of the test substrate during the polishing;
84. The method of claim 81, further comprising storing the plurality of spectra with the time each spectrum was obtained to create the spectral library. 85. The method of claim 84, further comprising creating an index in the spectral library, wherein the index represents a spectrum obtained from the configuration substrate at a particular time. A method for monitoring a chemical mechanical polishing process comprising:
Guiding a multi-wavelength light beam onto a substrate being polished and measuring a spectrum of light reflected by the substrate;
Moving the light beam in a path across the substrate surface;
Extracting a series of spectral measurements from the signal and determining a radial position of the substrate for each of the spectral measurements;
Storing the spectral measurements in a plurality of radial ranges according to the radial position;
Determining a polishing endpoint for the substrate from the spectral measurements in at least one of the plurality of radial ranges;
A method comprising: 99. The method of claim 96, further comprising determining a polishing rate adjusted to one of the radial ranges and applying the adjusted polishing rate to one of the radial ranges. A computer program product actually stored on a machine readable medium, the product comprising instructions that cause a processor to perform the following operations:
Obtaining a first spectrum of light reflected by the first zone on the substrate and a second spectrum of light reflected by the second zone on the substrate during the polishing sequence;
Comparing the first spectrum and the second spectrum with a spectrum library to determine a first index for the first spectrum and a second index for the second spectrum;
Obtaining a third spectrum of light reflected by the first zone and a fourth spectrum from the second zone at different times during the polishing sequence;
Comparing the third spectrum and the fourth spectrum with the library to determine a third index for the first zone and a fourth index for the second zone;
Determining the polishing rate in the first zone from the first index and the third index, and determining the polishing rate in the second zone from the second index and the fourth index;
Based on the first polishing rate and the second polishing rate, the first target relative thickness of the first zone, and the second target relative thickness of the second zone, the polishing rate adjusted to the second zone is By determining, the product is an operation of polishing the second zone to the second target relative thickness substantially simultaneously with the polishing of the first zone to the first target relative thickness. 99. The method of claim 98, wherein the first zone is an inner zone and the second zone is an outer annular zone. 99. The instructions for determining an adjusted polishing rate for the second zone include instructions for determining when the first target relative thickness is within a predetermined threshold from the second target relative thickness. Product described in.

Images