JP4996331B2 - Substrate polishing apparatus and substrate polishing method - Google Patents

Description

  The present invention relates to a substrate polishing apparatus such as a polishing apparatus for polishing and flattening a substrate such as a semiconductor wafer, and a substrate polishing method for polishing and flattening a substrate such as a semiconductor wafer.

  In recent years, semiconductor devices have become increasingly finer and the element structure has become more complex, and as the number of layers of logic-based multilayer wiring has increased, the unevenness of the surface of the semiconductor device has increased and the level difference has a tendency to increase. This is because, in the manufacture of semiconductor devices, the process of forming a thin film, micropatterning for patterning and opening and then forming the next thin film is repeated many times.

  If the irregularities on the surface of the semiconductor device increase, the film thickness at the stepped part will become thinner during thin film formation, open due to disconnection of the wiring, short circuit due to insulation failure between wiring layers, etc. There is a tendency to decrease. In addition, even if the device normally operates normally at the beginning, a problem of reliability occurs for a long time use. Furthermore, if the irradiation surface has irregularities at the time of exposure in the lithography process, the lens focus of the exposure system becomes partially unfocused. Therefore, if the irregularities on the surface of the semiconductor device increase, it becomes difficult to form a fine pattern itself.

Accordingly, in the semiconductor device manufacturing process, a planarization technique for the surface of the semiconductor device is becoming increasingly important. Among the planarization techniques, the most important technique is chemical mechanical polishing (CMP). In this chemical mechanical polishing, a polishing apparatus containing abrasive grains such as silica (SiO ₂ ) is supplied onto a polishing surface such as a polishing pad using a polishing apparatus, and a substrate such as a semiconductor wafer is slid onto the polishing surface. Polishing in contact.

  This type of polishing apparatus includes a polishing table having a polishing surface made of a polishing pad, and a substrate holding device called a top ring or carrier head for holding a semiconductor wafer. When polishing a semiconductor wafer using such a polishing apparatus, the semiconductor wafer is pressed against the polishing table with a predetermined pressure while the semiconductor wafer is held by the substrate holding apparatus. At this time, by moving the polishing table and the substrate holding device relative to each other, the semiconductor wafer comes into sliding contact with the polishing surface, and the surface of the semiconductor wafer is polished to a flat and mirror surface.

  In such a polishing apparatus, when the relative pressing force between the semiconductor wafer being polished and the polishing surface of the polishing pad is not uniform over the entire surface of the semiconductor wafer, it is applied to each part of the semiconductor wafer. Depending on the pressing force, insufficient polishing or overpolishing occurs. Therefore, the holding surface of the semiconductor wafer of the substrate holding device is formed of an elastic film made of an elastic material such as rubber, fluid pressure such as air pressure is applied to the back surface of the elastic film, and the pressing force applied to the semiconductor wafer is applied over the entire surface. It is also made uniform.

  In addition, since the polishing pad has elasticity, the pressing force applied to the outer peripheral edge of the semiconductor wafer being polished becomes non-uniform, so that only the outer peripheral edge of the semiconductor wafer is polished, so-called “edge fringing” is caused. May end up. In order to prevent such edge fringing, a substrate having a structure in which the outer peripheral edge of the semiconductor wafer is held by a guide ring or a retainer ring and the polishing surface located on the outer peripheral edge side of the semiconductor wafer is pressed by the guide ring or the retainer ring. A holding device is also used.

  By the way, the thickness of the thin film formed on the surface of the semiconductor wafer varies depending on the position of the semiconductor wafer in the radial direction, depending on the characteristics of the method and apparatus during film formation. That is, it has a film thickness distribution in the radial direction. For this reason, a polishing apparatus having a substrate holding device that divides a substrate that is in sliding contact with the polishing surface of the polishing table into a plurality of regions and adjusts a pressing force that presses the polishing surface for each region. Known (see Patent Documents 1 and 2). With this apparatus, it is possible to adjust the distribution of the pressing force in the radial direction, and it is possible to flatten the film thickness distribution as a whole.

JP 2003-106805 A JP 2002-187060 A

  However, the film thickness distribution on the surface of the semiconductor wafer described above varies depending on the film forming method and the type of film forming apparatus. That is, the radial position and the number of thick portions and the difference in film thickness between the thin and thick portions differ depending on the film forming method and the type of film forming apparatus. Therefore, there is a demand for a substrate polishing apparatus or a substrate polishing method that can easily and inexpensively cope with various film thickness distributions, not a substrate polishing apparatus that supports only a specific film thickness distribution.

  The present invention has been made in view of such problems of the prior art, and can accurately perform polishing corresponding to the film thickness distribution of a thin film formed on the surface of an object to be polished such as a semiconductor wafer. An object of the present invention is to provide a substrate polishing apparatus and a substrate polishing method capable of obtaining uniformity of film thickness after polishing.

A substrate polishing apparatus of the present invention that solves the above problems is formed on a polishing table having a polishing surface, a substrate holding device that holds a substrate to be polished and presses it against the polishing surface of the polishing table, and the substrate. A substrate polishing apparatus comprising an eddy current sensor for measuring a film thickness of a film, wherein the substrate holding device divides the substrate in sliding contact with the polishing surface of the polishing table into a plurality of regions, and Adjusting means for adjusting the pressing force applied to the polishing surface for each region, so that the gain of the eddy current sensor saturates more than a corresponding signal when the film on the substrate has a predetermined film thickness; The eddy current sensor detects when the film reaches the predetermined film thickness based on a first derivative value of the measured value, and the detection unit of the eddy current sensor crosses the substrate. The eddy current sensor By allocating the time-series film thickness measurement values obtained during the movement of the detection unit to each region of the substrate, the measurement information of the film thickness of each region is obtained, and the substrate holding device is The pressing force applied to each region of the substrate is adjusted based on measurement information on the substrate by the eddy current sensor .

In addition, the substrate polishing method of the present invention that solves the above-described problems is achieved by holding a substrate to be polished with a substrate holding device that can adjust the pressing force to the polishing surface of the polishing table for each of a plurality of regions of the substrate. A substrate polishing method for polishing a film formed on a substrate by pressing against a polishing surface, the film thickness of the substrate being measured using an eddy current sensor, and the gain of the eddy current sensor being The film on the substrate is set so as to saturate at a corresponding signal or more when the film on the substrate has a predetermined film thickness, and the film on the substrate is based on the first derivative of the measured value of the eddy current sensor. , The detection unit of the eddy current sensor is moved across the substrate, and the time series film thickness measurement values obtained during the movement of the detection unit are measured for each of the substrates. By allocating to the area, the film thickness of each area is measured. To obtain information, the pressing force applied to each region of the substrate of the substrate holding device, and adjusting based on the measurement information on the substrate by the eddy current sensor.

  According to the present invention, the pressing force to be brought into sliding contact with the polishing surface of the polishing table is adjusted for each region according to the film thickness for each of the plurality of regions of the substrate. Accordingly, the substrate can be polished at the polishing rate for each region, and the film thickness on the substrate can be processed with high accuracy as a whole. Here, as a film thickness measuring apparatus for measuring the film thickness of a plurality of regions of the substrate, it is preferable to use an eddy current sensor particularly for measurement during polishing because it is not necessary to prepare an opening on the polishing surface. is there. However, a sensor that outputs a signal corresponding to the film thickness on the substrate by optics, temperature, torque current, microwave, or the like may be used, or may be used in combination with these sensors.

  According to the present invention, the operation data of the substrate holding device is provided by including the substrate holding device capable of adjusting the distribution of the pressing force in the radial direction and the film thickness measuring device capable of measuring the film thickness distribution in the radial direction. By automatically adjusting (recipe), a more stable polishing result can be obtained. For example, when polishing a two-layer film of a Cu film and a barrier film such as Ta, this intermediate state is detected by a film thickness measuring device, and the polishing conditions such as pressing force are switched from the Cu film to the barrier film conditions. be able to. Moreover, the oscillation frequency of the oscillator of the film thickness measuring apparatus itself, for example, an eddy current sensor can be switched, and the film thickness measuring apparatus itself can be switched to a condition suitable for detection of the barrier film.

  According to the present invention, when it is necessary to switch the polishing mode with a predetermined film thickness, it is possible to accurately detect the film thickness on the order of angstroms. Therefore, by switching the operation mode (recipe) of the polishing apparatus to, for example, for the barrier layer, it is possible to execute a highly accurate polishing process. And by adjusting the pressing force for sliding contact with the polishing surface of the polishing table according to the film thickness for each region of the substrate for each region, the substrate can be polished at the polishing rate for each region. The film thickness on the substrate can be processed with high accuracy. Therefore, it is possible to adjust the polishing rate and the like within the polishing surface of the substrate to obtain a highly accurate film thickness including removal of the film material on the substrate.

  Embodiments of the present invention will be described below with reference to the accompanying drawings. 1 to 24 are views showing an embodiment of a substrate polishing apparatus for executing a substrate polishing method according to the present invention.

  FIG. 1 is a plan view showing an arrangement configuration of each part of a substrate polishing apparatus according to the present invention. The substrate polishing apparatus includes a polishing table having a polishing surface, a substrate holding device that holds a substrate to be polished and presses it against the polishing surface of the polishing table, and a film that measures the film thickness of a film formed on the substrate A thickness measuring device.

  In this substrate polishing apparatus, a transfer robot 1004 moving on a traveling rail 1003 takes out and stores a substrate such as a semiconductor wafer stocked in a cassette 1001 and also places the unpolished and polished substrate on a mounting table 1050. And it is relayed to the transport robot 1020 and reciprocated between the rotary transporter 1027. Then, the substrate on the rotary transporter 1027 is positioned on the polishing table 100 while being held on the top ring 1 of the substrate holding device described later, so that a plurality of substrates can be polished continuously. The substrate polishing apparatus is systematized. In FIG. 1, reference numerals 1005 and 1022 denote cleaning machines, which are configured to be able to clean and dry the polished substrate. Reference numeral 1036 denotes a polishing table, which is configured so that the substrate can be polished in two stages. Reference numeral 1038,3000 denotes a dresser for dressing the polishing tables 100, 1036, and reference numeral 1043 denotes a water tank for cleaning the dresser 1038.

  This substrate polishing apparatus includes an in-line film thickness measuring apparatus 200 'that measures the film thickness of a semiconductor wafer or the like that has been cleaned and dried after polishing. As shown in FIG. 1, before the transfer robot 1004 stores the polished wafer in the cassette 1001, or after the transfer robot 1004 takes out the unpolished wafer from the cassette 1001 (in-line), the sensor coil is used. A conductive film of a substrate such as a semiconductor wafer from an eddy current signal, an optical signal incident and reflected on a polished surface by optical means, a temperature signal of a polished surface, or a reflected signal of a microwave alone or in an appropriate combination A film thickness measuring device (measuring means) 200 ′ for measuring the film thickness of an insulating film such as a Cu film, a barrier layer, or an oxide film is disposed. Then, the substrate polishing apparatus can detect that the conductive film is removed except for a necessary region such as a wiring portion or the insulating film is removed during or after polishing the substrate. The measurement value is detected by monitoring, the end point of the CMP process is determined, and an appropriate polishing process can be repeated.

  Although not shown, the polishing table 100 includes an in-situ film thickness measuring device that measures the film thickness of a semiconductor wafer or the like being polished. These measurement results are transmitted to a controller, which will be described later, and used to correct operation data (recipe) of the polishing apparatus. In combination with the conditions of each polishing process in the polishing step, for example, the polishing table, the rotation speed of the top ring, the pressure, etc. By measuring the film thickness from a non-metallic thick film such as a film to a thin film and detecting a relative increase / decrease change, it is used for setting various conditions in the polishing process, for example, detecting a polishing end point. In these film thickness measuring devices, the film thickness of each region partitioned in the radial direction of the semiconductor wafer can be measured, and the pressing force applied to each region of the semiconductor wafer of the substrate holding device is the film thickness measuring device. Is adjusted based on the film thickness measurement information for each region.

  The substrate holding device of this substrate polishing apparatus is an apparatus that holds a substrate such as a semiconductor wafer to be polished as described above and presses it against the polishing surface on the polishing table for polishing. As shown in FIG. 2, a polishing table 100 having a polishing pad (polishing cloth) 101 attached to the upper surface is installed below the top ring 1 constituting the substrate holding device. A polishing liquid supply nozzle 102 is installed above the polishing table 100, and the polishing liquid Q is supplied onto the polishing pad 101 on the polishing table 100 by the polishing liquid supply nozzle 102.

  There are various types of polishing pads available on the market, such as SUBA800, IC-1000, IC-1000 / SUBA400 (double-layer cloth) manufactured by Rodel, Surfin xxx-5 manufactured by Fujimi Incorporated, Surfin 000 etc. SUBA800, Surfin xxx-5, and Surfin 000 are non-woven fabrics in which fibers are hardened with urethane resin, and IC-1000 is a hard foamed polyurethane (single layer). The polyurethane foam is porous (porous) and has a large number of fine dents or pores on its surface.

  The top ring 1 is connected to a top ring drive shaft 11 via a universal joint portion 10, and the top ring drive shaft 11 is connected to a top ring air cylinder 111 fixed to a top ring head 110. The top ring drive shaft 11 is moved up and down by the top ring air cylinder 111 to move the entire top ring 1 up and down and press the retainer ring 3 fixed to the lower end of the top ring body 2 against the polishing table 100. ing. The top ring air cylinder 111 is connected to the compressed air source 120 via the regulator RE1, and the regulator RE1 can adjust the fluid pressure such as the air pressure of the pressurized air supplied to the top ring air cylinder 111. . Thereby, the pressing force with which the retainer ring 3 presses the polishing pad 101 can be adjusted.

  The top ring drive shaft 11 is connected to the rotary cylinder 112 through a key (not shown). The rotating cylinder 112 includes a timing pulley 113 on the outer periphery thereof. A top ring motor 114 is fixed to the top ring head 110, and the timing pulley 113 is connected to a timing pulley 116 provided on the top ring motor 114 via a timing belt 115. Accordingly, when the top ring motor 114 is rotationally driven, the rotary cylinder 112 and the top ring drive shaft 11 are integrally rotated via the timing pulley 116, the timing belt 115, and the timing pulley 113, and the top ring 1 is rotated. The top ring head 110 is supported by a top ring head shaft 117 fixedly supported on a frame (not shown).

  Next, the top ring 1 constituting the substrate holding device will be described in more detail with reference to FIGS. FIG. 3 is a longitudinal sectional view showing the top ring 1 in this embodiment, and FIG. 4 is a bottom view of the top ring 1 shown in FIG.

  As shown in FIG. 3, the top ring 1 constituting the substrate holding device includes a cylindrical container-shaped top ring body 2 having an accommodation space inside, and a retainer ring 3 fixed to the lower end of the top ring body 2. ing. The top ring body 2 is made of a material having high strength and rigidity such as metal and ceramics. The retainer ring 3 is made of a highly rigid resin material or ceramics.

  The top ring body 2 is fitted to a cylindrical container-like housing part 2a, an annular pressure sheet support part 2b fitted inside the cylindrical part of the housing part 2a, and an outer peripheral edge part on the upper surface of the housing part 2a. And an annular seal portion 2c. The lower part of the retainer ring 3 fixed to the lower end of the housing part 2a of the top ring body 2 projects inward. The retainer ring 3 may be formed integrally with the top ring main body 2.

  The above-described top ring drive shaft 11 is disposed above the central portion of the housing portion 2a of the top ring main body 2. The top ring main body 2 and the top ring drive shaft 11 are connected by a universal joint portion 10. Yes. The universal joint portion 10 includes a spherical bearing mechanism that allows the top ring body 2 and the top ring drive shaft 11 to tilt relative to each other, and a rotation transmission mechanism that transmits the rotation of the top ring drive shaft 11 to the top ring body 2. The top ring drive shaft 11 transmits the pressing force and the rotational force while allowing the tilting of the top ring body 2.

  The spherical bearing mechanism includes a spherical recess 11a formed at the center of the lower surface of the top ring drive shaft 11, a spherical recess 2d formed at the center of the upper surface of the housing portion 2a, and the recesses 11a and 2d. And a bearing ball 12 made of a high hardness material such as ceramic. On the other hand, the rotation transmission mechanism includes a drive pin (not shown) fixed to the top ring drive shaft 11 and a driven pin (not shown) fixed to the housing portion 2a. Even if the top ring main body 2 is tilted, the driven pin and the driving pin can move relatively in the vertical direction, so that they engage with each other by shifting their contact points, and the rotation transmission mechanism rotates the top ring driving shaft 11. Torque is reliably transmitted to the top ring body 2.

  In the space defined inside the top ring body 2 and the retainer ring 3 fixed integrally to the top ring body 2, an elastic pad 4 that abuts on the semiconductor wafer W held by the top ring 1, and an annular shape A holder ring 5 and a generally disc-shaped chucking plate 6 that supports the elastic pad 4 are accommodated. The outer periphery of the elastic pad 4 is sandwiched between the holder ring 5 and the chucking plate 6 fixed to the lower end of the holder ring 5, and covers the lower surface of the chucking plate 6. Thereby, a space is formed between the elastic pad 4 and the chucking plate 6.

The chucking plate 6 may be formed of a metal material. However, the chucking plate 6 is formed on the surface of the semiconductor wafer to be polished by a film thickness measuring method using an eddy current described later with the top ring held. When measuring the thickness of a thin film, a material having no magnetism, for example, a fluorine-based resin such as a tetrafluoroethylene resin, or a ceramic such as SiC (silicon carbide) or Al ₂ O ₃ (alumina), etc. It is preferable that it is formed from the insulating material.

  A pressure sheet 7 made of an elastic film is stretched between the holder ring 5 and the top ring body 2. One end of the pressure sheet 7 is sandwiched between the housing portion 2a of the top ring body 2 and the pressure sheet support portion 2b, and the other end is sandwiched between the upper end portion 5a of the holder ring 5 and the stopper portion 5b. It is fixed. A pressure chamber 21 is formed inside the top ring body 2 by the top ring body 2, the chucking plate 6, the holder ring 5, and the pressure sheet 7. As shown in FIG. 3, a fluid path 31 including a tube, a connector, and the like communicates with the pressure chamber 21, and the pressure chamber 21 is connected to the compressed air source 120 via a regulator RE <b> 2 disposed on the fluid path 31. Has been. The pressure sheet 7 is formed of a rubber material having excellent strength and durability, such as ethylene propylene rubber (EPDM), polyurethane rubber, and silicon rubber.

  In the case where the pressure sheet 7 is an elastic body such as rubber, when the pressure sheet 7 is sandwiched and fixed between the retainer ring 3 and the top ring body 2, the pressure sheet 7 as an elastic body. Due to the elastic deformation, a preferable plane cannot be obtained on the lower surface of the retainer ring 3. Therefore, in order to prevent this, in the present embodiment, the pressure sheet support portion 2b is provided as a separate member and is fixed by being sandwiched between the housing portion 2a and the pressure sheet support portion 2b of the top ring body 2. is doing. As described in Japanese Patent Application No. 8-50956 (Japanese Patent Application Laid-Open No. 9-168964) and Japanese Patent Application No. 11-294503, the retainer ring 3 can be moved up and down with respect to the top ring body 2, or the retainer ring 3 Can be made a structure that can be pressed independently of the top ring main body 2. In such a case, the above-described fixing method of the pressure sheet 7 is not necessarily used.

  In the vicinity of the outer peripheral edge of the upper surface of the housing portion 2a to which the seal portion 2c of the top ring body 2 is fitted, a cleaning liquid path 51 formed of an annular groove is formed. The cleaning liquid path 51 communicates with the fluid path 32 through the through hole 52 of the seal portion 2 c, and the cleaning liquid (pure water) is supplied through the fluid path 32. Further, a plurality of communication holes 53 are provided from the cleaning liquid passage 51 through the housing part 2 a and the pressure sheet support part 2 b, and the communication holes 53 are slightly formed between the outer peripheral surface of the elastic pad 4 and the retainer ring 3. The gap G is communicated.

  In the space formed between the elastic pad 4 and the chucking plate 6, there are a center bag 8 (center contact member) as a contact member that contacts the elastic pad 4 and a ring tube 9 (outer contact). Member). In this embodiment, as shown in FIGS. 3 and 4, the center bag 8 is disposed at the center of the lower surface of the chucking plate 6, and the ring tube 9 surrounds the center bag 8 so as to surround the center bag 8. It is arranged outside. The elastic pad 4, the center bag 8, and the ring tube 9 are formed of a rubber material having excellent strength and durability, such as ethylene propylene rubber (EPDM), polyurethane rubber, and silicon rubber, like the pressure sheet 7. Yes.

  A space formed between the chucking plate 6 and the elastic pad 4 is partitioned into a plurality of spaces by the center bag 8 and the ring tube 9, and thus, a pressure is applied between the center bag 8 and the ring tube 9. A chamber 22 and a pressure chamber 23 are formed outside the ring tube 9, respectively.

  The center bag 8 includes an elastic film 81 that contacts the upper surface of the elastic pad 4 and a center bag holder 82 (holding portion) that holds the elastic film 81 in a detachable manner. A screw hole 82a is formed in the center bag holder 82, and the center bag 8 is detachably attached to the center portion of the lower surface of the chucking plate 6 by screwing a screw 55 into the screw hole 82a. . A center pressure chamber 24 is formed in the center bag 8 by an elastic membrane 81 and a center bag holder 82.

  Similarly, the ring tube 9 includes an elastic film 91 that contacts the upper surface of the elastic pad 4 and a ring tube holder 92 (holding part) that holds the elastic film 91 in a detachable manner. A screw hole 92 a is formed in the ring tube holder 92, and the ring tube 9 is detachably attached to the lower surface of the chucking plate 6 by screwing a screw 56 into the screw hole 92 a. An intermediate pressure chamber 25 is formed in the ring tube 9 by an elastic membrane 91 and a ring tube holder 92.

  The pressure chambers 22, 23, the central pressure chamber 24, and the intermediate pressure chamber 25 are in fluid communication with fluid paths 33, 34, 35, 36 made of tubes, connectors, etc. It is connected to a compressed air source 120 as a supply source via regulators RE3, RE4, RE5, and RE6 disposed on the fluid passages 33 to 36, respectively. The fluid paths 31 to 36 are connected to a pure water supply source and regulators RE2 to RE6 via a rotary joint (not shown) provided at the upper end of the top ring shaft 110, respectively.

  The pressure chamber 21 above the chucking plate 6 and the pressure chambers 22 to 25 are pressurized with pressurized air or the like via fluid passages 31, 33, 34, 35, and 36 communicated with the pressure chambers. Fluid, atmospheric pressure or vacuum is supplied. As shown in FIG. 2, the pressure of the pressurized fluid supplied to each pressure chamber is adjusted by regulators RE2 to RE6 arranged on the fluid passages 31, 33, 34, 35, and 36 of the pressure chambers 21 to 25. be able to. Thereby, the pressure inside each pressure chamber 21-25 can be controlled independently, respectively, or it can be made atmospheric pressure or a vacuum. In this way, by making the pressures inside the pressure chambers 21 to 25 independently variable by the regulators RE2 to RE6, the pressing force for pressing the semiconductor wafer W against the polishing pad 101 via the elastic pad 4 is changed. It can be adjusted for each part (partition area). In some cases, these pressure chambers 21 to 25 may be connected to the vacuum source 121.

  In this case, it is good also as controlling the temperature of the pressurized fluid and atmospheric pressure which are supplied to each pressure chamber 22-25. In this way, the temperature of the polishing object can be directly controlled from the back side of the surface to be polished of the polishing object such as a semiconductor wafer. In particular, if the temperature of each pressure chamber is controlled independently, the chemical reaction rate of chemical polishing in CMP can be controlled.

  The elastic pad 4 is provided with a plurality of openings 41 as shown in FIG. An inner periphery adsorbing portion 61 that protrudes downward from the chucking plate 6 is provided so as to be exposed from the opening 41 between the center bag 8 and the ring tube 9. An outer peripheral suction part 62 is provided so as to be exposed from the opening 41. In the present embodiment, the elastic pad 4 is provided with eight openings 41 so that the suction portions 61 and 62 are exposed in each opening 41.

  The inner periphery suction portion 61 and the outer periphery suction portion 62 are respectively formed with communication holes 61a and 62a communicating with the fluid paths 37 and 38, respectively. The inner periphery suction portion 61 and the outer periphery suction portion 62 are fluids. It is connected to a vacuum source 121 such as a vacuum pump through passages 37 and 38 and valves V1 and V2. And when the communication holes 61a and 62a of the inner peripheral part adsorption | suction part 61 and the outer peripheral part adsorption | suction part 62 are connected to the vacuum source 121, a negative pressure will be formed in the opening end of the communication holes 61a and 62a, and inner peripheral part adsorption | suction part The semiconductor wafer W is adsorbed by 61 and the outer periphery adsorbing part 62. In addition, elastic sheets 61b and 62b made of a thin rubber sheet or the like are attached to the lower end surfaces of the inner periphery suction portion 61 and the outer periphery suction portion 62, and the inner periphery suction portion 61 and the outer periphery suction portion 62 are The semiconductor wafer W is adsorbed and held flexibly.

  Further, as shown in FIG. 3, during polishing of the semiconductor wafer W, the inner periphery suction portion 61 and the outer periphery suction portion 62 are located above the lower end surface of the elastic pad 4, and the lower end surface of the elastic pad 4. There is no more protrusion. When adsorbing the semiconductor wafer W, the lower end surfaces of the inner periphery adsorbing portion 61 and the outer periphery adsorbing portion 62 are substantially flush with the lower end surface of the elastic pad 4.

  Here, since there is a slight gap G between the outer peripheral surface of the elastic pad 4 and the retainer ring 3, a member such as the holder ring 5, the chucking plate 6, and the elastic pad 4 attached to the chucking plate 6. Is movable in the vertical direction with respect to the top ring body 2 and the retainer ring 3 and has a floating structure. The stopper portion 5b of the holder ring 5 is provided with a plurality of protrusions 5c protruding outward from the outer peripheral edge thereof, and the protrusions 5c are related to the upper surface of the portion protruding inward of the retainer ring 3. By combining, the downward movement of the member such as the holder ring 5 is limited to a predetermined position.

  Next, the operation of the top ring 1 configured as described above will be described.

  In the substrate polishing apparatus having the above-described configuration, when the semiconductor wafer W is transferred, the entire top ring 1 is positioned at the transfer position of the semiconductor wafer, and the communication holes 61a and 62a of the inner periphery suction portion 61 and the outer periphery suction portion 62 are formed. It is connected to the vacuum source 121 via fluid paths 37 and 38. The semiconductor wafer W is vacuum-sucked to the lower end surfaces of the inner periphery suction portion 61 and the outer periphery suction portion 62 by the suction action of the communication holes 61a and 62a. Then, the top ring 1 is moved in a state where the semiconductor wafer W is attracted, and the entire top ring 1 is positioned above the polishing table 100 having the polishing surface (polishing pad 101). The outer peripheral edge of the semiconductor wafer W is held by the retainer ring 3 so that the semiconductor wafer W does not jump out of the top ring 1.

  At the time of polishing, the suction of the semiconductor wafer W by the suction portions 61 and 62 is released, the semiconductor wafer W is held on the lower surface of the top ring 1, and the top ring air cylinder 111 connected to the top ring drive shaft 11 is operated. Then, the retainer ring 3 fixed to the lower end of the top ring 1 is pressed against the polishing surface of the polishing table 100 with a predetermined pressing force. In this state, a pressurized fluid having a predetermined pressure is supplied to each of the pressure chambers 22 and 23, the central pressure chamber 24, and the intermediate pressure chamber 25 to press the semiconductor wafer W against the polishing surface of the polishing table 100. Then, by flowing the polishing liquid Q from the polishing liquid supply nozzle 102, the polishing liquid Q is held on the polishing pad 101, and the polishing liquid Q is interposed between the surface (lower surface) of the semiconductor wafer W to be polished and the polishing pad 101. Polishing is performed in the existing state.

  Here, the portions of the semiconductor wafer W positioned below the pressure chambers 22 and 23 are pressed against the polishing surface by the pressure of the pressurized fluid supplied to the pressure chambers 22 and 23, respectively. A portion of the semiconductor wafer W positioned below the central pressure chamber 24 is polished by the pressure of the pressurized fluid supplied to the central pressure chamber 24 through the elastic film 81 and the elastic pad 4 of the center bag 8. Pressed against the surface. A portion of the semiconductor wafer W positioned below the intermediate pressure chamber 25 is applied to the polishing surface by the pressure of the pressurized fluid supplied to the intermediate pressure chamber 25 via the elastic film 91 and the elastic pad 4 of the ring tube 9. Pressed.

  Therefore, the polishing pressure applied to the semiconductor wafer W can be adjusted for each part in the radial direction of the semiconductor wafer W by controlling the pressure of the pressurized fluid supplied to the pressure chambers 22 to 25, respectively. That is, a controller (control unit) 400, which will be described later, independently adjusts the pressures of the pressurized fluid supplied to the pressure chambers 22 to 25 by regulators RE3 to RE6 (adjusting means), and the semiconductor wafer W is placed on the polishing table 100. The pressing force that presses the polishing pad 101 is adjusted for each portion of the semiconductor wafer W. Thus, the semiconductor wafer W is pressed against the polishing pad 101 on the upper surface of the rotating polishing table 100 in a state where the polishing pressure is adjusted to a desired value for each portion of the semiconductor wafer W. Similarly, the pressure of the pressurized fluid supplied to the top ring air cylinder 111 can be adjusted by the regulator RE1, and the pressing force with which the retainer ring 3 presses the polishing pad 101 can be changed. As described above, by appropriately adjusting the pressing force with which the retainer ring 3 presses the polishing pad 101 and the pressing force with which the semiconductor wafer W is pressed against the polishing pad 101 during polishing, the central portion of the semiconductor wafer W (in FIG. 4). C1), distribution of polishing pressure in each part from the central part to the intermediate part (C2), the outer part (C3), the peripheral part (C4), and the outer peripheral part of the retainer ring 3 outside the semiconductor wafer W Can be a desired distribution.

  It should be noted that the portion of the semiconductor wafer W positioned below the pressure chambers 22, 23 has a pressure of the pressurized fluid, such as a portion where a pressing force is applied from the fluid via the elastic pad 4 and a location of the opening 41. Although there are portions that themselves are applied to the semiconductor wafer W, the pressing force applied to these portions may be the same pressure, and can be pressed even at any arbitrary pressure. Further, at the time of polishing, since the elastic pad 4 is in close contact with the back surface of the semiconductor wafer W around the opening 41, the pressurized fluid inside the pressure chambers 22 and 23 hardly leaks outside.

  Thus, the semiconductor wafer W can be partitioned into four concentric circles and ring portions (C1 to C4), and each portion (region) can be pressed with an independent pressing force. Although the polishing rate depends on the pressing force on the polishing surface of the semiconductor wafer W, since the pressing force of each part can be controlled as described above, the polishing rate of the four parts (C1 to C4) of the semiconductor wafer W is adjusted. It can be controlled independently. Therefore, even if the film thickness of the thin film to be polished on the surface of the semiconductor wafer W has a radial distribution, it is possible to eliminate insufficient polishing and overpolishing over the entire surface of the semiconductor wafer. That is, even if the thin film to be polished on the surface of the semiconductor wafer W has a different thickness depending on the position in the radial direction of the semiconductor wafer W, the surface of the semiconductor wafer W among the pressure chambers 22 to 25 described above. The pressure of the pressure chamber located above the thick part of the semiconductor wafer W is made higher than the pressure of the other pressure chambers, or the pressure chamber located above the thin part of the surface of the semiconductor wafer W By making the pressure lower than the pressure in the other pressure chambers, it is possible to make the pressing force on the polishing surface of the thick part thicker than the pressing force on the polishing surface of the thin part. The polishing rate can be selectively increased. Thus, it is possible to polish the entire surface of the semiconductor wafer W without excess or deficiency without depending on the film thickness distribution at the time of film formation.

  Here, drooling that occurs at the peripheral edge of the semiconductor wafer W can be prevented by controlling the pressing force of the retainer ring 3. If there is a large change in the film thickness of the thin film to be polished at the peripheral edge of the semiconductor wafer W, the peripheral edge of the semiconductor wafer W can be increased by intentionally increasing or decreasing the pressing force of the retainer ring 3. The polishing rate can be controlled. When pressurized fluid is supplied to each of the pressure chambers 22 to 25, the chucking plate 6 receives an upward force. In this embodiment, pressure fluid is supplied to the pressure chamber 21 via the fluid path 31. In addition, the chucking plate 6 is prevented from being lifted upward by the force from each pressure chamber 22-25.

  As described above, the pressing force of the retainer ring 3 to the polishing pad 101 by the top ring air cylinder 111 and the polishing pad 101 for each portion of the semiconductor wafer W by the pressurized air supplied to the pressure chambers 22 to 25 are applied. The semiconductor wafer W is polished by appropriately adjusting the pressing force. When the polishing is completed, the semiconductor wafer W is again vacuum-sucked to the lower end surfaces of the inner periphery suction portion 61 and the outer periphery suction portion 62. At this time, the supply of the pressurized fluid to the pressure chambers 22 to 25 that press the semiconductor wafer W against the polishing surface is stopped and the pressure is released to the atmospheric pressure, whereby the inner periphery suction portion 61 and the outer periphery suction portion 62. Is brought into contact with the semiconductor wafer W. Further, the pressure in the pressure chamber 21 is released to atmospheric pressure or negative pressure. This is because, if the pressure in the pressure chamber 21 is kept high, only the portion of the semiconductor wafer W that is in contact with the inner periphery suction portion 61 and the outer periphery suction portion 62 is strongly pressed against the polishing surface. This is because it becomes. Therefore, it is necessary to quickly reduce the pressure in the pressure chamber 21, and as shown in FIG. 3, a relief port 39 is provided so as to penetrate the top ring body 2 from the pressure chamber 21, so that the pressure in the pressure chamber 21 quickly decreases. You may do it. In this case, when applying pressure to the pressure chamber 21, it is necessary to always supply the pressure fluid from the fluid path 31. In addition, the relief port 39 includes a check valve so that outside air does not enter the pressure chamber 21 when the pressure chamber 21 is set to a negative pressure.

  After the semiconductor wafer W is adsorbed as described above, the entire top ring 1 is positioned at the transfer position of the semiconductor wafer, and the semiconductor wafer W is communicated from the communication holes 61a and 62b of the inner peripheral portion adsorbing portion 61 and the outer peripheral portion adsorbing portion 62. The semiconductor wafer W is released by spraying a fluid (for example, a mixture of compressed air or nitrogen and pure water).

  By the way, the polishing liquid Q used for polishing enters the slight gap G between the outer peripheral surface of the elastic pad 4 and the retainer ring 3. When this polishing liquid Q adheres, the holder ring 5, the chuck Smooth vertical movement of members such as the king plate 6 and the elastic pad 4 with respect to the top ring body 2 and the retainer ring 3 is hindered. Therefore, the cleaning liquid (pure water) is supplied to the cleaning liquid path 51 via the fluid path 32. As a result, pure water is supplied above the gap G from the plurality of communication holes 53, and the pure water is washed away from the gap G to prevent the above-described polishing liquid Q from sticking. This supply of pure water is preferably performed during the period from when the polished semiconductor wafer is released until the next polished semiconductor wafer is adsorbed. Moreover, it is preferable to provide the retainer ring 3 with a plurality of through holes 3a as shown in FIG. 3 so that all the pure water supplied until the next polishing is discharged to the outside. Furthermore, if pressure is accumulated in the space 26 formed by the retainer ring 3, the holder ring 5, and the pressure sheet 7, the chucking plate 6 is prevented from ascending. In order to raise, it is preferable to provide the through hole 3a and make the space 26 have the same pressure as the atmosphere.

  As described above, the pressure on the semiconductor wafer is controlled by independently controlling the pressure in the pressure chambers 22, 23, the pressure chamber 24 in the center bag 8, and the pressure chamber 25 in the ring tube 9. Can do. Furthermore, according to the substrate holding device, the range in which the pressing force is controlled can be easily changed by changing the positions and sizes of the center bag 8 and the ring tube 9.

  That is, the film thickness distribution of the thin film formed on the surface of the semiconductor wafer varies depending on the film forming method and the type of film forming apparatus, but according to the substrate holding apparatus according to the present invention, the pressure for applying a pressing force to the semiconductor wafer. The position and size of the chamber can be changed by simply replacing the center bag 8 and the center bag holder 82 or the ring tube 9 and the ring tube holder 92. Therefore, the position and range where the pressing force should be controlled in accordance with the film thickness distribution of the thin film to be polished can be changed easily and at low cost simply by exchanging a part of the top ring 1. In other words, even when there is a change in the film thickness distribution of the thin film to be polished on the surface of the semiconductor wafer to be polished, it can be handled easily and at low cost. If the shape and position of the center bag 8 or the ring tube 9 are changed, the size of the pressure chamber 22 sandwiched between the center bag 8 and the ring tube 9 and the pressure chamber 23 surrounding the ring tube 9 can be changed as a result. Become.

  A copper plating film for forming wiring is formed on a semiconductor wafer to be polished by this substrate polishing apparatus, and a barrier layer is formed as a base material thereof. When an insulating film such as silicon oxide is formed on the uppermost layer of a semiconductor wafer to be polished by this substrate polishing apparatus, the thickness of the insulating film is detected by an optical sensor or a microwave sensor. As the light source of the optical sensor, a halogen lamp, a xenon flash lamp, an LED, a laser light source, or the like is used. In order to remove a polishing target film such as an insulating film or a conductive film in an unnecessary region (outside of a wiring region, etc.) on the semiconductor wafer, the substrate polishing apparatus uses various sensors to determine whether or not the polishing target film exists. As shown in FIG. 2, the controller 400 controls the polishing process on the surface of the semiconductor wafer W while detecting the film thickness of the polishing target film 201 by the eddy current sensor (film thickness measuring device) 200.

  Hereinafter, the substrate polishing method executed by the controller 400 of the substrate polishing apparatus according to the present invention will be described in more detail with reference to FIGS.

  FIG. 5 is a block diagram showing a schematic overall configuration of the substrate polishing apparatus. Based on input from a man-machine interface 401 such as an operation panel or input from a host computer 402 that performs various data processing, the controller 400 converts the semiconductor wafer W into a target polishing rate so as to have a target profile such as a desired shape. Polish with (polishing amount). The closed-loop control system 403 automatically creates a polishing recipe (polishing conditions, etc.) corresponding to each of the regions C1 to C4 of the semiconductor wafer W according to simulation software prepared in advance in a hard disk device (not shown). ing. Then, the polishing recipe is temporarily stored in the memory (storage means) 404, and polishing control processing corresponding to the polishing recipe is executed. In this polishing process, the film thickness and polishing amount of the polishing target film 201 calculated by the arithmetic circuit 405 are compared with the target profile and the target polishing rate based on the measurement results from the film thickness measuring apparatuses 200 and 200 ′. The semiconductor wafer W is repeatedly subjected to the polishing process under the optimum conditions by examining and performing feedback processing for correcting the polishing recipe according to the result.

  The feedback process of the polishing process can be selectively set to be performed after the semiconductor wafer polishing process is completed or during the polishing process, and the controller 400 can select one or both of the polishing process and the polishing process according to the selected setting. The polishing recipe is corrected at.

  Specifically, as shown in FIG. 6, the operator selects and inputs a dry system (measurement of film thickness in a dry state after polishing) from the host computer 402, and also sets a target profile and polishing rate (removal rate). When input is set (step S1), a polishing recipe is automatically created by simulation software (step S2), and the polishing conditions of the recipe are displayed on the monitor of the host computer 402 and the like, and whether or not correction is necessary is determined and input. Is requested from the operator (step S3). If correction is required, the recipe is corrected according to the correction input (step S4), and then the polishing process is started (step S5).

  Then, after executing and completing the polishing process according to the polishing recipe and incrementing (+1) the number N of polishing processes (step S11), the polished semiconductor wafer W is cleaned and dried (steps S12 and S13).

Thereafter, in the dry system (measurement of film thickness in a dry state after polishing), the film thickness of the polishing target film 201 of the semiconductor wafer W is measured by the film thickness measuring apparatus 200 ′ (step S14), and the insulating film Alternatively, the polishing result is stored (stored) together with an identifier for specifying the semiconductor wafer W whose metal film has been polished. At the same time, the semiconductor wafer W is returned into the cassette 1001 and stored (step S15). In parallel with this storage process, the simulation software is used to push the polishing pad 101 for each polishing region or region C1 to C4 of the semiconductor wafer W according to the film thickness measurement result of the polishing target film 201 of the polished semiconductor wafer W. After performing the automatic creation process for correcting the polishing recipe under the polishing conditions such as pressure (step S16), the process returns to step S11 to repeat the polishing process for the next semiconductor wafer W or a wafer to be polished several times later. In addition, when a film to be polished such as an insulating film or a conductive film remains on the measured wafer without being completely removed, only the remaining portion of the film to be polished is polished (polished) Re-polishing conditions are created and re-polished, such as pressurizing only the pressure chamber in the region corresponding to the remaining portion so that the portion is not over-polished.
In addition, since the above dry system mainly measures the wafer after polishing, a film thickness measuring device that measures the wafer after polishing and before drying may be used instead of the wafer after drying. .

  On the other hand, when a wet system (film thickness measurement in a wet state during polishing) is selected and input from the host computer 402, a target profile and a polishing rate are also input and set as shown in FIG. (Step S1) After automatically creating a polishing recipe with simulation software and starting the polishing process (Steps S2 to S5), during the polishing process according to the polishing recipe, the number of polishing processes (the number of recipes created) N is incremented (+1) (step S21), and the film thickness of the polishing target film 201 of the semiconductor wafer W is measured by the eddy current sensor (film thickness measuring device) 200, the optical sensor, or the microwave sensor (step S22).

  When the film thickness measurement result of the film to be polished 201 of the semiconductor wafer W remains to the extent that the polishing process needs to be added, the film thickness of the film to be polished 201 of the semiconductor wafer W remains. After performing an automatic creation process of a polishing recipe for correcting the polishing conditions by simulation software according to the measurement result (step S23), the process returns to step S21 and the polishing process for the same semiconductor wafer W is repeated. On the other hand, if the film thickness measurement result of the polishing target film 201 of the semiconductor wafer W does not require an additional polishing process, the polished semiconductor wafer W is cleaned and dried (steps S24 and S25). Processing for storing the polished result of the polished semiconductor wafer W and storage processing in the cassette 1001 are performed (step S26), and the processing returns to step S21 to repeat the polishing processing of the next semiconductor wafer W.

  Here, the correction of the polishing recipe by the above simulation software compares the target profile with the actual profile (step S31), as shown in FIG. 8, and the difference in the polishing rate for each of the regions C1 to C4 of the semiconductor wafer W. Is converted into a pressure difference for each region (step S32), and the target polishing rate is compared with the actual polishing rate (step S33) to polish each region C1 to C4 of the semiconductor wafer W. The required time is calculated (step S34), and a polishing recipe for automatically adjusting the pressing force and polishing time of each region under the polishing conditions is automatically created. At the same time, automatic correction reflecting these is performed (step S35), and a corrected polishing recipe for executing the polishing process of the next semiconductor wafer W is automatically created (step S36). As a result, the semiconductor wafer can be uniformly polished along the radial direction.

  In addition, in the in-situ measurement of the film thickness 201 of the polishing target film 201 of the semiconductor wafer W, whether or not a desired polishing process has been completed in a specific region or all regions of the regions C1 to C4 of the semiconductor wafer. Therefore, the determination method may be selected and set as appropriate. For example, using the measurement result of a specific region, the measurement result of each region, the average value of the measurement result, etc., the removal of the polishing target film 201 or a predetermined polishing target film thickness value determined in advance is measured. It is determined by the temporal change pattern of. In addition, it is possible to easily determine the temporal change of the measurement result by performing first-order differentiation or n-th order differentiation.

  Specifically, as shown in FIG. 9, according to the film quality or film type of the polishing target film 201, the timing at which the measurement result, differential value, or the like becomes a set value or more (detection pattern 0), the set value Timing (detection pattern 1), timing for taking the maximum value (detection pattern 2), timing for taking the minimum value (detection pattern 3), timing to start the rise (detection pattern 4), timing to finish the rise (detection) It can be determined that the desired polishing process has been completed based on the pattern 5), the timing at which the descent starts (detection pattern 6), the timing at which the descent ends (detection pattern 7), and the like. It can also be determined that the desired polishing process has been completed when the differential value (gradient) falls within the specified range, or when the maximum value or the minimum value is taken (detection patterns 8 to 10). Furthermore, it can also be determined that the desired polishing process has been completed at a timing (detection pattern 11) where a specific measurement result falls within the specified range and converges. Further, when higher uniformity is required, it may be determined that the desired polishing process has been completed at the timing (detection pattern 12) when all the measurement results of all the regions are within the specified range (converged). it can.

  Further, for example, the first derivative value of the film thickness measurement value is set as a monitoring target. Calculates the film thickness measurement primary differential value difference between a specified wafer area and another wafer area within a plurality of previously specified areas (for example, a predetermined radius value range or a predetermined angle range viewed from a reference point) of the wafer to be measured. To do. It is also conceivable to detect a point where this difference falls within a predetermined threshold range as a feature point (for example, an end point). Furthermore, the impedance integral value Sz (or resistance value Sx, reactance value Sy, film thickness integral value St) from the start of polishing of the eddy current sensor is calculated and compared with the previously described impedance integral value S0 as a reference. It can be used for polishing state monitoring and polishing end point monitoring.

By determining the film thickness of the polishing target film 201 in this manner, the film thickness itself can be measured, and it is necessary to detect the end of polishing of the Cu layer or the barrier layer of the polishing target film during the polishing process. In some cases, the timing can be detected quickly, and the polishing process can be immediately terminated. In addition, in the eddy current sensor described later, the thickness of tungsten (W) used as a barrier layer is 1000 mm, and the absolute film thickness value of the metal film can be changed at any time even if there is a demand to slow down the polishing rate by changing to a low-pressure polishing process. Since it can be measured, it can be changed to a low-pressure polishing process while monitoring the film thickness value, and the amount of dishing and erosion can be reduced. By using an eddy current monitor, it has become possible to monitor changes in film thickness of thin barrier films and CVD films that were difficult to measure in-situ with an optical monitor.
In the eddy current sensor, if the metal exists in a solid film shape (a film shape that covers an entire region) in the region where the eddy current is formed, the end point of the metal barrier can be detected. In this case, if an abnormality occurs that results in film thickness measurement results such as in-plane uniformity or polishing rate (polishing rate) in a specific area of the wafer exceeding preset limit values or limit ranges, an abnormality will occur immediately. When the generation process is performed and the polishing operation is stopped, and the measurement result when the semiconductor wafer has a defect such as a flaw is obtained, it is preferable to attach the fact to the polishing result.

  As described above, in this embodiment, the pressing force applied to the polishing pad 101 can be adjusted for each of the regions C1 to C4 according to the film thickness of each of the regions C1 to C4 of the semiconductor wafer W. The polishing target film 201 can be polished at a polishing rate according to the shape such as the film thickness and the film quality. Therefore, the polishing target film 201 of the semiconductor wafer W can be polished and removed with high accuracy. Here, in the conductive film polishing process among the films to be polished, the film thickness measuring apparatus for the wet process is a semiconductor wafer with low cost and high accuracy by omitting the process of forming openings such as windows in the polishing pad 101. An eddy current sensor described in detail below is suitable because W can be polished. However, of course, a microwave sensor, an optical sensor, or the like may be used depending on the characteristics of the object to be polished.

  Hereinafter, the eddy current sensor (film thickness measuring apparatus) 200 constituting the film thickness measuring apparatus included in the substrate polishing apparatus according to the present invention will be described in more detail with reference to FIGS.

  In the eddy current sensor (film thickness measuring device) 200, as shown in FIG. 10A, a sensor coil 202 is disposed in the vicinity of a conductive film 201 ′ to be detected, and an AC signal source 203 is connected to the coil. ing. Here, the conductive film 201 ′ to be detected is, for example, a copper plating film (on a metal film such as Au, Cr, W, etc.) having a thickness of about 0 to 1 μm formed on the semiconductor wafer W, or The thickness formed on the underlying layer is an angstrom order barrier layer. The barrier layer is a high resistance layer made of Ta, TaN, Ti, TiN, WN or the like, and this film thickness detection is important for accurately detecting the end point in the chemical mechanical polishing described above. The sensor coil 202 is a coil for detection, and is disposed in the vicinity of, for example, about 1.0 to 4.0 mm with respect to the conductive film to be detected. Examples of eddy current sensor measurement targets include Al (aluminum film), polysilicon used for contact plugs, conductive materials such as CoFe and Zr (zirconia) used for hard disk magnetic heads, and metal materials. Of course, a semiconductor substrate including a metal film or metal wiring pattern formed on a semiconductor wafer is also an object to be measured.

In the eddy current sensor, an eddy current is generated in the conductive film 201 ′, whereby the oscillation frequency changes. The frequency type for detecting the film thickness from this frequency change and the impedance change, and the film thickness is determined from this impedance change. There are impedance types to detect. That is, in the frequency type, in the equivalent circuit shown in FIG. 10 (b), by eddy current I ₂ is changed, the impedance Z is changed, the oscillation frequency of the signal source (variable-frequency oscillator) 203 is changed, the detection circuit In 205, the change in the oscillation frequency is detected, and the change in the film thickness can be detected. In the impedance type, in the equivalent circuit shown in FIG. 10B, when the eddy current I ₂ changes, the impedance Z changes, and the impedance Z viewed from the signal source (fixed frequency oscillator) 203 changes. In 205, the change in impedance Z can be detected, and the change in film thickness can be detected.

  In an impedance type eddy current sensor, signal outputs X, Y, phase, and combined impedance Z are taken out as described later. Depending on the film thickness value converted from the frequency F or impedance X, Y, etc., the thickness of the metal film Cu, Al, Au, W, barrier film Ta, TaN, Ti, TiN, WN, polysilicon of the contact plug, etc. Measurement information is obtained. These can be used alone, in combination, or in combination for the polishing process such as end point detection. The eddy current sensor can be built in a position near the surface inside the polishing table, and is positioned so as to face the semiconductor wafer to be polished through the polishing pad, and the eddy current flowing in the conductive film on the semiconductor wafer. The film thickness can be detected from the current.

  The frequency of the eddy current sensor can be a single radio wave, mixed radio wave, AM modulated radio wave, FM modulated radio wave, function generator sweep output or multiple oscillation frequency sources, adapted to the film type of the metal film, It is preferable to select an oscillation frequency or modulation method with good sensitivity.

The impedance type eddy current sensor will be specifically described below. The AC signal source 203 is an oscillator having a fixed frequency of about 2 to 8 MHz. For example, a crystal oscillator is used. The current I ₁ flows through the sensor coil 202 due to the AC voltage supplied from the AC signal source 203. When a current flows through the coil 202 disposed in the vicinity of the conductive film 201 ′, this magnetic flux is linked to the conductive film 201 ′, thereby forming a mutual inductance M therebetween. eddy current _{I 2} flows. Where R1 is the equivalent resistance of the primary side including the sensor coil, L ₁ is self inductance of the primary side including the sensor coil as well. The conductive film 201 'side, R2 is equivalent resistance corresponding to eddy current loss, L ₂ is its self-inductance. The impedance Z when the sensor coil side is viewed from the terminals a and b of the AC signal source 203 varies depending on the magnitude of the eddy current loss formed in the conductive film 201 ′.

  FIG. 11 shows a configuration example of a sensor coil in the eddy current sensor of the present embodiment. The sensor coil 202 is obtained by separating a coil for forming an eddy current in a conductive film and a coil for detecting an eddy current in the conductive film, and a three-layer coil 312 wound around a bobbin 311. , 313, 314. Here, the center coil 312 is an oscillation coil connected to the AC signal source 203. The oscillation coil 312 forms an eddy current in the conductive film 201 ′ on the semiconductor wafer W disposed in the vicinity by a magnetic field generated by the voltage supplied from the AC signal source 203. A detection coil 313 is arranged on the upper side (conductive film side) of the bobbin 311 to detect a magnetic field generated by an eddy current formed in the conductive film. A balance coil 314 is disposed on the opposite side of the oscillation coil 312 from the detection coil 313.

  FIG. 12 shows a connection example of each coil. In this embodiment, the coils 312, 313, and 314 are formed by coils having the same number of turns (1 to 20 t), and the detection coil 313 and the balance coil 314 are connected to each other in the positive phase.

As described above, the detection coil 313 and the balance coil 314 constitute a positive-phase series circuit, and both ends thereof are connected to a resistance bridge circuit 317 including a variable resistor 316 as shown in FIG. The coil 312 is connected to the AC signal source 203 and generates an alternating magnetic flux, thereby forming an eddy current in the conductive film 201 ′ disposed in the vicinity. By adjusting the resistance value of the variable resistor 316, the output voltage of the series circuit composed of the coils 313 and 314 can be adjusted to zero when there is no conductive film. The variable resistors 316 (VR ₁ , VR ₂ ) entering in parallel to the coils 313 and 314 are adjusted so that the signals of L ₁ and L ₃ are in phase. That is, in the equivalent circuit of FIG.
VR _1-1 × (VR _2-2 + jωL ₃ ) = VR _1-2 × (VR _2-1 + jωL ₁ ) (1)
The variable resistors VR ₁ (= VR _1-1 + VR _1-2 ) and VR ₂ (= VR _2-1 + VR _2-2 ) are adjusted so that As a result, as shown in FIG. 12C, the L ₁ and L ₃ signals before adjustment (indicated by dotted lines in the figure) are made into signals having the same phase and amplitude (indicated by solid lines in the figure).

  When the conductive film is present in the vicinity of the detection coil 313, the magnetic flux generated by the eddy current formed in the conductive film is linked to the detection coil 313 and the balance coil 314, but the detection coil 313 is more conductive. Since it is arranged at a position close to the conductive film, the balance of the induced voltages generated in the coils 313 and 314 is lost, and thereby the interlinkage magnetic flux formed by the eddy current of the conductive film can be detected. That is, the zero point can be adjusted by separating the series circuit of the detection coil 313 and the balance coil 314 from the oscillation coil 312 connected to the AC signal source and adjusting the balance with a resistance bridge circuit. . Therefore, since it is possible to detect the eddy current flowing in the conductive film from the zero state, the detection sensitivity of the eddy current in the conductive film is enhanced. As a result, it is possible to detect the magnitude of the eddy current formed in the conductive film with a wide dynamic range.

FIG. 13 shows a measurement circuit example of the impedance Z when the sensor coil 202 side is viewed from the AC signal source side 203. In the impedance Z measurement circuit shown in FIG. 13, the resistance component (R), reactance component (X), amplitude output (Z), and phase output (tan ⁻¹ R / X) accompanying the change in film thickness are extracted. Can do. Therefore, by using these signal outputs, for example, it is possible to detect the progress of polishing in a multifaceted manner, such as measuring the film thickness based on the amplitude.

  As described above, the signal source 203 that supplies an AC signal to the sensor coil 202 disposed in the vicinity of the semiconductor wafer W on which the conductive film 201 ′ to be detected is formed is a fixed-frequency oscillator including a crystal oscillator. Yes, for example, voltages with fixed frequencies of 2 MHz and 8 MHz are supplied. The AC voltage formed by the signal source 203 is supplied to the sensor coil 202 via the band pass filter 302. The signal detected at the terminal of the sensor coil 202 passes through the high-frequency amplifier 303 and the phase shift circuit 304, and the cos component and the sin component of the detection signal are detected by the synchronous detection unit including the cos synchronous detection circuit 305 and the sin synchronous detection circuit 306. It is taken out. Here, the oscillation signal formed by the signal source 203 is formed into two signals of the in-phase component (0 °) and the quadrature component (90 °) of the signal source 203 by the phase shift circuit 304, and the cos synchronous detection circuit 305, respectively. And the sin synchronous detection circuit 306, and the synchronous detection described above is performed.

From the synchronously detected signal, unnecessary high frequency components equal to or higher than the signal component are removed by the low-pass filters 307 and 308, and the resistance component (R) output that is the cos synchronous detection output and the reactance component (X that is the sin synchronous detection output) ) Output is taken out respectively. Further, the vector operation circuit 309 obtains an amplitude output (R ² + X ² ) ^1/2 from the resistance component (R) output and the reactance component (X) output. Similarly, the vector operation circuit 310 obtains a phase output (tan ⁻¹ R / X) from the resistance component output and the reactance component output. Here, various filters are provided in the film thickness measuring device main body in order to remove noise components of the sensor signal. Each filter has a cut-off frequency corresponding to each filter. For example, by setting the cut-off frequency of the low-pass filter in a range of 0.1 to 10 Hz, noise components mixed in the sensor signal being polished can be set. The film thickness and the like of the measurement object can be measured with high accuracy by removing.

FIG. 14 shows changes in the impedance Z as viewed from the AC signal source side, the horizontal axis is the resistance component (R), and the vertical axis is the reactance component (X). Point A is when the film thickness is extremely large, for example, 100 μm or more. In this case, the impedance Z when the sensor coil 202 side is viewed from the terminals a and b of the AC signal source 203 has an extremely large eddy current in the conductive film 201 disposed in the vicinity of the sensor coil. The resistance component (R ₂ ) and the reactance component jω (M + L ₂ ) that are equivalently connected in parallel are extremely small. Accordingly, both the resistance component (R) and the reactance component (X) are reduced.

As the polishing progresses and the conductive film becomes thinner, the impedance Z viewed from the sensor coil input ends (terminals a and b) increases the equivalent resistance component (R ₂ ) and reactance component jω (M + L ₂ ). Increase. A point where the resistance component (R) of the impedance Z viewed from the sensor coil input end is maximized is indicated by B. At this time, the eddy current loss seen from the sensor coil input end is maximized. As the polishing progresses further and the conductive film becomes thinner, the eddy current decreases and the resistance component viewed from the sensor coil gradually decreases because the eddy current loss decreases gradually. . When all of the conductive film is removed by polishing, there is no eddy current loss, the resistance component (R ₂ ) equivalently connected in parallel becomes infinite, and the resistance component of the sensor coil itself ( Only R ₁ ) will remain. The reactance component (X) at this time is only the reactance component (L ₁ ) of the sensor coil itself. This state is indicated by point C.

  Actually, for example, when a copper wiring is formed in a groove provided in a silicon oxide film by a so-called damascene process, a barrier layer such as tantalum nitride (TaN) or titanium nitride (TiN) is provided on the silicon oxide film. On top of that, metal wiring such as copper or tungsten having high conductivity is provided. Therefore, in polishing these conductive films, it is important to detect the end point of polishing of the barrier layer. However, as described above, the barrier layer is made of a very thin film having a relatively low conductivity such as tantalum nitride (TaN) or titanium nitride (TiN) and having a film thickness on the order of angstroms.

  However, in the eddy current sensor of this embodiment, the thickness of the barrier layer in the vicinity of the polishing end point is detected and the thickness value during polishing (not the relative thickness data but the film as an absolute length). (Thickness value) can be easily measured. That is, the point D shown in FIG. 14 indicates a position where the film thickness is about 1000 mm, for example, and the resistance component changes extremely corresponding to the change in the film thickness toward the point C where the film thickness becomes zero. Large and changes substantially linearly. At this time, the reactance component (X) has a very small amount of change compared to the resistance component as shown in the figure. For this reason, in the eddy current sensor based on the principle that the film thickness is detected based on the change in the oscillation frequency caused by the change in the reactance, the change in the oscillation frequency is extremely small with respect to the change in the film thickness. For this reason, in order to increase the resolution of frequency change, it is necessary to increase the frequency. However, according to this eddy current sensor (film thickness measuring device) 200, it is possible to detect a change in film thickness by observing a change in the resistance component while keeping the oscillation frequency fixed. It becomes possible to clearly observe the polishing state of an extremely thin film thickness. Here, in this embodiment, a method of detecting the film thickness based on the change of the resistance component caused by the change of the reactance is adopted, but depending on the measurement target, the film thickness is detected based on the change of the oscillation frequency. It goes without saying that the film thickness may be detected based on the combined impedance of the reactance component and the resistance component.

  FIG. 15 shows the result of detecting the film thickness of a fine conductive layer on the order of angstroms. The horizontal axis represents the remaining film thickness, the solid line on the vertical axis represents the resistance component (R), and the dotted line on the vertical axis represents the reactance component (X). FIG. 15A shows data relating to the tungsten (W) film, and it can be seen that the change in the film thickness can be clearly detected by observing the change in the resistance component with a fine remaining film thickness of 1000 mm or less. FIG. 15B shows data relating to a titanium nitride (TiN) film, and similarly, a change in film thickness can be clearly detected in a region of 1000 mm or less. FIG. 15C shows data relating to a titanium (Ti) film. As shown in the figure, the resistance component changes drastically while the film thickness changes from 500 to 0 mm, so that the film thickness becomes clear. Changes can be detected.

  In each example shown in FIG. 15, the change in the reactance component (X) is extremely small with respect to the change in the resistance component (R). In the example of detecting the thickness of the barrier layer, the change in the reactance component (X) was 0.005% in the tantalum film when the remaining film thickness was 0 mm and 250 mm. On the other hand, the change of the resistance component (R) was 1.8%. Therefore, the detection sensitivity is improved by about 360 times as compared with the method of looking at the change of the reactance component.

  However, it is desirable that the oscillation frequency of the AC signal source is increased to, for example, about 8 to 16 MHz when detecting a barrier layer having a relatively low conductivity. By increasing the oscillation frequency, it is possible to clearly observe the change in the thickness of the barrier layer from 0 to 250 mm. On the other hand, for example, in a metal having a relatively high conductivity such as a copper film, it is possible to clearly detect a change in film thickness even at a low oscillation frequency of about 2 MHz. In the case of a tungsten film, an oscillation frequency of about 8 MHz is preferable. Thus, it is preferable to select the oscillation frequency, the sensor amplification degree, and the OFF-SET value of the sensor signal value in accordance with the type of film to be polished.

  Further, only when the wafer approaches or faces the eddy current sensor embedded in the polishing table 100, the eddy current sensor is applied to an alternating burst electromagnetic field or sine wave by an electric / electronic circuit (not shown), balanced modulation, amplitude modulation, An eddy current sensor module having a configuration in which an electromagnetic field subjected to pulse modulation is irradiated to a wafer measurement location may be used. Furthermore, the eddy current film thickness data measurement timing can be a circuit configuration that continuously irradiates an electromagnetic field, or a circuit configuration that irradiates the wafer with an electromagnetic field only when the wafer approaches or faces the eddy current sensor. When a circuit configuration that continues to irradiate an electromagnetic field is adopted, when the wafer is not in proximity to the eddy current sensor or when it is not facing it, the film thickness in the future can be complemented by the film thickness data predicted from the data acquired in the past. It can also be used for polishing abnormality detection, apparatus abnormality detection, etc. by predicting the time change and end point time or comparing with the actual polishing time. In some cases, the film thickness detection function of the sensor is disabled or the eddy current signal is not sampled at the time of non-polishing, dressing the polishing cloth, except when the wafer is close to or facing the eddy current sensor.

  FIG. 16 (a) is a longitudinal sectional view showing the main configuration of a substrate polishing apparatus provided with these eddy current sensors, and FIG. 17 is a plan view thereof. As shown in FIG. 16A, the polishing table 100 of the substrate polishing apparatus is rotatable about its axis as indicated by an arrow. A preamplifier integrated sensor coil 202 including an AC signal source and a synchronous detection circuit is embedded in the polishing table 100. The connection cable of the sensor coil 202 passes through the polishing table support shaft 321a of the polishing table 100, passes through the rotary joint 334 provided at the shaft end of the polishing table support shaft 321a, and passes through the main amplifier 200a. It is connected to the thickness measuring device main body (control unit) 200b.

  Here, various filters are provided in the film thickness measuring device main body 200b in order to remove noise components of the sensor signal. Each filter has a cut-off frequency corresponding to each filter. For example, by setting the cut-off frequency of the low-pass filter in a range of 0.1 to 10 Hz, noise components mixed in the sensor signal being polished can be set. The film thickness and the like of the measurement object can be measured with high accuracy by removing.

  FIG. 16B is an enlarged cross-sectional view of the eddy current sensor portion. When the polishing pad is peeled off by having a coating 202m of fluororesin such as tetrafluoroethylene resin on the end surface on the polishing pad side of the sensor coil 202 embedded in the polishing table 100, both the polishing pad and the eddy current sensor are used. It can be prevented from coming off. Further, the end surface on the polishing pad side of the eddy current sensor is installed at a position recessed from 0 to 0.05 mm from the surface of the polishing table 100 (surface on the polishing pad side) made of a material such as SiC near the polishing pad 101, This prevents contact with the wafer during polishing. The difference in position between the polishing table surface and the eddy current sensor surface is preferably as small as possible, but in an actual apparatus, it is often set to about 0.02 mm. For this position adjustment, adjustment by a shim (thin plate) 202n or adjustment means by a screw is taken.

  Here, the rotary joint connecting the sensor coil 202 and the film thickness measuring device main body 200b can transmit a signal even in the rotating part, but there is a limit to the number of signal lines to be transmitted. Therefore, the number of signal lines to be connected is limited to eight, and is limited to only the DC voltage source, the output signal line, and the transmission lines for various control signals. The sensor coil 202 can be switched at an oscillation frequency of 2 to 8 MHz, and the gain of the preamplifier can be switched according to the film quality to be polished.

  As shown in FIG. 17, the rotation of the polishing table 100 is such that the dog sensor 350 detects the dog 351 attached to the outer peripheral surface of the polishing table 100, and the film thickness measuring device main body 200 b includes the dog sensor 350. Signal processing of the semiconductor wafer W held by the top ring 1 is started. That is, as the polishing table 100 rotates, the sensor trace R is traced so as to cross the semiconductor wafer W.

  Accordingly, as shown in FIG. 18, the film thickness measuring device main body 200b first receives a signal from the dog sensor 350 while the polishing table 100 makes one round. At this time, since the semiconductor wafer W has not yet arrived on the sensor coil 202, a signal outside the wafer is received. Thereafter, when the sensor coil 202 is positioned under the semiconductor wafer W, a sensor signal having a level corresponding to an eddy current generated in the conductive film 201 ′ or the like is received. After the semiconductor wafer W passes over the sensor coil 202, a sensor signal outside the wafer at a level where no eddy current is generated is received.

  At this time, the film thickness measuring device main body 200b is in a state where the sensor coil 202 is always activated so as to be capable of sensing. However, if the film thickness of the conductive film 201 ′ of the semiconductor wafer W to be measured is measured as it is, the film When the thickness changes according to polishing and the level of the received sensor signal itself changes, the measurement timing becomes unstable. From this, water polishing is performed by supplying water from the polishing abrasive liquid supply nozzle 102 to perform dummy polishing of the reference wafer (for example, a reference wafer on which a 1000 nm Cu layer is formed is rotated 120 times by the polishing table 100 at 60 rpm. Second level water polishing) is performed, and the measurement start timing level of the polishing region of the semiconductor wafer W is acquired and set in advance. Specifically, an intermediate value of the sensor level based on the presence or absence of the semiconductor wafer after receiving the wafer detection signal by the dog sensor 350 is set as the arrival determination level of the edge of the semiconductor wafer W. Therefore, a sensor signal is acquired every 1 millisecond (msec), for example, at a timing exceeding the arrival determination level of the edge of the semiconductor wafer W after the arrival of the signal of the dog sensor. The acquisition of the sensor signal ends when the wafer leaves the sensor. The obtained sensor signal is assigned to each area in correspondence with the physical dimensions.

  As shown in FIG. 19A, the sensor signal received by the film thickness measuring device main body 200b is obtained by linearizing the sensor locus R in the semiconductor wafer W, as shown in FIG. 19A (C1 in FIG. 4) below the semiconductor wafer W. 19 to the peripheral edge (C4), for example, as shown in FIG. 19 (b), the semiconductor wafer W is divided into a central part (C1), an intermediate part (C2), and a peripheral part (C3). The thickness of the conductive film 201 for each C4) can be measured from before polishing to after polishing. The sensor signal of each region is subjected to arithmetic processing such as averaging, and is used as a measurement value of each region.

  Since the outermost peripheral region of the semiconductor wafer W is a portion where the conductive film 201 ′ is not formed, so-called edge cut processing is performed in which the sensor signal in the region is discarded and processed. In the present embodiment, the semiconductor wafer W is divided into three regions and the measurement values in the five regions are acquired. However, the semiconductor wafer W is divided into four regions C1 to C4 corresponding to the regions where the pressing force can be adjusted. Then, the measurement values in the seven regions may be acquired and controlled. Furthermore, it goes without saying that the polished surface of the semiconductor wafer W may be divided more finely or coarsely.

  As shown in FIG. 20, the acquired sensor signal is allocated to each of the areas C1 to C4, and the number of measurement values corresponding to the area width is calculated and acquired. For example, in the region 1 of the peripheral edge (C3, C4) at the time of entry from the entry to the withdrawal of the sensor coil 202 below the semiconductor wafer W, two measurement values are acquired, and then the intermediate part (C2). Two measurement values are acquired also in the area 2, and then one measurement value is acquired in the area 3 of the central portion (C 1). Thereafter, two measurement values are acquired again in the region 4 of the intermediate part (C2), and finally two measurement values are acquired also in the region 5 of the peripheral part (C3, C4).

Here, the film thickness measuring device main body 200b measures the film thickness of the conductive film 201 ′ for each passage (polishing) of the sensor coil 202 below the semiconductor wafer W based on the measurement value acquired for each region. Since processing such as displaying the film thickness of each region on a display device (display) is performed, as shown in FIG. 20, instead of the measurement value based on the sensor signal outside the target region (outside the wafer or outside the region). Thus, supplementary data is generated and displayed. Since this complementary data (value) is virtually processed when the conductive film 201 ′ exists so that the display data does not fluctuate greatly, for example, the preset effective number of recent effective measurements. The value is used to calculate from the following arithmetic expression.
Supplementary value = [measurement maximum value-measurement minimum value] x coefficient (conversion rate%) + measurement minimum value

  Here, measurement is performed by a batch method in which film thickness data is measured only during a period in which the eddy current sensor and the wafer to be polished are opposed each time the polishing table rotates. Further, the signal from the eddy current sensor corresponding to the change in the film thickness of the measurement object is continuously generated every 10 μs to several 100 μs (for example, 100 μs) by the signal-driven external synchronous A / D converter from the dog sensor 350 described above. Synchronous addition of a plurality of measured data (for example, this means that 10 consecutive data obtained every 100 μs obtained from the dog sensor 350 are added and averaged, and this data is used as 1 msec data. The noise of the data can be reduced by adding and averaging.

FIG. 21 shows another embodiment of the polishing table 100 shown in FIG. As shown in the figure, the sensor coils 202a to 202f are installed at six positions in this case at positions through which the center Cw of the semiconductor wafer W being polished held by the top ring 1 passes. Reference symbol _CT denotes the rotation center of the polishing table 100. While the sensor coils 202a to 202f pass through the central part (C1 in FIG. 4) below the semiconductor wafer W, the central part to the intermediate part (C2), the outer part (C3), and the peripheral part (C4), The film thicknesses of the conductive films such as the Cu layer and the barrier layer of the semiconductor wafer W can be detected continuously (without waiting for one round) on the passing trajectory. That is, the eddy current sensor (film thickness measuring device) 200 is configured such that the sensor coil (measuring means) 202 can measure the film thickness of the regions C1 to C4 partitioned so that the pressing force of the semiconductor wafer W can be adjusted. Has been. Here, a plurality of types of sensor coil frequencies may be used. Thereby, it is possible to perform management such as mainly detecting the change in the thickness of the barrier layer at the higher side and mainly detecting the change in the thickness of the Cu layer at the lower side.

  In this case, six sensor coils are arranged, but the number of arrangements can be changed as appropriate. Although an example in which a polishing pad is arranged on a polishing table will be described, a fixed abrasive plate may be used. In this case, a sensor coil may be arranged in the fixed abrasive plate.

  In the substrate polishing apparatus having the above configuration, the semiconductor wafer W is held on the lower surface of the top ring 1, and the semiconductor wafer W is pressed against the polishing pad 101 on the upper surface of the rotating polishing table 100 by the lifting cylinder. On the other hand, by supplying the polishing abrasive liquid Q from the polishing abrasive liquid supply nozzle 102, the polishing abrasive liquid Q is held on the polishing pad 101, and polishing is performed between the polishing surface (lower surface) of the semiconductor wafer W and the polishing pad 101. Polishing is performed in the state where the abrasive liquid Q is present.

  During this polishing, the sensor coils 202a to 202f pass directly under the surface to be polished of the semiconductor wafer W each time the polishing table 100 rotates once. In this case, since the sensor coils 202a to 202f are installed on a track passing through the center Cw of the semiconductor wafer W, the film is continuously formed on the arc-shaped track of the surface to be polished of the semiconductor wafer W as the sensor moves. Thickness detection is possible. In this case, since six sensor coils are provided, the progress of polishing can be detected by any one of the sensor coils at short intervals, although intermittently.

  As shown in FIGS. 22 (a) and 22 (b), as polishing progresses, the detection values for processing the signals of the sensor coils 202a to 202f by the film thickness measuring device main body 200b gradually decrease. That is, as the film thickness of the conductive film decreases, the detected value, which is a value obtained by processing the signals of the sensor coils 202a to 202f by the film thickness measuring device main body 200b, decreases. Therefore, if the value of the detection value when the conductive film is removed except for the wiring portion is examined in advance, the end point of the CMP process can be detected by monitoring the value of the detection output.

FIG. 23 shows an example in which the relationship between the film thickness and the resistance component output is calibrated. For example, a reference wafer such as 1000 ｔ (t ₁ ) or 200 Å (t ₂ ) is prepared, and the detection output on this reference wafer is measured, and these points are used as reference points. And the data of the film thickness change with respect to the detection output accompanying the actual progress of polishing are acquired, and this is indicated by a dotted line. In addition to the resistance component output, the detection output can be a reactance component output, an impedance (amplitude) output, and a phase output. A curve is formed for this data with respect to the reference point by a method such as a least square method. By calibrating and storing the characteristics of the eddy current sensor using such a method, the detection output can be amplified and offset as appropriate, without being affected by individual differences in the eddy current sensor. It is possible to accurately read the change in the film thickness directly from the change in the thickness.

  According to the substrate polishing apparatus provided with a large number of such eddy current sensors, end point detection can be performed on the entire surface of the semiconductor wafer, and can be performed at short time intervals. As described above, the polishing end point of the Ta, TaN, TiN layer or the like as the barrier layer can be detected, so that the polishing end point can be detected with extremely high accuracy. At this time, in the eddy current sensor configured as described above, even if a patch residue of the conductive film (metal that has not been polished and removed) occurs in the final stage of the polishing process, the polishing surface on the semiconductor wafer side and the sensor coil If the distance from the upper end is 3.5 mm or less, a patch residue having a diameter (φ) of 5 mm or more can be detected, and the patch residue can be reliably polished and removed in a polishing process described later. Furthermore, even in the case where the wiring made of the conductive material of the semiconductor wafer W to be polished is multi-layered, the eddy current sensor configured in this way can reliably ensure that the surface layer has a surface density of 90% or less. The conductive film can be detected, polished and removed.

  In the film thickness measuring apparatus main body 200b, when it is necessary to switch the polishing mode at a predetermined film thickness, the preamplifier or the main amplifier is selected and set from the beginning within a range in which the film thickness on the angstrom order can be measured. Therefore, it is possible to perform accurate film thickness confirmation processing. For example, when the purpose is to switch the polishing mode at about 300 mm, during the polishing of a W layer of about 300 mm or more, the measurement result of the film thickness to be measured is an overrange (saturation) that cannot be actually measured. So that linear characteristics can be obtained at about 300 mm or less.

  That is, as shown in FIG. 24A, the gain of the amplifier is set so as to be saturated at a signal equivalent to 300Å or more. For example, when the W layer film is polished, even if the W layer polishing progresses as shown by the dotted line as shown in FIG. 24B, the amplifier output is saturated as shown by the solid line in the figure. It becomes constant. However, when the film thickness is less than about 300 mm, the amplifier operates linearly, so that the amplifier output decreases as indicated by the solid line in the figure. Therefore, as shown in FIG. 24C, it is possible to clearly detect that the film thickness has reached about 300 mm by taking the first derivative.

  By obtaining this measurement result and switching the operation mode (recipe) of the polishing apparatus to that for the barrier layer, highly accurate polishing processing can be performed. Also, the operation mode (recipe) of the sensor can detect the presence or absence of a barrier layer with a small film thickness by switching the oscillation frequency or switching the amplification level, and can appropriately determine the polishing end timing. It becomes like this.

  Therefore, the controller 400 (see FIG. 2) of the substrate polishing apparatus includes a central portion (C1 in FIG. 4) and an intermediate portion (C1 in FIG. 4) of the film thickness measuring devices 200 and 200 ′ such as a microwave sensor and an eddy current sensor. C2), based on the measurement results of the film thickness of the peripheral portions (C3, C4), the pressures of the pressurized fluid supplied to the pressure chambers 22-25 of the top ring 1 by the regulators RE3-RE6 are independently adjusted, The pressing force applied to the polishing pad 101 on the polishing table 100 is optimized for each of the regions C1 to C4 of the semiconductor wafer W.

  Here, the film thickness measuring devices 200 and 200 ′ optimize the pressing force applied to the polishing pad 101 on the polishing table 100 for each of the regions C1 to C4 of the semiconductor wafer W. The controller 400 generates an instruction command to the film thickness measuring devices 200 and 200 ′ based on the film thickness measurement result. In other words, the film thickness measuring apparatuses 200 and 200 ′ selectively switch, for example, parameters suitable for the type of film to be measured, the type of multilayer film, and the like in accordance with an instruction command from the controller 400 based on its own film thickness measurement result. By executing the operation mode switching control, film thickness measurement is performed by performing sensor signal calculation processing using the parameters.

  Here, in this embodiment, the case where the polishing target film of the semiconductor wafer is removed by polishing has been described, but the film thickness of the polishing target film is similarly applied to so-called etching, electroless polishing, and ultrapure water electrolytic polishing. Measurements can be taken and process controlled. Needless to say, not only in the case of removing the film to be polished, but also in the film forming step, the film thickness of the film to be polished may be similarly measured to control the process.

  In addition, for example, an eddy current sensor (select as appropriate from oscillation frequencies of 2 MHz, 8 MHz, 20 MHz, and 160 MHz), slurry waste liquid on a pad that generates electromagnetic waves in the 30 GHz to 300 GHz band during the polishing process, and slurry reaction that is drained. Detects the demagnetizing field generated by applying electromagnetic waves to the liquid, the amplitude of the reflected wave, and the impedance change of the reflected wave, and compares it with the reference impedance before polishing, or observes the change in the time derivative of the impedance to end the wafer polishing. It can also be used as a polishing monitor signal for abnormality determination. In addition, the eddy current sensor and electromagnetic waves of the waste liquid and reaction liquid are observed in the plating process, ultrapure water electropolishing apparatus, electroless plating apparatus, electroless polishing apparatus, film removal process, membrane removal process electrolyte and ultrapure water. It can also be used in monitors.

  Although one embodiment of the present invention has been described so far, the present invention is not limited to the above-described embodiment, and needless to say, may be implemented in various forms within the scope of the technical idea. The substrate holding device and the eddy current sensor of the substrate polishing apparatus are not limited to the illustrated examples described above, and it is needless to say that various changes can be made without departing from the scope of the present invention.

It is a figure which shows one Embodiment of the substrate polishing apparatus which performs the substrate polishing method which concerns on this invention, and is a top view which shows the arrangement configuration of each part of the substrate polishing apparatus. It is a partial cross section explanatory view showing a schematic structure around the polishing table. It is a longitudinal cross-sectional view which shows the substrate holding apparatus of the substrate polishing apparatus. It is a bottom view showing a substrate holding device of the substrate polisher. It is a block diagram which shows schematic structure of the film thickness measuring apparatus of the substrate polishing apparatus. It is a flowchart explaining the grinding | polishing process by the substrate grinding | polishing apparatus. It is a flowchart explaining the grinding | polishing process by the substrate grinding | polishing apparatus. It is a flowchart explaining correction of the polishing process by the substrate polishing apparatus. It is a list which shows the detection pattern of the film thickness measurement by the film thickness measuring apparatus of the substrate polishing apparatus. It is a block diagram which shows the structure of the film thickness measuring apparatus of the substrate polishing apparatus. It is a perspective view which shows the sensor coil of the film thickness measuring apparatus of the board | substrate polish apparatus. It is a conceptual diagram which shows the example of a connection of the sensor coil of the film thickness measuring apparatus of the board | substrate polisher. It is a block diagram which shows the synchronous detection circuit of the film thickness measuring apparatus of the board | substrate polish apparatus. It is a graph which shows the transition locus of the resistance component (R) and the reactance component (X) of the film thickness measurement by the film thickness measuring apparatus of the substrate polishing apparatus. It is a graph which shows the example of a change of the resistance component (R) and reactance component (X) of the film thickness measurement by the film thickness measuring apparatus of the substrate polishing apparatus. It is a longitudinal cross-sectional view which shows the principal part structure of the substrate polishing apparatus. It is a top view explaining operation | movement of the substrate polishing apparatus. It is a graph explaining the sensor signal of the film thickness measuring apparatus of the substrate polisher. It is a conceptual diagram explaining the grinding | polishing of the board | substrate by the board | substrate polish apparatus. It is a graph explaining the sensor signal of the film thickness measuring apparatus of the substrate polisher. It is a top view explaining operation | movement of the substrate polishing apparatus. It is a graph explaining the sensor signal of the film thickness measuring apparatus of the substrate polisher. It is a graph explaining the detection output of the film thickness measuring apparatus of the substrate polishing apparatus. It is a graph explaining the sensor signal of the film thickness measuring apparatus of the substrate polisher.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Top ring 3 Retainer ring 5 Holder ring 6 Chucking plate 8 Center bag 9 Ring tube 10 Universal joint part 11 Top ring drive shaft 21-25 Pressure chamber 31-38 Fluid path 39 Relief port 41 Opening 61, 62 Adsorption part 81 , 91 Elastic film 82 Center bag holder 92 Ring tube holder 100 Polishing table 101 Polishing pad 102 Polishing liquid supply nozzle 120 Compressed air source 121 Vacuum source 200, 200 ′ Film thickness measuring device 201 Polishing target film 201 ′ Conductive film 202, 202a ˜202f Sensor coil 203 AC signal source 205 Synchronous detection circuit 302 Band pass filter 307, 308 Low pass filter 312 Oscillation coil 313 Detection coil 314 Balance coil 316 Variable resistance 317 Resistance bridge circuit 350 G-sensor 400 Controller 1001 Cassette 1003 Traveling rails 1004 and 1020 Transport robot 1027 Rotary transporter 1050 Mounting table W Semiconductor wafer

Claims (4)

A polishing table having a polishing surface, a substrate holding device that holds a substrate to be polished and presses it against the polishing surface of the polishing table, and an eddy current sensor that measures the film thickness of a film formed on the substrate A substrate polishing apparatus comprising:
The substrate holding device has an adjusting unit that divides the substrate to be in sliding contact with the polishing surface of the polishing table into a plurality of regions, and adjusts the pressing force that presses the polishing surface for each region.
The gain of the eddy current sensor is set so as to saturate with a corresponding signal or more when the film on the substrate has a predetermined film thickness, and the eddy current sensor is based on the first derivative of the measured value. Detecting when the film reaches the predetermined film thickness,
The detection unit of the eddy current sensor moves across the substrate, and the eddy current sensor transmits a time-series film thickness measurement value obtained during the movement of the detection unit to each region of the substrate. By allocating, obtain measurement information of the film thickness of each region,
The substrate holding apparatus adjusts a pressing force applied to each region of the substrate based on measurement information on the substrate by the eddy current sensor .   The substrate polishing apparatus according to claim 1, wherein the predetermined film thickness is 300 mm. It is formed on the substrate by holding the substrate to be polished and pressing it against the polishing surface of the polishing table with a substrate holding device capable of adjusting the pressing force to the polishing surface of the polishing table for each of a plurality of regions of the substrate. A substrate polishing method for polishing a film,
The thickness of the substrate using an eddy current sensor to measure the gain of the eddy current sensor, film on the substrate is set to be saturated at least signals corresponding to the time which is a predetermined thickness, said vortex Based on the first derivative value of the measured value of the current sensor , detects when the film on the substrate reaches the predetermined film thickness ,
By moving the detection unit of the eddy current sensor across the substrate and allocating the time series film thickness measurement values obtained during the movement of the detection unit to each region of the substrate, Get measurement information of the film thickness of the area,
A substrate polishing method , wherein a pressing force applied to each region of the substrate of the substrate holding device is adjusted based on measurement information on the substrate by the eddy current sensor . The substrate polishing method according to claim 3 , wherein the predetermined film thickness is 300 mm.

Images