JP5705093B2 - Polishing end point detection method and polishing apparatus - Google Patents

Description

  The present invention relates to a polishing end point detection method used in a polishing apparatus for polishing a polishing object (substrate) such as a semiconductor wafer, and more particularly to a polishing end point detection method using an eddy current sensor. The present invention also relates to a polishing apparatus capable of executing the above polishing end point detection method.

  In the manufacturing process of a semiconductor device, various materials are repeatedly formed in a film shape on a silicon wafer to form a multilayer wiring structure. In order to form such a multilayer wiring structure, CMP (Chemical Mechanical Polishing) is widely used. For example, the metal wiring is formed by forming a metal film on the surface of the substrate on which the wiring groove is formed, and then removing the unnecessary film while leaving only the metal film formed in the groove by CMP.

  CMP is performed by a polishing apparatus. This polishing apparatus includes a polishing table that supports a polishing pad, a top ring that holds a substrate (a wafer having a film), and a polishing liquid supply mechanism that supplies a polishing liquid onto the polishing pad. When polishing the substrate, the surface of the substrate is pressed against the polishing pad by the top ring while supplying the polishing liquid from the polishing liquid supply mechanism onto the polishing pad. By rotating the top ring and the polishing table to move the substrate and the polishing pad relative to each other, the film constituting the surface of the substrate is polished.

  In the metal wiring formation process described above, an eddy current sensor is widely used to detect whether or not an unnecessary metal film has been removed (that is, whether or not there is a metal residual film). However, in a substrate having a multilayer wiring structure, wiring existing in the lower layer of the metal film to be polished affects the output signal of the eddy current sensor and prevents detection of the remaining film.

In order to eliminate the influence of such lower layer wiring, the following method is conventionally employed.
(1) The average value of the output signals of the eddy current sensor acquired over the entire surface of the substrate is defined as the film thickness.
(2) The minimum output signal in a predetermined area within the substrate surface is defined as the film thickness.
(3) Top ring and polishing so that the trajectory drawn by the eddy current sensor on the substrate surface within a predetermined time (for example, within the moving average time of the output signal of the eddy current sensor) is distributed almost evenly over the entire circumference of the substrate. Adjust the rotation speed ratio of the table.

  However, with the conventional method described above, it is difficult to obtain film thickness information for each region in the surface of the substrate.

JP 2005-121616 A JP 2011-23579 A

  The present invention has been made to solve the above-described conventional problems, eliminates the influence of the metal material underlying the film to be polished, and uses an eddy current sensor to form a film in each region within the substrate surface. It is an object to provide a method capable of acquiring thickness information and determining a polishing end point of a substrate from the obtained film thickness information. It is another object of the present invention to provide a polishing apparatus capable of executing such a polishing end point detection method.

  In order to achieve the above-described object, according to a first aspect of the present invention, the substrate is rotated by rotating the top ring and the polishing table while pressing the substrate against a polishing pad on the polishing table. A method of detecting a polishing end point of a substrate polishing step for polishing a film of the substrate, wherein an eddy current sensor is moved across the surface of the substrate during polishing of the substrate, and a resistance component X of impedance of the eddy current sensor is detected. And the inductive reactance component Y is obtained, the coordinates X and Y comprising the resistance component X and the inductive reactance component Y are plotted on an XY coordinate system, and a plurality of impedance areas are provided on the XY coordinate system. The plurality of impedance areas are defined in advance and have a reference impedance area and at least one offset impeder. A plurality of film thickness index values for each of the plurality of impedance areas using the plurality of coordinates X and Y respectively belonging to the plurality of impedance areas, and using the plurality of film thickness index values The polishing end point of the substrate is determined for each of the plurality of impedance areas.

According to a second aspect of the present invention, there is provided a substrate polishing step of polishing the film of the substrate by rotating the top ring and the polishing table while pressing the substrate against a polishing pad on the polishing table by the top ring. A method for detecting a polishing end point, wherein a time for which the top ring makes one rotation is measured, a rotation speed of the top ring is calculated from the measured time, and the rotation speed of the top ring and the rotation of the polishing table are calculated. from the ratio of the speed, the eddy current sensor is embedded in the polishing table depicts the same trajectory to calculate the number of rotations of said polishing table for across the surface of the substrate, before the Kiuzu current sensor of the substrate Moving across the surface, obtaining an output signal of the eddy current sensor when the eddy current sensor crosses the surface of the substrate in the same locus, Calculating the film thickness index value from the output signal of Kiuzu current sensor, and determining a polishing end point of the substrate from a change in the film thickness index value.

  According to a third aspect of the present invention, there is provided a substrate polishing step of polishing the film of the substrate by rotating the top ring and the polishing table while pressing the substrate against the polishing pad on the polishing table by the top ring. A method for detecting a polishing end point, wherein an eddy current sensor embedded in the polishing table is moved across the surface of the substrate, an output signal of the eddy current sensor is obtained, and an output signal of the eddy current sensor is obtained. From the change in the position of the convex portion appearing in the film thickness profile, determine whether the convex portion appears due to either the residual film or the metal material existing in the lower layer of the film, The polishing end point of the substrate is determined based on the size of the convex portion that appears due to the remaining film.

  According to a fourth aspect of the present invention, there is provided a rotatable polishing table that supports the polishing pad, a top ring that presses the substrate against the polishing pad on the rotating polishing table while rotating the substrate, and the inside of the polishing table. And an eddy current sensor that moves across the surface of the substrate, and a polishing monitoring unit that monitors the film thickness of the substrate from an output signal of the eddy current sensor, the polishing monitoring unit comprising: The resistance component X and the inductive reactance component Y of the impedance of the current sensor are acquired, the coordinates X and Y consisting of the resistance component X and the inductive reactance component Y are plotted on the XY coordinate system, and the XY coordinate system A plurality of impedance areas are predefined above the plurality of impedance areas, the reference impedance area and at least one off-state. A plurality of film thickness index values for each of the plurality of impedance areas, using a plurality of coordinates X and Y respectively belonging to the plurality of impedance areas. In the polishing apparatus, a polishing end point of the substrate is determined for each of the plurality of impedance areas using a value.

One reference example of the present invention includes a rotatable polishing table that supports a polishing pad, a top ring that presses the substrate against the polishing pad on the rotating polishing table while rotating the substrate, An eddy current sensor that is installed and moves across the surface of the substrate, and a polishing monitoring unit that monitors a film thickness of the substrate from an output signal of the eddy current sensor, the polishing monitoring unit including the eddy current An output signal of the eddy current sensor when the sensor crosses the surface of the substrate in the same locus is obtained, a film thickness index value is calculated from the output signal of the eddy current sensor, and a change in the film thickness index value The polishing apparatus determines a polishing end point of the substrate.

  According to the first and fourth aspects of the present invention described above, each time the output signal of the eddy current sensor is acquired, the coordinates of the output signals X and Y are set to one of a plurality of impedance areas according to the value. Sorted. In other words, the sensor output signal is distributed to any one of a plurality of impedance areas based on the degree of influence of the underlying metal material. As described above, by providing a plurality of impedance areas in advance, it is possible to divide, that is, to reduce variations in the sensor output signal (X, Y). Therefore, in each impedance area, the film thickness index value obtained from the sensor output signal gradually decreases with the polishing time. Since such a plurality of impedance areas can be set for each region in the substrate surface, film thickness information in each region in the substrate surface can be acquired. Therefore, the polishing end point can be detected for each of a plurality of regions in the surface of the substrate.

  According to the second and fifth aspects of the present invention described above, the film thickness index value is acquired when the eddy current sensor scans the surface of the substrate along the same locus. Therefore, regardless of the presence of the metal material in the lower layer, the film thickness index value at each measurement point on the surface of the substrate decreases with the polishing time. That is, film thickness information in each region on the substrate surface can be acquired. Therefore, the polishing end point can be detected for each of a plurality of regions in the surface of the substrate.

  According to the third aspect of the present invention described above, the polishing of the substrate can be monitored on the basis of the convex portion caused by the remaining film. Therefore, it is possible to detect the precise polishing end point by eliminating the influence of the lower layer metal material.

It is a schematic diagram which shows the whole structure of a grinding | polishing apparatus. It is a top view which shows the relationship between a polishing table, an eddy current sensor, and a board | substrate. It is a figure which shows the equivalent circuit for demonstrating the principle of an eddy current sensor. It is a figure which shows the graph drawn by plotting X and Y which change with grinding | polishing time on an XY coordinate system. It is a figure which shows the graph which rotated the graph figure of FIG. 4 90 degree | times counterclockwise, and also translated it. It is a schematic diagram which shows an eddy current sensor. The structural example of the sensor coil in the eddy current sensor shown in FIG. 6 is shown. It is a schematic diagram which shows the detailed structure of an eddy current sensor. It is the figure which showed the locus | trajectory which an eddy current sensor scans a board | substrate. It is a figure which shows a mode that the film thickness index value obtained from the output signal of an eddy current sensor changes under the influence of lower wiring. FIG. 11A is a diagram showing an impedance curve when there is no influence of the underlying wiring structure, and FIG. 11B is a diagram showing a film thickness index value obtained from the impedance curve shown in FIG. is there. FIG. 12A is a diagram showing an impedance curve when there is an influence of the lower wiring structure, and FIG. 12B is a diagram showing a film thickness index value obtained from the impedance curve shown in FIG. is there. It is a figure which shows the example which divided | segmented the wide impedance curve shown in FIG.12 (b) into four impedance areas. It is a figure which shows the change of the film thickness index value determined from the coordinates X and Y which belong to each impedance area shown in FIG. It is a figure which shows the state in which the 1st-3rd offset impedance area overlapped on the reference | standard impedance area. It is a figure which shows the change of the film thickness index value determined from the coordinates X and Y which belong to each of four overlapping impedance areas shown in FIG. It is a figure which shows five area | regions defined in the surface of a board | substrate. It is a figure for demonstrating the method of calculating angle (theta) from the output signals X and Y of an eddy current sensor as a film thickness index value. It is a figure which shows the example from which angle (theta) changes resulting from presence of a lower layer wiring structure. It is a figure explaining the example which multiplies a coefficient to the angle calculated about the offset impedance area. It is a figure which shows the locus | trajectory on the board | substrate which an eddy current sensor draws when the rotational speed of a top ring is 77 min < ^-1 > and the rotational speed of a polishing table is 70 min < ^-1 >. It is a figure which shows the change of the film thickness profile on the same locus | trajectory of an eddy current sensor. It is a figure which shows the remaining film which exists on a board | substrate, and the film thickness profile of this board | substrate. It is a figure which shows the film thickness profile of the board | substrate which has both a lower layer wiring structure and a residual film. It is a schematic diagram which shows a table rotation detector and a top ring rotation detector. It is a time chart which shows a mode that a time measuring device receives a trigger signal and measures the rotation time of a polishing table and a top ring. It is a flowchart which shows the process of detecting a grinding | polishing end point. It is a figure for demonstrating the specific example of step 2-step 5 of FIG. It is sectional drawing which shows an example of the top ring shown in FIG.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a schematic diagram showing the overall configuration of the polishing apparatus. As shown in FIG. 1, the polishing apparatus includes a polishing table 1 and a top ring 10 that holds a substrate W that is an object to be polished and presses the polishing pad 2 on the polishing table 1. The polishing table 1 is connected to a table rotation motor (not shown) disposed below the table via a table shaft 1a, and is rotatable about the table shaft 1a.

  A polishing pad 2 is attached to the upper surface of the polishing table 1, and the surface of the polishing pad 2 constitutes a polishing surface 2a for polishing the substrate W. A polishing liquid supply nozzle 3 is installed above the polishing table 1, and the polishing liquid (slurry) is supplied to the polishing pad 2 on the polishing table 1 by the polishing liquid supply nozzle 3. An eddy current sensor 50 is embedded in the polishing table 1. The eddy current sensor 50 is connected to the polishing monitoring unit 53, and an output signal of the eddy current sensor 50 is sent to the polishing monitoring unit 53.

  The top ring 10 is connected to a top ring shaft 11, and the top ring shaft 11 moves up and down with respect to the top ring head 12. By moving the top ring shaft 11 up and down, the entire top ring 10 is moved up and down relative to the top ring head 12 for positioning. The top ring shaft 11 is rotated by driving a top ring rotation motor (not shown). The top ring 10 rotates about the top ring shaft 11 by the rotation of the top ring shaft 11.

The top ring 10 can hold a substrate (for example, a semiconductor wafer) W on its lower surface. The top ring head 12 is configured to be pivotable about the top ring head shaft 13, and the top ring 10 holding the substrate W on the lower surface is located above the polishing table 1 from the substrate receiving position by the rotation of the top ring head 12. It can be moved to the position. The top ring 10 holds the substrate W on the lower surface and presses the substrate W against the polishing surface 2 a of the polishing pad 2. At this time, the polishing table 1 and the top ring 10 are rotated, and the polishing liquid is supplied onto the polishing pad 2 from the polishing liquid supply nozzle 3. For the polishing liquid, a polishing liquid containing ceria (CeO ₂ ) or silica (SiO ₂ ) as abrasive grains is used. In this way, while supplying the polishing liquid onto the polishing pad 2, the substrate W is pressed against the polishing pad 2 to move the substrate W and the polishing pad 2 relative to each other to polish a conductive film such as a metal film on the substrate. . Specific examples of the metal film include a Cu film, a W film, a Ta film, and a Ti film.

  As shown in FIG. 1, the polishing apparatus includes a dressing apparatus 20 that dresses the polishing pad 2. The dressing device 20 includes a dresser arm 21, a dresser 22 rotatably attached to the tip of the dresser arm 21, a swing shaft 23 connected to the other end of the dresser arm 21, and a dresser centered on the swing shaft 23. A motor (not shown) as a drive mechanism for swinging the arm 21 is provided. The lower part of the dresser 22 is constituted by a dressing member 22a. The dressing member 22a has a circular dressing surface, and hard particles are fixed to the dressing surface by electrodeposition or the like. Examples of the hard particles include diamond particles and ceramic particles.

  A motor (not shown) is built in the dresser arm 21, and the dresser 22 is rotated by this motor. The swing shaft 23 is connected to an elevating mechanism (not shown), and the dresser member 21a is dressed by pressing the polishing surface 2a of the polishing pad 2 when the dresser arm 21 is lowered by the elevating mechanism. The dressing apparatus 20 can dress the polishing pad 2 when the substrate is not being polished, and can also dress the polishing pad 2 while the substrate is being polished.

  FIG. 2 is a plan view showing a relationship among the polishing table 1, the eddy current sensor 50, and the substrate W. As shown in FIG. As shown in FIG. 2, the eddy current sensor 50 is installed at a position passing through the center Cw of the substrate W being polished held by the top ring 10. The symbol CT is the rotation center of the polishing table 1. The eddy current sensor 50 can continuously detect the thickness of the conductive film of the substrate W on the passage locus (scanning line) while passing below the substrate W.

The eddy current sensor 50 induces an eddy current in the conductive film by causing a high-frequency alternating current to flow through the coil, and detects the thickness of the conductive film from a change in impedance caused by the magnetic field of the eddy current. FIG. 3 is a diagram showing a circuit for explaining the principle of the eddy current sensor. When a high-frequency AC current I _{1 is} passed from the AC power source E to the coil Q, the magnetic lines of force induced in the coil Q pass through the conductive film. Thus, mutual inductance is generated between the sensor-side circuit and the conductive film-side circuit, an eddy current I ₂ flows through the conductive film. The eddy current I ₂ generates magnetic lines of force, which change the impedance of the sensor side circuit. The eddy current sensor detects the film thickness of the conductive film from the change in impedance of the sensor side circuit.

In the sensor side circuit and the conductive film side circuit shown in FIG.
R ₁ I ₁ + L ₁ dI ₁ / dt + MdI ₂ / dt = E (1)
R ₂ I ₂ + L ₂ dI ₂ / dt + MdI ₁ / dt = 0 (2)
Here, M is a mutual inductance, R ₁ is an equivalent resistance of the sensor side circuit including the coil 1, and L ₁ is a self inductance of the sensor side circuit including the coil 1. R ₂ is an equivalent resistance corresponding to eddy current loss, and L ₂ is a self-inductance of the conductive film through which the eddy current flows.

Here, if I _n = A _n e ^jωt (sinusoidal wave), the above equations (1) and (2) are expressed as follows.
(R ₁ + jωL ₁ ) I ₁ + jωMI ₂ = E (3)
(R ₂ + jωL ₂ ) I ₂ + jωMI ₁ = 0 (4)
From these equations (3) and (4), the following equation is derived.
I ₁ = E (R ₂ + jωL ₂ ) / {(R ₁ + jωL ₁ ) (R ₂ + jωL ₂ ) + ω ² M ² }
= E / {(R ₁ + jωL ₁ ) + ω ² M ² / (R ₂ + jωL ₂ )} (5)

Therefore, the impedance Φ of the sensor side circuit is expressed by the following equation.
Φ = E / I ₁
= {R ₁ + ω ² M ² R ₂ / (R ₂ ² + ω ² L ₂ ² )}
+ Jω {L ₁ −ω ² L ₂ M ² / (R ₂ ² + ω ² L ₂ ² )} (6)
Here, when the real part (resistance component) and the imaginary part (inductive reactance component) of Φ are set as X and Y, respectively, the above equation (6) becomes as follows.
Φ = X + jωY (7)

FIG. 4 is a diagram showing a graph drawn by plotting X and Y that change with the polishing time on the XY coordinate system. The coordinate system of FIG. 4 is a coordinate system in which the Y axis is the vertical axis and the X axis is the horizontal axis. The coordinates of the point T∞ are the values of X and Y when the film thickness is infinite, that is, when R ₂ is 0, and the coordinates of the point T0 are assumed that the conductivity of the substrate can be ignored. The values of X and Y when the film thickness is 0, that is, when R ₂ is infinite. The point Tn positioned from the values of X and Y advances toward the point T0 while drawing an arc-shaped locus as the film thickness decreases. The symbol k shown in FIG. 4 is a coupling coefficient, and the following relational expression is established.
M = k (L ₁ L ₂ ) ^1/2 (8)

  FIG. 5 is a diagram showing a graph obtained by rotating the graph figure of FIG. 4 by 90 degrees counterclockwise and further translating it. That is, the point represented by the coordinates X and Y is rotated counterclockwise around the origin O of the XY coordinate system, and the rotated coordinates are moved so that the distance between the origin O and the coordinates X and Y is Generate a graph that shortens with decreasing film thickness. Hereinafter, the arc drawn by the coordinates X and Y is referred to as an impedance curve.

  Next, the eddy current sensor 50 will be described in more detail. FIG. 6 is a schematic diagram showing an eddy current sensor. The eddy current sensor 50 includes a sensor coil 102, an AC power source 103 connected to the sensor coil 102, and a resistance component X and an inductive reactance component Y of an electric circuit (sensor side circuit in FIG. 3) including the sensor coil 102. And a synchronous detection unit 105 for detection. The conductive film mf whose thickness is to be detected is a thin film made of a conductive material such as copper, tungsten, tantalum, or titanium formed on the substrate W, for example. The distance G between the sensor coil 102 and the conductive film mf is set to 0.5 mm to 5 mm, for example.

  FIG. 7 shows a configuration example of a sensor coil in the eddy current sensor shown in FIG. The sensor coil 102 includes three layers of coils 112, 113, and 114 wound around a bobbin 111. The central coil 112 is an exciting coil connected to the AC power source 103. The exciting coil 112 forms a magnetic field by an alternating current supplied from the alternating current power source 103 and generates an eddy current in the conductive film on the wafer. A detection coil 113 is disposed on the upper side (conductive film side) of the excitation coil 112 to detect magnetic flux generated by eddy current flowing through the conductive film. A balance coil 114 is disposed on the side opposite to the detection coil 113.

  The coils 113 and 114 are preferably formed by coils having the same number of turns (1 to 500), but the number of turns of the coil 112 is not particularly limited. The detection coil 113 and the balance coil 114 are connected to each other in opposite phases. When the conductive film exists in the vicinity of the detection coil 113, the magnetic flux generated by the eddy current formed in the conductive film is linked to the detection coil 113 and the balance coil 114. At this time, since the detection coil 113 is arranged closer to the conductive film, the balance of the induced voltages generated in the coils 113 and 114 is lost, thereby detecting the interlinkage magnetic flux formed by the eddy current of the conductive film. can do.

  FIG. 8 is a schematic diagram showing a detailed configuration of the eddy current sensor. The AC power supply 103 has a fixed frequency oscillator made of a crystal oscillator, and supplies, for example, an AC current having a fixed frequency of 1 to 50 MHz to the sensor coil 102. The alternating current formed by the alternating current power supply 103 is supplied to the sensor coil 102 via the band pass filter 120. A signal output from the terminal of the sensor coil 102 is sent to the synchronous detection unit 105 including the cos synchronous detection circuit 125 and the sin synchronous detection circuit 126 via the bridge circuit 121 and the high frequency amplifier 123. Here, from the oscillation signal formed by the AC power source 103, two signals of the in-phase component (0 °) and the quadrature component (90 °) of the AC power source 103 are formed by the phase shift circuit 124, and each is a cos synchronous detection circuit. 125 and the sin synchronous detection circuit 126. Then, the synchronous detection unit 105 extracts the impedance resistance component and the inductive reactance component.

  From the resistance component and the inductive reactance component output from the synchronous detection unit 105, unnecessary high frequency components (for example, a high frequency component of 5 KHz or more) are removed by the low-pass filters 127 and 128, and the signal X as the impedance resistance component and the inductive reactance are removed. A signal Y as a component is output. The polishing monitoring unit 53 processes the output signals X and Y of the eddy current sensor 50 by the same method as the processing (rotation processing, parallel movement processing, etc.) described in FIG. 5, and the distance Z (FIG. 5) is calculated. Then, the change in the film thickness is monitored based on the change in the distance Z. Note that predetermined processing such as rotation processing and parallel processing for the output signals X and Y of the eddy current sensor 50 may be performed electrically by the eddy current sensor 50 or by calculation by the polishing monitoring unit 53. Also good.

  Each time the eddy current sensor 50 crosses the surface of the substrate, the eddy current sensor 50 measures the film thickness of the substrate W at a plurality of measurement points. FIG. 9 is a diagram showing a trajectory that the eddy current sensor 50 scans the substrate W. FIG. As described above, when the polishing table 1 rotates, the eddy current sensor 50 scans the surface of the substrate W while drawing a trajectory passing through the center Cw of the substrate W. Since the rotation speed of the top ring 10 and the rotation speed of the polishing table 1 are usually different, the locus of the eddy current sensor 50 on the surface of the substrate W is scanned with the rotation of the polishing table 1 as shown in FIG. Lines SL1, SL2, SL3,. Even in this case, as described above, since the eddy current sensor 50 is disposed at a position passing through the center Cw of the substrate W, the locus drawn by the eddy current sensor 50 passes through the center Cw of the substrate W every time. In the present embodiment, the film thickness measurement timing by the eddy current sensor 50 is adjusted, and the film thickness at the center Cw of the substrate W is always measured by the eddy current sensor 50. In FIG. 9, symbol MPm-n represents the nth measurement point on the mth scan line SLm.

The output signals X and Y of the eddy current sensor 50 obtained at each measurement point are drawn on the XY coordinate system as coordinates X and Y. The output signals X and Y of the eddy current sensor 50 change according to the film thickness. Specifically, as shown in FIG. 5, the distance Z (= √ (X ² + Y ² )) between the origin O of the XY coordinate system and the point Tn specified from the coordinates X and Y has a film thickness. It becomes smaller as it decreases. Therefore, the distance Z obtained from the output signals X and Y can be referred to as a measured film thickness index value.

  As described above, the film thickness of the substrate W can be obtained from the output signals X and Y of the eddy current sensor 50, but the output signal of the eddy current sensor 50 varies greatly due to the influence of the metal material existing in the lower layer of the film. There are things to do. A substrate having a multilayer wiring structure has wiring (metal material) at each level. For this reason, the lower layer wiring affects the output signal of the eddy current sensor 50, and prevents accurate film thickness measurement.

  FIG. 10 is a diagram showing how the film thickness index value Z obtained from the output signal of the eddy current sensor 50 changes under the influence of the lower layer wiring. A plurality of wiring structures (for example, integrated circuits) 200 are formed under the film to be polished. Since these wiring structures 200 are covered with a film to be polished, they are indicated by dotted lines in FIG. The film thickness index value Z gradually decreases as a whole with the polishing time. However, the eddy current sensor 50 senses the film to be polished and the wiring structure 200 below the film, and as a result, the signal value of the eddy current sensor 50 is affected by the wiring structure 200 below.

  As shown in FIG. 10, the locus of the eddy current sensor 50 when the polishing table 1 is N-th rotation is different from the locus of the eddy current sensor 50 when the polishing table 1 is N + 1-th rotation. For this reason, the arrangement of the wiring structure 200 sensed by the eddy current sensor 50 varies depending on the rotation of the polishing table 1, and as a result, the film thickness profile generated from the film thickness index value Z (the film thickness along the radial direction of the substrate). Distribution) is also different. Thus, the output signals X and Y of the eddy current sensor 50 are affected by the lower wiring structure 200, and the impedance curve drawn on the XY coordinate system greatly fluctuates.

  FIG. 11A shows an impedance curve when there is no influence of the underlying wiring structure, and FIG. 11B shows a film thickness index value Z obtained from the impedance curve shown in FIG. When there is no influence of the underlying wiring structure, the impedance curve fluctuates to some extent due to system noise, but its width (indicated by dw) is small. Similarly, as shown in FIG. 11B, the width of the line drawn by the film thickness index value Z is also small. In the graph of FIG. 11B, the vertical axis represents the film thickness index value, and the horizontal axis represents the polishing time. Since the film thickness index value Z decreases with the polishing time, it is easy to detect the point where the film is removed, that is, the polishing end point.

  On the other hand, FIG. 12A shows an impedance curve when there is an influence of the lower wiring structure, and FIG. 12B shows a film thickness index value Z obtained from the impedance curve shown in FIG. . When there is an influence of the lower wiring structure, the impedance curve fluctuates greatly, and as a result, the width dw ′ of the impedance curve increases. Similarly, the line drawn by the film thickness index value Z is also wide. As a result, it becomes difficult to detect the polishing end point.

  Therefore, in the present embodiment, the wide impedance curve shown in FIG. 12A is divided into a plurality of regions (hereinafter referred to as impedance areas) along the longitudinal direction, and the film thickness is determined for each of the divided impedance areas. The index value is calculated, and the polishing of the substrate is monitored based on the film thickness index value for each impedance area. FIG. 13 is a diagram showing an example in which the wide impedance curve shown in FIG. 12A is divided into four impedance areas. In the following description, the narrow impedance curve shown in FIG. 11A is referred to as a reference impedance area, and the wide impedance curve shown in FIG. 12A is referred to as an initial impedance area, which is divided as shown in FIG. Among the impedance areas, an impedance area other than the reference impedance area is referred to as an offset impedance area.

  The four impedance areas, that is, the reference impedance area R0, the first offset impedance area R1, the second offset impedance area R2, and the third offset impedance area R3 have the same width, and the width is This is the width dw of the reference impedance area R0 obtained under the condition that there is no influence of the lower wiring structure. However, the width of the reference impedance area R0 and the width of the offset impedance areas R1 to R3 may be slightly different.

  The reference impedance area R0 is generated by using only the output signal (X, Y) of the eddy current sensor acquired at the center of the substrate. The eddy current sensor 50 always passes through the center of the substrate every time the polishing table 1 rotates once. Therefore, the film thickness index value Z acquired at the center of the substrate decreases with the polishing time regardless of the presence of the underlying metal material (wiring structure, etc.). In other words, the lower layer metal material does not affect the temporal change of the film thickness index value Z at the center of the substrate. Therefore, a narrow impedance area as shown in FIG. 11A can be generated from the sensor output signal (X, Y) obtained at the center of the substrate. The impedance area obtained at the center of the substrate is defined as the reference impedance area.

  The width dw of the reference impedance area is a difference between the minimum distance and the maximum distance from the center of the arc. More specifically, the width dw of the reference impedance area is obtained by obtaining the center of the arc drawn by the reference impedance area by a known method such as the least square method and obtaining the difference between the minimum distance and the maximum distance from the center. Desired. Similarly, the width dw ′ of the wide initial impedance area shown in FIG. 12A is also obtained by calculation. The initial impedance area is divided based on the width dw of the reference impedance area. The number of impedance areas to be divided is determined by the width dw ′ of the initial impedance area. That is, the number of impedance areas to be divided is determined by dividing the width dw ′ of the initial impedance area by the width dw of the reference impedance area. In the example shown in FIG. 13, four impedance areas R0, R1, R2, and R3 including the reference impedance area are created. Depending on the width of the initial impedance area, one of the impedance areas R1, R2, and R3 may have a slightly different width.

  The reference impedance area R0 and the offset impedance areas R1, R2, and R3 are acquired in advance by polishing a substrate having the same structure as the substrate to be polished. Usually, impedance areas R0, R1, R2, and R3 are created in advance by polishing one of a plurality of substrates having the same structure belonging to one lot.

  The plurality of impedance areas created as described above are defined on the XY coordinate system. Each time the output signals X and Y of the eddy current sensor are acquired, the coordinates of the output signals X and Y are distributed to any of the four impedance areas R0, R1, R2, and R3 according to the values. In other words, the sensor output signals X and Y are distributed to any one of the impedance areas R0, R1, R2, and R3 based on the degree of influence of the underlying wiring structure.

  FIG. 14 is a diagram showing a change in the film thickness index value Z (distance from the origin O to the coordinates X, Y) determined from the coordinates X, Y belonging to each impedance area. The film thickness index value Z draws four lines as the polishing time elapses. These four lines correspond to the four impedance areas R0, R1, R2, and R3 (see FIG. 13) to which the coordinates X and Y belong. The polishing of the substrate is monitored based on each of the four film thickness index values corresponding to the four impedance areas, and the polishing end point is determined based on the change of each film thickness index value.

  Furthermore, as shown in FIG. 15, the offset impedance areas R1, R2, and R3 may be translated so as to overlap the reference impedance area R0. The distance that each offset impedance area moves in parallel is determined from the distance between the arc centers of the four impedance areas, that is, the width dw of the reference impedance area R0. Specifically, the first offset impedance area R1 is translated by a distance of width dw × 1, the second offset impedance area R2 is translated by a distance of width dw × 2, and the third offset impedance area R3 Is translated by a distance of width dw × 3. By such an operation, as shown in FIG. 15, the first to third offset impedance areas R1, R2, and R3 overlap the reference impedance area R0.

  FIG. 16 is a diagram showing a change in the film thickness index value Z (distance from the origin O to the coordinates X, Y) determined from the coordinates X, Y belonging to each of the four overlapping impedance areas shown in FIG. As can be seen from FIG. 16, the four film thickness index values change in the same manner with the polishing time, and the four film thickness index values at each polishing time are also compared with the film thickness index values shown in FIG. Get closer to each other.

  The polishing end point of the substrate is determined for each impedance area. That is, four film thickness index values corresponding to the four impedance areas are separately monitored during polishing, and a point in time at which each film thickness index value reaches a predetermined threshold is determined as a polishing end point. The threshold value is set for each of the four film thickness index values. A time point when at least one film thickness index value among the four film thickness index values reaches a predetermined threshold value can be set as a final polishing end point. For example, when the film thickness index value corresponding to the reference impedance area reaches a predetermined threshold or when the film thickness index value corresponding to all impedance areas reaches the threshold, the polishing end point is set. Can do. Further, a point in time when at least two film thickness index values reach a predetermined threshold value can be set as a polishing end point.

  The first to third offset impedance areas R1, R2, and R3 are regions where the output signal of the eddy current sensor is affected by the underlying wiring structure. However, since the output signal of the eddy current sensor is approximately the same within the same offset impedance area and is affected by the lower wiring structure, the change in the film thickness index value is not affected by the influence of the lower wiring structure. Reflects the progress. Therefore, the polishing end point detection accuracy can be improved by monitoring the polishing of the substrate for each of the plurality of divided impedance areas.

  The plurality of impedance areas described above can be created for each of a plurality of regions (zones) in the substrate surface. Therefore, the polishing end point can be determined for each region of the substrate according to the method described above. These regions can be arbitrarily set in the surface of the substrate. For example, in the example shown in FIG. 17, five regions C1, C2, C3, C4, and C5 are set in the surface of the substrate. In this case, since four film thickness index values are acquired for each of the five regions C1 to C5, 20 (5 × 4) film thickness index values are monitored during the polishing of the substrate.

  By providing a plurality of impedance areas including the reference impedance area in advance, the variation in the sensor output signal (X, Y) can be divided, that is, reduced. Therefore, in each impedance area, the film thickness index value obtained from the sensor output signal decreases with the polishing time. Since such a plurality of impedance areas can be set for each region in the substrate surface, film thickness information in each region in the substrate surface can be acquired. Therefore, the polishing end point can be detected for each of a plurality of regions in the surface of the substrate.

In the embodiment described above, the film thickness index value Z (√ (X ² + Y ² )) is calculated from the output signals X and Y of the eddy current sensor, but the film thickness index value as in other embodiments described below. The angle may be calculated from the output signals X and Y of the eddy current sensor. FIG. 18 is a diagram for explaining a method of calculating the angle from the output signals X and Y of the eddy current sensor as the film thickness index value. As shown in FIG. 18, a line connecting a reference line FL passing through a preset reference point (fixed point) F and a point Tn determined from an output signal (X component, Y component) of the eddy current sensor and the reference point F. Is changed as the point Tn moves, that is, as the film thickness decreases. Therefore, the angle θ can be used as an index indicating the film thickness.

  In general, as the number of substrates to be polished increases, the polishing pad 2 gradually wears. As shown in FIG. 1, since the eddy current sensor 50 is embedded in the polishing table 1, the distance between the substrate W and the eddy current sensor 50 changes with wear of the polishing pad 2. It is known that the angle θ described above does not depend on the distance between the substrate W and the eddy current sensor 50 but changes depending on the film thickness (see Patent Document 1).

  However, as shown in FIG. 19, the angle θ may change due to the presence of the underlying wiring structure, and the film thickness may not be accurately reflected. Therefore, as shown in FIG. 20, the offset impedance areas R1, R2, and R3 are calculated so that the angles calculated for the reference impedance area R0 and the angles calculated for the offset impedance areas R1, R2, and R3 are equal. Multiply the resulting angle by a factor. The coefficients are set in advance for the respective offset impedance areas R1, R2, and R3. Since these coefficients can vary depending on the structure of the substrate, they are determined from the polishing result of the substrate having the same structure as the substrate to be polished. Based on the angle thus corrected, the polishing end point can be detected for each of the impedance areas R0, R1, R2, and R3.

Next, still another embodiment of the present invention will be described.
As shown in FIG. 10, the locus of the eddy current sensor 50 when the polishing table 1 is N-th rotation is different from the locus of the eddy current sensor 50 when the polishing table 1 is N + 1-th rotation. The position of the wiring structure 200 sensed by the eddy current sensor 50 differs depending on the number of rotations of the polishing table 1, and as a result, the film thickness profile acquired by the eddy current sensor 50 changes depending on the locus of the eddy current sensor 50. .

  The rotational speed of the top ring 10 and the rotational speed of the polishing table 1 are usually different. Under such conditions, the locus drawn by the eddy current sensor 50 on the substrate surface rotates around the center of the substrate. The locus of the eddy current sensor 50 makes a round on the surface of the substrate while the polishing table 1 rotates a certain number of times. The number of rotations of the polishing table 1 necessary for the sensor trajectory to make a round on the surface of the substrate is determined by the rotation speed ratio between the top ring 10 and the polishing table 1.

FIG. 21 is a diagram showing a locus on the substrate W drawn by the eddy current sensor 50 when the rotation speed of the top ring 10 is 77 min ⁻¹ and the rotation speed of the polishing table 1 is 70 min ⁻¹ . As shown in FIG. 21, under this condition, the trajectory of the eddy current sensor 50 rotates 36 degrees each time the polishing table 1 makes one rotation. Accordingly, the trajectory of the eddy current sensor 50 makes a round on the surface of the substrate W every time the polishing table 1 rotates 10 times. In this case, the sensor locus when the polishing table 1 is rotating for the first time and the sensor locus when the polishing table 1 is rotating for the 11th time are the same.

  FIG. 22 is a diagram showing a change in the film thickness profile on the same locus of the eddy current sensor 50. The film thickness profile is a film thickness distribution along the radial direction of the substrate. When the eddy current sensor 50 measures the film thickness of the substrate along the same locus, a convex portion appears at the same position of the film thickness profile due to the wiring structure 200. Since the eddy current sensor 50 scans the same portion of the substrate, the convex portions appear at the same position. Therefore, regardless of the presence of the lower wiring structure 200, the film thickness profile as a whole gradually decreases with the polishing time. That is, the film thickness index value decreases with the polishing time at each film thickness measurement point on the substrate. Therefore, the polishing end point can be determined based on the change in the film thickness profile (film thickness index value).

  When the purpose of polishing is to remove the metal film formed on the substrate, the film thickness profile does not change when the metal film is removed from the substrate. This is because the eddy current sensor 50 no longer reacts to the metal film. Therefore, the time point when the film thickness profile does not change (specifically, the time point when the film thickness index value stops decreasing) can be determined as the polishing end point. For example, the polishing end point can be the time when the difference between the current film thickness index value at the same position on the substrate and the previous film thickness index value has decreased to a predetermined value.

  The change in the film thickness profile (that is, the change in the film thickness index value) can be monitored for each of a plurality of regions defined in advance on the surface of the substrate as shown in FIG. In each region on the substrate, the film thickness index value decreases with the polishing time regardless of the presence of the underlying wiring structure. Therefore, the film thickness index value obtained from the eddy current sensor 50 can be used to detect the polishing end point in each region of the substrate.

  When the rotational speed of the top ring 10 and the rotational speed of the polishing table 1 are different, there are a plurality of trajectories of the eddy current sensor 50 that scans the surface of the substrate. In the example shown in FIG. 21, the sensor trajectory makes one round around the center of the substrate every time the polishing table 1 rotates 10 times, so there are 10 trajectories. A film thickness profile shown in FIG. 22 may be created for each of these ten trajectories. Even in this case, the surface of the substrate can be divided into five regions as in the example shown in FIG. Therefore, in this case, 50 (10 × 5) polishing end points can be detected.

  The above-described example is a case where the wiring structure existing in the lower layer of the film affects the output signal of the eddy current sensor 50. However, the film thickness profile is also obtained when the film to be polished remains locally on the substrate. The convex part appears. This example will be described with reference to FIG. FIG. 23 is a diagram showing the remaining film locally present on the substrate and the film thickness profile of the substrate. Usually, the remaining film is an annular film as shown in FIG. If such a residual film exists on the substrate, the eddy current sensor 50 reacts to the residual film and the film thickness index value becomes large. As a result, as in the example shown in FIG. Part appears.

  The film thickness profile shown in FIG. 23 is different from the film thickness profile shown in FIG. 22 in that a convex portion always appears at a fixed position regardless of the locus of the eddy current sensor 50. This is because the remaining film is an annular film extending in the circumferential direction of the substrate. The convex portion of the film thickness profile indicating the presence of the remaining film appears at the same location each time the polishing table 1 rotates, gradually decreases with the polishing time, and disappears when the remaining film is removed.

  FIG. 24 is a diagram showing a film thickness profile of a substrate having both a lower wiring structure and a remaining film. The convex portions of the film thickness profile due to the lower layer wiring structure appear at different positions as long as the scanning trajectories of the eddy current sensor 50 do not match, whereas the convex portions of the film thickness profile due to the remaining film are polished. Each time the table 1 rotates, it appears at the same position. Therefore, the polishing monitoring unit 53 can determine from the position of the convex portion appearing in the film thickness profile whether the convex portion is caused by the remaining film or the underlying wiring structure. Furthermore, the polishing end point can be determined from the change in the size of the convex portion due to the remaining film. For example, the polishing end point can be the time when the size of the convex portion becomes 0, or when the convex portion becomes smaller than a predetermined threshold value.

  A convex portion caused by the lower wiring structure and a convex portion caused by the remaining film can be distinguished as follows. Each time a film thickness profile is acquired, the number of protrusions appearing in the film thickness profile and the position of the protrusions in the radial direction of the substrate are acquired. As can be seen from FIG. 24, the convex portions due to the remaining film appear continuously at substantially the same position (similar position) every time the polishing table 1 makes one rotation, regardless of the sensor locus. On the other hand, the convex part resulting from the lower wiring structure appears at substantially the same position (similar position) at a constant period. Accordingly, the convex portions appearing at substantially the same position in succession are determined to be convex portions due to the remaining film, while the convex portions appearing at substantially the same position at a constant period are convex portions due to the underlying wiring structure. Part.

  When the polishing table 1 and the top ring 10 are rotating at the same rotational speed, the eddy current sensor 50 always crosses the substrate along the same locus. Therefore, in such a case, the convex portions can be distinguished based on whether or not the peak value of each convex portion decreases with the polishing time. That is, when the peak value of the convex portion does not decrease when the polishing time reaches a certain point, the convex portion can be determined as the convex portion due to the lower wiring structure. On the other hand, when the peak value of the convex portion decreases (even gradually) with the polishing time, it can be determined that the convex portion is a convex portion due to the remaining film.

  As described above, the number of scanning trajectories of the eddy current sensor 50 is determined by the ratio between the rotational speed of the top ring 10 and the rotational speed of the polishing table 1. In other words, by setting the rotation speed of the top ring 10 and the rotation speed of the polishing table 1, the substrate can be scanned with the same locus by the eddy current sensor 50 each time the polishing table 1 rotates a desired number of times. . However, the top ring 10 and the polishing table 1 do not always rotate at the set rotation speed. That is, there is an error between the set rotation speed and the actual rotation speed. Although this error is small, due to the error, the eddy current sensor 50 does not scan the substrate in a predetermined locus. As a result, it is not possible to obtain a film thickness profile in which convex portions appear at the same position as shown in FIG.

  Therefore, in this embodiment, the actual time for each rotation of the top ring 10 and the polishing table 1 is measured, and the rotation speed of the top ring 10 and the rotation speed of the polishing table 1 are calculated from the actual measurement time. FIG. 25 is a schematic diagram showing a table rotation detector 210 that measures the time for which the polishing table 1 makes one rotation, and a top ring rotation detector 220 that measures the time for which the top ring 10 makes one rotation. The table rotation detector 210 includes a sensor target 211 fixed to the outer peripheral surface of the polishing table 1, a sensor 212 for sensing the sensor target 211, and a time measuring device 213 connected to the sensor 212.

  The sensor target 211 rotates together with the polishing table 1, while the position of the sensor 212 is fixed. The sensor 212 is disposed close to the sensor target 211 and senses the sensor target 211 every time the polishing table 1 makes one rotation. When the sensor 212 senses the sensor target 211, a trigger signal is sent from the sensor 212 to the time measuring device 213. The time measuring device 213 measures the time from when the trigger signal is received until the next trigger signal is received. This measured time is the time for which the polishing table 1 rotates once.

  The top ring rotation detector 220 includes a sensor target 221 fixed to the top ring 10, a sensor 222 that senses the sensor target 221, and a time measuring device 223 connected to the sensor 222. The sensor 222 is fixed to the top ring head 12 (see FIG. 1). Since the operation of the top ring rotation detector 220 is the same as the operation of the table rotation detector 210 described above, the description thereof is omitted.

  FIG. 26 is a time chart showing how each time measuring device receives the trigger signal and measures the actual time per rotation. When receiving the trigger signal, the time measuring devices 213 and 223 start measuring time, and when receiving the next trigger signal, the time measuring devices 213 and 223 stop measuring time and simultaneously start measuring time again. . Since the trigger signal is input to the time measuring device 213 every time the polishing table 1 makes one rotation, the time interval between the trigger signal and the next trigger signal is the actual rotation time of the polishing table 1. Similarly, since the trigger signal is input to the time measuring device 223 every time the top ring 10 makes one rotation, the time interval between the trigger signal and the next trigger signal is the actual rotation time of the top ring 10. .

Rotational speed of the polishing table 1 the rotational speed of ^{(min -1)} and the top ring 10 ^{(min -1)} can be calculated from the respective measured rotation time. Thus, by acquiring the actual rotation speed of the polishing table 1 and the actual rotation speed of the top ring 10, the rotation speed ratio between the top ring 10 and the polishing table 1 can be accurately adjusted. Therefore, the eddy current sensor 50 can accurately scan the surface of the substrate with the same locus every time the polishing table 1 rotates a predetermined number of times. Note that either the top ring rotation detector 220 or the table rotation detector 210 may be omitted. In this case, since the actual rotational speed of the top ring 10 or the polishing table 1 cannot be measured, the set rotational speed is used instead.

  Next, the process for detecting the polishing end point will be described with reference to FIG. FIG. 27 is a flowchart showing a process of detecting the polishing end point. When the polishing of the substrate is started, the eddy current sensor 50 scans the surface of the substrate every time the polishing table 1 rotates, and outputs a signal X as a resistance component of impedance and a signal Y as an inductive reactance component. The polishing monitoring unit 53 receives the film thickness data including the output signals X and Y from the eddy current sensor 50 (step 1).

  The polishing monitoring unit 53 receives the measurement values of the rotation time of the polishing table 1 and the rotation time of the top ring 10 from the time measuring devices 213 and 223 (step 2), and the actual rotation speed of the top ring 10 and the polishing as described above. The actual rotational speed of Table 1 is calculated. Further, the polishing monitoring unit 53 calculates the number of rotations of the polishing table 1 necessary for the eddy current sensor 50 to draw the same locus from the rotation speed ratio between the top ring 10 and the polishing table 1 (step 3).

  The polishing monitoring unit 53 divides the film thickness data into a plurality of film thickness data groups according to a plurality of predefined areas (see FIG. 17) on the substrate surface (step 4), and further vortexes the film thickness data groups for each area. According to the trajectory of the current sensor 50, it is distributed into a plurality of film thickness data (step 5), and a film thickness profile for each sensor trajectory is created from each film thickness data.

A specific example of Step 2 to Step 5 will be described with reference to FIG. In the example shown in FIG. 28, the actual time for one rotation of the top ring 10 is 2000 milliseconds, and the actual time for one rotation of the polishing table 1 is 1000 milliseconds. In this case, the rotation speed of the top ring 10 is determined to be 30 min ⁻¹ , and the rotation speed of the polishing table 1 is determined to be 60 min ^−1, and the eddy current sensor 50 crosses the surface of the substrate twice each time the polishing table 1 rotates once. Accordingly, in this case, the eddy current sensor 50 has two trajectories. On the surface of the substrate, five regions are defined in advance along each locus.

The film thickness data obtained when the polishing table 1 is rotated 2N-1 are D1 _2N-1 , D2 _2N-1 , D3 _2N-1 , D4 _2N-1 , D5 _2N-1 , and the polishing table 1 is The film thickness data obtained at the 2N rotation is D1 _2N , D2 _2N , D3 _2N , D4 _2N , D5 _2N . N is a natural number. These film thickness data are divided into five data groups respectively belonging to the five regions on the substrate, that is, the first data group D1 _2N-1 , D1 _2N , the second data group D2 _2N-1 , D2 _2N , the third data group. Data groups D3 _2N-1 , D3 _2N , fourth data groups D4 _2N-1 , D4 _2N , and fifth data groups D5 _2N-1 , D5 _2N .

Furthermore, each data group is distributed for each same sensor locus. That is, the first data group is divided into film thickness data D1 _2N-1 and film thickness data D1 _2N, and the second data group is divided into film thickness data D2 _2N-1 and film thickness data D2 _2N. The third data group is divided into film thickness data D3 _2N-1 and film thickness data D3 _2N, and the fourth data group is divided into film thickness data D4 _2N-1 and film thickness data D4 _2N . The data group 5 is divided into film thickness data D5 _2N-1 and film thickness data D5 _2N . Then, a film thickness profile is generated from each film thickness data.

  Returning to FIG. 27, the polishing monitoring unit 53 compares the current film thickness obtained from each film thickness profile with the previous film thickness, and acquires a change in the film thickness profile (step 6). Specifically, the polishing monitoring unit 53 determines whether or not the difference between the current film thickness and the previous film thickness is less than a set value, or whether the film thickness reduction rate is less than a set value. In order to increase the accuracy of the polishing end point detection, these set values are preferably determined based on the magnitude of the system noise. When the difference between the current film thickness and the previous film thickness falls below the set value, or when the film thickness reduction rate falls below the set value, the polishing monitoring unit 53 reaches the end point of the substrate polishing process. (Step 7).

  In order to further improve the accuracy of detection of the polishing end point, it is preferable to determine that the polishing process of the substrate has reached the end point when the polishing end point of step 7 is detected for a plurality of sensor loci. Alternatively, when the polishing end point detection in step 7 is performed a plurality of times while the polishing table 1 rotates a plurality of times, it is preferable to determine that the substrate polishing process has reached the end point.

  The above-described embodiments can be applied to a top ring that can independently press a plurality of regions of a substrate against a polishing pad. 29 is a cross-sectional view showing an example of the top ring shown in FIG. The top ring 10 includes a top ring body 251 that is connected to the top ring shaft 11 via a free joint 250, and a retainer ring 252 that is disposed below the top ring body 251.

  Below the top ring body 251, a flexible membrane 256 that contacts the substrate W and a chucking plate 257 that holds the membrane 256 are disposed. Four pressure chambers (airbags) P1, P2, P3, and P4 are provided between the membrane 256 and the chucking plate 257. The pressure chambers P1, P2, P3, and P4 are formed by the membrane 256 and the chucking plate 257. The central pressure chamber P1 is circular, and the other pressure chambers P2, P3, P4 are annular. These pressure chambers P1, P2, P3, and P4 are arranged concentrically.

  Pressurized fluid such as pressurized air is supplied to the pressure chambers P1, P2, P3, and P4 by the pressure adjusting unit 270 via the fluid paths 261, 262, 263, and 264, respectively, or evacuated. ing. The internal pressures of the pressure chambers P1, P2, P3, and P4 can be changed independently of each other, so that four regions of the substrate W, that is, a central portion, an inner intermediate portion, an outer intermediate portion, and a peripheral edge can be obtained. The pressing force on the part can be adjusted independently. Further, by raising and lowering the entire top ring 10, the retainer ring 252 can be pressed against the polishing pad 2 with a predetermined pressing force.

  A pressure chamber P5 is formed between the chucking plate 257 and the top ring body 251. Pressurized fluid is supplied to the pressure chamber P5 through the fluid path 265 by the pressure adjusting unit 270, or vacuuming is performed. It has come to be. As a result, the chucking plate 257 and the entire membrane 256 can move in the vertical direction. The peripheral end of the substrate W is surrounded by a retainer ring 252 so that the substrate W does not jump out of the top ring 10 during polishing. An opening is formed in a portion of the membrane 256 constituting the pressure chamber P3, and the substrate W is attracted and held on the top ring 10 by forming a vacuum in the pressure chamber P3. Further, the substrate W is released from the top ring 10 by supplying nitrogen gas or clean air to the pressure chamber P3.

  The polishing monitoring unit 53 sets the target value of the internal pressure of each pressure chamber P1, P2, P3, P4 based on the film thickness index value in the region of the substrate surface corresponding to each pressure chamber P1, P2, P3, P4. decide. The polishing monitoring unit 53 sends a command signal to the pressure adjusting unit 270 and controls the pressure adjusting unit 270 so that the internal pressures of the pressure chambers P1, P2, P3, and P4 coincide with the target value. As described above, the top ring 10 having a plurality of pressure chambers can independently press each region on the surface of the substrate against the polishing pad 2 as the polishing progresses, so that the film can be uniformly polished.

  The embodiment described above is described for the purpose of enabling the person having ordinary knowledge in the technical field to which the present invention belongs to implement the present invention. Various modifications of the above embodiment can be naturally made by those skilled in the art, and the technical idea of the present invention can be applied to other embodiments. Accordingly, the present invention is not limited to the described embodiments, but is to be construed in the widest scope according to the technical idea defined by the claims.

DESCRIPTION OF SYMBOLS 1 Polishing table 1a Table axis | shaft 2 Polishing pad 2a Polishing surface 3 Polishing liquid supply nozzle 10 Top ring 11 Top ring shaft 12 Top ring head 20 Dressing device 21 Dresser arm 22 Dresser 22a Dressing member 23 Oscillating shaft 50 Eddy current sensor 53 Polishing monitoring Unit 102 Sensor coil 103 AC power source 105 Synchronous detection unit 111 Bobbin 112 Excitation coil 113 Detection coil 114 Balance coil 120 Band pass filter 121 Bridge circuit 123 High frequency amplifier 124 Phase shift circuit 125 Cos synchronous detection circuit 126 Sin synchronous detection circuit 127, 128 Low pass Filter 200 Wiring structure 210 Table rotation detector 211 Sensor target 212 Sensor 213 Time measuring device 220 Top ring rotation detector 221 Sensor target 222 Sensor 223 Time measuring device 251 Top ring body 252 Retainer ring 256 Membrane 257 Chucking plate 270 Pressure adjusting part P1, P2, P3, P4 Pressure chamber

Claims (13)

A method of detecting a polishing end point of a substrate polishing step of polishing a film of the substrate by rotating the top ring and the polishing table while pressing the substrate against a polishing pad on a polishing table by a top ring. ,
Moving the eddy current sensor across the surface of the substrate during polishing of the substrate;
Obtaining the resistance component X and the inductive reactance component Y of the impedance of the eddy current sensor;
Plotting the coordinates X, Y consisting of the resistance component X and the inductive reactance component Y on an XY coordinate system;
A plurality of impedance areas are defined in advance on the XY coordinate system, and the plurality of impedance areas include a reference impedance area and at least one offset impedance area,
Using a plurality of coordinates X, Y belonging to each of the plurality of impedance areas, a plurality of film thickness index values are calculated for each of the plurality of impedance areas,
A method of determining a polishing end point of the substrate for each of the plurality of impedance areas using the plurality of film thickness index values. The plurality of impedance areas are:
Polishing a substrate having the same structure as the substrate;
Obtaining the resistance component X and the inductive reactance component Y during polishing of the substrate of the same structure;
Plotting the obtained coordinates X and Y consisting of the resistance component X and the inductive reactance component Y on the XY coordinate system to form an initial impedance area on the XY coordinate system;
The method according to claim 1, wherein the initial impedance area is obtained by dividing the initial impedance area along a longitudinal direction thereof. The reference impedance area is
Polishing a substrate having the same structure as the substrate;
Obtaining the resistance component X and the inductive reactance component Y at the center of the substrate having the same structure;
It is an area drawn on the XY coordinate system by plotting the obtained coordinates X, Y consisting of the resistance component X and the inductive reactance component Y on the XY coordinate system. The method according to claim 1 or 2.   The method according to claim 1, wherein the offset impedance area has the same width as the reference impedance area.   5. The method according to claim 1, wherein the film thickness index value is a distance between a point specified by the coordinates X and Y and an origin of the XY coordinate system. .   The method according to claim 1, wherein the offset impedance area is translated until the offset impedance area overlaps the reference impedance area.   The film thickness index value is an angle formed by a straight line connecting a point specified by the coordinates X and Y and a predetermined reference point and a predetermined reference line passing through the reference point. The method according to any one of 1 to 4.   The coefficient obtained by multiplying the angle obtained for the offset impedance area by a coefficient that makes the angle obtained for the reference impedance area equal to the angle obtained for the offset impedance area. 8. The method according to 7. A method of detecting a polishing end point of a substrate polishing step of polishing a film of the substrate by rotating the top ring and the polishing table while pressing the substrate against a polishing pad on a polishing table by a top ring. ,
Measure the time for which the top ring makes one rotation, calculate the rotation speed of the top ring from the measured time,
From the ratio of the rotation speed of the top ring and the rotation speed of the polishing table, the number of rotations of the polishing table for the eddy current sensor embedded in the polishing table to draw the same trajectory and cross the surface of the substrate is determined. Calculate
Before the Kiuzu current sensor is moved across the surface of the substrate,
Obtaining an output signal of the eddy current sensor when the eddy current sensor crosses the surface of the substrate in the same locus;
Calculate the film thickness index value from the output signal of the eddy current sensor,
A method of determining a polishing end point of the substrate from a change in the film thickness index value. The method according to claim 9 , wherein the time for which the polishing table rotates once is measured, and the rotation speed of the polishing table is calculated from the measured time. The output signal of the eddy current sensor is distributed according to a plurality of predefined areas on the surface of the substrate,
The method according to claim 9 or 10, characterized in that from the distribution obtained output signal, calculates the thickness index value for each area of the substrate. A method of detecting a polishing end point of a substrate polishing step of polishing a film of the substrate by rotating the top ring and the polishing table while pressing the substrate against a polishing pad on a polishing table by a top ring. ,
Moving the eddy current sensor embedded in the polishing table across the surface of the substrate;
Obtaining an output signal of the eddy current sensor;
Create a film thickness profile from the output signal of the eddy current sensor,
From the change in the position of the convex portion appearing in the film thickness profile, it is determined whether the convex portion appears due to either the remaining film or the metal material present in the lower layer of the film,
A polishing end point of the substrate is determined based on a size of a convex portion that appears due to the remaining film. A rotatable polishing table that supports the polishing pad;
A top ring that presses the substrate against a polishing pad on the rotating polishing table while rotating the substrate;
An eddy current sensor installed in the polishing table and moving across the surface of the substrate;
A polishing monitoring unit that monitors the film thickness of the substrate from the output signal of the eddy current sensor,
The polishing monitoring unit
Obtaining the resistance component X and the inductive reactance component Y of the impedance of the eddy current sensor;
The coordinates X and Y composed of the resistance component X and the inductive reactance component Y are plotted on an XY coordinate system, and a plurality of impedance areas are defined in advance on the XY coordinate system. The impedance area includes a reference impedance area and at least one offset impedance area;
Using a plurality of coordinates X, Y belonging to each of the plurality of impedance areas, a plurality of film thickness index values are calculated for each of the plurality of impedance areas,
A polishing apparatus, wherein a polishing end point of the substrate is determined for each of the plurality of impedance areas using the plurality of film thickness index values.

Images