JP2007266068A - Polishing method and device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a polishing device and a polishing method which can provide polished article having a highly flat plane. <P>SOLUTION: A polishing pad 2 is provided on one side of a turn table 1 which can rotate in a predetermined direction, a spindle head 5 is provided at a position eccentric with respect to the rotating shaft of the turn table 1, and a back pad 6 is provided on the surface of the spindle head 5 opposite to the polishing pad 2. A wafer W is held by the back pad 6 as a thin plate polished article which is subjected to surface polishing through relative rotation with the polishing pad 2. The compression rate of the resin layer of the back pad 6 is 6-25%, and the product of the compression rate of the resin layer and the compression modulus of the polishing pad 2 is in the range of 600-2,100. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a polishing technique for an object to be polished, and more particularly, to a polishing method and an apparatus for polishing a surface of an object that requires a high level flat surface such as a semiconductor substrate or a precision glass substrate. is there.

  2. Description of the Related Art Conventionally, a semiconductor wafer (silicon wafer) used as a raw material wafer for manufacturing a semiconductor device is generally polished on its surface so as to be a mirror wafer.

  A wafer used as a raw material wafer for manufacturing a semiconductor device is a single crystal semiconductor ingot grown by the Czochralski method (CZ method) or the floating zone melting method (FZ method), and the outer periphery of the grown semiconductor ingot. Is shaped by grinding with a cylindrical grinder or the like, and is sliced with a wire saw in a slicing step to form a thin plate.

Thereafter, the wafer peripheral part is chamfered in a chamfering process, and after a flattening process and an etching process in a lapping process, after primary polishing and secondary polishing, a mirror surface wafer is obtained.
The mirror surface wafer obtained through such a process becomes a semiconductor device by forming a circuit on its surface.

  However, when the surface flatness of the mirror-finished wafer manufactured through the above steps is low, the lens is partially out of focus at the time of exposure in the photolithography process when forming the circuit, and it becomes difficult to form a fine pattern of the circuit. The problem arises. For this reason, extremely high flatness is required in recent high-precision device fabrication.

  In order to manufacture a mirror-finished wafer having such extremely high flatness, the surface polishing of the wafer is very important. As one of polishing apparatuses for performing the surface polishing of the wafer, a single-side polishing apparatus called a sheet type is known.

FIG. 8 shows an example of such a sheet-type single-side polishing apparatus. 8A is a perspective view of the polishing apparatus, and FIG. 8B is a longitudinal sectional view of the polishing apparatus.
In FIG. 8, a polishing apparatus 10 is disposed on a polishing surface plate 1 that rotates in a predetermined direction (for example, counterclockwise direction in the drawing) around the rotation shaft 3, and is disposed on the polishing surface plate 1. And a plurality of spindle heads 5 rotating in the counterclockwise direction in the figure.

  The upper surface of the polishing surface plate 1 is a flat surface, and a polishing pad 2 for polishing an object to be polished is attached and fixed with an adhesive or the like. For example, when the diameter of the wafer W is 200 mm, the polishing pad 2 is formed of a nonwoven fabric having a diameter of 620 mm. At the center of the bottom surface of the polishing surface plate 1, a rotating shaft 3 for rotating the polishing surface plate 1 by power transmission from a rotation drive device (not shown) is fixed.

  The load on the spindle head 5 is controlled by a weight support shaft 4 connected to the upper surface of the spindle head 5. The spindle head 5 is in a position eccentric from the axis of the rotary shaft 3 and is given a rotational driving force by a rotational driving device (not shown). Although FIG. 8 shows an example in which two spindle heads 5 are provided for one polishing pad 2, the number of spindle heads 5 is not limited to this and may be one or more.

  A back pad 6 is bonded and fixed to the bottom surface of the spindle head 5, and the back pad 6 is disposed so as to face the bottom surface of the spindle head 5 in parallel with the surface of the polishing pad 2. Around the wafer W held by the back pad 6, there is a retainer ring 7 arranged in an annular shape so as to surround the outer periphery of the wafer W. The retainer ring 7 is fixed to the lower surface of the back pad 6 along the outer periphery of the back pad 6. The retainer ring 7 includes a type that is fixed to the back pad 6 and a type that is fixed to the spindle head 5.

  As shown in FIG. 7, the back pad 6 is mainly composed of a base material 6a serving as a base for bonding to the spindle head 5 and a resin layer 6b not including the base material. When the diameter of the wafer W is 200 mm, the diameter is 245 mm. The thickness of the resin layer 6b is about 300 to 700 μm.

In order to polish the wafer W, the wafer W is disposed so that the surface to be polished faces the polishing pad 2, and the wafer W is adsorbed on the lower surface of the resin layer 6 b of the back pad 6. Then, the polishing surface plate 1 is rotated and the spindle head 5 holding the wafer W is rotated. In this state, the spindle head 5 is lowered to press the wafer W with a predetermined pressure on the upper surface of the polishing pad 2 to polish the surface to be polished of the wafer W. At this time, slurry is supplied from a nozzle (not shown) as necessary.
JP 2002-355755 A

  Since the polishing cloth 2 generally has elasticity, if the polishing is performed while pressing the wafer W against the polishing cloth 2, the wafer W slightly sinks into the polishing cloth 2. Similarly, since the resin layer 6 b of the back pad 6 is also elastic, if the wafer W is polished while being pressed against the polishing cloth 2, the wafer W slightly sinks into the back pad 6.

  FIG. 7A and FIG. 7B are enlarged longitudinal sectional views of a part of the polishing apparatus. 7A shows a case where the compression rate of the resin layer 6b of the back pad 6 is large, and FIG. 7B shows a case where the compression rate of the resin layer 6b of the back pad 6 is small.

  As shown in FIG. 7A, when the compressibility of the resin layer 6b is large, the sinking amount ct of the wafer W into the polishing pad 2 decreases, and the sinking amount bt into the back pad 6 increases. In this case, the surface pressure from the retainer ring 7 to the polishing pad 2 increases, and the surface pressure from the polishing pad 2 to the vicinity of the outer periphery of the wafer W decreases. As a result, the polishing allowance in the vicinity of the outer periphery of the wafer W is reduced, and the polishing allowance of the inner portion (hereinafter referred to as “inner peripheral portion”) relative to the vicinity of the outer periphery of the wafer W is relatively increased. Variation occurs in the machining allowance in the plane of W.

  On the other hand, as shown in FIG. 7B, when the compressibility of the resin layer 6b is small, the sinking amount bt of the wafer W into the back pad 6 decreases, and the sinking amount ct into the polishing pad 2 increases. . In this case, the surface pressure from the retainer ring 7 to the polishing pad 2 decreases, and the surface pressure from the polishing pad 2 to the vicinity of the outer periphery of the wafer W increases. As a result, the inner peripheral portion of the wafer W can have a relatively stable flatness, but the polishing allowance of the outer peripheral portion of the wafer W (especially in a range of 10 mm from the outer peripheral end surface) increases, and the outer peripheral portion of the wafer W There has been a problem that surface sagging occurs and the flatness of the portion, particularly SFQR, is deteriorated.

Here, the flatness includes the entire flatness of the wafer W and the flatness of each chip. The flatness calculated based on the flatness based on the back surface when it is assumed that the back surface of the wafer W is completely adsorbed and the reference plane defined by the least square method based on the surface of the wafer W is calculated. Degree (hereinafter simply referred to as “surface reference”).
In general, the flatness of the entire wafer based on the back surface is referred to as GBIR (Global Back Ideal Range), and the flatness of the back surface on a chip basis is referred to as SBIR (Site Back Ideal Range) or SBID (Site Back Ideal Deviation).
Further, the flatness of the entire wafer based on the surface is referred to as GFLR (Global Front Least Squares Range) or GFLD (Global Front Least Squares Deviation), and the flatness of the surface reference on a chip basis is SFQR (Site Front Least Squares Range) or SFQD. (Site Front Least Squares Deviation).

  In recent years, the flatness of the entire surface-based wafer (hereinafter referred to as “GFLR”) is 1.0 μm to 0.65 μm, and the flatness of the surface-based chip (hereinafter referred to as “SFQR”) is 0. Ultra-flattened wafers of 1 μm or less are required.

  However, conventionally, since the relative relationship between the compression rate of the resin layer 6b of the back pad 6 and the compression rate of the polishing pad 2 has not been studied, the two relationships shown in FIGS. Could not be resolved properly.

  Further, in the above-described polishing apparatus 10, by enclosing the outside of the wafer W with the retainer ring 7, local load concentration from the polishing pad 2 on the outer peripheral portion of the wafer W is eliminated, and SFQR is ensured to be ultra flat. Like to do. At this time, if the thickness of the retainer ring 7 is not thinner than the thickness of the wafer W, the point of polishing moves to the retainer ring 7, the progress of polishing of the outer peripheral portion of the wafer is hindered and the polishing allowance is reduced, and polishing is uniform. The nature will deteriorate.

  However, the thickness of the retainer ring 7 is made extremely smaller than the thickness of the wafer W, for example, as shown in FIG. 6A, and the difference (the thickness of the wafer W−the thickness of the retainer ring 7) dt. Is increased, the polishing pressure acting on the outer peripheral edge of the wafer W acts at a point, so that the polishing allowance of the outer peripheral portion becomes larger than that of the inner peripheral portion, and as a result, as shown in FIG. The result will be the same as the state.

  On the other hand, if the difference dt can be reduced as shown in FIG. 6B, the wafer W and the retainer ring 7 are as if they were connected to each other, and the polishing pressure acting on the outer peripheral edge of the wafer W acts on the surface. Will do.

  The invention according to the present application focuses on the balance of the above-described actions, and provides a polishing method and a polishing apparatus that can provide an object to be polished with more stable flatness.

In order to achieve the above object, a first invention according to the present application includes a polishing surface plate provided with a polishing pad, a spindle head arranged to face the polishing pad, and the spindle head and a thin plate-shaped substrate. A back pad disposed between the polishing object and a retainer ring disposed so as to surround the outer periphery of the object to be polished;
In a polishing apparatus for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad ,
Polishing, wherein the compression rate of the resin layer of the back pad is 6 to 25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600 to 2100. Device.
According to said invention, the dispersion | variation in the grinding | polishing pressure of the outer peripheral part of a to-be-polished object and an inner peripheral part can be reduced, and the to-be-polished surface which has high flatness can be provided.

In addition, a second invention according to the present application is directed to a polishing surface plate provided with a polishing pad, a spindle head disposed opposite to the polishing pad, and the spindle head and a thin plate-like object to be polished. A back pad disposed, and a retainer ring disposed so as to surround the outer periphery of the object to be polished,
In a polishing apparatus for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad ,
The retainer ring was set so that a difference in position in the height direction between the contacted surface of the retainer ring with the polishing pad and the polished surface of the object to be polished was in the range of 15 to 45 μm. A polishing apparatus characterized by being a retainer ring.
According to said invention, the dispersion | variation in the grinding | polishing pressure of the outer peripheral part of a to-be-polished object and an inner peripheral part can be reduced, and the to-be-polished surface which has high flatness can be provided.

Furthermore, a third invention according to the present application is directed to a polishing platen provided with a polishing pad, a spindle head disposed to face the polishing pad, and the spindle head and a thin plate-like object to be polished. A back pad disposed, and a retainer ring disposed so as to surround the outer periphery of the object to be polished,
In a polishing apparatus for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad ,
The compression rate of the resin layer of the back pad is 6 to 25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600 to 2100,
A polishing apparatus characterized in that a difference in position in a height direction between a surface to be contacted with the polishing pad of the retainer ring and a surface to be polished of the object to be polished is in a range of 15 to 45 μm. It is.
According to said invention, the dispersion | variation in the grinding | polishing pressure of the outer peripheral part of a to-be-polished object and an inner peripheral part can be reduced further, and a higher flat surface to be polished can be provided.

Moreover, 4th invention which concerns on this application WHEREIN: The grinding | polishing load at the time of grind | polishing the said to-be-polished object was 100-300 gf / cm < ² >, Any one of the said 1st-3rd invention characterized by the above-mentioned. The polishing apparatus described in 1. above.
According to said invention, the dispersion | variation in the grinding | polishing pressure of the outer peripheral part of a to-be-polished object and an inner peripheral part can be reduced further, and a higher flat surface to be polished can be provided.

Furthermore, a fifth invention according to the present application is directed to a polishing surface plate provided with a polishing pad, a spindle head disposed to face the polishing pad, and the spindle head and a thin plate-like object to be polished. A method of polishing the object to be polished by a polishing apparatus comprising at least a back pad disposed and a retainer ring disposed so as to surround an outer periphery of the object to be polished,
The compression rate of the resin layer of the back pad is 6-25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600-2100,
The polishing pad is a method for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished is in contact with the polishing pad. .
According to said invention, the dispersion | variation in the grinding | polishing pressure of the outer peripheral part of a to-be-polished object and an inner peripheral part can be reduced, and the to-be-polished surface which has high flatness can be provided.

According to a sixth aspect of the present application, there is provided a polishing surface plate provided with a polishing pad, a spindle head disposed opposite to the polishing pad, and the spindle head and a thin semiconductor substrate. A method of manufacturing a semiconductor wafer comprising a step of polishing the semiconductor substrate by a polishing apparatus comprising at least a back pad and a retainer ring disposed so as to surround an outer periphery of the semiconductor substrate,
The compression rate of the resin layer of the back pad is 6-25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600-2100,
The polishing surface is polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the polishing surface of the semiconductor substrate is in contact with the polishing pad. A method for manufacturing a semiconductor wafer.
According to said invention, the dispersion | variation in the grinding | polishing pressure of the outer peripheral part of a semiconductor substrate which is a to-be-polished object and an inner peripheral part can be reduced, and the semiconductor wafer which has high flatness can be provided.

  According to the polishing method and the polishing apparatus of the present invention, it is possible to process an object to be polished to a higher level.

  The terms compression rate (%) and compression modulus (%) in the present application are as defined in JIS L1096. Specifically, first, a sample for obtaining the compression rate and the compression elastic modulus is collected. Samples and specimens are collected and prepared in accordance with 5.3 (cloth-like specimens and specimens) of JIS L0105 (1) (for fabrics).

The compression rate (%) and the compression elastic modulus (%) were obtained by collecting 15 test pieces of about 50 mm × about 50 mm from the above sample, and using a compression elasticity tester, stacking 3 test pieces and standard pressure 4. 9kPa (50gf / cm ²⁾ measure the thickness _T 0 (mm) under, then, 29.4kPa (300gf / cm ²⁾ is left to a thickness of 1 minute under a constant pressure T ₁ (mm) Measure. Next, after removing the applied pressure and allowing to stand for 1 minute, the thickness T ₂ (mm) is measured again under a standard pressure of 4.9 kPa (50 gf / cm ² ). The above measured thicknesses are applied to the following formulas 1 and 2 to obtain the compression rate (%) and the compression elastic modulus (%), and an average value of 5 times is calculated and rounded to an integer.

Next, the polishing method and polishing apparatus of the present invention will be described with reference to the drawings. Since the entire structure of the polishing apparatus of the present invention is substantially the same as the polishing apparatus 10 described with reference to FIGS. 6 to 8, detailed description of the structure of the entire apparatus is omitted, and the reference numerals in the drawings are used. explain. The present invention is applied under the condition that a wafer W having a thickness of 350 μm or more is polished. Each experimental result to be described later is a result derived under conditions in which the polishing load is 100 to 300 gf / cm ² and the rotation speeds of the polishing platen 1 and the spindle head 5 are fixed to a constant rotation speed of 60 rpm.

  As the polishing pad 2, when the diameter of the wafer W is 200 mm, for example, a polishing cloth having a diameter of 620 mm is used. As a material of the polishing pad 2, a nonwoven fabric made of polyester resin or urethane resin can be used. The polishing pad 2 having a thickness of 600 to 1600 μm is used regardless of the diameter of the wafer W.

  When the diameter of the wafer W is 200 mm, the back pad 6 having a diameter of 245 mm can be used, for example. The thickness of the resin layer 6b of the back pad 6 is 300 to 700 μm regardless of the diameter of the wafer W.

  Specifically, a pressure-sensitive adhesive layer is provided on the back surface of the base material 6a of the back pad 6, and the pressure-sensitive adhesive layer is further separated from the back surface until it is attached to the bottom surface of the spindle head 5. The mold sheet is detachably attached. As the base material 6a described above, a polyester sheet, a polyvinyl chloride sheet, a nylon sheet, a polyimide sheet or the like is mainly used.

  The resin layer 6b of the back pad 6 includes an elastic layer laminated on the base material 6a and an adhesive resin layer laminated on the elastic layer as necessary. The adhesive resin layer is not essential, and the resin layer 6b may be composed only of an elastic layer.

  As the elastic layer, a member that exhibits a predetermined elastic force when the wafer W is held on the back pad 6 is used. For example, it can be formed by coating the substrate 6a with a DMF solution of urethane resin, coagulating the layer wet, washing in warm water, and drying with hot air. The surface of this elastic layer is smoothed to form a number of independent foam holes. The compressibility of the elastic layer is preferably 5 to 60% (measured according to JIS L1096).

  The adhesive resin layer laminated on the surface of the elastic layer as necessary is made of thermosetting resin or thermoplastic resin such as acrylic resin, polyurethane, polyester, epoxy resin, nylon, polyvinyl chloride, etc., and the surface is smooth. And has adhesiveness to the object to be polished. The hardness of the adhesive resin layer is preferably 48 to 88 (measured according to JIS-A, JISK6301 5.2 (A type)), and the compression ratio is preferably 3 to 52%.

  On the surface of the adhesive resin layer, minute irregularities having a surface roughness Ra of about 0.4 to 4.0 μm are formed over the entire surface. The minute irregularities are randomly (no directionality) formed on the surface of the adhesive resin layer. A more preferable range of the surface roughness is 0.5 to 2.0 (U70 to DM42) μm. If the surface roughness is less than 0.4 μm, the air escape rate is slow and air may be caught between the object to be polished and the back pad 6, and the handling by the machine is poor, and exceeds 4.0 μm. And the surface accuracy decreases.

  Such an adhesive resin layer can be manufactured as follows, for example. The surface of the elastic layer laminated on the surface of the substrate 6a is buffed to have a certain surface roughness (unevenness), and then the surface of this elastic layer is coated with a thermosetting urethane resin. Next, before the urethane resin is completely cured, a film with minute irregularities is pressure-bonded and laminated with a hot roll, and after the surface of the elastic layer and the urethane resin are adhered, the film is peeled off from the surface of the urethane resin. Let Then, the surface roughness of the film itself is transferred to the surface of the urethane resin, and a self-adsorption type back pad 6 having a certain roughness can be obtained on the surface non-foamed urethane layer.

  By producing the back pad 6 by such a manufacturing method, the surface roughness of the resin layer 6b of the back pad 6 can be controlled by controlling the surface roughness of the film. In addition, the manufacturing method of the resin layer 6b mentioned above is an example, and it cannot be overemphasized that the resin layer 6b and the back pad 6 which have a characteristic as shown in this application can be manufactured also by another method.

  The retainer ring 7 is attached along the outer periphery of the back pad 6 so as to surround the wafer W that is an object to be polished. The retainer ring 7 is a ring-shaped member or a plurality of block-shaped members arranged in an annular shape so as to surround the wafer W. For example, a ceramic material or resin, and a metal material coated with a ceramic material / resin Etc. As a material of the retainer ring 7, it is preferable to use a material having a hardness equivalent to or higher than that of the wafer W that is the object to be polished.

If the gap between the outer peripheral surface of the wafer W and the inner peripheral surface of the retainer ring 7 is too large, the effect of preventing the surface sag by the retainer ring 7 is reduced. Therefore, the gap between the wafer W and the retainer ring 7 is preferably 2 mm or less. In particular, it is desirable to be in the range of 0.5 to 2.0 mm. In the present embodiment, a ring-shaped retainer ring 7 having an inner diameter of 201 mm and an outer diameter of 241 mm is used.
The retainer ring 7 includes a type that is fixed to the back pad 6 and a type that is fixed to the spindle head 5. In this embodiment, a type that is fixed to the back pad 6 is adopted.

  In the present embodiment, the experiment was performed using the back pad 6 provided with the resin layer 6b having various characteristics as shown in Table 1 below.

  1 to 3 are comparative graphs showing the relationship between the compression ratio Cr (%) of the resin layer 6b and the compression elastic modulus Ce (%) of the polishing pad 2 and the polishing allowance. Here, a graph in the case of using a general wafer W having a diameter of 200 mm and a thickness of 725 μm is shown.

  FIG. 1 is a graph showing how the polishing allowance of the wafer W changes from the center of the wafer when the compression ratio Cr (%) of the resin layer 6b of the back pad 6 is 12%, 20%, and 30%. It is. The horizontal axis represents the distance (mm) from the center position of the wafer, and the vertical axis represents the polishing allowance (μm) of the wafer W.

  The compression rate Cr (%) of the resin layer 6b is 30%, which corresponds to a state where the wafer W is considerably submerged in the resin layer 6b as shown in FIG. %) Corresponds to a state in which the wafer W is hardly submerged in the resin layer 6b as shown in FIG. 7B.

  As shown by the broken line in FIG. 1, when the compressibility of the resin layer 6b is relatively large (soft) 30%, the polishing allowance at the center of the wafer W increases and the polishing allowance decreases toward the outer periphery. was gotten.

  On the other hand, as indicated by a two-dot chain line in FIG. 1, when the compressibility of the resin layer 6b is relatively small (hard) 12%, the result is that the change in the polishing allowance is generally small from the center of the wafer W to the vicinity of the outer periphery. was gotten. Further, as shown by the solid line in FIG. 1, when the compressibility of the resin layer 6b is 20%, the change in the polishing allowance is generally small from the center of the wafer W to the vicinity of the outer periphery as in the case of 12%. was gotten. In addition, the range shown by the arrow is a range actually used as a product.

  FIG. 2 is a graph showing the relationship between the compressibility Cr (%) of the resin layer 6 b and the flatness of the wafer W. The horizontal axis represents the compression ratio Cr (%) of the resin layer 6b, and the vertical axis represents the inclination a (in) of the inner peripheral portion machining allowance of the wafer W. Since the compression ratio Cr mainly contributes to the uniformity of the machining allowance of the inner peripheral portion of the wafer, it was calculated on the basis of the correlation with the inclination of the machining allowance of the inner peripheral portion.

  The inclination a (in) of the inner peripheral machining allowance is a linear regression equation of the machining allowance from the center of the wafer W to 90% in the radial direction from FIG. 1, and the inclination of the function is a measure of the uniformity of the wafer flatness. It is what. Accordingly, the closer to 0 means higher uniformity.

が得られる。 From the slope a (in) of the inner peripheral portion machining allowance plotted in FIG.

Is obtained.

  If the inclination a (in) of the inner peripheral portion machining allowance is in the range of ± 1, it can be shipped as a non-defective wafer having no problem in wafer flatness. When the value in which the slope a (in) of the inner peripheral portion machining allowance falls within the range of ± 1 is obtained from the regression equation y of Equation 3, the compression ratio Cr (%) of the resin layer 6b is 6 to 25%.

Therefore, it can be seen that if a back pad having a resin layer compressibility Cr (%) of 6 to 25% is used, a non-defective wafer having no problem in wafer flatness can be obtained.
Therefore, when a test was actually conducted using a back pad having a resin layer compression ratio Cr (%) of 6 to 25%, a wafer having particularly good GFLR (Global Front Least Squares Range) could be obtained. .

FIG. 3 is a graph showing the relationship between [compression rate Cr (%) of the resin layer 6b] × [compression elastic modulus Ce (%) of the polishing pad 2] and the flatness of the wafer W. Since each of [compressibility Cr (%) of the resin layer 6b] and [compression modulus Ce (%) of the polishing pad 2] has a correlation with the flatness of the wafer W, in this embodiment, The product of [compression rate Cr (%) of the resin layer 6b] and [compression modulus Ce (%) of the polishing pad 2] was adopted as a parameter.
The horizontal axis represents [compression rate Cr (%) of the resin layer 6b] × [compression elastic modulus Ce (%) of the polishing pad 2], and the vertical axis represents the inclination a (in) of the inner peripheral portion machining allowance of the wafer W. ing. [Compression rate Cr (%) of the resin layer 6b] × [Compression elastic modulus Ce (%) of the polishing pad 2] mainly contributes to the uniformity of the machining allowance of the inner peripheral portion of the wafer. It was calculated based on the correlation with the slope of.

  The inclination a (in) of the inner peripheral machining allowance is a linear regression equation of the machining allowance from the center of the wafer W to 90% in the radial direction from FIG. 1, and the inclination of the function is a measure of the uniformity of the wafer flatness. It is what. Accordingly, the closer to 0 means higher uniformity.

が得られる。 From the slope a (in) of the inner peripheral portion allowance plotted in FIG.

Is obtained.

  If the inclination a (in) of the inner peripheral portion machining allowance is in the range of ± 1, it can be shipped as a non-defective wafer having no problem in wafer flatness. When a value in which the slope a (in) of the inner peripheral portion machining allowance falls within a range of ± 1 is obtained from the regression equation y of Equation 4, [compression rate Cr (%) of the resin layer 6b] × [compression elastic modulus of the polishing pad 2] Ce (%)] was 600-2100.

Therefore, if the resin layer 6b and the polishing pad 2 are used such that the [compression rate Cr (%) of the resin layer 6b] × [compression elastic modulus Ce (%) of the polishing pad 2] is 600 to 2100, the wafer flatness is obtained. It can be seen that a non-defective wafer can be obtained.
Therefore, the test is actually performed using the resin layer 6b and the polishing pad 2 such that the [compression rate Cr (%) of the resin layer 6b] × [compression elastic modulus Ce (%) of the polishing pad 2] is 600 to 2100. As a result, a wafer having particularly good GFLR (Global Front Least Squares Range) could be obtained.

  Of course, it goes without saying that it is better to employ a resin layer and a polishing pad that satisfy the requirements of Example 1 and Example 2. That is, the compression rate Cr (%) of the resin layer is 6 to 25%, and the [compression rate Cr (%) of the resin layer] × [compression elastic modulus Ce (%) of the polishing pad] is 600 to 2100. It is desirable to employ such a resin layer and polishing pad.

  FIG. 4 is a graph showing how the polishing allowance of the wafer W changes from the center of the wafer when the thickness of the retainer ring 7 is 620 μm, 670 μm, and 720 μm. The horizontal axis represents the distance (mm) from the center position of the wafer, and the vertical axis represents the polishing allowance (μm) of the wafer W. Here, a graph in the case of using a general wafer W having a diameter of 200 mm and a thickness of 725 μm is shown.

  The thickness of the retainer ring 7 is 620 μm, which corresponds to a state where dt = (thickness of the wafer W−thickness of the retainer ring 7) is large as shown in FIG. 720 μm corresponds to a state where dt = (thickness of wafer W−thickness of retainer ring 7) is small as shown in FIG.

  As indicated by the broken line in FIG. 4, when the thickness of the retainer ring 7 is 720 μm, the polishing allowance on the outer peripheral portion of the wafer W does not increase. That is, the result that the surface sagging hardly occurs was obtained. Further, as shown by the solid line in FIG. 4, even when the thickness of the retainer ring 7 is 670 μm, the polishing allowance of the outer peripheral portion of the wafer W does not increase excessively, and the entire area from the center of the wafer W to the vicinity of the outer periphery. As a result, the change in polishing allowance was small.

  On the other hand, as shown by a two-dot chain line in FIG. 4, when the thickness of the retainer ring 7 is 620 μm, the result that the polishing allowance of the outer peripheral portion of the wafer W increases abruptly is obtained.

  FIG. 5 is a graph showing the relationship between dt = (thickness of wafer W−thickness of retainer ring 7) and flatness of wafer W. The horizontal axis represents dt = (thickness of wafer W−retainer ring 7) (μm), and the vertical axis represents the inclination a (out) of the outer peripheral portion of the wafer W. Since the difference dt between the thickness of the wafer W and the thickness of the retainer ring 7 mainly contributes to the uniformity of the machining allowance at the outer peripheral portion of the wafer, the difference dt was calculated based on the correlation with the inclination of the machining allowance at the outer peripheral portion.

  The inclination a (out) of the peripheral part machining allowance is a linear regression formula obtained from the peripheral part machining allowance from 90% of the radius of the wafer W from FIG. 1, and the slope of the function is used as a measure of the uniformity of the wafer flatness. Is.

が得られる。 From the slope a (out) of the outer peripheral portion machining allowance plotted in FIG.

Is obtained.

  If the inclination a (out) of the outer peripheral portion machining allowance is in the range of ± 5, the outer peripheral portion can also be shipped as a good wafer. From the regression equation y in Equation 5, when a value in which the inclination a (out) of the outer peripheral portion machining allowance falls within the range of ± 5 is obtained, dt = (thickness of the wafer W−thickness of the retainer ring 7) is 15 to 45 μm. It was.

Therefore, it can be seen that if the thickness of the retainer ring is set so that dt = (wafer thickness−retainer ring thickness) is 15 to 45 μm, a non-defective wafer having no problem in wafer flatness can be obtained. .
Therefore, when the test was actually performed by setting the thickness of the retainer ring so that dt = (wafer thickness−retainer ring thickness) was 15 to 45 μm, the SFQR (Site Front in the outer peripheral portion in particular was tested. A wafer with a good Least Squares Range was obtained.

  Of course, it goes without saying that it is better to employ a resin layer, a polishing pad and a retainer ring that satisfy the requirements of Example 3 and the requirements of Example 1 or Example 2.

  In the above embodiment, dt = (thickness of the wafer W−thickness of the retainer ring 7) is assumed to be a difference between the height direction position of the polished surface of the wafer W and the bottom surface of the retainer ring 7. Yes. However, in actual polishing, since the wafer W and the retainer ring 7 are sunk into the back pad 6 in accordance with the compression rate of the resin layer 6b, the thickness design of the retainer ring 7 in consideration of the sink into the resin layer 6b. Is required.

  Further, in the case of the type in which the retainer ring 7 is directly attached to the spindle head 5 instead of the back pad 6, dt '= [wafer thickness- (retainer ring thickness-back pad thickness)] is 15 to. If the thickness of the retainer ring is set so as to be 45 μm, a non-defective wafer having no problem in wafer flatness can be obtained.

  The present invention is novel in that it focuses on the compression rate Cr of the resin layer of the back pad, and in particular, [compression rate Cr (%) of the resin layer] × [compression elastic modulus Ce (%) of the polishing pad]. It has novelty in the point which paid attention to.

  Further, in combination with the above requirements, by paying attention to the difference in the height direction position between the surface to be polished of the wafer W and the bottom surface of the retainer ring, a higher flat wafer can be provided.

  In the above embodiment, a semiconductor wafer is described as an example. However, the present invention is not limited to a semiconductor wafer. For example, a thin plate-shaped precision glass substrate or a thin plate-shaped workpiece made of other materials is used. It can also be applied to.

  Further, the thin plate-like object to be polished is not limited to the disk-shaped wafer, but can be applied to a rectangular or polygonal object to be polished.

It is the graph which examined how the polishing allowance of a wafer changed from the center of a wafer when compression rate Cr of a resin layer of a back pad is 12%, 20%, and 30%. It is a graph which shows the relationship between the compressibility Cr of a resin layer, and the flatness of a wafer. 6 is a graph showing a relationship between [resin layer compressibility Cr] × [polishing pad compressive modulus Ce] and wafer flatness. It is the graph which investigated how the polishing allowance of a wafer changed from the center of a wafer when the thickness of a retainer ring is 620 micrometers, 670 micrometers, and 720 micrometers. It is a graph which shows the relationship between dt = (thickness of a wafer-thickness of a retainer ring) and the flatness of a wafer. FIG. 6A shows the relative relationship between the object to be polished and the retainer ring. FIG. 6A is an explanatory diagram of the main part when the thickness of the retainer ring is extremely thin with respect to the thickness of the object to be polished, FIG. FIG. 5 is an explanatory diagram of a main part when the thickness of the retainer ring is slightly smaller than the thickness of the object to be polished. FIG. 7A shows the relative relationship between the compression rate of the back pad and the compression elastic modulus of the polishing pad, FIG. 7A is an explanatory diagram of the main part when the compression rate of the back pad is high, and FIG. 7B is the compression of the back pad. It is explanatory drawing of the principal part in case a rate is low. FIG. 8A is a perspective view of the polishing apparatus, and FIG. 8B is a side view of the polishing apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Polishing surface plate 2 ... Polishing pad 3 ... Rotating shaft 4 ... Weight support shaft 5 ... Spindle head 6 ... Back pad 6a ... Base material 6b ... Resin layer 7 ... Retainer ring 10 ... Polishing apparatus W ... Wafer (to-be-polished object) .

Claims (6)

A polishing surface plate provided with a polishing pad; a spindle head arranged opposite to the polishing pad; a back pad arranged between the spindle head and a thin plate-like object; A retainer ring disposed so as to surround the outer periphery,
In a polishing apparatus for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad ,
Polishing, wherein the compression rate of the resin layer of the back pad is 6 to 25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600 to 2100. apparatus. A polishing surface plate provided with a polishing pad; a spindle head arranged opposite to the polishing pad; a back pad arranged between the spindle head and a thin plate-like object; A retainer ring disposed so as to surround the outer periphery,
In a polishing apparatus for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad ,
The retainer ring was set so that a difference in position in the height direction between the contacted surface of the retainer ring with the polishing pad and the polished surface of the object to be polished was in the range of 15 to 45 μm. A polishing apparatus that is a retainer ring. A polishing surface plate provided with a polishing pad; a spindle head arranged opposite to the polishing pad; a back pad arranged between the spindle head and a thin plate-like object; A retainer ring disposed so as to surround the outer periphery,
In a polishing apparatus for polishing the surface to be polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad ,
The compression rate of the resin layer of the back pad is 6 to 25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600 to 2100,
A polishing apparatus characterized in that a difference in position in a height direction between a surface to be contacted with the polishing pad of the retainer ring and a surface to be polished of the object to be polished is in a range of 15 to 45 μm. . The polishing apparatus according to any one of claims 1 to 3, characterized in that the grinding load at the time of polishing the object to be polished was 100~300gf / cm ^2. A polishing surface plate provided with a polishing pad; a spindle head arranged opposite to the polishing pad; a back pad arranged between the spindle head and a thin plate-like object; A retainer ring disposed so as to surround an outer periphery, and a method of polishing the object to be polished by a polishing apparatus comprising at least
The compression rate of the resin layer of the back pad is 6-25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600-2100,
A method of polishing the surface to be polished with the polishing pad by rotating at least one of the polishing surface plate or the spindle head in a state where the surface to be polished of the object to be in contact with the polishing pad. A polishing surface plate provided with a polishing pad, a spindle head arranged opposite to the polishing pad, a back pad arranged between the spindle head and a thin plate-like semiconductor substrate, and an outer periphery of the semiconductor substrate A method of manufacturing a semiconductor wafer comprising a step of polishing the semiconductor substrate by a polishing apparatus comprising at least a retainer ring disposed to surround the retainer ring,
The compression rate of the resin layer of the back pad is 6-25%, and the product of the compression rate of the resin layer and the compression elastic modulus of the polishing pad is in the range of 600-2100,
The polishing surface is polished with the polishing pad by rotating at least one of the polishing platen or the spindle head in a state where the polishing surface of the semiconductor substrate is in contact with the polishing pad. A semiconductor wafer manufacturing method.

Images