JPH0696225B2 - Polishing method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a polishing method particularly suitable for polishing a semiconductor wafer.

(Prior Art) FIG. 9 shows a principle diagram of a conventional polishing method. For example, a semiconductor wafer W is adhered to a lower surface of a plate 106 with a mounting material, wax or the like, and the lower surface is horizontally swung at a predetermined speed. The buff (polishing cloth) 102 attached on the rotary surface plate 101 is pressed by a weight 113 with a predetermined force, and the buff (abrasive cloth) is supplied from the nozzle 108 to receive the abrasive 109.
The lower surface of the semiconductor wafer W is mirror-polished due to relative slippage with the semiconductor wafer W.

By the way, with the recent high integration of semiconductor devices,
The semiconductor wafer is required to have high parallelism and high flatness, and in order to satisfy these requirements, the plate 106 must be kept in a high flatness without being flexibly deformed in the polishing state. .

(Problems to be Solved by the Invention) However, in the above-described conventional polishing method, the top ring 1 in which the plate 106 has an inverted dish shape as shown in FIG.
The upper edge is held by 03, and the weight 113
The load of ... locally acts on the peripheral edge of the upper surface of the plate 106, and upward polishing pressure acts on the lower surface of the plate 106, so that the plate 106 is projected upward as shown by the solid line in FIG. The plate is bent and deformed into a spherical shape
There is a problem in that the flatness of 106 cannot be kept high and the semiconductor wafer W cannot be polished with high precision.

The present invention has been made in view of the above problems, and an object of the present invention is to provide a polishing method capable of polishing an object to be polished with high accuracy while maintaining high flatness of the plate in the polishing state. .

(Means for Solving the Problems) In order to achieve the above object, the present invention holds an object to be polished on the lower surface of a plate rotatable about a central axis and presses the upper peripheral edge of the plate with a top ring. The lower surface of the object to be polished is pressed by a predetermined force on the buff attached to the rotating surface plate by a predetermined force to cause relative sliding between the object to be polished and the buff, and the lower surface of the object to be polished is a mirror surface. In the polishing method for polishing, a cooling medium is passed through the top ring,
The difference Δt between the upper surface temperature tp and the lower surface temperature tc of the plate is expressed by the following equation: Where a: Plate radius E: Young's modulus of pate h: Plate thickness P: Net load acting on the plate α: Linear expansion coefficient of the plate ν: Temperature control to satisfy the Poisson's ratio of the plate And

(Function) If no temperature control is performed, the plate is subjected to polishing pressure on the lower surface thereof and is flexibly deformed into a spherical surface having a convex shape as described above (see FIG. 10). If the upper surface of the plate is cooled by the cooling medium flowing in the ring to cause a temperature difference between the upper and lower surfaces of the plate, the plate will be deformed flexibly in the shape reverse to the above, and this thermal deformation and the mechanical The deformations cancel each other out, and the actual amount of bending deformation of the plate is suppressed to a small value.

However, the above equation is derived from the condition that the amount of flexural deformation at the center of the plate is zero, that is, the center and the peripheral edge of the plate are on the same plane, and the temperature difference between the upper and lower surfaces of the plate is If the temperature is controlled so that Δt becomes equal to the value calculated by the same equation, the amount of flexural deformation of the plate is suppressed to a minimum, its flatness is kept high, and the object to be polished held on the lower surface of the plate is It will be highly accurately polished.

(Example) Below, the Example of this invention is described based on an accompanying drawing.

First, a polishing apparatus for carrying out the method of the present invention will be described with reference to FIG. 1. In FIG. 1, reference numeral 1 is a rotary platen that is horizontally driven to rotate at a constant speed in a direction indicated by an arrow. A buff (polishing cloth) 2 having appropriate elasticity is attached on the surface plate 1. A top ring 3 is rotatably faced above the rotary platen 1, and a concave portion 3a having a downward opening is formed at a lower portion of the top ring 3 to form the concave portion 3a.
A water injection pipe 4 and a drainage pipe 5 are open in the. The water injection pipe 4 and the drain pipe 5 communicate with a cooling water supply device (not shown).

On the other hand, a plate 6 to which a plurality of semiconductor wafers W are adhered is placed under the buff 2 of the rotary platen 1 in advance, and the peripheral edge of the upper surface of the plate 6 has an elastic seal ring 7 interposed therebetween. And is pressed downward by the top ring 3 with a predetermined force. In FIG. 1, 8 is a nozzle that opens on the central axis of the rotary platen 1, and 9 is a polishing liquid ejected from the nozzle 8.

Thus, the rotary platen 1 is driven to rotate horizontally around its central axis at a constant speed by a drive device (not shown), and at this time, the cooling water is supplied from the water injection pipe 4 into the recess 3a of the top ring 3. The cooling water is discharged from the discharge pipe 5 after cooling the upper surface of the plate 6, and the cooling water constantly flows in the recess 3a. The leakage of cooling water from the recess 3a of the top ring 3 is prevented by the sealing effect of the elastic seal ring 7.

By the way, the peripheral speed difference is generated in the buff 2 on the rotary platen 1 in the radial direction, and the semiconductor wafer W, the plate 6, the top ring 3 and the like circulate due to the peripheral speed difference.
Due to this movement, relative slippage occurs between the semiconductor wafer W ... And the buff 2. Due to this relative slip, each semiconductor wafer W receives the supply of the abrasive liquid 9 from the nozzle 8 and the buff 2 gives a mirror surface. To be polished.

In the above, when the plate 6 is not cooled at all, that is, when the cooling water does not flow into the recess 3a of the top ring 3, the plate 6 is subjected to polishing pressure on its lower surface to form a convex shape upward. It bends and deforms into a spherical shape (see Fig. 10).

Therefore, in this embodiment, the upper surface of the plate 6 is cooled by the cooling water flowing in the recess 3a of the top ring 3,
A temperature difference is generated between the upper surface and the lower surface, and the temperature difference causes the plate 6 to be bent and deformed to form a convex spherical surface below the above-described shape. The deformations cancel each other out, and the actual amount of flexural deformation of the plate 6 is suppressed to a small level. As a result, the flatness of the plate 6 is kept high and the semiconductor wafer W held on the lower surface of the plate 6 is highly accurate. To be polished.

As described above, in this embodiment, the mechanical deformation due to the polishing pressure applied to the lower surface of the plate 6 is offset by the thermal deformation due to cooling, but the actual bending deformation amount of the plate 6 is minimized. The cooling condition (temperature control condition) for maintaining the temperature is analytically obtained below.

(1) Mechanical Deformation Amount of Plate First, the mechanical deformation amount due to the polishing pressure of the plate 6 is
We ask for a simplified model in the figure. That is, it is considered that the plate 6 is simply supported at its peripheral edge, and the uniformly distributed loads Pp and Pc in the directions of the arrows in the drawing act on the upper and lower surfaces of the plate 6.

Therefore, the mechanical deformation amount Wp due to the loads Pp and Pc of the plate 6
Is calculated by the following formula (Mechanical Engineering Handbook, revised 5th edition 4th
Hen: See Material Mechanics).

Where a: radius of plate r: radius of arbitrary point of plate h: plate thickness E: Young's modulus of plate P: net load (= Pp-Pc) ν: Poisson's ratio of plate (2) thermal of plate Deformation amount On the other hand, the difference Δt between the upper surface temperature tp and the lower surface temperature tc of the plate 5
= The thermal deformation amount W _H based on tp-tc required for approximately the following equation.

Here, α: Plate linear expansion coefficient (3) Actual deformation amount of plate Therefore, the actual deformation amount W of the plate 6 is the mechanical deformation amount Wp expressed by the equation (1) and the equation (2). It is calculated by the following formula as the sum of the thermal deformation amount W _H shown in.

(4) Temperature condition for keeping the actual amount W of deformation of the plate to a minimum As a condition for keeping the actual amount W of deformation of the plate 6 at a minimum,
The flexural deformation amount W _{r = 0} in the plate 6 is zero (W _{r = 0} =
0) is introduced and the temperature difference Δt _N that satisfies this condition is determined.

That is, When Δt _N is calculated from the above equation, Thus, when the temperature difference Δt between the upper and lower surfaces of the plate 6 satisfies the above equation (that is, the deformation amount W at the center point of the plate 6).
Actual flexural deformation amount W of the plate 6 when _{r = 0} = 0)
Is calculated by the following equation by substituting the equation (5) into the equation (3).

The deflection curve of the plate 6 expressed by the above equation is shown in FIG. 3, and the position where the maximum deflection W _MAX occurs should satisfy the following equation r
Is required as.

Since r> 0, the maximum deflection W _MAX is The maximum deflection W _MAX which occurs at the position of is calculated by the following equation by substituting the above equation into the equation (7).

(5) Usable Temperature Difference Next, the calculation formula of the usable temperature difference Δv of the upper and lower surfaces of the plate 6 (achievable temperature difference) is calculated.

First, the heat transfer amount Q of the semiconductor wafer W, the adhesive layer (wax layer) of the semiconductor wafer W, the plate 6, the cooling water boundary film (boundary layer), and the layers other than the cooling water boundary film is calculated by the following equation. To be

_{Q = AU (t 1 -t 3} ) ... (10) Where A: heat transfer area h ₁ : thickness of wafer h ₂ : thickness of adhesive layer k ₁ : thermal conductivity of wafer k ₂ : thermal conductivity of adhesive layer k: thermal conductivity of plate H: boundary Membrane heat transfer coefficient U: Overall heat transfer coefficient t ₁ : Wafer temperature t ₃ : Cooling water temperature On the other hand, the heat transfer amount Q of the plate 6 is obtained by the following equation.

Therefore, the temperature difference Δ that can be used from the above equations (10) and (11)
tv is calculated by the following equation.

After all, in order to adjust the upper and lower temperature differences Δt to the temperature difference Δt _N (see the formula (5)) at which the flexural deformation amount should be kept to a minimum, the following conditions are satisfied. There must be.

Δt _N ≦ tv (13) (6) Concrete example Next, the calculation results using concrete numerical values are shown. The calculation was performed based on the numerical values shown in the following table.

Then, the deformation amount W of the plate 6 and its maximum value W _MAX are respectively obtained by the above equations (6) and (9) as follows, and the results are graphed in FIGS. 4 and 5. Is displayed.

Further, the temperature difference Δt _N between the upper and lower surfaces of the plate 6 and the usable temperature difference Δtv, which are necessary to keep the deformation amount W of the plate 6 at the above result, are calculated by the following equations (5) and (12). And the result is displayed in a graph in FIG.

By the way, in order to improve the flatness of the plate 6 by keeping the deformation amount W of the plate 6 small according to the formula (6),
The following measures can be considered on the assumption that the temperature condition expressed by the formula is satisfied.

(A) To reduce the net load P as much as possible.

(B) Select a material with a small value of (1-ν ² ) / E.

(C) Use a plate with a large thickness h.

Here, when the influence of the thickness h on the maximum deflection W _MAX of the plate and the temperature difference Δt _N is calculated for a plate having a diameter of 2a = 52.5 cm, the following formula is obtained.

The above results are shown graphically in FIG. 7 and FIG. 8 with the net load P = 0.05, 0.1, 0.2 kg / cm ² as a parameter, respectively. The increase of the plate thickness h means the amount of bending deformation of the plate. It is extremely effective in reducing the required temperature difference Δ
It can be seen that there is a _trend to reduce the cooling problem by reducing t _N.

As described above, in this embodiment, the mechanical deformation due to the polishing pressure applied to the lower surface of the plate 6 is offset by the thermal deformation due to the temperature difference between the upper and lower surfaces of the plate 6, and the bending of the plate 6 is canceled. The amount of deformation is kept small, and in particular, the temperature difference Δt _N required to keep the amount of bending deformation to a minimum is analytically obtained. Therefore, if temperature control is performed to maintain this temperature difference Δt _{N, the} plate 6 The flatness is kept high,
The semiconductor wafer W held on the lower surface of the plate 6 is highly accurately polished so that its parallelism and flatness are kept high, and the accuracy required for the semiconductor wafer in accordance with the recent high integration of semiconductor devices and the like. Will be fully satisfied.

Although water is used as the cooling medium for cooling the upper surface of the plate 6 in the above embodiments, any cooling medium can be used. The temperature control is performed by adjusting the temperature, flow rate, etc. of the cooling medium.

(Effect of the invention) As is clear from the above description, according to the present invention, the object to be polished is held on the lower surface of the plate rotatable about the central axis,
By pressing the peripheral edge of the upper surface of the plate with a top ring, the lower surface of the object to be polished is pressed with a predetermined force onto the buff attached to the rotating surface plate so that the object to be polished and the buff are relatively opposed to each other. In a polishing method in which a lower surface of an object to be polished is mirror-polished by causing slippage, a cooling medium is passed through the top ring, and a difference Δt between an upper surface temperature tp and a lower surface temperature tc of the plate is expressed by the following formula: Where: a: radius of the plate E: Young's modulus of the plate h: thickness of the plate P: net load acting on the plate α: coefficient of linear expansion of the plate ν: temperature control to satisfy the Poisson's ratio of the plate As a result, the mechanical deformation of the plate is offset by the thermal deformation, the amount of flexural deformation of the plate is suppressed to a minimum, the flatness of the plate is kept high, and the workpiece held on the lower surface of the plate is highly accurate. It has the effect of being polished.

[Brief description of drawings]

1 is a configuration diagram of a polishing apparatus for carrying out the method of the present invention, FIG. 2 is a modeled diagram for calculating a mechanical deformation amount of a plate, FIG. 3 is a diagram showing a deflection curve of the plate, FIG. Fig. 4 is a graph showing the amount of deflection of the plate with respect to its radius, Fig. 5 is a graph showing the maximum amount of deflection of the plate with respect to its radius, and Fig. 6 is the temperature difference of the plate and the available temperature difference. The graph showing the radius of the plate,
FIG. 8 is a graph showing the maximum deflection of the plate with respect to the plate thickness, FIG. 8 is a graph showing the required temperature difference of the plate with respect to the plate thickness, and FIG. 9 is for explaining a conventional polishing method. FIG. 10 is a configuration diagram of the polishing apparatus of FIG. 10, and FIG. 10 is an explanatory view showing a state of flexural deformation of the plate. 1 ... Rotating surface plate, 2 ... Buff, 3 ... Top ring, 4
...... Concave part, 4 …… Water injection pipe, 5 …… Drainage pipe, 6 …… Plate, W …… Semiconductor wafer (abrasive).

Claims (1)

[Claims] 1. An object to be polished is held on a lower surface of a plate rotatable about a central axis, and a lower surface of the object is affixed to a rotary platen by pressing a peripheral edge of an upper surface of the plate with a top ring. In the polishing method, in which the lower surface of the object to be polished is mirror-polished by pressing the surface of the buff with a predetermined force to cause relative sliding between the object to be polished and the buff, a cooling medium is passed through the top ring. , The difference Δt between the upper surface temperature t _P and the lower surface temperature t _C of the plate is given by the following equation: Where a: Plate radius E: Young's modulus of the plate h: Plate thickness P: Net load acting on the plate α: Linear expansion coefficient of the plate ν: Temperature control to satisfy the Poisson's ratio of the plate And polishing method.