Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD FOR CUTTING A BLOCK OF SILICON INTO WAFERS
Document Type and Number:
WIPO Patent Application WO/2011/081533
Kind Code:
A1
Abstract:
There is provided a method for cutting a block (13) of silicon into a multiplicity of wafers, by movement of a planar array (11) of parallel fast moving wires relative to the block (13) (or vice versa) in a direction perpendicular to the plane of the wires and applying a slurry to the wires before the wires (11) pass through the block (13). The block (13) has a surface (15) into which the wires enter the block perpendicularly, the method further comprises applying a coating on said surface (15) prior to start of the cutting.

Inventors:
AAS, Kristian, Leonard (Ballestadvegen 25, Skien, N-3713, NO)
Application Number:
NO2010/000491
Publication Date:
July 07, 2011
Filing Date:
December 30, 2010
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
REC WAFER NORWAY AS (P.O.Box 594, Sandvika, N-1302, NO)
AAS, Kristian, Leonard (Ballestadvegen 25, Skien, N-3713, NO)
International Classes:
B28D5/00; B28D5/04
Attorney, Agent or Firm:
ONSAGERS AS et al. (P.O.Box 6963 St. Olavs plass, Oslo, N-0130, NO)
Download PDF:
Claims:
CLAIMS

1. Method for cutting a block of silicon into a multiplicity of wafers of thickness less than 200 μπι, where

- a planar array of parallel running wires is moved relative to the block in a direction perpendicular to the plane of the running wires,

- an abrasive slurry is applied to the wires before the wires pass through the block, characterised in that

- the surface into which the wires enters the block perpendicularly is protected by a polymer coating with thickness in the range from 20 to 1000 μπι.

2. Method according to claim 1 , wherein the thickness of the coating is 20 to 500 μιη, preferably 30 to 300 μπι and more preferably 40 to 250 μηι.

3. Method according to claim 1 or 2, wherein the polymer coating is made of the same coating as the glue employed to fasten the silicon block to the holding substrate for the cutting.

4. Method according to claim 3, wherein the polymer coating is made is an epoxy resin, acrylic resin, or made of plastic foils.

5. Method according to any one of the preceding claims, wherein the abrasive slurry comprises abrasive particles of silicon carbide or diamond.

Description:
Method for cutting a block of silicon into wafers

Field of the Invention

The present invention relates to the production of wafers cut out from a silicon ingot. More specifically it relates to wafers for use in the photovoltaic cell industries.

Background to the invention

Wafers are normally manufactured by cutting a silicon ingot first into blocks, and then by further cutting the blocks into a multiplicity of thin wafers. An example of an apparatus for producing a multiplicity of thin wafers from a block of solid material was described in UK Patent Specification 2,414,204.

In such machinery, a high tensile steel wire is passed under tension from a supply reel, around wire guide rollers, and is taken up on a collection reel. The surface of each of the wire guide rollers has a series of closely spaced 'V grooves at a pitch separation equal to the thickness of the wire plus the thickness of the required wafer. The configuration of the rollers is such that the wire can pass around them up to 3000 times producing a flat loom or web of wires, with the wire passing around the rollers at a speed of up to 20 meters per second.

The web is flooded with a suspension or slurry of finely divided abrasive powder such as e.g. 10 micron silicon carbide powder in a lubricating or cooling medium such as paraffin oil or polyethylene glycol. The material to be sliced is pressed against and slowly moved through the web of wires where the fast moving wires (together with the abrasive slurry) cut a regular series of thin slots through the material. This produces a multiplicity of wafers or plates of accurate thickness with a fine surface finish.

After the cutting, the block has been transformed into a wafer stack. After the cutting process, three of the block surfaces correspond to the edges of the newly created wafers in the wafer stack and the gaps between the wafers. The appearance of the wafer stack differ on the three newly cut surfaces. It has been observed that as the photovoltaic industry has employed thinner wafers, one of the surfaces has been showing a dull appearance with bright speckles. The other two surfaces are more uniform, and the edges constituting these surfaces have the same or almost the same shiny appearance as the original block surfaces before slicing of the block into wafers.

The shiny surface is made by block grinding and this operation is intended to control the level of subsurface damage and keep it on a low level. The abnormal dull surface with speckles is therefore a concern with respect to quality of the wafers.

By close inspection the surface with speckles is identified as the surface where the wires enter into the block perpendicularly, hereafter called the "wire in" surface. The edges of the wafers constituting the "wire in" surface are hereafter called the "wire in" edges. The "wire in" edges of the wafers have poorer quality compared to the other edges. Due to the damaged corner and especially the presence of relatively big cracks (>30 μιη) this will have an impact on breakage strength of the wafers. It appears that the problem with crack formation is due to the impinging force of abrasive particles coming from the steel wire of the wire saw and hits the edge of the wafers. This is a surprising discovery since the abrasive particles are very small, about 1-5 μηι in diameter, and are travelling at modest speeds of up to 10 - 15 m/s. However, the "wire in" surface is of major concern, but also on the "cutting in" surface some damage can be observed. The "cutting in" surface is the surface of the block which is closest to the wire web at the start of the sawing and also parallel to the wire web.

Summary of the Invention

The basic idea of the invention is to mask the "wire in" surface to provide a mechanical protection of the wafer edges.

Thus, the invention provides a method for cutting a block of silicon into a multiplicity of wafers of thickness less than 200 μπι, where

- a planar array of parallel running wires is moved relative to the block in a direction perpendicular to the plane of the running wires,

- an abrasive slurry is applied to the wires before the wires pass through the block, characterised in that

- the surface into which the wires enters the block perpendicularly is protected by a polymer coating with thickness in the range from 20 to 1000 μη .

The functionality of the coating is to provide a mechanical protection of the edge of the wafers during cutting. Thus the thickness of the coating should be sufficient to provide the mechanical protection, but on the other hand, the thickness should not constitute a barrier for the sawing slurry being carried along the cutting wire towards the cutting zone since reduction of the supply of sawing sludge may result in cutting marks or reduced cutting efficiencies. Thus the there is a trade-off between the need for mechanical protection and avoiding sawing sludge being hindered to reach the cutting zone, leading to a suitable thickness of the polymer coating in the range from 20 to 1000 μηι. Advantageously, the thickness of the coating is 20 μηι to 500 μηι, more advantageously 30 μπι to 300 μπι and most advantageously 40 μιη to 250 μηι. The invention may apply any polymer coating with a sufficient thickness and mechanical strength to provide an impact protection of the wafer edges from impinging abrasive particles loosened from the cutting wire, and which does not introduce non-acceptable pollution elements to the silicon wafers. Such polymers are known to those skilled in the art. However, from a cost perspective, the polymer coating should be removable during the conventional post-cutting washing process of the wafers in order to obtain an easy removal of the coating. Thus, it is

advantageous to apply the same polymer as coating of the "wire in"-surface as the polymer (glue) that is employed to fasten the silicon block to the holding substrate before and during the cutting process. Examples of suitable polymer coatings include, but is not limited to, epoxy resins, acrylic resins, and plastic foils.

It is also possible to apply a coating to the "cutting in" surface of the block prior to start of the cutting.

Preferably the abrasive slurry comprises abrasive particles and a solvent. The abrasive particles are typically silicon carbide particles or diamond particles.

The advantage of the invention is improvement of the quality of the "wire in" edges of the wafers. This gives wafers with higher breakage strength. The number of very weak wafers is reduced. It is also possible to improve the quality of the "cutting in" edges of the wafers which further contributes to increasing the quality of the wafers. Brief description of the figures

The invention will now be described by way of examples with reference to the accompanying figures where

Figs, la and lb show schematically a perspective view and an end view,

respectively, of a silicon block being sawn into wafers.

Figs. 2a, 2b, 2c and 2d show micrographs of four surfaces of a silicon block after it has been cut into wafers.

Fig. 3 is a micrograph illustrating the effect of a lacquer coating.

Fig. 4 is a micrograph illustrating the effect of a thinner layer of lacquer coating than in Fig. 3.

Fig. 5 is a micrograph illustrating the effect of an adhesive polymer coating.

Fig. 6 is a micrograph illustrating the effect of a glue coating.

Figs. 7a and 7b are two micrographs illustrating the effect of applying an adhesive tape with different thicknesses. Description of the invention

Figure la shows a perspective view of an example of a wafer cutting (sawing) apparatus. As shown in Fig. la, a first guide roller 10 feeds an array of cutting wires 1 1 to a second guide roller 12. The wire comes from a supply reel (not shown) and is drawn onto a collection reel (also not shown). An array of closely spaced cutting wires is used for producing the wafers. The guide rollers 10, 12 position the wires in a flat evenly spaced cutting array. An example of such an apparatus for producing a multiplicity of thin wafers from a block of solid material was described in UK Patent Specification 2,414,204. (It is known to have one or a few long wire(s) going more than 1000 times around the guide rollers.) There may be several thousand wires (or different sections of one or a few wires) spaced apart at a distance of 200 microns or less when measuring between adjacent wire surfaces. The arrow marked v w indicates the direction of the velocity of the wires. A silicon block 13, is slowly moved towards the array of wires (wire web) 1 1.

The apparatus has provision (not shown) for moving the block 13 downward through the array of wires 1 1. Fig. lb shows an end view of the same wafer cutting apparatus and silicon block 13 as in Fig. la. In Fig. lb a slurry device 14 is shown. The slurry device 14 (e.g. a manifold) is arranged above the array of wires 1 1 for supply of abrasive slurry. The flow of abrasive slurry is indicated by a downward pointing arrow from the slurry device 14 in Fig lb. The abrasive slurry typically comprises abrasive particles and a solvent. The abrasive particles can e.g. be silicon carbide particles or diamond particles. The solvent is typically Polyethylene glycol (PEG) or water.

In Figs, la and lb one wafer block 13 is shown. The wire saw is not limited to the type of saw illustrated in Figs, la and lb. One can, for example, use a saw which can cut two or more wafer blocks simultaneously. And, in Fig. l a, the block 13 is shown being moved down through the array of wires 1 1 (indicated by the vertical arrow). However, the block 13 could be moved in any direction provided that its movement is perpendicular to the parallel array of wires 1 1. Alternatively, the saw could be moved, what is important is the relative movement between the block(s) 13 and the array of wires 1 1.

The silicon block 13 has a "wire in" surface 15 and a "cutting in" surface 16. The "wire in" surface 15 is the surface where the wires enter into the block 13 perpendicularly. The "cutting in" surface 16 is the surface of the block 13 which is parallel to the wire web 1 1, and at the start of the cutting, closest to the wire web 1 1.

After the sawing process the block 13 has been cut into a multiplicity of wafers, a wafer stack. The "wire in" surface 15 has been transformed into a multiplicity of "wire in" edges of the resulting wafers. Likewise, the "cutting in" surface 16 has been transformed into a multiplicity of "cutting in" edges of the resulting wafers.

The appearance of the wafers in the wafer stack differs on the edges after the cutting, as shown in Figs. 2a, 2b 2c and 2d. Particularly one edge (Fig. 2a) shows a dull appearance with bright speckles in it. The other three edges are more uniform and have the same or almost the same shiny appearance as the original block surface before the block is sliced into wafers.

As seen in Fig. 2a the surface with speckles corresponds to the "wire in" edges of the wafers. The "wire in" edges have poorer quality compared to the other edges. Due to the damaged corner and especially the presence of relatively big cracks (>30 μηι) and indentations, this will have an impact on the breakage strength of the wafers. It is assumed that the indentations and cracks are caused by impact of the abrasive particles in the slurry hitting the block's "wire in" surface during the wafer cutting process.

As seen in Fig. 2b the "cutting in" edges of the wafers also show some cracks and damage, but to a significant less extent than the "wire in" edges. As seen in Figs. 2c and 2d the "cutting out" edges and the "wire out" edges of the wafers show no damage. The "cutting out" edge is the edge opposite of the "cutting in" edge of the wafer. The "wire out" edge is the edge opposite of the "wire in" edge of the wafer. According to the invention it has been found that application of a protective coating effectively prevents damage to the "wire in" surfaces of the silicon block (and the resulting wafer edges) when added to the surface before the wafer cutting takes place. The coating layer thickness will depend on the actual coating material chosen, but will normally be below 1 mm. It may also be possible to use a thicker coating, but this will likely not improve the results, and may cause increased chance of problems with the cutting process and can also imply more elaborate washing and cleaning procedures of the wafers.

Typically, the coating thickness is between 20 and 500 μπι. More preferably the coating thickness is between 30 and 300 μπι and most preferably the coating thickness is between 40 and 250 μηι.

The coating can be applied by hand or by an automated application device. The application device can either be separated from the cutting apparatus or can be a part of the cutting apparatus.

The coating can typically comprise a polymer material. The "cutting in" surface 16 of the silicon block can also be improved by application of a protective coating.

Removal of the protective coating from the silicon wafer edge after cutting is performed with standard washing and de-gluing procedure with good results. Examples

Examples are given below for typical coatings applied on the "wire in" surface of a silicon block. It is also possible to apply other materials as a protective coating than the example materials described below. Example 1

A region of the "wire in" surface of a silicon block was coated by spraying with lacquer (Rust-Oleum Hard Hat 2500 clear). The thickness of the coating was approx. 100 - 250 μηι. The rest of the "wire in" surface remained uncoated for comparison with the coated area. The silicon block was then cut into a multiplicity of 180 μηι thick wafers. After cutting of the silicon wafers an optical microscope photo was taken of the wafer edges while the wafers were still in the wafer stack, see Fig. 3. The dark stripes in the photo are the gaps between the wafers. In the photo the lacquer coated region of the "wire in" edges of the wafers is marked with an arrow A. The wafer edges in the lacquer coated region showed significantly lower damage than the part of the wafer edges that had not been coated. The protection provided by the lacquer coating is clearly illustrated.

Example 2

A region of the "wire in" surface of a silicon block was coated with the same material as in Example 1 , i.e. lacquer (Rust-Oleum Hard Hat 2500 clear), but a thinner layer was applied. The thickness of the coating was approx. 50 - 100 μπι.

The rest of the "wire in" surface remained uncoated. The silicon block was then cut into a multiplicity of 180 μηι thick wafers. Fig. 4 is a micrograph which shows the "wire in" edges of the wafers. The lacquer coated region of the "wire in" edges is marked with an arrow B. The wafer edges in lacquer coated region still showed significantly lower damage than the uncoated regions, but the thicker layer used in example 1 was slightly more efficient.

Example 3

In the next test an adhesive polymer film (Fellowes laminating pouch 80) was attached to a region of the "wire in" surface of the silicon block. The thickness of the polymer film was approx. 80 μηι. The rest of the "wire in" surface remained uncoated. The silicon block was then cut into a multiplicity of 180 μπι thick wafers. After cutting of the silicon wafers an optical microscope photo was taken of the wafer edges, see Fig. 5. The dark stripes in the photo are the gaps between the wafers. In the photo the coated region of the "wire in" edges of the wafers is marked with an arrow C. The wafer edges in the coated region showed significantly lower damage than the part of the wafer edges that had not been coated. The protection provided by the adhesive polymer film is clearly illustrated. Example 4

On the "wire in" surface of the silicon block a region of the surface was coated with epoxy glue. The thickness of the glue was approx. 50 - 250 μηι. The rest of the "wire in" surface remained uncoated. The silicon block was then cut into a multiplicity of 180 μπι thick wafers. Fig. 6 is a micrograph which shows the "wire in" edges of the wafers. The glue coated regions of the "wire in" edges are marked with an arrow D. The wafer edges in the glue coated region showed significantly lower damage than the part of the wafer edges that had not been coated. This illustrates the protection provided by the glue coating. Example 5

A region of the "wire in" surface of the silicon block was coated with adhesive tape (TartanTM invisible) with two different thicknesses. The rest of the "wire in" surface remained uncoated. The silicon block was then cut into a multiplicity of 180 μηι thick wafers. After cutting of the silicon wafers, two optical microscope photos were taken of the wafer edges, see Figs. 7a and 7b. In Fig. 7a the thickness of the adhesive tape was approx. 50 μηι (1 layer). In the photo the coated region of the "wire in" edges of the wafers is marked with an arrow E.

In Fig. 7b the thickness of the adhesive tape was approx. 500 μιη (10 layers of tape was applied). The wafer edges in the region covered with ten layers of adhesive tape on the "wire in" edges of the wafers are marked with an arrow F.

The wafer edges in the tape coated regions showed significantly lower damage than the uncoated regions. This illustrates the protection provided by the adhesive tape. The protective effect increased slightly by using a thicker coating (i.e. ten layers of adhesive tape compared to one).