WAFER PROCESSING PROCESS

Abstract

Method for processing a wafer (11) wherein components (15) are each formed in areas of a front surface side (11a) which is divided by several roads (13) and wherein patterns (17) which include a metal layer are formed on the roads (13), the method comprising: a laser processing step for applying a laser beam of such wavelength that it is absorbed in the wafer (11) along the roads (13) which are provided with the patterns (17) to form laser-machined grooves (11c) while the patterns (17) are removed; a training step for a cut groove (11d) to form cut grooves (11d) having a depth deeper than a finished thickness of the wafer (11) into which laser-machined grooves (11c) are cut by a cutting blade (28) thinner than a width of the laser-machined grooves (11c); an attachment step for a protective element (29) for attaching a protective element (29) to the front surface side (11a) of the wafer (11) which is formed with the cut grooves (11d); a grinding step to hold the wafer (11) by a clamping table (12) by a protective element (29), grinding a rear surface side (11b) of the wafer (11) to thin the wafer (11) to a finished thickness and to expose the cut grooves (11d) on the rear surface (11b) of the wafer (11), thereby dividing the wafer (11) into several component chips (31); a removal step for a damaged layer to remove a damaged layer that was formed on the rear surface side (11b) of the wafer (11) by grinding the wafer (11); and a training step for a deformed layer (33) to form a deformed layer (33) on the rear surface side (11b) of the wafer (11) from which the damaged layer (33) was removed, by plasma processing using an inert gas.

Description

BACKGROUND OF THE INVENTIONTechnical field

[0001] The present invention relates to a method for processing a wafer in which patterns including a metal layer are formed on streets. Description of the state of the art

[0002] A wafer formed with components such as integrated circuits (ICs) on the front surface is divided along streets (division lines), thereby obtaining multiple component dies, each including the components. Wafer division is performed, for example, using a cutting device to which a ring-shaped cutting blade is attached for cutting the wafer. The cutting blade is rotated and brought into contact with the wafer along the streets, thereby cutting and dividing the wafer.

[0003] The component chips obtained by dividing a wafer are incorporated into various electronic devices. In recent years, the reduction of the size and thickness of electronic devices has been emphasized, so reducing the size and thickness of the component chips has also been demanded. In view of this, a technique of grinding the back surface of the wafer with a grindstone to make the component chips thin has been used. When the back surface of the wafer is ground by the grindstone, small bumps (small protrusions and recesses) and cracks are formed in the ground area. Since the presence of the areas formed with bumps and cracks (damaged layer) would reduce the strength of the wafer, the damaged layer is removed by polishing with a polishing wheel or dry etching or the like after grinding.

[0004] On the other hand, it is known that when the damaged layer remains on the wafer, a gettering effect is obtained, by which metallic elements such as copper contained in the interior of the wafer are incorporated into the damaged layer. Therefore, when the damaged layer is formed on the back surface side of the wafer, the metallic elements are incorporated on the back surface side of the wafer, whereby the metallic elements can be prevented from moving into the devices formed on the front surface side of the wafer. As a result, the occurrence of defects (current leakage or the like) in the devices due to the metallic elements in the wafer can be prevented. If the damaged layer is removed to improve the strength of the wafer, the gettering effect is lost.Therefore, a method has been proposed in which small irregularities or cracks (stress) finer than those of the damaged layer are formed on the back surface of the wafer after removing the damaged layer, and metallic elements are incorporated into the area formed with this deformation (deformed layer). This allows a gettering effect to be maintained on the metallic elements without reducing the strength of the wafer die. Japanese Patent Application Laid-Open No. 2010-177430 A discloses a technique in which, after a damaged layer formed by grinding the back surface of the wafer is removed by plasma etching, an inert gas formed into plasma is applied to the back surface of the wafer to form a deformed layer (getter layer).

[0005] US 2009 / 0 186 465 A1 concerns a wafer dividing process.

[0006] US 2015 / 0 255 288 A1 relates to a machining method for a plate-like workpiece.

[0007] US 2004 / 0 112 880 A1 relates to a laser processing method for processing a workpiece by applying a laser beam.

[0008] US 2014 / 0 017 880 A1 discloses laser scribing and plasma etching, followed by grinding of a rear side to separate chips. DESCRIPTION OF THE INVENTION

[0009] Patterns including a metal layer that does not form device elements, such as a test element group (TEG) for device evaluation and columnar metallic patterns (studs) for supporting an insulation film, or the like, are often formed on the streets that divide the devices formed on the wafer at the time of device manufacturing. Since the patterns (test patterns) do not affect the use of the device chips obtained by dividing the wafer, they are cut and removed together with the wafer by a dicing blade at the time of dicing the wafer. However, when the test patterns including the metal layer are cut by the dicing blade, the dicing blade and the metal layer touch each other, and protrusions (burrs) of the metal are generated.When the plasma processing for forming the aforementioned deformed layer or the like is applied to the wafer formed with burrs, an electric discharge may occur in the region formed with the burrs, thereby making it possible to damage the components.

[0010] The present invention was made in consideration of the above-mentioned problem. Therefore, an object of the present invention is to provide a wafer processing method that can suppress the occurrence of electric discharge at the time of plasma processing of a wafer and prevent damage to components.

[0011] In accordance with one aspect of the present invention, there is provided a method for processing a wafer in which components are each formed in regions on a front surface side divided by a plurality of streets, and in which patterns including a metal layer are formed on the streets, the method including: a laser processing step of applying a laser beam of such a wavelength that it can be absorbed in the wafer along the streets formed with the patterns to form laser-machined grooves while removing the patterns; a cut groove forming step of forming cut grooves having a depth deeper than a finished thickness of the wafer in the laser-machined grooves by a cutting blade thinner than a width of the laser-machined grooves;a protective member attachment step for attaching a protective member to the front surface side of the wafer formed with the cut grooves; a grinding step for holding the wafer by a chuck table with the protective member, grinding a back surface side of the wafer to thin the wafer to a finished thickness and expose the cut grooves on the back surface side of the wafer, thereby dividing the wafer into a plurality of device chips; a removal step for removing a damaged layer formed on the back surface side of the wafer by grinding the wafer; and a deformed layer forming step for forming a deformed layer on the back surface side of the wafer where the damaged layer has been removed by plasma processing using an inert gas.

[0012] Note that preferably, in the damaged layer removing step, the damaged layer is removed by polishing with a polishing pad. Furthermore, in the damaged layer removing step, the damaged layer is preferably removed by plasma etching using a halogen-containing gas. In addition, the wafer processing method preferably further includes a protective film forming step of forming a water-soluble protective film on the front surface side of the wafer before the laser processing step and a protective film removing step of removing the protective film from the front surface side of the wafer after the laser processing step.

[0013] In the wafer processing method according to the described aspect of the present invention, after the test patterns including a metal layer formed on the streets of the wafer are removed by applying the laser beam, the wafer is cut along the streets. This prevents the test patterns from being cut by the cutting blade and suppresses the occurrence of metal burrs at the time of cutting the wafer. Therefore, at the time of applying plasma etching to the wafer to form the deformed layer on the back surface of the wafer, the occurrence of electric discharge is prevented and damage to the components is prevented.

[0014] The above and other objects, features and advantages of the present invention and the mode of carrying them out will become clearer and the invention itself best understood by studying the following description and appended claims with reference to the appended figures which show a preferred embodiment of the invention. SHORT DESCRIPTION OF THE CHARACTERS Fig. 1 is a plan view showing a configuration example of a wafer; Fig. Fig. 2 is a perspective view of the wafer in the state where it is supported by an annular frame; Fig. 3 is a side sectional view, partially in cross section, illustrating the manner in which the protective film is applied to the wafer; Fig. 4A is a side view, partially in cross section, illustrating the manner in which a laser beam is applied to the wafer; Fig. 4B is an enlarged plan view of a road formed with patterns including a metal layer; Fig. 4C is an enlarged plan view of a road from which patterns including a metal layer have been removed and in which a laser-machined groove has been formed; Fig. 5 is a side view, partially in cross section, illustrating the manner in which a protective film on the wafer is removed; Fig. 6A is a side view, partially in cross section, illustrating the manner in which a cut groove is formed on the wafer; Fig. 6B is an enlarged plan view of the road provided with the cut groove; Fig. 7A is a perspective view illustrating the manner in which a protective member is attached to the wafer; Fig. 7B is a perspective view of the wafer to which the protective member has been attached; Fig. 8 is a side view illustrating a manner in which the back surface side of the wafer is ground; Fig. 9 is a side view illustrating the manner in which the back surface side of the wafer is polished; Fig. 10 is a schematic sectional view illustrating a configuration example of a plasma processing apparatus; and Fig. 11 is an enlarged sectional view of the wafer in the state where it is formed with a deformed layer. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

[0015] An embodiment of the present invention will be described below with reference to the attached figures. First, an example of a wafer to be processed by the wafer processing method according to the present embodiment will be described. Fig. 1 is a plan view showing a configuration example of a wafer 11.

[0016] The wafer 11 is formed in a disc shape from a material such as silicon and divided into a plurality of regions by a plurality of streets (dividing lines) 13 arranged in a lattice pattern. In addition, a component 15 including an IC or the like is formed in each of the regions divided by the streets 13 on the front surface side of the wafer 11. Note that while the disc-shaped wafer 11 formed from a material such as silicon is used in the present embodiment, the material, shape, structure, size, and the like of the wafer 11 are not limited. For example, the wafer 11 formed from a material such as a semiconductor other than silicon, a ceramic, a plastic, or a metal can also be used. Similarly, the type, number, shape, structure, size, layout, and the like of the components 15 are not limited.

[0017] In addition, at least a portion of the streets 13 are formed with patterns 17 including a metal layer. The patterns 17 are patterns (test patterns) including a metal layer that is not a forming material of the components 15 and are formed, for example, from a TEG for component evaluation or columnar metal patterns (pillars) for supporting an insulation film formed at the time of manufacturing the components, or the like on the streets 13. Note that an example in which the patterns 17 are formed on a portion of the streets 13 is shown in Fig. 1, wherein the patterns 17 can be arranged on all the streets 13.

[0018] By dividing the wafer 11 along the streets 13, a plurality of device chips each including the devices 15 are obtained. In the present embodiment, an example will be described in which grooves having a depth less than the thickness of the wafer 11 are first formed (half-cut) on the front surface side of the wafer 11, and thereafter, the back surface side of the wafer 11 is ground by a grindstone, thereby dividing the wafer 11 into a plurality of device chips.

[0019] At the time of forming the grooves on the front surface side of the wafer 11, for example, a cutting device equipped with an annular cutting blade may be used. In the case of using the cutting device, the cutting blade is rotated and brought into contact with the wafer 11 to thereby form the grooves in the wafer 11. However, it should be noted that since at least a part of the streets 13 are formed with the patterns 17 including the metal layer, forming the grooves along the streets 13 by this method results in patterns 17 being cut by the cutting blade. The patterns 17 include the metal layer, so when the metal layer is cut by the cutting blade, the metal layer is expanded by the cutting blade, so that metal projections (burrs) are formed.Where the burrs remain on the wafer 11 when plasma processing is performed in a later step (for example, in a strained layer forming step described later), electrical discharge may occur in the areas where the burrs are formed, possibly resulting in damage to the device 15.

[0020] In view of this, in the wafer processing method according to the present embodiment, before cutting the wafer 11 with the cutting blade, a laser beam is applied to the wafer 11 to remove the patterns 17. This can avoid cutting the patterns 17 by the cutting blade, thereby preventing the occurrence of burrs, and suppressing the occurrence of electric discharge at the time of plasma processing. The wafer processing method according to the present embodiment will be described in detail below.

[0021] First, in order to hold the wafer 11 on each of various processing devices such as a cutting device and a laser processing device, the wafer 11 is supported by an annular frame. Fig. Fig. 2 is a perspective view of the wafer 11 in the state in which it is supported by the annular frame 21. Note that the patterns 17 in the illustration in Fig. 2 are omitted. The annular frame 21 is attached along an outer periphery of a disk-shaped adhesive tape 19 formed of a plastic or the like, and a rear surface 11b side of the wafer 11 is attached to an adhesive tape 19. As a result, the wafer 11 is supported on the annular frame 21 in the state where its front surface 11a is exposed upward.

[0022] Next, a water-soluble protective film is formed on the front surface side 11a of the wafer 11 (protective film forming step). Fig. 3 is a side view, partially in cross section, illustrating a manner in which the water-soluble protective film 23 is formed on the front surface 11a of the wafer 11. The protective film 23 can be formed using a spin coater 2, for example. The spin coater 2 includes a turntable 4 supporting the wafer 11, a plurality of clamps 6 fixing the annular frame 21 supporting the wafer 11, and a nozzle 8a for ejecting a material for the protective film 23, which includes a water-soluble resin or the like, toward the wafer 11.

[0023] First, the wafer 11 is placed on the turntable 4, and the annular frame 21 is fixed by the clamps 6. A part of the upper surface of the turntable 4 forms a holding surface 4a for suction-holding the wafer 11 by the adhesive tape 19. The holding surface 4a is connected to a suction source (not shown) through a suction passage (not shown) or the like formed in the turntable 4. Negative pressure of the suction source is caused to act on the holding surface 4a, whereby the wafer 11 is held under suction by the turntable 4. Note that a chuck table for holding the wafer 11 by a mechanical method, an electrical method, or the like may be used instead of the turntable 4.

[0024] Then, while the turntable 4, which suction-holds the wafer 11 by the holding surface 4a, is rotated about a rotation axis substantially parallel to a vertical direction, a water-soluble resin such as polyvinyl alcohol (PVA) or polyethylene glycol (PEG) is ejected from the nozzle 8a disposed on the upper side of the turntable 4. As a result, the water-soluble resin attached to the wafer 11 flows toward an outer peripheral portion of the wafer 11 under centrifugal force, thereby forming the water-soluble protective film 23 on the front surface 11a side of the wafer 11. The protective film 23 serves to prevent particles (contaminants) scattered from a processing area at the time of applying the laser beam from the front surface 11a of the wafer 11 in a later step from adhering to the front surface 11a of the wafer 11.It should be noted, however, that the protective film forming step need not necessarily be performed, and the protective film forming step may be omitted, for example, in the case where the application of the laser beam is performed under such processing conditions that contamination hardly occurs or in the case where the adhesion of contamination is unlikely to be a problem.

[0025] Subsequently, a laser beam having a wavelength such that it is absorbed in the wafer 11 is applied along the streets 13 formed with the patterns 17 to form laser-machined grooves while removing the patterns 17 (laser machining step). The application of the laser beam to the wafer 11 is performed using a laser machining apparatus. Fig. 4A is a side view, partially in cross section, illustrating the manner in which the laser beam is applied to the wafer 11.

[0026] The laser processing apparatus 10 includes a chuck table 12 that holds the wafer 11 under suction, a laser forming unit 14 that applies the laser beam of such a wavelength that it is absorbed in the wafer 11, and a plurality of clamps 16 that fix the annular frame 21 that supports the wafer 11.

[0027] First, the wafer 11 is placed on the chuck table 12, and the annular frame 21 is fixed by the clamps 16. A part of the upper surface of the chuck table 12 forms a holding surface 12a that holds the wafer 11 by suction through the adhesive tape 19. The holding surface 12a is connected to a suction source (not shown) through a suction passage (not shown) or the like formed in the chuck table 12. Note that a chuck table for holding the wafer 11 by a mechanical method, an electrical method, or the like may be used instead of the chuck table 12. In a state where the wafer 11 is placed on the holding surface 12a of the chuck table 12 by the adhesion-promoting tape 19, a negative pressure of the suction source is caused to act on the holding surface 12a, whereby the wafer 11 is held under suction by the chuck table 12.Then, the chuck table 12 with the wafer 11 held thereon is moved to a position below the laser deposition unit 14.

[0028] Next, the laser beam is applied from the laser application unit 14 onto the wafer 11. The laser application unit 14 has a function of applying a pulsed laser beam of such a wavelength that it can be absorbed in the wafer 11. Note that in the case where the protective film 23 is applied to the front surface 11a of the wafer 11, the laser beam is applied to the wafer 11 through the protective film 23. During the application of the laser beam from the laser application unit 14 onto the wafer 11, the chuck table 12 is moved along the longitudinal direction of the streets 13, so that the laser beam is applied along the streets 13. Thereby, the wafer 11 is subjected to ablation processing and a straight laser-machined groove 11c is formed along the street 13 on the front surface side 11a of the wafer 11.

[0029] In addition, when the laser beam is applied to the road 13 formed with the pattern 17, the patterns 17 are removed by the application of the laser beam. Fig. Figure 4B is an enlarged plan view of the road 13 formed with the patterns 17. Fig. 4B illustrates an example in which TEG 25 for component evaluation and pillars 27 as columnar metal patterns for supporting an insulation film or the like on the road, which are formed at the time of forming the components 15, are formed as the patterns 17. Note that in Fig. 4B the areas where the TEG 25 and the columns 27 are formed are hatched.

[0030] When the laser beam is applied along the road 13, the laser is also applied to the TEG 25 and the pillars 27, thereby removing the TEG 25 and the pillars 27. Fig. Figure 4C is an enlarged top view of street 13 from which the TEG 25 and the columns 27 have been removed and in which the laser-machined groove 11c has been formed. In Fig. In Figure 4C, the area where the laser-machined groove 11c was formed is shaded. Note that the patterns 17 are mainly used in the production process of the components 15 and do not affect the operation of the component chips obtained by dividing the wafer 11, so removing the patterns 17 does not affect the function of the component chips. Laser beam application conditions (the power, beam diameter, and repetition frequency of the laser beam, etc.) are set in such a way that the laser-machined grooves 11c are formed in the wafer 11 and the patterns 17 can be removed.

[0031] Note that in the laser processing step, the laser beam may be applied multiple times along the same streets 13. In this case, if the application position of the laser beam is gradually changed in the width direction of the street 13, the width of the laser-processed groove 11c formed along the street 13 can be controlled. In addition, it is not necessary to completely remove the patterns 17 by applying the laser beam, and it is sufficient to apply the laser beam in such a manner that at least the width of the laser-processed groove 11c is larger than the width of the cutting blade used for cutting the wafer 11 in a later step. Fig. 4C shows an example in which a part of the pillars 27 remain on the roads 13. Moreover, the laser beam application in the laser processing step may be performed for all the roads 13 or only on the roads 13 formed with the pattern 17. In the case where the laser beam application is performed only on the roads 13 on which patterns 17 are formed, the roads 13 formed with the patterns 17 are detected in advance.

[0032] Subsequently, the protective film 23 is removed from the front surface side 11a of the wafer 11 (protective film removing step). Fig. 5 is a side view, partially in cross section, illustrating the manner in which the water-soluble protective film 23 formed on the front surface side 11a of the wafer 11 is removed.

[0033] First, in the protective film forming step, the wafer 11 is held under suction by the turntable 4. Then, while the turntable 4, with the wafer 11 held under suction by the holding surface 4a, is rotated about a rotation axis substantially parallel to the vertical direction, pure water is ejected from the nozzle 8b located on the upper side of the turntable 4. Thereby, the water-soluble protective film 23 formed on the front surface 11a of the wafer 11 is removed together with the contaminants deposited on the wafer 11. Consequently, since the protective film 23 is formed using a water-soluble resin (protective film forming step), it can be washed away by pure water. Therefore, the protective film 23 can be removed very easily.Note that the removal of the protective film 23 can be performed by ejecting the wafer 11 with a mixed fluid of pure water and a gas (e.g., air, etc.). Note that in the case where the protective film forming step is not performed and no protective film is formed on the front surface 11a of the wafer 11, it is not necessary to perform the protective film removing step.

[0034] Next, using a cutting blade thinner than the width of the laser-machined grooves 11c formed in the laser-machining step, cut grooves are formed in the laser-machined grooves 11c having a depth deeper than a finished thickness of the wafer 11 (cut groove forming step). Fig. 6A is a side view, partially in cross section, showing the manner in which the cut groove 11d is formed in the wafer 11.

[0035] In the cut groove forming step, cut grooves are formed using a cutting device 18. As shown in Fig. 6A, the cutting device 18 includes a chuck table 20 that holds the wafer 11 under suction, a cutting unit 22 that cuts the workpiece held by the chuck table 20, and a plurality of clamps 24 that fix the annular frame 21 that supports the wafer 11.

[0036] First, the wafer 11 is placed on a chuck table 20, and the annular frame 21 is fixed by the clamps 24. A part of an upper surface of the chuck table 20 forms a holding surface 20a for suction-holding the wafer 11 by the adhesive tape 19. The holding surface 20a is connected to a suction source (not shown) through a suction passage (not shown) or the like formed in the chuck table 20. A negative pressure of the suction source is caused to act on the holding surface 20a, whereby the wafer 11 is held under suction by the chuck table 20. Note that the chuck table for holding the wafer 11 may be used by a mechanical method, an electrical method, or the like instead of the chuck table 20.

[0037] The cutting unit 22 for cutting the wafer 11 is arranged on the upper side of the chuck table 20. The cutting unit 22 includes a spindle 26 having an axis in a direction substantially parallel to the holding surface 20a, and an annular cutting blade 28 is fixed to a tip end portion of the spindle 26. The cutting blade 28 is formed, for example, from an electrodeposited grindstone in which abrasive grains are bonded by nickel plating. In addition, the spindle 26 is connected to a rotational drive source (not shown) such as a motor, and the cutting blade 28 fixed to the spindle 26 is rotated by a power transmitted from the rotational drive source.

[0038] In the cut groove forming step, first, the wafer 11 is placed on the holding surface 20a of the chuck table 20 by the adhesive tape 19, and a negative pressure of the suction source is applied. As a result, the wafer 11 is held by the chuck table 20 in a state where the front surface 11a thereof is exposed upward.

[0039] Next, the spindle 26 is rotated, the cutting blade 28 is caused to cut into the lower portion of the laser-machined groove 11c from the front surface 11a of the wafer 11 along the road 13, and the chuck table 20 is cut in a machining feed direction, namely, a direction substantially parallel to the support surface 20a and substantially perpendicular to the axis of the spindle 26. As a result, a straight-cut groove 11d is formed in the lower portion of the laser-machined groove 11c along the road 13. At the time the cutting blade 28 is caused to cut the wafer 11, the distance from the lower surface 11a of the wafer 11 to a lower end of the cutting blade 28 is set to a value that is less than the thickness of the wafer 11 and exceeds a finished thickness of the wafer 11.The finished thickness 11 corresponds to the thickness of the component chips when the wafer 11 is finally processed into the component chips.

[0040] Note that the width of the cutting blade 28 is smaller than the width of the laser-machined grooves 11c formed in the laser-machined groove forming step. In the cut groove forming step, the cutting blade 28 is caused to cut into a part of the lower portion of the laser-machined groove 11. Therefore, the cut groove 11d is formed inside the laser-machined groove 11c. Fig. Fig. 6B is an enlarged plan view of the road 13 formed with the cut groove 11d. The cut groove 11d has a narrower width than the width of the laser-machined groove 11c and is formed in a region within the laser-machined groove 11c in a plan view. Note that in Fig. 6B the area formed with the cut groove 11d is hatched.

[0041] The lower portion of the laser-machined groove 13c formed in the laser processing step is a region where the patterns 17 have been removed and no metal layer remains. In the cut groove forming step, the lower portion of the laser-machined groove 13c is cut by the cutting blade 28. Therefore, cutting of the patterns 17 by the cutting blade 28 does not occur, so that the formation of burrs due to cutting of the metal layer can be prevented.

[0042] Subsequently, a protective member is attached to the front surface side 11a of the wafer 11 (protective member attachment step). Fig. 7A is a perspective view illustrating the manner in which the protective member 29 is attached to the wafer 11. As in Fig. As shown in Fig. 7A, the protective member 29 is a disk-shaped member attached to the front surface 11a of the wafer 11 so as to cover the components 15. As the protective member 29, for example, a tape formed of a flexible plastic or the like can be used. Fig. Fig. 7B is a perspective view illustrating the wafer 11 to which the protective member 29 has been attached. The protective member 29, the components 15 formed on the front surface 11a of the wafer 11, are protected in later steps (grinding step, a damaged layer removal step, and a deformed layer formation step). Note that Fig. 7B illustrates a state in which the wafer 11 has been peeled off from the adhesive tape 19 and detached from the annular frame 21.

[0043] Next, the back surface side 11b of the wafer 11 is ground to thin the wafer 11 to a finished thickness and expose the cut grooves 11d on the back surface 11b, thereby dividing the wafer 11 into a plurality of device chips (grinding step). Fig. Fig. 8 is a side view illustrating the manner in which the back surface side 111b of the wafer 11 is ground.

[0044] The grinding of the wafer 11 is carried out, for example, using a grinding device 30 which is Fig. 8. The grinding apparatus 30 includes a chuck table 32 for suction-holding the wafer 11. The chuck table 32 is connected to a rotational drive source (not shown) such as a motor and rotates about a rotation axis substantially parallel to the vertical direction. Additionally, a moving mechanism (not shown) is provided below the chuck table 32, and the moving mechanism has a function of moving the chuck table 32 in a horizontal direction.

[0045] An upper surface of the chuck table 32 forms a holding surface 32a for suction-holding the wafer 11. The holding surface 32a is connected to a suction source (not shown) through a suction passage (not shown) or the like formed in the chuck table 32. In the state where the wafer 11 is placed on the holding surface 32a through the protective member 29, a negative pressure of the suction source is caused to act on the holding surface 32a, whereby the wafer 11 is held under suction by the chuck table 32. Note that a chuck table for holding the wafer 11 by a mechanical method, an electrical method, or the like may be used instead of the chuck table 32.

[0046] A grinding unit 34 is arranged on the upper side of the chuck table 32. The grinding unit 34 includes a spindle housing (not shown) supported by a lifting mechanism (not shown). A spindle 36 is housed in the spindle housing, and a disk-shaped fixture 38 is fixed to a lower end portion of the spindle 36. A grinding wheel 40, which has a diameter substantially the same as the fixture 38, is fixed to a lower surface of the fixture 38. The grinding wheel 40 includes a wheel base 42 formed of a metallic material such as stainless steel or aluminum. A plurality of grindstones 44 are arranged on a lower surface of the wheel base 42.A rotational drive source (not shown) such as a motor is connected to the upper end (base end) of the spindle 36, and the grinding wheel 40 is rotated about a rotation axis substantially parallel to the vertical direction by a force generated by the rotational drive source. A nozzle (not shown) for supplying a grinding liquid such as pure water to the wafer 11 and the like is provided inside or near the grinding unit 34.

[0047] At the time of grinding the wafer 11, first, in a state where the wafer 11 is held on the holding surface 32a by a protective member 29, a negative pressure of the suction source is caused to act on the holding surface 32a. As a result, the wafer 11 is held under suction by the chuck table 32 in the state where the back surface side 11b is exposed upward.

[0048] Subsequently, the chuck table 32 is moved to a position below the grinding unit 34. Then, while the chuck table 32 and the grinding wheel 40 are respectively rotated and while the grinding fluid is supplied to the back surface 11b of the wafer 11, the spindle 36 is lowered. Note that the position and lowering speed of the spindle 36 are adjusted in such a manner that the lower surface of the grindstones 44 are pressed against the back surface 11b of the wafer 11 with an appropriate force. Thereby, the back surface 11b is ground and the wafer 11 is thinned. When the wafer 11 is thinned and the cut grooves 11d are exposed to the back surface 11b of the wafer 11, the wafer 11 is divided into the plurality of device chips, each including the devices 15. Then, when the wafer 11 is thinned to the finished thickness, the grinding is completed.Note that the wafer 11 is ground using one set of grinding units in the present embodiment, but the wafer 11 may be ground using two or more sets of grinding units. In that case, for example, rough grinding is performed using grindstones containing abrasive grains with a larger diameter, and finish grinding is performed using grindstones containing abrasive grains with a smaller diameter, whereby the flatness of the back surface 11b can be improved without increasing the time required for grinding.

[0049] When the back surface 11b of the wafer 11 is ground by the grindstones, small asperities (small protrusions and recesses) or cracks are formed on the back surface 11b of the wafer 11. The presence of a region (damaged layer) where strength or cracks are formed leads to a decrease in the strength of the device chips obtained by dividing the wafer 11. Therefore, the damaged layer formed on the back surface 11b of the wafer 11 is removed (damaged layer removal step). The damaged layer can be removed, for example, by polishing performed using a polishing wheel. Fig. Fig. 9 is a side view illustrating the manner in which the back surface side 11b of the wafer 11 is polished by a polishing pad.

[0050] The polishing of the wafer 11 is carried out, for example, using a polishing device 46 which is Fig. 9. The polishing apparatus 46 includes a chuck table 48 for suction-holding the wafer 11. The chuck table 48 is connected to a rotational drive source (not shown) such as a motor and is rotated about a rotation axis substantially parallel to the vertical direction. In addition, a moving mechanism (not shown) is provided below the chuck table 48, and the moving mechanism has a function of moving the chuck table 48 in a horizontal direction. An upper surface of the chuck table 48 forms a holding surface 48a for suction-holding the wafer. The holding surface 48a is connected to a suction source (not shown) through a suction passage (not shown) or the like formed in the chuck table 48.In the state where the wafer 11 is held on the holding surface 48a by the protective member 29, a negative pressure of the suction source is caused to act on the holding surface 48a, whereby the wafer 11 is held under suction by the chuck table 48. Note that a chuck table for holding the wafer 11 by a mechanical method, an electrical method, or the like may be used instead of the chuck table 48.

[0051] A polishing unit 50 is arranged on the upper side of the chuck table 48. The polishing unit 50 includes a spindle housing (not shown) supported by a lifting mechanism (not shown). A spindle 52 is housed in the spindle housing, and a disk-shaped fixture 54 is fixed to a lower end portion of the spindle 52. A polishing disk 56 is fixed to a lower surface of the fixture 54. The polishing disk 56 includes a polishing cloth formed of a non-woven fabric, a polyurethane foam, or the like. A rotational drive source (not shown) including a motor or the like is connected to the upper end (base end) side of the spindle 52, and the polishing disk 56 is rotated about a rotation axis substantially parallel to the vertical direction by a force generated by the rotational drive source.Note that the polishing unit 50 may be formed with a supply passage (not shown) for supplying a polishing liquid to the wafer 11 held on the chuck table 48. For example, a suspension containing abrasive grains dispersed in a liquid that reacts with the wafer 11 may be supplied as the polishing liquid to the wafer 11 through the supply passage.

[0052] In the damaged layer removing step, first, the protective member 29 is brought into contact with the holding surface 48a of the chuck table 48, and the negative pressure of the suction source is caused to act thereon. As a result, the wafer 11 is held under suction by the chuck table 48 in the state where the back surface 11b thereof is exposed upward. Note that the wafer 11 has been divided into a plurality of device chips 31 by the grinding step. Next, the chuck table 48 is moved to a position below the polishing unit 50. Then, while the polishing liquid is supplied to the back surface 11b of the wafer 11, the chuck table 48 and the polishing disk 56 are respectively rotated, and the spindle 52 is lowered.Note that the lowering amount of the spindle 52 is adjusted to such an extent that a lower surface (polishing surface) of the polishing pad 56 is pressed against the back surface 11b of the wafer 11. In this way, the back surface 11b of the wafer 11 is polished, thereby removing the damaged layer formed on the back surface 11b of the wafer 11 by grinding in the grinding step. By removing the damaged layer, the strength of the device chips 31 can be improved.

[0053] Note that while wet polishing, in which the polishing liquid is supplied to the wafer 11, was described above, dry polishing, in which the polishing liquid is not used, can also be used to polish the wafer 11. In addition, the method for removing the damaged layer is not limited to polishing by the polishing pad 56. For example, plasma etching using a halogen gas can be used to remove the damaged layer. The details of plasma etching will be described later.

[0054] Next, using plasma etching using an inert gas, deformations, small asperities (small protrusions and depressions), or cracks are formed on the back surface 11b of the wafer 11 (deformation layer forming step). A plasma treatment apparatus can be used to form the deformation. Fig. 10 is a schematic sectional view illustrating a configuration example of a plasma treatment apparatus 60 that can be used to form the deformation.

[0055] The plasma treatment apparatus 60 includes a vacuum chamber 64 defining a treatment space 62. The vacuum chamber 64 is formed in a rectangular parallelepiped shape including a bottom wall 64a, a top wall 64b, a first side wall 64c, a second side wall 64d, a third side wall 64e, and a fourth side wall (not shown). The second side wall 64b is provided with an opening 66 for removing and inserting the wafer 11. A door 68 is provided on the exterior of the opening 66 for opening and closing the opening 66. The door 68 is moved up and down by an opening / closing mechanism 70. The opening / closing mechanism 70 includes an air cylinder 72 and a piston rod 74. The air cylinder 72 is fixed to the bottom wall 64a of the vacuum chamber 64 by a bracket 76, and a tip of the piston rod 74 is connected to a lower portion of the door 68.When the door 68 is opened by the opening / closing mechanism 70, the wafer 11 can be inserted into the processing space 62 of the vacuum chamber 64 through the opening 66, or the wafer 11 can be removed from the processing space 62 of the vacuum chamber 64. An ejection port 78 is formed in the bottom wall 64a of the vacuum chamber 64. The ejection port 78 is connected to an ejection mechanism 80 such as a vacuum pump.

[0056] A lower electrode 82 and an upper electrode 84 are arranged opposite each other in the treatment space 62 of the vacuum chamber 64. The lower electrode 82 is formed of a conductive material and includes a disc-shaped holding portion 86 and a cylindrical support portion 88 protruding from the center of the lower surface of the holding portion 86. The support portion 88 is inserted into an opening 90 formed in the bottom wall 64a of the vacuum chamber 64. An annular insulating member 92 is formed in the opening 90 between the bottom wall 64a and the support portion 88, thereby isolating the vacuum chamber 64 and the lower electrode from each other. The lower electrode 82 is connected to a high-frequency power source 94 outside the vacuum chamber 64.

[0057] An upper surface of the holding portion 86 is formed with a recess, and a table 96 on which the wafer 11 is to be placed is provided in the recess. The table 96 is formed therein with a suction passage (not shown) connected to a suction source 100 through a flow path 98 formed in the lower electrode 82. In addition, the holding portion 86 is formed therein with a cooling flow path 102. One end of the cooling flow path 102 is connected to a coolant circulation mechanism 106 through a coolant introduction passage 104 formed in the support portion 88, whereas the other end of the cooling flow path 102 is connected to a coolant circulation mechanism 106 through a coolant outlet passage 108 formed in the support portion 88.When the coolant circulation mechanism 106 is actuated, a coolant flows through the coolant introduction passage 104, the coolant flow path 102, and the coolant outlet passage 108 in this order to cool the lower electrode 82.

[0058] The upper electrode 84 is formed of a conductive material and includes a disk-shaped gas ejection portion 110 and a cylindrical support portion 112 protruding from the center of an upper surface of the gas ejection portion 110. The support portion 112 is inserted into an opening 114 formed in the upper wall 64b of the vacuum chamber 64. An annular insulating member 116 is disposed in the opening 114 between the upper wall 64b and the support portion 112, thereby isolating the vacuum chamber 64 and the upper electrode 84 from each other. The upper electrode 84 is connected to a high-frequency power source 118 outside the vacuum chamber 64. In addition, a support arm 122 connected to a lifting mechanism 120 is attached to an upper end portion of the support portion 112, and the upper electrode 84 is moved up and down by the lifting mechanism 120 and the support arm 122.A lower surface of the gas ejection section 110 is provided with a plurality of outlet openings 124. The outlet openings 124 are connected to a first gas supply source 130 and a second gas supply source 132 through a flow path 126 formed in the gas ejection section 110 and a flow path 128 formed in the support section 112. The first gas supply source 130, the second gas supply source 132, the flow paths 126 and 128, and the ejection openings 124 form a gas introduction section for introducing the gas into the vacuum chamber 64.

[0059] The opening / closing mechanism 70, the ejection mechanism 80, the high-frequency power source 94, the suction source 100, the coolant circulation mechanism 106, the high-frequency power source 118, the lifting mechanism 120, the first gas supply source 130, the second gas supply source 132, and the like are connected to a controller 134. Information about the pressure in the treatment chamber 62 is input from the ejection mechanism 80 to the controller 134. In addition, information about the temperature of the coolant or information about the temperature of the lower electrode 82 is input from the coolant circulation mechanism 106 to the controller 134. Further, information about the flow rates of the gas from the first gas supply source 130 and the second gas supply source 132 is input to the controller 134.Based on this information and other information input by an operator or the like, the controller 134 outputs control signals for controlling the aforementioned components of the plasma treatment apparatus 60.

[0060] In the deformed layer forming step, first, the door 68 of the plasma processing apparatus 60 is lowered by the opening / closing mechanism 70. Next, the wafer is inserted into the processing space 62 of the vacuum chamber 64 through the opening 66 and placed on the table 96 of the lower electrode 82 with its rear surface 11b exposed upward. Note that at the time of inserting the wafer 11, it is preferable to preliminarily raise the upper electrode 84 by the lifting mechanism 120 to preliminarily increase the space between the lower electrode 82 and the upper electrode 84. Thereafter, negative pressure of the suction source 100 is applied, thereby fixing the wafer 11 to the table 96. Furthermore, the door 68 is raised by the opening / closing mechanism 70 to hermetically close the processing space 62.Furthermore, the height position of the upper electrode 84 is adjusted by the lifting mechanism 120 in such a way that the upper electrode 84 and the lower electrode 82 are brought into a predetermined positional relationship suitable for plasma processing. In addition, the ejection mechanism 80 is actuated to form a vacuum (low pressure) in the processing space 62. Note that in the case where it is difficult to hold the wafer by the negative pressure of the suction source 100 after the pressure in the processing space 62 is lowered, the wafer is held on the table 96 by an electric force (typically an electrostatic force or the like). For example, by embedding electrodes in the table 96 and supplying the electrodes with electric power, an electric force can be caused to act between the chuck table 96 and the wafer 11.

[0061] In this state, while a gas for plasma processing is supplied at a predetermined flow rate, a predetermined high-frequency power is supplied to the lower electrode 82 and the upper electrode 84. In the deformed layer forming step according to the present embodiment, while maintaining a predetermined pressure (for example, a pressure of 5 to 50 Pa) in the processing space 62, a predetermined high-frequency electric power (for example, 1000-3000 W) is supplied to the lower electrode 82 and the upper electrode 84, while an inert gas such as a rare gas is supplied at a predetermined flow rate from the first gas supply source 130.As a result, a plasma is generated between the lower electrode 82 and the upper electrode 84. Ions generated by the inert gas formed as plasma are attracted to the lower electrode 82 side and deposited on the rear surface 11b of the wafer 11. Then, the rear surface 11b of the wafer 11 is sputtered, thereby forming small asperities (small protrusions and recesses or cracks (deformations)) on the rear surface 11b. The region provided with the deformation (deformation layer) serves as a getter layer for containing metallic elements contained in the wafer 11.

[0062] Fig. Fig. 11 is an enlarged sectional view of the wafer 11 in the state in which it has been formed with a deformed layer 33. The wafer 11, which has been divided into the plurality of device chips 31 by the grinding step, is arranged on the table 96 through the protective member 29, and the rear surface side 11b of the wafer 11 faces the upper electrode 84 (see Fig. 10). When plasma processing using the inert gas is applied to the wafer 11, the deformed layer 33 is formed on the back surface 11b of the wafer 11. In the deformed layer forming step, plasma processing is applied to the wafer 11, which involves removing the patterns 17 (see Fig. 1, etc.) by applying the laser beam and then cutting by a cutting blade (laser processing step, the cut groove forming step), and in which occurrence of burrs was suppressed. Therefore, occurrence of electrical discharge at the time of plasma processing is suppressed, and damage to the components 15 is prevented. By forming the deformed layer 33, a gettering effect can be obtained, whereby metallic elements contained in the interior of the wafer 11 are accumulated on the back surface 11b of the wafer 11. Note that the deformed layer 33 formed by the plasma processing has an extremely small thickness compared to the damaged layer formed by the grinding step (for example, 1 / 10 of that or less).Therefore, the formation of a deformed layer 33 does not lead to a significant reduction in the strength of the wafer 11.

[0063] As described above, in the wafer processing method according to the present embodiment, the wafer 11 is cut along the streets 13 after the patterns 17 formed on the streets 13 of the wafer 11 are removed by applying the laser beam. As a result, cutting of the patterns 17 by the cutting blade 28 is prevented, and occurrence of burrs at the time of cutting the wafer 11 is suppressed. Therefore, at the time of applying plasma processing to the wafer 11 to form the deformed layer 33 on the back surface 11b of the wafer 11, damage to the components due to electric discharge at the burrs can be prevented.

[0064] Note that while the damaged layer removing step for removing the damaged layer by polishing with the polishing pad 56 has been described in the present embodiment (see Fig. 9), the damaged layer can also be plasma etched using the plasma treatment device 60 shown in Fig. 10 in the damaged layer removal step. In the case of removing the damaged layer by plasma etching, the wafer 11, which has been divided into a plurality of component chips 31, is fixed to the table 96 of the plasma treatment apparatus 60 through the protective member 29 as shown in Fig.11. In this state, high-frequency electric power is supplied to the lower electrode 82 and the upper electrode 84 while supplying a gas for etching. Specifically, the interior of the treatment space 62 is maintained at a predetermined pressure (for example, 50 to 300 Pa), and a predetermined high-frequency electric power (2000 to 3000 W) is supplied to the lower electrode 82 and the upper electrode 84, while supplying a halogen-containing gas such as SF6 from the second gas supply source 132 at a predetermined flow rate. Note that the distance between the lower electrode 82 and the upper electrode 84 is set larger than the distance at the time of performing the strained layer forming step. This allows the voltage applied to the wafer 11 to be lowered compared to the time of performing the strained layer forming step.

[0065] Through the above step, a plasma is generated between the lower electrode 82 and the upper electrode 84, and an active substance generated by the plasma acts on the back surface 11b of the wafer 11, thereby etching the back surface 11b of the wafer 11. In this way, the damaged layer formed on the back surface 11b of the wafer 11 is removed. This plasma etching is also performed on the wafer 11 that has been subjected to the laser processing step and the cut groove forming step in which the occurrence of burrs is suppressed. Therefore, even in the damaged layer removal step, electric discharge can be suppressed, and damage to the devices 15 is prevented.

[0066] Other than the above, structures, methods or the like according to the present embodiment may be modified as necessary in carrying out the present invention within the scope of the object of the invention.

Claims

[1] A method for processing a wafer (11) in which components (15) are each formed in regions of a front surface side (11a) divided by a plurality of streets (13), and in which patterns (17) including a metal layer are formed on the streets (13), the method comprising: a laser processing step of applying a laser beam of such a wavelength that it is absorbed in the wafer (11) along the streets (13) provided with the patterns (17) to form laser-machined grooves (11c) while removing the patterns (17); a cut groove forming step (11d) of forming cut grooves (11d) having a depth deeper than a finished thickness of the wafer (11) in the laser-machined grooves (11c) by a cutting blade (28) thinner than a width of the laser-machined grooves (11c); a protective member (29) attaching step for attaching a protective member (29) to the front surface side (11a) of the wafer (11) formed with the cut grooves (11d); a grinding step of holding the wafer (11) by a chuck table (12) through a protective member (29), grinding a rear surface side (11b) of the wafer (11) to thin the wafer (11) to a finished thickness and expose the cut grooves (11d) on the rear surface (11b) of the wafer (11), thereby dividing the wafer (11) into a plurality of device chips (31); a damaged layer removing step for removing a damaged layer formed on the back surface side (11b) of the wafer (11) by grinding the wafer (11); and a deformed layer (33) forming step of forming a deformed layer (33) on the back surface side (11b) of the wafer (11) from which the damaged layer (33) has been removed by plasma processing using an inert gas. [2] A method of processing a wafer (11) according to claim 1, wherein in said damaged layer removing step, said damaged layer is removed by polishing with a polishing pad (56). [3] A method of processing a wafer (11) according to claim 1, wherein in said damaged layer removing step, said damaged layer is removed by plasma etching using a halogen-containing gas. [4] A method for processing a wafer (11) according to any one of the preceding claims, further comprising: a protective film forming step (23) for forming a water-soluble protective film (23) on the front surface side (11a) of the wafer (11) before the laser processing step; and a protective film removing step (23) for removing the protective film (23) from the front surface side (11a) of the wafer (11) after the laser processing step.

Images