JP7353712B2 - Wafer processing method - Google Patents

Description

The present invention relates to a wafer processing method for processing a wafer using plasma-generated gas.

As a method for manufacturing semiconductor device chips from a semiconductor wafer, for example, SDBG (Stealth Dicing Before Grinding), in which a wafer is divided by laser processing and grinding processing, is known. In SDBG, a laser beam having a wavelength that is absorbed by the wafer is positioned inside the wafer, and the laser beam is irradiated along the planned dividing line (street) of the wafer.

As a result, multiphoton absorption occurs in the vicinity of the focal point, forming a modified layer with lower strength than in other regions away from the focal point. When a wafer with a modified layer formed thereon is ground, stress is applied to the wafer during grinding, so cracks develop starting from the modified layer. When the growing crack reaches the front and back sides of the wafer, the wafer is divided into multiple chips.

DBG (Dicing Before Grinding) is also known as another method for manufacturing semiconductor device chips, in which a wafer is divided by cutting and grinding. In DBG, cutting grooves are formed on the front side of the wafer along the planned dividing line, and then the back side of the wafer is ground. Once the wafer is thinned down to the bottom of the front side kerf, the wafer is divided into chips.

On the back side of a wafer that has been subjected to mechanical processing such as grinding, a strained layer whose crystal structure is distorted due to the machining is formed. The strained layer causes a decrease in the bending strength of the chip. Therefore, a method was devised to remove the strained layer on the back side of the wafer by direct plasma etching, in which plasma gas generated in the chamber of a plasma processing equipment is supplied to the back side of the wafer placed in the chamber. (For example, see Patent Document 1).

In addition, remote plasma etching is also known, in which plasma generated in a discharge space provided outside the chamber is carried into the chamber by airflow, and gas in a plasma state is supplied to the wafer placed inside the chamber. ing. Remote plasma etching can reduce damage to the wafer compared to direct plasma etching, so remote plasma etching is more suitable for removing strained layers and improving the bending strength of chips, for example.

By the way, in the process of transporting a wafer after machining into a chamber of a plasma processing apparatus, a natural oxide film having in-plane thickness variations may be formed on the back side of the wafer. If remote plasma etching is performed on the back side where a natural oxide film with uneven thickness is formed, minute pits, roughness, etc. will be formed on the back side, which will change the appearance of the back side of the chip. It may become cloudy due to diffused reflection. Furthermore, micro pits, roughness, etc. may reduce the bending strength of the chip.

Differences in etching rate are considered to be the cause of the formation of minute pits and the like. For example, when a wafer is mainly made of silicon, the etching rate of a silicon oxide film is about 30 to 100 times lower than the etching rate of a silicon wafer in remote plasma etching using a fluorine-based gas. Therefore, pits are more likely to be formed in areas where the natural oxide film is thin.

A natural oxide film with in-plane thickness variations is particularly likely to be formed when a wafer is transported with a vacuum-type transport pad (ie, suction pad) in contact with the wafer. Additionally, in this case, fine foreign matter tends to adhere to the wafer. Therefore, when the transfer pad portion is brought into contact with the wafer, minute pits, roughness, etc. are particularly likely to be formed, and problems such as a cloudy appearance and a decrease in transverse strength become significant.

Japanese Patent Application Publication No. 2003-173987

However, after SDBG or DBG has been applied to the wafer, the wafer is divided by cracks or dividing grooves that extend from the front surface to the back surface of the wafer. Therefore, in order to transfer a wafer subjected to SDBG or DBG into a chamber of a plasma processing apparatus, it is necessary to bring the entire surface of the wafer into contact with the transfer pad section and to adsorb it.

The present invention has been made in view of the above problems, and it is possible to suppress the cloudy appearance of chips and the decrease in bending strength even when the transfer pad portion is brought into contact with the entire surface of the wafer. purpose.

According to one aspect of the present invention, there is provided a method for processing a wafer in which a device is formed in each of a plurality of regions partitioned by a plurality of dividing lines set on the front side of the wafer, the method comprising: transmitting light through the wafer; Forming a modified layer by irradiating the wafer with a laser beam having a wavelength such that the laser beam is focused inside the wafer to form a modified layer inside the wafer along the planned dividing line. a back side processing step of grinding the back side of the wafer after the modified layer forming step ; a cleaning step of cleaning the back side of the wafer with a cleaning unit after the back side processing step; After the cleaning step, the wafer is transferred from the cleaning unit to the plasma processing chamber while the wafer is suctioned and held by the transfer pad so that the holding surface of the transfer pad is in contact with the back side of the wafer. a step of transporting the wafer; and after the transport step, supplying a first gas turned into plasma in the plasma processing chamber to the back side of the wafer to remove an oxide film formed by oxidizing the back side of the wafer; After the oxide film removal step, a second gas that has been turned into plasma outside the plasma processing chamber is supplied to the back side of the wafer from which the oxide film has been removed, and the back side of the wafer is removed. a back side etching step for etching the back side, and in the back side processing step, stress is applied to the wafer to cause cracks that have grown from the modified layer to reach the back side of the wafer, and the back side is etched. A wafer processing method is provided in which the etching step removes processing distortion formed on the back side of the wafer, as well as at least a portion of the cracked region and the modified layer.

Preferably, in the oxide film removal step, the first gas turned into plasma in the plasma processing chamber and the second gas turned into plasma outside the plasma processing chamber are transferred to the back side of the wafer. supply to.

Preferably, in the oxide film removal step, foreign particles having a size of 1 μm or less are also removed together with the oxide film.

In the wafer processing method according to one embodiment of the present invention, the back side of the wafer is sucked and held by a transfer pad, the wafer is transferred to a plasma processing chamber, and then a first gas that has been turned into plasma in the plasma processing chamber is Supply to the back side of the wafer. As a result, the oxide film formed on the back side of the wafer is removed (oxide film removal step). Next, a second gas turned into plasma outside the plasma processing chamber is supplied to the back side from which the oxide film has been removed to etch the back side (back side etching step).

In this manner, in the oxide film removal step, direct plasma etching is performed on the back surface side by turning the first gas into plasma in the plasma processing chamber. In direct plasma etching, the energy of active species is relatively high and the amount of incident active species is relatively large, so that natural oxide films and fine foreign matter can be removed.

Further, in the back side etching step after the oxide film removal step, remote plasma etching is performed on the back side by turning the second gas into plasma outside the plasma processing chamber. In remote plasma etching, the energy of active species is lower than that in direct plasma, and the amount of incident active species is small, so it is possible to remove strained layers and the like on the back side of the wafer while reducing damage to the wafer.

Therefore, even when the wafer is transported with the transport pad portion in contact with the entire back surface of the wafer, the natural oxide film and the like can be removed in the oxide film removal step (direct plasma etching). Therefore, cloudy appearance of the chip after the back side etching step (remote plasma etching) can be reduced. Furthermore, in the backside etching step, the strained layer and the like on the backside are removed, so it is possible to suppress a decrease in the bending strength of the chip.

FIG. 1(A) is a perspective view of a wafer held by a chuck table, and FIG. 1(B) is a partially sectional side view of the laser processing apparatus. FIG. 2 is a partially sectional side view of a grinding device and the like. FIG. 2 is a partially cross-sectional side view of the cleaning device and the like. FIG. 3 is a partially cross-sectional side view of the conveyance device. FIG. 2 is a partially cross-sectional side view of a plasma processing apparatus and the like. FIG. 2 is a partially cross-sectional side view of a plasma processing apparatus, etc., showing an oxide film removal step. FIG. 2 is a partially cross-sectional side view of the plasma processing apparatus, etc., showing a back side etching step. FIG. 8(A) is a diagram showing the modified layer forming step, FIG. 8(B) is a partially cross-sectional side view of the wafer etc. after the back side processing step, and FIG. 8(C) is a diagram showing the wafer etc. being transported to the chamber. FIG. 2 is a partially cross-sectional side view of a wafer, etc. FIG. 9(A) is a diagram showing an oxide film removal step, and FIG. 9(B) is a diagram showing an example of a back side etching step. FIG. 2 is a flow diagram of a wafer processing method. FIG. 7 is a diagram showing another example of the back side etching step.

Embodiments according to one aspect of the present invention will be described with reference to the accompanying drawings. In the first embodiment, a method for processing the wafer 11 will be described in which the wafer 11 is divided by SDBG (Stealth Dicing Before Grinding) and then subjected to plasma treatment.

First, a laser processing apparatus 2 for forming a modified layer 21 inside a wafer 11 will be described using FIGS. 1(A) and 1(B). The laser processing apparatus 2 includes a chuck table 4 for sucking and holding the wafer 11. FIG. 1A is a perspective view of the wafer 11 held by the chuck table 4. FIG.

The wafer 11 is formed into a disk shape using a material such as silicon, and each has a substantially circular front surface 11a and a substantially circular back surface 11b. The front surface 11a side of the wafer 11 is divided into a plurality of regions by a plurality of dividing lines (streets) 13 arranged in a grid pattern so as to intersect with each other.

A device 15 made of an IC (Integrated Circuit), an LSI (Large Scale Integration), or the like is formed on the surface 11a side of each area partitioned by a plurality of planned division lines 13. When the wafer 11 is divided along the planned dividing line 13, a plurality of device chips each having a device 15 are obtained.

Note that there are no restrictions on the material, shape, structure, size, etc. of the wafer 11. For example, the wafer 11 may be made of a material such as glass. Furthermore, there are no restrictions on the type, quantity, shape, structure, size, arrangement, etc. of the device 15.

Before processing the wafer 11 with the laser processing apparatus 2, a protective member 17 made of resin and having approximately the same diameter as the wafer 11 is pasted on the front surface 11a side of the wafer 11 in order to protect the devices 15. wear. Thereby, a wafer unit 19 is formed in which the protective member 17 is attached to the wafer 11.

The protective member 17 has, for example, a laminated structure of a base material layer and an adhesive layer (glue layer). The base layer is made of polyolefin (PO), for example. An adhesive layer is formed on a part or all of one surface of the base material layer. The adhesive layer is, for example, an ultraviolet curable resin, and is formed of a rubber-based, acrylic-based, silicone-based, or the like resin.

However, the protective member 17 is not limited to a laminated structure of a base material layer and an adhesive layer. For example, the protective member 17 may have only a base material layer. In this case, the protective member 17 is attached to the wafer 11 by thermocompression bonding the base material layer to the front surface 11a side of the wafer 11.

When processing the wafer 11, the protection member 17 side of the wafer unit 19 is held by suction with the chuck table 4. The chuck table 4 includes, for example, a disc-shaped porous plate (not shown).

The lower surface side of the porous plate is connected to a suction source (not shown) such as an ejector via a flow path (not shown) formed inside the chuck table 4. When the suction source is operated, negative pressure is generated on the upper surface side of the porous plate. Therefore, the upper surface side of the porous plate functions as a holding surface 4a that attracts and holds the wafer unit 19.

A rotational drive source (not shown) such as a motor is connected to the lower part of the chuck table 4. The rotational drive source rotates the chuck table 4 around a rotation axis that is generally parallel to the vertical direction (Z-axis direction, vertical direction).

A horizontal movement mechanism (not shown) is provided below the rotational drive source. The horizontal movement mechanism moves the chuck table 4 and the rotational drive source along the processing feed direction (X-axis direction) and the indexing feed direction (Y-axis direction).

A laser irradiation unit 6 is provided above the chuck table 4. FIG. 1(B) is a partially sectional side view of the laser processing device 2. FIG. The laser irradiation unit 6 has a laser oscillator (not shown) that generates a pulsed laser beam 8.

A condenser (not shown) is connected to the laser oscillator via a predetermined optical system. The laser beam 8 emitted from the laser oscillator is focused at a predetermined position by a condenser. The laser beam 8 has, for example, a wavelength (for example, 1342 nm) that transmits through the wafer 11, and is irradiated from a condenser to the wafer 11 so as to be focused inside the wafer 11 held by the holding surface 4a. be done.

The modified layer forming step (S10) of forming the modified layer 21 inside the wafer 11 using the laser processing apparatus 2 will be described with reference to FIGS. 1(B) and 8(A). FIG. 8(A) is a diagram showing the modified layer forming step (S10). Note that FIG. 10 is a flow diagram of a method for processing the wafer 11.

In the modified layer forming step (S10), first, the wafer unit 19 is placed on the holding surface 4a so that the back surface 11b side is exposed. Then, the suction source is operated to hold the front surface 11a side of the wafer 11 on the holding surface 4a.

Next, with the focal point of the laser beam 8 positioned at one end of the planned dividing line 13 and inside the wafer 11, the horizontal movement mechanism is operated to move the chuck table 4 in the X-axis direction. As a result, the focal point is moved from one end of the planned division line 13 to the other end.

Multiphoton absorption occurs near the focal point, forming a region with lower intensity (i.e., brittle) compared to other regions, so a modified layer with relatively low strength is formed along the path of the focal point movement. 21 is formed. Note that two or more modified layers 21 may be formed along one planned dividing line 13 so that the two or more modified layers 21 are located at different depths.

After forming one or more modified layers 21 along one dividing line 13, the laser irradiation unit 6 is indexed and sent by a predetermined length along the Y-axis direction, and the irradiation position of the laser beam 8 is moved to another position. It is positioned at the planned division line 13.

Next, the focal point of the laser beam 8 is moved along another planned dividing line 13 to form one or more modified layers 21 inside the wafer 11. After forming one or more modified layers 21 inside the wafer 11 along each dividing line 13 substantially parallel to the X-axis direction, the chuck table 4 is rotated 90 degrees.

Similarly, one or more modified layers 21 are formed inside the wafer 11 along all the dividing lines 13 substantially parallel to the X-axis direction. As a result, the modified layer 21 is formed inside the wafer 11 along all the dividing lines 13 arranged in a grid pattern.

After the modified layer forming step (S10), the back surface 11b side of the wafer 11 is processed using a grinding device (back surface side processing step (S20)). FIG. 2 is a partially sectional side view of the grinding device 12 and the like.

The grinding device 12 includes a chuck table 14 for sucking and holding the wafer 11. The structure of the chuck table 14 is the same as the chuck table 4 of the laser processing device 2. A rotational drive source (not shown) such as a motor is connected to the lower part of the chuck table 14 .

A grinding unit 16 is provided above the chuck table 14. The grinding unit 16 has a spindle housing (not shown), and a lifting mechanism (not shown) for moving the grinding unit 16 up and down along the Z-axis direction is connected to this spindle housing.

A spindle 18 is rotatably housed in the spindle housing. A motor for driving the spindle 18 is connected to one end of the spindle 18 . The other end of the spindle 18 protrudes from the spindle housing, and a disc-shaped wheel mount 20 is fixed to this other end.

A grinding wheel 22 having approximately the same diameter as the wheel mount 20 is mounted on the lower surface of the wheel mount 20. The grinding wheel 22 has an annular wheel base 24 made of a metal material such as aluminum or stainless steel.

The wheel base 24 is attached to the spindle 18 by fixing the upper surface side of the wheel base 24 to the wheel mount 20. A plurality of grinding wheels 26 are provided on the lower surface side of the wheel base 24. The plurality of grinding wheels 26 are arranged in a ring shape such that gaps are provided between adjacent grinding wheels 26 in the circumferential direction of the lower surface of the wheel base 24.

Each grinding wheel 26 is formed by mixing abrasive grains such as diamond and cBN (cubic boron nitride) into a binding material such as metal, ceramics, or resin. However, there are no restrictions on the binding material or the abrasive grains, and they can be selected as appropriate depending on the specifications of the grinding wheel 26.

A grinding water supply nozzle 28 is provided above the holding surface 4a. A grinding water supply source (not shown) is connected to the grinding water supply nozzle 28, and grinding water 30 such as pure water is supplied from the grinding water supply source during grinding.

In the back side processing step (S20), first, the back side 11b of the wafer 11 is ground (grinding step). In the grinding step, the front surface 11a side of the wafer 11 is suctioned and held by the holding surface 14a of the chuck table 14 via the protection member 17.

Then, while rotating the chuck table 14 and the grinding wheel 22 in predetermined directions, the lifting mechanism moves the grinding wheel 22 downward at a predetermined speed. Thereby, the grinding wheel 26 is pressed against the back surface 11b side of the wafer 11, and the back surface 11b side of the wafer 11 is ground to a predetermined finishing thickness. Note that during grinding, grinding water 30 is supplied from the outer peripheral side of the grinding wheel 22 to the back surface 11b side of the wafer 11.

When stress is applied to the wafer 11 during grinding, cracks 23 (see FIG. 8(B)) develop starting from the modified layer 21. By making the cracks 23 reach the front surface 11a and the back surface 11b of the wafer 11, the wafer 11 is divided into a plurality of chips.

FIG. 8(B) is a partially sectional side view of the wafer 11 etc. after the back side processing step (S20). Note that in the back side processing step (S20), due to mechanical processing of the back side 11b of the wafer 11, a strained layer 25 with a distorted crystal structure (namely, processing distortion) is formed.

After the back side processing step (S20), the back side 11b of the wafer 11 is cleaned using, for example, a cleaning unit provided near the grinding device 12 (cleaning step (S30)). FIG. 3 is a partially sectional side view of the cleaning unit 32 and the like.

The cleaning unit 32 includes a spinner table 34 for sucking and holding the wafer 11. The spinner table 34 has, for example, a disc-shaped porous plate (not shown). The lower surface side of the porous plate is connected to a suction source (not shown) such as an ejector via a flow path (not shown) formed inside the spinner table 34 .

When the suction source is operated, negative pressure is generated on the upper surface side of the porous plate. Therefore, the upper surface side of the porous plate functions as a holding surface 34a. A rotational drive source (not shown) such as a motor is connected to the lower part of the spinner table 34 .

A cleaning liquid supply nozzle 36 is provided above the spinner table 34. A cleaning liquid 38 such as pure water is supplied to the cleaning liquid supply nozzle 36 from a cleaning liquid supply source (not shown). Further, a drive source (not shown) for moving the cleaning liquid supply nozzle 36 is attached to the cleaning liquid supply nozzle 36 . When the drive source is operated, the cleaning liquid supply nozzle 36 reciprocates in an arc shape on the holding surface 34a.

In the cleaning step (S30), first, the front surface 11a side of the wafer 11 is suctioned and held by the holding surface 34a of the spinner table 34 through the protection member 17. Then, while the spinner table 34 is being rotated, the cleaning liquid 38 is sprayed from the cleaning liquid supply nozzle 36 onto the back surface 11b of the wafer 11.

At this time, the cleaning liquid 38 is ejected from the cleaning liquid supply nozzle 36 while reciprocating the cleaning liquid supply nozzle 36 on the holding surface 34a so that the movement locus of the cleaning liquid supply nozzle 36 becomes an arc shape.

The cleaning liquid 38 cleans the back surface 11b side of the wafer 11 while spreading over the entire back surface 11b side due to centrifugal force. As a result, machining debris etc. remaining on the back surface 11b side after the back surface side processing step (S20) are removed from the back surface 11b side.

After the cleaning step (S30), the wafer 11 is transported from the cleaning unit 32 to the outside of the cleaning unit 32 using the transport device 40 (transport step (S40)). FIG. 4 is a partially sectional side view of the conveyance device 40.

The transport device 40 has an elongated flat arm 42 . A moving unit (not shown) is provided at one end of the arm 42 to move the arm 42 in the horizontal direction (XY plane direction) and the height direction (Z axis direction).

At the other end of the arm 42, a substantially disc-shaped suction pad section (transport pad section) 44 is provided. The suction pad section 44 has a housing in which a substantially disc-shaped recess is formed. A substantially disc-shaped porous plate (not shown) is provided in the recess of the housing so as to be exposed from the housing.

The other surface of the porous plate opposite to the exposed surface is connected to a suction source (not shown) such as an ejector via a channel (not shown) formed inside the casing. . When the suction source is operated, negative pressure is generated on the exposed surface side of the porous plate. Therefore, the exposed surface side of the porous plate functions as a suction surface (holding surface) 44a that attracts and holds the wafer 11.

In the conveying step (S40), first, the suction pad section 44 is moved onto the spinner table 34 which is in a stationary state and the suction on the holding surface 34a is released. Next, the back surface 11b side of the wafer 11 is sucked and held by the suction surface 44a so that the back surface 11b side contacts the suction surface 44a of the suction pad section 44.

Then, with the back surface 11b side and the suction surface 44a in contact with each other, the moving unit located at one end of the arm 42 is operated to transport the wafer 11 to the chamber of the plasma processing apparatus. FIG. 5 is a partially sectional side view of the plasma processing apparatus 46 and the like.

The plasma processing apparatus 46 includes a chamber (plasma processing chamber) 48 having a processing space inside. The chamber 48 is made of, for example, a conductive metal material and is grounded.

A part of the side wall of the chamber 48 is provided with a carry-in/out port 48a through which the suction pad section 44 passes. A sliding door 50 is provided on the outside of the side wall of the chamber 48 to close the loading/unloading entrance 48a.

The door 50 is provided with an opening/closing unit (not shown) consisting of an air cylinder or the like. The opening/closing unit opens and closes the carry-in/out entrance 48a by, for example, sliding the door 50 in the vertical direction.

An exhaust port 48b is provided on the bottom side of the side wall of the chamber 48 located on the opposite side from the carry-in/out port 48a. An exhaust unit 54 is connected to this exhaust port 48b via an exhaust pipe 52. The exhaust unit 54 includes an exhaust valve 54a such as a solenoid valve having one end of a flow path connected to the exhaust pipe 52, and an exhaust pump 54b connected to the other end of the flow path of the exhaust valve 54a.

A table base 56 for holding the wafer 11 and the like is provided in the processing space within the chamber 48 . The table base 56 includes a disk portion 56a made of a conductive material such as metal, and a column portion 56b extending downward from the center of the lower surface of the disk portion 56a.

An electrostatic chuck (not shown) is provided on the upper surface side of the disk portion 56a. The electrostatic chuck includes a disc-shaped main body made of an insulating material and a plurality of electrodes embedded within the main body. A DC power supply section (not shown) capable of generating high voltage is connected to each of the plurality of electrodes.

The electrostatic chuck of this embodiment is a bipolar electrostatic chuck, and includes a first electrode to which a positive potential is supplied and a second electrode to which a negative potential is supplied. For example, a high voltage of +3 kV is applied to the first electrode from the DC power supply, and a high voltage of -3 kV is applied to the second electrode from the DC power supply.

Note that a flow path (not shown) for supplying a coolant is formed in the main body of the electrostatic chuck. A refrigerant controlled at a predetermined temperature is supplied to the flow path from a refrigerant circulation device (not shown). Thereby, the temperature of the electrostatic chuck is maintained at a predetermined temperature (for example, 35° C.).

A bias electrode 58 to which a high frequency voltage is applied is provided on the disk portion 56a in a manner electrically separated from the plurality of electrodes of the electrostatic chuck. A high frequency voltage application unit 60 is connected to the bias electrode 58 .

The high-frequency voltage application unit 60 includes, for example, a high-frequency power supply 60a that can apply a high-frequency voltage of 13.56 MHz to the bias electrode 58, and a blocking capacitor 60b for cutting DC that is provided between the bias electrode 58 and the high-frequency power supply 60a. including.

A plasma diffusion member 62 made of metal is provided inside the chamber 48 above the table base 56 . The plasma diffusion member 62 has a mesh-like region, and this mesh-like region divides the processing space within the chamber 48 into a first region on the upper side and a second region on the lower side. .

However, a plurality of through openings are formed in the mesh area to spatially connect the first area and the second area. The plasma diffusion member 62 has, for example, a function of supplying gas supplied to the first region that has been turned into plasma (that is, radicalized, ionized, etc.) to the second region so as to disperse it.

A gas inlet 48c is provided in the upper wall of the chamber 48. A substantially cylindrical introduction tube 64a is connected to the gas introduction port 48c so as to protrude from the upper wall of the chamber 48. Note that the introduction tube 64a is connected to the processing space of the chamber 48, but is located outside the chamber 48. The introduction tube 64a is made of a material (sapphire, crystal, ceramics, etc.) that allows microwaves to pass therethrough.

A gas supply unit 66 is provided above the introduction tube 64a. The gas supply unit 66 has an inert gas supply source 66a. The inert gas supply source 66a includes an inert gas such as helium (He), argon (Ar), nitrogen ( _N2 ), or the like.

The inert gas supply source 66a is connected to the introduction tube 64a via a first valve 68a, a first flow controller (not shown), and the like. Inert gases are used, for example, as carrier gases for carrying other gases. However, inert gas may be used for the purpose of stabilizing the discharge.

The gas supply unit 66 further includes a fluorine gas supply source 66b. The fluorine-based gas supply source 66b includes fluorine-based gas such as sulfur hexafluoride (SF ₆ ) and tetrafluoromethane (CF ₄ ).

The fluorine-based gas supply source 66b is connected to the introduction tube 64a via a second valve 68b, a second flow controller (not shown), and the like. The fluorine-based gas is, for example, a gas used to etch the wafer 11.

The gas supply unit 66 further includes an oxygen gas supply source 66c having oxygen (O ₂ ) gas. The oxygen gas supply source 66c is connected to the introduction tube 64a via a third valve 68c, a third flow rate controller (not shown), and the like.

Oxygen gas is used, for example, to control the etching rate of the wafer 11. In the process of oxidizing fluorine-containing gas molecules, active species of fluorine atoms (fluorine radicals, fluorine ions, etc.) that contribute to etching are generated, which may increase the etching rate of the wafer 11.

By adjusting the first, second, and third flow rate controllers, a plurality of types of gases are supplied from the gas supply unit 66 to the introduction tube 64a at predetermined flow rates. Note that the number of gas supply sources, the types of gases, and the flow rates of each gas can be changed as appropriate depending on the type of wafer 11 and the like.

An applicator 64b including a housing made of a conductive material such as metal is provided on the side of the introduction tube 64a so as to surround the introduction tube 64a. The housing of the applicator 64b includes, for example, a waveguide for irradiating the introduction tube 64a with microwaves generated by a high frequency generation source such as a magnetron.

Microwaves are electromagnetic waves with a frequency of 300 MHz or more and 300 GHz or less (for example, 2.45 GHz). The gases supplied from the gas supply unit 66 are turned into plasma by irradiating the plurality of types of gases flowing through the introduction tube 64a through the waveguide of the applicator 64b with microwaves.

The plasma-converted gas is supplied from the gas inlet 48c to the first region of the processing space in the chamber 48, and further supplied to the second region of the processing space via the plasma diffusion member 62. When the wafer 11 is placed in the processing space, the wafer 11 is etched with a gas turned into plasma outside the chamber 48 (remote plasma etching).

However, when the gas supplied from the gas supply unit 66 is not irradiated with microwaves (that is, the applicator 64b is in the off state), the gas that has not been turned into plasma flows from the gas inlet 48c into the first region of the processing space. Supplied.

In this case, when the high frequency voltage application unit 60 connected to the bias electrode 58 is operated, the gas supplied from the gas introduction port 48c is turned into plasma within the chamber 48. When the wafer 11 is placed in the processing space, the wafer 11 is etched with gas turned into plasma in the chamber 48 (direct plasma etching).

Next, a procedure for plasma processing the wafer 11 using the plasma processing apparatus 46 will be explained. In this embodiment, first, the wafer 11 is transported to the chamber 48 while the back surface 11b side is held by the suction pad section 44. FIG. 8C is a partially cross-sectional side view of the wafer 11 and the like transferred to the chamber 48.

When the suction pad section 44 of the transport device 40 physically contacts the back surface 11b and transports the wafer 11, the back surface 11b side of the wafer 11 is oxidized. At this time, a natural oxide film 27 with in-plane thickness variations is likely to be formed. Furthermore, there is a possibility that fine foreign matter (not shown) made of oxide, metal, etc. and having a size of 1 μm or less may adhere to the back surface 11b.

Note that the foreign matter having a size of 1 μm or less means, for example, that the particle diameter (i.e., length) of the foreign matter (i.e., particle) is 1 μm or less. There are known methods for expressing the particle size, such as geometric diameter and equivalent diameter. Geometric diameters include Feret diameter, maximum direction diameter (i.e. Krummbein diameter), Martin diameter, sieve diameter, etc. Equivalent diameters include projected area circle equivalent diameter (i.e. Heywood diameter), There are equivalent diameters of equal surface area spheres, equivalent diameters of equal volume spheres, Stokes diameters, light scattering diameters, etc.

In this embodiment, the natural oxide film 27, foreign matter, etc. are removed by direct plasma etching in which the energy of active species is relatively high and the amount of incident active species is relatively large (oxide film removal step (S50)). FIG. 6 is a partially cross-sectional side view of the plasma processing apparatus 46 and the like showing the oxide film removal step (S50), and FIG. 9A is a diagram showing the oxide film removal step (S50).

In the oxide film removal step (S50), first, the front surface 11a side of the wafer 11 is placed on the disk portion 56a of the table base 56 using the transfer device 40. Thereafter, the transport device 40 is evacuated from the chamber 48 and the door 50 is closed.

Next, the DC power supply unit connected to the electrostatic chuck is turned on. As a result, a voltage is applied to each electrode of the electrostatic chuck, and an electrostatic force is generated between each electrode and the wafer 11, so that the surface 11a side of the wafer 11 is held by the electrostatic chuck through the protective member 17. Ru.

Thereafter, the exhaust unit 54 is operated to exhaust the processing space within the chamber 48 . As a result, the processing space within the chamber 48 is reduced to a first predetermined pressure (for example, 50 Pa). Note that during the oxide film removal step (S50), the pressure inside the chamber 48 is maintained at a first predetermined pressure.

Next, for example, He gas is supplied at 200 sccm from the inert gas supply source 66a, and SF ₆ gas is supplied at 400 sccm from the fluorine-based gas supply source 66b, respectively, to the processing space of the chamber 48. In addition, the high frequency power supply 60a is operated to apply high frequency power of 13.56 MHz and 300 W to the bias electrode 58.

An electric field is applied to the gas existing in the processing space, and a first plasma 70 (ie, the first gas turned into plasma in the chamber 48), which is a capacitively coupled plasma (CCP), is generated.

The first plasma 70 is supplied to the entire back surface 11b side of the wafer 11 for, for example, 20 seconds. As a result, the natural oxide film 27 formed on the back surface 11b side and fine foreign matter adhering to the back surface 11b side are removed. Note that a portion of the strained layer 25 may be removed by the first plasma 70.

In the oxide film removal step (S50), the natural oxide film 27 and fine foreign matter can be removed by direct plasma etching in which the energy of active species is relatively high and the amount of incident active species is relatively large.

In particular, in direct plasma etching, ions in the plasma tend to collide with the back surface 11b of the wafer 11 due to the electric field formed in the processing space. Can be removed efficiently.

After the oxide film removal step (S50), while the wafer 11 is held by the electrostatic chuck, a second gas that has been turned into plasma outside the chamber 48 is supplied to the back surface 11b side of the wafer 11, and the second gas is supplied to the back surface 11b side of the wafer 11. (back side etching step (S60)).

FIG. 7 is a partially sectional side view of the plasma processing apparatus 46 and the like showing the back side etching step (S60). FIG. 9B is a diagram showing an example of the back side etching step (S60).

Specifically, after the oxide film removal step (S50), the operation of the high frequency power source 60a is stopped. Then, He gas is supplied at 3000 sccm from the inert gas supply source 66a, SF ₆ gas at 165 sccm from the fluorine gas supply source 66b, and O ₂ gas at 55 sccm from the oxygen gas supply source 66c, respectively, to the processing space of the chamber 48. do.

In addition, the applicator 64b is turned on to generate microwaves with high frequency power of 2.45 GHz and 2000 W. By irradiating the gas supplied from the gas supply unit 66 to the introduction tube 64a with this microwave, the gas is turned into plasma in the introduction tube 64a.

The second gas (that is, the second plasma 72) turned into plasma outside the chamber 48 is supplied to the first region of the processing space from the gas introduction port 48c along with the airflow, and is further supplied to the first region of the processing space through the plasma diffusion member 62. and is then supplied to the second region of the processing space.

The second plasma 72 is supplied to the back surface 11b side of the wafer 11 held by the electrostatic chuck for, for example, 5 minutes. Note that during the back side etching step (S60), the exhaust unit 54 is operated to maintain the processing space in the chamber 48 at a second predetermined pressure (for example, 150 Pa).

In the back side etching step (S60), the strained layer 25 (ie, processing strain) is removed by etching using the second plasma 72 (ie, remote plasma etching). Furthermore, in the back side etching step (S60), the groove 11c is formed by the second plasma 72 etching the crack 23, so that the region where the crack 23 is formed and at least a part of the modified layer 21 (for example, , all) are removed.

In the backside etching step (S60), remote plasma etching is performed on the backside 11b of the wafer 11, in which the energy of active species is lower and the amount of incident active species is smaller than that of direct plasma. Thereby, the strained layer 25 and the like on the back surface 11b side of the wafer 11 can be removed while reducing damage to the wafer 11.

In this way, even when the wafer 11 is transported with the suction pad section 44 in contact with the entire back surface 11b, the natural oxide film 27 and the like can be removed in the oxide film removal step (S50) (direct plasma etching). Therefore, the cloudy appearance of the chip after the back side etching step (S60) (remote plasma etching) can be reduced.

Furthermore, by performing remote plasma etching, in which oxides are difficult to remove, after direct plasma etching, in which oxides are easily removed, the strained layer 25, the region where the cracks 23 are formed, the modified layer 21, etc. can be efficiently removed. can. Therefore, a decrease in the bending strength of the chip can be suppressed.

Note that in the backside etching step (S60) of this embodiment, He gas is used at a flow rate ten times or more as compared to the oxide film removal step (S50). The discharge starting voltage of He gas is lower than that of SF ₆ gas and the like. Therefore, by increasing the flow rate of He gas, it is possible to reduce the occurrence of electric discharge in localized spaces where the electric field is strong. In other words, the spatial uneven distribution of the second plasma 72 can be reduced.

Next, another example of the back side etching step (S60) will be explained. In this other example, the second plasma 72 is supplied to the back surface 11b side of the wafer 11 for a period longer than 5 minutes. As a result, the groove 11c is formed, and a chamfered region where the end of the groove 11c on the back surface 11b side is chamfered is also formed.

FIG. 11 is a diagram showing another example of the back side etching step (S60). The chamfered region 11d is a slope or curved surface extending from the back surface 11b to the front surface 11a. By forming the chamfered region 11d on the back surface 11b side, the bending strength of the chip is improved compared to the case where the chamfered region 11d is not formed.

Next, a second embodiment will be described. In the oxide film removal step (S50) according to the second embodiment, a first plasma 70 and a second plasma 72 are supplied to the back surface 11b side of the wafer 11. Although this point differs from the first embodiment, other points are the same as the first embodiment.

In the oxide film removal step (S50) of the second embodiment, the high frequency power source 60a is operated while supplying He gas and SF ₆ gas to the processing space of the chamber 48 from an introduction route (not shown) different from the introduction tube 64a. By doing so, the first plasma 70 is generated in the processing space.

In the oxide film removal step (S50) of the second embodiment, while supplying He gas, SF ₆ gas, and O ₂ gas from the introduction tube 64a to the processing space of the chamber 48, the applicator 64b is turned on and introduced. A second plasma 72 is generated in the cylinder 64a. In this way, while the first plasma 70 is generated in the processing space of the chamber 48, the second plasma 72 is supplied to the processing space.

In the oxide film removal step (S50) according to the second embodiment, the natural oxide film 27 and fine foreign matter can also be removed. Therefore, even when the wafer 11 is transported with the suction surface 44a of the suction pad section 44 in contact with the entire back surface 11b, the formation of pits, roughness, etc. on the back surface 11b side can be suppressed, resulting in a cloudy appearance of the chip. can be reduced.

Furthermore, by performing the backside etching step (S60) after the oxide film removal step (S50), the strained layer 25, the region where the cracks 23 are formed, the modified layer 21, etc. can be appropriately removed. Therefore, a decrease in the bending strength of the chip can be suppressed.

In addition, the structure, method, etc. according to the above embodiments can be modified and implemented as appropriate without departing from the scope of the objective of the present invention. For example, the SDBG embodiments described above may be applied to a DBG process. In this case, instead of the modified layer forming step (S10), cutting grooves deeper than the finished thickness of the chip are formed on the surface 11a side (cutting groove forming step).

In the first modification, after the back side processing step (S20) in the SDBG and DBG and before the cleaning step (S30), the back side 11b of the wafer 11 is polished using a polishing device (not shown). A polishing step may be further added.

Furthermore, in the oxide film removal step (S50) in the second modification, CF ₄ gas may be used instead of SF ₆ gas. CF ₄ gas has a higher etching rate for silicon oxide films and a lower etching rate for silicon than SF ₆ gas. Therefore, the natural oxide film 27 can be removed more efficiently.

11 Wafer 11a Front surface 11b Back surface 11c Groove 11d Chamfering area 13 Planned dividing line 15 Device 17 Protective member 19 Wafer unit 21 Modified layer 23 Crack 25 Strain layer 27 Natural oxide film 2 Laser processing device 4 Chuck table 4a Holding surface 6 Laser irradiation unit 8 Laser beam 12 Grinding device 14 Chuck table 14a Holding surface 16 Grinding unit 18 Spindle 20 Wheel mount 22 Grinding wheel 24 Wheel base 26 Grinding wheel 28 Grinding water supply nozzle 30 Grinding water 32 Cleaning unit 34 Spinner table 34a Holding surface 36 Cleaning liquid supply Nozzle 38 Cleaning liquid 40 Transport device 42 Arm 44 Suction pad section (transport pad section)
44a Adsorption surface (holding surface)
46 Plasma treatment equipment 48 Chamber (plasma treatment chamber)
48a Carrying in/out port 48b Exhaust port 48c Gas inlet 50 Door 52 Exhaust stack 54 Exhaust unit 54a Exhaust valve 54b Exhaust pump 56 Table base 56a Disk section 56b Pillar section 58 Bias electrode 60 High frequency voltage application unit 60a High frequency power source 60b Blocking capacitor 62 Plasma diffusion member 64a Introduction cylinder 64b Applicator 66 Gas supply unit 66a Inert gas supply source 66b Fluorine gas supply source 66c Oxygen gas supply source 68a First valve 68b Second valve 68c Third valve 70 First plasma 72 Second plasma

Claims (3)

A method for processing a wafer in which devices are formed in each of a plurality of regions defined by a plurality of dividing lines set on the front side of the wafer, the method comprising:
Irradiate the wafer with a laser beam having a wavelength that transmits the wafer so that the laser beam is focused inside the wafer, and form a modified layer inside the wafer along the planned dividing line. , a modified layer forming step;
After the modified layer forming step, a back side processing step of grinding the back side of the wafer;
After the back side processing step, a cleaning step of cleaning the back side of the wafer with a cleaning unit;
After the cleaning step, the wafer is transferred from the cleaning unit to the plasma processing chamber while the wafer is suctioned and held by the transfer pad so that the holding surface of the transfer pad is in contact with the back side of the wafer. a conveyance step for conveying the
After the transfer step, an oxide film removing step of supplying a first gas that has been turned into plasma in the plasma processing chamber to the back side of the wafer to remove an oxide film formed by oxidizing the back side of the wafer. ,
After the oxide film removal step, a second gas that has been turned into plasma outside the plasma processing chamber is supplied to the back side of the wafer from which the oxide film has been removed, thereby etching the back side of the wafer. comprising a step ;
In the back side processing step, stress is applied to the wafer to cause cracks that have grown from the modified layer to reach the back side of the wafer,
A method for processing a wafer, characterized in that in the back side etching step, at least a part of the region where the crack is formed and the modified layer are removed together with the processing strain formed on the back side of the wafer. In the oxide film removal step, the first gas turned into plasma in the plasma processing chamber and the second gas turned into plasma outside the plasma processing chamber are supplied to the back side of the wafer. The method of processing a wafer according to claim 1 , characterized in that: 3. The wafer processing method according to claim 1, wherein in the oxide film removing step, foreign particles having a size of 1 μm or less are also removed together with the oxide film.

Images