JP4640945B2 - Laser processing method - Google Patents

Description

  The present invention relates to a laser processing method for irradiating a workpiece with a pulsed laser beam having a wavelength having transparency, thereby forming a deteriorated layer inside the workpiece.

  In the semiconductor device manufacturing process, divided division lines called streets arranged in a lattice pattern on the surface of a wafer including an appropriate substrate such as a silicon substrate, a sapphire substrate, a silicon carbide substrate, a lithium tantalate substrate, a glass substrate, or a quartz substrate. A plurality of regions are partitioned by this, and devices such as ICs and LSIs are formed in these partitioned regions. Then, each semiconductor device is manufactured by dividing the region in which the device is formed by cutting the wafer along the planned dividing line.

In recent years, as a method of dividing a plate-like workpiece such as a semiconductor wafer, a pulse laser beam having a wavelength that is transparent to the workpiece is used, and a pulse is focused on the inside of the region to be divided. A laser processing method for irradiating a laser beam has also been attempted. The dividing method using this laser processing method irradiates a pulsed laser beam in an infrared region having transparency to the work piece by aligning a condensing point from one side of the work piece to the inside. The workpiece is divided by continuously forming a deteriorated layer along the planned division line inside the workpiece and applying external force along the planned division line whose strength has been reduced by the formation of this modified layer. To do. (For example, refer to Patent Document 1.)
Japanese Patent No. 3408805

Thus, in the laser processing method described above, when the thickness of the workpiece is 100 μm or less, an altered layer having a predetermined depth can be formed relatively easily inside the workpiece. When the thickness of the workpiece becomes as thick as 500 μm, for example, it becomes difficult to form a deteriorated layer having a predetermined thickness inside the workpiece.
In the case of focusing with a normal objective condenser lens, the laser beam is focused at a position away from the objective condenser lens by a focal length. However, when a condensing point is provided inside a member having a high refractive index such as a silicon wafer. It is generally known that the position of the condensing point is deepened by the refractive index times. For example, when a laser beam is focused through an objective condenser lens inside a workpiece with a high refractive index, such as a silicon wafer, the low incident angle component near the center and the NA (open number) of the objective condenser lens. In the near outside high incident angle component, the condensing point position is modulated due to the difference in the refraction angle on the surface of the workpiece. This modulation effect can be interpreted as a light beam distortion effect caused by a laser beam entering a substance having a high refractive index from the air layer. For this reason, when the position of the condensing point of the laser beam is set to a deep position of the workpiece, it becomes difficult to form a deteriorated layer having a predetermined thickness by modulating the condensing point position.

  The present invention has been made in view of the above-mentioned facts, and a main technical problem thereof is a laser capable of forming an altered layer having a predetermined thickness in a deep position from a light incident surface of a laser beam inside a workpiece. It is to provide a processing method.

In order to solve the above-mentioned main technical problem, according to the present invention, the laser beam oscillated from the laser beam oscillation means is condensed inside the workpiece through the objective condenser lens, and the altered layer is formed inside the workpiece. A laser processing method,
By changing the optical path length from the laser beam oscillating means to the objective condenser lens, the laser beam is adjusted so that the loss of kicking by the objective condenser lens is 25 to 35%, and the laser beam is adjusted to the objective condenser lens. Shine in
A laser processing method is provided.

  It is desirable to set the kicking loss to 30%.

In the laser processing method according to the present invention, the laser beam is adjusted by changing the optical path length from the laser beam oscillation means to the objective condenser lens so that the loss of the laser beam by the objective condenser lens is 25 to 35%. Is incident on the objective condenser lens, so that a deteriorated layer having a predetermined thickness can be formed at a deep position from the incident surface of the laser beam even for a workpiece having a high refractive index. In other words, the laser beam in the outer peripheral region where the energy of the laser beam having a Gaussian distribution is weak is cut, and the laser beam having a relatively strong energy is incident on the outer peripheral portion of the effective region of the objective condenser lens and is collected at a focal point with a short focal depth. Since the light is emitted, an altered layer having a predetermined thickness can be formed at a deep position from the incident surface of the laser beam even for a workpiece having a high refractive index.

  Hereinafter, a preferred embodiment of a laser processing method according to the present invention will be described in detail with reference to the accompanying drawings.

  FIG. 1 shows a schematic configuration diagram of a laser processing apparatus for carrying out a laser processing method according to the present invention. The laser processing apparatus shown in FIG. 1 includes a chuck table 3 for holding a wafer 2 as a workpiece, and a laser beam irradiation means generally indicated by numeral 4.

  As shown in FIG. 8, the wafer 2 is made of, for example, a silicon wafer having a thickness of 500 μm, and a plurality of division lines 21 are formed in a lattice pattern on the surface 2a. A circuit 22 as a functional element is formed on the surface 2 a of the wafer 2 in a plurality of regions partitioned by a plurality of division lines 21. In order to perform laser processing on the wafer 2 formed in this manner, a protective tape 20 for protecting the circuit 22 is attached to the surface 2a of the wafer 2 as shown in FIG.

  Returning to FIG. 1, the description will be continued. The chuck table 3 includes, for example, a suction chuck 31 formed of a porous member or formed with a plurality of suction holes or grooves, and the suction chuck 31 is not shown. It communicates with the suction means. Therefore, the wafer 2 is sucked onto the chuck table 3 by placing the protective tape 20 attached to the surface 2a of the wafer 2 as a workpiece on the suction chuck 31 and operating a suction means (not shown). Retained. The chuck table 3 configured as described above is configured to be moved in a processing feed direction indicated by an arrow X in FIG. Accordingly, the chuck table 3 and the laser beam irradiation means 4 are relatively movable in the processing feed direction indicated by the arrow X.

  The laser beam irradiation unit 4 includes a pulse laser beam oscillation unit 41, an output adjustment unit 42, a processing head 43, and an optical path length adjustment unit 44. The pulse laser beam oscillating means 41 includes a pulse laser beam oscillator composed of a YAG laser oscillator or a YVO4 laser oscillator and a repetition frequency setting means. The pulse laser beam oscillating means 41 is configured to be movable and adjustable in the left-right direction in FIG. That is, when the pulse laser beam oscillating means 41 moves to the right from the position indicated by the solid line in FIG. 1 as indicated by the two-dot chain line, the optical path length from the pulse laser beam oscillating means 41 to the objective condenser lens (to be described later) of the processing head 43 is increased. Increased. The output adjustment unit 42 adjusts the output of the pulse laser beam LB oscillated from the pulse laser beam oscillation unit 41. The processing head 43 includes a direction changing mirror 431 and a condenser lens 432 in the illustrated embodiment. The direction conversion mirror 431 converts the direction of the pulse laser beam LB adjusted to a predetermined output by the output adjusting unit 42 toward the objective condenser lens 432. The objective condenser lens 432 collects the pulsed laser beam LB that has passed through the direction conversion mirror 431 and irradiates the wafer 2 held on the chuck table 3.

  In the laser beam irradiation means 4 configured as described above, the diameter D of the pulse laser beam LB entering the objective condenser lens 432 forms a focal point called a pupil of the objective condenser lens 432 (see FIG. 2). Effective area) is set to be larger than E. Therefore, the pulse laser beam LB is lost because the region F outside the effective region E of the objective condenser lens 432 is kicked. In other words, the pulse laser beam LB is kicked at the bottom of the Gaussian distribution as shown in FIG. In the illustrated embodiment, the kicking loss is set to be 25 to 35%.

  The ratio of the kicking loss can be adjusted by changing the optical path length from the pulse laser beam oscillation means 41 to the objective condenser lens 432 of the processing head 43. That is, the pulse laser beam LB oscillated from the pulse laser beam oscillating means 41 has a property of spreading slightly (about 0.05 degrees) toward the downstream side, so that the diameter increases as the optical path length increases. Therefore, in the embodiment shown in FIG. 1, when the pulse laser beam oscillating means 41 is moved rightward in FIG. 1, the optical path length becomes longer, so the diameter D of the pulse laser beam LB entering the objective condenser lens 432 becomes larger, Kicked and loss increases.

Another embodiment of the optical path length adjusting means for changing the optical path length from the pulse laser beam oscillation means 41 to the objective condenser lens 432 will be described with reference to FIG.
The optical path length adjusting means 45 shown in FIG. 4 includes a first direction conversion mirror 451, a second direction conversion mirror 452, a third direction conversion mirror 453, a fourth direction conversion mirror 454, The first optical path adjustment mirror means 455 disposed between the direction conversion mirror 451 and the second direction conversion mirror 452 and the third direction conversion mirror 453 and the fourth direction conversion mirror 454 are arranged. The second optical path adjusting mirror means 456 is provided. The first direction conversion mirror 451 changes the direction of the pulse laser beam LB oscillated from the pulse laser beam oscillation unit 41 and adjusted to a predetermined output by the output adjustment unit 42 toward the first optical path adjustment mirror unit 455.

  The first optical path adjustment mirror means 455 includes a first mirror 455a and a second mirror 455b, and is configured to be movable and adjustable in the vertical direction in FIG. The first optical path adjusting mirror means 455 configured in this way changes the direction of the pulse laser beam LB, which has been direction-converted by the first direction-converting mirror 451, toward the second mirror 455b by the first mirror 455a, Further, the direction is changed toward the second direction changing mirror 452 by the second mirror 455b. The first optical path adjusting mirror means 455 can increase the optical path length of the pulsed laser beam LB by moving from the position indicated by the solid line to the position indicated by the two-dot chain line in FIG.

  The second direction conversion mirror 452 changes the direction of the pulse laser beam LB that has passed through the first optical path adjustment mirror means 455 toward the third direction conversion mirror 453. The third direction conversion mirror 453 changes the direction of the pulse laser beam LB that has passed through the second direction conversion mirror 452 toward the second optical path adjustment mirror means 456. The third optical path adjustment mirror means 456 includes a first mirror 456a and a second mirror 456b, similar to the first optical path adjustment mirror means 455, and is configured to be movable and adjustable in the vertical direction in FIG. ing. The second optical path adjustment mirror means 456 configured in this way changes the direction of the pulse laser beam LB that has passed through the third direction conversion mirror 453 toward the second mirror 456b by the first mirror 456a, and further The direction is changed toward the fourth direction changing mirror 454 by the second mirror 456b. The second optical path adjustment mirror means 456 can increase the pulse laser beam LB optical path length by moving from the position indicated by the solid line in FIG. 4 to the position indicated by the two-dot chain line.

  The fourth direction changing mirror 454 changes the direction of the pulse laser beam LB that has passed through the second optical path adjusting mirror means 456 toward the processing head 43.

Next, a deteriorated layer forming step of forming a deteriorated layer along the planned dividing line 21 inside the wafer 2 held on the chuck table 3 as shown in FIG. 1 will be described.
The chuck table 3 holding the wafer 2 is moved to a laser beam irradiation region where the processing head 43 of the laser beam irradiation unit 4 is located by a processing feed unit (not shown), and a predetermined division scheduled line 21 is formed as shown in FIG. One end (the left end in FIG. 5A) is positioned directly below the machining head 43. The chuck table 3, that is, the wafer 2 is moved at a predetermined processing feed speed in the direction indicated by the arrow X 1 in FIG. Let me. Then, as shown in FIG. 5B, when the irradiation position of the processing head 43 reaches the position of the other end of the wafer 2 (the right end in FIG. 5B), the irradiation of the pulse laser beam is stopped and the chuck table is stopped. 3. That is, the movement of the wafer 2 is stopped. In this deteriorated layer forming step, the condensing point P of the pulsed laser beam is matched with the vicinity of the surface 10 a (lower surface) of the wafer 2. As a result, the altered layer 210 is formed from the surface 2a toward the inside while being exposed on the surface 2a (lower surface) of the wafer 2. This altered layer 210 is formed as a melt-resolidified layer.

The processing conditions in the deteriorated layer forming step are set as follows, for example.
Wavelength: 1064nm
Output: 1W
Repeat frequency: 100kHz
Pulse width: 40ns
Condensing spot: Approximately φ1μm
NA of objective condenser lens: 0.7
Focal length of objective condenser lens: 2mm
Diameter when laser beam LB is emitted: φ500μm
Processing feed rate: 100mm / sec

Here, the relationship between the kicking loss and the thickness of the altered layer will be described.
On the condition that the output of the laser beam LB is constant on a silicon wafer having a thickness of 600 μm, the focal point P of the pulse laser beam LB is set at a position 530 μm from the incident surface (upper surface) of the laser beam LB on the silicon wafer, An experiment was conducted in which a deteriorated layer was formed by changing the ratio of the kicked loss under the same processing conditions as the layer forming step. In this experiment, the pupil (effective area) of the objective condensing lens 432 in the pulse laser beam LB is changed by changing the optical path length from the pulse laser beam oscillating means 41 to the condensing lens 432 of the processing head 43 in the range of 400 to 900 mm. Adjusted the rate of kicking outside. FIG. 6 is a graph showing the results of measuring the thickness of the altered layer formed by this experiment. In FIG. 6, the horizontal axis represents kicking loss (%), and the vertical axis represents the thickness (μm) of the deteriorated layer.

  From the experimental results shown in FIG. 6, when the kick loss at the objective condenser lens 432 is 0%, the deteriorated layer is not formed, and the thickness of the deteriorated layer suddenly increases from about 10% when the kick loss is about 10%. It has been found that when the loss is around 30%, the thickness of the deteriorated layer exceeds 60 μm and reaches a maximum and saturates. Therefore, an altered layer having a thickness of about 60 μm can be formed by setting the kicking loss to 25 to 35%. On the other hand, it was found that the output of the laser beam LB must be increased to 2.5 W in order to form the altered layer with a thickness of 60 μm under the condition that the kicking loss is 0%. This suggests that the effect of improving the light condensing performance in the vicinity of the condensing point is superior to the loss due to kicking by increasing the ratio of the high incident angle component. That is, when the kicking loss is set to about 30%, the laser beam in the outer peripheral region where the energy in the laser beam having the Gaussian distribution is weak is cut, and the laser beam having relatively strong energy in the outer peripheral portion of the effective region of the objective condenser lens 432 Is incident and focused on a focal point with a short focal depth, so that even a workpiece with a high refractive index can form an altered layer having a predetermined thickness deep from the incident surface of the laser beam. .

Next, the relationship between the kicking loss and the thickness of the deteriorated layer with respect to the focal point P position of the pulse laser beam LB will be described.
FIG. 7 shows experimentally the condensing point P position of the pulse laser beam LB and the thickness of the altered layer formed on the silicon wafer when the kicking loss is 30% and when the kicking loss is 10%. It is a graph which shows the measurement result. When the kick loss is 30%, the optical path length from the pulse laser beam oscillation means 41 to the condenser lens 432 of the machining head 43 is 900 mm, and when the kick loss is 10%, the pulse laser beam oscillation means 41 to the machining head 43 The optical path length to the condenser lens 432 was 450 mm. In this experiment, the condensing point P position of the pulse laser beam LB was adjusted by changing the distance between the objective condensing lens 432 and the chuck table 3. The irradiation conditions of the pulsed laser beam LB were the same as those in the above-described deteriorated layer forming step.

  As shown in FIG. 7, when the above-mentioned kick loss is 30%, it has been found that an altered layer having a thickness of about 60 μm is formed regardless of the position P of the pulse laser beam LB. On the other hand, when the kicking loss is 10%, an altered layer having a thickness of about 90 μm is formed when the condensing point P position of the pulse laser beam LB is about 100 μm from the incident surface. It was found that the deteriorated layer was hardly formed when the position of the point P was about 400 μm from the incident surface.

1 is a configuration block diagram of a laser processing apparatus that implements a laser processing method according to the present invention. Explanatory drawing which shows the kicking loss of the laser beam which injects into the objective condensing lens with which the laser processing apparatus shown in FIG. 1 is equipped. Explanatory drawing which shows the kicking loss of the laser beam shown in FIG. 2 by Gaussian distribution. The schematic block diagram which shows other embodiment of the optical path length adjustment means with which the laser processing apparatus shown in FIG. 1 is equipped. Explanatory drawing of the deteriorated layer formation process which forms a deteriorated layer in the inside of a wafer with the laser processing method by this invention. The graph which shows the relationship between the kicking loss and the thickness of an altered layer. The graph which shows the relationship between a condensing point position and the thickness of an altered layer. The perspective view of the wafer laser-processed by the laser processing method by this invention. The perspective view which shows the state which affixed the protective tape on the surface of the wafer shown in FIG.

Explanation of symbols

2: Wafer 20: Protective tape 3: Chuck table 4: Laser beam irradiation means 41: Pulse laser beam oscillation means 42: Output adjustment means 43: Processing head 431: Direction conversion mirror 432: Objective condenser lens 44: Optical path length adjustment means 45: Optical path length adjusting means 451: first direction conversion mirror 452: second direction conversion mirror 453: third direction conversion mirror 454: fourth direction conversion mirror 455: first optical path adjustment mirror means 456: second Optical path adjusting mirror means

Claims (2)

A laser processing method of condensing a laser beam oscillated from a laser beam oscillation means through an objective condensing lens into a workpiece and forming a deteriorated layer inside the workpiece,
By changing the optical path length from the laser beam oscillating means to the objective condenser lens, the laser beam is adjusted so that the loss of kicking by the objective condenser lens is 25 to 35%, and the laser beam is adjusted to the objective condenser lens. Shine in
The laser processing method characterized by the above-mentioned.   The laser processing apparatus according to claim 1, wherein the kicking loss is set to 30%.

Images