JP2011187760A - Method for manufacturing semiconductor device, and laser annealing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for manufacturing a semiconductor device, capable of suppressing variations in temperature history with respect to in-plane of a workpiece. <P>SOLUTION: The surface of an irradiation object is scanned with laser beam. A physical quantity depending on a surface temperature of the irradiation object, at a position where the laser beam is incident is measured, and an energy density to the irradiation object in an unscanned region in front of the scanning direction is set lower than that to a scanned region, on the basis of the measured physical quantity. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a method of manufacturing a semiconductor device including a step of annealing an irradiation object with a laser beam and a laser annealing apparatus.

  In the manufacturing process of an insulated gate bipolar transistor (IGBT), a collector is formed on the surface (second surface) opposite to the surface (first surface) on which the gate electrode and emitter are formed. By irradiating the second surface of the semiconductor wafer with the laser beam, the impurity implanted into the collector can be activated.

In order to activate the impurities implanted deep in the semiconductor wafer, the wavelength 690Nm～900nm, irradiation time 10Myuesu～100myuesu, a method of performing laser annealing under the conditions in the power density 250kW ^/ cm ² ~750kW ^/ cm 2 are known ( Patent Document 1). A method of controlling the irradiation time of a laser beam by using a continuous wave laser light source and controlling the beam spot size on the surface of a semiconductor wafer is known (Patent Document 2).

JP 2006-351659 A Japanese Patent No. 4117020

  When a semiconductor wafer is scanned with a laser beam, heat accumulates in the semiconductor wafer over time. For this reason, the average temperature of the wafer increases as scanning progresses. When the average temperature of the wafer rises, the temperature history varies depending on the position of the semiconductor wafer. When the variation of the temperature history occurs, the depth at which the impurity is activated varies in the plane.

  An object of the present invention is to provide a method of manufacturing a semiconductor device and a laser annealing apparatus that can suppress variations in temperature history with respect to the in-plane of a processing object.

According to one aspect of the invention,
Scanning the surface of the irradiation object with a laser beam;
Measuring a physical quantity depending on a surface temperature of the irradiation object at a position where the laser beam is incident;
And a step of lowering an energy density input to the irradiation object in the unscanned area forward in the scanning direction based on the measured physical quantity to be lower than an energy density input to the already scanned area. A manufacturing method is provided.

According to another aspect of the invention,
A laser light source for emitting a laser beam;
A stage for holding an irradiation object at a position where a laser beam emitted from the laser light source is incident;
A scanning mechanism for moving one of the path of the laser beam and the irradiation object relative to the other so that the incident position of the laser beam moves on the surface of the irradiation object held on the stage;
A measuring device for measuring a physical quantity depending on a surface temperature of a position where the laser beam is incident among the irradiation object;
There is provided a laser annealing apparatus having an energy density adjusting device for changing an energy density input to the irradiation object by the laser beam based on a measurement result of the physical quantity by the measuring device.

  By executing the process of reducing the energy density, excessive melting of the surface of the irradiation object can be suppressed.

1 is a schematic view of a laser annealing apparatus according to Example 1. FIG. (2A) and (2B) are plan views showing the positional relationship between the beam spot of the annealing laser beam and the beam spot of the reference laser beam. It is a graph which shows the simulation result of the relationship between the irradiation time until the surface temperature of an irradiation target object becomes 1673K, and a power density. It is a graph which shows the result of having calculated | required with the simulation the relationship between the maximum temperature of the position of 3 micrometers depth when laser irradiation is performed until the surface temperature of an irradiation target object is 1673K, and a power density. It is a graph which shows the result of having calculated | required the relationship between the temperature of the back surface when laser irradiation is performed until the surface temperature of an irradiation target object is 1673K, and elapsed time by simulation. 3 is a flowchart showing a laser annealing method according to Example 1. It is a graph which shows the result of having calculated | required the time history of surface temperature when performing laser irradiation on the time until the surface temperature becomes 1673K when the initial temperature Ti of the irradiation object is 300K, and the power density is constant, by simulation. is there. It is a graph which shows the result of having calculated | required the time history of the surface temperature when performing laser irradiation by varying the power density and the time until the surface temperature reaches 1673K when the initial temperature Ti of the irradiation object is 300K. is there. 7 is a plan view showing a positional relationship between a beam spot of an annealing laser beam and a beam spot of a reference laser beam according to a modification of Example 1. FIG. 6 is a flowchart showing a laser annealing method according to Example 2. 6 is a schematic view of a laser annealing apparatus according to Example 3. FIG. The graph which shows the result which calculated | required the time history of the surface temperature when performing laser irradiation on the conditions which stop laser irradiation when the surface temperature of an irradiation target object becomes 1673K, and a power density is constant, by simulation is there. 6 is a schematic view of a laser annealing apparatus according to Example 4. FIG. (14A) is a graph showing the power density distribution on the surface of the irradiation object, (14B) is a graph showing the energy density distribution, and (14C) is when the laser irradiation is intermittently performed several times. It is a graph which shows energy density distribution. It is sectional drawing of IGBT.

[Example 1]
FIG. 1 shows a schematic diagram of a laser annealing apparatus according to the first embodiment. The laser light source 10 emits a continuous wave laser beam. For the laser light source 10, for example, a semiconductor laser is used. The oscillation wavelength is, for example, 808 nm. A semiconductor laser having an oscillation wavelength in the range of 650 nm to 850 nm may be used. An irradiation object 60 is held on the stage 35. The irradiation object 60 is, for example, a silicon wafer before impurities are implanted and activation annealing is performed. Laser annealing is performed when the laser beam is incident on the irradiation object 60. The laser beam emitted from the laser light source 10 is referred to as “annealing laser beam”.

  The laser beam emitted from the laser light source 10 enters the irradiation object 60 via the power adjusting device 20, the beam shaping optical system 25, the homogenizing optical element 26, the folding mirror 30, and the lens 31.

  The power adjustment device 20 includes a control device 21, a half-wave plate 22, a beam splitter 23, and a beam damper 24. The laser beam emitted from the laser light source 10 enters the beam splitter 23 via the half-wave plate 22. The beam splitter 23 transmits the P component of the incident laser beam and reflects the S component. The laser beam reflected by the beam splitter 23 enters the beam damper 24. The laser beam transmitted through the beam splitter 23 is emitted from the power adjustment device 20 and is incident on the beam shaping optical system 25.

  The control device 21 rotates the half-wave plate 22 around the center beam of the laser beam. When the half-wave plate 22 is rotated, the ratio of the P component and S component of the laser beam incident on the beam splitter 23 changes. Thereby, the power of the laser beam emitted from the power adjusting device 20 can be adjusted. An optical attenuator with variable attenuation may be used as the power adjustment device 20.

  The beam shaping optical system 25 shapes the beam cross section of the laser beam. For the beam shaping optical system 25, for example, a beam expander is used. The laser beam whose beam cross section has been shaped by the beam shaping optical system 25 enters the uniformizing optical system 26. The uniformizing optical element 26 uniformizes the light intensity distribution in the beam cross section of the laser beam. For the uniformizing optical element 26, for example, a diffractive optical element (DOE), an array lens, a fly-eye lens, or the like is used.

  The laser beam transmitted through the homogenizing optical element 26 is reflected by the folding mirror 30 and enters the irradiation object 60 via the lens 31. The central beam of the laser beam incident on the irradiation target 60 is perpendicular to the surface of the irradiation target 60. The positions of the homogenizing optical system 26, the lens 31, and the stage 35 are adjusted with respect to the traveling direction of the laser beam so that the light intensity distribution is uniform on the surface of the irradiation object 60 held on the stage 35.

  The reference light source 40 emits a reference laser beam. For the reference light source 40, for example, a HeNe laser having a wavelength of 633 nm is used. Note that a laser in the near-infrared region having an oscillation wavelength of visible light or a wavelength of 1 μm or less may be used. The reference laser beam is reflected by the folding mirror 41 and obliquely enters the irradiation object 60. The beam spot of the reference laser beam on the surface of the irradiation object 60 is included or partially overlaps with the beam spot of the annealing laser beam. A preferred positional relationship between the beam spots will be described later.

  A part of the reference laser beam that has entered the irradiation target 60 is reflected by the surface of the irradiation target 60, is reflected by the folding mirror 42, and enters the reference light detector 43. For the reference light detector 43, for example, a photodiode is used. The intensity data of the reflected light detected by the reference light detector 43 is input to the control device 21.

  Of the surface of the irradiation object 60, blackbody radiation from a specific position within the beam spot of the annealing laser beam is condensed by the condenser lens 50 and enters the blackbody radiation detector 51. Black body radiation intensity data detected by the black body radiation detector 51 is input to the control device 21. The control device 21 controls the attitude of the half-wave plate 22 in the rotation direction based on at least one of reflected light intensity data and blackbody radiation intensity data. In the first embodiment, the condenser lens 50 and the black body radiation detector 51 are not essential.

  The scanning mechanism 36 moves the stage 35 in the X axis direction and the Y axis direction parallel to the surface of the irradiation target 60. By moving the stage 35 while irradiating the annealing laser beam, the surface of the irradiation object 60 can be scanned with the laser beam. For example, let the X-axis direction and the Y-axis direction be the main scanning direction and the sub-scanning direction, respectively.

  FIG. 2A shows the positional relationship between the beam spot of the annealing laser beam and the beam spot of the reference laser beam. The beam spot 61 of the annealing laser beam is, for example, a rectangle that is long in the Y-axis direction. The dimension in the X-axis direction (hereinafter referred to as “beam width”) Wx is, for example, 50 μm. When the beam spot 61 of the annealing laser beam moves in the positive direction of the X axis (the stage 35 moves in the negative direction of the X axis with respect to the path of the annealing laser beam), the beam spot of the reference laser beam 45 is arranged at a position in contact with the rear edge (edge on the negative side of the X axis) of the beam spot 61 of the annealing laser beam with respect to the traveling direction (main scanning direction). The shape of the beam spot 45 is, for example, a circle having a diameter of 10 μm.

  As shown in FIG. 2B, when the moving direction of the beam spot 61 of the annealing laser beam is the negative direction of the X axis, the beam spot 45 of the reference laser beam is rearward with respect to the traveling direction (main scanning direction). It is arranged at a position in contact with the side edge (edge on the positive side of the X axis).

  Next, with reference to FIG. 3 to FIG. 5, preferable annealing conditions for performing activation annealing of a semiconductor wafer, for example, a silicon wafer will be described. When annealing the IGBT collector, the annealing should be performed under the condition that the temperature on the front side where the gate electrode, the emitter, etc. are formed does not increase excessively and is activated as deep as possible from the back surface. preferable.

  An ultraviolet laser such as an excimer laser is absorbed by the extreme surface of the semiconductor wafer, and is not suitable for annealing to activate to a deep position. Further, when the wavelength is in the far infrared region, the absorption coefficient of the semiconductor wafer is low, so that it is not suitable for annealing. For activation annealing of the IGBT collector, it is preferable to use a laser having a wavelength of 650 nm to 850 nm.

FIG. 3 shows a simulation result of the relationship between the power density of the laser beam on the surface of the semiconductor wafer and the irradiation time until the surface temperature rises to 1673K. The horizontal axis represents the power density in the unit “kW / cm ² ”, and the vertical axis represents the irradiation time in the unit “μs”. The wavelength of the laser beam was 800 nm, and the semiconductor wafer material was silicon. The melting point of silicon is 1687K. As the power density increases, the irradiation time until the surface temperature reaches 1673K is shortened.

FIG. 4 shows the maximum temperature reached at a depth of 3 μm from the surface on the laser irradiation side in the simulation of FIG. The horizontal axis represents the power density in the unit “kW / cm ² ”, and the vertical axis represents the maximum attained temperature in the unit “K”. Under any irradiation condition, the maximum temperature reached on the surface is 1673K, but the maximum temperature reached at a depth of 3 μm varies depending on the power density. As the power density increases, the maximum temperature reached tends to decrease.

  As shown in FIG. 3, when the power density is increased, the irradiation time is shortened, so that the vicinity of the surface of the semiconductor wafer is rapidly heated. Thereby, the temperature rise width inside the semiconductor wafer is reduced, and the time during which the temperature is kept high is shortened. It can be seen that when the power density increases, it is difficult to activate deep region impurities under the condition that the surface is not melted.

  FIG. 5 shows a simulation result of the relationship between the elapsed time from the start of laser irradiation and the temperature of the back surface of the semiconductor wafer (the surface opposite to the surface irradiated with laser). The horizontal axis represents the elapsed time in the unit “μs”, and the vertical axis represents the back surface temperature in the unit “K”. The numerical value attached to each curve indicates the power density of the laser beam on the surface of the semiconductor wafer. The laser irradiation was stopped when the surface temperature of the semiconductor wafer reached 1673K. As boundary conditions, the thickness of the semiconductor wafer was set to 100 μm, and the back surface was insulative.

  As the power density decreases, the temperature reached on the back surface of the semiconductor wafer increases. This is because when the power density is lowered, the irradiation time becomes longer as shown in FIG. 3, so that the total energy input to the semiconductor wafer is increased.

  In order to increase the activation depth, it is required to perform annealing at a high temperature for a long time. In order to perform long-time annealing at a high temperature, it is preferable to lower the power density from the results shown in FIGS. However, as shown in FIG. 5, when the power density is lowered, the back surface temperature of the semiconductor wafer is greatly increased. The configuration of the back surface of the semiconductor wafer, that is, the surface side of the IGBT, restricts the temperature increase range allowed during laser annealing. When the allowable temperature rise is large, deep region impurities can be activated by laser annealing at a low power density.

On the other hand, when the allowable temperature rise is small, it is necessary to anneal at a high power density, and it becomes difficult to increase the activation depth. As shown in FIG. 5, when the power density is less than 150 kW / cm ² , the increase in the back surface temperature of the semiconductor wafer becomes too large. Therefore, it is understood that the power density is preferably 150 kW / cm ² or more.

The upper limit of the preferred value of the power density depends on the target value of the activation depth. When the power density exceeds 1 MW / cm ² , the order is about the same as that in the annealing method using a pulse laser, and it becomes difficult to perform activation under non-melting conditions. Therefore, the power density is preferably 1 MW / cm ² or less.

In general, it is desirable that the maximum temperature reached on the back surface of the semiconductor wafer be 600K or less. For this reason, it is preferable that the power density is 250 kW / cm ² or more from the simulation result of FIG. Moreover, when the target value of the activation depth is set to about 3 μm, it is preferable that the maximum temperature reached at the position of the depth of 3 μm exceeds 1000 ° C. (1273 K). Therefore, from the simulation result of FIG. 4, it is preferable to set the power density to 500 kW / cm ² or less.

  With reference to FIG. 6, the laser annealing method according to the first embodiment will be described. FIG. 6 is a flowchart of the laser annealing method according to the first embodiment.

  In step SA1, a reflected light intensity measurement position is set. Specifically, as shown in FIGS. 2A and 2B, the beam spot 45 of the reference laser beam is arranged at a position in contact with the rear edge in the scanning direction in the beam spot 61 of the annealing laser beam. . The beam spot 45 of the reference laser beam can be moved relative to the beam spot 61 by moving the folding mirrors 41 and 42 in the X-axis direction with respect to the path of the annealing laser beam shown in FIG. it can.

  In step SA2, the stage 35 is moved in the X-axis direction. When the edge of the region to be annealed reaches the incident position of the annealing laser beam, the irradiation of the annealing laser beam is started. At this time, a running period is provided so that the moving speed of the stage 35 reaches the target speed. The irradiation of the annealing laser beam is started by actually starting laser oscillation of the laser light source 10. Alternatively, the laser beam may be oscillated before the start of irradiation, and the irradiation of the annealing laser beam may be started by changing the orientation of the half-wave plate 22 in the rotational direction.

At the start of irradiation, as an example, the power density on the surface of the irradiation object 60 is set to 400 kW / cm ² . From FIG. 3, when the power density is 400 kW / cm ₂ , the surface temperature reaches 1673 K when the irradiation time is 14.1 μs. The irradiation time t is determined by using the beam width Wx of the annealing laser beam and the scanning speed v shown in FIGS. 2A and 2B.

It is expressed. When the beam width Wx is 50 μm and the irradiation time t is desired to be 14.1 μs, the scanning speed v may be set to about 3.5 m / s.

  When scanning with the annealing laser beam is continued, energy is accumulated in the irradiation object 60 and its average temperature rises. At the start of scanning, the laser irradiation conditions are set so that the maximum temperature reached on the surface of the irradiation object 60 is 1673 K. However, when the average temperature of the irradiation object 60 increases, The temperature may be higher than the melting point.

  In step SA3, it is determined whether or not main scanning for one row has been completed. If the main scanning for one row has not been completed, the intensity of the reflected light is measured by the reference light detector 43 in step SA4. Thereafter, in step SA5, the measured intensity of reflected light is compared with a threshold value. When an extremely shallow region on the surface of the irradiation object 60 is melted and enters a liquid phase state, the reflectance increases. By measuring the intensity of the reflected light, it can be detected that the surface of the irradiation object 60 has melted. Specifically, an intermediate intensity between the intensity of the reflected light in the solid phase state and the intensity of the reflected light in the liquid phase state is stored in the control device 21 as a threshold value. When the measured intensity of the reflected light is equal to or higher than this threshold value, it is considered that an extremely shallow region on the surface of the irradiation object 60 has melted.

  If it is determined that the intensity of the measured reflected light is greater than or equal to this threshold value, that is, if it is determined that the extremely shallow region of the surface of the irradiation object 60 is melted, in step SA6, The energy density of the laser energy thrown into the irradiation object 60 is reduced. Specifically, the power density on the surface of the irradiation object 60 is reduced while the scanning speed v is constant. The reduction width of the power density is set in the control device 21 in advance. After reducing the power density, the process returns to step SA3.

  If it is determined in step SA5 that the intensity of the reflected light is less than the threshold value, the process returns to step SA3 without correcting the energy density.

  If it is determined in step SA3 that the main scanning for one row has been completed, the irradiation of the annealing laser beam is stopped in step SA7. Next, in step SA8, it is determined whether scanning of all rows to be annealed has been completed. When the scanning of all the rows is finished, the annealing process is finished. If unscanned rows remain, in step SA9, the stage 35 is moved in the Y-axis direction to perform sub-scanning for one row.

  In step SA10, the reflected light intensity measurement position is reset. Specifically, since the direction of main scanning is reversed, the position of the beam spot 45 of the reference laser beam shown in FIGS. 2A and 2B is moved from the position in contact with one edge of the beam spot 61 of the annealing laser beam. And move to a position in contact with the other edge. Thereafter, the process returns to step SA2 to start main scanning.

  Next, effects of the laser annealing method according to the first embodiment will be described with reference to FIGS.

FIG. 7 shows a simulation result of the relationship between the elapsed time from the start of irradiation of the annealing laser beam and the surface temperature of the semiconductor wafer. The horizontal axis represents elapsed time in units of “μs”, and the vertical axis represents surface temperature in units of “K”. The solid line, broken line, and dotted line in FIG. 7 indicate changes in the surface temperature when the surface temperature (initial temperature) Ti at the start of irradiation is 300K, 323K, and 373K, respectively. The power density on the surface of the semiconductor wafer was 400 kW / cm ² . When the laser irradiation was performed under the condition that the surface temperature at the start of irradiation was 300K, the laser irradiation was stopped at an irradiation time of 14.1 μs until the surface temperature reached 1673K.

  When the initial temperature Ti of the semiconductor wafer is 323 K and 373 K, the surface temperature reaches 1673 K at 13.4 μs and 12.0 μs, respectively. Thereafter, when the irradiation is continued until the irradiation time reaches 14.1 μs, not only the extremely shallow region of the surface but also the deep region is melted. At this time, since the heat energy input to the semiconductor wafer is consumed as heat of fusion, the surface temperature is maintained at the melting point.

FIG. 8 shows a simulation result of the relationship between the elapsed time from the start of irradiation of the annealing laser beam and the surface temperature of the semiconductor wafer when the power density is optimized. The power density when the initial temperature Ti is 300K and 400 kW / cm ^2, the initial temperature Ti is set to 389.2kW / cm ² power density at 323 K, the initial temperature Ti is 367.5kW / cm ² power density at the time of 373K It was.

  Under these conditions, in any case, the surface temperature reaches 1673 K when 14.1 μs has elapsed from the start of irradiation. Thus, by optimizing the power density, even when the average temperature of the semiconductor wafer rises, the maximum surface temperature can be made slightly lower than the melting point without changing the irradiation time.

  As shown in FIG. 7, when scanning is performed with a constant power density, the surface layer portion at the position irradiated with the laser in a state where the average temperature of the irradiation object 60 is increased is melted. The depth of the melted region depends on the time from when the surface temperature reaches the melting point to when the laser irradiation is stopped. This time depends on the initial temperature Ti of the semiconductor wafer. As the scanning progresses and the average temperature of the semiconductor wafer rises, the activation depth varies.

  In the step SA5 shown in FIG. 6, the state where the reflected light intensity is determined to be equal to or higher than the threshold value means that the surface temperature has risen to the melting point. In step SA6, by reducing the power density, it is possible to prevent the surface from being melted in a region scanned after the position where the annealing laser beam is currently incident. If the scanning is continued, the average temperature of the irradiation object 60 further rises, and the surface may rise to the melting point even with the power density once lowered. In this case, the power density is further reduced in step SA6. In Example 1, since the melt | dissolution beyond that will be suppressed if the very shallow area | region of a surface layer part melt | dissolves, the dispersion | variation in the depth of activation can be suppressed.

If the step size for reducing the power density is too large, the amount of energy input to the irradiation object 60 becomes too small, resulting in insufficient annealing. If the step size of the power density reduction is too small, the power density reduction cannot follow the increase in the average temperature of the object 60 to be processed. It is preferable to determine a suitable value for the step size for power density reduction by actually conducting an evaluation experiment. For example, when the elapsed time and the change in the surface temperature have the relationship shown in FIG. 7, the power density reduction step size may be 0.5% to 2% of the current power density, or 2 to 10 kW / it may be used as the cm ^2.

  In the first embodiment, as shown in FIGS. 2A and 2B, the beam spot of the reference laser beam is formed in the region adjacent to the rear edge in the scanning direction in the beam spot 61 of the annealing laser beam. 45 was placed. Thereby, the reflectance of the region having the highest surface temperature in the beam spot 61 can be measured.

  As shown in FIG. 9, the beam spot 45 of the reference laser beam may be arranged so as to straddle in the width direction (main scanning direction) from one edge of the beam spot 61 of the annealing laser beam to the other edge. Good. When melting starts in a part of the area within the beam spot 45, the average reflectance increases, and the intensity of the reflected light increases. Accordingly, it is possible to detect that melting has started in the vicinity of the rear edge in the scanning direction. In this case, even if the scanning direction is reversed, it is not necessary to move the position of the beam spot 45 of the reference laser beam.

  In the first embodiment, the reference laser beam is obliquely incident on the irradiation target 60. However, the reference laser beam may be vertically incident on the annealing laser beam. In the case of perpendicular incidence, for example, a dichroic mirror may be disposed between the folding mirror 30 and the lens 31 in FIG. 1, and the reference laser beam may be superimposed on the annealing laser beam path. The dichroic mirror transmits the annealing laser beam and reflects the reference laser beam. The reference laser beam reflected by the irradiation object 60 is reflected by the dichroic mirror and propagates in the direction opposite to the incident path. By inserting a beam splitter and a quarter-wave plate between the dichroic mirror and the reference light source 40, the path of the reflected light can be branched from the path of the incident light. A reference light detector 43 is arranged on the reflected light path after branching.

  When the path of the laser beam for reference is superimposed on the path of the laser beam for annealing, the lens 31 is subjected to antireflection processing in two wavelength regions of the laser beam for annealing and the laser beam for reference. It is preferable to use one that has been designed to be erased.

  In the configuration in which the reference laser beam is incident obliquely, the incident position of the reference laser beam also changes when the height of the surface of the irradiation target 60 changes. For this reason, the position of the optical system of the reference laser beam must be readjusted. In the configuration in which the path of the reference laser beam is superimposed on the path of the annealing laser beam, this readjustment is not necessary.

[Example 2]
FIG. 10 shows a flowchart of the laser annealing method according to the second embodiment. Hereinafter, description will be made by paying attention to differences from the flowchart of the first embodiment shown in FIG. 6, and description of common points will be omitted. The configuration of the laser annealing apparatus is the same as that shown in FIG.

  Instead of step SA1 of the first embodiment, a black body radiation intensity measurement position is set in step SB1. Specifically, as shown in FIGS. 2A and 2B, the region in contact with the rear edge in the scanning direction within the beam spot 61 of the annealing laser beam (the same as the beam spot 45 of the reference laser beam). The positions of the condenser lens 50 and the black body radiation detector 51 are adjusted so that the black body radiation light from the region) is collected by the condenser lens 50 and enters the black body radiation detector 51. Steps SA2 and SA3 are the same as those in the first embodiment.

  Instead of step SA4 of the first embodiment, the black body radiation intensity is measured by the black body radiation detector 51 in step SB4. The measurement result is input to the control device 21. By measuring the black body radiation intensity, the surface temperature of the irradiation object 60 can be calculated. A method for calculating the surface temperature of the object to be processed from the black body radiation intensity is described in JP-A-2008-116269.

  In step SB5, it is determined whether or not the black body radiation intensity is greater than or equal to a threshold value. This threshold value may be equal to the melting point of the processing object 60 or the black body radiation intensity from the surface at a temperature slightly lower than the melting point. For example, this threshold value is set to be equal to the black body radiation intensity when the surface temperature is 1673K. When it is determined that the measured black body radiation intensity is equal to or higher than the threshold value, that is, the surface temperature is equal to or higher than 1673 K, the energy density of the laser energy input to the irradiation target 60 is decreased in step SA6. This process is common to the process in step SA6 of the first embodiment shown in FIG. When it is determined that the measured black body radiation intensity is less than the threshold value, that is, the surface temperature is less than 1673K, the process returns to step SA3.

  Steps SA7, SA8, and SA9 are the same as those in the flowchart according to the first embodiment shown in FIG. Instead of step SA10, the measurement position of the black body radiation intensity is reset in step SB10. Specifically, the measurement position of the black body radiation intensity is moved to a position in contact with the rear edge in the scanning direction within the beam spot 61 of the annealing laser beam shown in FIGS. 2A and 2B. Specifically, the positions of the condenser lens 50 and the blackbody radiation detector 51 shown in FIG. 1 are readjusted.

  Also in Example 2, since the energy density of laser irradiation is reduced as the average temperature of the irradiation object 60 increases, the surface can be prevented from melting. Thereby, the variation in the activation depth can be suppressed.

  As an example, when the laser irradiation is performed under the condition where the power density and the pulse width shown in FIG. 7 are constant, the highest temperature reached at the position of 3 μm depth when the initial temperature Ti of the irradiation object is 300K, 323K, and 373K. The simulation results were 1372K, 1400K, and 1428K, respectively. On the other hand, in the method according to Example 2 shown in FIG. 8, the simulation results of the maximum temperature reached at the position of 3 μm depth when the initial temperature Ti of the irradiation object is 300K, 323K, and 373K are 1372K and 1373K, respectively. 1374K. From the simulation results, it can be seen that the variation in the maximum temperature reached inside the irradiation object is reduced by the method of Example 2. Since the variation in the maximum temperature is suppressed, it can be expected that the variation in the activation depth is also reduced.

  Since black body radiation is radiated from the surface of the irradiation object 60 into a space with a solid angle of 2π steradians, there is little black body radiation incident on the condenser lens 50. Therefore, it is preferable to use a detector having a multiplication function such as an avalanche photodiode (APD) or a photomultiplier tube (PMT) as the black body radiation detector 51. Note that some of these detectors cannot detect CW component light, and when the irradiation target is large and the measurement time of the black body radiation light intensity is long, the measured intensity may not be stable. is there. In this case, a black body radiation light input to the detector may be appropriately blocked using a chopper, a high-speed shutter, or the like. In order not to be affected by the scattered light of the annealing laser beam, the black body radiation detector 51 is disposed in the dark room, and the black body radiation detector 51 is disposed in the dark room with an optical fiber. It is preferable to transmit up to.

  The intensity of the reflected light measured in Example 1 and the black body radiation intensity measured in Example 2 are both physical quantities that depend on the surface temperature of the irradiation object. In addition to the intensity of reflected light and the black body radiation intensity, other physical quantities that depend on the surface temperature may be measured. For example, the surface temperature can be calculated by measuring the extinction coefficient.

  In addition, you may determine whether an energy density is reduced based on both the measurement result of reflected light intensity, and the measurement result of black body radiation intensity.

[Example 3]
FIG. 11 shows a schematic diagram of a laser annealing apparatus according to the third embodiment. In the first embodiment and the second embodiment, as shown in FIG. 1, the power adjusting device 20 including the control device 21 adjusts the power of the annealing laser beam, so that the laser energy input to the irradiation target 60 is increased. Reduced energy density. In the third embodiment, the control device 21 can control the scanning mechanism 36 to change the scanning speed.

  The energy density E input to the irradiation target 60 is P, the power density on the surface of the irradiation target 60, the beam width Wx, and the scanning speed v.

It is expressed. In Example 1, the energy density E was reduced by reducing the power density P on the right side of Formula 4 in Step SA6 of FIG. In the third embodiment, the energy density E is decreased by increasing the scanning speed v on the right side of Equation 4 in step SA6 in FIG.

FIG. 12 shows a simulation result of the relationship between the elapsed time from the start of irradiation of the annealing laser beam and the surface temperature. The horizontal axis represents elapsed time in units of “μs”, and the vertical axis represents surface temperature in units of “K”. The power density on the surface of the irradiation start object 60 is 400 kW / cm ² . The solid line, broken line, and dotted line in FIG. 12 indicate changes in the surface temperature when the initial temperature Ti of the irradiation object 60 is 300K, 323K, and 373K, respectively. Laser irradiation was stopped when the surface temperature reached 1673K.

  When the initial temperature Ti is 300 K, 323 K, and 373 K, the surface temperature reaches 1673 K when 14.1 μs, 13.4 μs, and 12.0 μs have elapsed from the start of irradiation, respectively. When laser irradiation is stopped, the surface temperature gradually decreases.

  As the scanning speed v increases, the time (irradiation time) during which the annealing laser beam is incident on a minute region on the surface of the irradiation object 60 is shortened. Thus, by shortening the irradiation time, it is possible to control the surface temperature so as not to exceed the melting point. Thereby, similarly to the case of the first embodiment and the second embodiment, variation in the activation depth can be suppressed.

  As an opportunity to reduce the scanning speed, the intensity of reflected light may be used as in step SA5 in FIG. 6, the black body radiation intensity may be used as in step SB5 in FIG. 10, or both may be used. Also good.

  As can be seen from Equation 4, the energy density E can also be reduced by making the power density P constant and narrowing the beam width Wx. The beam width Wx can be adjusted by using a variable-width slit arranged in the annealing laser beam path and an imaging lens that forms an image of the slit on the surface of the irradiation object 60.

  Thus, the energy density E can be adjusted by adjusting one of the scanning speed v and the beam width Wx to control the irradiation time of the annealing laser beam.

  In the method for adjusting the power density according to the first embodiment, as shown in FIG. 8, the surface temperature when the initial temperature Ti of the irradiation object 60 is high is the laser beam irradiation period and the surface temperature drop after the end of irradiation. In any period, it is higher than the surface temperature when the initial temperature Ti is low. On the other hand, in Example 3, as shown in FIG. 12, the surface temperature when the initial temperature Ti of the irradiation object 60 is high is the surface temperature when the initial temperature Ti is low during the laser beam irradiation period. taller than. However, the vertical relationship of the surface temperature is reversed during a certain period during which the surface temperature is lowered after the laser irradiation is stopped. For this reason, the integrated value of the temperature history is leveled.

  The calculation result of the activation depth when activation annealing was performed under the three types of laser irradiation conditions shown in FIG. 12 was 2.27 μm in any condition. Thus, the method according to the third embodiment has a higher effect of suppressing the variation in activation depth than the method according to the first embodiment.

[Example 4]
FIG. 13 shows a schematic diagram of a laser annealing apparatus according to the fourth embodiment. In the fourth embodiment, an electro-optic element (EOM) 22A is used instead of the half-wave plate 22 shown in FIG. The control device 21, the electro-optic element 22A, the beam splitter 23, and the beam damper 24 constitute a blocking device 20A. For example, a Pockels cell is used as the electro-optic element 22A. The electro-optic element 22A changes the polarization direction of the laser beam under the control of the control device 21. In the fourth embodiment, it is possible to select a state in which all the laser beams incident on the beam splitter 23 are P components and a state in which all laser beams are S components.

  When the laser beam incident on the beam splitter 23 is only the P component, the laser beam passes through the beam splitter 23 and enters the irradiation object 60. When the laser beam incident on the beam splitter 23 is only the S component, the laser beam is reflected by the beam splitter 23 and therefore does not enter the irradiation target 60.

  The electro-optic element 22A can switch between the state of only the P component and the state of only the S component at a higher speed than when the half-wave plate 22 is rotated. For this reason, the blocking device 20A can block the annealing laser beam at high speed. In place of the electro-optic element 22A and the beam splitter 23, an acousto-optic element (AOM) may be used. In the fourth embodiment, the state where the irradiation object 60 is irradiated with the annealing laser beam and the state where the annealing laser beam is blocked are alternately repeated.

FIG. 14A shows an example of the power density distribution of the annealing laser beam incident on the surface of the irradiation object 60. Since the light intensity distribution of the annealing laser beam is uniformized by the homogenizing optical element 26, the power density distribution is assumed to be constant. The horizontal axis represents the position in the scanning direction (X-axis direction), and the vertical axis represents the power density. The power density distribution of the annealing laser beam incident at the irradiation start time (t = t ₀ ) is represented by a thin solid line P0, and the power of the annealing laser beam incident at the irradiation interruption time (t = t ₂ ). The density distribution is represented by a thick solid line P1. The moving distance L1 of the beam spot from the time when irradiation is started to the time when irradiation is interrupted is sufficiently shorter than the beam width Wx.

In FIG. 14B, the distribution of energy density of the laser beam incident on the irradiation object 60 during the irradiation period from time t ₀ to t ₂ is indicated by a thick solid line E1. In the region where the laser beam continuously from the irradiation start time t ₀ to cutoff time t ₂ is incident, energy density is constant. On both sides of the region where the energy density is constant, there are inclined regions where the energy density decreases toward the outside.

FIG. 14C shows an energy density distribution input to the irradiation object 60 when the annealing laser beam is intermittently irradiated. The energy density distribution E1 (t ₀ -t ₂ ) of laser energy input by irradiation from time t ₀ to t ₂ and the energy density distribution E1 of laser energy input by irradiation from time t ₃ to t ₄ ( t ₃ −t ₄ ) partially overlap. Similarly the energy density distribution of laser energy E1 _(t 3 -t ₄₎ to be introduced by irradiation to _{t 4} from time _{t 3,} the energy density distribution of laser energy to be introduced by irradiation from time _{t 5} to _{t 6} E1 (t ₅ -t ₆ ) partially overlaps.

  The energy density distribution of the laser energy actually input to the irradiation object 60 is a distribution obtained by accumulating these energy density distributions. In this manner, the laser beam for annealing that oscillates continuously may be irradiated in a pulsed manner by intermittently blocking it.

As shown in FIG. 14A, showing the power density distribution in the cutoff time when the cut-off time was set earlier t ₁ than t ₂ by a broken line P2. In FIG. 14B, the energy density distribution of the laser energy input to the irradiation object 60 during the irradiation period from time t ₀ to t ₁ is indicated by a broken line E2. Shorter irradiation time as compared to blocking time when the t _2, the energy density of the region where the energy density is constant is lower than when the cut-off time was t _2. Thus, by shortening the irradiation time, the energy density of the laser energy input to the irradiation object 60 can be reduced.

  In Step SA6 of Example 1 shown in FIG. 6 and Step SA6 of Example 2 shown in FIG. 10, the energy density can be reduced by shortening the irradiation time.

  In the method according to the first to third embodiments, as shown in Expression 3, when the beam width Wx and the time t to be irradiated are determined, the scanning speed v is also determined. For example, when the time t to be irradiated is 50 μs and the beam width Wx is 50 μm, the scanning speed v is 1 m / s. When the scanning speed increases, a space for accelerating the stage 35 until reaching a constant speed and a space for decelerating to reverse the scanning direction must be secured. For this reason, an annealing apparatus will enlarge.

  In Example 4, the irradiation time can be changed regardless of the scanning speed. For this reason, it becomes possible to make a scanning speed slow compared with the case of Example 1-3. Since the space to be secured for acceleration and deceleration may be small, the apparatus can be miniaturized.

  In the fourth embodiment, a configuration in which a continuously oscillated laser beam is blocked is employed. In addition, a laser beam that performs quasi-continuous oscillation (QCW) may be used, or a pulse laser beam having a pulse width of 1 μs or more may be used. When a pulse laser beam having a pulse width of 1 μs or more is used, the irradiation time can be adjusted by blocking a part of one pulse with the blocking device 20A. In the case where the laser light source 10 is a laser oscillator having a variable pulse width, the laser light source 10 may be directly controlled by the control device 21 without disposing the electro-optical element 22A, the polarizer 23, and the beam damper 24.

  FIG. 15 is a schematic diagram of a cross section of an IGBT to which the laser annealing method according to the first to fourth embodiments is applied. An IGBT is manufactured by forming an emitter and a gate on one surface of an n-type semiconductor substrate and forming a collector on the other surface. The structure of the surface on which the emitter and the gate are formed is manufactured in the same process as a general MOSFET manufacturing process. For example, as shown in FIG. 15, by disposing a p-type base region 74, an n-type emitter region 75, a gate electrode 81, a gate insulating film 82, and an emitter electrode 80, a current at a voltage between the gate and the emitter can be obtained. ON / OFF control can be performed.

  A p-type collector layer 73 is formed on the opposite surface of the semiconductor substrate 71. If necessary, an n-type buffer layer 72 may be formed between the collector layer 73 and the semiconductor substrate 71. The collector layer 73 and the buffer layer 72 are formed by implanting boron and phosphorus as impurities, respectively, and performing activation annealing. For this activation annealing, the methods according to Examples 1 to 4 are applied. A collector electrode 83 is formed on the surface of the collector layer 73 after the activation annealing.

  In the first to fourth embodiments, the laser annealing for activating the impurities implanted into the semiconductor wafer has been described. The laser annealing method can be applied to a method of performing annealing by scanning with a laser beam without excessively melting the surface of the irradiation object. In this case, the temperature history can be leveled by annealing in the plane of the irradiation object.

  Although the present invention has been described with reference to the embodiments, the present invention is not limited thereto. It will be apparent to those skilled in the art that various modifications, improvements, combinations, and the like can be made.

DESCRIPTION OF SYMBOLS 10 Laser light source 20 Power adjustment apparatus 20A Blocking device 21 Control apparatus 22 Half wave plate 22A Electro-optic element 23 Beam splitter 24 Beam damper 25 Beam shaping optical system 26 Uniformation optical system 30 Folding mirror 31 Lens 35 Stage 36 Scan mechanism 40 Reference light source 41, 42 Folding mirror 43 Reference light detector 45 Beam spot 50 of reference laser beam Condensing lens 51 Black body radiation detector 60 Irradiation target 61 Beam spot 70 of annealing laser beam Silicon substrate 71 n ⁻ type region 72 Buffer layer 73 Collector layer 74 Base region 75 Emitter region 80 Emitter electrode 81 Gate electrode 82 Gate insulating film 83 Collector electrode

Claims (11)

Scanning the surface of the irradiation object with a laser beam;
Measuring a physical quantity depending on a surface temperature of the irradiation object at a position where the laser beam is incident;
And a step of lowering an energy density input to the irradiation object in the unscanned area forward in the scanning direction based on the measured physical quantity to be lower than an energy density input to the already scanned area. Manufacturing method.   In the step of reducing the energy density, when it is detected that the physical quantity exceeds a threshold value corresponding to an upper limit value of a temperature allowable range, the energy density input to the irradiation object is reduced. The manufacturing method of the semiconductor device of description. The step of measuring the physical quantity further includes:
A step of causing reference light to be incident on an incident position of the laser beam;
The method for manufacturing a semiconductor device according to claim 1, further comprising: measuring the intensity of reflected light reflected by the surface of the irradiation object by the reference light and setting it as the physical quantity. The step of measuring the physical quantity further includes:
4. The method of manufacturing a semiconductor device according to claim 1, further comprising a step of measuring an intensity of black body radiation from an incident position of the laser beam to obtain the physical quantity. 5.   5. The method of manufacturing a semiconductor device according to claim 1, wherein the step of reducing the energy density includes a step of reducing a power density of the laser beam on the surface of the irradiation object.   5. The method of manufacturing a semiconductor device according to claim 1, wherein the step of reducing the energy density includes a step of reducing a scanning speed of the laser beam on the surface of the irradiation object. In the step of scanning the surface of the irradiation object with the laser beam, the laser beam is intermittently incident on the irradiation object,
5. The method of manufacturing a semiconductor device according to claim 1, wherein the step of reducing the energy density includes a step of shortening a time during which the laser beam is continuously incident on the irradiation object. . A laser light source for emitting a laser beam;
A stage for holding an irradiation object at a position where a laser beam emitted from the laser light source is incident;
A scanning mechanism for moving one of the path of the laser beam and the irradiation object relative to the other so that the incident position of the laser beam moves on the surface of the irradiation object held on the stage;
A measuring device for measuring a physical quantity depending on a surface temperature of a position where the laser beam is incident among the irradiation object;
A laser annealing apparatus comprising: an energy density adjusting device that changes an energy density input to the irradiation object by the laser beam based on a measurement result of the physical quantity by the measuring device.   The laser annealing apparatus according to claim 8, wherein the energy density adjusting device includes a power adjustment device with a variable amount of attenuation for attenuating the power of the laser beam based on a measurement result of the physical quantity by the measuring device.   The energy density adjusting device includes a control device that controls the scanning mechanism so that a scanning speed of the laser beam on a surface of the irradiation object changes based on a measurement result of the physical quantity by the measuring device. 9. The laser annealing apparatus according to 8.   The energy density adjusting device is a laser emitted from the laser light source so that the time during which the laser beam is continuously incident on the irradiation object changes based on the measurement result of the physical quantity by the measuring device. The laser annealing apparatus according to claim 8, wherein a beam is incident on the irradiation object intermittently.