TW202046408A - Laser annealing method and laser control device capable of suppressing temperature rise of a non-irradiated surface of a semiconductor wafer and efficiently heating the irradiated surface - Google Patents

Abstract

The object of the present invention is to provide a laser annealing method that can suppress temperature rise of a non-irradiated surface of a semiconductor wafer and efficiently heat the irradiated surface. A laser pulse is periodically incident to the semiconductor wafer for annealing. At this time, the laser pulse of the next cycle is incident to a position in the cooling process following the temperature rise caused by the incidence of the just previous laser pulse.

Description

Laser annealing method and laser control device

The invention relates to a laser annealing method and a laser control device. This application claims priority based on Japanese Patent Application No. 2019-107012 filed on June 7, 2019. The entire contents of this Japanese application are incorporated in this specification by reference.

In order to recrystallize and activate semiconductor wafers such as silicon wafers doped with impurities, the semiconductor wafers need to be heated (annealed). In the manufacturing steps of insulated gate bipolar transistors (IGBT), etc., there is a step of forming circuit elements on one surface (non-irradiated surface) of a semiconductor wafer, and doping with impurities on the other surface (irradiated surface) annealing. At this time, in order to protect the formed circuit elements, it is expected to suppress the temperature rise of the non-irradiated surface.

In order to sufficiently heat the irradiated surface and suppress the temperature rise of the non-irradiated surface, laser annealing in which laser light is irradiated to the irradiated surface is used (for example, Patent Document 1, etc.). As a laser oscillator for annealing, pulsed lasers such as continuous oscillation (CW) lasers, Q-switched lasers, and excimer lasers are used. Patent Document 1 discloses a laser annealing technique, which uses a laser diode to excite an all-solid pulse laser oscillator. [Prior Technical Literature]

[Patent Document 1] Japanese Patent Application Publication No. 2011-114052

[The problem to be solved by the invention]

When the thickness of the semiconductor wafer becomes thinner, the laser irradiation conditions for suppressing the temperature rise of the non-irradiated surface of the semiconductor wafer and sufficiently increasing the temperature of the irradiation surface becomes more severe. The object of the present invention is to provide a laser annealing method and a laser control device that can suppress the temperature rise of the non-irradiated surface of a semiconductor wafer and efficiently heat the irradiated surface. [Technical means to solve the problem]

According to a viewpoint of the present invention, Provided is a laser annealing method, which is a method of annealing a semiconductor wafer by periodically making a laser pulse incident on a semiconductor wafer, and making the next period of laser pulse incident on the temperature rise caused by the incident of the immediately preceding laser pulse Position in the cooling process afterwards.

According to another viewpoint of the present invention, Provided is a laser control device, which controls a pulse laser oscillator that makes a pulsed laser beam incident on a semiconductor wafer to be annealed, and controls the pulse laser oscillator so that the repetition frequency of the laser pulse becomes 15kHz or more. [Effects of Invention]

It can effectively use the energy of the laser pulse incident on the semiconductor wafer. Thereby, the amount of energy input to the semiconductor wafer can be reduced, and as a result, the temperature rise of the non-irradiated surface can be suppressed. If the repetition frequency of the laser pulse is 15 kHz or more, it is easy to effectively use the energy of the laser pulse.

The laser annealing method and laser annealing device based on the embodiment will be described with reference to the diagrams of FIGS. 1 to 5.

Fig. 1 is a schematic diagram of a laser annealing apparatus based on an embodiment. The laser annealing device based on the embodiment includes a laser oscillator 10, a laser control device 30, a transmission optical system 40 and a chamber 50. The holding table 53 is supported in the chamber 50 by the scanning mechanism 52. The scanning mechanism 52 receives an instruction from the laser control device 30 and can move the holding table 53 in a horizontal plane. The semiconductor wafer 60 as an annealing target is held on the holding table 53. The pulsed laser beam output from the laser oscillator 10 passes through the transmission optical system 40, transmits through the laser transmission window 51 provided on the ceiling of the chamber 50, and is incident on the semiconductor wafer 60. The laser annealing apparatus based on this embodiment, for example, performs activation annealing of the dopants doped in the semiconductor wafer 60. The semiconductor wafer 60 is, for example, a silicon wafer.

The transmission optical system 40 includes, for example, a beam homogenizer, a lens, a mirror, and the like. The beam homogenizer and the lens shape the beam spot on the surface of the semiconductor wafer 60 and make the beam profile uniform.

Next, the configuration of the laser oscillator 10 will be described. As the laser oscillator 10, a fiber laser oscillator is used. One end of the gain fiber 11 doped with a laser active medium is connected to an input side optical fiber 12, and the other end is connected to an output side optical fiber 15. A high-reflectivity fiber Bragg grating (Fiber Bragg grating) 13 is formed on the input-side optical fiber 12, and a low-reflectivity fiber Bragg grating 16 is formed on the output-side optical fiber 15. The optical resonator is composed of the high reflectivity type fiber Bragg grating 13 and the low reflectivity type fiber Bragg grating 16.

The excitation light output from the laser diode 20 is guided to the gain fiber 11 through the input side fiber 12. The laser active medium doped in the gain fiber 11 is excited by the excitation light. When the laser active medium transitions to a low-energy state, stimulated emission occurs, generating laser light. The laser light generated in the gain fiber 11 is incident on the wavelength conversion element 22 through the output side fiber 15. The laser beam subjected to wavelength conversion by the wavelength conversion element 22 is incident on the semiconductor wafer 60 via the transmission optical system 40. The gain fiber 11 outputs laser light in the infrared region, for example, and the wavelength conversion element 22 converts the laser light in the infrared region into laser light in the green wavelength region.

The driver 21 drives the laser diode 20 in accordance with instructions from the laser control device 30. The command received from the laser control device 30 includes information specifying the repetition frequency of the laser pulse output from the laser diode 20. The driver 21 outputs laser light for excitation from the laser diode at the repetition frequency of the laser pulse commanded by the laser control device 30. As a result, the laser oscillator 10 outputs laser pulses at the commanded repetition frequency.

The laser control device 30 includes a console 31 operated by an operator. The operator operates the console 31 to input information specifying the repetition frequency of the laser pulse. The laser control device 30 provides the driver 21 with information specifying the repetition frequency of the input laser pulse.

According to the research of the inventor of the present application, it is found that as the laser light used for laser annealing, the pulse laser is better than the CW laser. Next, from the viewpoint of the temperature change when the laser light is incident on the semiconductor wafer 60, the reason why the pulse laser is better than the CW laser as the laser light for laser annealing will be explained. For the sake of convenience, the case where a laser pulse with a uniform power density P is incident on the semiconductor wafer 60 will be described.

在此，t為從加熱開始的經過時間，C為半導體晶圓60的比熱，ρ為半導體晶圓60的密度，λ為半導體晶圓60的導熱率。例如，表面溫度T的單位為“K”，功率密度P的單位為“W/cm² ”，經過時間t的單位為“秒”，比熱C的單位為“J/g·K”，密度ρ的單位為“g/cm³ ”，導熱率λ的單位為“W/cm·K”。The surface temperature T of the surface (irradiated surface) of the semiconductor wafer 60 can be expressed by the following equation. [Numerical formula 1]

Here, t is the elapsed time from the start of heating, C is the specific heat of the semiconductor wafer 60, ρ is the density of the semiconductor wafer 60, and λ is the thermal conductivity of the semiconductor wafer 60. For example, the unit of surface temperature T is "K", the unit of power density P is "W/cm ² ", the unit of elapsed time t is "second", the unit of specific heat C is "J/g·K", and the density ρ The unit of is "g/cm ³ ", and the unit of thermal conductivity λ is "W/cm·K".

能量密度E的單位例如為“J/cm² ”。The energy density E applied to the semiconductor wafer 60 during the period from the start of heating (the rising point of the laser pulse) to the elapsed time t ₀ can be expressed by the following equation. [Numerical formula 2]

The unit of energy density E is "J/cm ² ", for example.

若從式(2)和式(3)刪除經過時間t₀ ，則可獲得下式。 [數式4]

It can be seen from equation (1) that in order for the surface temperature of the semiconductor wafer 60 to reach Ta when the elapsed time from the start of heating is t ₀ , the following conditions must be satisfied. [Numerical formula 3]

If the elapsed time t ₀ is deleted from equations (2) and (3), the following equation can be obtained. [Equation 4]

In order to suppress the temperature rise of the back surface of the semiconductor wafer 60 under the condition of increasing the surface temperature to Ta and perform efficient annealing, it is preferable to reduce the energy density E input to the semiconductor wafer 60. The parameters in the parentheses on the right side of the formula (4) are constants. Therefore, it can be seen that in order to reduce the energy density E, it is sufficient to increase the power density P. The power density level required for actual use is several MW/cm ² or more.

In order to achieve a power density of several MW/cm ² or more with a CW laser, for example, the beam spot of a laser beam with a power of several tens of W has to be reduced to an area of approximately 1×10 ³ μm ² . Technically, this is not easy.

The peak power of a pulse laser is defined by the value obtained by dividing the energy per pulse (hereinafter referred to as pulse energy) by the pulse width. Furthermore, the average power of a pulsed laser is defined by the product of pulse energy and pulse repetition frequency. The power density P of formula (4) is equivalent to the value obtained by dividing the peak power by the area of the beam spot.

The pulse energy of Q-switch lasers and excimer lasers usually used for annealing is tens of mJ, and the pulse width is about 100ns. Therefore, the peak power is in the order of hundreds of kW. In this way, a sufficiently large peak power can be achieved compared with the power of a CW oscillating laser. Therefore, it can be seen that as a laser light source for annealing to increase the power density P, pulsed lasers are more suitable than CW oscillation lasers.

Next, the cooling process of the semiconductor wafer 60 after laser pulse irradiation will be described. If t represents the elapsed time from the rising point of the laser pulse, t ₀ represents the pulse width, and Ta represents the surface temperature at the falling point of the laser pulse (elapsed time t = t ₀ ), the surface temperature of the semiconductor wafer 60 T is represented by the following formula. [Equation 5]

The repetition frequency of the laser pulse of the pulse laser generally used for annealing of semiconductor wafers is in the order of kHz, so the time interval of the laser pulse is in the order of ms. If the pulse width t ₀ of the pulse laser used for annealing is set to the general pulse width of Q-switched lasers and excimer lasers, that is, 100ns, the rise of a certain laser pulse to that of the next cycle of laser pulse The time interval to the rise is set to 1 ms, and the surface temperature T _n of the semiconductor wafer 60 at the time when the laser pulse of the next cycle rises immediately after the incident laser pulse is approximated by the following equation. [Equation 6]

It can be seen from equation (6) that when the laser pulse is incident, the temperature rise due to the incident laser pulse immediately before decreases to almost the same temperature as the temperature immediately before the incident laser pulse.

Fig. 2 is a graph showing the calculated value of the time change of the surface temperature when one laser pulse is incident on the silicon wafer. On the horizontal axis, the elapsed time t from the rising point of the laser pulse is represented by the unit "ns", the surface temperature T of the semiconductor wafer 60 is represented by the unit "°C" on the left vertical axis, and the surface temperature T of the semiconductor wafer 60 is represented by the unit "℃" on the right vertical axis. MW/cm ² "represents the power density of laser light. The broken line in the graph shows the time change of the power density of the laser pulse, and the solid line shows the time change of the surface temperature T of the semiconductor wafer 60. The pulse width of the laser pulse is t ₀ and the peak power density is 5 MW/cm ² .

During the laser pulse incidence period (0≤t≤t ₀ ), the surface temperature T rises according to formula (1). The surface temperature T at the time point (t=t ₀ ) when the time corresponding to the pulse width t _{0 has} elapsed from the rising time point of the laser pulse is equal to the highest reached temperature Ta. After the laser pulse drops (t≥t ₀ ), the surface temperature T decreases according to equation (5).

Next, with reference to FIGS. 3A and 3B, the temporal change of the surface temperature of the semiconductor wafer 60 when the activation annealing is actually performed will be described. During activation annealing, a laser pulse is incident on the surface of the semiconductor wafer 60 periodically, that is, at a constant period, and the beam spot of the laser beam is moved, thereby performing annealing of a desired area. The moving speed of the beam spot is set in such a way that the beam spot of the current laser pulse partially overlaps the beam spot of the immediately preceding laser pulse.

FIG. 3A is a diagram showing the movement of the beam spot BS on the surface of the semiconductor wafer 60. FIG. 3A shows the beam spots BS ₁ , BS ₂ , and BS ₃ of the laser pulses LP ₁ , LP ₂ , and LP ₃ incident on the specific point P from the first to the _third . Each beam spot BS has a shape elongated in one direction (the longitudinal direction in FIG. 3A). During the annealing, the beam spot BS is moved relative to the semiconductor wafer 60 in the width direction orthogonal to the length direction. The distance that the beam spot BS moves during one period from the point of incidence of the immediately preceding laser pulse to the point of incidence of the next laser pulse is, for example, 1/3 of the width of the beam spot BS. That is, the overlap rate is about 67%.

At this time, if one point P on the surface of the semiconductor wafer 60 is focused, the point P is included in the three beam spots BS ₁ , BS _2, and BS ₃ . That is, during the annealing, three laser pulses LP ₁ , LP ₂ , and LP ₃ will be incident on one point P. The pulse energy of the three laser pulses LP ₁ , LP ₂ and LP ₃ are the same.

FIG. 3B is a graph showing the time change of the surface temperature at the position of the point P. The horizontal axis represents the elapsed time, and the vertical axis represents the surface temperature at the point P. During the period when the laser pulses LP ₁ , LP ₂ , and LP _{3 are} incident, the surface temperature rises, and during the period when the laser pulses are not incident, the surface temperature decreases. The surface temperature at the point P immediately before the incident laser pulse LP ₁ is denoted as T ₀ , and the highest reached temperature reached by the incident laser pulse LP ₁ is denoted as Ta ₁ .

In the embodiment, the laser pulse LP _{2 of} the next cycle is incident on the position (the position of the point P) in the cooling process after the temperature rise caused by the incident of the laser pulse LP ₁ immediately before. That is, before the surface temperature drops to the surface temperature T ₀ before the laser pulse LP _{1 is} incident, the next laser pulse LP _{2 is} incident on the position of the point P. The surface temperature at the point P immediately before the laser pulse LP ₂ is incident is denoted as T ₁ . The surface temperature of the incident point of the laser pulse LP ₂ than T ₁ of the incident laser pulse LP time ₁ surface temperature T _1, so _{2 2} is higher than the maximum reached temperature Ta reaches the laser pulse LP from the immediately preceding Ray shot pulse LP ₁ reaches the maximum reaching temperature Ta _1. Therefore, the surface temperature T ₂ at the point P immediately before the incident laser pulse LP ₃ is higher than the surface temperature T ₁ . Thus, to the laser pulse sent by the third maximum temperature reached the LP ₃ Ta ₃ is higher than the maximum temperature reached Ta _2.

The increase in surface temperature caused by the incidence of the laser pulse is expressed as ΔTu, and the decrease in surface temperature until the incident of the laser pulse in the next cycle is expressed as ΔTd. The temperature retention rate Tr is defined by the following formula. [Equation 7]

Fig. 4 is a graph showing the calculation result of the relationship between the pulse repetition frequency and the temperature retention rate Tr. On the horizontal axis, the pulse repetition frequency is represented by the unit "kHz", and on the vertical axis, the temperature retention rate Tr is represented by the unit "%". The solid line and the dashed line in the graph of FIG. 4 show the calculation results when the pulse width is 10 ns and when the pulse width is 100 ns, respectively.

Under the condition of constant pulse width, as the pulse repetition frequency increases, the temperature retention rate Tr also increases. Under the condition that the pulse repetition frequency is constant, the longer the pulse width, the higher the temperature retention rate Tr. This means that the higher the pulse repetition frequency or the longer the pulse width, the shorter the non-irradiation time interval between the rise of the next laser pulse after the fall of the laser pulse, and the further the reduction of the surface temperature ΔTd. .

Next, the excellent effects of the above-mentioned embodiment will be described. In the embodiment, the laser pulse of the next cycle is incident on a position in the cooling process after the temperature rise caused by the incident of the laser pulse immediately before. For example, as shown in Figure 3B, the second laser pulse is sent to the LP ₂ enters in the cooling process after the incident laser pulse sent by the first of the LP ₁ caused a temperature rise in the position (the position of the point P). In addition, the third laser pulse LP _{3 is} incident on a position (the position of the point P) during the cooling process after the temperature rise caused by the incident of the second laser pulse LP ₂ . Therefore, a part of the energy input by the immediately preceding laser pulse overlaps the energy input by the laser pulse into the semiconductor wafer, and the semiconductor wafer is heated. In this way, by overlapping the energy of a plurality of laser pulses input to the semiconductor wafer, the same degree of thermal effect can be obtained as the pulse energy of one laser pulse is substantially increased, and the laser energy can be effectively used.

In addition, the pulse energy required to heat the semiconductor wafer to the target temperature can be reduced. As an example, in order to reach the maximum temperature reaches the laser pulse sent by the third LP ₃ Ta ₃ to the target value of the heating will be described. When the repetition frequency of the laser pulse is so low that the effect of energy overlap by the laser pulse input cannot be obtained (for example, in the case of about 1kHz), the semiconductor wafer has to be transferred with one laser pulse The surface is heated to the target value. That is, the highest reached temperature Ta ₁ of the first laser pulse LP ₁ has to be increased to the target heating value.

In contrast to this, in the embodiment, even when the highest reached temperature Ta _{1 of the} first laser pulse LP ₁ is lower than the heating target value, the highest reached temperature of the third laser pulse LP ₃ can be set Ta ₃ becomes more than the heating target value. Therefore, the pulse energy of each of the laser pulses LP ₁ , LP ₂ , and LP ₃ can be reduced.

By reducing the pulse energy of each laser pulse, the total amount of energy invested in the semiconductor wafer is reduced. Reducing the total amount of input energy helps suppress the temperature rise of the non-irradiated surface of the semiconductor wafer. Therefore, annealing can be performed under the condition that the temperature rise of the non-irradiated surface of the semiconductor wafer can be suppressed.

The temperature rise of the non-irradiated surface of the semiconductor wafer can be suppressed, so restrictions on the thickness of the semiconductor wafer and the materials that can be used are alleviated.

In the past, the laser pulse of the next cycle was incident in a state where the thermal influence of the immediately preceding laser pulse was not substantially retained. For example, as shown in formula (6), annealing is performed under the condition that the temperature retention Tr is 0.5% or less. In order to fully obtain the effect of making the laser pulse of the next cycle incident on the position in the cooling process after the temperature rise caused by the incident of the laser pulse immediately before, annealing is performed under the condition that the temperature retention Tr is 1% or more For better. That is, it is preferable to make the laser pulse of the next cycle incident before the point of time when the temperature amplitude corresponding to 99% of the increase in the surface temperature of the semiconductor wafer caused by the incident laser pulse immediately before is reduced. In order to easily satisfy this condition, the pulse repetition frequency is preferably set to 15 kHz or more. By setting the time from the incidence of the immediately preceding laser pulse to the incidence of the next cycle of the laser pulse or the pulse repetition frequency as described above, it is possible to sufficiently suppress the non-irradiation of the semiconductor wafer compared with the conventional method. The effect of the temperature rise of the noodles.

Next, referring to FIG. 5, the results of the evaluation experiment of actually performing annealing of the silicon wafer by changing the pulse repetition frequency will be described.

Figure 5 is a graph showing the relationship between the relative pulse energy density of the pulsed laser beam used for annealing and the relative melting depth of the silicon wafer. On the horizontal axis, the relative pulse energy density is represented by arbitrary units, and on the vertical axis, the relative melting depth is represented by arbitrary units.

The circular mark in the graph shows the experimental result when the pulse repetition frequency is 1 kHz, and the triangular mark shows the experimental result when the pulse repetition frequency is 150 kHz. In the evaluation experiment when the pulse repetition frequency is 1kHz, the 2 times higher harmonic wave (wavelength 532nm) of the all-solid laser was used, and the pulse width was set to 100ns. In the evaluation experiment when the pulse repetition frequency is 150kHz, the 2 times higher harmonic wave (wavelength 530nm) of the fiber laser was used, and the pulse width was set to 10ns. Make the average power of the two pulse laser beams almost the same. The overlap ratio during annealing of both is set to 67%.

As the pulse energy density increases, the melting depth becomes deeper. In addition, it can be seen that if the pulse repetition frequency is increased, the pulse energy density required to obtain the same melting depth decreases. In other words, if the pulse energy density is the same, even if the average power is constant, the higher the pulse repetition frequency, the deeper the melting depth. This is because the more the pulse repetition frequency is increased, the more effectively the energy of the laser light can be used.

From the viewpoint of effectively using the energy of the laser light, the next step is to reduce the temperature range corresponding to 95% of the increase in the surface temperature of the semiconductor wafer caused by the incident laser pulse immediately before Periodic laser pulse incidence is preferable. In order to easily satisfy this condition, the pulse repetition frequency is preferably 100 kHz or more. By setting the time from the incidence of the immediately preceding laser pulse to the incidence of the next cycle of the laser pulse or the pulse repetition frequency as described above, compared with the conventional method, it is possible to fully utilize the laser light. The effect of energy.

Next, a modification of the above-mentioned embodiment will be described. In the above-mentioned embodiment, the fiber laser oscillator is used as the laser oscillator, but other pulse laser oscillators, such as mode-locked laser oscillators, which can easily achieve the repetition frequency of pulses above 15kHz, can also be used. .

In the above-mentioned embodiment, by moving the beam spot on the surface of the semiconductor wafer 60, annealing of the desired area is performed. However, the position of the beam spot can also be fixed relative to the semiconductor wafer so that the required number of mines can be emitted. The radio pulse is incident. In the case where the maximum reached temperature of the semiconductor wafer reached by the incidence of multiple laser pulses is to be the same, it is only necessary to make the pulse energy of the second laser pulse lower than the pulse energy of the first laser pulse. The thermal effect caused by the incidence of the first laser pulse will remain at the point of incidence of the second laser pulse. Therefore, even if the pulse energy of the second laser pulse is reduced, it can be heated to the same level as the first laser pulse. The highest temperature reached the same degree.

The above-mentioned embodiments are examples, and the present invention is not limited by the above-mentioned embodiments. For example, relevant technical personnel should be able to make various changes, improvements, combinations, etc.

10: Laser oscillator 11: gain fiber 12: Input side fiber 13: Fiber Bragg Grating 15: Output side fiber 16: Fiber Bragg Grating 20: Laser diode 21: drive 22: Wavelength conversion element 30: Laser control device 31: console 40: Transmission optical system 50: chamber 51: Laser transmission window 52: Scanning mechanism 53: hold the stage 60: Semiconductor wafer

[Fig. 1] A schematic diagram of the laser annealing apparatus based on the embodiment. [Fig. 2] A graph showing the calculated value of the time change of the surface temperature when one laser pulse is incident on the silicon wafer. [FIG. 3A] is a graph showing the movement of the beam spot on the surface of the semiconductor wafer 60, and [FIG. 3B] is a graph showing the time change of the surface temperature at the point P. [Figure 4] is a graph showing the calculation result of the relationship between the pulse repetition frequency and the temperature retention Tr. [Figure 5] is a graph showing the relationship between the relative pulse energy density of the pulsed laser beam used for annealing and the relative melting depth of the silicon wafer.

BS: beam spot

BS ₁ : beam spot

BS ₂ : beam spot

BS ₃ : beam spot

LP ₁ : The first laser pulse

LP ₂ : The second laser pulse

LP ₃ : The third laser pulse

P: point

T ₀ : Elapsed time

T ₁ : Elapsed time

T ₂ : Elapsed time

Ta ₁ : Maximum reach temperature

Ta ₂ : Maximum reach temperature

Ta ₃ : Maximum reach temperature

△Td: Decrease range of surface temperature

△Tu: The increase in surface temperature

Claims (8)

A laser annealing method is a method of annealing a semiconductor wafer by periodically causing a laser pulse to be incident on the semiconductor wafer. The laser pulse of the next cycle is incident after the temperature rise caused by the incident laser pulse immediately before The position in the cooling process. The laser annealing method according to claim 1, wherein before the time point when the temperature range corresponding to 99% of the increase range of the surface temperature of the semiconductor wafer caused by the incident of the immediately preceding laser pulse is reduced, the lower One cycle of laser pulse is incident. The laser annealing method according to claim 2, wherein, before the temperature range corresponding to 95% of the increase range of the surface temperature of the semiconductor wafer caused by the incident of the immediately preceding laser pulse is reduced, the lower One cycle of laser pulse is incident. The laser annealing method according to any one of claim 1 to claim 3, wherein the repetition frequency of the laser pulse is 15 kHz or more. The laser annealing method according to claim 4, wherein the repetition frequency of the laser pulse is 100 kHz or more. A laser control device controls a pulse laser oscillator that makes a pulsed laser beam incident on a semiconductor wafer to be annealed, and controls the pulse laser oscillator in such a way that the repetition frequency of the laser pulse becomes 15kHz or more. The laser control device according to claim 6 controls the pulse laser oscillator so that the repetition frequency of the laser pulse becomes 100 kHz or more. The laser control device according to claim 6 or 7, wherein the pulsed laser oscillator is a fiber laser oscillator or a mode-locked laser oscillator.

Images