JP2005353979A - Manufacturing method of semiconductor thin film - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enlarge the single crystallization region of a crystalline thin film and to improve its productivity in a method for manufacturing the crystalline thin film by irradiating a precursor semiconductor thin film with a main beam and a sub beam. <P>SOLUTION: The manufacturing method of the crystalline thin film by irradiating the precursor semiconductor thin film with the main beam and the sub beam in superposition in time and/or in space comprises a first process for casting the sub beam at least once before casting the main beam and the sub beam in superposition, and a second process for casting the main beam and the sub beam in superposition. In the second process, the main beam is cast just before finishing of irradiation of the sub beam. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a method for manufacturing a semiconductor thin film. More specifically, the present invention relates to a method for manufacturing a semiconductor thin film having a polycrystalline semiconductor region formed by melting and recrystallization of a precursor semiconductor thin film by laser light irradiation.

  In a thin film transistor (TFT) used in a display device using liquid crystal or electroluminescence (EL), an amorphous or crystalline silicon film is often used as an active layer. In this case, a crystalline silicon thin film such as a polycrystal or a single crystal has a higher electron mobility than an amorphous silicon thin film, and thus can exhibit many advantages.

  For example, when a crystalline silicon TFT is used, not only can a switching element be formed in a pixel portion of a display device, but also a drive circuit and a peripheral circuit can be formed in a peripheral portion of the pixel. They can be formed on the same substrate. That is, since it is not necessary to mount a separate driver IC or drive circuit board on the display device, it is possible to provide these display devices at a low price.

  In addition, if a crystalline silicon thin film is used, the size of the TFT can be reduced, so that the switching element formed in the pixel portion can be reduced and the aperture ratio of the display device can be increased. As a result, it is possible to provide a display device with high brightness and high definition.

  As described above, a TFT using a crystalline silicon thin film has many advantages over a TFT using an amorphous silicon thin film.

  By the way, as a method for producing a crystalline silicon thin film such as a polycrystal or a single crystal, many recrystallization techniques using laser light have been proposed.

  The earliest proposed method is to scan continuously emitted laser light to melt and recrystallize the precursor semiconductor thin film on the substrate, and is generally referred to as ZMR (Zone Melting Recrystallization). . In recent years, a technique for polycrystallizing an amorphous silicon thin film at a low substrate temperature of 600 ° C. or less using an excimer laser has been generalized, and a display device including a polycrystalline silicon TFT formed on a low-cost glass substrate Can be offered at a low price.

  In the crystallization technique using an excimer laser, a glass substrate on which an amorphous silicon thin film is deposited is heated to about 400 ° C., and the glass substrate is moved at a constant speed, and is 200 to 400 mm in length and 0.2 to 1 in width. An excimer laser beam having a linear cross section of about 0.0 mm is pulsed onto the amorphous silicon thin film on the glass substrate. By this method, a polycrystalline silicon thin film having an average crystal grain size comparable to the thickness of the amorphous silicon thin film can be formed.

  Thus, the recrystallization technique using an excimer laser is generally referred to as an ELA (Excimer Laser Annealing) method and is industrially used as a laser crystallization technique with excellent productivity. However, silicon crystal grains formed by the ELA method are generally small with a diameter of several tens of nanometers, and it is difficult to form a high performance TFT on such a polycrystalline silicon thin film with a small particle diameter.

In order to improve this problem, various methods for expanding the single crystallization region by irradiating the main beam and the sub beam to increase the crystal grain size are disclosed in Patent Document 1, Patent Document 2, Patent Document 3, and the like. Proposed.
Japanese Patent No. 2709376 Japanese Patent Publication No. 64-12088 Japanese Patent No. 3221149

  In Patent Document 1, an excimer laser is used for both the main laser (second light beam of Patent Document 1) and the sub laser (first light beam of Patent Document 1), and the main laser and the sub laser are irradiated simultaneously. In addition, there is a description that the main beam is moved and irradiated in the sub beam irradiation region.

  In Patent Document 2, a ruby laser is used as the main laser (second laser in Patent Document 2), and a carbon dioxide gas laser is used as the sub laser (first laser in Patent Document 2), and the main laser and the sub laser are irradiated simultaneously. In addition, there is a description that the main beam is irradiated during the sub beam irradiation.

  In Patent Document 3, an excimer laser or a YAG laser is used as a main laser (second laser in Patent Document 3), and a YAG laser or a carbon dioxide gas laser is used as a sub laser (first laser in Patent Document 3). There is a description that the main laser is irradiated after a delay time τ after completion of the laser irradiation.

As a result of repeatedly conducting detailed experiments on a method of expanding a single crystallized region by irradiating a main beam and a sub beam as described in Patent Documents 1-3, the following points were found. It was revealed.
(1) The single crystallized area differs greatly depending on the irradiation timing of the main laser and the sub laser. If the main laser is irradiated immediately before the completion of the sub laser irradiation, the single crystallized region is maximized.
(2) The precursor semiconductor thin film generally has a very high reflectivity with respect to laser light when melted. For example, in the case of silicon, the reflectivity of amorphous silicon with respect to a carbon dioxide gas laser is about 15%, whereas molten silicon. The reflectance is about 80%.
(3) By performing preheating with a sub beam before irradiating the main laser, and further irradiating with the main laser and the sub laser in combination, the single crystallized region can be greatly expanded. Alternatively, the energy efficiency of the sub beam can be apparently increased by the preliminary heating, so that it is possible to irradiate the sub beam extended over a wide area.

  Compared with the above experimental results (1) to (3) by the present inventors, in Patent Document 1, the main beam is moved within the spatial intensity distribution of the main beam and the sub beam and the irradiation region of the sub beam. However, the concept of the most important irradiation time is not included in the laser crystallization method using the main beam and the sub beam. Therefore, in Patent Document 1, the expansion of the single crystallization region is limited. In Patent Document 2, although it is described that the main beam is irradiated during the irradiation of the sub beam, the above-described decrease in irradiation efficiency (absorption rate) due to the increase in reflectance is not taken into consideration. Therefore, also in Patent Document 2, the expansion of the single crystallization region is limited. In Patent Document 3, since the main laser light is irradiated after a time after the completion of the sub beam irradiation, the effect of expanding the single crystallized region is extremely small. Further, in the prior art, various problems such as no proposal from an industrially very important viewpoint such as expanding the area that can be processed by one pulse irradiation in consideration of the energy efficiency of the sub-laser. Is present.

  In view of the current state of the prior art, the main object of the present invention is to manufacture a semiconductor thin film that expands a single crystallization region and is excellent in industrial productivity in a laser crystallization method using a main laser and a sub laser. Is to provide a method.

  In order to solve the above-mentioned problem, the present inventors basically improved the method of expanding the single crystallized region by using the main laser and the sub laser together, further expanding the single crystallized region, and Under the policy of finding a production method with excellent productivity, we have intensively researched and developed to find such a laser recrystallization process.

  As a result, the present inventors irradiate the precursor semiconductor thin film with a main pulse laser absorbed in the precursor semiconductor thin film and a sub-pulse laser absorbed in a layer under the precursor semiconductor thin film, thereby forming a polycrystalline semiconductor thin film. In addition to irradiating the sub-pulse laser at least once prior to the main laser light irradiation, the single laser region can be enlarged by irradiating the main laser light substantially simultaneously with the completion of the sub-laser light irradiation. The present invention was completed by finding that a polycrystalline semiconductor thin film can be formed with high productivity.

  That is, according to one aspect of the present invention, a precursor semiconductor is formed by superimposing a main pulse laser absorbed in a precursor semiconductor thin film and a sub-pulse laser absorbed in a lower layer of the precursor semiconductor thin film in time and space. A method of manufacturing a polycrystalline semiconductor thin film by irradiating a thin film, characterized in that a sub-pulse laser is irradiated at least once prior to main laser light irradiation.

  According to another aspect of the present invention, a method for producing a polycrystalline semiconductor thin film by irradiating a precursor semiconductor thin film with a main pulse laser and a sub-pulse laser while relatively moving the precursor semiconductor thin film and the laser beam. The main pulse laser is characterized in that the region irradiated with the sub-pulse laser is irradiated at least once.

Note that the main pulse laser is preferably irradiated immediately before the completion of the sub-pulse laser irradiation. That is, when irradiating the main pulse laser and the sub pulse laser in an overlapping manner, the delay time from the oscillation trigger pulse of the main laser to the start of main laser radiation is DL1, the fluctuation of DL1 is J1, and the sub laser oscillation trigger pulse is When the delay time until the start of laser emission is DL2, the fluctuation of DL2 is J2, and the oscillation trigger delay time of the main laser for the sub laser is Ts, the time Tr from the start of sub laser pulse irradiation to the start of main laser pulse emission is DL1-DL2- (J1 + J2) + Ts <Tr <DL1-DL2 + (J1 + J2) + Ts
It is preferable to satisfy.

  The main pulse laser has a wavelength absorbed by the precursor semiconductor thin film and energy capable of melting the precursor semiconductor thin film and the polycrystalline semiconductor thin film, and the sub-pulse laser has at least one of the precursor semiconductor thin film and its lower layer. It is preferable to have a wavelength that can be heated. Further, the main pulse laser preferably has a wavelength in the range from the ultraviolet region to the visible region, and the sub-pulse laser preferably has a wavelength in the range from the infrared region to the far infrared region.

  According to the present invention, in the laser crystallization method using the main laser and the sub laser, it is possible to provide a method for manufacturing a semiconductor thin film that expands a single crystallization region and is excellent in industrial productivity.

<Outline of Manufacturing Method of Crystalline Semiconductor Thin Film According to the Present Invention>
As shown in the schematic cross-sectional view of FIG. 2, in the method for producing a semiconductor thin film according to the present invention, a precursor semiconductor thin film 23 (hereinafter referred to as “substrate composite”) deposited on an underlayer 22 on a substrate 21 is combined. The precursor semiconductor thin film 23 is melted and recrystallized by irradiating the main laser beam 24 and the sub-laser beam 25. The main laser beam 24 has a wavelength that is absorbed by the precursor semiconductor thin film 23 and is used to melt the precursor semiconductor thin film 23. The sub laser 25 has a wavelength that is absorbed by the lower layer 22 of the precursor semiconductor thin film 23, preheats the precursor semiconductor thin film 23, extends the time from the molten state of the precursor semiconductor thin film 23 to solidification, It is used for expanding the single crystallized region of the formed polycrystalline semiconductor film.

  FIG. 1 is a timing chart for explaining an irradiation mode of the main laser pulse 11 and the sub laser pulse 12 according to the present invention. In the present invention, with respect to one irradiation of the main laser pulse, the auxiliary laser pulse is irradiated twice or more. In one embodiment of the present invention shown in FIG. 1, a single step of irradiating the main laser pulse 11 has a series of steps of irradiating the sub laser pulse 12 three times, and the polycrystalline semiconductor is formed in a desired region by repeating the steps. A thin film can be formed. Although the main laser irradiation is performed in synchronization with the sub laser irradiation, in FIG. It has a series of steps including a second step of irradiating the precursor semiconductor thin film 23 with the laser pulse 12 superimposed temporally and spatially.

  By irradiating the sub laser pulse 12 prior to the irradiation of the main laser pulse 11, the lower layer 22 of the precursor semiconductor thin film 23 is heated, and the precursor semiconductor thin film 23 is indirectly preheated to form a multi-layer formed. It becomes possible to expand the single crystallization region of the crystalline semiconductor thin film. Further, by performing the irradiation of the sub laser pulse 12 and the irradiation of the main laser pulse 11 in time, it becomes possible to further promote the preheating effect and until the molten precursor semiconductor thin film is completely solidified. Since the time of 1 is delayed, the single crystallized region can be further expanded.

  As a result of careful examination of a method for expanding the single crystallized region by controlling the irradiation timing of the main laser and the sub laser in the second step, the present inventors have determined that the time immediately before the completion of the sub laser irradiation (to be precise, It was found that the single crystallized region can be expanded most by controlling so that the output of the main laser is maximized at the moment when the output of the sub-laser starts to drop rapidly. The graph of FIG. 5 shows an example of the change in the specific intensity of the sub laser pulse with time and the change in the crystal grain size with the main laser irradiation time. In FIG. 5A, the sub-laser specific intensity starts to rapidly decrease at a time of 500 μs immediately before the sub-laser irradiation end time (600 μs). This time corresponds to the time when the high voltage applied to the laser oscillator is turned off. FIG. 5B shows that the crystal grain size becomes maximum when the main laser irradiation time is 500 μs. In this experiment, the pulse width (full width at half maximum) of the main laser is about 15 ns, which is sufficiently smaller than the pulse width of the sub laser, so that the output of the sub laser starts in a moment. You can safely assume that the pulse irradiation is complete.

Next, a general laser oscillation control method will be described. The following two methods are generally used as a method for controlling the laser light oscillation time, repetition frequency, number of repetitions, and the like.
(A) Oscillation trigger pulse controls oscillation time, repetition frequency / number of repetitions, etc. (peak) output and pulse width are uniquely dependent on the characteristics of the high voltage power supply circuit of the oscillator and other characteristics of the oscillator How to be decided.
(B) In addition to controlling the oscillation time, repetition frequency / number of repetitions, etc. by the oscillation trigger pulse, a method for controlling the (peak) output and the pulse width (the output and pulse width are controlled by the high voltage of the oscillator) Achieved by defining the voltage and application time of the power supply with a trigger pulse).

The graph of FIG. 3 illustrates the laser oscillation trigger pulse (upper stage) and the temporal change (lower stage) of the laser pulse intensity, and shows an example of the laser oscillation control method (b). In general, laser oscillation starts rising after a certain delay time (t ₃ -t ₁ ) has elapsed since the oscillation trigger pulse was input to the high voltage power supply circuit of the laser oscillator (time t ₁ ) ( At time t ₃ ), at the end of the high voltage application time (t ₂ −t ₁ ), the output starts to fall (time t ₄ ). Since the high voltage application time is from the start of laser rise to the start of laser fall, t ₂ −t ₁ = t ₄ −t ₃ . In the present invention, the time during which the high voltage power source is applied is defined as the pulse width of the laser.

  The timing chart of FIG. 4 illustrates the main laser and sub laser irradiation control method according to the present invention. 4A shows the oscillation trigger pulse input to the sub laser oscillator, FIG. 4B shows the temporal intensity change of the sub laser pulse, and FIG. 4C shows the oscillation trigger pulse input to the main laser oscillator. And FIG. 4 (d) shows the temporal intensity change of the main laser pulse. In the present embodiment, a case where an excimer laser is used as the main laser and a carbon dioxide gas laser is used as the sub laser will be described.

  In general, the excimer laser, as in the above-described oscillation control method (a), “controls oscillation time, repetition frequency / number of repetitions by oscillation trigger pulse, and (peak) output and pulse width are the high voltage power supply circuit of the oscillator. The laser oscillates by a method uniquely determined depending on the characteristics of the oscillator and other characteristics of the oscillator, and the pulse width is about 50 to 200 ns. In general, a carbon dioxide laser has a (peak) output and a pulse width in addition to controlling the oscillation time, repetition frequency / number of repetitions, etc. by the oscillation trigger pulse as in the oscillation control method (b) described above. Laser oscillation is performed by a “control method”, and in this embodiment, oscillation from pulse width 1 μs to continuous oscillation is used.

  An excimer laser widely used industrially for melting a precursor semiconductor thin film has a sufficient peak intensity and pulse width, and its oscillation wavelength is in the ultraviolet region, and is therefore a typical precursor semiconductor thin film. Absorbed by the amorphous silicon thin film. For these reasons, an excimer laser can be preferably used also in the present invention.

  Carbon dioxide laser, which is widely used industrially to heat the lower layer of the precursor semiconductor thin film, has a far-infrared wavelength and is therefore a representative material that transmits the amorphous silicon thin film and forms the lower layer. It has the characteristic of being absorbed by silicon oxide or glass. Further, according to the experiments by the present inventors, the irradiation condition of the carbon dioxide laser for effectively expanding the single crystallized region needs to increase the laser pulse width to at least several μs or more. For these reasons, the carbon dioxide laser can be preferably used also in the present invention.

  From (a) to (d) in FIG. 4, the following equation (1) is established.

DL ₂ + Tr = Ts + DL ₁ + δτ (1)
However, δτ is the time required to reach the maximum peak intensity from the start of irradiation of the main laser pulse. By transforming this equation (1), the following equation (2) is obtained.

Tr−Ts = (DL ₁ −DL ₂ ) + δτ (2)
When the pulse width of the main laser is sufficiently smaller than the pulse width of the sub laser, it can be regarded as δτ = 0 (it is appropriate to think in this way from the relationship of the pulse width described above). In addition, the single crystallized region can be expanded most by controlling the main laser output to be maximized at the time immediately before the completion of the secondary laser irradiation (more precisely, the moment when the secondary laser output starts to drop sharply). The following equation (3) is established under these conditions.

Tr-Ts = DL ₁ -DL ₂ (3)
In consideration of fluctuations (jitter (± J ₁ , ± J ₂ )) of the time (DL1, DL2) required from the trigger input to the start of laser rise, the following equation (4) is obtained from FIGS. And (5) are obtained.

DL ₂ + J ₂ + Tr _min = Ts + DL ₁ −J ₁ (4)
DL ₂ −J ₂ + Tr _max = Ts + DL ₁ + J ₁ (5)
By transforming these equations (4) and (5), the following equations (6) and (7) are obtained.

Tr _min = (DL ₁ −DL ₂ ) − (J ₁ + J ₂ ) + Ts (6)
Tr _max = (DL ₁ −DL ₂ ) + (J ₁ + J ₂ ) + Ts (7)
Here, since Tr _min <Tr <Tr _max , the following inequality (8) is obtained.
(DL ₁ −DL ₂ ) − (J ₁ + J ₂ ) + Ts <Tr <(DL ₁ −DL ₂ ) + (J ₁ + J ₂ ) + Ts (8)
That is, in order to enlarge the single crystallized region as much as possible, the irradiation timing of the main laser and the sub laser may be determined under the condition satisfying the inequality (8).

<Details of precursor semiconductor thin film>
The schematic perspective view of FIG. 8 shows a substrate composite used in the method for manufacturing a semiconductor thin film of the present invention, and a material containing glass, quartz, or the like can be suitably used as the insulating substrate 801. Among these materials, it is desirable to use glass because it is inexpensive and can easily produce a large-area insulating substrate. There is no particular limitation on the thickness of the substrate, but if it is thin, the substrate is lighter, but it has the disadvantage of being easily broken. As the insulating substrate of the present embodiment, a glass substrate having a thickness of 0.7 mm that is currently most commonly used in industry can be used.

  In the substrate composite, a buffer layer 802 (corresponding to the above-described base layer) is preferably formed. This is because the buffer layer acts so that the thermal effect of the precursor semiconductor thin film 803 melted by laser light irradiation does not reach the glass substrate 801, and further prevents impurity diffusion from the glass substrate to the precursor semiconductor thin film. It is because it can act. There is no particular limitation on the material of the buffer layer, and an insulating film such as silicon oxide or silicon nitride is preferably employed industrially. The thickness of the buffer layer is preferably reduced so as not to impair the effect of suppressing thermal influence on the substrate and preventing impurity diffusion from the viewpoint of industrially efficient production by shortening the film formation time as much as possible. Empirically, a thickness of 100 to 500 nm is suitable. As a film forming method, a CVD (chemical vapor deposition) method, a sputtering method, a PVD (physical vapor deposition) method or the like can be employed. Among these, the CVD method capable of forming the film at the highest speed is preferably used industrially. In the present embodiment, a 200 nm thick silicon oxide layer formed by a CVD method can be used as the buffer layer.

  Further, in the substrate composite, the material of the precursor semiconductor thin film 803 is not particularly limited as long as it is an amorphous semiconductor or a crystalline semiconductor, and any semiconductor material can be used. As a specific example of the material of the precursor semiconductor thin film 803, a material containing amorphous silicon such as hydrated amorphous silicon (a-Si: H) is preferable, but polycrystalline silicon having poor crystallinity is used. It may be a material containing, or a material containing microcrystalline silicon. Furthermore, the material of the precursor semiconductor thin film is not limited to a material made only of silicon, and may be a material mainly composed of silicon containing other elements such as germanium. For example, the forbidden band width of the precursor semiconductor thin film can be arbitrarily controlled by adding germanium.

  Further, the thinner the precursor semiconductor thin film, the shorter the film formation time and the better the productivity. On the other hand, there is a drawback that the electrical characteristics of the semiconductor thin film are lowered. Therefore, it is desirable that the thickness of the precursor semiconductor thin film be as thin as possible within a range that can satisfy desired electrical characteristics. The thickness is generally empirically within the range of 30 nm to 200 nm, but is not particularly limited to this range. In this embodiment, a 50 nm thick precursor semiconductor thin film made of amorphous silicon formed by a CVD method can be used.

<Details of semiconductor thin film manufacturing equipment>
FIG. 6 is a block diagram showing an example of a semiconductor thin film manufacturing apparatus suitably used in the present invention. In this semiconductor thin film manufacturing apparatus, the main laser oscillator 601 is not particularly limited as long as it can emit a laser beam to melt the precursor semiconductor thin film. For example, an excimer laser or a YAG laser (second and second lasers) An oscillator that emits ultraviolet laser light, such as various solid-state lasers represented by the third harmonic), is desirable. This is because the amorphous silicon thin film used as the precursor semiconductor thin film of this embodiment has a characteristic of absorbing light in the ultraviolet region or light on the short wavelength side even in the visible region.

  The variable attenuator 602 has a function of adjusting the energy density of the laser light that reaches the substrate composite 607. A beam shaping optical system 603 and a uniform illumination optical system 604 provided in the laser beam path shape the main laser beam emitted from the main laser oscillator 601 to an appropriate size, and then uniformly apply the main laser beam to the pattern forming surface of the mask 605. It has a function to illuminate. The image of the mask 605 provided in the laser beam path is imaged on the substrate complex 607 at a predetermined magnification by the imaging lens 606.

  In the semiconductor thin film manufacturing apparatus shown in FIG. 6, the magnification of the imaging lens 606 can be set to 1/5, for example. It is also possible to install a field lens between the mask 605 and the uniform illumination optical system 604 so that the imaging system is an image side telecentric system. The plurality of mirrors 608 provided in the laser beam path are used for folding the laser beam, but there are no restrictions on the arrangement location or quantity thereof, and the mirror 608 can be appropriately arranged according to the optical design and mechanism design of the apparatus. Is possible.

  On the other hand, the secondary laser oscillator 611 is not particularly limited as long as it can emit a laser beam and heat the lower layer of the precursor semiconductor thin film. A laser having an outer wavelength is desirable. This is because infrared to far-infrared light is transmitted through the amorphous silicon thin film used as the precursor semiconductor thin film of this embodiment, and a buffer layer (silicon oxide) formed under the amorphous silicon thin film and This is because it is absorbed by the substrate (glass).

  The beam shaping optical system 612 is arranged to adjust the diameter of the laser light incident on the intensity uniformizing optical system 613. This is because, in order to make the intensity distribution of the laser beam irradiated to the substrate composite 607 uniform with high accuracy, it is necessary to precisely adjust the size of the incident laser beam. The intensity uniformizing optical system 613 has a function of irradiating the substrate composite 607 with the sub laser light having a uniform intensity distribution.

  A plurality of mirrors 614 provided in the laser beam path are also used to fold the laser beam, but there are no restrictions on the arrangement location and quantity thereof, and it is possible to arrange them appropriately according to the optical design and mechanism design of the apparatus. is there. Although a variable attenuator arranged in the optical path of the main laser light can be installed, in the present embodiment, the auxiliary laser oscillator 611 having an output variable function is employed.

  6 is configured to irradiate the main laser light from the vertical direction of the substrate composite and irradiate the sub laser light from the oblique direction of the substrate composite, as shown in the block diagram of FIG. In addition, it is possible to couple the main laser beam and the sub laser beam by the beam splitter 709 and irradiate both laser beams from the vertical direction of the substrate composite 707. Here, in the relationship between the apparatus in FIG. 6 and the apparatus in FIG. 7, the constituent elements 701 to 708 correspond to the constituent elements 601 to 608, respectively, and the constituent elements 711 to 714 correspond to the constituent elements 611 to 614, respectively. .

Example 1
Example of applying the present invention to a conventional ELA method:
FIG. 9A is a schematic plan view showing an example in which the present invention is applied to a conventional ELA method. As described above, in the ELA method, an excimer laser having a linear cross section having a length of 200 to 400 mm and a width of about 0.2 to 1.0 mm is pulsed on the precursor semiconductor thin film on the substrate while moving the substrate at a constant speed. Irradiated. When the precursor semiconductor thin film is irradiated with laser light having energy that cannot be melted over the entire thickness, the buffer layer 122 and the precursor semiconductor thin film 123 are formed on the substrate 121 as shown in the schematic cross-sectional view of FIG. Innumerable crystal nuclei are formed in the vicinity of the boundary, and crystals grow in the direction of the arrow in the recrystallization region 124.

  In this embodiment, the main laser irradiation pattern 901 in FIG. 9A is a rectangle having a length of 300 mm and a width of 0.5 mm, and the sub laser irradiation pattern 902 has a length of 300.2 mm and a width of 0.7 mm. The main laser is irradiated over the sub-laser irradiation area. By making the sub-laser irradiation area larger than the main laser irradiation area, a polycrystal with a uniform grain size can be formed throughout the main laser irradiation area Is possible. In this case, the desired region of the precursor semiconductor thin film is crystallized by repeating the series of steps shown in FIG. 1, and irradiation is performed 0.1 mm in the direction of the arrow in FIG. Laser irradiation was performed while stepping the area. Laser irradiation conditions are summarized in Table 1. However, the delay time from the sub laser irradiation to the main laser irradiation was set to satisfy the above inequality (8).

  Table 2 shows a comparison of crystal grain sizes between the polycrystalline semiconductor thin film obtained by this example and the polycrystalline semiconductor thin film obtained by the conventional technique. Table 2 also shows the area that can be irradiated by the secondary laser in relation to the comparison of productivity between the prior art 2 and the method according to the present embodiment. In Table 2, the conventional technique 1 means a conventional ELA method, and the conventional technique 2 means a conventional method of irradiating a main laser and a sub laser in an overlapping manner, and means the technique described in Patent Document 2. .

  As is clear from Table 2, the single crystallized region is greatly expanded by the present invention as compared with the prior art 1. Compared with the prior art 2, in the present invention, in addition to the effect of expanding the single crystallized region by about 8 times, the energy density of the sub laser can be reduced. The ability to reduce the energy density of the secondary laser means that the irradiation area of the secondary laser can be enlarged under the same laser source, and a more efficient crystallization process is possible.

  FIG. 11 is a schematic plan view showing another laser irradiation region in the present embodiment. As shown in FIG. 11, if the irradiation areas and irradiation positions of the main laser 31 and the sub laser 32 are determined in advance, the sub laser is at least once at the main laser irradiation position 31 with the relative movement of the substrate and the laser light. The light is irradiated into the irradiated position 32. As a result, even if the first step of FIG. 1 (multiple pulse irradiation of the sub-laser in FIG. 1) is omitted, an equivalent effect is recognized in the expansion of the single crystal region.

(Example 2)
Example of applying the present invention to a conventional super lateral growth method:
The super lateral growth method is a method of performing crystallization by irradiating a precursor semiconductor thin film with a slit-shaped laser beam having a fine width on the order of several μm and melting the precursor semiconductor thin film over the entire thickness. As shown in the schematic cross-sectional view of FIG. 10A, if the precursor semiconductor thin film 103 formed on the buffer layer 102 on the substrate 101 is melted over the entire thickness in a region of several μm in length, the molten precursor An infinite number of crystal nuclei are formed at the boundary between the body semiconductor thin film region and the precursor semiconductor thin film region that does not melt, and crystals grow in the direction of the arrow shown in the crystallization region 104.

  The schematic plan views of FIGS. 9B to 9E illustrate the irradiation patterns of the main laser and the sub laser when the present invention is applied to the conventional super lateral growth method. In this example, the main laser irradiation pattern 911 in FIG. 9B is a rectangle having a length of 1 mm and a width of 0.04 mm, and the sub laser irradiation pattern 912 is a rectangle having a length of 1.2 mm and a width of 0.14 mm. did. An irradiation pattern in which three rectangular patterns were arranged at equal intervals was employed. Similar to the first embodiment, also in the second embodiment, the irradiation region 912 of the sub laser is made larger than the irradiation region 911 of the main laser. Each time the main laser 911 was irradiated, laser irradiation was performed while stepping the irradiation region by 0.015 mm in the direction of the arrow in FIG. The laser irradiation conditions in this case are summarized in Table 3. However, also in this case, the delay time from the sub laser irradiation to the main laser irradiation was set to satisfy the above inequality (8).

  Table 4 shows a comparison of crystal grain sizes between the polycrystalline semiconductor thin film obtained by this example and the polycrystalline semiconductor thin film obtained by the prior art. Table 4 also shows the area that can be irradiated with the secondary laser in relation to the comparison of productivity between the prior art 4 and the present embodiment. In Table 4, the conventional technique 3 means a conventional super lateral growth method, and the conventional technique 4 is a method of irradiating a conventional main laser and a sub laser in an overlapping manner. It means an example applied to the super lateral growth method. In Table 4, the average crystal length represents the length of crystal grown per pulse of the main laser.

  As is apparent from Table 4, the single crystallized region is greatly expanded by the present invention as compared with the prior art. Furthermore, compared with the prior art 4, the effect of expanding the single crystallization region is about 7 times. Further, similarly to the first embodiment, the energy density of the sub-laser can be reduced and the irradiation area of the sub-laser can be increased, so that a more efficient crystallization process can be performed.

  As in the case of the first embodiment, also in the second embodiment, as shown in FIG. 11, the region 31 irradiated with the sub laser at least once with the relative movement of the substrate and the laser light. It is also possible to set so that the main laser 31 is irradiated inside. Further, in the second embodiment, the pattern of FIG. 9B is described, but the effects of the present invention can be similarly obtained for the irradiation patterns of FIGS. 9C to 9E. That is, in FIGS. 9C to 9E, reference numerals 921, 931, 941, and 951 represent main laser irradiation areas, and reference numerals 922, 932, 942, and 952 represent sub laser irradiation areas.

Example 3
Example of applying the present invention to a super lateral growth method combined with a conventional capping technique:
The capping technique is a method of performing crystallization by providing a cap layer 814 on a precursor semiconductor thin film 813 formed on a buffer layer 812 on a substrate 811 as shown in a schematic perspective view of FIG. It is. As the cap layer 814, a thin film having an antireflection effect on the main laser light, such as silicon oxide, can be preferably used. Only in the region under the cap layer 814, the precursor semiconductor thin film is selectively melted and crystallized.

  The schematic cross-sectional view of FIG. 10B illustrates the outline of the super lateral growth method using the capping technique together. 10B, the substrate composite includes a substrate 111, a buffer layer 112, a precursor semiconductor thin film 113, and a cap layer 115. Since the cap layer 115 has an antireflection effect on the laser beam, the precursor semiconductor thin film 113 can absorb a larger amount of laser energy in the region covered with the cap layer than in other regions. Therefore, by setting an appropriate laser energy, the precursor semiconductor thin film 113 can be melted and crystallized in a region limited only under the cap layer 115. In this case, as in the case of Example 2, if the precursor semiconductor thin film 113 is melted over the entire thickness in a region having a length of several μm, the boundary between the melted region and the non-melted region of the precursor semiconductor thin film is innumerable. Crystal nuclei are formed and crystals grow in the direction of the arrow shown in the crystallization region 114.

  The schematic plan view of FIG. 9 (f) illustrates the irradiation pattern of the main laser and the sub laser when the present invention is applied to a super lateral growth method using a conventional capping technique together. In this embodiment, the main laser irradiation pattern 951 is a rectangle of 0.8 mm × 1.8 mm, and the sub laser irradiation pattern is a rectangle of 1.0 mm × 2.0 mm. As in the case of Examples 1 and 2, also in Example 3, the sub laser irradiation region 952 was made larger than the main laser irradiation region 951. Each time the main laser was irradiated, the irradiation area was moved stepwise so that the irradiation areas did not overlap. Laser irradiation conditions are summarized in Table 5. However, also in this case, the delay time from the sub laser irradiation to the main laser irradiation was set to satisfy the above inequality (8).

  Table 6 shows a comparison of crystal grain sizes between the polycrystalline semiconductor thin film obtained by the present invention and the polycrystalline semiconductor thin film obtained by the prior art. Table 6 also shows the area that can be irradiated with the sub-laser in relation to the comparison of productivity between the prior art 6 and the present invention. In Table 6, the prior art 5 means a super lateral growth method using a conventional capping technique together, and the prior art 6 is based on a conventional method of irradiating a main laser and a sub laser in an overlapping manner. Is applied to a super lateral growth method using a conventional capping technique together.

  As is apparent from Table 6, it can be seen that the single crystallized region can be greatly expanded by the present invention as compared with the prior art. Compared with the prior art 6, in the present invention, the effect of expanding the single crystallized region is about four times. Further, as in the case of the first and second embodiments, also in the third embodiment, the energy density of the sub laser can be reduced and the irradiation area of the sub laser can be expanded, so that a more efficient crystallization process is performed. Is possible.

  As described above, according to the method for manufacturing a semiconductor thin film of the present invention, it becomes possible to expand a recrystallization region using a laser as compared with the prior art, and manufacture thin film semiconductor devices of various sizes. It becomes possible to do.

It is a timing chart illustrating the irradiation mode of the main laser pulse and the sub laser pulse by this invention. It is typical sectional drawing illustrating the manufacturing method of the polycrystalline-semiconductor thin film by irradiation of a main laser and a sublaser. It is a graph which shows the time change of the oscillation trigger pulse and laser pulse intensity of a laser. It is a timing chart illustrating the irradiation control method of the main laser and the sub laser according to the present invention. It is a graph which shows the time change of the specific intensity of a sublaser pulse, and the change of the crystal grain size depending on the main laser irradiation time. It is a block diagram which shows an example of the manufacturing apparatus of the semiconductor thin film which can be used suitably for this invention. It is a block diagram which shows the other example of the manufacturing apparatus of the semiconductor thin film used suitably for this invention. It is a typical perspective view which shows the board | substrate composite body used suitably for this invention. It is a typical top view which shows the laser irradiation pattern in the various Example of this invention. It is a typical sectional view illustrating the direction in which solidification recrystallization grows in the fusion region of a precursor semiconductor thin film. It is a typical top view which shows the other laser irradiation pattern in the Example of this invention.

Explanation of symbols

  11 Main laser pulse, 12 Sub laser pulse, 21 Insulating substrate, 22 Buffer layer, 23 Precursor semiconductor thin film, 24 Main laser beam, 25 Sub laser beam, 601 and 701 Main laser oscillator, 602 and 702 Variable attenuator, 603 , 612, 703, 712 Beam shaping optical system, 604, 704 Uniform illumination optical system, 605, 705 Mask, 606, 706 Imaging lens, 607, 707 Substrate complex, 608, 708 Mirror, 611, 711 Sub laser oscillator, 613, 713 Intensity equalizing optical system, 614, 714 Mirror, 709 Beam splitter, 801, 811 Insulating substrate, 802, 812 Buffer layer, 803, 813 Precursor semiconductor thin film, 814 Cap layer, 901, 911, 921, 931 941, 951 Main laser irradiation pattern, 02, 912, 922, 932, 942, 952 Sub laser irradiation pattern, 101, 111, 121 Insulating substrate, 102, 112, 122 Buffer layer, 103, 113, 123 Precursor semiconductor thin film, 104, 114, 124 Crystallization Region, 115 cap layer.

Claims (6)

  A polycrystalline semiconductor in which a main pulse laser absorbed in a precursor semiconductor thin film and a sub-pulse laser absorbed in a layer below the precursor semiconductor thin film are temporally and spatially overlapped to irradiate the precursor semiconductor thin film A method for manufacturing a thin film, comprising: irradiating the sub-pulse laser at least once prior to the irradiation with the main laser beam.   A method of manufacturing a polycrystalline semiconductor thin film by irradiating the precursor semiconductor thin film with a main pulse laser and a sub-pulse laser while relatively moving a precursor semiconductor thin film and a laser beam, the main pulse laser comprising: A method of manufacturing a semiconductor thin film, wherein the region irradiated with the sub-pulse laser is irradiated at least once.   3. The method of manufacturing a semiconductor thin film according to claim 1, wherein the main pulse laser is irradiated immediately before the completion of irradiation of the sub-pulse laser. When irradiating the main pulse laser and the sub pulse laser, the delay time from the oscillation trigger pulse of the main laser to the start of main laser radiation is DL1, the fluctuation of DL1 is J1, and the sub laser oscillation trigger pulse is When the delay time until the start of laser emission is DL2, the fluctuation of DL2 is J2, and the oscillation trigger delay time of the main laser for the sub laser is Ts, the time Tr from the start of sub laser pulse irradiation to the start of main laser pulse emission is ,
DL1-DL2- (J1 + J2) + Ts <Tr <DL1-DL2 + (J1 + J2) + Ts
The method of manufacturing a semiconductor thin film according to claim 1, wherein the inequality is satisfied.   The main pulse laser has a wavelength absorbed by the precursor semiconductor thin film and energy capable of melting the precursor semiconductor thin film and the polycrystalline semiconductor thin film, and the sub-pulse laser includes the precursor semiconductor thin film 5. The method for producing a semiconductor thin film according to claim 1, wherein the semiconductor thin film has a wavelength capable of heating at least one of the first layer and the lower layer.   6. The main pulse laser has a wavelength in a range from an ultraviolet range to a visible range, and the sub-pulse laser has a wavelength in a range from an infrared range to a far infrared range. The manufacturing method of the semiconductor thin film of description.

Images