JP2002261015A - Semiconductor thin film, method of manufacturing it, manufacturing device, semiconductor element and method of manufacturing it - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make large-sized crystal grains in a silicon thin film in a polycrystalline silicon thin film transistor to obtain a semiconductor element having a high field-effect mobility and the like. SOLUTION: A semiconductor element is constituted in a structure that an upper insulating film 203 is formed on a lower insulating film 202, which is formed on its side coming into contact with a transparent insulative substrate 201 and has a heat conductivity higher than that of the film 203, and a two-layer structure insulating film consisting of a material having a heat conductivity higher than that of the film 202 is formed on the substrate 201. After that, the film 203 is formed by patterning into the form of a plurality of stripes and thereafter, an amorphous silicon thin film 204 is formed on the patterned insulating film. Moreover, a laser beam is irradiated in scanning in parallel to the strip pattern of the film 203 to turn the thin film 204 into a polycrystalline silicon thin film 210.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a thin film transistor (TFT) used for a liquid crystal display, an optical sensor such as a line sensor, a photovoltaic element such as a solar cell, an SRA
Memory LS such as M (Static Random Access Memory)
The present invention relates to a semiconductor film applied to I and the like, a manufacturing method thereof, and a manufacturing apparatus. More specifically, the semiconductor film is, for example, a semiconductor thin film having crystallinity formed on a glass substrate or the like and formed by subjecting an amorphous material or the like to laser annealing. The present invention also relates to a semiconductor device using such a semiconductor thin film and a method for manufacturing the same.

[0002]

2. Description of the Related Art Conventionally, as a method and an apparatus for forming a high-quality silicon semiconductor thin film constituting a thin film transistor (TFT) on an amorphous insulating substrate, a plasma CVD method using a glow discharge and a plasma CVD apparatus have been known. Used. The hydrogenated amorphous silicon (a-Si) film obtained by these manufacturing methods and apparatuses has reached the level of functioning as a high-quality semiconductor thin film through years of vigorous research and development. Practical use in electro-optical devices such as switching transistors for pixels of active matrix liquid crystal displays in personal computers, engineering workstations, car navigation devices, optical sensors for facsimile image sensors, solar cells for calculator batteries, and various integrated circuits. Have been. The greatest advantage of the hydrogenated amorphous silicon is that
That is, it can be stably manufactured with good reproducibility on a large-area substrate at a process temperature of about 00 ° C.

On the other hand, in recent years, as the size of displays and image sensors has increased and the density of pixels has increased (higher definition), a silicon semiconductor thin film capable of following higher-speed driving has been required. In addition, in order to reduce the weight and cost, it is necessary to be able to operate at high speed in order to apply to a driver element formed in a peripheral circuit portion of a liquid crystal display. However, the field-effect mobility of the hydrogenated amorphous silicon is, for example, at most 1.0 cm ² / V · sec, and electrical characteristics that sufficiently satisfy the above requirements cannot be obtained.

[0004] Therefore, a technique for forming a semiconductor thin film having crystallinity to improve the field effect mobility and the like has been studied. As the process, (1) hydrogen or SiF _{4 is} added to silane gas.
And a method of crystallizing a thin film to be deposited by using a plasma CVD method and (2) a method of attempting crystallization using amorphous silicon as a precursor.

In the above (1), the crystallization is performed simultaneously with the formation of the semiconductor thin film.
(00 ° C. or higher). Therefore, it is necessary to use an expensive quartz substrate or the like that can withstand high temperatures as the substrate, and it is difficult to use an inexpensive glass substrate, and there is a disadvantage that the manufacturing cost is high. Specifically, for example, Corning 7059 glass often used for an active matrix type liquid crystal display device has a glass strain point of 59%.
When the heat treatment is performed at 3 ° C. and 600 ° C. or more, mechanical deformation such as shrinkage or distortion of the glass substrate becomes remarkable, so that an appropriate process for forming a semiconductor circuit or a process for manufacturing a liquid crystal panel is difficult. become. In addition, if multi-dimensional integration is attempted, there is a possibility that the previously formed circuit region may be thermally damaged.

The above (2) is to form an amorphous silicon thin film on a substrate and heat it to form a polycrystalline silicon (polysilicon: p-Si) thin film.
A solid phase growth method in which heat treatment is performed at a temperature of about 600 ° C. for a long time and a laser annealing method (particularly, an excimer laser annealing method) are used.

In the former solid phase growth method, it is necessary to heat the substrate on which the amorphous thin film is formed and maintain the substrate at a temperature of 600 ° C. or more for 20 hours or more.

On the other hand, the latter excimer laser annealing method is described, for example, in IEEE Electron Device Letters, 7 (1986) pp. 2
76-278 and IEEE Transactions on Electron Devices, 42
As disclosed in (1995) pp. 251-257, an amorphous silicon thin film is irradiated with excimer laser light, which is a UV light having a large light energy, to be crystallized, without directly heating the glass substrate. And a polycrystalline silicon thin film having high field-effect mobility (> 100 cm ² / V · sec) and relatively good electrical characteristics.
That is, amorphous silicon has transmittance characteristics as shown in FIG. 1 and, for example, has an absorption coefficient of about 10 ⁶ cm ^{−1 with} respect to a laser beam having a wavelength of 308 nm by a XeCl excimer laser. Is almost absorbed in a region of about 100 ° from the surface of amorphous silicon, the temperature of the substrate hardly rises (about 600 ° C. or less), and only the amorphous silicon becomes high in temperature and is crystallized (polycrystal or single crystal). ). Therefore, an inexpensive glass substrate can be used, and a light beam can be locally irradiated to be crystallized. Crystallizing to form a driver circuit that requires high-speed operation, and successively improving the quality of a specified area on the same substrate without thermally damaging already formed circuits. Can be formed. Further, according to this technology, it becomes possible to integrate a CPU (Central Processing Unit) and the like on the same substrate.

Here, a general structure and a manufacturing method of a TFT as an example of a semiconductor element using p-Si as described above will be described.

FIG. 2 shows a TF having a coplanar structure.
FIG. 2A is a schematic diagram showing an outline of T110, and FIG.
FIG. 2B is a plan view of T110, and FIG.
It is sectional drawing in the -P 'arrow direction. As shown in FIG.
T110 is formed by forming an undercoat layer 112, a p-Si film 113, a first insulating film (gate insulating film) 114, a second insulating film 116, a gate electrode 115, a source electrode 117s and drain electrode 117d
And three electrodes are provided. p-Si
The film 113 is a crystalline semiconductor layer made of Si (silicon). The p-Si film 113 is formed on the undercoat layer 112 by patterning it into a predetermined shape. Further, the p-Si film 113 is
3a, the source region 113b and the drain region 113c
The source region 113b and the drain region 113c are located on both sides of the channel region 113a. The source region 113b and the drain region 113
c is formed by doping an impurity ion such as phosphorus or boron.

The first insulating film 114 is made of, for example, silicon dioxide (SiO ₂ ), and is formed above the p-Si film 113 and the undercoat layer 112. The gate electrode 115 is a metal thin film made of, for example, aluminum (Al). The gate electrode 115 is provided above the first insulating film 114 and at a position corresponding to the channel region 113a of the p-Si film 113. The second insulating film 116 is made of, for example, SiO ₂ ,
It is stacked above the gate electrode 115 and the first insulating film 114.

First insulating film 114 and second insulating film 11
6 are source regions 113 of the p-Si film 113, respectively.
Contact holes 118 reaching the drain region 113b or the drain region 113c are formed. Source electrode 117s
And the drain electrode 117d is formed in the contact hole 1
It is formed so as to be in contact with the source region 113b or the drain region 113c via the lines 18 and 118. The gate electrode 115, the source electrode 117s, and the drain electrode 117d form a wiring pattern by being patterned into a predetermined shape at a portion other than the illustrated cross section.

The TFT 110 is manufactured as follows. First, on the insulating substrate 111, for example, S
An undercoat layer 112 made of iO _{2 is formed} . Thus, diffusion of impurities into the p-Si film 113 or the like formed later is prevented. Next, an a-Si film (not shown) as amorphous silicon is formed on the undercoat layer 112 by, for example, a plasma CVD method, and the a-Si film is patterned into a predetermined shape by etching. I do. Note that the patterning may be performed after crystallization. Subsequently, the a-Si film is irradiated with a short-wavelength excimer laser or the like and allowed to cool (laser annealing). Thereby, the reforming, that is, the a-Si film is polycrystallized and the p-Si film 113 is formed. Here, since an a-Si film has a large light absorption coefficient in a short wavelength region, when an excimer laser is used as an energy beam,
Only the film can be selectively heated. Therefore,
Since the temperature rise of the insulating substrate 111 is small, there is an advantage that a low-cost glass substrate or the like can be used as the material of the insulating substrate 111.

On the p-Si film 113 formed above,
The first insulating film 114 is formed by atmospheric pressure CVD (Chemical Vapor Dep.
osition) method, and a gate electrode 115 is formed over the first insulating film 114. Next, using the gate electrode 115 as a mask, impurity ions serving as donors or acceptors, specifically, impurity ions such as phosphorus or boron are implanted into the p-Si film 113 by, for example, an ion doping method. Thereby, the p-Si film 11
3, a channel region 113a, a source region 113b, and a drain region 113c are formed.

Next, after a second insulating film 116 is formed on the gate electrode 115, contact holes 118 are formed.
Is formed, and, for example, aluminum is vapor-deposited and patterned to form a source electrode 117s and a drain electrode 117d.
To form

A pulse oscillation type laser such as an excimer laser in the process of forming a p-Si film as described above has a large output. For example, a laser beam in the form of a line beam is irradiated while scanning by moving a substrate or the like. Thereby, amorphous silicon having a large area can be crystallized at one time, which is advantageous for mass production of semiconductor devices, but has a problem that it is difficult to improve crystal quality. That is, in this type of laser, the irradiation time of one pulse is extremely short, about several tens of nanoseconds, and the temperature difference between irradiation and non-irradiation becomes large, so that the molten silicon film is rapidly cooled. Crystallizes. Therefore, it is difficult to control the degree of crystal growth and the crystal orientation, it is difficult to perform sufficient crystal growth, and the crystal grain size tends to be small, the crystal grain boundary density tends to be large, and the variation tends to be large. At the same time, crystal defects tend to increase. More specifically, in the cooling process after laser irradiation, crystal nuclei are randomly generated, and each of the randomly generated crystal nuclei grows in a random direction. Then, the crystal growth stops in a state where the crystal grains collide with each other. The crystal grains generated through such a growth process are small and have a random shape. This results in a poly-Si film having a large number of crystal grain boundaries. In such a poly-Si film, charge carriers cannot move smoothly, and thus TFT characteristics such as field-effect mobility are inferior.

Hereinafter, the mechanism of crystal growth and the reason why it is difficult to perform good crystal growth will be described in more detail. The excimer laser is generated by a method in which a mixed gas of a rare gas such as Xe or Kr and a halogen such as Cl or F is excited by an electron beam. For this reason, a light beam shaped into a rectangular or linear uniform light intensity of several cm square using an optical system called a beam homogenizer is used in the laser annealing method. In crystallization of a non-single-crystalline thin film (usually, an amorphous thin film) formed on a substrate, a method of irradiating the shaped light beam while scanning is used.

However, this method also has some problems to be solved, such as poor uniformity of crystal grain size and crystallinity, unstable transistor characteristics, and low field-effect mobility. I have. For this reason,
As a method for solving these problems, (1) an intensity distribution is formed by covering a part of an irradiation surface with a reflection film or an absorption film and controlling the light absorption of the thin film surface, thereby controlling the crystal growth direction. (2) Technology for smoothing crystallization by applying laser irradiation while heating the substrate (400 ° C) (Extended Abstracts of the 1991 Inter
national Conference on Solid State Devices and Mat
erials, Yokohama, 1991, pp 623-625) have been proposed. Also, for example, Jpn.J.Appl.Phys.31 (1992) p4
The one disclosed in 550-4554 is also known. This is, for example, as shown in FIG.
Is placed on a substrate stage 124, and a substrate heater 1
The laser light 123 a of the excimer laser 123 is irradiated to the precursor semiconductor thin film 122 while the substrate 24 is heated to about 400 ° C. In this way, by using the method of heating the glass substrate at the time of laser beam irradiation,
High crystal quality, that is, relatively large and uniform crystal grains are obtained, and the electrical characteristics are improved.

Among them, the former (1) can achieve single crystallization, and the latter (2) can be applied relatively easily, and when this technique is used, the variation in the field effect mobility is ± 10%.
It is said that it can be suppressed within. However, the above-described technology has the following problems, and cannot sufficiently respond to recent technical trends for multidimensional integration and further cost reduction.

That is, the technique (1) requires a step of applying a reflective film or the like, which complicates the manufacturing process and increases the cost. In addition, since it is not easy to apply a reflection film or the like to a limited narrow area, it is difficult to crystallize a minute specific area.

On the other hand, in the technique (2), since a heating step for heating the substrate is required, the productivity is reduced accordingly. That is, although the substrate is not heated to a high temperature as in the solid phase growth method, the process of heating and cooling the substrate still requires time (for example, about 30 minutes to 1 hour), and thus has a problem that the throughput is reduced. . The problem is that the larger the area of the substrate, the longer the time required for heating and cooling to reduce the distortion of the substrate,
It will be even more noticeable. Further, this technique can reduce the variation in the field effect mobility to some extent, but cannot sufficiently increase the field effect mobility itself, so that it is insufficient for forming a circuit requiring high speed. . That is, the glass substrate cannot be heated to about 550 ° C. or higher in order to prevent the distortion or the like of the glass substrate, and it is difficult to obtain higher crystal quality or the like. Further, since this technique is a method of heating the entire substrate, it is not suitable for crystallizing only a limited area (specific area) on the substrate.

As described above, each of the techniques (1) and (2) has a problem such as an increase in manufacturing cost. In particular, the techniques (1) and (2) (the same applies to other conventional techniques) have an essential problem that it is difficult to realize various and multidimensional lamination. That is, the means for controlling the temperature distribution employed by these technologies includes a circuit region (polycrystalline region) requiring high speed,
This is an unsuitable means for selectively forming a circuit region (amorphous region) other than the above on the same substrate. Therefore, it is difficult to simultaneously achieve high integration and cost reduction by these techniques.

Here, the usefulness of the technique capable of crystallizing only a limited arbitrary region will be described. In the laser annealing method according to the prior art, as shown in FIG.
A light beam having a steep side portion (edge portion) and a flat top (the same energy intensity per unit area) is used. Poly-S using light beam with such characteristics
Even an i-thin film was sufficient because it was conventionally used to form, for example, a switching circuit for a pixel electrode, which does not require a very high-speed operation.

However, when elements requiring high-speed operation, such as a gate drive circuit and a source drive circuit, and even a CPU or the like, are to be integrally formed on the same substrate, the above-described conventional technique can be used. Polycrystalline thin films of moderate quality are not sufficient. Specifically, for example, a mobility of about 0.5 to 10 cm ² / Vs is sufficient in a pixel region of an LCD, but a mobility of about 100 to 10 cm ² / Vs is used in a peripheral driving circuit such as a gate circuit or a source circuit for controlling pixels. 300c
Mobility of about m ² / Vs is required. However, the conventional technique using the light beam having the above characteristics cannot stably obtain a high mobility. That is, in general, in a polycrystalline silicon thin film, transistor characteristics such as mobility increase as the crystal grain size increases, but sufficient transistor characteristics cannot be obtained by the above polycrystallization treatment.

The reason for this is that, with a light beam having the above characteristics, the crystal grains and the degree of non-uniformity of crystallinity increase, and in order to enlarge the crystal grains and increase the crystallinity, the irradiation intensity is increased or the number of irradiations is increased. This is because, when the number is increased, the size of the crystal grains becomes more irregular and the degree of crystallinity varies. Hereinafter, this cause will be discussed in detail.

FIG. 5 is a schematic diagram showing the distribution of crystallinity when the rectangular light beam is irradiated on the amorphous silicon thin film formed on the substrate. In FIG. 5, 1
Reference numeral 701 denotes a boundary of the irradiation light, 1702 and 1704 denote low crystallinity portions, and hatched portions 1703 denote high crystallinity portions. As shown in this figure, according to the conventional method using an excimer laser having a uniform energy intensity, a hatched portion 1703 slightly inside the boundary 1701 of the irradiation light.
Alone increases the crystallinity, and the other parts (170 around the boundary)
2 and the central portion 1704), a characteristic crystallinity distribution pattern of low crystallinity is formed. This is confirmed by micro-Raman spectroscopy.

That is, FIG. 6 shows the measurement results of the Raman intensity of the portion along the line AA in FIG. 5. In FIG. 6, the presence of a steep peak slightly inside the boundary indicates the presence of this portion. It can be seen that the degree of crystallinity is high. Further, since there is no peak in the central portion, it is understood that the crystallinity of this portion is low.

Next, with reference to FIG. 7, the mechanism by which such non-uniform crystallinity occurs will be considered. After irradiating the amorphous silicon thin film with a light beam and heating the thin film temperature to a temperature higher than the melting temperature of silicon (about 1400 ° C. or higher), when the light irradiation is stopped, the thin film temperature decreases due to heat radiation,
In this process, the molten silicon precipitates and crystallizes.
Here, when a light beam having a uniform light intensity distribution as shown in FIG. 7A is applied, a temperature distribution pattern as shown in FIG. 7B is formed on the thin film surface. That is, a flat temperature region without a temperature gradient is formed in the central portion, and a steep temperature gradient is formed in the peripheral portion because heat escapes to the surroundings. In this case, the temperature in the central portion is the melting temperature of silicon. If this is the case, after the end of the irradiation, first, the temperature near the intersection (near the boundary) between the temperature distribution curve 1901 and the crystallization temperature line 1902 reaches the crystallization temperature. Therefore, the crystal nucleus (1
903) occurs (FIG. 7C). That is, the amorphous silicon thin film exceeds the melting point, and crystallization occurs when the amorphous silicon thin film is melted and solidified in a region beyond the melting point, whereby polycrystallization is achieved. Next, when the temperature further decreases (FIG. 7 (d)), crystallization proceeds from the crystal nucleus 1903 as a starting point toward a central portion which has not yet reached the crystallization temperature (FIG. 7 (e)). 7), when a light beam having a uniform light energy intensity is used, FIG.
As shown in (b, d, f), the temperature drops in a state where there is almost no temperature gradient in the center in the plane direction. Therefore, at a certain point of the temperature drop, a relatively wide range reaches the crystallization temperature at the same time (FIG. 7 (f)), and crystal nuclei can be generated with equal probability in any part of this range (1904). Become. For this reason, as shown in FIG. 7G, fine crystal nuclei are simultaneously generated on the entire surface of 1904, and as a result, a poly-Si thin film composed of many fine crystal grains is formed.
Such a poly-Si thin film naturally has a high density of crystal grain boundaries. Therefore, the degree to which carriers are trapped at the crystal grain boundaries increases, and the field-effect mobility decreases. Note that reference numeral 1900 in FIG. 7C indicates a cross section of the thin film.

The mechanism by which the degree of crystallinity becomes non-uniform as described above is the same as in the case of irradiating a linear laser beam as shown in FIG. Here, in FIG. 8, (a) shows the energy density distribution in the x and y directions of the excimer laser to be used, and (b) shows the excimer laser having such an energy density distribution on an amorphous silicon thin film. The temperature rise distribution of the amorphous silicon thin film when irradiated is shown, and (c) is a perspective view of a polycrystalline silicon thin film transistor irradiated with laser as shown in (a) and (b) above. That is, although the temperature distribution in the y-direction of the irradiated area is substantially uniform due to the use of the laser having the energy distribution as shown in FIG. 8A, as shown in FIG. In the x direction, a temperature distribution that is high at the center and low at both sides is generated. As a result of such a temperature distribution, the crystallization proceeds from the peripheral portion to the center in the x direction, and the respective growth surfaces meet in the central portion due to generation of a large number of nuclei. As shown schematically, the crystal grain in the portion where the line beam energy density of the laser beam is low is large, but the crystal grain in the portion where the energy density is high (center portion) is small. In FIG. 8, 131 denotes a transparent insulating substrate, 134 denotes a polycrystalline silicon thin film, and 141 denotes crystal grains. Reference numeral 139 denotes an insulating film, which is generally a silicon dioxide (SiO ₂ ) film, and reference numeral 140 denotes an amorphous silicon thin film.

In the above example, the case where the energy beam is irradiated once is shown for the sake of simplicity, but the same applies to the case where the energy beam is irradiated a plurality of times.

In conventional laser annealing,
In addition to the difficulty in improving the field-effect mobility as described above, it is difficult to improve the uniformity of the film quality of the semiconductor film, and in particular, it is difficult to achieve both. Had.

Here, the configuration of a conventional laser annealing apparatus will be described with reference to FIG. In FIG. 9, 15
1 is a laser oscillator, 152 is a reflecting mirror, 153 is a homogenizing device, 154 is a window, 155 is a substrate on which an amorphous silicon layer is formed, 156 is a stage, and 157 is a control device. During the laser annealing of the amorphous silicon layer, the laser beam oscillated from the laser oscillator 151 is guided to the homogenizer 153 by the reflecting mirror 152, and the laser beam shaped into a predetermined shape with uniform energy is applied to the window. 154
Substrate 1 fixed to stage 156 in the processing chamber through
55 is illuminated.

When an annealing process is performed using the above-described laser annealing apparatus, it is difficult to irradiate the entire surface of the substrate with a laser beam at one time. The entire surface of the substrate is irradiated under the same conditions while shifting (for example, I. Asai, N. Kato,
M. Fuse and T. Hamano, Japan J Appli Physics (Jpn. J. Appl. Phys.) 32 (1993) 474). However, in the laser annealing method of irradiating the laser beam while shifting the irradiation area sequentially while overlapping the irradiation areas of the laser beam, if the laser energy density is increased, the mobility, which is one of the evaluation criteria of the semiconductor film characteristics, is increased. Although the film quality is improved as a whole, the film quality becomes non-uniform at the joint of the irradiation area, and the uniformity of the entire semiconductor film is reduced. On the other hand, when laser irradiation is performed at a relatively low energy density, it is easy to improve the uniformity of the film quality, but it is difficult to increase the field-effect mobility because the energy density is low.

Therefore, for example, when a substrate on which a TFT is formed is used for a liquid crystal display, as shown in FIG. 10, uniformity of film quality required for a relatively large area image display region 158 and peripheral circuit portion ( It has been difficult to form a semiconductor film that satisfies the field effect mobility required for the driver circuit 159. In order to solve such a problem, for example, as disclosed in U.S. Pat. No. 5,756,364, it is proposed to make the laser beam intensity different between the image display area 158 and the peripheral circuit section 159. It is difficult to make the peripheral circuit portion 159 have sufficient field effect mobility only by changing the intensity of the laser beam in this way.

In this specification, crystallization is used to include both single crystallization and polycrystallization. However, the method for producing a crystalline semiconductor thin film of the present invention is poly-crystalline.
It is particularly useful for producing a Si thin film.

[0036]

In the conventional laser annealing method, as described above, it is difficult to control the crystal grain size and crystal orientation, and high crystal quality, that is,
It is difficult to form a semiconductor thin film having a large and uniform crystal grain size and few crystal defects, and it is also difficult to improve the throughput and reduce the manufacturing cost. There is a problem that it is not possible to simultaneously improve the characteristics (such as the field effect mobility) and uniformity of the film quality.

In view of the above, the present invention can form a semiconductor thin film of high crystal quality without lowering the throughput, and can further improve the film characteristics of the semiconductor thin film and improve the uniformity of the film quality. It is an object of the present invention to provide a method and an apparatus for manufacturing a semiconductor thin film capable of simultaneously achieving the above, and a thin film transistor using such a semiconductor thin film and having excellent TFT characteristics such as field-effect mobility, and a method for manufacturing the same.

[0038]

In order to solve the above-mentioned problems, the present inventors have made various studies and found that the crystal grains of the polycrystalline silicon thin film become small due to the fact that the silicon thin film when heated by excimer laser irradiation. Focusing on the temperature distribution of the thin film, a method of forming a polycrystalline silicon thin film having a large grain size at least in a region where a transistor is formed has been devised.

That is, the present inventors provide high heat conduction regions on both sides of a region where a transistor is to be formed when performing a polycrystallization process using a laser, so that the region where the transistor is formed is formed. By increasing the temperature of the surrounding area and lowering the temperature of the transistor forming area relatively to the surrounding area, the silicon thin film in the transistor forming area is first crystallized to increase the grain size. I came to think I could get it.

Therefore, the invention of claim 1 has a first insulating film having a first thermal conductivity on a substrate and a second thermal conductivity different from the first thermal conductivity, Laminating a second insulating film selectively formed in a partial region, laminating a non-single-crystal semiconductor thin film on the first insulating film and the second insulating film, Irradiating the single crystal semiconductor thin film with an energy beam to grow crystals.

More specifically, for example, by making the thermal conductivity of the insulating film of the amorphous silicon thin film different between the region where the transistor is formed and the other region, the amorphous silicon film in the region where the transistor is formed is formed. The thermal conductivity of the thin film is made higher than that of the amorphous silicon thin film in other regions.

According to this configuration, when performing polycrystallization,
Since the temperature of the silicon thin film in the transistor formation region is lower than that in the other regions, crystallization occurs from the transistor formation region, and the grain size of polycrystalline silicon in the transistor formation region can be increased.

According to a twelfth aspect of the present invention, at least a part of the periphery of the surface of the semiconductor film is provided with one or more protrusions extending in a direction substantially horizontal to the semiconductor film.

Here, an approach to the present invention will be briefly described for the purpose of understanding the present invention. First, the present inventors have repeatedly studied to find the causes of the above-mentioned problems of the conventional technology, and have not sufficiently clarified them, but have assumed the following items as the causes. That is,
Generally, generation of crystal nuclei and crystal growth are performed by heating a semiconductor film once by an annealing process and then cooling the semiconductor film.

By the way, in the prior art, the semiconductor film after the annealing treatment is cooled substantially uniformly regardless of the central part and the peripheral part, and as a result, the crystal nuclei are located at random positions almost simultaneously. Therefore, it is assumed that it is difficult to control the crystal grain size and crystal orientation. In addition, this may cause crystal nuclei to be generated almost simultaneously at relatively close positions. In this case, the crystals interfere with each other during the crystal growth process, making it difficult to obtain a sufficient crystal grain size. ing.

As a result of intensive studies with the above matters in mind, the present inventors have found that “crystal nuclei in the peripheral portion of the semiconductor film are generated earlier than crystal nuclei in the central portion, and thereafter, The crystal nuclei generated in the part are grown toward the central part before the crystal nucleus is generated or grown in the central part, so that the crystal grain size and the crystal orientation can be controlled and the crystal growth The technical idea of the present invention is to create a sufficient crystal grain size by preventing interference between crystals in the process.

That is, according to the twelfth aspect, in the semiconductor film after the annealing process, the heat accumulated in the peripheral protrusions is transferred to a plurality of directions on the horizontal surface outside (for example, when the protrusions have a rectangular shape, three times). Direction), while the heat accumulated in the central part has an escape area only on the peripheral side that has not yet been cooled in the horizontal plane, so the peripheral edge including the protrusion cools down much faster than the central part Is done.

As a result, the crystal nuclei at the periphery are generated earlier than the crystal nuclei at the center, and the crystal nuclei generated at the periphery are formed at the center before the crystal nuclei are generated or grown at the center. Because the crystal grows toward
The crystal grain size and crystal orientation can be controlled. This prevents interference between crystals in the course of crystal growth,
A sufficient crystal grain size can be obtained.

According to the thirteenth aspect, only one crystal nucleus is generated at the projection, and the crystal nucleus grows. According to the fourteenth and fifteenth aspects, the grain size of the crystal is further adjusted, and one crystal nucleus is reliably generated for each projection.

According to the sixteenth aspect, the crystal generated and grown on the protrusion further grows toward the center. In this case, the crystal grows from the adjacent protrusion toward the center and the crystal grows toward the center. It is anticipated that the crystal growth will be minimized from interfering with the crystal growth from the protrusion on the side to the center.

According to the eighteenth aspect, since the protrusion is provided in the region corresponding to the gate electrode, good conductive characteristics can be obtained.

In another aspect of the present invention, the step of forming the amorphous semiconductor film and the step of selectively forming one or more protrusions extending in a substantially horizontal direction on at least a part of the periphery of the amorphous semiconductor film are provided. A method of manufacturing a semiconductor film, comprising a step of forming and a step of annealing and crystallizing the amorphous semiconductor film on which the protrusions are formed.

The semiconductor film manufactured according to the nineteenth aspect has the same effect as the twelfth aspect.

Further, according to claim 22, a crystal nucleus in a peripheral portion of the semiconductor film is generated earlier than a crystal nucleus in a central portion, and thereafter, the crystal nucleus generated in the peripheral portion is replaced with the central portion. Since the crystal is grown toward the center before the generation of the crystal nucleus or the crystal growth, the crystal grain size and the crystal orientation can be controlled.

As a result, interference between crystals in the course of crystal growth is prevented, and a sufficient crystal grain size can be obtained.

According to another aspect of the present invention, a crystalline semiconductor layer having a channel region, and a source region and a drain region disposed on both sides of the channel region is provided. In a crystalline thin film transistor formed on a substrate, the crystalline semiconductor layer is formed by crystallizing a non-single crystalline thin film, and at least a channel region of the crystalline semiconductor layer controls a crystal growth direction. Characterized in that a gap for controlling the crystal growth direction is provided.

According to the above configuration, the crystal growth direction control gap formed in the channel region controls the crystal growth direction of the channel region when crystallization of the non-single crystalline thin film. Therefore, in the crystalline semiconductor layer having such a crystal growth direction control gap, the crystal shape and the crystal grain boundary density are appropriately regulated. Excellent TFT characteristics such as effective mobility.

Here, the above-mentioned crystal growth direction control gap is:
A depression (concave) formed on the surface of the crystalline semiconductor layer (non-monocrystalline thin film in the manufacturing stage), and the depression reaches the lower layer (substrate surface or undercoat layer) of the crystalline semiconductor layer. Or may not reach the lower layer. Also, the size and shape of the depression are not particularly limited. Therefore, it can be appropriately set in consideration of the size and thickness of the surface area of the crystalline semiconductor layer, the desired field-effect mobility, and the like. For example, a hole having a circular or square surface shape or an elongated groove can be exemplified, and the cross-sectional shape of the hole or groove can be exemplified by a C-shape, a V-shape, or a U-shape. The details of the role and function of the crystal growth direction control gap will be described later.

According to a twenty-seventh aspect of the present invention, there is provided a channel region, a source region disposed on both sides of the channel region,
And a crystalline semiconductor layer having a drain region and a crystalline semiconductor layer formed on a substrate, wherein the semiconductor layer is obtained by crystallizing a non-single crystalline thin film, and at least a channel of the crystalline semiconductor layer is formed. The region is characterized in that two or more rows of groove-shaped voids are provided in a direction connecting the source region and the drain region.

According to this configuration, the groove-shaped voids provided in two or more rows function to guide the crystal growth direction to the direction connecting the source region and the drain region when the non-single-crystalline thin film is crystallized. So the resulting poly-
The Si film is an aggregate of large crystal grains that are long in the direction connecting the source region and the drain region. Such poly-S
Since the i-film has a low crystal grain boundary density in the direction connecting the source region and the drain region, the carrier movement speed in this direction is high. That is, the crystalline thin film transistor having the above structure is excellent in characteristics such as carrier mobility.

The reason why large crystal grains with controlled crystal growth direction can be obtained by providing the crystal growth direction control voids will be described in detail with reference to FIGS. 11 and 20.

As shown in FIG. 20, a groove-shaped crystal growth direction control gap (reference numeral 411) is formed on the surface of the non-single crystalline thin film which is a precursor of the crystalline semiconductor layer in a direction connecting the source region and the drain region. When two or more rows are formed, and then the thin film is irradiated with an absorbable energy beam according to a conventional method,
The temperature of the thin film surface is low in the crystal growth direction control gap, its vicinity, and the periphery of the semiconductor thin film, and has a high temperature distribution in the channel region main body (the thin film portion where the crystal growth direction control gap is not formed).

This is because the groove portion (the crystal growth direction control gap) has a smaller thickness of the thin film than other portions or does not have the thin film, so that the absorption of the energy beam is small. This is because the temperature is lower than other parts. In addition, since there is usually no thin film outside the semiconductor thin film, the absorption of the energy beam is small, and the heat is diffused outside at the peripheral portion, so that the temperature is lower than that at the central portion of the thin film.

Next, the crystal growth process in a non-single-crystal thin film having a crystal growth direction control gap and a low temperature distribution in the periphery will be described. Since the temperature of the peripheral portion of the non-single-crystalline thin film is low also in the conventional technique, the relationship between the crystal growth direction control gap and the crystal growth direction will be described with reference to FIG. I do.

FIG. 11 is a diagram for conceptually explaining the manner of crystal growth. First, crystal nuclei are generated around the crystal growth direction control gaps whose temperature is lower than that of the main body. The crystal nuclei grow in a direction of higher temperature, that is, in a direction away from the groove-shaped crystal growth direction control gap (a direction perpendicular to the grooves) as the temperature of the entire thin film drops. Here, in the above configuration, since the crystal growth direction control gaps are provided in two or more rows in the direction connecting the source region and the drain region, the crystal nuclei generated near the two opposing crystal growth direction control gaps are: Each grows from the opposite direction toward the center of the channel region main body.

As a result, the crystal grains collide with each other near the center of the channel region main body. However, the temperature near the center far from the crystal growth direction control gap is higher than other portions, and the molecules still move freely. In a state of gaining. Therefore, crystal growth is induced in a direction to avoid collision, that is, in a direction connecting the source region and the drain region (a direction parallel to the groove) (see FIG. 11A). As a result, large crystal grains long in the direction connecting the source region and the drain region are formed (see FIG. 11B). In the case of a channel region composed of an aggregate of crystal grains having such a shape, a crystalline thin film transistor having excellent TFT characteristics such as field-effect mobility because a crystal grain boundary density in a direction connecting a source region and a drain region is small. Can be configured.

According to a twenty-eighth aspect of the present invention, in the twenty-sixth aspect, the plurality of crystal growth direction control gaps are discontinuously provided in a direction connecting the source region and the drain region. Features.

With this configuration in which a plurality of crystal growth direction control gaps are discontinuously arranged, crystal growth can be more finely controlled. In particular, when two or more rows of crystal growth direction control gaps are arranged, the crystal growth direction can be reduced. The size and shape of the grains can be more precisely controlled. The reason will be described below.

As described in the twenty-seventh aspect, crystal nuclei are generated in the vicinity of the crystal growth direction control gap where the temperature is lowered to the crystallization temperature sooner. Before crystal growth, it collides with other crystal grains to stop crystal growth, so that it becomes polycrystalline composed of a large number of fine crystal grains, and the crystal structure becomes distorted near boundaries where crystal grains collide with each other. Therefore, desired TFT characteristics cannot be obtained. Therefore, in order to enhance the TFT characteristics such as the field-effect mobility, it is necessary to control not only the crystal growth direction but also the generation density of crystal nuclei appropriately.

Here, if the voids are arranged discontinuously, crystal nuclei are generated in the vicinity of the voids, but it is difficult to generate crystal nuclei in the intermediate portion between the voids and the adjacent voids. Therefore, the density of crystal nuclei can be controlled by adjusting the number of voids and the spacing between voids. The reason that crystal nuclei are not easily generated in the intermediate portion between the void and the adjacent void is that this portion (the portion where the thin film substance exists) is sufficiently heated by laser irradiation.

According to a twenty-ninth aspect of the present invention, there is provided a crystalline thin film transistor in which a semiconductor layer having a channel region, and a source region and a drain region disposed on both sides of the channel region is formed on a substrate. The semiconductor layer is obtained by crystallizing a non-single-crystalline thin film,
At least the channel region is provided with an early crystallization region having a higher crystallization start temperature than the channel region main body.

According to the above configuration, the early crystallization region functions to control the crystal growth of the channel region main portion, so that a high-quality crystalline semiconductor layer having a small crystal grain boundary density can be formed. The reason is as follows.

Since the early crystallization region is a portion where the crystallization start temperature is higher than that of the channel region main portion, crystal nuclei are first generated in the early crystallization region during the cooling process. Therefore, thereafter, crystal growth is performed centering on the crystal nucleus. Therefore, by providing the early crystallization region, a phenomenon that many crystal nuclei are generated at one time can be prevented, and as a result, a polycrystalline semiconductor layer in which larger crystal grains are aggregated can be formed.

Here, at least one early crystallization region may be arranged at least in the channel region. However, the plurality of early crystallization regions are located at positions where the movement of carriers in the direction connecting the source region and the drain region is not hindered. May be provided. If a plurality of early crystallized regions are scattered at appropriate positions and intervals on the thin film surface, the density of crystal nuclei can be appropriately controlled, so that better results can be obtained. Note that “high crystallization start temperature” in the above configuration means that crystallization is started at a higher temperature than the channel region main body.

According to a thirtieth aspect of the present invention, in the crystalline thin film transistor according to the twenty-ninth aspect, the early crystallization region has a shape elongated in a direction connecting the source region and the drain region.

Since the early crystallization region is not a region for moving carriers, this region is preferably narrow in a direction connecting the source region and the drain region. This is because if the early crystallized region has a shape that is long in a direction orthogonal to the direction connecting the source region and the drain region, the early crystallized region may cause the carrier mobility to be impaired.

According to a thirty-first aspect of the present invention, in the crystalline thin film transistor according to the twenty-ninth aspect, the early crystallization region includes a component constituting a channel region main body and an impurity. I do.

If the means for increasing the crystallization start temperature by including impurities in the semiconductor layer, the early crystallization region can be formed relatively easily. Therefore, the crystalline thin film transistor having the above configuration is excellent in TFT characteristics such as the field effect mobility and is inexpensive.

A thirty-second aspect of the present invention is the crystalline thin-film transistor according to the twenty-sixth aspect, wherein the crystalline semiconductor layer is mainly composed of silicon or a compound of silicon and germanium.

Silicon or a compound of silicon and germanium is easily available and easily crystallized. Therefore, with the above structure, a high-quality crystalline thin film transistor can be provided at low cost.

The invention of claims 33 to 39 described below relates to the method of manufacturing the crystalline thin film transistor of claims 26 to 32. The functions and effects of the inventions of claims 33 to 39 are substantially the same as those described in the description of claims 26 to 32. Therefore, a detailed description of the operation and effect will be omitted below.

According to a thirty-third aspect of the present invention, there is provided a method of manufacturing a crystalline thin film transistor including a crystalline semiconductor layer having a channel region, and a source region and a drain region disposed on both sides of the channel region. Depositing a non-single crystalline thin film on the non-single crystalline thin film, forming a crystal growth direction control gap in the non-single crystalline thin film, Irradiating an energy beam to crystallize the thin film.

[0083] In the above invention, the crystal growth direction control gap may be formed in a groove shape in a direction connecting the source region and the drain region. A plurality can be formed discontinuously in the direction connecting the drain region. With these configurations, the above-described crystalline thin film transistors according to claims 26 to 28 can be manufactured.

According to a thirty-sixth aspect of the present invention, there is provided a method for manufacturing a crystalline thin film transistor comprising a crystalline semiconductor layer having a channel region and a source region and a drain region disposed on both sides of the channel region. Depositing a non-single-crystalline thin film on the insulating substrate; and ion-implanting an impurity for increasing the crystallization start temperature of the non-single-crystalline semiconductor thin film into a part of the non-single-crystalline semiconductor thin film, thereby performing early crystallization including the impurity. The present invention relates to a method for manufacturing a crystalline thin film transistor, comprising: an early crystallization region forming step of forming a region; and, after the early crystallization region forming step, a step of irradiating an energy beam to crystallize the thin film.

In the invention according to claim 36, in the early crystallization region forming step, a band-like early crystallization region long in a direction connecting the source region and the drain region can be formed, and the early crystallization region can be formed. The active region can be discontinuously arranged in a direction connecting the source region and the drain region. With these configurations, the above-described crystalline thin film transistors according to claims 29 to 31 can be manufactured.

Further, in each of the above inventions relating to the manufacturing method, an excimer laser beam can be used as the energy beam.

The excimer laser has a high light energy and is a UV light, so that it is well absorbed by silicon.
Therefore, with the use of an excimer laser beam, the non-single-crystalline semiconductor layer can be efficiently crystallized, particularly when the non-single-crystalline semiconductor layer is composed of an ultraviolet absorbing material such as silicon. Alternatively, only the semiconductor layer can be selectively heated and melted. Therefore, the semiconductor layer can be crystallized without adversely affecting a portion other than the irradiation region, and the rise in the substrate temperature is small, so that an inexpensive glass substrate can be used. Further, when a combination of an excimer laser and a UV-absorbing thin film material is used, the temperature difference between the crystal growth direction control gap and the main body of the semiconductor absorbing layer increases, so that the function of the crystal growth direction control gap (crystal growth direction Control function) is fully exhibited.

The inventors of the present invention have intensively studied a method for sufficiently growing a crystal based on the above consideration of the crystallization mechanism. As a result, they have found that, by intentionally making the light intensity distribution within the light beam width non-uniform, crystallization proceeds smoothly, and as a result, a high-quality crystalline thin film can be obtained. Based on such knowledge, the present invention having the following configuration has been completed.

That is, a non-single-crystal thin film formed on a substrate is irradiated with a light beam, whereby the non-single-crystal is crystallized or recrystallized. In the method for producing a crystalline semiconductor thin film to be a thin film, the light beam whose light energy intensity distribution pattern is adjusted so that a temperature gradient or a non-uniform temperature distribution is generated on the surface of the thin film that is the surface to be irradiated as the light beam. It is characterized by using a beam.

With this configuration, a temperature gradient or non-uniform temperature distribution occurs on the surface of the non-single-crystal thin film irradiated with the light beam, so that fine crystal nuclei are generated simultaneously in a wide range. The phenomenon described in 7f and g can be prevented. Therefore, relatively large crystal grains can be obtained, and the uniformity of crystallinity can be improved. As a result, the density of crystal grain boundaries decreases, and the field-effect mobility improves.

According to a forty-first aspect of the present invention, in the method for producing a crystalline thin film according to the forty-ninth aspect, the light intensity within the beam width monotonically increases from one side to the other or monotonically from one side to the other side. The method is characterized in that a light beam having a decreasing distribution pattern is used.

With this configuration, a temperature gradient is formed on the surface of the non-single-crystalline thin film, which is the surface to be illuminated, in accordance with the intensity of the light energy, and the crystallization proceeds in a direction from a low temperature to a high temperature. Be guided. Thus, generation of disordered crystal nuclei and disordered crystal growth are prevented, and the above-described FIGS.
Can be reliably prevented.

Here, the crystalline thin film is formed, for example, in the source region-
When used in a semiconductor circuit comprising a channel region and a drain region, an intensity gradient of light energy is preferably formed in a direction parallel to the source-drain direction.
In this way, the direction of crystal growth is regulated in a direction parallel to the direction of carrier movement, and the density of crystal grain boundaries in this direction decreases. Therefore, by adopting this method,
For example, a mobility of about 300 cm ² / Vs or more can be realized.

According to a forty-second aspect, in the method for manufacturing a crystalline thin film according to the forty-ninth aspect, the light beam includes a portion having a relatively high light intensity and a portion having a relatively low light intensity within a beam width. Are characterized by using light beams having distribution patterns alternately arranged in a plane.

When a light beam having a striped pattern composed of a portion having a high light intensity and a portion having a low light intensity is irradiated, a striped temperature distribution pattern composed of a high temperature portion and a low temperature portion can be formed on the irradiated surface. . In such a striped temperature distribution pattern, crystal growth is induced from a low-temperature portion (usually a strip) to a high-temperature portion.
Then, the crystal grains collide near the center of the high temperature portion (band), and a continuous line of the crystal grain boundary (continuous line like a mountain range) is formed here, and in a direction parallel to this continuous line. Slightly longer crystal grains are formed.

Therefore, even with this configuration, the phenomenon described with reference to FIGS. 7F and 7G can be prevented, and the same effect as that of the invention according to claim 41 can be obtained. That is, crystallization is performed by arranging a region having a relatively high light intensity and a region having a relatively low light intensity in parallel with the source-drain direction. In this case, the collision line of the crystal grain becomes parallel to the source-drain direction, and the collision line of the crystal grain (boundary line of the crystal grain boundary) causing the carrier to greatly reduce the mobility.
Will not cross. Therefore, a channel region having high mobility can be formed.

According to the present invention of claim 43, in the method for producing a crystalline thin film according to claim 42, the light beam is
It is characterized by using a light beam formed by simultaneously irradiating at least two mutually coherent lights to cause light interference.

With this configuration utilizing light interference, a fine light intensity distribution can be formed, and as a result, a fine striped temperature distribution can be formed on the irradiation surface. Therefore, according to this configuration, crystallization of a relatively wide region can be smoothly advanced.

According to a forty-fourth aspect of the present invention, in the method for manufacturing a crystalline thin film according to the forty-fourth aspect, the light beam is
A dynamic light interference pattern characterized by using a wave-like interference pattern formed by simultaneously irradiating at least two mutually coherent lights and dynamically modulating the phase of at least one of the lights. In this configuration using the method, the energy intensity distribution of the light beam changes in a wave-like manner, and the temperature of the irradiation surface changes in a wave-like manner in response to the change. Therefore, with this configuration, impurities contained in the amorphous thin film can be gradually driven out of the effective region, and as a result, a crystalline thin film having high purity and excellent mobility can be formed.

In the method for producing a crystalline thin film according to any one of claims 40 to 44, the light beam may be irradiated while relatively moving the non-single crystalline thin film on the substrate. A light beam whose light energy intensity distribution pattern is adjusted so as to cause a temperature gradient or a non-uniform temperature distribution on the irradiation surface (non-single-crystalline thin film surface) is irradiated while moving relatively to the thin film surface. With this configuration, the crystal growth direction can be guided finely. Therefore, a high-quality crystalline thin film having high crystallinity uniformity and a small grain boundary density in a certain direction can be obtained.

A forty-fifth aspect of the present invention is directed to a crystal for irradiating a non-single crystalline thin film formed on a substrate with a light beam and then radiating heat to crystallize or recrystallize the non-single crystalline. In the method for manufacturing a high quality semiconductor thin film, an uneven temperature distribution is generated on a thin film surface irradiated with a light beam by maintaining an ambient atmospheric pressure at a predetermined value or more.

With this configuration, when gas molecules constituting the atmospheric gas collide with the surface of the thin film and desorb, the heat of the thin film is taken away, and a locally low-temperature portion is formed. Therefore, a crystal nucleus is generated at this portion, and the crystal nucleus promotes the growth of the crystal, so that the phenomenon described with reference to FIGS. 7F and 7G can be prevented.

The invention according to claim 46 is the method for forming a crystalline thin film according to claim 45, wherein the atmospheric pressure equal to or higher than the predetermined value is 10-5 torr when the atmospheric gas is hydrogen gas.
r or more.

When the laser annealing treatment is performed in a hydrogen gas pressure of 10 -5 torr or more, the action and effect as described in claim 45 can be reliably obtained by the movement of hydrogen molecules having high specific heat.

According to another aspect of the present invention, there is provided a method for manufacturing a semiconductor film, wherein at least the precursor semiconductor film formed on a substrate is crystallized. A first energy beam that gives energy that can be converted to the precursor semiconductor film, and an energy that allows the precursor semiconductor film to have a lower absorptivity than the first energy beam and to crystallize the precursor semiconductor film. A step of irradiating the precursor semiconductor film with a second energy beam that gives smaller energy to the precursor semiconductor film to crystallize the precursor semiconductor film.

Thus, the second energy beam easily reaches the lower part of the precursor semiconductor film and the substrate, and the precursor semiconductor film is heated in the thickness direction and the substrate is also heated. The temperature difference between the irradiation of the energy beam and the end of the irradiation is reduced. Therefore, the precursor semiconductor film heated and melted by irradiation with the first energy beam is crystallized while being gradually cooled after the irradiation is completed. Therefore, crystal growth is promoted, relatively large crystal grains are formed, crystal defects are reduced, and electric characteristics of the semiconductor film are improved.

[0107] The invention of claim 48 is the method of manufacturing a semiconductor film of claim 47, wherein the precursor semiconductor film is an amorphous silicon thin film.

As a result, it is possible to easily form a polycrystalline silicon thin film having good crystal quality and good electric characteristics such as field effect mobility.

The invention of claim 49 is the method of manufacturing a semiconductor film of claim 47, wherein the first energy beam is
The absorption coefficient of the precursor semiconductor film is equal to or greater than the reciprocal of the thickness of the precursor semiconductor film, and the second energy beam has an absorption coefficient of the precursor semiconductor film of the precursor semiconductor film. It is characterized in that it is not more than the reciprocal of the film thickness.

As a result, while many first energy beams are absorbed near the surface of the precursor semiconductor film, many second energy beams reach the lower part of the precursor semiconductor film and the substrate. The body semiconductor film is efficiently heated, the substrate is also heated, and after the first energy beam irradiation is completed, the substrate is gradually cooled to promote crystal growth, so that relatively large crystal grains are reliably formed. Thus, a semiconductor film having good crystal quality is formed.

The invention of claim 50 is the method of manufacturing a semiconductor film of claim 47, wherein the first energy beam is
The absorption coefficient of the precursor semiconductor film is approximately 10 times or more the reciprocal of the thickness of the precursor semiconductor film, and the second energy beam has an absorption coefficient of the precursor semiconductor film of the precursor semiconductor film. It is characterized in that the thickness is substantially the reciprocal of the thickness of the semiconductor film.

As a result, the precursor semiconductor film is more efficiently heated, and a semiconductor film having better crystal quality is formed.

The invention of claim 51 is the method of manufacturing a semiconductor film of claim 47, wherein the first energy beam and the second energy beam are light beams having different wavelengths. .

Thus, the difference in the absorptance as described above can be easily provided.

In the light having different wavelengths as described above, for example, the first energy beam is a single-wavelength energy beam, and the second energy beam is
Light containing at least a wavelength component in the visible light region can be applied.

More specifically, as the first energy beam and the second energy beam, for example, laser light and infrared lamp, laser light and incandescent light, or laser light and excimer lamp light are used. be able to. Further, as the light having different wavelengths as described above, for example, the second energy beam may be light including a wavelength component in at least a visible light region to an ultraviolet light region, such as a xenon flash lamp light.

Further, the first energy beam and the second energy beam may be laser beams.

That is, by using a laser beam, it is possible to easily irradiate an energy beam having a large energy density, so that the precursor semiconductor film and the substrate can be easily heated efficiently.

More specifically, for example, when the precursor semiconductor film is an amorphous silicon thin film, the first energy beam is an argon fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser, or Any of a xenon fluorine excimer laser and a laser beam of an argon laser can be used as the second energy beam.

Further, when the substrate is a glass substrate and the precursor semiconductor film is an amorphous silicon thin film, the first energy beam is an argon fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser. A laser beam of any of a laser and a xenon fluorine excimer laser, and a laser beam of a carbon dioxide laser can be used as the second energy beam.

Each of the above excimer lasers easily obtains a large output and is easily absorbed in the vicinity of the surface of the amorphous silicon thin film. The amorphous silicon thin film is easily absorbed in the thickness direction of the amorphous silicon thin film, and the laser beam of the carbon dioxide gas laser passes through the amorphous silicon thin film relatively well and is easily absorbed by the glass substrate. Heating can be performed well, and a polycrystalline silicon thin film having good crystal quality can be easily formed, and productivity can be easily improved.

The invention of claim 61 is the method of manufacturing a semiconductor film of claim 47, wherein the first energy beam and the second energy beam irradiate a strip-shaped region in the precursor semiconductor film. It is characterized by:

By irradiating the belt-shaped region in this manner, heating can be easily performed with a uniform temperature distribution, a semiconductor film having a uniform crystal quality can be easily formed, and the crystallization process can be easily performed. The required time can be easily reduced.

The invention of claim 62 is the method of manufacturing a semiconductor film according to claim 47, wherein the irradiation region of the precursor semiconductor film with the second energy beam is the same as the precursor region of the first energy beam. It is a region that is larger than the irradiation region on the body semiconductor film and includes the irradiation region of the first energy beam.

As a result, heating can be easily performed with a uniform temperature distribution, and a semiconductor film having uniform crystal quality can be easily formed.

The invention of claim 63 is the method of manufacturing a semiconductor film of claim 47, wherein the first energy beam and the second energy beam are incident almost perpendicularly on the precursor semiconductor film. Irradiation is performed.

As described above, since each energy beam is substantially perpendicularly incident on the precursor semiconductor film, the irradiation unevenness of each energy beam is reduced, so that a semiconductor film having uniform crystal quality can be easily formed. can do.

[0128] The invention of claim 64 is the method of manufacturing a semiconductor film of claim 47, wherein the second energy beam is irradiated at least prior to the irradiation of the first energy beam. And Irradiation of the second energy beam prior to the first energy beam,
In addition to controlling the irradiation timing of each energy beam, for example, while moving the substrate on which the precursor semiconductor film is formed, the second energy beam is used for the second energy beam in the precursor semiconductor film. 1
It can also be performed by, for example, irradiating a position on the forward side in the moving direction with respect to the irradiation position of the energy beam.

By performing such irradiation, crystallization is performed by the first energy beam in a state where the semiconductor film or the substrate is sufficiently heated by the second energy beam. It can be carried out.

The invention of claim 66 is the method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam is
While irradiating intermittently, the second energy beam is
It is characterized by continuous irradiation. In particular,
For example, pulse oscillation laser light can be used as the first energy beam, and continuous oscillation laser light or lamp light can be used as the second energy beam.

As described above, by continuously irradiating the second energy beam, the substrate and the precursor semiconductor film can be easily heated to a predetermined stable temperature, and the first energy beam can be intermittently irradiated. By irradiating the substrate, it is possible to easily conduct crystallization of the precursor semiconductor film while suppressing transmission of heat to the substrate, preventing melting and distortion caused by excessive heating of the substrate. .

The invention of claim 69 is the method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam are intermittently irradiated in synchronization with each other. Features. As a specific irradiation timing, for example, the period of irradiating the first energy beam is within the period of irradiating the second energy beam, and is 分 の of the irradiation period of the second energy beam. It is preferable to set the period to 2 or less. Specifically, each energy beam uses a pulsed laser beam as the first energy beam, and uses a pulsed laser beam or a lamp light that is intermittently turned on as the second energy beam. be able to.

As described above, by intermittently irradiating the first energy beam and the second energy beam, it is possible to easily irradiate a large amount of light per unit area and to melt the substrate by excessive heating. Heating can be performed by applying a large amount of energy while preventing generation of heat or strain, so that crystallization of the precursor semiconductor film can be easily performed with certainty. In particular, a pulsed laser beam can easily be obtained with a high output and can be easily heated to a high temperature over a wide area, so that the time required for the crystallization step can be shortened and the productivity can be easily improved. it can.

The invention of claim 73 or claim 74 is
48. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam are formed such that the precursor semiconductor film has a temperature of 300C to 1200C, more preferably 600C to 1100C. Irradiation is performed so as to be heated to the following temperature.

By heating the precursor semiconductor film to a temperature in such a range, it is possible to prevent crystal defects and uneven crystallization due to the generation of partially fine crystals, and to change the temperature during crystallization. To promote crystal growth,
Large crystal grains can be easily formed.

[0136] The invention of claim 75 is the method of manufacturing a semiconductor film of claim 47, further comprising a step of heating the substrate on which the precursor semiconductor film is formed by a heater. Specifically, for example, it is preferable to heat the substrate on which the precursor semiconductor film is formed to a temperature of 300 ° C. or more and 600 ° C. or less.

As described above, by heating the substrate with the heater in addition to the second energy beam, the precursor semiconductor film can be further efficiently heated, and the crystal growth is promoted by gradual heating. Can be done easily. Moreover, since the substrate can be heated to a predetermined temperature in a short time as compared with the case where the substrate is heated only by a conventional heater, productivity can be easily improved.

[0138] The invention of claim 77 is the method of manufacturing a semiconductor film of claim 47, wherein the first energy beam is:
In addition to irradiating a plurality of regions in the precursor semiconductor film, the second energy beam irradiates only a part of the plurality of regions.

By partially irradiating the second energy beam as described above, for example, the crystallinity can be improved only in a region where particularly high electric characteristics are required. The necessary and sufficient crystallization can be easily performed to improve the productivity.

The invention of claim 78 is the method of manufacturing a semiconductor film according to claim 47, wherein the second energy beam is
An absorptance in the substrate is larger than an absorptivity in the precursor semiconductor film. Further, the first energy beam preferably has an absorption coefficient of the precursor semiconductor film that is at least about 10 times the reciprocal of the thickness of the precursor semiconductor film.

More specifically, for example, when the substrate is a glass substrate and the precursor semiconductor film is an amorphous silicon thin film, the first energy beam is an argon fluorine excimer laser, a krypton fluorine excimer laser, or the like. A laser beam of any of a laser, a xenon chlorine excimer laser, and a xenon fluorine excimer laser, and a laser beam of a carbon dioxide gas laser can be used as the second energy beam.

Thus, while many first energy beams are absorbed near the surface of the precursor semiconductor film, many second energy beams are absorbed by the substrate, so that the precursor semiconductor film is efficiently heated. At the same time, the substrate is also heated, and after the irradiation of the first energy beam is completed, the substrate is gradually cooled to promote crystal growth, so that relatively large crystal grains are reliably formed, and a semiconductor with good crystal quality is formed. A film is formed.

An invention according to claim 81 is an apparatus for manufacturing a semiconductor film for crystallizing a precursor semiconductor film formed on a substrate, comprising: a first irradiation means for irradiating a first energy beam; And a second irradiating means for irradiating the first energy beam with a second energy beam having a lower absorption rate of the precursor semiconductor film.

As a result, the second energy beam easily reaches the lower portion of the precursor semiconductor film and the substrate, and the precursor semiconductor film is heated in the thickness direction and the substrate is also heated. The temperature difference between the irradiation of the energy beam and the end of the irradiation is reduced. Therefore, the precursor semiconductor film heated and melted by irradiation with the first energy beam is crystallized while being gradually cooled after the irradiation is completed. Therefore, crystal growth is promoted, relatively large crystal grains are formed, crystal defects are reduced, and a semiconductor film with improved electrical characteristics can be manufactured.

An invention according to claim 82 is the apparatus for manufacturing a semiconductor film according to claim 81, wherein the second irradiating means is a lamp which emits a second energy beam in a radial manner, and further comprises the second irradiating means. And a concave reflector for condensing the energy beam.

Accordingly, the substrate and the like can be efficiently heated, and the temperature distribution can be made uniform to easily form a semiconductor film having uniform crystal quality.

An invention according to claim 83 is the apparatus for manufacturing a semiconductor film according to claim 81, further comprising reflecting one of the first energy beam and the second energy beam and the other. And a reflecting plate that transmits the first energy beam and the second energy beam.
Both are characterized in that they are configured to be incident on the precursor semiconductor film almost perpendicularly.

As described above, since each energy beam is substantially perpendicularly incident on the precursor semiconductor film, irradiation unevenness of each energy beam is reduced, so that a semiconductor film having uniform crystal quality can be easily formed. can do.

The first irradiation means as described above, and the second irradiation means
Specifically, for example, when the precursor semiconductor film is an amorphous silicon thin film, the first irradiation means is an argon fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser. Or the xenon fluorine excimer laser, an argon laser can be used as the second irradiation means.

When the substrate is a glass substrate and the precursor semiconductor film is an amorphous silicon thin film, the first energy beam is an argon fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser. A laser beam of any of a laser and a xenon fluorine excimer laser, and a laser beam of a carbon dioxide gas laser can be used as the second energy beam.

Further, in order to solve the above-mentioned problem, the invention of claim 86 relates to a method of irradiating a non-single-crystal semiconductor thin film formed on a substrate having an image display area and a drive circuit section area with an energy beam. A method of manufacturing a semiconductor thin film including a step of growing a crystal, wherein the first irradiation on the image display area is performed using an energy beam having a linear beam cross-sectional shape, while the first irradiation on the drive circuit section area is performed. The second irradiation uses an energy beam having a square cross section, and
The first irradiation is performed at a higher energy density than the first irradiation.

The invention of claim 87 is directed to a semiconductor thin film having a step of irradiating a non-single-crystal semiconductor thin film formed on a substrate having an image display region and a drive circuit region with an energy beam to grow a crystal. In the manufacturing method, the first irradiation on the image display area is performed by scanning an energy beam relative to the substrate and irradiating the energy beam while shifting the irradiation area of the energy beam by a predetermined overlap amount. On the other hand, the second irradiation on the drive circuit section area
It is characterized in that stationary irradiation is performed with an energy beam fixed relative to the substrate, and the energy irradiation is performed at a higher energy density than in the first irradiation.

Specifically, for example, in a thin film transistor constituting a liquid crystal display device, a pixel portion in which uniformity of semiconductor film characteristics is required and a drive circuit portion in which characteristics (particularly, high mobility) are required, are formed by a laser. Use different irradiation methods. In other words, when laser light is irradiated to amorphous silicon formed on a substrate to melt and crystallize the amorphous silicon to form polycrystalline silicon, a driving circuit portion in the substrate surface is used. Laser annealing is performed by setting the energy density of the laser light applied to the area higher than the energy density of the laser light applied to the pixel area, thereby forming polycrystalline silicon having different characteristics between the drive circuit area and the pixel area. Things. More specifically, for example, after the first laser light irradiation is performed only on the pixel portion region or on the entire surface of the substrate, the driving circuit region is irradiated with the first laser light more than the first laser light irradiation. A second laser beam irradiation with a high energy density is performed.

According to this structure, while the mobility of the polycrystalline silicon in the drive circuit portion region is higher than the mobility of the polycrystalline silicon in the pixel portion region, the characteristics of the polycrystalline silicon in the pixel portion region are reduced in-plane. Can be made uniform.

Further, the stage for fixing the substrate is rotated by 90 degrees by making the laser beam for the first laser beam irradiation linear and the laser beam for the second laser beam irradiation square. Laser annealing can be performed without the need.

Further, the first laser light irradiation is scanning irradiation in which the irradiation position of the laser beam is shifted a plurality of times while being shifted, and the second laser light irradiation is stationary irradiation in which the irradiation position of the laser beam is fixed. Thereby, the mobility of the polycrystalline silicon in the drive circuit section region can be increased and uniformity can be achieved.

It is also possible to irradiate a plurality of regions in the drive circuit portion region with laser beams having different energy densities to perform laser annealing, thereby forming polycrystalline silicon having different characteristics in the drive circuit portion region. It is. In this case, it is preferable to perform laser annealing by irradiating a laser beam having a different energy density in a region where a transfer gate is formed in a latch or a shift register from another region.

In the above laser annealing method, it is preferable to irradiate a laser beam so that the laser beam end does not come on the TFT pattern.

The apparatus for producing a semiconductor thin film of the present invention
An energy beam generating means; and a uniformizing means for shaping the energy beam emitted from the energy beam generating means into a predetermined beam cross-sectional shape having uniform energy, wherein the shaped energy beam is formed on a substrate. A semiconductor thin film manufacturing apparatus for irradiating the non-single-crystal semiconductor thin film to grow a crystal, further comprising a filter having regions in which the energy beam transmittances are different from each other; A plurality of regions in the semiconductor thin film are irradiated with the energy beam at different energy densities.

With this configuration, it is possible to form a plurality of polycrystalline semiconductor films having different characteristics on the same substrate surface.

In the above configuration, by using a laser annealing apparatus in which the transmittance of the mask is changed by the optical thin film or the like, it is possible to accurately create the transmittance distribution, and to make the mask and the laser beam By using a laser annealing apparatus in which the window for irradiating the substrate in the processing chamber is the same, the structure of the apparatus is simplified, and
It is possible to reduce the attenuation of the light amount.

Further, there is provided an energy beam generating means, and a uniformizing means for shaping the energy beam emitted from the energy beam generating means into a predetermined beam cross section having uniform energy. An apparatus for manufacturing a semiconductor thin film for irradiating a non-single-crystal semiconductor thin film formed on a substrate to grow a crystal, wherein the uniformizing means can selectively switch an energy beam into a plurality of beam cross-sectional shapes to shape the beam. It is characterized by having such a configuration.

With this configuration, it is possible to irradiate a laser beam having an optimal shape to each location on the substrate.

[0164]

DESCRIPTION OF THE PREFERRED EMBODIMENTS (Embodiment 1-1) An example in which regions having different thermal conductivities are provided on a substrate, a semiconductor thin film is provided with a temperature distribution, and crystal growth is controlled with reference to FIG. I do.

As shown in FIG. 12C, a lower insulating film 202 is formed on the entire surface of a transparent insulating substrate 201 such as a glass substrate. On the lower insulating film 202, a stripe-shaped upper insulating film 203 made of a material having a lower thermal conductivity than the lower insulating film 202 is partially formed. Further, an amorphous silicon thin film 204 is formed on the lower insulating film 202 and the upper insulating film 203.
Are formed.

The amorphous silicon thin film 204 shown in FIG.
By irradiating a linear laser beam having an energy density distribution in the x and y directions as shown in FIG. At this time, since the thermal conductivity of the upper insulating film 203 is lower than the thermal conductivity of the lower insulating film 202 as described above, as shown in FIG. The temperature in the upper region is higher than that in the region between the upper insulating films 203. Therefore, the crystallization of the amorphous silicon thin film 204 starts from a region between the upper insulating films 203 and grows toward a region on the upper insulating film 203. therefore,
In the region between the upper insulating films 203, collision between crystal grains hardly occurs, and the large crystal region 210b having relatively large crystal grains is formed.
On the other hand, in the region above the upper insulating film 203, crystal grains grown from both sides of the upper insulating film 203 collide with each other, so that a small crystal grain region 210a is formed.

The degree of crystallization of the polycrystalline silicon thin film 210 formed as described above and the polycrystalline silicon thin film polycrystallized by the conventional method were compared by the peak intensity of Raman spectroscopic analysis. FIG. 13 shows the results.

Here, in polycrystalline silicon thin film 210 of the present embodiment, lower insulating film 202 has a thickness of 200
nm silicon nitride thin film (thermal conductivity: 0.19 W / cm
° C.), a silicon oxide thin film having a thickness of 30 nm, a width of about 5 μm, and an interval of 20 μm (thermal conductivity:
0.014 W / cm · ° C.). On the other hand, in the conventional method, a single-layer silicon oxide thin film having a thickness of 200 nm is used as the insulating film. The thickness of the amorphous silicon thin film is 85 nm in both cases. The measurement location of the Raman peak intensity is the center of the irradiation area in the x direction in FIG.

As is apparent from FIG. 13, the crystallinity is small as a whole in the conventional method, whereas in the method of the present invention, the portions on the upper insulating film 203 indicated by A, B and C are removed. Although the crystallinity is low, the Raman peak intensity of the portion on the lower insulating film 202 sandwiched between the upper insulating films 203 is large, and it is recognized that the crystallinity is greatly improved.

The interval between the stripe patterns of the upper insulating film 203 is determined by the lower insulating film 202 and the upper insulating film 20.
Although the optimum value changes in accordance with the thermal conductivity of No. 3, the irradiation energy density, and the like, in the above example, 5 to 50 μm, more preferably 10 to 30 μm, was desirable as the range in which a large crystal can be stably obtained. .

In the above description, y in FIG.
In the example described above, the surface of the silicon thin film has a temperature distribution with respect to the direction. However, when the laser beam is irradiated with the laser beam stationary, the temperature distribution may be similarly provided with respect to the x direction. In the case of scanning the laser beam in the x-direction, the influence on the temperature distribution due to the sequential movement of the irradiation region may be considered. Further, the temperature distribution may be adjusted by utilizing the difference in thermal conductivity as described above and further by making the energy density distribution of the laser beam different for each region.

In the above example, the thermal conductivity of the upper insulating film 203 is lower than that of the lower insulating film 202, and the crystal grain size in the region where the upper insulating film 203 does not exist is increased.
Conversely, the thermal conductivity of the upper insulating film may be made larger than that of the lower insulating film to increase the crystal grain size in the region where the upper insulating film is formed. However, the former generally has higher thermal conductivity (the surface temperature of the silicon thin film is lower)
Since it is easy to increase the area of the region, it is easy to increase the temperature gradient in the temperature distribution on the silicon surface.

The relationship between the magnitude of the thermal conductivity and the vertical relationship of the lamination is not limited to the above, but may be reversed, as long as a predetermined temperature distribution is formed.

When the insulating film has a two-layer structure as described above, the desired shape (thickness) can be obtained by setting a large etching selectivity (ratio of etching rate) between the upper insulating film and the lower insulating film. ), The upper insulating film can be formed easily with a uniform thickness over a large area, and as a result, a polycrystalline silicon thin film having a uniform grain size can be easily obtained over the entire surface of the substrate. Can be. On the other hand, in order to provide regions having different thermal conductivities, for example, the thickness of the silicon thin film may be changed by etching or the like. In this case, the accuracy of the etching process needs to be relatively high, but there is no need to form two insulating films as described above, so that the manufacturing process can be simplified.

Further, instead of making the thermal conductivity different as described above, the temperature distribution may be generated by forming regions having different heat capacities, thereby similarly improving the crystallinity.

(Embodiment 1-2) An example of a polycrystalline silicon thin film transistor formed using the semiconductor thin film formed as described above will be described.

FIG. 14A is a plan view of a polycrystalline silicon thin film transistor, and FIG. 14B is a cross-sectional view taken along the line AA ′ in FIG. 14, 201 is a transparent insulating substrate, 202 is a lower insulating film, 203 is an upper insulating film, 205 is a gate insulating film, 206 is a source electrode film,
207 is a drain electrode film, 208 is a gate electrode film, 21
0b is a large grain region 210 of the polycrystalline silicon thin film 210.
b. That is, this polycrystalline silicon thin film transistor uses only the large crystal grain region 210b of the region sandwiched by the upper insulating film 203 in the polycrystalline silicon thin film 210 polycrystallized as described in Embodiment 1-1. The upper insulating film 203 is formed so that it is selectively used while being left by etching or the like, and the direction of the source-drain is parallel to the direction of the stripe pattern of the upper insulating film 203. As a method for forming the gate insulating film 205, the source electrode film 206, the drain electrode film 207, and the gate electrode film 208, a thin film deposition and patterning method similar to a conventional thin film transistor can be applied.

The polycrystalline thin film transistor thus obtained has a field effect mobility of about 180 cm ² / V · se.
c, and the field effect mobility of the transistor manufactured by the conventional method is 70 cm ² / V · sec.
The TFT characteristics were significantly improved.

The direction of the upper insulating film 203 and the source
The relationship with the direction of the drain is not limited to the one as described above, and the long direction of the crystal grains formed according to the distance between the upper insulating films 203 and the like may be the direction of the source-drain. .

(Embodiment 1-3) Embodiment 1-
An example of forming a polycrystalline silicon thin film transistor having a size larger than 2 will be described.

This polycrystalline silicon thin film transistor
As shown in FIG. 15, the present embodiment is different from the above-described Embodiment 1-2 in that it is formed using two large crystal grain regions 210b formed between three upper insulating films 203. As the lower insulating film 202, a silicon nitride oxide thin film formed by plasma CVD and having a thickness of about 200 nm is used. As the upper insulating film 203, a silicon oxide thin film having a thickness of about 40 nm is used. In the large crystal grain region 210b, an amorphous silicon thin film 204 having a thickness of 85 nm is formed on the upper insulating film 203 formed by patterning, and irradiated with excimer laser light in the same manner as in Embodiment 1-1. To form a polycrystalline silicon thin film.

That is, if the interval between the upper insulating films 203 is increased in order to increase the size of the transistor, the difference between the transistor formation region and the region around the transistor formation region on the surface of the silicon thin film when performing the polycrystallization process. Therefore, it may be difficult to sufficiently increase the temperature gradient between them, and as a result, it may not be possible to sufficiently increase the crystal grain size of silicon in the transistor formation region. Therefore, the temperature gradient is positively increased without setting the interval between the upper insulating films 203 wide as described above, and a plurality of large crystal grain regions 210b in a favorable crystal state are formed, and by combining them, A thin film transistor having a large size and good characteristics can be formed. Specifically, for example, the field effect mobility is about 200 cm.
A thin film transistor having excellent characteristics of ² / V · sec was obtained.

As described above, in the method of manufacturing a polycrystalline silicon thin film transistor according to the present invention, only a region where a transistor is to be formed can be made to have large crystal grains. However, as an insulating film formed on a transparent insulating substrate, The material is not limited to silicon nitride, silicon nitride oxide, and silicon oxide, and is not particularly limited as long as it is a combination having different thermal conductivities and capable of selective etching.

(Embodiment 2-1) An example of a thin film transistor as a semiconductor element having a large crystal grain will be described as the semiconductor element of the embodiment 2-1.

FIGS. 16A and 16B are schematic views of a thin film transistor. FIG. 16A is a plan view, and FIG.
FIG. 3A is a cross-sectional view taken along the line AA ′ in FIG.

Referring to FIG. 16, reference numeral 301 denotes an insulating substrate, and the undercoat layer 3 is provided above the insulating substrate 301.
02, and a semiconductor layer 303 formed by crystallizing an amorphous semiconductor film made of Si is provided thereabove.
As shown in FIG. 16A, the semiconductor layer 303
The semiconductor layer 303 is provided on a pair of opposite sides of the semiconductor layer 303.
A plurality of protrusions 303a extending outward in the same plane are formed at predetermined intervals. The protrusion 303
a is formed in a substantially rectangular shape, and its length (the semiconductor layer 3
03 and the width (length in the direction perpendicular to the protrusion length) are set to 1 μm. Further, a first insulating layer 304 is provided above the semiconductor layer 303 so as to cover the semiconductor layer 303, and a gate electrode 305 serving as a first electrode is provided at a predetermined position on the first insulating layer 304. Have been. Then, the second electrode is formed so as to cover the gate electrode 305.
Is provided at a predetermined position on the second insulating layer 306, a source electrode 307 s and a drain electrode 30, which are a pair of second electrodes electrically contacting the semiconductor layer 303.
7d is provided.

Here, the width of the projection 303a is 1 μm.
The thickness of the semiconductor layer 303 is not less than 0.05 μm (for example, 0.05 μm) in order to further match the crystal grain diameter and generate one crystal nucleus for each protrusion 303 a. It is desirable that the thickness be about 3 μm or less. The technical reason for adopting the above numerical range is that, when the width of the protrusion 303a is smaller than the film thickness, the crystal nucleus generated in the protrusion 303a is subjected to the effect of surface tension, is drawn into the semiconductor layer 303, and On the other hand, if the width of the projection 303a is larger than 3 μm, the projection 303
This is because two or more crystal nuclei may be generated in a. The shape of the protrusion 303a is not limited to a rectangle, but may be another shape such as a semicircle or a triangle. The protrusion 303a is not limited to be formed over the entire length of the opposing side of the semiconductor layer 303.
05 may be formed only in the portion corresponding to
What is necessary is just to form in the channel part which affects the characteristic of an element. Further, it may be formed so as to be located near the middle between the source and the drain. Further, the interval between the adjacent protrusions 303a can be appropriately selected depending on conditions such as a desired particle size. In the present embodiment, the interval between the protrusions 303a is used as the interval between the protrusions 303a.
Is set to be substantially equal to the length (w) of the side orthogonal to the side where is provided. It should be noted that such a setting is preferable in that large crystal grains having substantially the same length in the vertical and horizontal directions are easily formed. However, even when the setting is not performed, crystal growth is regularly performed from the peripheral portion. By doing so, the effect of forming relatively large crystal grains can be obtained.

Since the semiconductor layer 303 is irradiated with a laser beam and heated by the formation of the projections 303a as described above, the projections 303a are cooled earlier and crystal nuclei are generated. The crystal grows from the crystal nucleus toward the center of the semiconductor layer 303. Further, at this time, relatively large crystal grains are formed because the crystal grains grown from the adjacent projections 303a and the projections 303a on the opposite side easily grow to the vicinity of the center of the semiconductor layer 303 without interfering with each other. Is done. Therefore, the field effect mobility can be increased and the TFT characteristics can be easily improved.

Next, a method of manufacturing the above thin film transistor will be described with reference to FIG. Figure 1
8 is a process chart showing a method for manufacturing a thin film transistor.

First, as shown in FIG. 18A, an undercoat layer 302 is formed on an insulating substrate 301, and silicon is deposited on the undercoat layer 302 to form an amorphous (non-single crystal) ) Is formed. next,
A photoresist (not shown) is selectively formed on the semiconductor layer 303 in a predetermined shape, and using this photoresist as a mask, as shown in FIG. The photoresist is formed in a shape having a protrusion 303a extending in the same plane over the entire length, and then the photoresist is removed.

Next, as shown in FIG. 18B, the amorphous semiconductor layer 303 is crystallized by irradiating an excimer laser beam as an energy beam to form a poly-Si modified layer. Here, after the irradiation with the laser beam, the heat accumulated in the protrusions 303a on the peripheral edge is diffused in three directions on the outside in a plane parallel to the semiconductor layer 303, whereas the heat accumulated in the center is still diffused. Since there is only a relief area on the peripheral side that has not been cooled, the peripheral part including the protrusion 303a is cooled sufficiently faster than the central part. Therefore, the crystal nucleus in the protrusion 303a occurs earlier than the crystal nucleus in the central part, and before the crystal nucleus is generated or crystal-grows in the central part, the crystal nucleus generated on the periphery is directed toward the central part. Since the crystal grows, the crystal grain size and crystal orientation can be controlled. This prevents interference between crystals in the course of crystal growth and facilitates obtaining a sufficient crystal grain size.

Subsequently, as shown in FIG. 18C, a first insulating layer 304 is formed on the semiconductor layer 303 and the undercoat layer 302, and a first electrode is formed on the first insulating layer 304. A certain gate electrode 305 is selectively formed.

Thereafter, as shown in FIG. 18D, the semiconductor layer 3 is formed using the gate electrode 305 as a mask.
The source region 303s and the drain region 303d are formed by adding an impurity serving as a donor or an acceptor to the substrate 03 by ion implantation or ion doping without mass separation.

Finally, as shown in FIG.
After forming the insulating layer 306, a contact hole is opened, and a source electrode 307s and a drain electrode 307d are selectively formed to obtain a thin film transistor.

In the above example, Si is used for the semiconductor layer 303, but another material such as a compound of Si and Ge may be used. Further, another combination of the IV groups such as SiC, a combination of the III group and the V group such as GaAs, or a combination of the II group and the VI group such as CdSe may be used. In addition, although a polycrystalline silicon thin film transistor has been described as an example, the present invention is not limited to this, and it is of course possible to apply the present invention to various other semiconductor elements.

Further, an excimer laser was used as an energy beam when the amorphous semiconductor layer 303 was polycrystallized. However, other energy beams such as a laser beam such as an Ar laser and a YAG laser, an ion beam, and an electron beam were used. Beams and the like can also be used.

(Embodiment 2-2) An example of an inverted staggered thin film transistor will be described as a semiconductor element of Embodiment 2-2.

FIG. 17 is a schematic view of a thin film transistor, FIG. 17 (a) is a plan view, and FIG.
FIG. 3A is a cross-sectional view taken along the line AA ′ in FIG.

This thin film transistor has an inverted staggered structure as compared with the embodiment 2-1.
The difference is that the protrusion 303 a is formed over the entire circumference of the semiconductor layer 303.

In FIG. 17, reference numeral 301 denotes an insulating substrate, and the undercoat layer 3 is provided above the insulating substrate 301.
02, a gate electrode 305 as a first electrode is provided thereabove. Further, a first covering the gate electrode 305
Is provided, and a semiconductor layer 303 is provided over the first insulating layer 304. In the semiconductor layer 303, as shown in FIG. 17A, a plurality of protrusions 303a extending outward in the same plane as the semiconductor layer 303 are formed at predetermined intervals all around the semiconductor layer 303. Have been.
The shape and the like of the projection 303a are the same as those in Embodiment 2-1. Here, in the same figure, for convenience, the interval between the respective projections 303a is drawn narrower, however, in Embodiment 2-1.
It is preferable that the width is set to be substantially equal to the width of the semiconductor layer 303 in the same manner as described above. However, as shown in FIG.
Conversely, even when the interval is long, the effect of forming relatively large crystal grains can be obtained by regularly growing crystals from the periphery. A source electrode 307 s and a drain electrode 307 d which are a pair of second electrodes which are in electrical contact with the semiconductor layer 303 are formed over the semiconductor layer 303.

Next, a method of manufacturing the above-described thin film transistor will be described with reference to FIG. Figure 1
9 is a process chart showing a method for manufacturing a thin film transistor.

First, as shown in FIG. 19A, an undercoat layer 302 is formed on an insulating substrate 301, and a gate electrode 305 as a first electrode is selectively formed on the undercoat layer 302.

Next, as shown in FIG. 19B, a first insulating layer 304 is formed on the gate electrode 305 and the undercoat layer 302, and silicon is deposited on the first insulating layer 304. Then, the amorphous (non-single-crystal) semiconductor layer 3
03 is formed. Next, a photoresist (not shown) is selectively formed on the semiconductor layer 303 in a predetermined shape, and the photoresist is used as a mask to completely cover the amorphous semiconductor layer 303 as shown in FIG. The photoresist is formed in a shape having a protrusion 303a extending in the same plane over the circumference, and then the photoresist is removed.

Next, as shown in FIG. 19 (c), the amorphous semiconductor layer 303 is irradiated with an excimer laser beam as an energy beam to be crystallized to obtain a poly-Si modified layer. Here, by forming the projections 303a as described above, it is easy to obtain a sufficient crystal grain size, as described in Embodiment 2-1.

Thereafter, as shown in FIG. 19D, a resist 308 as a mask for doping is selectively formed on the semiconductor layer 303 in a predetermined shape, and ion implantation is performed on the semiconductor layer 303 using the resist 308 as a mask. The source region 303s and the drain region 303d are formed by adding an impurity serving as a donor or an acceptor by an implantation method or an ion doping method without mass separation, and thereafter, the resist 308 is removed.

Finally, as shown in FIG. 19E, a source electrode 307s and a drain electrode 307d are selectively formed to obtain a thin film transistor.

[0207] In the present embodiment 2-2, various modifications similar to those described in the embodiment 2-1 can be applied.

The same effect can be obtained by forming a staggered thin film transistor similar to that of the embodiment 2-1 without being limited to the above-mentioned inverted staggered thin film transistor. Further, instead of forming the protrusions 303a over the entire circumference of the semiconductor layer 303 as described above, the protrusions 303a may be formed only on opposing sides as in Embodiment 2-1.

(Embodiment 3-1) FIGS. 20 to 22
It will be described based on. First, the structure of a thin film transistor (TFT: Thin Film Transistor) according to the present embodiment will be described.

FIG. 20 is a schematic view showing the outline of a forward stagger type TFT 410. FIG. 20 (a) is a plan view of the TFT 410, and FIG. 20 (b) is AA ′ in FIG. 20 (a).
It is arrow sectional drawing. FIG. 21 shows B in FIG.
FIG. 4 is a cross-sectional view taken along the arrow B ′. As shown in FIG.
Reference numeral 410 denotes an undercoat layer 4 on the insulating substrate 401.
02, a p-Si film 403, a first insulating film 404,
A second insulating film 406 and three electrodes of a gate electrode 405, a source electrode 407s, and a drain electrode 407d are provided.

The insulating substrate 401 has a strain point 5
It is a glass substrate having a temperature of 93 ° C. and a thickness of 1.1 mm, and the undercoat layer 402 is a thin film made of, for example, SiO ₂ .
In addition, the p-Si film 403 is
The polycrystalline semiconductor layer main body formed by applying the method of the present invention above. The p-Si film 403 includes a channel region 403a, a source region 403b, and a drain region 40.
3c, and the source region 403b and the drain region 403c are located on both sides of the channel region 403a. The source region 403b and the drain region 4
03c is formed by doping impurity ions such as phosphorus or boron.

As a material of the p-Si film 403, for example, silicon (Si) or a compound of silicon and germanium (Ge) is used. Also, the p-Si film 403
Is preferably in the range of 200 ° to 1500 °, more preferably 300 ° to 1000 °. 2
When the thickness is less than 00 °, a problem occurs in the uniformity of the film thickness. When the thickness exceeds 1500 °, a problem of so-called photoconduction in which a current flows between the source and the drain due to light irradiation occurs. 300 to 1000 mm
Within the range, both uniformity of film thickness and photoconduction can be achieved at the same time.

Further, the channel region 403 shown in FIG.
The width of arrow a in the direction of arrow X is, for example, about 12 μm, and p-S
The width of the i film 403 in the direction of the arrow Y is, for example, about 14 μm.
m.

Here, in the channel region 403a,
As shown in FIGS. 20A and 21, a plurality of groove-shaped crystal growth direction control gaps 411 are formed in parallel with the direction connecting the source region 403b and the drain region 403c. The crystal growth direction control gap 411 has a semicircular shape at both ends in the longitudinal direction and a rectangular parallelepiped shape at the center, and a groove width at the center (a groove width in a direction orthogonal to the longitudinal direction) is about 1.
μm. However, the shape of the crystal growth direction control gap 411 is not particularly limited. For example, it may be formed in a rectangular shape in the direction from the source region 403b to the drain region 403c.

The crystal grains in the channel region 403a have a shape that is elongated in the direction of the source region 403b or the drain region 403c. Is configured. In the channel region 403a having such a polycrystalline structure, the density of crystal grain boundaries in the direction connecting the source region 403b and the drain region 403c is small, so that charge carriers can move at high speed.

The first insulating film 404 is an insulating film made of, for example, SiO ₂ , and is formed above the p-Si film 403 and the undercoat layer 402. Gate electrode 40
Reference numeral 5 is made of, for example, aluminum (Al) or the like, and is provided above the first insulating film 404 and at a position corresponding to the channel region 403a of the p-Si film 403. The second insulating film 406 is made of, for example, SiO ₂ , and is laminated above the first insulating film 404 and the gate electrode 405.

The source region 4 of the p-Si film 403 is formed on the first insulating film 404 and the second insulating film 406, respectively.
The contact holes 408 reaching the drain region 03b or the drain region 403c are formed. Source electrode 40
The drain electrode 7s and the drain electrode 407d are made of, for example, Al and are formed to be in contact with the source region 403b or the drain region 403c through the contact holes 408 and 408. The gate electrode 405, the source electrode 407s, and the drain electrode 407d are patterned into a predetermined shape at a portion other than the illustrated cross section,
The wiring pattern is configured.

Next, a method of manufacturing the TFT 410 according to the present embodiment will be described.

FIG. 22 is a schematic sectional view showing a manufacturing process of the TFT 410. First, as shown in FIG.
An undercoat layer 402 is formed over an insulating substrate 401 by a normal pressure CVD method. The thickness of the undercoat layer 402 is, for example, 3000 °.

[0220] A Si layer is formed on the undercoat layer 402 by, for example, a plasma CVD method, and a photoresist (not shown) is selectively formed in a predetermined shape on the Si layer. Next, after exposing using the photoresist as a mask, the photoresist is patterned into a predetermined shape by etching, and then the photoresist is removed.

As a result, the crystal growth direction control gap 4
An a-Si film 413 as a non-single-crystal semiconductor layer having 11 is formed. Here, the thickness of the a-Si film 413 is, for example, 650 °. In the case where the crystal growth direction control gap 411 is minutely formed, exposure using a high-precision photoresist and interference fringes of coherent light may be used.

Following the formation of the a-Si film 413, FIG.
As shown in (b), the entire surface of the a-Si film 413 is irradiated with an excimer laser for one shot, and
13 is heated and melted, and then left to cool. Thus, a p-Si film 403 as a crystalline semiconductor layer is formed.

Here, according to the crystallization method using an excimer laser, since the a-Si film 413 has a large absorption coefficient in the ultraviolet region, the temperature of the main body of the a-Si film 413 can be sufficiently increased. On the other hand, since the laser light is not absorbed in the crystal growth direction control gaps 411 from which a-Si has been removed, the temperature can be kept low. Therefore, in the cooling step, the temperature near the crystal growth direction control gap 411 (and the peripheral portion of the a-Si film 413) reaches the crystallization start temperature, and the first crystal nucleus is generated here. After that, crystal growth will be performed centering on this crystal nucleus, but as already explained,
The crystal growth direction is regulated by the crystal growth direction control gaps 411 provided in parallel and guided in a direction connecting the source region 403b and the drain region 403c. As a result, a p-Si film having a small grain boundary density in a direction connecting the source region 403b and the drain region 403c is formed.

As an irradiation condition of the energy beam, for example, in the case of an excimer laser such as XeCl (wavelength 308 nm), a laser light pulse of 50 ns whose cross-sectional shape is a square of several millimeters on one side is used. The energy density of laser light (irradiation energy per unit area: mJ / cm ² )
The temperature may be set appropriately so that 13 can be heated to a temperature suitable for crystallization.

Note that the excimer laser is X
In addition to eCl, an excimer laser such as ArF, KrF, or XeF may be used. A plurality of crystal growth direction control gaps 4
11 can be appropriately set in consideration of the thickness of the a-Si film, irradiation conditions, and a desired moving speed of the charge carrier. In this embodiment, the distance is set to about 2 μm. is there. In addition, the width of the crystal growth direction control gap 411 can be appropriately set according to the thickness of the a-Si film, the type and intensity of the energy beam to be irradiated, and the like. 1 μm
And

After the above crystallization, as shown in FIG. 22C, a first insulating film 40 is formed on the p-Si film 403.
4 is formed by a normal pressure CVD method so that the film thickness becomes 1000 °. Further, on the first insulating film 404, for example, Al
The film is sputtered to a thickness of 2000
The gate electrode 405 and the wiring pattern are formed by patterning into a predetermined shape by performing wet etching using an etchant for about 1 minute.

Next, as shown in FIG. 22D, a p-Si film 403 is formed on the p-Si film 403 using the gate electrode 405 as a mask.
Impurity ions serving as donors or acceptors, specifically, impurity ions such as phosphorus or boron are implanted by an ion implantation method or an ion doping method without mass separation. Thus, a channel region 403a, a source region 403b, and a drain region 403c are formed in the p-Si film 403.

[0228] As further shown in FIG. 22 (e), on the gate electrode 405, is deposited for example the second insulating film 406 made of SiO _2, to a thickness 5000Å at atmospheric pressure CVD . Subsequently, contact holes 408, 408 reaching the source region 403b or the drain region 403c of the p-Si film 403 are opened in the first insulating film 404 and the second insulating film 406, respectively. Then, Al
After the film is sputtered so as to have a film thickness of 3000 ° and 3000 °, respectively, the film is patterned into a predetermined shape by, for example, dry etching using a BCl 3 / Cl 2 gas. Thus, the source electrode 407s and the drain electrode 407d and their wiring patterns are formed.

According to Embodiment 3-1 described above, large crystal grains having a long shape can be formed in the direction connecting source region 403b and drain region 403c. Type TFT is obtained. In this embodiment, the insulating substrate 401 or p
-Since an expensive material is not used for the Si film 403, a TFT having excellent field-effect mobility can be provided at low cost.

(Embodiment 3-2) An embodiment 3-2 according to the present invention will be described with reference to FIGS. Note that among the constituent elements of the thin film transistor according to Embodiment 3-2, the same components and functions as those of Embodiment 3-1 are denoted by the same names and reference numerals, and detailed description is omitted.

FIG. 23 is a schematic diagram schematically showing an inverted staggered TFT 420 according to Embodiment 3-2, and FIG. 23A is a plan view of the TFT 420.
FIG. 23B is a cross-sectional view taken along the line AA ′ in FIG. FIG. 24 is a cross-sectional view taken along the line BB ′ in FIG. As shown in FIG.
0 denotes an undercoat layer 402 on an insulating substrate 401.
, P-Si film 403, first insulating film 404, gate electrode 405, source electrode 407 s and drain electrode 4
07d are provided.

The gate electrode 405 is formed on the insulating substrate 40
1 is formed on the undercoat layer 402. The first insulating film 404 is formed of the undercoat layer 402
And on the gate electrode 405. Further, a p-Si film 403 is formed on the first insulating film 404.

Here, in the channel region 403a of the p-Si film 403, a plurality of groove-shaped crystal growth direction control gaps 411 are formed in the channel region 403a from the source region 403b to the drain region 403c in the same manner as in the embodiment 3-1. (See FIGS. 23A and 24). The source electrode 407s and the drain electrode 407d are connected to the source region 403b or the drain region 403 on the p-Si film 403.
and c. The gate electrode 405, the source electrode 407s, and the drain electrode 407
“d” forms a wiring pattern by being patterned into a predetermined shape at a portion other than the illustrated cross section.

A method for manufacturing the TFT 420 according to this embodiment will be described with reference to FIG. FIG. 25 is a schematic sectional view showing a manufacturing process of the TFT 420. First, in the same manner as in Embodiment 3-1, the insulating substrate 4
An undercoat layer 402 is formed on the substrate 01. further,
The gate electrode 405 and the wiring pattern are formed by patterning the undercoat layer 402 to have a predetermined shape (see FIG. 25A).

Next, as shown in FIG. 25B, a first layer is formed on the gate electrode 405 and the undercoat layer 402.
Of the insulating film 404 is formed. Further, Embodiment 3
In the same manner as in 1, a Si layer is formed on the first insulating film 404 by, for example, a plasma CVD method. After a photoresist is selectively formed on the Si layer into a predetermined shape, the photoresist is exposed using the photoresist as a mask, and then patterned into a predetermined shape by etching. afterwards,
The photoresist is removed. Thereby, an a-Si film 413 having crystal growth direction control gaps 411... Is formed, and as shown in FIG.
Is irradiated with an excimer laser on the entire surface of the a-Si film 41.
3 is crystallized to form a p-Si film 403.

Here, since crystal growth direction control gaps 411 are provided in channel region 403a in p-Si film 403, crystal grains to be formed are formed in source region as in Embodiment 3-1. It has a shape that is elongated in the direction of 403b or the drain region 403c. Therefore,
Since the crystal grain boundaries in the linear direction connecting the source region 403b and the drain region 403c are substantially reduced, the field effect mobility can be improved.

After that, as shown in FIG. 25D, a resist agent is applied on the p-Si film 403, and is patterned into a predetermined shape by exposure and development to form a resist mask 414 as an ion shielding film. I do. The resist mask 414 is not particularly limited as long as it blocks impurity ions, and various known masks can be employed. Specifically, for example, a positive resist (trade name: OFPR-5000, manufactured by Tokyo Ohka Co., Ltd.) and the like can be mentioned. Further, the material is not limited to a photosensitive material such as a resist agent, and may be a material that can be patterned by photolithography.

Using the resist mask 414 as a mask, impurity ions such as phosphorus or boron are implanted into the p-Si film 403 by, for example, an ion doping method. Thereby, the channel region 403 is formed in the p-Si film 403.
a and source regions 40 on both sides of the channel region 403a.
3b and the drain region 403c are formed. Thereafter, the resist mask 414 is peeled off, and further, FIG.
As shown in (e), the source electrode 407s and the drain electrode 407d are selectively formed to form the third embodiment 3-2.
Is obtained.

The inverted staggered TF thus produced
Also at T, effects such as improvement of the field effect mobility can be obtained as in Embodiment 3-1.

(Embodiment 3-3) Embodiment 3
No. 3 is characterized in that an early crystallization region where crystallization is started at a higher temperature than other regions is provided instead of the crystal growth direction control gap in Embodiments 3-1 and 3-2. Hereinafter, the crystalline thin-film semiconductor transistor according to the embodiment 3-3 will be described with reference to FIG. It is to be noted that, except that an early crystallization region is provided in place of the crystal growth direction control gap, the same as in the embodiment 3-1 described above. Omitted. Embodiment 3-1
Alternatively, the same reference numerals are given to components having functions similar to those of the thin film transistor of Embodiment 3-2.

As shown in FIG. 26, in the p-Si film 403, a band-like early crystallization region 421 in which impurity ions other than phosphorus or boron are implanted from the source region to the drain region in the channel region. The TFT 430 formed and having such a structure can be manufactured as follows.

First, the undercoat layer 402 is formed on the insulating substrate 401 in the same manner as in the embodiment 3-1.
Is formed by a normal pressure CVD method. Next, Si is formed on the undercoat layer 402 by, for example, a plasma CVD method.
A layer is formed, and a photoresist is selectively formed on the Si layer in a predetermined shape. After exposing using the photoresist as a mask, an a-Si film 413 is formed by patterning into a predetermined shape by etching.

Next, in the channel region 403a of the a-Si film 413, impurity ions other than phosphorus or boron and capable of increasing the crystallization start temperature are implanted in a band shape in the direction from the source region 403b to the drain region 403c. Thus, an early crystallization region 421 is formed. Then, the a-Si film 41 on which the early crystallization region 421 is formed
3 was irradiated with an excimer laser beam as an energy beam for about 50 ns, and then allowed to cool to a-Si
The film 413 is crystallized.

Here, when the entire surface of the a-Si film 413 is irradiated with an energy beam, the temperature of the surface of the a-Si film 413 increases, and the temperature of the a-Si film 413 gradually decreases due to subsequent cooling. In the course of the temperature drop, the first crystal nucleus is generated in the early crystallization region 421 prior to the other regions. This is because, by implanting impurity ions into the early crystallization region 421, crystallization is started at a higher temperature than in other regions.

Thereafter, crystal growth is performed centering on the crystal nuclei generated in the early crystallization region 421. Therefore, a poly-Si film in which large crystal grains are aggregated can be formed.

Steps after crystallization are the same as those in the embodiment 3-1.

The method of injecting the impurity ions capable of increasing the crystallization start temperature into the a-Si film is not particularly limited, and various conventionally known methods can be employed. In this embodiment, the example of the staggered type has been described. However, the same effect can be obtained with the inverted staggered type. In addition, the early crystallization region is not limited to the region into which impurity ions are implanted as described above, but a region (pre-crystal) partially crystallized in advance is formed, and a melting point (crystallization) depending on the degree of crystallinity is formed. Temperature). Further, in order to form such a pre-crystal finely, for example, irradiation of interference fringes of coherent light may be used.

(Other Matters Regarding Embodiments 3-1 to 3-3) The above-described Embodiments 3-1 and 3-
In 2, the a-Si film 413 is provided with a long groove-shaped crystal growth direction control space 411 in the direction connecting the source region and the drain region, but the present invention is not limited to this embodiment. For example, as shown in FIG. 27, discontinuous crystal growth direction control gaps 431 may be provided in a direction connecting the source region and the drain region. In this embodiment,
The space between the crystal growth direction control gaps 431 in the direction connecting the source region and the drain region, or together with this space,
By appropriately adjusting the distance between adjacent crystal growth direction control gaps 431 in a direction perpendicular to the direction, the grain size of the crystal grains in the direction can be controlled.

In the present invention, as shown in FIG.
A void having a depth that does not penetrate the a-Si film may be provided in the channel region of the a-Si film. Further, a gap having such a shape may be formed discontinuously in an island shape.

In the case where the void does not penetrate, the convex portion forming the void is removed by etching or the like after the formation of the p-Si film, and the surface of the p-Si film may be flattened. Good.

Further, for example, a bar-shaped member having a specific heat different from that of the main body may be placed in the channel region of the a-Si film 413. It is also possible to form crystal growth direction control regions having different specific heats in the channel region. For example, when a member having a higher specific heat than the a-Si film is placed and irradiated with an energy beam for a short time, the a-Si film in contact with the member
Since the temperature rise in the film portion is small, crystal nuclei are generated earlier than in other regions. On the other hand, for example, when a plurality of members having a lower specific heat than the a-Si film are placed in a direction connecting the source region and the drain region and irradiated with an energy beam for a short time, the temperature of the member is lower than that of the a-Si film. Therefore, the temperature of the intermediate portion of the members arranged in a plurality of rows becomes relatively low. Therefore, the first crystal nucleus is generated in this portion, and the effect of preventing the generation of disordered crystal nuclei is obtained.

In the above embodiment, an embodiment using Si or Si and Ge as the material of the p-Si film 403 has been described, but in the present invention, in addition to these, silicon carbide (SiC) may be used. Such as a compound of a combination of the IV groups, and II such as gallium arsenide (GaAs).
Compounds based on a combination of Group I and Group V, and compounds based on a combination of Group II and Group V such as cadmium selenide (CdSe) can also be used.

Further, in the present invention, the gate electrode 40
5, the embodiment using Al as the material of the source electrode 407s and the drain electrode 407d was described. In addition, chromium (Cr), molybdenum (Mo),
a), a metal such as titanium (Ti), or an alloy thereof may be used.

Further, in the present invention, the a-Si film 41
Although the embodiment using an excimer laser as an energy beam when crystallizing 3 is shown, laser light such as an Ar laser or a YAG laser, an ion beam, an electron beam, or the like may be used. Even if these energy beams are used, high-density energy can be easily locally irradiated in a short time, so that crystallization can be performed with the substrate temperature kept relatively low.

(Embodiment 4-1) In this embodiment 4-1, the light energy intensity (light energy per unit area, hereinafter simply referred to as light intensity) within the beam width monotonically increases from one to the other. Alternatively, crystallization is performed using a light beam having a distribution pattern that monotonically decreases from one side to the other side.

A typical light beam having a distribution pattern of monotonically increasing from one side to the other side or monotonically decreasing from one side to the other side has a linear light intensity gradient as shown in FIG. The light intensity may increase or decrease exponentially.

Light source for generating the light beam (before shaping)
For example, a He-Ne laser, an argon laser,
Various lasers such as a carbon dioxide laser, a ruby laser, and an excimer laser can be used. However, an excimer laser is preferably used because a high output is obtained and the silicon is well absorbed. Hereinafter, a laser annealing method according to the present invention using an excimer laser will be described.

FIG. 42 is a perspective view schematically showing a crystallization operation using the laser annealing method.
400 denotes a light beam irradiation device, 1410 denotes an irradiation target to which a light beam is irradiated, and 1401 denotes, for example, XeCl
A laser light generator using an excimer laser, 140
2 is a mirror, 1403 is a beam homogenizer.
In this light beam irradiation device 1400, the laser light generator 1
The light generated in 401 is guided to a beam homogenizer 1403 via a mirror 1402, where the light is shaped into a predetermined light intensity pattern and then output. The beam homogenizer 1403 incorporates an optical system for shaping a light beam. In this embodiment, a transmission filter having a light transmittance gradient as shown in FIG. 31 is arranged at the most downstream side of the optical path. (Not shown). Therefore, the light generated by the laser light generator 1401 passes through this transmission filter and is shaped into a light beam having a pattern as shown in FIG. 29A.

[0259] In the light beam irradiation apparatus 1400, for example, the average energy density is 300 mJ / cm ² (irradiation energy per unit area), low energy density regions L is 250 mJ / cm ^2, a high energy density region H is 350mJ / cm ² , beam cross-sectional shape is 7mm ×
A light beam shaped to 7 mm can be output, and this light beam is irradiated on a surface to be crystallized such as an amorphous silicon thin film to crystallize the material to be crystallized. The crystallization step will be described more specifically. First, as shown in an irradiation object 1410 in FIG. 42, a non-monocrystalline silicon film 1412 having a thickness of 85 nm is formed on a glass substrate 1411 by, for example, a low pressure CVD method.
Is formed. More specifically, for example, a monosilane gas (SiH ₄ ) or a disilane gas (Si ₂
H ₆ ) at a pressure of several Torr and the glass substrate 1
A non-single-crystal silicon film 1412 is formed with 411 heated to 350 ° C. to 530 ° C.

Here, an underlayer 1413 made of, for example, SiO ₂ may be formed on a glass substrate 1411, and a non-single-crystal silicon film 1412 may be formed on this underlayer. Further, the method for forming the non-single-crystal silicon film 1412 is not limited to the low pressure CVD method, and for example, a plasma CVD method may be used. Further, the thickness of the non-single-crystal silicon film 1412 is not limited to 85 nm,
It may be set appropriately.

A specific region 1404 of the non-single-crystalline silicon film 1412 thus formed is irradiated with, for example, 10 shots of excimer laser light shaped from the light beam irradiation device 1400 to melt the portion. rear,
Crystallizes by releasing heat. In this embodiment, upon irradiation with a light beam, the irradiation object 1410 is placed in an airtight container having a window made of a quartz plate, and the inside is evacuated (about 10 ⁻⁶ torr).
r), a method of irradiating a specific area 1404 with a light beam through the window at room temperature (about 23 ° C.) (FIG. 43)
42), but in FIG. 42, the airtight container is omitted.

The above-described conditions are merely examples, except that a light beam having a distribution pattern in which the light intensity within the beam width monotonically increases from one side to the other side or monotonically decreases from one side to the other side is used. Other conditions are not particularly limited. For example, the light energy density may be any light intensity that is sufficient to crystallize the non-single-crystal silicon film 1412 and has a light intensity gradient.

Also, the degree of the light intensity gradient is not particularly limited, and a gradient that can suitably induce and control crystallization may be set in consideration of the material and thickness of the non-single-crystal thin film. Further, the beam width of the light beam to be irradiated and the number of irradiations (the number of shots) are not limited to the above, and for example, a laser beam having a higher intensity may be irradiated only for one shot.

Also, the sectional shape of the light beam is not particularly limited, and may be, for example, a triangle or a circle.

Next, the crystal growth behavior when a light beam having a light intensity gradient is used will be described with reference to FIGS.

When the non-single-crystal silicon thin film is irradiated with a light beam having the light intensity pattern shown in FIG. 29A, the temperature of the irradiated surface becomes as shown by 701 (temperature distribution curve) in FIG.
The pattern has a temperature gradient rising to the right in the center and a sharp temperature gradient in the periphery. The sharp temperature gradient is formed in the peripheral part because the heat radiation to the surroundings is large. Next, when the light irradiation is stopped, first, the vicinity of the intersection (near the boundary) between the temperature distribution curve 701 and the crystallization temperature line 702 becomes lower than the melting temperature. Therefore, a fine crystal 704 is generated in the vicinity (703 indicates a cross section of the thin film). Then, with this crystal 704 as a nucleus, crystal growth proceeds in the right direction in the drawing, which is still at or above the crystallization temperature. Here, unlike the case of FIG. 7,
In FIG. 29b, since a temperature gradient is formed in the center,
Heat flows from the high-temperature region side (H side) toward the low-temperature region side (L side), and this heat acts to alleviate a sharp drop in temperature and induce crystal growth toward the high-temperature side (right side in the drawing). . Therefore, the generation and growth of crystal nuclei proceed smoothly, and as a result, the uniformity of the grain size and crystallinity of the crystal grains is increased, and the crystal grains longer from the L side to the H side (crystal growth direction) are formed. Generate. That is, when a light beam having a light intensity gradient is used, a crystalline thin film having high mobility in the crystal growth direction can be manufactured.

By the way, the irradiation method of the light beam is as follows.
The irradiation may be performed in a fixed state (immobile state) on both the irradiation side and the substrate side, but either the light beam or the substrate side may be moved, and this movement may be reciprocated. In the method of irradiating while moving or reciprocating, preferably, as shown in FIG. 30, the light intensity gradient direction (L → H
(Or H → L direction). In this direction, the crystal growth direction can be finely guided, the uniformity of the crystal grain size and the crystallinity can be increased, and the movement can be adjusted in accordance with the light intensity gradient and the light irradiation intensity. By adjusting the speed, the crystal growth direction can be more finely guided.

Note that the arrow in FIG.
Reference numeral 712 denotes an irradiation surface before and after the movement, and 713 (shaded portion) denotes an overlap irradiation region. Although the light beam is moved in FIG. 30, the substrate side may be moved, and when the light beam is shot a plurality of times, for example, the irradiation position may be several to several tens% of the irradiation area. Irradiation can be performed while shifting.

In the poly-Si thin film manufactured as described above, the source region and the drain region are generally formed by ion-implanting impurities such as phosphorus and boron into both ends of the poly-Si thin film. To form a TFT. The light beam having the energy intensity pattern described in this embodiment (FIG. 29A) is used for the AM-LCD.
(Active Matrix Liquid Crystal Display) It is effective for crystallization of a relatively narrow area to form peripheral circuits.

(Embodiment 4-2) This embodiment (the same applies to Embodiment 4-3 described later) is an example effective for crystallizing a relatively wide region.

FIG. 32A shows a light intensity distribution pattern of a light beam used in the present embodiment. As shown in the figure, the light beam according to the embodiment 4-2 has a high light intensity H
The region 721 and the small L region 722 have a pattern in which they are alternately arranged on a plane. Here, the light intensity ratio between the H region and the L region is not particularly limited, and may be appropriately set. However, the total amount of light energy is generally defined so that the entire irradiation surface (L region and H region) is melted within the number of irradiations. Here, the H region is 300 mJ / cm ² , the L region is 200 mJ / cm ² ,
The thickness of the amorphous silicon thin film was set to 50 nm, and other conditions were the same as those in Embodiment 4-1.

The crystallization behavior of this embodiment will be described below with reference to FIGS. First,
When a light beam having the light distribution characteristic shown in FIG. 32A is irradiated, the temperature of the thin film surface has a distribution pattern as shown in FIG. 32B. Then, in the process in which the light irradiation is completed and the temperature of the irradiation surface decreases, when the temperature of the L region 722 approaches the crystallization temperature line 723, FIG.
As shown in FIG.
24 occur (725 indicates the cross section of the thin film). When the temperature is further lowered (FIG. 32D), the crystal growth is induced toward the high-temperature region H by the heat transferred from the high-temperature region H to the low-temperature region L, and new crystal nuclei are also generated in this process. It grows (FIG. 32e).

The generation and growth of such crystals continue until the temperature of the high-temperature region H becomes lower than the melting temperature (FIG. 32f, FIG.
g) In this embodiment, the direction of crystal growth is guided from L to H. Therefore, crystal grains grow from both the low-temperature regions L sandwiching the high-temperature region H, and as a result, the crystal grains collide with each other near the center 726 of the high-temperature region H (FIG. 32g). As a result, a crystal grain boundary line is formed near the center of the high-temperature region H, and the collision induces further crystal growth in a direction parallel to the L → H direction. Therefore, after the collision, the crystal grows slightly in the direction perpendicular to the L ← → H direction, and thus crystal grains having a long diameter in the direction perpendicular to the L ← → H are formed.

According to the above-described mechanism, according to this embodiment, crystallization in a relatively wide irradiation area of, for example, several cm square can be smoothly advanced. Further, as described above, when the light beam is set so that the direction perpendicular to the moving direction of the carrier (the source-drain direction) becomes the L ← → H direction, and the light is irradiated, the carrier crosses the grain boundary line. Since it can be moved without any problem, a high-speed TFT can be realized.

Also in this embodiment, the fourth embodiment
In the same manner as −1, as shown in FIG.
= T1 to t = t2), the irradiation may be performed while moving either the light beam or the substrate (including the reciprocating motion). By doing so, it is possible to further improve the uniformity of the crystallinity. In FIG. 33, reference numerals 731 and 732 denote irradiation surface positions before and after the movement, 733 (shaded area) denotes an overlap irradiation area, and arrows indicate a movement direction. However, it is needless to say that the movement is not limited to such movement.

By the way, as shown in FIG. 32A, a light beam in which a portion H having a high light intensity and a portion L having a weak light intensity are arranged in a stripe pattern can be easily formed by a known technique without requiring any special technique. It can be realized, and the means for realizing it is not limited at all. For example, filters that absorb light to be used to some extent are arranged at predetermined intervals, and a comb-shaped transmission filter having a transmission distribution as shown in FIG. 34 is manufactured. This filter can be realized by installing the filter in the optical path of the light beam irradiation device (for example, in a beam homogenizer).
Further, for example, it can also be realized by means for arranging, in the optical path, a filter in which a large number of metal fibers are arranged vertically or horizontally in parallel. Further, it can also be realized by means for producing a striped light intensity pattern by a method of arranging a slit in the optical path and causing diffraction interference.

(Embodiment 4-3) Embodiment 4-3
Is a method for giving non-uniformity to the light intensity distribution using light interference. Since this method can control the light intensity distribution pattern relatively freely, it is suitable for crystallization of a relatively large area as in the case of Embodiment 4-2.

FIG. 35 shows a pattern of the light intensity distribution of the light beam used in the present embodiment. As shown in FIG. 36, such a light intensity distribution pattern can be easily formed by means of simultaneously irradiating two coherent light beams 801 and 802 to cause light interference. Specifically, for example, interference can be generated by dividing a laser beam generated from the same light beam into two optical paths by a semi-transmissive mirror and generating a relative angle between the optical paths using a reflecting mirror. .

When two coherent lights interfere with each other, a portion having a high light intensity (a bright line portion H) and a portion having a low light intensity (a dark line portion L) are formed. The degree of modulation (which affects the ratio of the light intensity in the light and dark lines) can be freely changed by the angle at which the two light beams intersect, and the energy intensity of the two light beams can be changed. Therefore, the distance between the bright line portion H and the dark line portion L and the intensity ratio can be set relatively freely. Therefore, the distance and the intensity ratio are appropriately set in consideration of the thickness and the like of the amorphous thin film which is the surface to be irradiated.

Hereinafter, the crystal growth behavior when a light beam having such an interference pattern is irradiated will be described with reference to FIGS. The operating conditions in this embodiment are the same as those in Embodiment 4-1 and the like.

When the non-single-crystal silicon thin film is irradiated with the light beam characterized by FIG. 37A, a high temperature distribution pattern (curve 901) is formed in the bright line portion H and low in the dark line portion L on the thin film (curve 901). Figure 37b). As shown in FIG. 37c, in the process of ending the light irradiation and lowering the temperature, the crystal nucleus 9 is located at the portion where the curve 901 first intersects the crystallization temperature line 902 (the lowest temperature portion of the low temperature region L).
03 occurs. When the temperature further decreases (see FIG. 3)
7d) By the heat transmitted from the high temperature region H to the low temperature region L, the crystal growth is induced in the L → H direction, and a new nucleus is generated and similarly induced to grow (FIG. 37).
e). Such generation and growth of crystals continue until the temperature of the high-temperature region H corresponding to the bright line portion H falls below the melting temperature (FIGS. 37f, g).

From the crystallization mechanism described above, according to Embodiment 4-3, a crystalline semiconductor thin film having a uniform crystallinity and a large field-effect mobility can be manufactured in a relatively wide range. Also, in this embodiment, similarly to the above-mentioned Embodiment 4-2, the boundary line of the crystal grain boundary (collision line of the crystal grains) is formed at the center of the high temperature region. As described in 2, when the direction orthogonal to the arrangement direction of H → L → H → L is used as the carrier movement direction, high field effect mobility can be obtained.

(Embodiment 4-4) Embodiment 4-4
Is basically the same as in the case of Embodiment 4-3.
However, in this embodiment, the period of the interference pattern and the modulation degree are dynamically adjusted so that the bright line portion H and the dark line portion L
And change in a wave-like manner. Hereinafter, the contents of Embodiment 4-4 will be described.

As shown in FIG. 38, at least one of the two coherent light beams is dynamically phase-modulated to change the positions of the bright and dark lines of the interference fringes in a wave-like manner. Forming a light beam. As the phase modulation, for example, the phase of one light beam is sequentially changed to 0, π / 2, π relative to the other light beam. In this way, the positions of the bright line portion and the dark line portion of the interference fringes are shifted in time series, and a light beam can be formed in which the striped pattern composed of the bright line portion and the dark line portion changes dynamically (FIG. 38). .

As means for modulating the phase, for example, a method of dynamically changing the optical path length of one of two light beams by using a mirror to change the phase, the refractive index of a transparent body disposed in the optical path, Can be exemplified, and such an optical system is incorporated in, for example, the beam homogenizer (1403 in FIG. 42).

According to this embodiment, since the temperature distribution pattern in which the high-temperature region H and the low-temperature region L alternate with each other is formed on the thin film surface as the irradiation surface, the effect of inducing the crystal growth in a certain direction is obtained. large. In addition, this method has an effect of driving impurities out of the effective region, so that a high-quality crystalline thin film can be formed while purifying the thin film. The effect of driving impurities out of the effective region is based on the following principle. That is, the melting point of the thin film component and the impurity,
Since physical properties such as specific gravity are different, when a wave-like temperature change is applied, a difference in traveling speed occurs between the two. Therefore, by performing irradiation many times, a minute amount of impurities and a thin film component are separated.

Here, the adjustment of the period and the modulation degree of the interference pattern may be performed during one irradiation or for each irradiation of a large number of irradiations. Furthermore, it is also possible to control the degree of modulation in accordance with each stage of crystal growth, and according to such a method, it is possible to more suitably induce crystal growth.

Also, in any of Embodiments 4-3 and 4-4, similarly to Embodiment 4-1 or 4-2, either the optical laser or the substrate side is moved (reciprocating motion or the like). The crystal growth can be appropriately controlled by this method. When this movement is performed in a direction parallel to the direction of the striped pattern including the bright line portion H and the dark line portion L, even in the case of the above-described Embodiment 4-3, the effect of driving out impurities is achieved. Is obtained.

Although the above description has been made mainly on the assumption of the temperature distribution in the plane direction of the irradiation area, as shown in FIG. 39, it is also possible to form a light intensity distribution due to light interference in the thickness direction of the thin film to be irradiated. it can. FIG. 39 shows a thin film 110 of an amorphous silicon layer in order from the irradiation direction (above in the figure).
1. Underlayer (SiO ₂ ) 1102, glass substrate 1103
FIG. 4 is a schematic diagram showing a state in which a light beam is irradiated on an irradiation target made of. When a light beam having a light intensity distribution in the vertical direction (thickness direction) is incident on the irradiation object, a temperature distribution corresponding to this waveform is formed in the thickness direction. However, the thickness of the silicon thin film used for the TFT is usually Since it is as thin as several tens of nanometers and the distance is shorter than the period of the interference fringes, it is difficult to form a periodic temperature distribution in the thickness direction.

However, the upper surface of the thin film 1101 is cooled by radiating heat to the surrounding environment, and the lower surface (substrate side) is dissipated by conducting heat to the base layer 1102 and the glass substrate 1103. There is a temperature distribution, and it is possible to increase this temperature distribution. The above-mentioned interference pattern can be used as means for expanding the temperature distribution in the thickness direction by light irradiation. Specifically, for example, a reflection mirror is provided on the lower surface of the glass substrate 1103 to cause interference, or the thin film 1101 and the underlayer 11
02 or the glass substrate 1103, the difference in the refractive index is increased, so that light incident from the thin film 1101 side and light reflected at the interface between the layers cause interference. Further, the temperature distribution can be formed in the thickness direction (vertical direction) by adjusting the period of the interference fringes, and thereby, the crystal growth in the thickness direction can be controlled.

In the case of controlling the temperature distribution in the thickness direction, it is preferable to specifically set suitable setting conditions in consideration of the thickness of the non-single-crystalline thin film and the thermal conductivity of the underlayer and the substrate. . Alternatively, light emitted from one light source may be split into two, and one may be irradiated from the thin film surface side (above) and the other may be irradiated from the substrate side (below) to cause interference inside the thin film. However, in this case, the substrate and the base layer are made of a material through which a light beam is transmitted.

(Embodiment 4-5) This embodiment is characterized in that a temperature gradient is formed on the surface to be crystallized by suitably setting the pressure of the atmospheric gas in the crystallization process. On the other hand, Embodiments 4-1 to 4-
In No. 4, the crystallinity is improved and uniformized by adjusting and controlling the light intensity pattern of the light beam.
Therefore, the present embodiment and the embodiments 4-1 to 4-4 have completely different ideas. Hereinafter, the contents of the present embodiment will be described.

FIG. 40 is a cross-sectional view of an object to be irradiated (laminated body) similar to FIG. 39.
201 indicates a thin film, 1202 indicates a base layer, 1203 indicates a substrate, and arrows indicate the direction of heat transfer (radiation direction) of the thin film. As shown in FIG. 40, part of the heat diffuses in the ambient atmosphere (upward) and out of the irradiation region of the thin film (left-right direction in the figure), but most of the heat has a large contact area and a low thermal conductivity. It is transmitted to the large substrate side (downward). Here, in the conventional laser annealing method, a method of irradiating a light beam having a uniform light intensity distribution in a high vacuum atmosphere is used. Therefore, as described in Embodiment 4-1 above, since a temperature gradient is hardly formed at the center of the irradiation surface, nuclei hardly occur at the center in the initial stage of heat radiation. On the other hand, a phenomenon occurs in which a large number of nuclei are generated simultaneously and frequently at a certain stage of the heat release process.

This embodiment is characterized in that the ambient atmosphere is not set to a high vacuum unlike the above-mentioned conventional method.
The motion of gas molecules forming the surrounding atmosphere is used to generate a portion having a non-uniform temperature on the light irradiation surface.

First, the principle of Embodiment 4-5 will be described. Generally, the time of one shot (one pulse) of pulse light such as an excimer laser is 20 to 50 ns.
c is extremely short. Therefore, it is necessary to raise the temperature of silicon or the like to a temperature equal to or higher than the melting point within this short irradiation time. Therefore, it is general that silicon or the like is formed into a very thin film of several tens of nm. In the case of such an extremely thin thin film, the influence of the surrounding gas molecules in the heat radiation process becomes extremely large.

That is, the gas molecules forming the surrounding atmosphere and the gas molecules existing in the thin film make a motion of colliding with the surface of the thin film with a certain probability and moving away. Since it is smaller than the heated and thin film, it loses the heat of the thin film when it collides with the thin film surface and separates. Considering the action of such gas molecules, a perturbed temperature distribution should have occurred on the thin film surface. Therefore, if the ambient atmosphere pressure and the type of gas molecules constituting the ambient atmosphere are properly set, even when a light beam having a uniform light intensity distribution is irradiated, a portion having a non-uniform temperature in the light irradiation area. (Low temperature part) can be formed. It is considered that if such sites can be formed, generation of nuclei and smoothing of crystal growth can be realized. Based on this idea, the following experiment was conducted.

The crystal nuclei in the crystallization process are as follows:
It is thought that it plays a role similar to the water vapor core when the moisture in the atmosphere condenses.

(Experiment 1) Experimental conditions First, a 200 nm-thick SiO ₂ film was placed on a substrate (thickness: 1.1 mm) made of Corning # 7059 glass.
A layer (base layer) was formed, and an amorphous silicon layer having a thickness of 50 nm was further formed thereon to prepare an irradiation object. Next, the irradiation target on which the amorphous silicon layer 1503 is formed is placed in an airtight container 1500 provided with a window 1501 made of quartz glass as shown in FIG. 43, and the air in the airtight container 1500 is removed. Then, hydrogen gas was introduced from the hydrogen gas cylinder 1502, and the inside of the airtight container was set to a predetermined hydrogen gas pressure. Next, the laser irradiation device 1
The amorphous silicon layer 1503 of the irradiation object was irradiated with the excimer laser generated at 510 through the window 1501, and then heat was released for crystallization.

The predetermined hydrogen gas pressure (atmospheric pressure) is 5 × 10 ⁻⁶ torr and 1 × 10 ⁻⁵ torr.
r, 1 × 10 ⁻² torr, 1 × 10 ⁻¹ torr, 1
Torr, 10 torr. As the laser irradiation condition, one pulse (one shot) is 30 n ·
sec, beam width 7mm x 7mm, light intensity 350m
A conventional light beam having a uniform light intensity distribution of J / cm ² was used. After irradiating 100 pulses of this light beam,
Dissipate heat in a room temperature environment to form an amorphous silicon layer 15
03 was polycrystallized.

In FIG. 43, reference numeral 1511 denotes an excimer laser light generator, 1512 denotes a mirror, and 1513 denotes a beam homogenizer.

Relationship between Atmospheric Pressure and Crystallinity Six types of crystalline silicon thin films (poly
-Si) was visually observed by oblique light. The Raman intensity was measured by microscopic Raman spectroscopy, and the Raman intensity was 1 when the hydrogen gas pressure was 5 × 10 ⁻⁶ torr.
The crystallinity of each was evaluated. The results are listed in Table 1.

[Table 1]

As shown in Table 1, it was confirmed by visual observation that the state of the silicon thin film after the laser annealing changed according to the atmospheric pressure at the time of manufacturing. That is, when the hydrogen gas pressure (atmospheric pressure) is on the order of 10 ⁻⁶ torr, only a small amount of bluish scattered light is observed, but on the order of 10 ⁻⁵ torr, the scattered light shifts to the green side. The whole became bright. Further, when the hydrogen gas pressure is 10
^{When the value} was increased to ^-1 , scattering became remarkable and a cloudy state was observed, and thereafter, a similar state was confirmed up to about 10 torr.

On the other hand, according to the evaluation of crystallinity by micro-Raman spectroscopy, when the Raman intensity of a sample crystallized at a hydrogen gas pressure of 5 × 10 ⁻⁶ torr is used as a reference, the atmospheric pressure is 1 × 10 ^−5. At torr, the intensity is four times higher, and
6 times to 7 between × 10 ^-5 torr and 10 torr
The Raman intensity was doubled. The following becomes clear from these results.

Conventionally, light irradiation is generally performed in a state where the atmospheric pressure is reduced as much as possible (high vacuum state), except for a special case such as a reaction between molecules in an atmosphere and a thin film substance. As is evident from Table 1, a high degree of vacuum does not provide good crystallinity. On the other hand, as the hydrogen gas pressure increases, the crystallinity increases. From the experimental results, in the crystallization by laser annealing, it is preferable to set the ambient atmospheric pressure to a certain value or more,
Preferably, the atmospheric pressure is 1 × 10 ⁻² torr or more.

[0305] It is considered that the reason why good crystallinity cannot be obtained in a high vacuum state is that perturbation of temperature non-uniformity due to movement of gas molecules cannot be formed. On the other hand, the crystallinity was improved in the presence of hydrogen gas pressure because hydrogen molecules collide with the surface of the thin film and take away the heat of the thin film when it departs, causing local and perturbative temperature non-uniformity. it is conceivable that. That is, the results in Table 1 support the above considerations.

(Experiment 2) The hydrogen gas pressure as the atmospheric pressure was set to 5 × 10 ⁻⁶ torr and 1 torr, and the number of light beam irradiation was changed to 1, 10, 100 and 500 under each hydrogen gas pressure condition. Other than that, the same procedure as in Experiment 1 described above was performed to remove the crystalline silicon thin film (poly-S
i) was prepared. Then, the Raman intensity was measured in the same manner as described above, and the influence of the hydrogen gas pressure on the relationship between the number of irradiation times and the crystallinity was examined. The result is shown in FIG.

As is clear from FIG. 41, when the hydrogen gas pressure was 1 torr, it was recognized that the Raman intensity increased and the crystallinity improved as the number of irradiations increased. On the other hand, when the hydrogen gas pressure was 5 × 10 ⁻⁶ torr, the Raman intensity did not increase even if the irradiation frequency was increased beyond 10 times, and the crystallinity was not improved.

From this result, it is understood that it is better not to set the atmospheric pressure to a high vacuum even in the method of crystallization by irradiating the light beam many times. Then, it was confirmed from these results that when at least the hydrogen gas pressure was set to 1 torr, the crystallinity was improved as the number of irradiations was increased.

[0309] In Experiment 1, no results are shown under the condition exceeding 10 torr, but it is considered that a good quality crystalline thin film is formed even under the condition exceeding 10 torr. The reason is as follows. It is considered that when the hydrogen gas pressure increases, the number of hydrogen molecules that collide with and depart from the thin film surface increases, so that the effect of causing non-uniform temperature distribution is considered to be weakened. However, the following effect is added. That is, when the temperature of the thin film is heated to a temperature equal to or higher than the melting point temperature by light irradiation, the vapor pressure inside the thin film increases, and this vapor pressure inhibits the growth of the crystal, and the material constituting the thin film is scattered. When the pressure is high, scattering of the thin film material or the like is suppressed by the pressure, and as a result, the progress of crystallization becomes smooth.

In the experiments 1 and 2, hydrogen gas (H ₂ ) having a large specific heat and a large thermal cooling effect was used as a gas constituting the ambient atmosphere. However, the gas constituting the ambient atmosphere was limited to hydrogen gas. It is not something to be done. For example, N
An inert gas such as ₂ or He or Ar can be used, and a mixed gas obtained by mixing two or more of these gas molecules may be used. However, since the influence (including the adverse effect) on the specific heat and the thin film material differs depending on the type of the gas molecule, it is preferable to set a suitable gas pressure according to the type of the gas molecule.

In the above description, an excimer laser is used as a light beam, but the light beam used in the present invention is not limited to an excimer laser. For example, not only a laser that emits continuously such as a He-Ne laser and an argon laser as described above, but also light such as an ultraviolet lamp can be used.

Although the present invention is particularly useful as a polycrystallization method, it is needless to say that the present invention can be used as a method for producing single crystallization.

Further, the content of the present invention has been described above mainly with respect to the method of forming a crystalline semiconductor thin film. However, the technology according to the present invention relates to the modification of a substance using light energy, for example, the melt molding of a polymer, It can be widely applied to thermal annealing operation for alloys and the like.

(Embodiment 5-1) An embodiment 5-1 of the present invention will be described with reference to FIG. 44 and FIG.

First, as shown in FIG.
At 21, an amorphous silicon thin film 522 as a precursor semiconductor thin film is formed by a plasma CVD method.
The thickness of the amorphous silicon thin film 522 is not particularly limited, but usually differs depending on the application.
The thickness is about 300 to 1000 ° when used for an optical sensor, and about 1 μm or more when used for an optical sensor or a photovoltaic element (such as a solar cell).

Next, the amorphous silicon thin film 52
2 is formed on the substrate stage 535.
And a laser beam 531 a of a XeCl excimer laser 531 as a first energy beam for crystallization and a laser beam 532 a of an Ar laser 532 as a second energy beam for preheating on the amorphous silicon thin film 522. Is irradiated for 5 seconds in a stationary state. More specifically, the XeCl excimer laser 531 has an oscillation frequency of 50 Hz, a wavelength of 308 nm, and an irradiation energy of 30.
Was 0 mJ / cm ^2, Ar laser 532, continuous wave, wavelength 488 nm, output is 20W / cm ^2.
The laser beam 531a is transmitted through the half mirror 533 and irradiated, while the laser beam 532a is reflected by the half mirror 533 and irradiated. Each laser beam 5 on the amorphous silicon thin film 522
Irradiation areas 531b and 532b of the laser beams 31a and 532a are band-shaped, and the laser beam 532a is
Wider irradiation area 5 including irradiation area 531b of 31a
Irradiated at 32b.

It should be noted that making the amorphous silicon thin film 522 vertically irradiated with the laser beams 531a and 532a by using the half mirror 533 as described above can prevent the dispersion of the crystal grain size and the field effect mobility. Although it is preferable in that the reduction is easy, the irradiation is not necessarily performed exactly vertically, but may be performed substantially perpendicularly by, for example, disposing two mirrors slightly shifted. It may be. In place of the XeCl excimer laser 531, various lasers having a wavelength of 400 nm or less, such as ArF, KrF, and XeF excimer lasers, may be used, while various lasers having a wavelength of 450 to 550 nm may be used instead of the Ar laser 532. A laser may be used.

Here, the amorphous silicon thin film 522
Has, for example, a transmittance characteristic as shown in FIG. 1 when the film thickness is 1000 °. That is, for example, when the wavelength is about 5
For light with a wavelength of 00 nm, the absorption coefficient is 10 ⁵ cm ^−1, which is about the reciprocal of the film thickness, and for light with a wavelength shorter than 400 nm, the absorption coefficient is 10 ⁶ cm ⁻¹ or more, and is hardly transmitted. . Therefore, the laser beam 531a of the XeCl excimer laser 531 having a wavelength of 308 nm is almost absorbed near the surface of the amorphous silicon thin film 522, and the temperature rise and heat conduction thereby cause the amorphous silicon thin film 522 to mainly reach about 1200 ° C. Heated. On the other hand, an Ar laser 5 having a wavelength of 488 nm
The 32 laser beams 532a are absorbed in almost the entire region in the thickness direction of the amorphous silicon thin film 522, and the glass substrate 521 is heated to about 400 ° C. by conduction heat. Therefore, after the irradiation of the laser beams 531a and 532a is completed, the amorphous silicon thin film 522 is gradually cooled,
Crystal growth is promoted, and a polysilicon thin film 523 having large crystal grains is formed.

Raman scattering measurement was performed to evaluate the crystallization state of the polysilicon thin film 523 formed as described above and the polysilicon thin film crystallized only by the XeCl excimer laser 531 in the same manner as before. Was. Each measurement result is indicated by a symbol P or R in FIG. As is clear from the figure, the laser beam 532 of the Ar laser 532 is larger than the case where only the laser beam 531a of the XeCl excimer laser 531 is irradiated (R).
It was confirmed that Raman scattering intensity was higher and the crystallinity was better in the case of irradiation (a) when a was also applied.

The glass substrate 521 on which the amorphous silicon thin film 522 is formed is moved at a speed of, for example, 3 mm / sec.
2 were similarly irradiated with laser beams 531a and 532a, and Raman scattering measurement was performed on a plurality of regions of the formed polysilicon thin film 523 to examine the distribution of crystallization. confirmed.

In order to improve the uniformity of crystallization, the irradiation area 53 of the laser beam 532a is increased as described above.
2b is preferably a wider area including the irradiation area 531b of the laser beam 531a.

(Embodiment 5-2) Embodiment 5
Instead of one Ar laser 532, as shown in FIG.
For example, an infrared lamp 534 having a wavelength of 4 μm may be used. That is, the glass substrate 521 has a transmittance characteristic as shown in FIG. 47, for example, and the infrared ray 534a emitted from the infrared lamp 534 passes through the amorphous silicon thin film 522 and is mostly absorbed by the glass substrate 521. . Therefore, when the laser beam 531a of the XeCl excimer laser 531 is applied, the amorphous silicon thin film 522 is mainly heated by the laser beam 531a, as in Embodiment 5-1. The substrate 521 is heated. Therefore, the laser beam 531
After the irradiation of the amorphous silicon thin film 522 is completed, the crystal growth is promoted.
A polysilicon thin film 523 having large crystal grains is formed.

The results of Raman scattering measurement of the polysilicon thin film 523 formed under the same conditions as in Embodiment 5-1 except that the infrared lamp 534 is used as described above are shown in FIG. Indicated by As is clear from the figure, the XeCl excimer laser 53
It was confirmed that the Raman scattering intensity was higher and the crystallinity was better than in the case where only one laser beam 531a was irradiated (R).

Further, the uniformity of the crystal grains is also determined in the fifth embodiment.
It was confirmed to be as high as 1.

[0325] In addition to the infrared ray 534a, the laser beam 532a of the Ar laser 532 may be irradiated similarly to the embodiment 5-1. In addition, the infrared ray 534a may be irradiated perpendicularly to the amorphous silicon thin film 522 by using a half mirror as in Embodiment 5-1.

(Embodiment 5-3) An embodiment 5-3 of the present invention will be described with reference to FIGS.

[0327] First, as shown in FIG.
At 21, a microcrystalline silicon thin film 524 as a precursor semiconductor thin film is formed by an inductively coupled plasma CVD apparatus. More specifically, for example, a mixed gas obtained by mixing monosilane gas (SiH ₄ ) and hydrogen gas at a ratio of 2: 3 is used as the reaction gas, and the substrate temperature (reaction temperature) is 350 ° C. to 53 ° C.
Under the reaction conditions of 0 ° C. and a pressure of several mTorr, the film thickness is 85 n
Then, a microcrystalline silicon thin film 524 is formed. Note that, instead of the microcrystalline silicon thin film 524, an amorphous silicon thin film 522 may be formed as in Embodiment 5-1. Also, instead of a plasma CVD apparatus, LP (Low Po
wer) A CVD apparatus, a sputtering apparatus, or the like may be used.

Next, the glass substrate 521 on which the microcrystalline silicon thin film 524 is formed is heated at 400 ° C. to 500 ° C. for 30 minutes.
Heat treatment is performed for more than one minute to perform dehydrogenation treatment for releasing hydrogen in the microcrystalline silicon thin film 524. That is, at the time of laser annealing to be described later, it is possible to prevent the hydrogen taken in the microcrystalline silicon thin film 524 from being rapidly released and causing damage to the microcrystalline silicon thin film 524.

Next, laser annealing is performed. That is,
As shown in FIG. 49, the glass substrate 521 is set in a chamber 541 in which an irradiation window 541a made of a quartz plate is formed, and a laser beam 5 of a XeCl excimer laser 531 is placed.
31a and the incandescent light 542a of the incandescent lamp 542 are irradiated to crystallize the microcrystalline silicon thin film 524 to form a polysilicon thin film 523. More specifically, the laser beam 531a has a pulse width of several tens ns, a wavelength of 308 nm, and an irradiation energy of 350 mJ /.
cm ² , and the number of irradiations is 10. The laser beam 531a is irradiated via a laser beam attenuator 543, a homogenizer (laser beam homogenizer) 544, and a reflecting mirror 545. On the other hand, the incandescent light 542a is applied so that the microcrystalline silicon thin film 524 is heated to about 400 ° C.

Further, by heating the polysilicon thin film 523 to 350 ° C. or more in a hydrogen plasma atmosphere, a hydrogenation treatment for terminating the broken bonds in the polysilicon thin film 523 with hydrogen is performed.

When the crystal grain size of the polysilicon thin film 523 formed as described above was measured using SEM and TEM, it was 0.7 μm, which was smaller than 0.3 μm in the conventional polysilicon film. It was confirmed that the particle size had increased. The field-effect mobility is 50 c
m ² / V · sec to 80 cm ² / V · sec, and the total defect density in the interface and the film of the polysilicon thin film 523 is 1.3 × 10 ¹² cm ⁻² eV ⁻¹ . It decreased to 1.0 × 10 ¹² cm ⁻² eV ⁻¹ . That is, the incandescent lamp 542 is irradiated when the laser beam 531a is irradiated.
In addition, the crystal grain size of the polysilicon thin film 523 is increased and the film quality is improved.

The experiment was carried out with the laser beam 531a under various irradiation conditions.
Crystallization occurs in mJ / cm ² or more, the microcrystalline silicon is lost at 500 mJ / cm ² or more. In addition, 300mJ /
In the range of not less than cm ^{2 and not} more than 450 mJ / cm ² , crystal growth is sufficiently performed and the crystal grain size is large. Further, when the number of irradiations is 5 or more, generation of crystal defects is suppressed, and crystallinity is improved.

Next, a process of forming a predetermined insulating film or conductive film,
A thin film transistor (TFT) is formed by performing patterning by etching, ion implantation, and the like.
The patterning of the polysilicon thin film 523 may be performed before laser annealing.

When the gate voltage (Vg) -drain current (Id) characteristics of the TFT formed as described above were measured, as shown in FIG.
The rise of the drain current with respect to the gate voltage became steep, and it was confirmed that the subthreshold characteristic was improved, and the threshold voltage was also lowered.

Note that the laser beam 531 a
51, the glass substrate 521 is placed on a horizontally movable substrate stage 535 as shown in FIG. 51 without being irradiated with the incandescent light 542a, and is moved in the direction indicated by the arrow A in FIG. However, by irradiating the incandescent light 542a on the microcrystalline silicon thin film 524 with the incandescent light 542a ahead of the irradiation area of the laser beam 531a in the movement direction, heating by the incandescent light 542a may be performed prior to heating by the laser beam 531a. The same effect can be obtained.

(Embodiment 5-4) A TFT according to an embodiment 5-4 of the present invention will be described with reference to FIGS. 52 and 50. FIG.

First, as shown in FIG.
At 21, an amorphous silicon thin film 522 is formed using a plasma CVD method. More specifically, for example, a mixed gas of monosilane gas (SiH ₄ ) and hydrogen gas is used as a reaction gas, the substrate temperature is 180 ° C. to 300 ° C., the pressure is 0.8 Torr, and the amorphous silicon thin film 522 having a thickness of 85 nm is formed. To form

Next, as in Embodiment 5-3, the glass substrate 521 on which the amorphous silicon thin film 522 is formed is subjected to a dehydrogenation treatment.

Next, the glass substrate 521 is moved to the direction indicated by the arrow A in FIG.
Are irradiated with the laser beam 531a of the XeCl excimer laser 531 and the excimer lamp light 551a of the excimer lamp 551 to crystallize the amorphous silicon thin film 522 to form the polysilicon thin film 523. More specifically, the laser beam 531 a has an irradiation energy of 350 mJ / cm ² and an irradiation region 531 in the amorphous silicon thin film 522.
Irradiation is performed so that b has a band shape of 500 μm × 70 mm. Further, as the glass substrate 521 moves, the irradiation regions 531b of the respective pulses of the laser beam 531a overlap each other by 90%, so that the amorphous silicon thin film 522 is formed.
Laser beam 53 in each area 10 times
1a is irradiated. On the other hand, the excimer lamp light 551a is light in the range from visible light to ultraviolet light, and the laser beam 531 directly and via the concave reflecting mirror 552.
In the irradiation area 551b of 5 mm × 70 mm including the irradiation area 531b of FIG.
Irradiation is performed so as to be heated to about 0 ° C.

Further, hydrogenation treatment is performed in the same manner as in Embodiment 5-3.

When the crystal grain size of the polysilicon thin film 523 formed as described above was measured using SEM and TEM, it was 1 μm, which was 0.3 μm of the conventional polysilicon film. Was confirmed to have increased. The field-effect mobility is 50 cm ²
/ V · sec to 120 cm ² / V · sec, and the total defect density at the interface of the polysilicon thin film 523 and in the film is 1.3 × 10 ¹² cm ⁻² eV ⁻¹ to 1. It was reduced to 1 × 10 ¹² cm ⁻² eV ⁻¹ . That is, the excimer lamp 5 is irradiated when the laser beam 531a is irradiated.
By using the heating by 51 together, the crystal grain size of the polysilicon thin film 523 is increased and the film quality is improved.

Next, as in Embodiment 5-3, a TFT is formed by performing a process of forming a predetermined insulating film or conductive film, patterning by etching, ion implantation, and the like.

When the gate voltage (Vg) -drain current (Id) characteristics of the TFT formed as described above were measured, as shown in FIG. 50, the drain voltage with respect to the gate voltage was again higher than that of the conventional TFT. The current rises sharply, the sub-threshold characteristic improves,
It was confirmed that the threshold voltage was reduced from 5.0 V to 4.2 V.

(Embodiment 5-5) Embodiment 5-5 of the present invention will be described with reference to FIG. 53 and FIG.

First, as in Embodiment 5-4, an amorphous silicon thin film 522 is formed on a glass substrate 521, and a dehydrogenation process is performed.

Next, as shown in FIG.
The laser beam 531a of the XeCl excimer laser 531 and the excimer lamp light 551a of the excimer lamp 551 are irradiated while moving the substrate 21 in the direction indicated by the arrow A, and the glass substrate 52 is heated by the heater 561.
1 is heated from the bottom side to form a polysilicon thin film 523. That is, the irradiation conditions and the like of the laser beam 531a and the excimer lamp light 551a are the same as those in the embodiment 5-4, but the embodiment is further characterized in that the entire glass substrate 521 is heated to 450 ° C. by the heater 561. Different from 5-4.

Further, hydrogenation treatment is performed in the same manner as in Embodiment 5-3.

When the crystal grain size of the polysilicon thin film 523 formed as described above was measured using SEM and TEM, it was 1.5 μm, which was smaller than 0.3 μm in the conventional polysilicon film. It was confirmed that the particle size had increased. The field-effect mobility is 50 c
m ² / V · sec to 150 cm ² / V · sec, and the total defect density at the interface and the film of the polysilicon thin film 523 is 1.3 × 10 ¹² cm ⁻² eV ⁻¹ . It decreased to 8.7 × 10 ¹¹ cm ^-2 eV ^-1 . That is, by using both the excimer lamp 551 and the heater 561 when irradiating the laser beam 531a, the crystal grain size of the polysilicon thin film 523 and the film quality are further improved.

Also, as in Embodiment 5-3, the TFT
And gate voltage (Vg) -drain current (Id)
When the characteristics were measured, as shown in FIG. 50, it was confirmed that the rise of the drain current with respect to the gate voltage became even steeper than that in Embodiment 5-4, and the sub-threshold characteristics and the like were improved.

Note that when an experiment was conducted by setting the temperature of the glass substrate 521 variously, the temperature of the glass substrate 521 was set to 30.
Heating at 0 ° C. or higher can improve the crystal quality. However, at 600 ° C. or higher, distortion occurs in the glass substrate 521, making it difficult to manufacture elements such as TFTs.

(Embodiment 5-6) Embodiment 5-6 of the present invention will be described with reference to FIG. 54 and FIG.

First, similarly to Embodiment 5-4, an amorphous silicon thin film 522 is formed on a glass substrate 521, and dehydrogenation is performed.

Next, as shown in FIG.
The laser beam 571a of the KrF excimer laser 571 and the excimer lamp light 551a of the excimer lamp 551 are irradiated while moving the substrate 21 in the direction indicated by the arrow A, and the glass substrate 521 is heated by the heater 561.
Is heated from the bottom side to form a polysilicon thin film 523. Here, as compared with Embodiment 5-5, the excimer lamp light 551a is mainly used for the glass substrate 521.
Irradiating through the wavelength-selective reflector 572 from directly above, and K instead of the XeCl excimer laser 531
An rF excimer laser 571 is used, and a laser beam 5
71 a is different in that it is irradiated via a wavelength-selective reflector 572. The excimer lamp light 551a is 5 mm × 1 including the irradiation area 571b of the laser beam 571a.
The irradiation area 551b of 00 mm is irradiated. Other heating conditions and the like are the same as in Embodiment 5-5.

The wavelength-selective reflection plate 572 is 280n
Those that reflect light having a wavelength shorter than m and transmit light having a wavelength longer than 280 nm are used. Therefore, a KrF excimer laser 5 in which KrF is used for discharge is used.
The 71 laser beam 571a (wavelength is 248 nm)
The light is reflected by the wavelength-selective reflector 572 and irradiates the amorphous silicon thin film 522 almost perpendicularly, and the excimer lamp light 551a in the visible to ultraviolet region passes through the wavelength-selective reflector 572 and passes through the amorphous silicon thin film 522. Irradiates almost vertically.

Further, hydrogenation treatment is performed in the same manner as in Embodiment 5-3.

As described above, the crystal grain size, field effect mobility, and defect density of the polysilicon thin film 523 formed by irradiating the amorphous silicon thin film 522 vertically with the laser beam 571a and the excimer lamp light 551a are respectively As in Embodiment 5-5, 1.5 μm,
150 cm ² / V · sec, and 8.7 × 10 ¹¹ cm
^Although it was ⁻² eV ⁻¹ , the variation in the crystal grain size and the field effect mobility in each region of the polysilicon thin film 523 was further reduced, and almost uniform characteristics were obtained over the entire surface of the polysilicon thin film 523.

Also, as in Embodiment 5-3, the TFT
And gate voltage (Vg) -drain current (Id)
When the characteristics were measured, the same characteristics as those of Embodiment 5-5 were obtained as shown in FIG.

The selective reflection / transmission according to the wavelength of the laser beam or the like by the wavelength selective reflection plate 572 can be easily performed by using the KrF excimer laser 571 as described above. Xe
A short-wavelength laser using Br, KrCl, ArF, ArCl, or the like may be used.

(Embodiment 5-7) Embodiment 5-7 of the present invention will be described with reference to FIGS. 54 to 56 and FIG.

In this Embodiment 5-7, as compared with Embodiment 5-6, the irradiation area 551b of the excimer lamp light 551a is a 5 mm × 70 mm area, and the amorphous silicon thin film 5 formed by the excimer lamp 551
22 is different in that the heating temperature is set variously, and other heating conditions and the like are the same as those in Embodiment 5-6. That is, the irradiation intensity of the excimer lamp light 551a in FIG. 54 is adjusted, the heating temperature of the amorphous silicon thin film 522 is variously set in a range from room temperature to 1200 ° C., and the polysilicon thin film 523 is formed. Was measured for the crystal grain size and the field-effect mobility.

As shown in FIG. 55, when the amorphous silicon thin film 522 is heated to about 300 ° C. or higher, the crystal grain size of the polysilicon thin film 523 increases in accordance with the heating temperature. 5 μm or more. If the temperature exceeds 1000 ° C., the surface of the glass substrate 521 is partially melted and crystal growth is hindered, so that the crystal grain size becomes small.

Also, as shown in FIG. 56, when the amorphous silicon thin film 522 is heated to about 300 ° C. or higher, the field effect mobility of the polysilicon thin film 523 increases in accordance with the heating temperature. To a maximum of 450 cm ² / V · sec.
If it exceeds 0 ° C., it will decrease.

That is, in addition to the irradiation of the laser beam 571a, the glass substrate 521 is heated by the heater 561, and the amorphous silicon thin film 522 is irradiated with the excimer lamp light 551a.
By heating in the range of 00 ° C., in particular, the effect of expanding the crystal grain size of the polysilicon thin film 523 and improving the film quality can be obtained.

A TFT was formed on the polysilicon thin film 523 in the same manner as in Embodiment 5-3, and the gate voltage (Vg) -drain current (Id) characteristics were measured. As shown in FIG. 50 in the case where the heating temperature of 522 is 600 ° C., even better TFT characteristics were obtained than in Embodiments 5-5 and 5-6.

(Embodiment 5-8) Embodiment 5-8 of the present invention will be described with reference to FIGS. 57, 58, and 50. FIG.

In the embodiment 5-8, as compared with the embodiment 5-7, the excimer lamp 5 is mainly used.
Xe flash lamp 58 that emits pulse light instead of 51
1 is used.

More specifically, as shown in FIG. 57, the laser beam 571a of the KrF excimer laser 571 similar to that of the embodiment 5-7 is applied to the amorphous silicon thin film 52.
2 is 500 μm × 200 mm
Irradiation so as to form a band. On the other hand, Xe in the range from visible light to ultraviolet light emitted from the Xe flash lamp 581
The flash lamp light 581 a
An irradiation region 581b of 5 mm × 200 mm including the irradiation region 571b of 1a is irradiated so that the amorphous silicon thin film 522 is heated to about 1000 ° C. As shown in FIG. 58, the Xe flash lamp light 581a is synchronized with the irradiation pulse of the laser beam 571a,
Irradiation is performed as a pulse having a width extending before and after the irradiation pulse. Further, the irradiation pulse of the laser beam 571a is applied so as to have a pulse width of 2/3 or less of the irradiation cycle. Other heating conditions and the like are the same as in Embodiment 5-7.

The crystal grain size and the field-effect mobility of the polysilicon thin film 523 formed as described above are almost the same as those in the case where the amorphous silicon thin film 522 is heated to 1000 ° C. in the embodiment 5-7. However, in the embodiment 5-7, the glass substrate 521 is slightly distorted, whereas in the embodiment 5-8, the distortion is not caused, so that an appropriate semiconductor circuit can be more reliably formed. And so on. In addition, since the Xe flash lamp 581 has high heating efficiency and can heat a large area at a time, productivity can be easily improved.

Also, as in Embodiment 5-3, the TFT
And gate voltage (Vg) -drain current (Id)
When the characteristics were measured, as shown in FIG. 50, better TFT characteristics were obtained than in Embodiment 5-7.

(Embodiment 5-9) Embodiment 5-9 of the present invention will be described with reference to FIG. 59 and FIG.

In Embodiment 5-9, FIG.
As shown in the figure, a laser beam 571a emitted from a KrF excimer laser 571 similar to that in the embodiment 5-8.
Is reflected by the wavelength-selective reflection plate 572, and is 500 μm × 200 in the amorphous silicon thin film 522.
mm and a laser beam emitted from a YAG laser is irradiated by a KTP.
A laser beam 591a from a YAG laser device 591 that converts the wavelength into half using a crystal is reflected by a reflection plate 592, and is irradiated on an irradiation region 591b of 5 mm × 200 mm in the amorphous silicon thin film 522. As described above, the laser beam 571a and the laser beam 591a are vertically incident on the amorphous silicon thin film 522 by the wavelength-selective reflector 572 and the reflector 592. The irradiation timing and pulse width of laser beam 571a and laser beam 591a, the irradiation energy of laser beam 571a, the heating temperature of glass substrate 521 by heater 561, and the like are the same as those in Embodiment 5-8.

The heating temperature of the amorphous silicon thin film 522 by the laser beam 591a of the YAG laser device 591 is variously set in the range from room temperature to 1200 ° C. to form the polysilicon thin film 523, and the crystal grain size and the field effect mobility are formed. When the heating temperature of the amorphous silicon thin film 522 was 1100 ° C., the maximum values were 5.5 μm and 600 cm ² / V · sec, respectively. That is, by using the YAG laser device 591 for preheating, even if the amorphous silicon thin film 522 is heated to a relatively high temperature, the glass substrate 521 is distorted or melted and impurities are mixed into the polysilicon thin film 523. Thus, a more excellent crystalline polysilicon thin film 523 was obtained than in the case where the amorphous silicon thin film 522 was heated by the excimer lamp 551 in the embodiment 5-7. However, when the amorphous silicon thin film 522 was heated to 1200 ° C., both the crystal grain size and the field-effect mobility decreased. This is because microcrystalline silicon has already been formed at the time of preheating by the YAG laser device 591, and this has an adverse effect on crystal growth by the KrF excimer laser 571.

Also, as in Embodiment 5-3, the TFT
And gate voltage (Vg) -drain current (Id)
When the characteristics were measured, as shown in FIG. 50, better TFT characteristics were obtained than in Embodiment 5-8.

The laser device for preheating includes:
As described above, not only the YAG laser device 591 but also a pulse laser such as a XeCl excimer laser, for example, may use a KrF excimer laser 5 having a wavelength different from that of the KrF excimer laser 571 and a mixing ratio of gas.
A similar effect can be obtained by irradiating with a pulse width longer than 71 and at the timing shown in FIG. 58, for example. Further, a continuous oscillation laser device such as an Ar laser may be used.

In each of the above embodiments, an example in which silicon (Si) is used as a semiconductor has been described. However, the present invention is not limited to this. For example, germanium (Ge), gallium arsenide (GaAs), etc. Even when a group V semiconductor, a II-VI group semiconductor such as zinc selenium (ZnSe) or the like is used,
Conditions such as the heating temperature are not necessarily the same, but it has been confirmed that similar effects can be obtained. Further, silicon carbon (SiC), silicon germanium (SiGe), or the like may be used.

The irradiation of the amorphous silicon thin film 522 with the laser beam 531a is performed by the glass substrate 521.
Or from both sides of the amorphous silicon thin film 522 and the glass substrate 521.

In place of the glass substrate 521, quartz,
Alternatively, a substrate such as an organic material such as plastic may be used, or a conductive substrate having an insulating film formed on a surface may be used.

Also, a laser beam 53 for preheating is used.
2a and the like do not irradiate the entire region of the amorphous silicon thin film 522, but irradiate only the region where the high TFT characteristics are required.
Alternatively, only the laser beam 531a for crystallization may be irradiated.

(Embodiment 6-1) An example in which a thin film transistor as a semiconductor element is applied to a liquid crystal display device will be described below.

In the active matrix type liquid crystal display device, the thin film transistor provided in the image display area needs to have high uniformity of transistor characteristics in order to reduce unevenness of a display image, while the peripheral portion of the image display area is required. A thin film transistor used for a drive circuit (driver circuit) arranged in the semiconductor device needs high responsiveness.
However, it is not easy to achieve both uniformity of characteristics and high responsiveness despite various crystal growth methods and the like being studied. Therefore, in the present embodiment, a semiconductor film (amorphous silicon layer) formed on a substrate
By making the laser irradiation method different for each area of
The characteristics required for each area are obtained. That is, after irradiating the first laser light to the entire surface of the substrate or only to the image display area, the second laser light is applied to the drive circuit section area at a higher energy density than the first laser light irradiation. Irradiation. Hereinafter, the laser annealing apparatus and the laser annealing method will be specifically described with reference to the drawings.

The laser annealing apparatus used in this embodiment can be basically the one having the same configuration as the conventional apparatus shown in FIG. In FIG. 9, 151 is a laser oscillator, 152 is a reflecting mirror, and 15
Reference numeral 3 denotes a uniformizing device, 154 denotes a window, 155 denotes a substrate on which an amorphous silicon layer is formed, 156 denotes a stage, and 157 denotes a control device. During the laser annealing of the amorphous silicon layer, the laser beam oscillated from the laser oscillator 151 is guided to the homogenizer 153 by the reflecting mirror 152, and the laser beam shaped into a predetermined shape with uniform energy is applied to the window. The substrate 155 fixed to the stage 156 in the processing chamber is irradiated through 154. However, the control device 157 can irradiate the laser beam only for each predetermined region on the substrate 155, and can control the irradiation conditions to be different in each region.

Using the above laser annealing apparatus, first, a laser beam having a linear beam cross section (for example, a width of 300 μm and a length of 10 cm) is converted to a laser beam having an energy density of 280 mJ / cm ² through a homogenizer 153. While the substrate 155 is being moved, first laser light irradiation is performed to irradiate the entire surface of the substrate 155 while partially overlapping the irradiation areas (scanning irradiation using linear laser light). Note that this laser beam irradiation may be performed only on the image display area 155a shown in FIG. Next, second laser light irradiation for irradiating the drive circuit portion regions 155b and 155c with laser light is performed at an energy density of 400 mJ / cm ² higher than the above (scanning irradiation using linear laser light).

Here, as the substrate 155, for example, a substrate obtained by forming an amorphous silicon layer with a thickness of 500 ° on a glass substrate by plasma CVD and then performing a dehydrogenation treatment at 450 ° C. for 1 hour is used. Was. The laser beam oscillated at an interval of 300 Hz with a pulse width of, for example, 25 ns, and relatively scanned with the laser beam while moving the substrate 155 at a predetermined speed. Further, in the second laser beam irradiation, as shown in FIG. 61, scanning was performed so that the irradiation regions were overlapped by 30 μm (the overlap ratio was 10%). In this case, although the characteristics such as the mobility are uneven between the region where the laser light is repeatedly irradiated and the region where the laser light is not irradiated, as shown in FIG.
If 610 or the like is formed so as not to join the joint and the joint is used for a wiring pattern or the like, it becomes easy to reduce variations in TFT characteristics and the like. Also, the second
In the laser beam irradiation, the linear beam direction is set in a direction parallel to each side of the substrate 155 (the drive circuit section region 155b,
If the scanning is performed in a direction perpendicular to each side in the direction indicated by the solid line in 155c), the total time required for irradiation can be shortened. For this purpose, laser irradiation may be performed by rotating the stage on which the substrate 155 is fixed by 90 degrees (the linear beam direction of the laser light may be rotated by 90 degrees, but this is generally difficult. .).

While the first laser light irradiation crystallization is performed so that the uniformity of the semiconductor film characteristics required for the image display area 155a is maintained, the second laser light irradiation drives the drive circuit. In the partial regions 155b and 155c, high field-effect mobility can be obtained. That is, when the present inventors performed laser irradiation under various irradiation conditions, when the scanning irradiation was performed at an energy density of 300 mJ / cm ² or more, the electric field effect movement at the joint of the irradiation region in each scan was performed. It has been found that the degree of unevenness is likely to occur.
Thus, in the image display region 155a where the in-plane uniformity of the characteristics of the polycrystalline silicon is required as described above, the laser light is irradiated at an energy density lower than 300 mJ / cm ^2, while the area is larger than the pixel region. 300 mJ / cm in the drive circuit section regions 155b and 155c which require small characteristics and high characteristics such as high field-effect mobility.
By irradiating the laser light with an energy density higher than ² , even if the uniformity and improvement of the film characteristics are not compatible,
Image display area 155a and drive circuit section areas 155b, 15
5c, it is possible to form polycrystalline silicon layers having different characteristics, which meet respective needs.

(Embodiment 6-2) Another example in which a thin film transistor is applied to a liquid crystal display device will be described.

In this example, when the first laser beam is radiated, the beam cross-sectional shape is linear as in Embodiment 6-1. On the other hand, when the second laser beam is radiated, The difference is that the beam cross-sectional shape is square.

The laser annealing apparatus used in this embodiment is different from the apparatus shown in FIG. 9 in that, as shown in FIG.
The difference is that, instead of the uniformizing device 153, a uniformizing device A621 for shaping the beam cross-sectional shape of the laser beam into a linear shape and a homogenizing device B622 for shaping the beam into a square shape (for example, 1 cm square) are provided. (Note that the same components as those in FIG. 9 are denoted by the same reference numerals and description thereof is omitted.) First, as shown in FIG. 62A, the substrate is passed through a homogenizing device A621 using the laser annealing device. Laser annealing is performed on the entire surface 155 or only the image display area 155a with linear laser light at an energy density of 280 mJ / cm ² that maintains uniformity (scanning irradiation using linear laser light). Then, as shown in FIG. 62 (b), an energy density of 400 mJ / cm is applied to the irradiation areas 631, 632 of the drive circuit section areas 155b, 155c by using the homogenizing device B622 as shown in FIG. ^2. Irradiate the angular laser light (scanning irradiation using the angular laser light).

As described above, in the case where the laser light for the second laser light irradiation is made angular, the driving circuit portion area is not rotated by 90 degrees as in the embodiment 6-1. Laser annealing of 155b and 155c can be performed. Therefore, similarly to Embodiment 6-1, polycrystalline silicon having different characteristics between the pixel portion region and the drive circuit portion region can be obtained, and the device and the manufacturing process can be easily simplified.

(Embodiment 6-3) Embodiment 6-
The second laser beam irradiation in 2 may be performed a plurality of times. That is, the substrate 155 is not moved at the time of the second laser beam irradiation as in the case of 6-2, and each irradiation corresponding to the square laser beam shape in the drive circuit section regions 155b and 155c as shown in FIG. For each of the regions 631 and 632, the irradiation position of the laser light may be fixed and stationary irradiation may be performed. For example, each irradiation region 631 and 632 has a square shape of 1 cm, When the laser beams overlap each other, it is easy to greatly increase the field-effect mobility in a region where laser light is not repeatedly irradiated, and uniformity in the region can be improved. Here, when the energy density of the laser light is high as described above, the characteristics such as the mobility are uneven between the region irradiated with the laser light and the other region, but the driving circuit region 155b , 155c,
Uniform uniformity over the entire area such as the image display area 155a is not always necessary. That is, a driving circuit may be formed so that a pattern (TFT pattern) of a semiconductor film does not overlap a joint (laser beam end) of a laser shot, and the joint may be used as a wiring pattern or the like. That is, only a portion of the polycrystalline silicon having uniform characteristics may be used for forming a TFT or the like. Even when such a region to be illuminated in a superimposed manner is not used, its area is relatively small.
The usage efficiency of 55b and 155c does not decrease so much.

The irradiation of the laser beam is performed in each irradiation area 63.
By performing a plurality of times (for example, 30 times) for each 1,632, the film characteristics can be further improved. Here, the relationship between the number of times of stationary irradiation and the obtained mobility of polycrystalline silicon is shown in FIG. As is clear from the figure, there is an appropriate range for the number of laser irradiations, and the mobility decreases even if the number of irradiations is smaller or larger than this range. When the energy density of the stationary irradiation is 400 mJ / cm ² , the number of times of irradiation is 50 or more, preferably 80
Polycrystalline silicon having high field-effect mobility can be obtained in one to four hundred times. Note that the effect of improving the field effect mobility by performing the stationary irradiation of the laser light on the drive circuit portion region as described above is less than the case where the linear laser light is used, even though the linear laser light is used. can get.

(Embodiment 6-4) The laser beam irradiation conditions are different only in the image display area 155a and the drive circuit section areas 155b and 155c as in the above embodiments 6-1 to 6-3. Instead, different regions may be used for different regions, and polycrystalline silicon having different characteristics may be formed in the driving circuit region. That is, for example, at the time of the second laser beam irradiation in Embodiment 6-3, a high mobility is required in a region where a transfer gate of a latch or a shift register is formed in the driving circuit portion regions 155b and 155c. Irradiation at a high energy density (for example, 400 mJ / cm ² ), and for other parts, priority should be given to uniformity due to reduction of noise and variation in adjustment, and to improvement in productivity by expanding the irradiation area by expanding the beam. Irradiation may be performed at about 330 mJ / cm ² . The same effect can be obtained by changing the number of irradiations as the difference in the irradiation conditions.

(Embodiment 6-5) More examples of the laser annealing apparatus will be described.

This laser annealing apparatus is different from the apparatus shown in FIG. 9 in that, as shown in FIG.
6 in that a mask member 641 having a partially different transmittance of laser light is provided. The mask member 6
As shown in FIG. 66, 41 includes an attenuation region 641a corresponding to the image display region 155a of the substrate 155, and a transmission region 641b corresponding to the drive circuit unit regions 155b and 155c. Specifically, for example, by partially covering the quartz plate with an optical thin film such as an ND filter or a dielectric multilayer film, the transmittance of the laser light is partially set to a predetermined value, and the transmittance of the pixel portion is reduced. The energy density of laser light irradiation can be reduced.

By using the laser annealing apparatus as described above, a laser beam or the mask member 641 and the stage 156 are moved to irradiate, for example, a linearly shaped laser beam on the entire surface of the substrate. is irradiated at an energy density of 400 mJ / cm ^2,
Embodiment 6-1 is applied to the image display area 155a.
Similarly, laser annealing can be performed at an energy density of 280 mJ / cm ² . That is, it becomes possible to simultaneously form semiconductor films having different characteristics in the pixel portion and the driver circuit portion.

Note that the mask member 641 is not limited to being arranged with the window 154 or the substrate 155 as shown in FIG. 65, but may be brought into close contact with the substrate 155 to improve the flatness of the laser light irradiation surface. Alternatively, the mask member 641 and the window 154 may be the same, and may be provided in the homogenizing device 153. Further, the laser light may not be attenuated by a refraction optical system. A device that changes the strength may be used.

(Embodiment 6-6) An example of a laser annealing apparatus capable of further improving the uniformity in the image display area will be described.

In this laser annealing apparatus, as shown in FIG. 67, a uniformizing optical element 651 for scattering an incident laser beam is provided above a substrate 155. Accordingly, it is possible to reduce unevenness in the amount of light generated due to diffraction or the like in the shape of the laser beam, and to prevent the reflected light from the substrate 155 from returning to the laser oscillator to make the laser pulse unstable.

Also, as shown in FIG.
By using a compound uniform optical element 652 having the transparent region 52a and the mirror-finished transmissive region 652b, a region having high uniformity and a region having high crystallinity in the semiconductor layer may be formed at the same time. .

Each of the above embodiments can obtain the above-described effects. However, as long as the operations do not contradict each other, the configurations of the embodiments are combined to obtain the effects and synergistic effects of the respective operations. The effect may be obtained.

[0400]

The present invention is embodied in the form described above, and has the following effects.

That is, according to the present invention, a region where a transistor is to be formed can be a polycrystalline silicon thin film having a larger grain size, and transistor characteristics such as field-effect mobility can be greatly improved. For example, there is an effect that a large-scale driving circuit can be built in a liquid crystal display device or the like. In addition, by using a silicon nitride oxide thin film in which oxygen is added to silicon nitride as an insulating film, the hydrogen content in the film can be reduced and stress can be reduced, so that a more stable transistor can be obtained. In addition, the crystal grain size and crystal orientation can be controlled, and interference between crystals in the course of crystal growth can be prevented, so that a sufficient crystal grain size can be obtained. Further, according to the present invention, the timing at which crystal nuclei are generated in the peripheral portion is earlier than in the related art, so that crystal growth can be performed earlier than in the related art.

Further, according to the present invention, a crystal growth direction control region such as a crystal growth direction control gap for controlling the crystal growth direction in the direction of the source region and the drain region is provided at least in the channel region of the non-single-crystal semiconductor layer. In addition, since large crystal grains having a long grain size are formed in a direction connecting the source region and the drain region, a crystalline thin film transistor having a small grain boundary density in this direction can be obtained. Excellent in TFT characteristics such as mobility.

Also, since the uniformity of crystal grains and the degree of crystallinity are improved by means of appropriately adjusting the intensity pattern of the light beam, according to the present invention, such a circuit can be obtained without adversely affecting other circuits. In addition, a crystallization region having a higher field effect mobility can be formed only in a limited specific portion on the substrate. Therefore, for example, it is possible to integrally form a pixel transistor and a driver circuit which requires a mobility several tens to several hundred times higher than this on the same substrate. In addition, since the CPU and the like can be formed integrally on the same substrate, according to the present invention, high performance,
An excellent effect is obtained in that a highly integrated AM-LCD or the like can be provided at low cost.

By irradiating at least two types of energy beams having different absorptivity of the precursor semiconductor film, the precursor semiconductor film is heated in the thickness direction and the substrate is also heated. The body semiconductor film is crystallized while being gradually cooled. Therefore, crystal growth is promoted, relatively large crystal grains are formed, crystal defects are reduced, and the electrical characteristics of the semiconductor film are improved. In addition, since the substrate can be heated in a shorter time than when a heater or the like is used, productivity can be improved.

Further, it is possible to form a plurality of regions having different characteristics, that is, a region having a high characteristic of the semiconductor film and a region having a high characteristic uniformity in the substrate surface. In a built-in thin film transistor array for a liquid crystal panel, high characteristics required for a circuit portion and high uniformity required for a pixel portion can be realized.

[Brief description of the drawings]

FIG. 1 is a graph showing transmittance characteristics of an amorphous silicon thin film.

FIG. 2 is a plan view and a cross-sectional view schematically showing a conventional thin film transistor (TFT).

FIG. 3 is an explanatory view showing a conventional method of manufacturing a polysilicon thin film.

FIG. 4 is a diagram showing an intensity pattern of a light beam having a flat light intensity distribution according to the related art.

FIG. 5 is a schematic diagram showing non-uniformity of crystallinity in a crystallized region according to the related art.

FIG. 6 is a diagram showing a Raman intensity curve at the line AA in FIG. 5;

FIG. 7 is an explanatory diagram for explaining the progress of crystallization when a light beam having a flat light intensity distribution is used.

FIG. 8 is an explanatory view showing the principle of polycrystallization by laser light irradiation according to a conventional method.

FIG. 9 is a schematic diagram of a conventional laser annealing apparatus.

FIG. 10 is an explanatory view showing a laser annealing region of a liquid crystal display.

FIG. 11 is an explanatory view showing a crystal growth direction in an a-Si film provided with a crystal growth direction control gap.

FIG. 12 is an explanatory diagram illustrating the principle of polycrystallization in Embodiment 1-1.

FIG. 13 is a graph showing the degree of crystallization of the polycrystalline silicon thin film of the embodiment 1-1.

14A and 14B are a plan view and a cross-sectional view of a TFT in Embodiment 1-2.

FIG. 15 is a plan view and a cross-sectional view of a TFT of Embodiment 1-3.

16A and 16B are a plan view and a cross-sectional view of a TFT in Embodiment 2-1.

17A and 17B are a plan view and a cross-sectional view of a TFT in Embodiment 2-2.

FIG. 18 is an explanatory diagram showing a manufacturing process of the TFT in Embodiment 2-1.

FIG. 19 is an explanatory diagram illustrating a manufacturing process of a TFT in Embodiment 2-2.

20A and 20B are a plan view and a cross-sectional view illustrating a structure of a TFT in Embodiment 3-1 (FIG. 20A is a plan view, and FIG. 20B is a cross-sectional view taken along line AA ′ in FIG. 20A).

FIG. 21 is a cross-sectional view taken along line BB ′ in FIG.
Sectional view

FIG. 22 is an explanatory diagram illustrating a manufacturing process of the TFT in Embodiment 3-1.

23A and 23B are a plan view and a cross-sectional view illustrating a structure of a TFT in Embodiment 3-2 (FIG. 23A is a plan view, and FIG. 23B is a CC ′ cross-sectional view in FIG. 23A).

FIG. 24 is DD ′ of FIG. 23A of the embodiment 3-2.
Sectional view

FIG. 25 is an explanatory diagram showing a manufacturing process of the TFT of Embodiment 3-2.

FIG. 26 is a plan view showing the structure of a TFT in Embodiment 3-3.

FIG. 27 is a plan view showing a configuration of a TFT having a crystal growth direction control gap according to a modified example of the embodiments 3-1 to 3-3.

FIG. 28 is a cross-sectional view showing a configuration of a TFT having a crystal growth direction control gap according to another modification of the embodiments 3-1 to 3-3.

FIG. 29 is a diagram illustrating the progress of crystallization when a light beam having a light intensity gradient is used.

FIG. 30 is a diagram schematically showing a state in which a light beam having a light intensity gradient is irradiated while being moved.

FIG. 31 is a diagram showing light transmission characteristics of a filter for producing a light beam having a light intensity gradient.

FIG. 32 is a view for explaining the progress of crystallization in the case where a light beam in which portions having relatively high light intensity and portions having relatively low light intensity are alternately arranged in a plane is used.

FIG. 33 is a diagram schematically showing a state in which the light beam having the distribution pattern shown in FIG. 32a is irradiated while moving.

FIG. 34 is a view showing light transmission characteristics of a filter for producing the light beam shown in FIG. 32;

FIG. 35 is a diagram showing a light intensity distribution pattern in another mode of a light beam in which a portion having a relatively high light intensity and a portion having a relatively low light intensity are alternately arranged in a plane.

FIG. 36 is a schematic view showing the principle of creating the light intensity distribution pattern shown in FIG. 35 by light interference.

FIG. 37 is a view for explaining the progress of crystallization when the light beam shown in FIG. 35 is used.

FIG. 38 is a schematic diagram for explaining a method for producing a light beam from a dynamic interference pattern in which a bright line portion and a dark line portion wave.

FIG. 39 is a schematic view showing a state in which a light interference pattern is formed in the thickness direction of the thin film.

FIG. 40 is a diagram showing a state in which heat flows from a thin film heated by light irradiation to the surroundings.

FIG. 41 is a diagram showing the relationship between the atmospheric pressure and the number of times of irradiation and the degree of crystallization (Raman intensity) during light irradiation.

FIG. 42 is a schematic view showing a state where crystallization is performed using an excimer laser.

FIG. 43 is a view showing an experimental apparatus for examining a relationship between an ambient pressure and a crystallinity in laser annealing.

FIG. 44 is an explanatory diagram showing the method for manufacturing the polysilicon thin film of the embodiment 5-1.

FIG. 45 is a graph showing Raman scattering measurement results of the polysilicon thin films of the embodiments 5-1 and 2;

FIG. 46 is an explanatory diagram showing the method for manufacturing the polysilicon thin film of the embodiment 5-2.

FIG. 47 is a graph showing transmittance characteristics of glass.

FIG. 48 is a perspective view showing a structure of a glass substrate over which a microcrystalline silicon thin film of Embodiment 5-3 is formed.

FIG. 49 is an explanatory diagram showing the method of manufacturing the polysilicon thin film of the embodiment 5-3.

FIG. 50 is a graph showing the characteristics of the TFTs of the embodiments 5-3 to 9;

FIG. 51 is an explanatory view showing another method of manufacturing the polysilicon thin film of the embodiment 5-3.

FIG. 52 is an explanatory diagram showing the method of manufacturing the polysilicon thin film of the embodiment 5-4.

FIG. 53 is an explanatory view showing the method of manufacturing the polysilicon thin film of the embodiment 5-5.

FIG. 54 is an explanatory view showing a method of manufacturing a polysilicon thin film of the embodiments 5-6 and 7;

FIG. 55 is a graph showing the relationship between the heating temperature and the crystal grain size in Embodiment 5-7.

FIG. 56 is a graph showing the relationship between the heating temperature and the field-effect mobility in Embodiment 5-7.

FIG. 57 is an explanatory view showing the method of manufacturing the polysilicon thin film of the embodiment 5-8.

FIG. 58 is an explanatory diagram showing irradiation timing in Embodiment 5-8.

FIG. 59 is an explanatory view showing the method of manufacturing the polysilicon thin film of the embodiment 5-9.

FIG. 60 illustrates an irradiation region of a liquid crystal display with laser light in Embodiment 6-1.

FIG. 61 is an explanatory diagram illustrating a method for applying laser light in Embodiment 6-1.

FIG. 62 is a schematic view of a laser annealing apparatus according to Embodiment 6-2.

FIG. 63 is an explanatory diagram illustrating an irradiation region of a laser beam in Embodiments 6-2 and 3;

FIG. 64 is a graph showing dependence of mobility on the number of laser irradiations in Embodiment 6-3.

FIG. 65 is a schematic view of a laser annealing apparatus according to Embodiment 6-5.

FIG. 66 is a plan view showing a configuration of a mask member according to Embodiment 6-5.

FIG. 67 is an explanatory view showing the laser annealing method of the embodiment 6-6.

FIG. 68 is an explanatory view showing another laser annealing method of the embodiment 6-6.

[Explanation of symbols]

 151 Laser oscillator 152 Reflector 153 Uniform device 154 Window 155 Substrate 155a Image display area 155b, 155c Driving circuit section area 156 Stage 157 Controller 158 Image display area 159 Peripheral circuit section 201 Transparent insulating substrate 202 Lower insulating film 203 Upper insulating Film 204 amorphous silicon thin film 205 gate insulating film 206 source electrode film 207 drain electrode film 208 gate electrode film 210 polycrystalline silicon thin film 210a small crystal grain region 210b large crystal grain region 301 insulating substrate 302 undercoat layer 303 semiconductor layer 303a Projection 303d Drain region 303s Source region 304 First insulating layer 305 Gate electrode 306 Second insulating layer 307d Drain electrode 307s Source electrode 308 Resist 401 Insulating substrate 40 2 Undercoat layer 403 p-Si film 403a Channel region 403b Source region 403c Drain region 404 First insulating film 405 Gate electrode 406 Second insulating film 407d Drain electrode 407s Source electrode 408 Contact hole 410 TFT 411 Crystal growth direction control gap 413 a-Si film 414 resist mask 420 TFT 421 early crystallization region 430 TFT 431 crystal growth direction control gap 521 glass substrate 522 amorphous silicon thin film 523 polysilicon thin film 524 microcrystalline silicon thin film 531 excimer lasers 531a, 532a laser beams 531b, 532b Irradiation area 532 Laser 533 Half mirror 534 Infrared lamp 534a Infrared 535 Substrate stage 541 Chamber 541a Irradiation 542 Incandescent lamp 542a Incandescent light 543 Laser light attenuator 544 Homogenizer (laser light homogenizer) 545 Reflector 551 Excimer lamp 551a Excimer lamp light 551b Irradiation area 552 Concave reflector 561 Heater 571 KrF excimer laser 571a Laser beam 571a Laser beam 571a Wavelength-selective reflector 581 Flash lamp 581a Flash lamp light 581b Irradiated area 591 YAG laser device 591a Laser beam 591b Irradiated area 592 Reflector 607 Control unit 610 TFT 621 Uniform unit A 622 Uniform unit B 631,632 Irradiated region 641 Mask Member 641a Attenuation area 641b Transmission area 651 Uniform optical element 652 Composite uniform optical element 652a Scattering area 652b Transmission area 7 1 Temperature Distribution Curve 702 Crystallization Temperature Line 703 Thin Film Cross Section 704 Crystal 711 Irradiation Surface 712 Irradiation Surface 713 Overlap Irradiation Area 721 H Area 722 L Area 723 Crystallization Temperature Line 724 Crystal Nucleus 725 Thin Film Cross Section 726 Near Center 731 Irradiation Surface Position 732 Irradiation surface position 733 Overlap irradiation region 801, 802 Light beam 901 Curve 902 Crystallization temperature line 903 Crystal nucleus 1101 Thin film 1102 Underlayer 1103 Glass substrate 1200 Irradiation surface 1201 Thin film 1202 Underlayer 1203 Substrate 1400 Light beam irradiation device 1401 Laser light generator 1402 mirror 1403 beam homogenizer 1404 region 1410 irradiation object 1411 glass substrate 1412 non-single-crystalline silicon film 1413 underlayer 1500 airtight container 1501 window 1502 hydrogen gas Cylinder 1503 amorphous silicon layer 1510 laser irradiation apparatus 1511 excimer laser beam generator 1512 mirrors 1513 beam homogenizer 1701 boundaries 1702 near the boundary 1703 hatched portion 1704 central 1901 Temperature distribution curve 1902 the crystallization temperature line 1903 crystal nuclei

 ──────────────────────────────────────────────────続 き Continuation of the front page (31) Priority claim number Japanese Patent Application No. 10-67993 (32) Priority date March 18, 1998 (March 18, 1998) (33) Priority claim country Japan (JP) (31) Priority claim number Japanese Patent Application No. 10-138318 (32) Priority date May 20, 1998 (May 20, 1998) (33) Priority claim country Japan (JP) (31) Priority claim number Japanese Patent Application No. 10-163130 (32) Priority Date June 11, 1998 (June 11, 1998) (33) Countries claiming priority Japan (JP) (72) Inventor Yuji Satani 1006 Kazuma Kadoma, Kadoma City, Osaka Prefecture Address Matsushita Electric Industrial Co., Ltd. (72) Inventor Yoshinao Taketomi 1006 Ojidoma, Kadoma-shi, Osaka, Japan Matsushita Electric Industrial Co., Ltd. 72) Inventor Akira Nishitani 1006 Kadoma, Kadoma, Osaka Prefecture Matsushita Electric Within Sangyo Co., Ltd. (72) Inventor Mikihiko Nishitani 1006 Kazuma Kadoma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Inventor Masashi Goto 1006 Oji Kadoma, Kadoma City, Osaka Prefecture Matsushita Electric Industrial Co., Ltd. (72) Invention Person Miko Mino 1006 Kazuma Kadoma, Kadoma City, Osaka Prefecture F-term in Matsushita Electric Industrial Co., Ltd. (reference) DA06 DB02 DB03 EA01 EA11 EA12 FA02 FA04 FA06 FA25 FA26 HA06 JA01 5F110 AA16 AA17 AA30 BB02 BB04 BB05 BB07 BB09 BB10 CC02 CC05 CC07 CC08 DD02 DD13 DD15 DD17 EE03 EE04 EE06 EE09 GG01 GG02 GG02 GG02 GG02 GG02 GG02 GG02 GG04 GG04 GG02 GG04 GG04 PP02 PP03 PP04 PP05 PP06 PP08 PP13 PP24 PP29 PP36 QQ02 QQ11 QQ19 QQ25

Claims (97)

[Claims] A first heat conductive material having a first thermal conductivity on a substrate;
Laminating an insulating film having a second thermal conductivity different from the first thermal conductivity and a second insulating film selectively formed in a partial region; Laminating a non-single-crystal semiconductor thin film on the insulating film and the second insulating film; and irradiating the non-single-crystal semiconductor thin film with an energy beam to grow crystals. Production method. 2. The method of manufacturing a semiconductor thin film according to claim 1, wherein said second insulating film is laminated on said substrate after said first insulating film is laminated on said substrate. A method for producing a semiconductor thin film, wherein the rate is set lower than the first thermal conductivity. 3. The method for manufacturing a semiconductor thin film according to claim 2, wherein said first insulating film is made of one of a silicon nitride compound and a silicon nitride oxide compound, Comprises a silicon oxide compound. 4. The method according to claim 1, wherein the first heat conductive material has a first thermal conductivity on the substrate.
Laminating an insulating film having a second thermal conductivity different from the first thermal conductivity and a second insulating film selectively formed in a partial region; Stacking a non-single-crystal semiconductor thin film on the insulating film and the second insulating film, and irradiating the non-single-crystal semiconductor thin film with an energy beam to grow a crystal; Forming a semiconductor element using a region corresponding to a higher thermal conductivity of the first insulating film and the second insulating film. 5. The method for manufacturing a semiconductor device according to claim 4, wherein, in the semiconductor thin film grown by crystal growth, the one having a lower thermal conductivity among the first insulating film and the second insulating film. A method for manufacturing a semiconductor device, comprising removing a corresponding region and forming a semiconductor device using the remaining region. 6. The method for manufacturing a semiconductor device according to claim 4, wherein after laminating the first insulating film on the substrate, the second insulating film is laminated, and the second heat conduction is performed. A method for manufacturing the semiconductor device, wherein the rate is set lower than the first thermal conductivity. 7. The method of manufacturing a semiconductor device according to claim 6, wherein the first insulating film is made of one of a silicon nitride compound and a silicon nitride oxide compound, and the second insulating film Comprises a silicon oxide compound. 8. The method for manufacturing a semiconductor device according to claim 4, wherein the second insulating film is selectively formed in a region patterned in a stripe shape. 9. The method for manufacturing a semiconductor device according to claim 8, wherein the irradiation of the energy beam is performed while scanning in a longitudinal direction of the stripe pattern of the second insulating film, and the scanning direction is substantially the same. To match the direction of the current path,
A method for manufacturing a semiconductor element, comprising forming the semiconductor element. 10. A first insulating film having a first thermal conductivity, laminated on the substrate and the substrate, having a second thermal conductivity different from the first thermal conductivity, A second insulating film selectively formed in a partial region; and a semiconductor in which a non-single-crystal semiconductor thin film laminated on the first insulating film and the second insulating film is crystal-grown by irradiation with an energy beam. A semiconductor element having a thin film, wherein a region corresponding to a higher thermal conductivity of the first insulating film and the second insulating film in the crystal-grown semiconductor thin film is formed in a current path. A semiconductor element characterized by being provided in a semiconductor device. 11. The semiconductor device according to claim 10, comprising a plurality of regions in the crystal-grown semiconductor thin film, wherein each of the regions is provided in a plurality of current paths in which current flows in parallel. A semiconductor element characterized by the following. 12. A semiconductor thin film in which a non-single-crystal semiconductor thin film is crystal-grown by irradiation with an energy beam, and a projection extending outward in the same plane as the semiconductor thin film is formed at a peripheral portion of the semiconductor thin film. A semiconductor thin film characterized by being made. 13. The semiconductor thin film according to claim 12, wherein said projection is formed to a size such that one crystal nucleus is generated during crystal growth by irradiation of said energy beam. Thin film. 14. The semiconductor thin film according to claim 13, wherein a length of the protrusion in the protrusion direction is not less than the thickness of the semiconductor thin film and not more than 3 μm. Semiconductor thin film. 15. The semiconductor thin film according to claim 13, wherein a length of the protrusion in a width direction orthogonal to a protruding direction is formed to be not less than the thickness of the semiconductor thin film and not more than 3 μm. Characteristic semiconductor thin film. 16. The semiconductor thin film according to claim 12, wherein the semiconductor thin film is formed in a shape having a pair of opposed sides, and the plurality of protrusions are formed on each of the opposed sides. A semiconductor thin film, wherein an interval between the adjacent protrusions formed on each side is set substantially equal to an interval between the opposing sides. 17. A semiconductor device having a semiconductor thin film in which a non-single-crystal semiconductor thin film is crystal-grown by irradiation of an energy beam. A semiconductor element, wherein a projection is formed. 18. The semiconductor device according to claim 17, wherein a thin film transistor having a source region, a gate region, and a drain region formed by the semiconductor thin film is formed, and the protrusion is formed at least in the gate region. A semiconductor element formed on a peripheral portion. 19. A step of forming a non-single-crystal semiconductor thin film on a substrate, the non-single-crystal semiconductor thin film having a protrusion extending outward in the same plane as the non-single-crystal semiconductor thin film; Growing the non-single-crystal semiconductor thin film by energy beam irradiation. 20. The method for manufacturing a semiconductor thin film according to claim 19, wherein the energy beam includes at least one of a laser beam, an electron beam, and an ion beam. Method. 21. The method for manufacturing a semiconductor thin film according to claim 20, wherein the energy beam includes an excimer laser beam. 22. A method of manufacturing a semiconductor thin film in which a non-single-crystal semiconductor thin film is crystallized by an annealing treatment, wherein a crystal nucleus in a peripheral portion of the non-single-crystal semiconductor thin film is generated earlier than a crystal nucleus in a central portion. And then
A method of manufacturing a semiconductor thin film, wherein the crystal nucleus generated in the peripheral portion is crystal-grown toward a central portion before the crystal nucleus is generated or crystal-grown in the central portion. 23. The method of manufacturing a semiconductor thin film according to claim 22, wherein, in the annealed semiconductor thin film, the crystal nuclei in the peripheral portion of the semiconductor thin film are cooled by cooling the peripheral portion faster than the central portion. Characterized in that it is generated earlier than the crystal nuclei in (1). 24. The method of manufacturing a semiconductor thin film according to claim 23, wherein the peripheral portion includes a peripheral edge having a substantially protruding shape, and heat generated and accumulated by annealing at the peripheral portion is parallel to the semiconductor thin film. A method for manufacturing a semiconductor thin film, wherein a plurality of escape directions are provided in a plane direction, and a peripheral portion is cooled faster than a central portion. 25. A semiconductor device having a semiconductor thin film obtained by crystallizing a non-single-crystal semiconductor thin film by annealing, wherein a crystal nucleus in a peripheral portion of the non-single-crystal semiconductor thin film is earlier than a crystal nucleus in a central portion. And then
A semiconductor element having a semiconductor thin film in which the crystal nuclei generated in the peripheral portion are crystal-grown toward the central portion before the crystal nuclei are generated or crystal-grown in the central portion. 26. A semiconductor device in which a crystalline semiconductor layer having a channel region, and a source region and a drain region arranged on both sides of the channel region is formed on a substrate, wherein the crystalline semiconductor layer comprises: A semiconductor element, which is obtained by crystallizing a non-single-crystal thin film, wherein a crystal growth direction control gap for controlling a crystal growth direction is provided in at least a channel region of the crystalline semiconductor layer. 27. The semiconductor device according to claim 26, wherein said crystal growth direction control gap is formed by providing two or more rows of groove-shaped gaps in a direction connecting a source region and a drain region. A semiconductor element characterized by the above-mentioned. 28. The semiconductor device according to claim 26, wherein the plurality of crystal growth direction control gaps are discontinuously provided in a direction connecting a source region and a drain region. 29. A semiconductor device in which a crystalline semiconductor layer having a channel region, and a source region and a drain region disposed on both sides of the channel region is formed on a substrate, wherein the crystalline semiconductor layer comprises: A semiconductor device comprising a non-monocrystalline thin film crystallized, wherein at least a channel region is provided with an early crystallization region having a higher crystallization start temperature than a channel region body. . 30. The semiconductor device according to claim 29, wherein the early crystallization region is long in a direction connecting the source region and the drain region. 31. The semiconductor device according to claim 29, wherein the early crystallization region is formed by adding an impurity to a component constituting a channel region main body. 32. The semiconductor device according to claim 26, wherein the crystalline semiconductor layer is mainly composed of silicon or a compound of silicon and germanium. 33. A method for manufacturing a semiconductor device, comprising: forming, on a substrate, a crystalline semiconductor layer having a channel region, and a source region and a drain region disposed on both sides of the channel region. Depositing a non-single crystalline thin film on a substrate; forming a crystal growth direction control gap in the non-single crystalline thin film; Irradiating an energy beam to crystallize the thin film. 34. The method of manufacturing a semiconductor device according to claim 33, wherein the crystal growth direction control gap is formed in a groove shape in a direction connecting a source region and a drain region. Method. 35. The method of manufacturing a semiconductor device according to claim 33, wherein the plurality of crystal growth direction control gaps are discontinuously formed in a direction connecting a source region and a drain region. Manufacturing method. 36. A method for manufacturing a semiconductor device comprising a crystalline semiconductor layer having a channel region, and a source region and a drain region disposed on both sides of the channel region, the method comprising the steps of: Depositing a non-single-crystalline thin film; and ion-implanting an impurity that raises the crystallization start temperature of the non-single-crystalline semiconductor thin film into a part of the non-single-crystalline semiconductor thin film to form an early crystallization region containing the impurity. A method of manufacturing a semiconductor device, comprising: a crystallization region forming step; and, after the early crystallization region forming step, irradiating an energy beam to crystallize the thin film. 37. The method for manufacturing a semiconductor device according to claim 36, wherein in the early crystallization region forming step, a band-like early crystallization region long in a direction connecting the source region and the drain region is formed. A method for manufacturing a semiconductor device, comprising: 38. The method of manufacturing a semiconductor device according to claim 36, wherein said early crystallization region is discontinuously arranged in a direction connecting said source region and said drain region. Manufacturing method. 39. The method for manufacturing a semiconductor device according to claim 33, wherein the energy beam is an excimer laser beam. 40. Production of a semiconductor thin film formed by irradiating a non-single crystalline thin film formed on a substrate with a light beam to crystallize or recrystallize the non-single crystalline material to form a crystalline semiconductor thin film. In the method, as the light beam, a light beam whose light energy intensity distribution pattern is adjusted so that a temperature gradient or a non-uniform temperature distribution is generated on the surface of the thin film that is the irradiated surface, and the light beam is stationary. A method for producing a semiconductor thin film, comprising irradiating in a state. 41. The method of manufacturing a semiconductor thin film according to claim 40, wherein the distribution pattern of the light energy intensity is such that the light intensity within the beam width monotonically increases from one side to the other side or monotonically decreases from one side to the other side. A method for manufacturing a semiconductor thin film, wherein the semiconductor thin film has a distribution pattern. 42. The method of manufacturing a semiconductor thin film according to claim 40, wherein the distribution pattern of the light energy intensity includes a portion having a relatively high light intensity and a portion having a relatively low light intensity within a beam width. A method for manufacturing a semiconductor thin film, wherein the semiconductor thin film has a distribution pattern alternately arranged in a plane. 43. The method of manufacturing a semiconductor thin film according to claim 42, wherein the distribution pattern of the light energy intensity is at least two.
A method of manufacturing a semiconductor thin film, wherein the semiconductor thin film is formed by simultaneously irradiating two coherent lights to cause light interference. 44. The method of manufacturing a semiconductor thin film according to claim 42, wherein the energy intensity distribution pattern simultaneously irradiates at least two coherent lights and shifts the phase of at least one of the lights. A method for manufacturing a semiconductor thin film, wherein the method is a wave interference pattern formed by dynamic modulation. 45. Production of a crystalline semiconductor thin film for irradiating a non-single crystalline thin film formed on a substrate with a light beam and then radiating heat to crystallize or recrystallize the non-single crystalline material The method of manufacturing a semiconductor thin film according to claim 1, wherein the manufacturing method includes generating a non-uniform temperature distribution on a surface of the thin film irradiated with the light beam by maintaining an ambient atmosphere pressure at a predetermined value or more. 46. The method of manufacturing a semiconductor thin film according to claim 45, wherein the atmospheric pressure equal to or higher than the predetermined value is equal to or higher than 10 -5 torr when an atmospheric gas is a hydrogen gas. 47. A first energy beam for applying at least an energy capable of crystallizing the precursor semiconductor film to the precursor semiconductor film formed on the substrate, and the first energy beam. Irradiating the precursor semiconductor film with a second energy beam that gives the precursor semiconductor film an energy smaller than the energy that can crystallize the precursor semiconductor film, and the absorptivity of the precursor semiconductor film is smaller than the beam; A method for manufacturing a semiconductor film, comprising a step of crystallizing a precursor semiconductor film. 48. The method for manufacturing a semiconductor film according to claim 47, wherein the precursor semiconductor film is an amorphous silicon thin film. 49. The method of manufacturing a semiconductor film according to claim 47, wherein in the first energy beam, the absorption coefficient of the precursor semiconductor film is substantially equal to or greater than the reciprocal of the thickness of the precursor semiconductor film. And a second energy beam, wherein the absorption coefficient of the precursor semiconductor film is not more than the reciprocal of the thickness of the precursor semiconductor film. 50. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam has an absorption coefficient of the precursor semiconductor film that is approximately 10 times the reciprocal of the thickness of the precursor semiconductor film.
A method of manufacturing a semiconductor film, wherein the second energy beam has an absorption coefficient of the precursor semiconductor film that is substantially the reciprocal of a thickness of the precursor semiconductor film. 51. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam are light beams having different wavelengths from each other. . 52. The method of manufacturing a semiconductor film according to claim 51, wherein the first energy beam is a single-wavelength energy beam, and the second energy beam is at least a wavelength component in a visible light region. A method for manufacturing a semiconductor film, characterized by comprising light. 53. The method according to claim 52, wherein the first energy beam is a laser beam, and the second energy beam is an infrared lamp. Manufacturing method of membrane. 54. The method according to claim 52, wherein the first energy beam is a laser beam and the second energy beam is an incandescent beam. Manufacturing method of membrane. 55. The method of manufacturing a semiconductor film according to claim 52, wherein said first energy beam is a laser beam, and said second energy beam is an excimer lamp beam. A method for manufacturing a semiconductor film. 56. The method of manufacturing a semiconductor film according to claim 51, wherein the second energy beam is light containing at least a wavelength component from a visible light region to an ultraviolet light region. Production method. 57. The method for manufacturing a semiconductor film according to claim 56, wherein the first energy beam is a laser beam, and the second energy beam is a xenon flash lamp light. Of manufacturing a semiconductor film. 58. The method of manufacturing a semiconductor film according to claim 51, wherein the first energy beam and the second energy beam are laser beams. 59. The method of manufacturing a semiconductor film according to claim 58, wherein said precursor semiconductor film is an amorphous silicon thin film, and said first energy beam is an argon fluorine excimer laser or a krypton fluorine excimer laser. A laser beam of any one of xenon chlorine excimer laser and xenon fluorine excimer laser, and the second energy beam is a laser beam of an argon laser. 60. The method of manufacturing a semiconductor film according to claim 58, wherein said substrate is a glass substrate, said precursor semiconductor film is an amorphous silicon thin film, and said first energy beam is argon. A laser light of any one of a fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser, and a xenon fluorine excimer laser, wherein the second energy beam is a laser beam of a carbon dioxide gas laser. Of manufacturing a semiconductor film. 61. The method for manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam irradiate a strip-shaped region in the precursor semiconductor film. A method for manufacturing a semiconductor film. 62. The method of manufacturing a semiconductor film according to claim 47, wherein an irradiation area of the second energy beam on the precursor semiconductor film is formed on the precursor semiconductor film by the first energy beam. A method for manufacturing a semiconductor film, which is larger than an irradiation area and includes an irradiation area of the first energy beam. 63. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam are irradiated so as to enter the precursor semiconductor film substantially perpendicularly. A method for manufacturing a semiconductor film. 64. The method of manufacturing a semiconductor film according to claim 47, wherein the second energy beam is irradiated at least prior to irradiating the first energy beam. Production method. 65. The method of manufacturing a semiconductor film according to claim 64, wherein the substrate on which the precursor semiconductor film is formed is moved, and the second energy beam is applied to the first semiconductor film in the precursor semiconductor film. A method for manufacturing a semiconductor film, comprising irradiating a position on the forward side in the moving direction with respect to the irradiation position of the energy beam. 66. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam is applied intermittently, while the second energy beam is applied continuously. Of manufacturing a semiconductor film. 67. The method of manufacturing a semiconductor film according to claim 66, wherein the first energy beam is a pulsed laser beam, and the second energy beam is a continuous oscillation laser beam. A method for manufacturing a semiconductor film. 68. The method of manufacturing a semiconductor film according to claim 66, wherein the first energy beam is a pulsed laser beam, and the second energy beam is a lamp beam. A method for manufacturing a semiconductor film. 69. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam are intermittently irradiated in synchronization with each other. Manufacturing method. 70. The method of manufacturing a semiconductor film according to claim 69, wherein the irradiation with the first energy beam is performed during the second energy beam irradiation.
Within the period of irradiating the energy beam of
A method of manufacturing the semiconductor film, wherein the period is not more than two thirds of the irradiation cycle of the energy beam. 71. The method of manufacturing a semiconductor film according to claim 69, wherein the first energy beam and the second energy beam are pulsed laser light. . 72. The method of manufacturing a semiconductor film according to claim 69, wherein the first energy beam is a pulsed laser beam, and the second energy beam is intermittently lit. A method of manufacturing a semiconductor film, wherein the light is a light of the following order. 73. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam heat the precursor semiconductor film to a temperature of 300 ° C. or more and 1200 ° C. or less. The method for manufacturing a semiconductor film, comprising: 74. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam and the second energy beam heat the precursor semiconductor film to a temperature of 600 ° C. or more and 1100 ° C. or less. The method for manufacturing a semiconductor film, comprising: 75. The method for manufacturing a semiconductor film according to claim 47, further comprising a step of heating the substrate on which the precursor semiconductor film is formed by a heater. 76. The method of manufacturing a semiconductor film according to claim 75, wherein the substrate on which the precursor semiconductor film is formed has a temperature of 300 ° C. or higher.
A method for manufacturing a semiconductor film, comprising heating to a temperature of 00 ° C. or lower. 77. The method of manufacturing a semiconductor film according to claim 47, wherein the first energy beam irradiates a plurality of regions in the precursor semiconductor film, and the second energy beam irradiates the plurality of regions. A method of manufacturing a semiconductor film, which comprises irradiating only a part of the region. 78. The method of manufacturing a semiconductor film according to claim 47, wherein the second energy beam has an absorptance in the substrate larger than an absorptivity in the precursor semiconductor film. A method for manufacturing a semiconductor film. 79. The method of manufacturing a semiconductor film according to claim 78, wherein the first energy beam has an absorption coefficient of the precursor semiconductor film that is approximately 10 times the reciprocal of the thickness of the precursor semiconductor film.
A method for manufacturing a semiconductor film, wherein the method is at least twice as large. 80. The method according to claim 78, wherein the substrate is a glass substrate, the precursor semiconductor film is an amorphous silicon thin film, and the first energy beam is argon. A laser light of any one of a fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser, and a xenon fluorine excimer laser, wherein the second energy beam is a laser beam of a carbon dioxide gas laser. Of manufacturing a semiconductor film. 81. An apparatus for manufacturing a semiconductor film for crystallizing a precursor semiconductor film formed on a substrate, comprising: first irradiation means for irradiating a first energy beam; A second irradiating means for irradiating the precursor semiconductor film with a second energy beam having a small absorption rate. 82. The semiconductor film manufacturing apparatus according to claim 81, wherein said second irradiation means is a lamp which emits a second energy beam radially, and further collects said second energy beam. An apparatus for manufacturing a semiconductor film, comprising a concave reflecting mirror that emits light. 83. The semiconductor film manufacturing apparatus according to claim 81, further comprising: a reflecting plate that reflects one of the first energy beam and the second energy beam and transmits the other. An apparatus for manufacturing a semiconductor film, wherein the first energy beam and the second energy beam are configured to be incident substantially perpendicularly on the precursor semiconductor film. 84. The semiconductor film manufacturing apparatus according to claim 81, wherein said precursor semiconductor film is an amorphous silicon thin film, and said first irradiation means is an argon fluorine excimer laser or a krypton fluorine excimer laser. An apparatus for producing a semiconductor film, wherein the second irradiation means is an argon laser or any one of a xenon chlorine excimer laser and a xenon fluorine excimer laser. 85. The apparatus for manufacturing a semiconductor film according to claim 81, wherein said substrate is a glass substrate, said precursor semiconductor film is an amorphous silicon thin film, and said first energy beam is argon. A laser light of any one of a fluorine excimer laser, a krypton fluorine excimer laser, a xenon chlorine excimer laser, and a xenon fluorine excimer laser, wherein the second energy beam is a laser beam of a carbon dioxide gas laser. Semiconductor film manufacturing equipment. 86. A method of manufacturing a semiconductor thin film, comprising: irradiating a non-single-crystal semiconductor thin film formed on a substrate having an image display region and a drive circuit region with an energy beam to grow crystals. The first irradiation on the image display area is performed using an energy beam having a linear beam cross section, while the second irradiation on the driving circuit section area is performed using an energy beam having a square cross section. A method of manufacturing a semiconductor thin film, wherein the first irradiation is performed at a higher energy density than the first irradiation. 87. A method for manufacturing a semiconductor thin film, comprising: irradiating a non-single-crystal semiconductor thin film formed on a substrate having an image display region and a drive circuit region with an energy beam to grow a crystal, The first irradiation to the image display area is scanning irradiation in which an energy beam is scanned relative to the substrate and the energy beam irradiation area is shifted while being shifted by a predetermined overlap amount, while the driving circuit The second irradiation to the partial region is stationary irradiation performed by fixing an energy beam relative to the substrate,
A method of manufacturing a semiconductor thin film, wherein the method is performed at a higher energy density than the first irradiation. 88. The method for manufacturing a semiconductor thin film according to claim 87, wherein the second irradiation is performed a plurality of times with an energy beam fixed relative to the substrate. Manufacturing method of thin film. 89. A method for manufacturing a semiconductor thin film, comprising: irradiating an energy beam on a non-single-crystal semiconductor thin film formed on a substrate having an image display region and a drive circuit portion region to grow a crystal, The energy beam is applied to the image display area and a plurality of predetermined areas in the drive circuit section area at different energy densities, and at a higher energy density in the drive circuit section area than in the image display area. Irradiating the semiconductor thin film. 90. The method for manufacturing a semiconductor thin film according to claim 89, wherein a transfer gate forming at least one of a latch circuit and a shift register is formed in each of the regions in the drive circuit section region. A method of manufacturing a semiconductor thin film, wherein irradiation of an energy beam to a region is performed at a higher energy density than irradiation of an energy beam to another region. 91. A method of manufacturing a semiconductor thin film, comprising: irradiating an energy beam on a non-single-crystal semiconductor thin film formed on a substrate having an image display region and a drive circuit portion region to grow a crystal, Simultaneously irradiating the image display area and the drive circuit section area with the energy beam through a filter whose area corresponding to the image display area has a lower transmittance of the energy beam than the area corresponding to the drive circuit section area. A method of manufacturing a semiconductor thin film. 92. A method for manufacturing a semiconductor thin film, comprising a step of irradiating a non-single-crystal semiconductor thin film formed on a substrate with an energy beam to grow a crystal, wherein the irradiation of the energy beam is performed by scattering energy beams. A method for producing a semiconductor thin film, wherein the method is performed via a uniformizing element having the following. 93. The method for manufacturing a semiconductor thin film according to claim 92, wherein the equalizing element has a partially transparent region of the energy beam, and the energy beam incident on the transparent region. And irradiating the non-single-crystal semiconductor thin film with light as it is. 94. An energy beam generating means, and uniformizing means for shaping the energy beam emitted from the energy beam generating means into a predetermined beam cross-sectional shape with uniform energy, wherein the shaped energy beam is An apparatus for manufacturing a semiconductor thin film for irradiating a non-single-crystal semiconductor thin film formed on a substrate to grow a crystal, further comprising: a filter having regions having different energy beam transmittances from each other; And a plurality of regions in the non-single-crystal semiconductor thin film are irradiated with the energy beams at different energy densities from each other. 95. The apparatus for manufacturing a semiconductor thin film according to claim 94, wherein said filter is constituted by an optical thin film so as to have regions in which transmittances of said energy beams are different from each other. Thin film manufacturing equipment. 96. An apparatus for manufacturing a semiconductor thin film according to claim 94, further comprising a chamber in which said substrate is disposed, wherein irradiation of said energy beam is performed through a window formed in said chamber. An apparatus for manufacturing a semiconductor thin film, wherein said filter is provided in said window. 97. An energy beam generating means, comprising: an energy beam generating means; and a uniformizing means for shaping the energy beam emitted from the energy beam generating means into a predetermined beam cross-sectional shape having uniform energy. An apparatus for manufacturing a semiconductor thin film for growing a crystal by irradiating a non-single-crystal semiconductor thin film formed on a substrate, wherein the uniformizing means can selectively switch an energy beam to a plurality of beam cross-sectional shapes to shape the beam. An apparatus for manufacturing a semiconductor thin film characterized by being configured as described above.