JP2001127304A - Semiconductor device and manufacturing method therefor, and electronic device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a laser annealing method for a crystalline semiconductor film with large crystal size. SOLUTION: When an amorphous semiconductor film is crystallized under laser beam irradiation, the front and rear surfaces of the amorphous semiconductor film are irradiated with laser beam. The relation 0<I0'/I0<1 or 1<I0'/I0 is established where I0'/I0 is a ratio between an effective energy intensity of laser beam projected on the front surface and an effective energy intensity I0 of laser beam projected on the rear surface.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to annealing of a semiconductor film using laser light (hereinafter referred to as laser annealing).
And a laser apparatus (an apparatus including a laser and an optical system for guiding laser light output from the laser to an object to be processed). Further, the present invention relates to a semiconductor device formed by such a laser annealing method and a method for manufacturing the same.

[0002]

2. Description of the Related Art In recent years, thin film transistors (hereinafter referred to as TFTs) have been developed.
Has been developed, and a TFT using a polycrystalline silicon film (polysilicon film) as a crystalline semiconductor film has attracted attention. In particular, liquid crystal displays (liquid crystal displays)
And EL (electroluminescence) display devices (EL displays) are used as elements for switching pixels and elements for forming a drive circuit for controlling the pixels.

As a means for obtaining a polysilicon film, a technique of crystallizing an amorphous silicon film (amorphous silicon film) into a polysilicon film is generally used. In particular, recently, a method of crystallizing an amorphous silicon film using laser light has attracted attention. In this specification, means for crystallizing an amorphous semiconductor film with laser light to obtain a crystalline semiconductor film is called laser crystallization.

[0004] Laser crystallization is a technique capable of instantaneously heating a semiconductor film and effective as an annealing means for a semiconductor film formed on a substrate having low heat resistance such as a glass substrate or a plastic substrate. Further, the throughput is much higher than that of a heating means using a conventional electric furnace (hereinafter, referred to as furnace annealing).

There are various types of laser light, and generally, laser crystallization using laser light (hereinafter referred to as excimer laser light) using a pulse oscillation type excimer laser as an oscillation source is used. . An excimer laser has an advantage that it has a large output and can be repeatedly irradiated at a high frequency, and has an advantage that an excimer laser beam has a high absorption coefficient with respect to a silicon film.

At present, the most noticeable problem is how to increase the crystal grain size of a crystalline semiconductor film crystallized by laser light. As a matter of course, as one crystal grain (also referred to as a grain) becomes larger, the number of crystal grain boundaries that traverse the TFT, particularly, the channel formation region decreases. Therefore, it is possible to improve variations in typical electric characteristics of the TFT such as the field-effect mobility and the threshold voltage.

Further, the inside of each crystal grain maintains relatively clean crystallinity, and in order to improve the above-mentioned various characteristics of the TFT, the channel forming region is completely contained within one crystal grain. It is desirable to form a TFT in such a manner.

However, it is difficult to obtain a crystalline semiconductor film having a sufficiently large crystal grain size with the current technology.
Although there are reports that they were obtained experimentally, they have not yet reached the level of practical use.

Experimentally, "" High-Mobility Poly-Si Thi
n-Film Transistors Fabricated bya Novel Excimer La
ser Crystallization Method ", K.Shimizu, O.Sugiura
andM.Matumura, IEEE Transactions on Electron Devic
es vol.40, No.1, pp112-117, 1993 ”. This reference discloses that Si / SiO
A three-layer structure of ₂ / n ⁺ Si is formed, and excimer laser light is irradiated from both directions of the Si side and the n ⁺ Si side. It is shown that such a configuration can increase the crystal grain size.

[0010]

SUMMARY OF THE INVENTION The present invention is directed to solving the above-mentioned problems, and is directed to a laser annealing method for obtaining a crystalline semiconductor film having a large crystal grain size and a laser apparatus used for the laser annealing method. The task is to provide Another object is to provide a semiconductor device using such a laser annealing method and a manufacturing method thereof.

[0011]

SUMMARY OF THE INVENTION The gist of the present invention is that when crystallizing an amorphous semiconductor film, a laser beam is applied to the surface of the amorphous semiconductor film (the surface on which the thin film is superposed) and the back surface ( And the effective energy intensity of the laser light (hereinafter referred to as primary laser light) applied to the front surface and the laser light applied to the back surface (hereinafter referred to as second laser light). (Referred to as secondary laser light).

That is, when the effective energy intensity of the primary laser light is (I ₀ ) and the effective energy intensity of the secondary laser light is (I ₀ ′), the effective energy intensity ratio (I ₀ )
₀ ′ / I ₀ ) to “0 <I ₀ ′ / I ₀ <1” or “1 <I ₀ ′ /
It is characterized by irradiating a laser beam so that the relationship of “I ₀ ” is satisfied. Of course, I ₀ · I ₀ ′ ≠ 0.

In this specification, the term “effective energy intensity” refers to the energy intensity of a laser beam when it reaches the front or back surface of the amorphous semiconductor film, and is an energy intensity in consideration of energy loss due to reflection or the like. Here, the unit is defined as density: mJ / cm ² . Although it cannot be measured, it can be obtained from the calculation of reflectance and transmittance if the medium existing in the path of the laser beam is known.

For example, a specific method of calculating the effective energy intensity when the present invention is applied to the structure shown in FIG. 6 will be described. In FIG. 6, 601 is a reflector made of aluminum, 602 is a Corning # 1737 substrate (0.7 mm thick), 603 is a 200-nm-thick silicon nitride oxide film (hereinafter referred to as SiON film), and 604 is 55 nm
It is a thick amorphous silicon film. A case where such a sample is irradiated with XeCl excimer laser light having a wavelength of 308 nm in the air will be taken as an example.

The energy intensity of the laser light (wavelength 308 nm) immediately before reaching the amorphous silicon film 604 is defined as (I _a ). At this time, the effective energy intensity (I ₀ ) of the primary laser light is given by I ₀ = I _{0 in} consideration of the reflection of the laser light on the surface of the amorphous silicon film.
_a is represented by (1−R _Si ). Here, R _Si is the reflectance of the laser beam. In this case, the I ₀ = 0.45I _a in the calculation.

The effective energy intensity (I ₀ ′) of the secondary laser light is I ₀ ′ = I _a T ₁₇₃₇ R _Al T ₁₇₃₇ (1-R
_SiON-Si ). Here, T ₁₇₃₇ is the transmittance of the # 1737 substrate, R _Al is the reflectance on the aluminum surface, R
_SiON-Si is the reflectivity when entering the amorphous silicon film from inside the SiON film. In addition, SiON
Reflectance when entering the film, transmittance in SiON film, Si
The reflectivity when entering the # 1737 substrate from inside the ON film and the reflectivity when entering the SiON film from inside the # 1737 substrate were not included in the calculation because it was experimentally found to be negligible. In this case, the I ₀ '= 0.13I _a in the calculation.

Accordingly, in the case of the structure shown in FIG. 6, the effective energy intensity (I ₀ ) of the primary laser light is 0.45I _a , and the effective energy intensity (I ₀ ′) of the secondary laser light is 0.13.
Obtained as I _a. That is, the effective energy intensity ratio (I ₀ ′
/ I ₀ ) is 0.29. One of the features of the present invention is that the effective energy intensity ratio obtained as described above satisfies 0 <I ₀ ′ / I ₀ <1.

Further, the present invention can be realized even when the intensity of the primary laser light is smaller than the intensity of the secondary laser light. That is, the effective energy intensity ratio is 1 <I ₀ ′ / I ₀
The present invention is also satisfied when the condition is satisfied.

In order to make the effective energy intensities of the primary laser light and the secondary laser light different, the following method can be used. 1) When irradiating laser light to the front and back surfaces of the amorphous semiconductor film using a reflector provided below the substrate, the effective energy intensity of the secondary laser light is adjusted by adjusting the reflectance of the reflector. Attenuating the laser beam so that it is relatively smaller than the effective energy intensity of the primary laser beam. 2) The primary laser light is split in the middle to form a secondary laser light, and the effective energy intensity of the primary laser light or the secondary laser light is attenuated by a filter (variable attenuator, etc.). And making the effective energy intensities of the two relatively different. 3) A method in which the effective energy intensity of the secondary laser light is attenuated depending on the material of the substrate on which the amorphous semiconductor film is formed, and is relatively smaller than the effective energy intensity of the primary laser light. 4) An insulating film is interposed between the substrate and the amorphous semiconductor film, and the effective energy intensity of the secondary laser light is attenuated by the insulating film so as to be relatively smaller than the effective energy intensity of the primary laser light. Method. 5) By covering the surface of the amorphous semiconductor film with an insulating film and reducing the reflectance of the primary laser light on the surface of the amorphous semiconductor film, the effective energy intensity of the primary laser light is relatively reduced. Method to make the effective energy intensity of the secondary laser light larger than the effective energy intensity. 6) A method in which the amorphous semiconductor film is covered with an insulating film, and the effective energy of the primary laser light is attenuated by the insulating film so as to be relatively smaller than the effective energy of the secondary laser light. . 7) A method in which the primary laser light and the secondary laser light are formed using different lasers as oscillation sources, and the effective energy intensities thereof are different.

The present invention does not depend on the type of laser, but generally known excimer laser (typically, KrF laser or XeCl laser), solid laser (typically, Nd: YAG laser or ruby laser). A gas laser (typically, an argon laser or a helium-neon laser), a metal vapor laser (typically, a copper vapor laser or a helium-cadmium laser), or a semiconductor laser can be used.

When a laser beam having a long fundamental wave (first harmonic: wavelength 1064 nm), such as an Nd: YAG laser, is used, the second, third, or fourth harmonic is used. Is preferred. These harmonics can be obtained using a non-linear crystal (non-linear element).
Further, a known Q-switch method may be used.

[Circumstances leading up to the invention] Here, the background that the present applicant came to the present invention will be described based on experimental results. SEM (Scanning Electron Mi) shown in FIG.
The croscopy photograph is a photograph after the polysilicon film formed by laser crystallization is subjected to Secco etching.
For more information on Seco etching technology, see "F.Secco d'A
ragona: "Dislocation Etch for (100) Planes in Silic
on ".J.Electrochem.soc.Vol.119.No.7.pp.948-950 (197
2) ”.

In both cases, a silicon oxide film (thickness: 200 mm) was placed on a Corning # 1737 substrate (thickness: 0.7 mm).
nm) through an amorphous silicon film (thickness 55n).
m) is formed, and is obtained by irradiating an excimer laser beam. The excimer laser light used in this experiment was a pulse laser light having a wavelength of 308 nm using XeCl gas as an excitation gas, a pulse width was 30 ns, the number of shots was 20, and the energy density was 370 mJ / cm.
^{And 2} .

FIG. 7A shows a polysilicon film (average crystal grain diameter is about 0.3 μm) obtained by irradiating laser light only to the surface of the amorphous silicon film, and FIG. 7B shows the amorphous silicon film. A polysilicon film obtained by irradiating a laser beam to the front and back surfaces of the semiconductor device (the average crystal grain size is about 1.5 μm)
m). According to this, the polysilicon film obtained by irradiating the laser light to the front and back surfaces of the amorphous silicon film has a crystal grain size about 5 times larger than that of the amorphous silicon film, and it has been confirmed that irradiation from both surfaces is very effective. Was done.

In the present specification, the definition of the average crystal grain size is based on the definition of the average diameter of the crystal grain region in the specification of Japanese Patent Application Laid-Open No. 11-219133.

As described above, it has been confirmed that the crystal grain size can be increased by irradiating the front and back surfaces of the amorphous semiconductor film with laser light. In the experiments in the literature described in the conventional example, the back surface of the semiconductor film to be crystallized is not directly irradiated with the laser beam, but aims at the heat storage effect using the residual heat of n ⁺ Si. The configuration is completely different from the experiment performed by humans.

Next, the present applicant conducted a similar experiment using a quartz substrate instead of a glass substrate (provided that the energy density of the laser beam was 200 mJ / cm ² ).
As a result, the result as shown in FIG. 8 (SEM photograph after Seco etching) was obtained.

FIG. 8A shows a polysilicon film obtained by irradiating only the surface of the amorphous silicon film with a laser beam, and FIG. 8B shows the polysilicon film obtained by irradiating the laser beam on the front and back surfaces of the amorphous silicon film. This is a polysilicon film obtained by the above method. According to this, when a quartz substrate was used as the substrate, the average crystal grain size was at most about 0.4 to 0.5 μm, and the increase in grain size as shown in FIG. 7B could not be confirmed. In addition, no difference was observed in the crystal grain size when the irradiation was performed from one side or both sides of the substrate. That is, as described above, the effect of increasing the average crystal grain size was not confirmed despite the laser light irradiation on the front and back surfaces of the amorphous semiconductor film.

Then, the present applicant considered the above experimental results, and the difference between the experiments shown in FIGS. 7 and 8 is that the transmittance of the glass substrate (about 50%) and the transmittance of the quartz substrate (about 93%).
%), That is, the difference in the effective energy intensity of the laser light applied to the back surface of the amorphous semiconductor film.
The following experiment was performed for confirmation.

First, in this experiment, a sample having the structure shown in FIG. 6 was manufactured using a quartz substrate as the substrate 602 and a tantalum nitride film as the reflector 601. Then, this sample was irradiated with XeCl excimer laser light under the same conditions as those for obtaining the photograph of FIG. 7 (B), and the average crystal grain size of the obtained polysilicon film was shown on the SEM photograph after Seco etching. Confirmed. The result is shown in FIG.

As can be seen from FIG. 9, it is confirmed that the crystal grains of the obtained polysilicon film are distributed in almost the same state as the polysilicon film of FIG. 7B. In the case of the sample obtained in the photograph of FIG. 7B, the effective energy intensity ratio between the primary laser light and the secondary laser light is 0.5.
29 has already been mentioned. This is substantially the result of the attenuation of the secondary laser light on the glass substrate. Similarly, as a result of calculating the effective energy intensity ratio for the sample of the present experiment, a value of 0.33 was obtained. This is a result of the secondary laser light being substantially attenuated by the reflector.

The sample of FIG. 8B (combination of a reflector made of quartz and aluminum) and the sample of FIG. 9 (combination of a reflector made of quartz and tantalum nitride) are the same except that the material of the reflector surface is different. The structure is different from that of the sample of FIG. 9 only in that the reflectance of the reflector surface is smaller than that of the sample of FIG.

Considering the above results, when the front and back surfaces of the amorphous semiconductor film are crystallized by irradiating laser light, the effective energy intensity of the laser light (primary laser light) on the front surface side is larger than the effective energy intensity. It was found that when the effective energy intensity of the laser beam (secondary laser beam) on the back side was small, an increase in the average crystal grain size was confirmed.

[0034]

[Embodiment 1] One embodiment of the present invention will be described. FIG. 1A is a diagram showing a configuration of a laser device of the present invention. This laser apparatus includes a laser 101, an optical system 201 that linearly transforms laser light using the laser 101 as an oscillation source, and a stage 102 for fixing a light-transmitting substrate.
03 and the heater controller 104 are provided to keep the substrate at a temperature in the range of room temperature to 550 ° C. A reflector 105 is provided on the stage 102, and a substrate 106 on which an amorphous semiconductor film is formed is provided.

Next, a method for holding the substrate 106 in the laser device having the structure shown in FIG. 1A will be described with reference to FIG. Substrate 106 held on stage 102
Is installed in the reaction chamber 107, and is irradiated with a linear laser beam having the laser 101 as an oscillation source. The inside of the reaction chamber can be reduced in pressure or in an inert gas atmosphere by an exhaust system or a gas system (not shown), and can be heated to 100 to 450 ° C. without contaminating the semiconductor film.

The stage 102 is mounted on the guide rail 10.
8, the substrate can be moved in the reaction chamber, and the entire surface of the substrate can be irradiated with linear laser light. The laser light enters from a quartz window (not shown) provided on the upper surface of the substrate 106. In FIG. 1B, a transfer chamber 109, an intermediate chamber 110, and a load / unload chamber 111 are connected to the reaction chamber 107, and these chambers are separated by gate valves 112 and 113.

In the load / unload chamber 111, a cassette 114 capable of holding a plurality of substrates is provided.
Transfer robot 11 provided in transfer room 109
The substrate is transported by 5. The substrate 106 'represents the substrate being transported. With such a configuration, laser annealing can be continuously performed under reduced pressure or in an inert gas atmosphere.

Next, an optical system 20 for linearizing the laser light
1 will be described with reference to FIG. FIG. 2A is a diagram of the optical system 201 viewed from the side, and FIG. 2B is a diagram of the optical system 201 viewed from the top.

A laser beam having the laser 101 as an oscillation source is split vertically by a cylindrical lens array 202. The split laser light is further split in the horizontal direction by the cylindrical lens 203. That is,
The laser light is applied to the cylindrical lens arrays 202 and 20.
3 will eventually be divided into a matrix.

Then, the laser light is once collected by the cylindrical lens 204. At this time, the light passes through the cylindrical lens 205 immediately after the cylindrical lens 204. After that, the light is reflected by the mirror 206, passes through the cylindrical lens 207, and reaches the irradiation surface 208.

At this time, the laser light projected on the irradiation surface 208 shows a linear irradiation surface. That is, it means that the cross-sectional shape of the laser beam transmitted through the cylindrical lens 207 is linear. Homogenization of the linearly deformed laser light in the width direction (short direction) is performed by the cylindrical lens array 202, the cylindrical lens 204, and the cylindrical lens 207. The homogenization in the length direction (long direction) of the laser beam is performed by the cylindrical lens array 203 and the cylindrical lens 2.
05.

Next, a configuration for irradiating a laser beam from the front and back surfaces of the film to be processed formed on the substrate will be described with reference to FIG. FIG. 3 is a diagram showing a positional relationship between the substrate 106 and the reflector 105 in FIG.

In FIG. 3, reference numeral 301 denotes a light-transmitting substrate, and an insulating film 302 and an amorphous semiconductor film (or a microcrystalline semiconductor film) 303 are provided on the surface thereof (the surface on which a thin film or element is formed). Is formed. In addition, the light-transmitting substrate 301
A reflector 304 for reflecting the laser light is disposed below.

As the translucent substrate 301, a glass substrate, a quartz substrate, a crystallized glass substrate, or a plastic substrate is used. The transmissive substrate 301 itself can adjust the effective energy intensity of the secondary laser light. Further, as the insulating film 302, an insulating film containing silicon such as a silicon oxide film or a silicon nitride oxide film (SiOxNy) may be used, and the effective energy intensity of the secondary laser light may be adjusted with the insulating film 302. Further, the amorphous semiconductor film 3
03 includes a compound semiconductor film such as an amorphous silicon germanium film in addition to the amorphous silicon film.

The reflector 304 may be a substrate having a metal film formed on the surface (reflection surface of laser light) or a substrate made of a metal element. In this case, any material may be used for the metal film. Typically, silicon (Si), aluminum (Al), silver (Ag), tungsten (W), titanium (Ti), tantalum (Ta)
A metal film containing any one of the above elements is used. For example, tungsten nitride (WN), titanium nitride (TiN), and tantalum nitride (TaN) may be used.

Further, the reflector 304 is formed on the transparent substrate 3.
01 may be provided, or may be provided separately. Instead of disposing the reflector 304, it is also possible to form the above-described metal film directly on the back surface (the surface opposite to the front surface) of the substrate 301 and reflect the laser light there. In any case, the reflectivity of the reflector 304 can be used to adjust the effective energy intensity of the secondary laser light. When the reflector 304 is provided separately from the light-transmitting substrate 301, the energy intensity of the secondary laser light can be controlled by a gas filling the gap.

Then, the amorphous semiconductor film 303 is irradiated with the linearly deformed laser light via the optical system 201 described with reference to FIG. The irradiation of the linearly deformed laser light is performed by scanning the laser light.

In any case, the cylindrical lens 2
07, the first laser light 305 irradiating the front surface of the amorphous semiconductor film 303 and the second laser light once reflected by the reflector 304 and irradiating the back surface of the amorphous semiconductor film 303. 306 and the effective energy intensity ratio (I ₀ ′ / I
₀ ) satisfies the relationship of 0 <I ₀ ′ / I ₀ <1 or 1 <I ₀ ′ / I ₀ . For this purpose, the reflector 30
The reflectivity of the laser light of No. 4 is preferably 20 to 80%. At this time, a desired intensity ratio may be obtained by combining a plurality of means for attenuating the effective energy intensity of the secondary laser light described in this embodiment.

The laser beam having passed through the cylindrical lens 207 has an incident angle of 45 to 90 ° with respect to the substrate surface in the process of being focused. Therefore, the secondary laser light 306 is also applied to the back surface side of the amorphous semiconductor film 303 while wrapping around. In addition, the secondary laser light 306 can be obtained more efficiently by providing an uneven portion on the reflection surface of the reflector 304 and irregularly reflecting the laser light.

[Embodiment 2] An embodiment different from Embodiment 1 in this embodiment will be described. In the present embodiment, an example is shown in which two systems of laser light separated in the middle of the optical system are irradiated from the front surface and the back surface of the amorphous semiconductor film without using the reflector as in the first embodiment.

FIG. 4A is a diagram showing the configuration of the laser device of this embodiment. Since the basic configuration is the same as that of the laser apparatus of FIG. 1 described in the first embodiment, the description will be made with the reference numerals of the different parts changed.

This laser apparatus comprises a laser 101, an optical system 401 for linearly transforming a laser beam emitted from the laser 101 as an oscillation source and splitting the laser light into two systems, and a translucent stage for fixing a translucent substrate. 402. A substrate 403a is provided on the stage 402, and an amorphous semiconductor film 403b is formed thereon.

In the case of this embodiment, since the amorphous semiconductor film 403b is irradiated with the laser beam transmitted through the stage 402, the stage 402 must have a light transmitting property. In addition, since the laser light (secondary laser light) emitted from the side of the stage 402 passes through the stage 402, the effective energy intensity of the laser light needs to consider the attenuation when transmitting through the stage 402.

FIG. 4B is a view for explaining a method for holding the substrate 403a in the laser device shown in FIG. 4A, except that a light-transmitting stage 402 is used. Since the configuration is the same as that shown in FIG.

Next, the configuration of the optical system 401 shown in FIG. 4A will be described with reference to FIG. FIG. 5 shows an optical system 40.
FIG. 1 is a side view of 1. Laser light having a laser 501 as an oscillation source is split in a vertical direction by a cylindrical lens array 502. The split laser beam is further split laterally by the cylindrical lens 503. In this manner, the laser light is divided into a matrix by the cylindrical lens arrays 502 and 503.

Then, the laser light is once collected by the cylindrical lens 504. At this time, the light passes through the cylindrical lens 505 immediately after the cylindrical lens 504. Up to this point, it is the same as the optical system shown in FIG.

Thereafter, the laser light is applied to the half mirror 506.
, Where the laser beam is the primary laser beam 507
And a secondary laser beam 508. Then, the primary laser light 507 is reflected by the mirrors 509 and 510, passes through the cylindrical lens 511, and reaches the surface of the amorphous semiconductor film 403b.

The secondary laser light 508 split by the half mirror 506 is reflected by mirrors 512, 513, 514.
After being reflected by and passing through the cylindrical lens 515,
The light passes through the substrate 403a and reaches the back surface of the amorphous semiconductor film 403b.

At this time, similarly to the first embodiment, the laser light projected on the irradiation surface of the substrate shows a linear irradiation surface. The homogenization in the width direction (short direction) of the linearly deformed laser light is performed by the cylindrical lens array 502,
This is performed by the cylindrical lens 504 and the cylindrical lens 515. Further, homogenization in the length direction (long direction) of the laser light is performed by the cylindrical lens array 503, the cylindrical lens 505, and the cylindrical lens 511.

In any case, the cylindrical lens 5
A primary laser beam transmitted through the laser beam 11 and irradiated on the surface of the amorphous semiconductor film 403b;
Energy intensity ratio (I ₀ ′ / I) with the secondary laser light that is transmitted through the laser beam and irradiates the back surface of the amorphous semiconductor film 403b.
₀ ) satisfies the relationship of 0 <I ₀ ′ / I ₀ <1 or 1 <I ₀ ′ / I ₀ .

In the present embodiment, the above relational expression is satisfied by using a glass substrate (made of a material having a laser beam transmittance of about 50% used here) as the substrate 403a. Of course, in addition to the substrate, the transmittance of the insulating film (not shown) provided on the substrate 403a or the stage (not shown) on which the substrate 403a is installed and the reflectance of the interface are adjusted to adjust the secondary laser light. The effective energy intensity may be attenuated.

Further, in the optical path of the secondary laser light of the optical system 401, a neutral density filter is provided at an arbitrary position,
It is also possible to attenuate the effective energy intensity of the secondary laser light. Conversely, in the optical path of the primary laser light of the optical system 401, a neutral density filter is provided at an arbitrary position to provide the primary laser light. Can be attenuated.

Further, a plurality of means for attenuating the effective energy intensity of the primary laser light or the secondary laser light described in this embodiment may be used in combination to obtain a desired intensity ratio.

[Example 1] In this example, an example in which an amorphous silicon film is crystallized with the configuration shown in the first embodiment will be described. FIG. 3 is used for the description.

In this embodiment, the substrate 301 is 1.1 m
m-thick quartz substrate, 200-nm-thick silicon nitride oxide film (SiON film) as the insulating film 302, and the amorphous semiconductor film 30
As No. 3, an amorphous silicon film was used. At this time,
The SiON film 302 and the amorphous silicon film 303 were formed using a plasma CVD method.

In this embodiment, first, ₄ SCCM of SiH ₄ and N ₂
O was introduced into the reaction chamber at 400 SCCM,
The SiON film 302 was formed at a temperature of 30 ° C., a reaction pressure of 30 Pa, a discharge power density of 0.41 W / cm ² , and a discharge frequency of 60 MHz.
Next, SiH ₄ was introduced into the reaction chamber at 100 SCCM, the film formation temperature was 300 ° C., the reaction pressure was 45 Pa, and the discharge power density was 0.037.
An amorphous silicon film 303 was formed at W / cm ² and a discharge frequency of 13.56 MHz. Actually, the amorphous silicon film was patterned into an island pattern.

Next, laser crystallization of the amorphous silicon film 303 was performed using an excimer laser apparatus as shown in FIG. At this time, the reflector 30 of FIG.
As No. 4, a tungsten nitride film formed on a silicon substrate was used. The reflector 304 and the quartz substrate 3
A gap of 150 μm was provided between the gap and 01.

In this state, as shown in FIG. 3, excimer laser light (primary laser light 305 and secondary laser light 3
06) with respect to the amorphous silicon film 303 at room temperature,
Irradiation was performed in an air atmosphere. The cross section of the excimer laser light is linear (0.4 mm × 16 mm) by the optical system shown in FIG.
0 mm) and scanned from one end of the substrate to the other.
The scanning speed was set to 1 mm / s, and the energy density (energy intensity imagining _Ia in FIG. 6) was 336 mJ / c.
m ² , pulse width 30 ns, repetition frequency 30H
z and the overlapping ratio were 90%. As a result, one shot could be irradiated with 20 shots of laser light.

When laser crystallization is performed in the configuration of this embodiment, the effective energy intensity (I
₀ ) is 151.2 mJ / cm ² , and the effective energy intensity (I ₀ ′) of the secondary laser beam is 77.3 mJ / cm ²
Met. Therefore, the effective energy intensity ratio (I ₀ ′ /
I ₀ ) was 0.51.

FIG. 10 shows an SEM photograph of the polysilicon film crystallized according to this embodiment. Note that FIG.
Is the state after the Seco etching. This secco etching was performed using 50 cc of hydrofluoric acid solution, 25 cc of water and 1.1 cc.
A room temperature etchant to which 4 g of potassium chromate (divalent) was added was used.

As a result, as shown in FIG. 10, relatively large crystal grains having an average crystal grain size of about 0.5 to 0.6 μm were confirmed near the center of the island pattern. Although crystal grains having a small crystal grain size exist at the end of the island-shaped pattern, the positions formed by changing the laser energy density change. In the case where the polysilicon film formed according to the present embodiment is actually used as an active layer of a TFT, the design should be such that such a portion having a small crystal grain size does not hit the channel formation region.

[Embodiment 2] In this embodiment, an example in which an amorphous silicon film is crystallized with the structure shown in the first embodiment will be described. In the laser crystallization performed in this embodiment, the film formed on the surface of the reflector 304 in the first embodiment is a tungsten film, and the laser energy density is 369 mJ /.
Since it was merely changed to cm ² , the detailed description of the other conditions may be referred to Example 1.

FIG. 11 shows an SEM photograph of the polysilicon film crystallized according to this embodiment. Note that FIG. 11 shows a state after the secco etching as in the first embodiment. Seco
The etching conditions may be referred to Embodiment 1.

When laser crystallization is performed in the configuration of this embodiment, the effective energy intensity (I
₀ ) is 166.1 mJ / cm ² and the effective energy intensity (I ₀ ′) of the secondary laser beam is 88.6 mJ / cm ^2.
Met. Therefore, the effective energy intensity ratio (I ₀ ′ /
I ₀ ) was 0.53.

As a result, as shown in FIG. 11, relatively large crystal grains having an average crystal grain size of about 0.6 to 0.7 μm were confirmed in the entire island pattern. In FIG. 11, FIG.
Small crystal grains at the edge of the island-like pattern as shown in FIG. 0 were not remarkably observed. However, there are some conditions that are conspicuous when the laser energy density is changed, so that the laser energy density needs to be optimized.
Even if there is a portion having a small crystal grain size, there is no problem if the design is made so as not to hit the channel forming region of the TFT as in the first embodiment.

[Embodiment 3] In this embodiment, an example in which an amorphous silicon film is crystallized with the structure shown in the first embodiment will be described. In the laser crystallization performed in this embodiment, the film formed on the surface of the reflector 304 in the first embodiment is a titanium nitride film, and the laser energy density is 384 mJ / c.
Since only m ² has been changed, the detailed description of the other conditions may be referred to the first embodiment.

FIG. 12 shows an SEM photograph of the polysilicon film crystallized according to this embodiment. FIG. 12 shows a state after Seco etching as in the first embodiment. Seco
The etching conditions may be referred to Embodiment 1.

When laser crystallization is performed in the configuration of this embodiment, the effective energy intensity (I
₀ ) is 172.8 mJ / cm ² , and the effective energy intensity (I ₀ ′) of the secondary laser light is 57.6 mJ / cm ²
Met. Therefore, the effective energy intensity ratio (I ₀ ′ /
I ₀ ) was 0.33.

As a result, as shown in FIG. 12, large crystal grains having an average crystal grain size of about 0.8 to 1.0 μm were confirmed in the entire island pattern. These crystal grains have a shape that is longer in the horizontal direction toward the paper surface, and it is considered that this suggests the possibility that crystallization has progressed from the horizontal ends of the island pattern. This tendency is slightly confirmed also in FIG.

Further, there are also conditions that are remarkably observed when the laser energy density is changed, so that the laser energy density needs to be optimized. Even if there is a portion having a small crystal grain size, there is no problem if the design is made so as not to hit the channel forming region of the TFT as in the first embodiment.

[Embodiment 4] In this embodiment, an example of forming a polysilicon film to be an active layer of a TFT by the method of Embodiment 1 or Embodiment 2 will be described. FIG. 13 is used for the description.

First, a 200-nm-thick silicon nitride oxide film (not shown) is formed on a glass substrate, and a 50-nm-thick silicon nitride oxide film is formed thereon.
An amorphous silicon film (not shown) having a thickness of nm is formed. Next, the amorphous silicon film is patterned to form an island pattern 1301 made of an amorphous silicon film.
a and 1301b are formed. (FIG. 13A)

Next, island-like patterns 1301a and 1301b
Is laser-crystallized by the method of the first or second embodiment. The island-shaped patterns 1302a and 1302b made of a polysilicon film obtained by laser crystallization may have regions 1303a and 1303b with small crystal grains at end portions. In addition, island-like patterns 1302a and 1302b
Is also a region containing a lot of crystal defects and lattice distortion.
(FIG. 13 (B))

The dotted lines 1304a and 1304b indicate the island-shaped patterns 13 made of an amorphous silicon film.
01a and 1301b, which means that the island pattern is reduced by about 1 to 15% by laser crystallization. This reduction is considered to occur due to the densification and vaporization of the silicon film, but details are not clear.

The island-like patterns 1301a and 1301a
Traces 1304a and 1304b of 01b remain as steps of a silicon nitride oxide film (not shown) formed below.

Next, the island-like patterns 1302a and 1302b made of a polysilicon film are patterned again to form the active layer 1
305a and 1305b are formed. Note that 1306a, 1
A dotted line indicated by 306b indicates a small region 1303 of a crystal grain.
a, 1303b. Also, the active layers 1305a, 1
By forming 305b, a step is again formed in the silicon nitride oxide film (not shown) formed below.
(FIG. 13 (C))

That is, the island pattern 1 formed in the previous step
A first step formed along the shapes of 301a and 1301b and a second step formed along the shapes of the active layers 1305a and 1305b are formed on the silicon nitride oxide film on the glass substrate. become. As described above, a technique of forming a first shape semiconductor pattern and then laser crystallization and further patterning the first shape semiconductor pattern to form a second shape semiconductor pattern is disclosed in JP-A-8-228006. It is disclosed in the gazette. The use of this technique leaves two steps as described above.

Next, a gate insulating film made of a silicon nitride oxide film having a thickness of 80 nm is formed so as to cover the active layers 1305a and 1305b, and a gate electrode 1307 is formed thereon. The gate electrode 1307 is formed with a stacked structure of a tungsten nitride film and a tungsten film, and has a thickness of 300 nm.
(FIG. 13D)

After the gate electrode 1307 is formed, a step of adding an impurity element for imparting n-type is performed, and the source region 13 is formed.
08a, a drain region 1309a, and an LDD region 1310 are formed. Further, a step of selectively adding an impurity element imparting p-type is performed, so that a source region 1308b and a drain region 1309b are formed. At the same time, channel formation regions 1311a and 1311b (regions to which an impurity element is not added in the active layer) are formed.

Next, after an interlayer insulating film (not shown) made of a silicon oxide film is formed to a thickness of 1 μm, contact holes are opened to form source wirings 1312a, 1312b and drain wiring 1313. These wirings may be formed of a low-resistance conductive film mainly composed of an aluminum film.
(FIG. 13E)

Through the above steps, a CMOS circuit 1316 in which an n-channel TFT 1314 and a p-channel TFT 1315 having a structure as shown in FIG.

This embodiment is an embodiment in which the present invention is implemented when forming an active layer of a TFT, and it is not necessary to limit the present invention to this manufacturing process. The present invention relates to any known TF
It can be used for the manufacturing process of T. However, this excludes the case where a light-shielding film or the like is provided below the active layer, that is, the case where it is impossible to simultaneously perform laser annealing on the front and back surfaces of the amorphous semiconductor film.

In this embodiment, an example in which a CMOS circuit is formed is shown. However, a pixel TFT provided in a pixel portion of an active matrix image display device can be easily manufactured by using a known technique. It is.

Example 5 In this example, the crystalline semiconductor film formed according to the first or second embodiment is
An arrangement in the case of using as an FT active layer will be described.

In the polysilicon film obtained by performing laser annealing under the conditions of Example 2 while changing the laser density to 384 mJ / cm ² , crystal grains are distributed as shown in FIG. It could be confirmed. That is, a first region having a larger average crystal grain size is formed at the end of the island-shaped pattern, and a second region having a smaller average crystal grain size than the first region is formed at the center. The average crystal grain size in the second region is 1/3 or less of the first average crystal grain size. In such a case, as shown in FIG.
It is necessary to design the arrangement of the active layer so as not to use the region.

In FIG. 14A, reference numeral 1401 denotes a schematic diagram of an island pattern formed under the conditions of the present embodiment.
Reference numeral 402 denotes a second area. 1403a, 140
3b is a first region (a region used as an active layer),
Dotted lines 1404a and 1404b correspond to portions where an active layer is formed.

Also, as shown in FIG.
Two types of TFs having active layers arranged as in (A)
It is also possible to form a CMOS circuit by combining T (n-channel TFT and p-channel TFT).

In FIG. 14B, n-channel type TF
T1405 indicates that the gate electrode 1406 and the source region 140
7a, a drain region 1408a, an LDD region 1409, a channel formation region 1410a, a source wiring 1411a, and a drain wiring 1412. Further, the p-channel TFT 1413 includes a gate electrode 1406 and a source region 1.
407b, drain region 1408b, channel formation region 1
410b, source wiring 1411b, and drain wiring 141
It consists of two.

Then, the n-channel TFT 1405 and p-type TFT
By sharing the gate electrode 1406 and the drain wiring 1412 with the channel type TFT 1413, a CMOS circuit can be formed. Of course, other electric circuits or electric elements can be formed.

As described in the fourth embodiment, when the present embodiment is carried out, the step corresponding to the island pattern 1401 and the step corresponding to the active layers 1404a and 1404b become the insulating film (base) of the polysilicon film. Or a substrate). This can be said to be a feature of the present embodiment.

The concept common to FIGS. 14A and 14B is that an active layer is arranged so that at least a channel forming region is formed in a first region having a large crystal grain size. Most preferably, the crystal grain boundaries included in the channel formation region are arranged to be one, preferably zero.
By doing so, it is possible to improve typical electric characteristics such as a field effect mobility and a threshold voltage of the TFT.

[Embodiment 6] In this embodiment, a case will be described in which an active layer is cut out from the crystalline island-like pattern as described with reference to FIG. Specifically, the pixel TFT and the storage capacitor of the pixel portion,
N-channel type TF of a driving circuit provided around the pixel portion
A method for simultaneously manufacturing a T and a p-channel TFT will be described. 15 to 17 are used for the description.

In FIG. 15A, a substrate 701 is made of a glass substrate such as barium borosilicate glass or aluminoborosilicate glass typified by Corning # 7059 glass or # 1737 glass, etc., and polyethylene terephthalate (PET). ), Polyethylene naphthalate (PEN), polyethersulfone (PES), and other plastic substrates having no optical anisotropy. Further, a quartz substrate or a crystallized glass substrate may be used.

Then, a base film 702 such as a silicon oxide film, a silicon nitride film, or a silicon oxynitride film is formed on the surface of the substrate 701 where a TFT is to be formed, in order to prevent impurity diffusion from the substrate 701. SiH ₄ by plasma CVD in this embodiment, NH _3, silicon oxynitride is formed from N ₂ O film 702a of 10 to 200 nm (preferably 50 to 100 nm), is similarly prepared from SiH _4, N ₂ O 50 to 200 n of silicon oxynitride hydrogenated film 702b
m (preferably 100 to 150 nm).

The silicon oxynitride film is a conventional parallel plate type
It is formed by a plasma CVD method. Silicon oxynitride
The film 702a is made of SiH_FourTo 10 SCCM, NH_ThreeTo 100 SCC
M, N _TwoO was introduced into the reaction chamber at 20 SCCM, and the substrate temperature was 3
25 ° C, reaction pressure 40Pa, discharge power density 0.41W / c
m^TwoAnd the discharge frequency is 60 MHz. On the other hand, hydrogen oxynitride
The silicon film 702b is made of SiH_FourTo 5 SCCM, N_TwoO to 12
0 SCCM, H_TwoInto the reaction chamber as 125 SCCM
Temperature 400 ° C, reaction pressure 20Pa, discharge power density 0.41
W / cm^TwoAnd the discharge frequency is 60 MHz. These films are
Change the temperature and change the reaction gas
It can also be done.

The silicon oxynitride film 702a is formed so that its internal stress becomes a tensile stress, considering the substrate as a center. The silicon oxynitride film 702b also has internal stress in the same direction.
The stress should be smaller than the absolute value.

Next, 25 to 80 nm (preferably 30 to 80 nm)
An amorphous semiconductor film 703 having a thickness of 60 nm is formed by a known method such as a plasma CVD method or a sputtering method. For example, an amorphous silicon film is 55 nm thick by a plasma CVD method.
Formed to a thickness of At this time, both the base film 702 and the amorphous semiconductor film 703 can be formed continuously. For example, as described above, the silicon oxynitride film 702a
And hydrogenated silicon oxynitride film 702b by plasma CVD
After continuous film formation by the method, SiH ₄ , N ₂ O, H ₂
If you switch to only SiH ₄ and H ₂ or SiH ₄ from
It can be formed continuously without once exposing it to the atmosphere. As a result, contamination of the surface of the hydrogenated silicon oxynitride film 702b can be prevented, and variation in characteristics of a TFT to be manufactured and fluctuation in threshold voltage can be reduced.

Then, first, island-shaped semiconductor layers 704 to 706 having a first shape are formed from the semiconductor layer 703 having an amorphous structure as shown by a dotted line in FIG. FIG. 18A is a top view of the island-shaped semiconductor layer 704 in this state, and similarly, FIG.
FIG.

In FIG. 18 and FIG. 19, the island-shaped semiconductor layer is rectangular and formed so that one side is 50 μm or less. The shape of the island-shaped semiconductor layer can be arbitrary, preferably the center thereof. Any polygon or circle may be used as long as the minimum distance from the part to the end is 50 μm or less.

Next, such island-like semiconductor layers 704 to 7
The crystallization process is performed on the 06. In the crystallization step, either of the methods described in the first and second embodiments can be used, but in this embodiment, laser annealing is performed on the island-shaped semiconductor layers 704 to 706 by the method of the first embodiment. Thus, island-shaped semiconductor layers 707 to 711 each formed of a crystalline silicon film indicated by a solid line in FIG.

In this embodiment, an example is shown in which one island-shaped semiconductor layer is formed corresponding to two TFTs. However, when the occupied area of the island-shaped semiconductor layer becomes large (one TFT-shaped semiconductor layer).
In the case where the size becomes larger, it is also possible to divide the semiconductor layer into a plurality of island-shaped semiconductor layers and connect a plurality of TFTs connected in series to function as one TFT.

At this time, as the amorphous silicon film is crystallized, the film becomes denser and shrinks by about 1 to 15%. And
A region 712 in which distortion has occurred due to shrinkage is formed at an end of the island-shaped semiconductor layer. Further, a region having a small crystal grain (a region having a small average crystal grain size) is provided at the center of the island-shaped semiconductor layer.
13 are formed. FIG. 18 (B) and FIG. 19 (B)
Shows top views of the island-shaped semiconductor layers in this state, respectively.
In the drawing, regions 704 and 706 indicated by dotted lines indicate the sizes of the originally formed island-shaped semiconductor layers 704 and 706.

When the region 712 in which the strain is accumulated and the region 713 in which the crystal grains are small are included in the channel formation region, a large number of defect levels cause deterioration of TFT characteristics. For example, it is not preferable because the off-state current value increases or the current concentrates in this region to locally generate heat.

Therefore, as shown in FIG. 15C, island-shaped semiconductor layers 715 and 716 of the second shape are formed so that the above-described region is removed. However, as shown in FIG. 19, in the case of the island-shaped semiconductor layer 719 serving as an active layer of the pixel TFT, the above-described region 713 in which crystal grains are small is included in a region serving as a source region, a drain region, and an electrode of a storage capacitor. I have. Such a region where the crystal grains are small is not a problem as long as it is not included in the channel formation region.

A region 712 in which a strain is accumulated and a region 713 in which crystal grains are small exist in a region indicated by a dotted line 714 ′, and an island-shaped semiconductor layer 715 having a second shape is located inside the region. , 716 and 719 are shown. The island-shaped semiconductor layers 715, 71 of the second shape
The shape of 6, 719 may be any shape. FIG. 18C shows an island-shaped semiconductor layer 715 in this state.
716 shows a top view. Similarly, FIG. 19C is a top view of the island-shaped semiconductor layer 719.

Thereafter, the island-shaped semiconductor layers 715 to 719 are formed.
And cover 5 by plasma CVD or sputtering.
A mask layer 720 made of a silicon oxide film having a thickness of 0 to 100 nm is formed. In this state, the island-like semiconductor layer
In order to control the threshold voltage (Vth) of the FT, the impurity element imparting p-type is set to 1 × 10 ^{16 to} 5 × 10 ¹⁷ atoms / cm ^3.
It may be added to the entire surface of the island-shaped semiconductor layer at a concentration of about the same.

As the impurity element imparting p-type to the semiconductor, an element belonging to Group 13 of the periodic table such as boron (B), aluminum (Al), and gallium (Ga) is known.
As the method, an ion implantation method or an ion doping method can be used, but the ion doping method is suitable for treating a large-area substrate. In the ion doping method, diborane (B
Boron (B) is added using ₂ H ₆ ) as a source gas. Such implantation of the impurity element is not always necessary and may be omitted, but it is particularly effective for keeping the threshold voltage of the n-channel TFT within a predetermined range.

In order to form an LDD region of an n-channel TFT of a driver circuit, an impurity element imparting n-type is selectively added to the island-shaped semiconductor layers 716 and 718. Therefore, resist masks 721a to 721e are formed in advance. As an impurity element imparting n-type, phosphorus (P) or arsenic (As) may be used. Here, an ion doping method using phosphine (PH ₃ ) is used to add phosphorus (P).

The formed impurity regions are low-concentration n-type impurity regions 722 and 723, and the phosphorus (P) concentration is 2 ×
The range may be in the range of 10 ^{16 to} 5 × 10 ¹⁹ atoms / cm ³ . In this specification, the impurity regions 722, 7 formed here are
The concentration of the impurity element imparting n-type contained in
^- ). The impurity region 724 is a semiconductor layer for forming a storage capacitor in the pixel portion, and phosphorus (P) is added to this region at the same concentration (FIG. 15D).

Next, the added impurity element is activated.
Perform the process. Activation is performed in a nitrogen atmosphere at 500 to 600
C. for 1 to 4 hours or by laser activation
Can be performed. It is also acceptable to use both together
No. In the case of laser activation method, KrF excimer
Form a linear beam using laser light (wavelength 248 nm)
Oscillation frequency 5-50Hz, energy density 10
0-500mJ / cm ^TwoAs linear beam overlay
The scan is performed with the tip ratio being 80 to 98%, and the island-shaped semiconductor layer is scanned.
The entire surface of the substrate on which is formed is processed. Note that the laser light
There are no restrictions on the firing conditions, and
It should be set. This step is performed while leaving the mask layer 720.
May be performed, or may be performed after removal.

In FIG. 15E, a gate insulating film 72 is formed.
5 is formed of an insulating film containing silicon with a film thickness of 40 to 150 nm by using a plasma CVD method or a sputtering method. For example, a silicon oxynitride film with a thickness of 120 nm is preferably used. A silicon oxynitride film formed by adding O ₂ to SiH ₄ and N ₂ O is a preferable material for this application because the fixed charge density in the film is reduced. Needless to say, the gate insulating film 725 is not limited to such a silicon oxynitride film, and another insulating film containing silicon may be used as a single layer or a stacked structure. In any case, the gate insulating film 725 is formed to have a compressive stress considering the substrate as a center.

Then, as shown in FIG. 15E, a heat-resistant conductive layer for forming a gate electrode is formed over the gate insulating film 725. The heat-resistant conductive layer may be formed as a single layer, or may be formed as a multilayer structure including a plurality of layers such as two layers or three layers as necessary. Using such a heat-resistant conductive material, for example, a conductive layer (A) 726 made of a conductive nitride metal film and a conductive layer (B) 72 made of a metal film
7 is preferably laminated.

The conductive layer (B) 727 is made of tantalum (Ta),
An element selected from titanium (Ti), molybdenum (Mo), and tungsten (W), an alloy containing the above element as a main component, or an alloy film combining the above elements (typically, a Mo-W alloy film, Mo -Ta alloy film), and the conductive layer (A) 726 is formed using tantalum nitride (TaN), tungsten nitride (WN), titanium nitride (TiN), molybdenum nitride (MoN), or the like. Further, as the conductive layer (A) 726, tungsten silicide, titanium silicide, or molybdenum silicide may be used.

It is preferable that the impurity concentration of the conductive layer (B) 727 be reduced in order to reduce the resistance, and it is particularly preferable that the oxygen concentration be 30 ppm or less. For example, tungsten (W) can realize a specific resistance value of 20 μΩcm or less by setting the oxygen concentration to 30 ppm or less.

The conductive layer (A) 726 has a thickness of 10 to 50 nm (preferably 20 to 30 nm), and the conductive layer (B) 727 has a thickness of 200 to 400 nm (preferably 250 to 350 nm).
It is good. When W is used as the gate electrode, an argon (Ar) gas and a nitrogen (N ₂ ) gas are introduced by sputtering using W as a target, and the conductive layer (A) 726 is made of tungsten nitride (WN) of 50 nm. The conductive layer (B) 727 is formed with W to a thickness of 250 nm. As another method, the W film is made of tungsten hexafluoride (WF ₆ )
Can also be formed by a thermal CVD method.

In any case, in order to use the gate electrode as a gate electrode, it is necessary to reduce the resistance.
It is desirable to set it to 0 μΩcm or less. The resistivity of the W film can be reduced by enlarging the crystal grains.
When there are many impurity elements such as oxygen therein, crystallization is inhibited and the resistance is increased. Thus, in the case of using the sputtering method, a W target having a purity of 99.9999% is used, and further, the W film is formed with sufficient care so as not to mix impurities from the gas phase during film formation. 9-20μ
Ωcm can be realized.

On the other hand, a TaN film is formed on the conductive layer (A) 726,
When a Ta film is used for the conductive layer (B) 727, it can be formed by a sputtering method in the same manner. TaN film is T
The target film a is formed using a mixed gas of Ar and nitrogen as a sputtering gas, and the Ta film uses Ar as a sputtering gas. When an appropriate amount of Xe or Kr is added to these sputter gases, the internal stress of the film to be formed can be relaxed and the film can be prevented from peeling. The α-phase Ta film has a resistivity of about 20 μΩcm and can be used as a gate electrode, but the β-phase Ta film has a resistivity of about 180 μΩcm and is not suitable for a gate electrode. TaN film is α
Since it has a crystal structure close to a phase, if a Ta film is formed thereon, an α-phase Ta film can be easily obtained.

Although not shown, the conductive layer (A) 726
It is effective to form a silicon film doped with phosphorus (P) with a thickness of about 2 to 20 nm under the layer. Accordingly, the adhesion of the conductive film formed thereover is improved and oxidation is prevented, and at the same time, a small amount of an alkali metal element contained in the conductive layer (A) 726 or the conductive layer (B) 727 is added to the gate insulating film 725. Spreading can be prevented. In any case, the conductive layer (B) 727 has a resistivity of 10 to 50.
It is preferable to be in the range of μΩcm.

Next, using a photomask, the resist masks 728a to 728a are formed by photolithography.
28f, and the conductive layer (A) 726 and the conductive layer (B) 7
27 are collectively etched to form gate electrodes 729-73.
3 and a capacitor wiring 734 are formed. Gate electrodes 729-7
33 and a capacitor wiring 734 are formed of a conductive layer (A) 729
a to 733a and 729b to 73 made of a conductive layer (B)
3b are integrally formed (FIG. 16A).

In this state, the island-like semiconductor layer 71
FIG. 18D shows the positional relationship between the gate electrodes 5 and 716 and the gate electrodes 729 and 730. Similarly, FIG. 19D shows the relationship between the island-shaped semiconductor layer 719, the gate electrode 733, and the capacitor wiring 734. In FIG. 18D and FIG. 19D, the gate insulating film 725 is omitted.

The method of etching the conductive layer (A) and the conductive layer (B) may be appropriately selected by a practitioner. However, when the conductive layer (A) and the conductive layer (B) are formed of a material containing W as a main component as described above, It is desirable to apply a dry etching method using high-density plasma in order to perform etching at high speed and with high accuracy. Microwave plasma or inductively coupled plasma (Inductively Coup
It is preferable to use an led plasma (ICP) etching apparatus.

For example, W using an ICP etching apparatus
Is an etching method using CF ₄ and Cl ₂ as etching gas.
A seed gas is introduced into the reaction chamber, and the pressure is 0.5 to 1.5 Pa.
(Preferably 1 Pa), and 200 to 10
A high-frequency (13.56 MHz) power of 00 W is applied.
At this time, a high frequency power of 20 W is applied to the stage on which the substrate is placed, and the stage is charged to a negative potential by a self-bias, so that positive ions are accelerated and anisotropic etching can be performed. By using an ICP etching apparatus, even a hard metal film such as W can obtain an etching rate of 2 to 5 nm / sec. In order to perform etching without leaving a residue, overetching is preferably performed by increasing the etching time at a rate of about 10 to 20%. At this time, however, it is necessary to pay attention to the etching selectivity with the base. For example, since the selectivity of the silicon oxynitride film (gate insulating film 725) to the W film is 2.5 to 3, the exposed surface of the silicon oxynitride film is etched by about 20 to 50 nm by such an over-etching process. Being substantially thinner.

Then, the n-channel type TFT of the pixel TFT is used.
In order to form an LDD region, a step of adding an impurity element imparting n-type (n ⁻ doping step) is performed. Using the gate electrodes 729 to 733 as a mask, an impurity element that imparts n-type may be added in a self-aligned manner by an ion doping method. n
The concentration of phosphorus (P) added as an impurity element for imparting a mold is added in a concentration range of 1 × 10 ^{16 to} 5 × 10 ¹⁹ atoms / cm ³ . Thus, low-concentration n-type impurity regions 735 to 739 are formed in the island-shaped semiconductor layer as shown in FIG.

Next, in the n-channel TFT, a high-concentration n-type impurity region functioning as a source region or a drain region is formed (n ⁺ doping step). First, using a photomask, resist masks 740a to 740d are used.
Is formed, and an impurity element imparting n-type is added to form high-concentration n-type impurity regions 741 to 746. Phosphorus (P) is used as an impurity element for imparting n-type and its concentration is 1 ×
The ion doping method using phosphine (PH ₃ ) is performed so as to have a concentration of 10 ²⁰ to 1 × 10 ²¹ atoms / cm ³ (FIG. 16C).

Then, high-concentration p-type impurity regions 748 and 749 serving as a source region and a drain region are formed in the island-shaped semiconductor layers 715 and 717 forming the p-channel TFT. Here, an impurity element imparting p-type is added using the gate electrodes 729 and 731 as a mask, and a high-concentration p-type impurity region is formed in a self-aligned manner. At this time, the island-shaped semiconductor films 716, 718, 7 forming the n-channel TFT
Reference numeral 19 denotes a resist mask 747a using a photomask.
747c is formed and the entire surface is covered.

The high-concentration p-type impurity regions 748 and 749 are formed by an ion doping method using diborane (B ₂ H ₆ ).
The boron (B) concentration in this region is 3 × 10 ^{20 to} 3 × 10 ^21.
atoms / cm ³ (FIG. 16D).

The high-concentration p-type impurity regions 748 and 749
Has phosphorus (P) added in the previous step, and 1 × 10 ²⁰ is added to the high-concentration p-type impurity regions 748a and 749a.
The concentration is about 1 × 10 ²¹ atoms / cm ³ , and the high-concentration p-type impurity regions 748b and 749b have 1 × 10 ^{16 to} 5 × 10 ¹⁹ atoms.
Phosphorus is contained at a concentration of s / cm ^{3, and} the concentration of boron (B) added in this step is set to 1.5 to 3 times the concentration of the contained phosphorus, so that the source region of the p-channel TFT and It can function as a drain region without any problem.

Thereafter, as shown in FIG. 17A, a protective insulating film 750 is formed over the gate electrode and the gate insulating film. The protective insulating film may be formed using a silicon oxide film, a silicon oxynitride film, a silicon nitride film, or a stacked film including a combination thereof. In any case, the protective insulating film 75
0 is formed from an inorganic insulating material. The thickness of the protective insulating film 750 is 100 to 200 nm.

Here, when a silicon oxide film is used, tetraethyl orthosilicate (TEOS) and O ₂ are mixed by plasma CVD, the reaction pressure is 40 Pa, the substrate temperature is 300 to 400 ° C., and the high frequency (13.56MHz) Power density 0.5 ~ 0.8W
/ cm ² and can be formed by discharging. In the case of using a silicon oxynitride film, Si
A silicon oxynitride film formed from H ₄ , N ₂ O, and NH ₃ , or a silicon oxynitride film formed from SiH ₄ , N ₂ O may be used. The manufacturing conditions in this case are a reaction pressure of 20 to 200 Pa, a substrate temperature of 300 to 400 ° C., and a high frequency (60 MHz) power density of 0.1 to 1.0 W / cm ² . Alternatively, a hydrogenated silicon oxynitride film formed from SiH ₄ , N ₂ O, and H ₂ may be used. Similarly, the silicon nitride film is formed by SiH ₄ , N
It can be prepared from H _3. Such a protective insulating film is formed so as to have a compressive stress considering the substrate as a center.

Thereafter, a step of activating the impurity elements imparting n-type or p-type added at the respective concentrations is performed. This step is performed by a furnace annealing method using an electric furnace. In addition, a laser annealing method or a rapid thermal annealing method (RTA method) can be applied. In the furnace annealing method, the oxygen concentration is 40 ppm or less in a nitrogen atmosphere of 1 ppm or less, preferably 0.1 ppm or less.
The heat treatment is preferably performed at 0 to 700 ° C, typically 500 to 600 ° C. In this embodiment, the heat treatment is performed at 550 ° C for 4 hours. In the case where a plastic substrate having a low heat-resistant temperature is used as the substrate 701, a laser annealing method is used (FIG. 17B).

After the activation step, an additional 3 to 100%
1 to 12 at 300 to 450 ° C. in an atmosphere containing hydrogen
A heat treatment is performed for a long time to perform a step of hydrogenating the island-shaped semiconductor layer. This step is a step of terminating dangling bonds in the island-shaped semiconductor layer by thermally excited hydrogen. As another means of hydrogenation, plasma hydrogenation (using hydrogen excited by plasma) may be performed. If the heat resistance of the substrate 701 permits, a heat treatment at 300 to 450 ° C. is performed to form the silicon oxynitride nitride film 702 of the base film 702.
b, The island-shaped semiconductor layer may be hydrogenated by diffusing hydrogen in the silicon oxynitride film of the protective insulating film 750 into hydrogen.

When the activation and hydrogenation steps are completed,
The interlayer insulating film 751 made of an organic insulator is formed to a thickness of 1.0 to 2.0.
It is formed with an average thickness of μm. As an organic insulator,
Polyimide, acrylic, polyamide, polyimide amide, BCB (benzocyclobutene), or the like can be used. For example, when using a polyimide that is thermally polymerized after being applied to a substrate, a 300 mm clean oven is used.
It is formed by firing at ℃. Also, when using acrylic, use a two-liquid type, after mixing the main material and the curing agent,
After coating the entire surface of the substrate using a spinner, preheating is performed at 80 ° C. for 60 seconds on a hot plate, and further, firing is performed at 250 ° C. for 60 minutes in a clean oven.

The surface can be satisfactorily planarized by forming the interlayer insulating film with an organic insulator. Also,
Since organic insulators generally have a low dielectric constant, parasitic capacitance can be reduced. However, since it has a hygroscopic property and its effect as a protective film is weak, it is preferable to use it in combination with a silicon oxide film, a silicon oxynitride film, a silicon nitride film, or the like formed as the protective insulating film 750 as in this embodiment.

Thereafter, a resist mask having a predetermined pattern is formed using a photomask, and a contact hole reaching a source region or a drain region formed in each island-like semiconductor film is formed. The formation of the contact hole is performed by a dry etching method. In this case, the interlayer insulating film 751 made of an organic insulator is first etched by using a mixed gas of CF ₄ , O ₂ , and He as an etching gas, and then the protective insulating film 750 is formed by using the etching gas as CF ₄ and O _2. Etch. Further, by switching the etching gas to CHF ₃ and etching the gate insulating film 725 in order to increase the selectivity with respect to the island-shaped semiconductor layer, a contact hole can be formed favorably.

Then, a conductive metal film is formed by a sputtering method or a vacuum evaporation method, a resist mask is formed by using a photomask, and the source wirings 752 to 752 are formed by etching.
6 and drain wirings 757 to 761 are formed. A drain wiring 762 indicates a drain wiring of an adjacent pixel. Here, the drain wiring 761 functions as a pixel electrode. Although not shown, in this embodiment, this electrode is
A Ti film is formed to a thickness of 50 to 150 nm, a contact is formed with the semiconductor film forming the source or drain region of the island-shaped semiconductor layer, and aluminum (Al) is formed on the Ti film to a thickness of 300 to 400 nm. Then, it is formed as a wiring.

FIG. 18E shows island-like semiconductor layers 715 and 716, gate electrodes 729 and 730, source wirings 752 and 753, and drain wirings 757 and 75 in this state.
8 shows a top view. The source wirings 752 and 753 are connected to the island-shaped semiconductor layers 715 and 716 by contact holes provided in an interlayer insulating film and a protective insulating film (not shown).
And 830 and 833 respectively. The drain wirings 757 and 758 are connected to the island-shaped semiconductor layers 715 and 716 at 831 and 832.

Similarly, in FIG. 19E, the island-shaped semiconductor layer 7 is formed.
19, a gate electrode 733, a capacitor wiring 734, a source wiring 756, and a drain wiring 761 are shown in top views. The source wiring 756 is a contact portion 834 and the drain wiring 7
Numeral 61 denotes contact portions 835 each of which is an island-shaped semiconductor layer 71
9 is connected.

In any case, in the region inside the island-shaped semiconductor layer having the first shape, the region where the strain remains is removed, and the island-shaped semiconductor layer having the second shape is formed. T
Form FT.

When hydrogenation is performed in this state, favorable results can be obtained for improving the characteristics of the TFT. For example, 3 ~
Heat treatment may be performed at 300 to 450 ° C. for 1 to 12 hours in an atmosphere containing 100% hydrogen, or the same effect can be obtained by using a plasma hydrogenation method. In addition, by such heat treatment, the protective insulating film 750 and the base film 702 are formed.
Can be diffused into the island-shaped semiconductor films 715 to 719 for hydrogenation. In any case, the island-like semiconductor layer is desirably defect density and 10 ¹⁶ / cm ³ or less in the 715-719, hydrogen 5 × 10 in order that ¹⁸
It is preferable to provide about 5 × 10 ¹⁹ atoms / cm ³ .
(FIG. 17C).

Thus, the TF of the driving circuit is provided on the same substrate.
A substrate having T and a pixel TFT in a pixel portion can be completed. The drive circuit includes a first p-channel TFT 8
00, a first n-channel TFT 801, a second p-channel TFT 802, a second n-channel TFT 80
3. In the pixel portion, a pixel TFT 804 and a storage capacitor 805 are formed. In this specification, such a substrate is referred to as an active matrix substrate for convenience.

First p-channel TFT 80 of drive circuit
0, the channel-forming region 80 is formed in the island-shaped semiconductor film 715.
6. Source region 807 including high concentration p-type impurity region
a, 807b and drain regions 808a, 808b.

In the first n-channel TFT 801, a channel formation region 809, an LDD region 810 overlapping the gate electrode 730, and a source region 81 are formed on the island-shaped semiconductor film 716.
2. It has a drain region 811. In this LDD region, the length in the channel length direction of the LDD region overlapping with the gate electrode 730 is set to 0.5 to 3.0 μm, preferably 1.0 to 2.0 μm. By setting the length of the LDD region in the n-channel TFT in this way, a high electric field generated near the drain region can be reduced, hot carriers can be prevented from being generated, and deterioration of the TFT can be prevented.

Second p-channel TFT 80 of drive circuit
Similarly, a channel forming region 8 is formed in the island-shaped semiconductor film 717.
13. Source region 814 including high-concentration p-type impurity region
a, 814b and drain regions 815a, 815b.

In the second n-channel TFT 803, a channel forming region 816, LDD regions 817 and 818 partially overlapping the gate electrode 732, a source region 820, and a drain region 819 are formed on the island-shaped semiconductor film 718. . The length of the LDD region overlapping with the gate electrode 732 of this TFT is also 0.5 to 3.0 μm, preferably 1.0 to 2.0 μm.
0 μm. The length in the channel length direction of the LDD region that does not overlap with the gate electrode is 0.5 to 4.0 μm, preferably 1.0 to 2.0 μm.

The pixel TFT 804 includes the island-shaped semiconductor film 71.
9, the channel forming regions 821 and 822 and the LDD region 82
3-825, source or drain regions 826-828
have. The length of the LDD region in the channel length direction is 0.5 to 4.0 μm, preferably 1.5 to 2.5 μm. Further, the storage capacitor 80 is formed from the capacitor wiring 734, the insulating film made of the same material as the gate insulating film, and the semiconductor layer 829 connected to the drain region 828 of the pixel TFT 804.
5 are formed. In FIG. 17C, the pixel TFT 80
4 has a double gate structure, but may have a single gate structure or a multi-gate structure provided with a plurality of gate electrodes.

FIG. 20 is a top view showing almost one pixel of the pixel portion. The AA ′ cross section in the drawing corresponds to the cross-sectional view of the pixel portion in FIG. The gate electrode 733 of the pixel TFT 804 intersects the island-shaped semiconductor layer 719 thereunder via a gate insulating film (not shown). Although not shown, a source region, a drain region, and an LDD region are formed in the island-shaped semiconductor layer. Also, 83
4 is a contact portion between the source wiring 756 and the source region 826, and 835 is a drain wiring 761 and the drain region 82.
8 is a contact portion. The storage capacitor 805 is connected to the pixel T
A semiconductor layer 829 extending from the drain region 828 of the FT 804 is formed in a region overlapping with the capacitor wiring 734 with a gate insulating film interposed therebetween.

The active matrix substrate is completed as described above. In the active matrix substrate manufactured according to this embodiment, a TFT having an appropriate structure is arranged according to the specifications of the pixel portion and the driving circuit. Therefore, it is possible to improve the operation performance and reliability of the electro-optical device using the active matrix substrate.

In the present embodiment, the drain wiring 761 of the pixel TFT 804 is used as it is as a pixel electrode, and has a structure corresponding to a reflection type liquid crystal display device.
However, by forming a pixel electrode made of a transparent conductive film so as to be electrically connected to the drain wiring 761, it is possible to cope with a transmission type liquid crystal display device.

This embodiment is an example of a manufacturing process of a semiconductor device using the present invention, and it is not necessary to limit the materials and numerical ranges shown in this embodiment. Furthermore, the arrangement of the LDD region and the like may be appropriately determined by the practitioner.

[Embodiment 7] In this embodiment, a process of manufacturing an active matrix type liquid crystal display device from an active matrix substrate manufactured according to Embodiment 6 will be described.
First, as shown in FIG. 21A, spacers 901a to 901f made of a resin material are formed on the active matrix substrate in the state of FIG. 17C by patterning. In addition, well-known spherical silica or the like can be sprayed and used as the spacer.

In this embodiment, the spacer 9 made of a resin material is used.
Using NN700 manufactured by JSR as 01a to 901f,
After application by a spinner, a predetermined pattern is formed by exposure and development. In a clean oven etc.
Heat and cure at 50-200 ° C. The spacer produced in this manner can have a different shape depending on the conditions of exposure and development processing.However, it is preferable that the top be a columnar shape and the top be flat, when the substrates on the opposite side are combined. Mechanical strength as a liquid crystal display panel can be secured.

The shape is not particularly limited, such as a conical shape or a pyramid shape. For example, when the shape is a conical shape, specifically, the height H is set to 1.2 to 5 μm and the average radius L1 is set to 5 to 7 μm.
m, the ratio of the average radius L1 to the bottom radius L2 is 1: 1.5.
And At this time, the taper angle of the side surface is set to ± 15 ° or less.

The arrangement of the spacers 901a to 901f may be determined arbitrarily. However, as shown in FIG. 21A, it is preferable that the spacer overlap the contact portion 835 of the drain wiring 761 (pixel electrode) in the pixel portion. It is good to form so that the part may be covered. Since the flatness of the contact portion 835 is impaired and the liquid crystal is not well aligned in this portion, disclination and the like can be prevented by filling the contact portion 835 with a resin for a spacer.

After that, an alignment film 902 is formed. Usually, a polyimide resin is used for the alignment film of the liquid crystal display element. After forming the alignment film, a rubbing treatment is performed so that the liquid crystal molecules are aligned with a certain pretilt angle. It is preferable that the area not rubbed in the rubbing direction from the ends of the spacers 901a to 901f provided in the pixel portion is 2 μm or less. In the rubbing process, the generation of static electricity often poses a problem.
If the spacers 901a to 901e are formed on at least the source wiring and the drain wiring on T, the original role as the spacer in the rubbing step and the effect of protecting the TFT from static electricity can be obtained.

On the opposite substrate 903, a light-shielding film 904, an opposite electrode 905 made of a transparent conductive film, and an alignment film 906 are formed. The light-shielding film 904 is made of Ti, Cr, Al,
It is formed with a thickness of 300 nm. Then, the active matrix substrate on which the pixel portion and the driver circuit are formed and the counter substrate are attached with a sealant 907. A filler 908 is mixed in the sealant 907, and the counter substrate and the active matrix substrate are bonded to each other at a uniform interval by the filler 908 and the spacers 901a to 901f.

After that, a liquid crystal material 909 is injected between the two substrates, and is completely sealed with a sealing agent (not shown). A known liquid crystal material may be used as the liquid crystal material. For example, T
In addition to the N liquid crystal, a thresholdless antiferroelectric mixed liquid crystal exhibiting electro-optical response in which the transmittance changes continuously with an electric field can be used. Some thresholdless antiferroelectric mixed liquid crystals exhibit a V-shaped electro-optical response characteristic. See `` H. Furue et al .; Charakteristics and Drivng Scheme '' for details.
of Polymer-Stabilized Monostable FLCD Exhibiting
Fast Response Time and High Contrast Ratio with Gr
ay-Scale Capability, SID, 1998 '', `` T. Yoshida et a
l.; A Full-Color Thresholdless Antiferroelectric LC
D Exhibiting Wide Viewing Angle with Fast Response
Time, 841, SID97DIGEST, 1997 '', `` S.Inui et al.; Thre
sholdless antiferroelectricity in liquid crystals
and its application to displays, 671-673, J.Mater.Ch
em. 6 (4), 1996 "or U.S. Pat. No. 5,594,569.

Thus, the active matrix type liquid crystal display device shown in FIG. 21B is completed. In FIG. 21, the spacers 901a to 901e are formed separately on at least the source wiring and the drain wiring on the TFT of the driving circuit. However, the spacers 901a to 901e may be formed so as to cover the entire surface of the driving circuit.

FIG. 22 is a top view of the active matrix substrate, showing a positional relationship between the pixel portion and the drive circuit portion, the spacers, and the sealant. Pixel section 1500
, A scanning signal driving circuit 1501 and an image signal driving circuit 1502 are provided as driving circuits. Further, a signal processing circuit 1503 such as a CPU and a memory may be added.

These drive circuits are connected to the connection wiring 15.
11 is connected to the external input / output terminal 1510. In the pixel portion 1500, a gate wiring group 1504 extending from the scanning signal driving circuit 1501 and the image signal driving circuit 15
02 are formed in a matrix by intersecting a source wiring group 1505 extending in a matrix form.
T804 and a storage capacitor 805 are provided.

Spacer 150 provided in pixel portion
Reference numeral 6 corresponds to the spacer 901f shown in FIG. 21, and may be provided for all pixels, or may be provided for every several to several tens of pixels arranged in a matrix. That is, the ratio of the number of spacers to the total number of pixels constituting the pixel portion is preferably 20 to 100%. Further, the spacers 1507 to 1509 provided in the driver circuit portion may be provided so as to cover the entire surface thereof, or may be provided in plural pieces in accordance with the positions of the source and drain wirings of each TFT as shown in FIG. May be.

The sealant 907 is formed outside the pixel portion 1500 and the scanning signal control circuit 1501, the image signal control circuit 1502, and other signal processing circuits 1503 on the substrate 701 and inside the external input / output terminal 1510. I do.

The structure of such an active matrix type liquid crystal display device will be described with reference to the perspective view of FIG. FIG.
In 3, the active matrix substrate is a glass substrate 7
01, a pixel portion 1500, a scanning signal driving circuit 1501, an image signal driving circuit 1502, and another signal processing circuit 1503.

The pixel portion 1500 is provided with a pixel TFT 804 and a storage capacitor 805, and a driving circuit provided around the pixel portion is basically formed of a CMOS circuit. Scanning signal driving circuit 1501 and image signal driving circuit 1502
Are connected to the pixel TFT 804 via a gate wiring 733 and a source wiring 756, respectively. In addition, Flexible Printed Circuit (FPC)
Reference numeral 1513 is connected to the external input terminal 1510 and used to input an image signal or the like. The flexible print circuit 1513 is fixed with a reinforcing resin 1512 to increase the adhesive strength. Then, a connection wiring 1511 is connected to each drive circuit. Although not shown, the opposing substrate 903 is provided with a light-shielding film and a transparent electrode.

A liquid crystal display having such a structure can be formed using the active matrix substrate described in the sixth embodiment. For example, a reflective liquid crystal display device can be obtained by using an active matrix substrate having the structure in FIG. 17C, and transmission can be achieved by using an active matrix substrate using a transparent conductive film as a pixel electrode as described in Embodiment 6. Type liquid crystal display device.

[Eighth Embodiment] In the first embodiment, an example is shown in which an amorphous semiconductor film is crystallized by laser annealing according to the method described in the first or second embodiment. Laser annealing can also be performed on the semiconductor film at the stage where crystallization has progressed.

That is, the laser annealing of the present invention is also effective when the crystalline semiconductor film obtained by crystallizing the amorphous semiconductor film by furnace annealing is further subjected to laser annealing to improve the crystallinity.

Specifically, JP-A-7-161634, JP-A-7-321339 or JP-A-7-13-1
It is possible to use the laser annealing method of Embodiments 1 to 3 in the laser irradiation step (laser annealing step) in the application such as 31034.

After using the present invention in the above publication, a TFT using the formed crystalline semiconductor film can be manufactured. That is, it is possible to combine this embodiment with the fourth to seventh embodiments.

[Embodiment 9] Embodiments 1 to 3 show examples in which the present invention is applied to a liquid crystal display device.
The present invention can be applied to any semiconductor device using a TFT.

Specifically, an active matrix type E
When manufacturing an L (electroluminescence) display device or an active matrix type EC (electrochromic) display device, the present invention can be implemented in a laser annealing step of a semiconductor film. In that case, either of the configurations of the first and second embodiments may be used.

Furthermore, SRAM used for ICs and LSIs
The present invention can be implemented when forming a load transistor of the present invention, and a TFT having a three-dimensional structure can be formed on an IC or LSI.
The present invention is also effective in forming

Since the present invention is an invention of a laser annealing step, a known TFT manufacturing process can be applied to other parts. Therefore, the active matrix EL
When a display device or an active matrix type EC display device is manufactured, the present invention may be applied to a known technique.
Of course, it is also possible to manufacture by referring to the manufacturing steps described with reference to FIGS.

[Embodiment 10] The present invention can be applied to an electronic device (also referred to as electronic equipment) having an electro-optical device such as an active matrix liquid crystal display device or an active matrix EL display device as a display. It is possible. Electronic devices include personal computers, digital cameras, video cameras, personal digital assistants (mobile computers, mobile phones, e-books, etc.),
Navigation system and the like are raised.

FIG. 24A shows a personal computer, which comprises a main body 2001 provided with a microprocessor and a memory, an image input section 2002, a display section 2003, and a keyboard 2004. The present invention relates to a display unit 2003.
And other signal processing circuits.

FIG. 24B shows a video camera, which includes a main body 2101, a display section 2102, an audio input section 2103, operation switches 2104, a battery 2105, and an image receiving section 210.
6. The present invention can be implemented when the display portion 2102 and other signal control circuits are manufactured.

FIG. 24C shows a goggle type display, which comprises a main body 2201, a display section 2202, and an arm section 220.
Consists of three. The present invention can be implemented when manufacturing the display portion 2202 and other signal control circuits (not shown). .

FIG. 24D shows an electronic game machine such as a video game or a video game, and includes a main body 23 on which an electric circuit 2308 such as a CPU, a recording medium 2304, and the like are mounted.
01, controller 2305, display unit 2303, main body 2
The display unit 2302 is incorporated in the display unit 301. The display unit 2303 and the display unit 23 incorporated in the main body 2301
02 may display the same information, display the information of the recording medium 2304 as the main display unit and the latter as the sub display unit, display the operation state of the device, or display the touch sensor. An operation panel can be provided with additional functions. Further, the main body 2301, the controller 2305, and the display unit 2303 may be wired communication for transmitting signals to each other, or may be wireless communication or optical communication by providing the sensor units 2306 and 2307. The present invention can be implemented when the display portions 2302 and 2303 are manufactured. Further, the display portion 2303 can use a conventional CRT.

FIG. 24E shows a player that uses a recording medium (hereinafter, referred to as a recording medium) on which a program is recorded, and includes a main body 2401, a display section 2402, and a speaker section 24.
03, a recording medium 2404, and operation switches 2405. The recording medium is a DVD (Digital Versati
le Disc) and compact disc (CD)
Playback of music programs, video display, video games (or video games), information display via the Internet, and the like can be performed. The present invention can be implemented when the display portion 2402 and other signal control circuits are manufactured.

FIG. 24F shows a digital camera, which comprises a main body 2501, a display section 2502, an eyepiece section 2503, operation switches 2504, and an image receiving section (not shown).
The present invention can be implemented when the display portion 2502 and other signal control circuits are manufactured.

FIG. 25A shows a front type projector, which comprises an optical engine 2601 and a screen 2602. FIG. 25B shows a rear projector, which includes a main body 2701, an optical engine 2702, and a mirror 27.
03, and a screen 2704.

FIG. 25C shows an example of the structure of the optical engine (system including the light source optical system and the display device) 2601 and 2702 in FIGS. 25A and 25B. The optical engines 2601 and 2702 include a light source optical system 2801 and mirrors 2802 and 2804 to 280.
6, dichroic mirror 2803, beam splitter 2807, liquid crystal display 2808, retardation plate 280
9. The projection optical system 2810. Projection optical system 28
Reference numeral 10 includes a plurality of optical lenses.

FIG. 25C shows an example of a three-plate type using three liquid crystal display devices 2808. However, the present invention is not limited to such a type, and a single plate type optical system may be used. Also,
An optical path indicated by an arrow in FIG. 25C may be provided with a suitable optical lens, a film having a polarizing function, a film for adjusting a phase, an IR film, or the like.

FIG. 25D shows an example of the structure of the light source optical system 2801 in FIG. 25C.
In this embodiment, the light source optical system 2801 is
11, light source 2812, lens arrays 2813, 281
4. It is composed of a polarization conversion element 2815 and a condenser lens 2816. Note that the light source optical system shown in FIG. 25D is an example and is not limited to the illustrated configuration.

Although not shown here, the present invention can also be implemented when manufacturing a navigation system or a reading circuit of an image sensor. . Thus, the applicable range of the present invention is extremely wide,
The present invention can be implemented when manufacturing electronic devices in all fields.

[Embodiment 11] In this embodiment, FIG.
A case where the structure shown in (A) and (B) is irradiated with laser light under the conditions shown in Example 1 will be described.

In the structure of FIG. 27A, reference numeral 1601 denotes a 1.1-mm-thick quartz substrate, 1602 denotes a 200-nm-thick silicon nitride oxide film, and 1603 denotes a 55-nm-thick amorphous silicon film. Thus, in the structure of FIG. 27A, normal laser crystallization was performed.

In the structure shown in FIG.
04 is a reflector whose surface (reflection surface) is a tantalum nitride film,
1605 is a 1.1 mm thick quartz substrate, 1606 is 200 mm
A silicon nitride oxide film having a thickness of nm is provided, and 1607 is an amorphous silicon film having a thickness of 55 nm. Thus, FIG.
In the structure of (B), laser crystallization was performed according to the present invention.

The TEM of the resulting polysilicon film
(Transmission Electron Microscopy) Fig. 28
(A) and (B) show. FIG. 28 (A) shows the state shown in FIG.
28B is a TEM photograph of a polysilicon film obtained by crystallizing the amorphous silicon film 1603 with the structure shown in FIG.
Is an amorphous silicon film 16 having a structure shown in FIG.
11 is a TEM photograph of a polysilicon film obtained by crystallizing 07.

By comparing FIG. 28A and FIG. 28B, it can be confirmed that the polysilicon film of FIG. 28B according to the present invention has a clearly larger crystal grain size. As described above, it was also confirmed from the TEM photograph that the present invention can increase the average crystal grain size of the crystalline semiconductor film.

Example 12 According to the experiment performed by the present applicant,
The effective energy intensity ratio (I ₀ ′ / I ₀ ) is 0 <I ₀ ′ / I ₀ <
When 1 or 1 <I ₀ ′ / I ₀ is satisfied,
In particular, there was a condition under which the expansion of the average crystal grain size was remarkable.

In this embodiment, an experiment will be described in which the material shown in FIG. 6 is changed variously for the substrate (all are 1.1 mm thick) or the reflector (strictly, the reflection surface of the reflector). First, Table 1 shows the substrates and reflectors and the effective energy intensity ratios at that time in the samples (A) and (B) used in the experiment.

[0201]

[Table 1]

In Table 1, # 1737 is a product name of a glass substrate manufactured by Corning Incorporated, and AN100 is a product name of a glass substrate manufactured by Asahi Glass Co., Ltd.

Thus, the effective energy intensity ratio is 0.0
Examples 1 to 7 for samples prepared in the range of 7 to 1.0
XeCl excimer laser light was irradiated under the same conditions as in No. 3, and the resulting polysilicon film was observed in an SEM photograph.

As a result, the effective energy intensity ratio was 0.2
At 9, 0.33, 0.53 or 0.67, the average grain size was confirmed to be about 1 μm, and when the effective energy intensity ratio was 1.0, 0.16, 0.11, 0.07. In, it was confirmed that the average crystal grain size was about 0.3 μm. That is, under the condition that the effective energy intensity of the primary laser light differs from that of the secondary laser light by 20% or more, it is considered that the average crystal grain size significantly increases. Therefore, the above results indicate that the effective energy intensity ratio is I
₀ ′ / I ₀ = 0.2 to 0.9 (preferably 0.3 to ₀ .
It is considered that when the condition (7) is satisfied, it indicates that the optimal crystallization condition exists.

[Embodiment 13] In this embodiment, Embodiment 12 will be described.
In this example, an example in which the effective energy intensity ratio is obtained in consideration of the influence of multiple reflection on the reflecting surface of the reflector will be described. In addition,
Samples (A) to (H) used in the experiment are the same as those in Example 12. In the case of the present embodiment, the effective energy intensity (I ₀ ′) of the secondary laser light is I ₀ ′ = I _a T _sub R
_mirror T _sub (1-R _SiON-Si ) / 1-R _SiON-Si T _sub R
It is represented by _mirror T _sub .

Here, T _sub is the transmittance of the substrate, R _mirror is the reflectance on the surface of the reflector, and R _SiON-Si is the reflectance when entering the amorphous silicon film from inside the SiON film. The reflectance when entering the SiON film from the air, the transmittance in the SiON film, the reflectance when entering from the SiON film to the substrate, and the reflectance when entering from the substrate to the SiON film are experimentally determined. Since it turned out to be negligible, it was not included in the calculation.

Table 2 shows data calculated from the above equations. The data shown in Table 2 is obtained by modifying the data in Table 1 in consideration of the influence of multiple reflection.

[0208]

[Table 2]

Example 7 is based on the data shown in Table 2.
The optimum crystallization condition described in the above, that is, the effective energy intensity ratio is I ₀ ′ / I ₀ = 0.2 to 0.9 (preferably 0.3 to 0.9).
The condition satisfying 0.7) was not changed.

[Embodiment 14] In this embodiment, effects of the present invention will be described based on experimental results. In this example, the crystallinity was evaluated in five stages relatively.
In the present specification, the crystalline state is evaluated as distinguished as follows.

Crystal state (0): state in which the film has disappeared by ablation. Crystal state (1): As shown in FIG. 29A, a fine crystal state in which fine crystal grains are observed. Crystal state (2): As shown in FIG. 29B, a crystal state in which crystal grains having an average crystal grain size of about 300 to 450 nm are observed. Crystal state (3): As shown in FIG. 30A, a crystal state in which relatively large crystal grains having an average crystal grain size of about 600 to 800 nm are observed. Crystal state (4): As shown in FIG. 30 (B), a crystal state in which very large crystal grains having a major axis exceeding about 3 μm are observed. In this embodiment, the crystal grains in this state are defined as S
The crystal grains are formed by LG (Super Lateral Growth).

On the basis of the above evaluation, the relationship between the laser crystallization conditions and the crystal state was examined. The data shown in Figure 31, (corresponding to an energy intensity I _a of the laser beam just before reaching the amorphous silicon film) irradiation energy
4 shows the results of comparison between the single irradiation and the dual irradiation for the relationship between and single crystal irradiation. Note that single irradiation refers to a case where only the front surface is irradiated with laser light, and dual irradiation refers to a case where the front and back surfaces are irradiated with laser light.

As is apparent from FIG. 31, a film having a good crystal state can be obtained with dual irradiation at a lower irradiation energy. In other words, in the case of single irradiation, irradiation energy of about 510 mJ / cm ² is required to cause SLG,
In the case of dual irradiation, irradiation energy of about 440 to 460 mJ / cm ² is sufficient. This indicates that a semiconductor film having higher crystallinity can be obtained with lower irradiation energy in the dual irradiation used in the present invention than in the conventional single irradiation.

It has been experimentally found that the higher the irradiation energy is, the higher the effective energy of the primary laser light is, and the surface roughness of the formed crystalline semiconductor film is increased. This suggests that, in obtaining crystals formed by SLG, dual irradiation can reduce damage to the film surface.

Next, in the case of dual irradiation, experimental results are shown in which the reflectance of the reflector is changed to change the effective energy intensity ratio. FIG. 32A shows the relationship between the irradiation energy and the crystal state, and FIG. 32B shows the relationship between the effective incident energy and the crystal state.

As shown in FIG. 32 (A), the higher the reflectivity of the reflector (the higher the effective energy intensity of the secondary laser beam), the better the crystal state was obtained with the same irradiation energy. . This is probably because, for the same irradiation energy, the dual irradiation has a higher effective incident energy. The effective incident energy is
This is the sum of the effective energies incident on the amorphous semiconductor film, and corresponds to the sum of the effective energy intensity of the primary laser light and the secondary effective energy intensity.

Therefore, the relationship between the effective incident energy and the crystal state was examined with the same irradiation energy fixed. Then, as shown in FIG. 32B, the higher the reflectance, the higher the SLG
The effective incident energy required to obtain the crystal (crystal state 4) formed by the above was shifted to a higher energy side. In other words, the lower the reflectivity of the reflector, the easier it is to obtain crystal grains formed by SLG with less effective incident energy, that is, it means that crystallization with less energy loss is possible.

As shown in FIG. 32 (B), as the reflectance of the reflector decreases, the effective incident energy required for the SLG also decreases. However, when the reflectance is zero, SLG does not occur. Has been confirmed. From this, it is considered that the reflectance of the reflector has an optimal value for generating SLG.

[0219]

As shown in the present invention, when the amorphous semiconductor film is subjected to laser crystallization, the laser light is simultaneously irradiated on the front and back surfaces of the amorphous semiconductor film and is irradiated on the back surface side. By making the effective energy intensity different from the effective energy intensity applied to the surface side, it is possible to obtain a crystalline semiconductor film having an average crystal grain size larger than that of the related art.

By obtaining a crystalline semiconductor film having a large crystal grain size, the performance of a TFT or a semiconductor device typified by an active matrix display device formed of a TFT can be significantly improved.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a laser device.

FIG. 2 is a diagram showing a configuration of an optical system of the laser device.

FIG. 3 is a view showing a laser annealing method.

FIG. 4 is a diagram showing a configuration of a laser device.

FIG. 5 is a diagram showing a configuration of an optical system of the laser device.

FIG. 6 is a diagram for explaining a primary laser beam and a secondary laser beam.

FIG. 7 is an SE showing crystal grains in a polysilicon film.
M photo.

FIG. 8 shows SE showing crystal grains of a polysilicon film.
M photo.

FIG. 9 is an SE showing crystal grains in a polysilicon film.
M photo.

FIG. 10 is an SE showing crystal grains of a polysilicon film.
M photo.

FIG. 11 shows an SE showing crystal grains of a polysilicon film.
M photo.

FIG. 12 shows SE showing crystal grains in a polysilicon film.
M photo.

FIG. 13 is a view showing a manufacturing process of a CMOS circuit using a TFT.

FIG. 14 is a diagram showing an example of an arrangement of an active layer.

FIG. 15 illustrates a manufacturing process of an active matrix substrate.

FIG. 16 illustrates a manufacturing process of an active matrix substrate.

FIG. 17 illustrates a manufacturing process of an active matrix substrate.

FIG. 18 illustrates a manufacturing process of an active matrix substrate.

FIG. 19 illustrates a manufacturing process of an active matrix substrate.

FIG. 20 illustrates a pixel structure.

FIG. 21 illustrates a cross-sectional structure of an active matrix liquid crystal display device.

FIG. 22 is a diagram illustrating a top structure of an active matrix liquid crystal display device.

FIG. 23 is a perspective view of an active matrix liquid crystal display device.

FIG. 24 illustrates an example of an electronic device.

FIG. 25 illustrates an example of a projector.

FIG. 26 illustrates a sample structure.

FIG. 27 shows SE showing crystal grains in a polysilicon film.
M photo.

FIG. 28 is a view showing TE showing a state of crystal grains of a polysilicon film.
M photo.

FIG. 29 is an SEM photograph showing a crystal state of a polysilicon film.

FIG. 30 is an SEM photograph showing a crystal state of a polysilicon film.

FIG. 31 is a diagram showing a relationship between irradiation energy and a crystal state.

FIG. 32 is a diagram showing a relationship between irradiation energy or effective incident energy and a crystal state.

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Shunpei Yamazaki 398 Hase, Atsugi-shi, Kanagawa Japan Semiconductor Energy Laboratory Co., Ltd.

Claims (14)

[Claims] An island-shaped semiconductor layer having a first region and a second region sandwiched between the first regions and having a smaller average crystal grain size than the first region is obtained by patterning. A semiconductor device, comprising: an active layer formed on the substrate; and at least a channel forming region of the active layer is formed by the first region. 2. An island-like semiconductor layer having a first region and a second region sandwiched between the first regions and having a smaller average crystal grain size than the first region is obtained by patterning. A semiconductor device, comprising: an active layer formed on the active layer; and at least a channel forming region of the active layer is a combination of an n-channel TFT and a p-channel TFT formed in the first region. 3. The semiconductor device according to claim 1, wherein an average crystal grain size of the second region is equal to or less than ３ of an average crystal grain size of the first region. 4. An electronic device comprising the semiconductor device according to claim 1. 5. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to the front and back surfaces of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer; and when the laser light is applied to the front and back surfaces of the semiconductor film, the effective energy intensity of the laser light applied to the surface of the semiconductor film (I ₀ ) is different from an effective energy intensity (I ₀ ′) of a laser beam applied to the back surface of the semiconductor film. 6. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to the front and back surfaces of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer; and when the laser light is applied to the front and back surfaces of the semiconductor film, the effective energy intensity of the laser light applied to the surface of the semiconductor film (I ₀₎ and the 'to 0 <I ₀ between / I ₀ <1 or 1 <I _0' / I ₀ of the relationship the effective energy intensity of the laser light irradiated on the back surface of the semiconductor film (I ₀₎ ' A method for manufacturing a semiconductor device. 7. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to a front surface and a back surface of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer, wherein the laser light applied to the back surface of the semiconductor film is reflected by a reflector provided on the back surface side of the semiconductor film to form the semiconductor film. A method for manufacturing a semiconductor device, comprising irradiating a back surface of a semiconductor device. 8. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to a front surface and a back surface of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer, wherein the laser light applied to the back surface of the semiconductor film is reflected by a reflector provided on the back surface side of the semiconductor film to form the semiconductor film. And that the effective energy intensity (I ₀ ) of the laser light applied to the back surface of the semiconductor film is different from the effective energy intensity (I ₀ ) of the laser light applied to the back surface of the semiconductor film. A method for manufacturing a semiconductor device. 9. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to the front and back surfaces of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer, wherein the laser light applied to the back surface of the semiconductor film is reflected by a reflector provided on the back surface side of the semiconductor film to form the semiconductor film. is the irradiation on the back, between the effective energy intensity of the laser light irradiated on the surface of the semiconductor film (I ₀₎ and the effective energy intensity of the laser light irradiated on the back surface of the semiconductor film (I ₀ ') 0 <
A method for manufacturing a semiconductor device, wherein a relationship of I ₀ ′ / I ₀ <1 or 1 <I ₀ ′ / I ₀ is satisfied. 10. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to the front and back surfaces of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer, wherein the laser light is converted into a laser light applied to a surface of the semiconductor film by an optical system and a laser light applied to a back surface of the semiconductor film. A method for manufacturing a semiconductor device, which is divided. 11. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to the front and back surfaces of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer, wherein the laser light is converted into a laser light applied to a surface of the semiconductor film by an optical system and a laser light applied to a back surface of the semiconductor film. The effective energy intensity (I ₀ ) of the laser light irradiated on the front surface of the semiconductor film is different from the effective energy intensity (I ₀ ) of the laser light irradiated on the back surface of the semiconductor film. A method for manufacturing a semiconductor device. 12. A step of forming a semiconductor film on a substrate, a step of irradiating a laser beam to a front surface and a back surface of the semiconductor film to crystallize the semiconductor film, and a step of crystallizing the semiconductor film. Forming a TFT having a semiconductor film as an active layer, wherein the laser light is converted into a laser light applied to a surface of the semiconductor film by an optical system and a laser light applied to a back surface of the semiconductor film. is divided, 0 between the effective energy strength of the semiconductor film laser light irradiated on the surface of the (I ₀₎ and the effective energy intensity of the laser light irradiated on the back surface of the semiconductor film (I ₀ ') <
A method for manufacturing a semiconductor device, wherein a relationship of I ₀ ′ / I ₀ <1 or 1 <I ₀ ′ / I ₀ is satisfied. 13. The method for manufacturing a semiconductor device according to claim 5, further comprising the step of linearly deforming the laser light. 14. The method for manufacturing a semiconductor device according to claim 5, wherein the semiconductor film is an amorphous semiconductor film or a microcrystalline semiconductor film.