JP2007059695A - Crystallization method, thin film transistor manufacturing method, crystallized substrate, thin film transistor and display device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a crystallization method that is capable of performing crystallization with a large particle diameter and implements excellent electrical characteristics; and to provide a thin film transistor manufacturing method, a crystallized substrate, a thin film transistor, and a display device. <P>SOLUTION: The crystallization method irradiates a non-single crystal semiconductor film with a laser beam for crystallization. The laser beam has a light intensity distribution with a plurality of peak patterns of inverted triangular shape in cross section, and the non-single crystal semiconductor film is a fluorine-containing non-single crystal semiconductor film. The concentration of fluorine contained in the film of the fluorine-containing non-single crystal semiconductor film is preferably from 10<SP>18</SP>cm<SP>-3</SP>to 10<SP>20</SP>cm<SP>-3</SP>. A gap film is preferably formed on the fluorine-containing non-single crystal semiconductor film. This method allows the crystallized film to have the large particle diameter and the excellent electrical characteristics to be implemented. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a crystallization method, a method for manufacturing a thin film transistor, a substrate to be crystallized, a thin film transistor, and a display device suitable for use in a display device such as a liquid crystal or organic EL.

  A driving circuit of a display device such as a liquid crystal display device is formed on an amorphous semiconductor film formed on a glass substrate. Since information handled by the expansion of the IT market is digitized and speeded up, display devices are also required to have high image quality. As means for satisfying this requirement, for example, a switching transistor for switching each pixel is formed in a crystalline semiconductor, whereby the switching speed is increased and the image quality can be improved.

An excimer laser annealing method (ELA method) is known as a means for crystallizing an amorphous silicon layer formed on a glass substrate. However, the grain size of crystals obtained by this ELA method is about 0.1 μm, and when a thin film transistor (TFT) is formed in this crystallized region, a large number of crystal grains are formed in the channel region of one thin film transistor. The world is included. As a result, the field effect mobility of the thin film transistor is 200 cm ² / Vs, which is significantly inferior to the field effect mobility of the MOS transistor formed in single crystal Si.

  The present inventors have developed an industrialization technique for forming crystal grains large enough to form a channel portion of at least one thin film transistor by irradiating an amorphous silicon layer with laser light. Unlike the conventional transistor in which the crystal grain boundary is formed in the channel region, the TFT is formed in a single crystal grain without adversely affecting the characteristics on the crystal grain boundary, and the TFT characteristics are greatly improved. Functional elements such as a processor, a memory, and a sensor can be formed. As such a crystallization method, the present inventors have proposed a crystallization method described in Non-Patent Document 1, Non-Patent Document 2, and the like.

In the former Non-Patent Document 1, an amorphous silicon film is irradiated with a phase-modulated laser beam having a fluence of 0.8 J / cm 2 through a SiON / SiO ₂ cap layer or a SiO ₂ film (silicon oxide film) cap layer. Thus, a method of crystallizing an amorphous silicon film by laterally growing crystal grains in a direction parallel to the film is described.

Further, in the latter non-patent document 2, the amorphous silicon film is irradiated laterally by irradiating the amorphous silicon film with the intensity-modulated laser light through the SiO ₂ cap layer under substrate heating. A method for crystal growth is described.
W. Yeh and M.M. Matsumura Jpn. Appl. Phys. Vol. 41 (2002) 1909. Proceedings of the 63rd Annual Meeting of the Japan Society of Applied Physics, Autumn 2, 2002, P779, 26a-G-2. Masato Hiramatsu and others

  However, in the method of Non-Patent Document 1, although a crystal grain having a large grain size of 10 μm or more can be obtained, a fine crystal grain having a small grain size is generated in the vicinity of the enlarged crystal grain. Therefore, it is desired to form crystal grains having a large grain size as a whole in a film structure and to form them relatively uniformly (that is, densely).

  Further, in the methods of Non-Patent Documents 1 and 2, there is a problem that the substrate needs to be heated to a high temperature region in order to increase the crystal grain size, and the requirement for low temperature processing cannot be sufficiently satisfied. .

  The present invention has been made in view of the above points, and a crystallization method and a thin film transistor for producing a crystallized film capable of obtaining a crystal having a large particle size even at room temperature and having good electrical characteristics. It is an object to provide a method, a crystallized substrate, a thin film transistor, and a display device.

  In the crystallization method in which the non-single-crystal semiconductor film according to the present invention is crystallized by irradiating the non-single-crystal semiconductor film, the laser light is laser light having a light intensity distribution having a plurality of cross-section inverted triangular peak patterns. The non-single-crystal semiconductor film is a fluorine-containing non-single-crystal semiconductor film.

The fluorine concentration in the fluorine-containing non-single-crystal semiconductor film is preferably 10 ¹⁸ cm ^{−3 to} 10 ²⁰ cm ⁻³ . Furthermore, the oxygen concentration and nitrogen concentration in the fluorine-containing non-single-crystal semiconductor film are preferably 10 ¹⁸ cm ⁻³ or less, and the carbon concentration in the film is preferably 10 ¹⁸ cm ⁻³ or less.

The fluorine-containing non-single crystal semiconductor film is preferably deposited using SiH ₄ and SiF ₄ . The gas flow rate ratio of the SiF _{4 to} the SiH ₄ is preferably in the range of 0.008% to 0.3%.

  The fluorine-containing non-single-crystal semiconductor film is preferably formed by providing a cap film on the laser light incident surface. Further, the cap film is preferably deposited continuously after the deposition of the fluorine-containing non-single crystal semiconductor film. Further, the cap film is preferably at least one of a silicon oxide film, a silicon nitride film, and a silicon oxynitride film.

  The laser light is preferably pulsed laser light that is modulated using a phase shifter so that the light intensity has a repetitive pattern of light intensity distribution that repeats monotonous increase and monotonous decrease. Further, the laser light is preferably light that is made uniform with respect to incident angle and light intensity.

A method of manufacturing a thin film transistor according to the present invention includes a step of depositing a fluorine-containing non-single crystal semiconductor film using SiH ₄ and SiF ₄ on a substrate, and a laser beam incident surface of the fluorine-containing non-single crystal semiconductor film. Forming a cap film at least one layer, and irradiating the fluorine-containing non-single-crystal semiconductor film with the homogenized pulsed laser light through the cap film, A step of forming a crystallized region in the fluorine-containing non-single-crystal semiconductor film after being melted and interrupted by a pulsed laser beam; and a step of forming a thin film transistor in alignment with the crystallized region. To do.

  The substrate to be crystallized according to the present invention includes a fluorine-containing non-single-crystal semiconductor film provided on the substrate, and at least one cap film on the laser light incident surface of the fluorine-containing non-single-crystal semiconductor film. It is characterized by.

  A thin film transistor according to the present invention is characterized in that a channel region, a source region, and a drain region are formed in a crystallization region manufactured by any of the crystallization methods described above.

  A display device according to the present invention is characterized in that a thin film transistor for switching pixels is formed in a crystallization region manufactured by any one of the crystallization methods described above.

  A semiconductor device according to the present invention includes a fluorine-containing non-single-crystal semiconductor film directly or indirectly provided on a substrate, and a crystallization region provided in an island shape on the fluorine-containing non-single-crystal semiconductor film. It is characterized by.

  A thin film transistor according to the present invention includes a fluorine-containing non-single-crystal semiconductor film directly or indirectly provided on a substrate, a crystallization region provided in an island shape on the fluorine-containing non-single-crystal semiconductor film, and the crystallization And at least a channel region provided in the region.

  A display device according to the present invention includes a fluorine-containing non-single-crystal semiconductor film directly or indirectly provided on a substrate, a crystallization region provided in an island shape on the fluorine-containing non-single-crystal semiconductor film, and the crystal A thin film transistor having at least a channel region provided in the conversion region constitutes at least one drive circuit among drive circuits such as a drive circuit for switching the pixel electrode, a vertical line drive circuit, and a horizontal line drive circuit. It is characterized by becoming.

  Hereinafter, a preferred embodiment of the present invention will be described with reference to FIGS.

  First, terms used in this specification are defined as follows. “Lateral growth (lateral growth)” means that the growth of crystal grains proceeds laterally along the film surface in the process of melting and solidifying the crystallization target film. “Input fluence” is a scale representing the energy density of laser light for crystallization, and refers to the energy amount of one pulse per unit area. Specifically, in a light source or an irradiation region (irradiation field). It means the average light intensity of the measured laser beam.

  “Phase shifter” is an example of a phase modulation optical system, which refers to a spatial intensity modulation optical element for modulating the phase of laser light. What is a phase shift mask used in the exposure process of a photolithography process? It is a distinction. In the phase shifter, for example, a desired step is formed at a desired location on a quartz substrate as a transparent body. The step of the phase shifter is formed by a process such as etching so that incident light has a size that produces a phase difference of a predetermined phase angle, for example, 180 degrees.

  A “projection lens” is an optical system for projecting an image created by a phase shifter onto a substrate surface. When the irradiation size is small, a telecentric lens is generally used. By using a telecentric lens, it is possible to prevent the size of the projected image from changing even if the distance between the substrate and the lens changes slightly. Therefore, a one-telecentric lens system in which only the substrate side is parallel or a bi-telecentric lens system in which both the substrate side and the light source side are parallel is used. A “scale” beam is often used and is not shown in this specification, but uses a “kamaboko” type projection lens in which only the short side has a telecentric structure and the long side has no lens effect. This is feasible.

  First, the structure of a projection type crystallization apparatus for explaining the crystallization method will be described with reference to FIG. The crystallization apparatus 1 includes an excimer laser apparatus 2 that emits laser light, a concave lens 3 that is sequentially provided on the optical axis of the apparatus 2, a convex lens 4, a homogenizer 5, a phase shifter 6, and a projection lens 7. The mounting table 9 for mounting the crystallized substrate 8, the XYZθ stage 10, and the controller 11. That is, the crystallization apparatus 1 crystallizes the pulse laser beam 12 on the mounting table 9 by the crystallization optical system including the concave lens 3, the convex lens 4, the homogenizer 5, the phase shifter 6, and the projection lens 7. The substrate 8 is irradiated.

  The excimer laser device 2 that emits laser light outputs laser light with small light absorption in the concave lens 3, the convex lens 4, the homogenizer 5, the phase shifter 6, and the projection lens 7 that are crystallization optical systems. As this excimer laser device 2, for example, an XeCl excimer laser device that oscillates laser light with a wavelength of 308 nm or a KrF excimer laser device that oscillates laser light with a wavelength of 248 nm is optimal.

  The exit optical path of the excimer laser device 2 is provided with an attenuator for adjusting a laser light amount (not shown) to a desired light amount. A homogenizer 5 is provided on the outgoing optical path of the attenuator via a concave lens 3 and a convex lens 4. The homogenizer 5 has a function of leveling the pulse laser beam 12 in the irradiation region. That is, the homogenizer 5 is an optical system for homogenizing (homogenizing) the incident angle and light intensity of the pulse laser beam 12 passing through the homogenizer 5 to the phase shifter 6.

  Further, the homogenized pulse laser beam 12 is caused to generate a phase difference of, for example, 180 degrees by the phase shifter 6 shown in FIG. The phase shifter 6 is made of a transparent body. As shown in FIGS. 2A and 2B, the phase shifter 6 having a plurality of linear steps 6a arranged in parallel causes a phase difference in the pulse laser beam 12 at the steps 6a. Due to this phase difference, the pulse laser beam 12 is phase-modulated, and the pulse laser beam 12 is modulated in light intensity. As a result, as shown in FIG. 2D, a light intensity distribution BP having a repeating pattern that repeats monotonous increase and monotonous decrease is formed in the irradiated portion. In the present embodiment, the interval W between the phase shifters 6 is set to 100 μm, for example.

  2A is a plan view of the phase shifter 6, FIG. 2B is a cross-sectional view of FIG. 2A, and FIG. 2C is a pulse laser beam 12 transmitted through the phase shifter 6. FIG. 2 (d) is a perspective view of FIG. 2 (c). The light intensity distribution BP shown in FIG. 2C is a light intensity distribution of a cross-section inverted triangular peak pattern, for example, a three-dimensional view of the light intensity distribution of a V-shaped groove as shown in FIG. . Such a phase shifter 6 is, for example, an area ratio (duty) modulation type phase shifter 6. In order to obtain a light intensity distribution having a plurality of cross-section inverted triangular peak patterns shown in FIG. 2 (c), the area ratio (duty) modulation pattern has a step 6a as shown in FIG. 2 (a). The surface area is changed. The light intensity distribution of each inverted triangular peak pattern has the same amplitude PH and the same pitch interval PW.

The pulse laser beam 12 whose light intensity is modulated by the phase shifter 6 enters the projection lens 7. The projection lens 7 is provided so as to form an image of the phase shifter 6 on the upper surface of the crystallized substrate 8. The projection lens 7 is an optical system in which the reduction ratio of an image that has been reduced, for example, reduced by the same magnification, is, for example, 1/5. The projection image corresponding to the surface area 100 μm ² of the step 6 a of the phase shifter 6 is 4 μm ² , for example.

The structure of the crystallized substrate 8 has the configuration shown in FIG. That is, the crystallized substrate 8 is formed by sequentially laminating a base protective film 16, a non-single crystal semiconductor film such as an amorphous semiconductor film 17, and a cap insulating film 18 on a substrate 15 made of an insulator or a semiconductor, such as a silicon substrate. It is a structure. The base protective film 16 is a material that prevents impurities from penetrating from the substrate 15 and stores heat generated during the crystallization process of the amorphous semiconductor film 17. The base protective film 16 is a silicon oxide film having a thickness of 1000 nm, for example. The base protective film 16 is not limited to one layer and may be two or more layers. In particular, in order to prevent the permeation of impurities from the substrate 15, for example, a configuration in which a SiNx layer is formed on the substrate 15 and a SiO ₂ layer is further formed is effective. The amorphous semiconductor film 17 is a film to be crystallized, and is made of amorphous silicon having a thickness of, for example, 50 nm to 200 nm. The cap insulating film 18 is made of, for example, a silicon oxide film having a thickness of 20 to 200 nm, for example, 30 nm.

  Next, an embodiment of a method for manufacturing such a crystallized substrate 8 will be described more specifically. An insulating layer as a base protective film 16 is formed on an insulating substrate made of a substrate 15 such as a glass substrate. As the substrate 15, for example, an insulating substrate, in addition to an insulating substrate such as a glass substrate, a quartz substrate, or a plastic substrate, a metal substrate, a silicon substrate, or a ceramic substrate having an insulating film formed on the surface can be applied. . As the glass substrate, it is desirable to use a low alkali glass substrate, for example, represented by Corning # 1737 substrate. As the base protective film 16, a silicon oxide film having a thickness of 50 to 2000 nm, for example, 1000 nm is formed by plasma chemical vapor deposition.

  For example, an amorphous silicon film is formed as an amorphous semiconductor film on the base protective film 16. As a method for forming an amorphous silicon film, an amorphous silicon film having a film thickness of 200 nm is formed by, for example, a plasma chemical vapor deposition method. A cap film 18 is formed on the amorphous silicon film.

As the cap film 18, a silicon oxide film is formed on the amorphous silicon film. This silicon oxide film is a film having a film thickness of, for example, 30 nm, which is formed by, for example, a plasma chemical vapor deposition method of SiH ₄ and N ₂ O and is close to the stoichiometric composition ratio.

  Next, a dehydrogenation process is performed on the thin films 16 to 18 formed on the substrate 15. This dehydrogenation treatment is, for example, a heat treatment at 580 ° C. for 2 hours in a nitrogen atmosphere. In this way, the crystallized substrate 8 is formed.

  The mounting table 9 is mounted on an XYZθ stage 10, can move in the X-axis and Y-axis directions within the horizontal plane, can move in the Z-axis direction orthogonal to the horizontal plane, and rotates around the Z-axis. Is possible. The power supply circuit of the XYZθ stage 10 is connected to the output unit of the controller 11 so that the X-axis drive mechanism, the Y-axis drive mechanism, the Z-axis drive mechanism, and the θ-rotation drive mechanism are automatically controlled. The power supply circuit of the excimer laser device 2 is connected to the output unit of the controller 11 so that the oscillation timing, pulse interval, output magnitude, etc. of the pulse laser beam 12 are controlled.

  Next, the optical system of the homogenizer 5 will be specifically described with reference to FIG. The same parts as those in FIG. 1 are denoted by the same reference numerals, and detailed description thereof will be omitted because it is duplicated. The optical system includes an excimer laser device 2 such as an XeCl excimer laser light source that emits excimer pulse laser light having a wavelength of 308 nm or a KrF excimer laser light source that emits laser light having a wavelength of 248 nm. The laser light emitted from the excimer laser light source is expanded through the optical systems 3 and 4 including a beam expander, and then enters the homogenizer 5. The homogenizer 5 includes a first cylindrical lens 25, a first condenser optical system 26, a second cylindrical lens 27, and a second condenser optical system 28.

  The pulsed laser light incident on the homogenizer 5 is incident on the first cylindrical lens 25. A plurality of light sources are formed on the rear focal plane of the first cylindrical lens 25, and light beams from these plurality of light sources pass through the incident surface of the second cylindrical lens 27 via the first condenser optical system 26. Illuminate in a superimposed manner. As a result, more light sources are formed on the rear focal plane of the second cylindrical lens 27 than on the rear focal plane of the first cylindrical lens 25. The light beam from the light source formed on the rear focal plane of the second cylindrical lens 27 illuminates the phase modulation element 6 (phase shifter) in a superimposed manner via the second condenser optical system 28.

  Here, the first cylindrical lens 25 and the first condenser optical system 26 constitute a first homogenizer, and the incident angle on the phase shifter 6 is made uniform by the first homogenizer.

  The second cylindrical lens 27 and the second condenser optical system 28 constitute a second homogenizer, and the second homogenizer relates to the light intensity (laser fluence) at each position in the plane on the phase shifter 6. Uniformity is achieved. In this way, the illumination system irradiates the phase shifter 6 with light having a substantially uniform light intensity distribution (light intensity distribution).

  The pulsed laser light 12 incident on the phase shifter 6 is made uniform with respect to the incident angle by the first cylindrical lens 25 and the first condenser optical system 26 as a homogenizing optical system (homogenizer), and further, the second cylindrical lens 27. It is desirable that the light intensity is made uniform by the second condenser optical system 28.

  That is, when the pulse laser beam 12 made uniform with respect to the incident angle and the light intensity by the homogenizer 5 passes through the phase shifter 6, the light intensity repeatedly increases and decreases monotonously as shown in FIG. The light intensity distribution BP is obtained. The light intensity distribution BP in FIG. 2 (c) has an inverted triangular cross-section, the maximum peak value and the minimum peak value are protruding, and have no flat portion. Moreover, it has an equal amplitude PH and an equal pitch interval PW. That is, since the phase-modulated homogenized laser beam does not contain higher-order vibration components, theoretically, when this pulsed laser beam 12 irradiates the crystallized substrate 8, the width interval of the steps 6 a-6 a of the phase shifter 6. Large crystal grains having a size corresponding to W can be laterally grown.

  At this time, heat energy is replenished to the film to be crystallized by the heat generation effect due to the light absorption of the amorphous semiconductor film 17 and the heat storage effect of the insulating layer, so that a series of processes of melting → solidification crystallization → crystal grain lateral growth Is promoted and the size of the crystal grains is increased. In addition, in the light intensity distribution BP of FIG. 2C, if the peak portion angle θ becomes sharp, film breakage is likely to occur. Therefore, the light intensity distribution BP is set so that the peak portion angle θ is as gentle as possible. It is desirable.

  In the above embodiment, the projection method for projecting the modulated light by the phase shifter 4 onto the substrate 5 has been described. However, the present invention is not limited to this, and the phase shifter 4 is separated from the substrate 5 at a desired distance. It can also be applied to the proximity method.

  Next, the crystallization method by the crystallization apparatus 1 will be specifically described. The same parts as those in FIGS. 1 to 4 are denoted by the same reference numerals, and detailed description thereof will be omitted because it is duplicated. It is automatically controlled by a program stored in the controller 11 in advance. The controller 11 controls the conveyance of the crystallized substrate 8 to a predetermined position on the mounting table 9 and temporarily fixes the crystallized substrate 8 such as an electrostatic chuck or a vacuum chuck. The controller 11 aligns the temporarily fixed substrate 8 to be crystallized by a predetermined procedure.

The controller 11 performs control for causing the excimer laser device 2 to oscillate. As a result, the excimer laser device 2, for example, the XeCl or KrF excimer laser device 2 oscillates and emits the pulse laser 12. The pulse laser 12 emits a pulse laser beam 50 having a pulse width of, for example, 30 nsec and an irradiation laser fluence of, for example, 940 mJ / cm ² . The pulse laser beam 12 is diverged and converged by the concave lens 3 and the convex lens 4 and enters the homogenizer 5. The homogenizer 5 homogenizes the incident angle and light intensity of the incident pulse laser beam 12.

  The homogenizer 5 causes the uniformed pulse laser beam 12 to enter the phase shifter 6, and the phase shifter 6 emits the pulse laser beam 12 having a light intensity distribution having a plurality of cross-section inverted triangular peak patterns. The pulse laser beam 12 emitted from the excimer laser device 2 is made uniform by the homogenizer 5 after the light intensity and the incident angle are uniformed, and then modulated by the phase shifter 6 into a light intensity distribution having a plurality of cross-section inverted triangular peak patterns, This light intensity distribution is imaged on the crystallized substrate 8 by the projection lens 7. As a result, the amorphous semiconductor film 17 in the imaging portion is melted and crystallized after the laser beam is blocked.

  In this crystallization process, the amorphous semiconductor film 17 is crystallized in the crystallized substrate 8 in the following process. That is, when the pulse laser beam 12 having a light intensity distribution having a plurality of cross-section inverted triangular peak patterns is incident on the crystallized substrate 8, the pulse laser beam 12 of the amorphous semiconductor film 17 of the crystallized substrate 8 is Then, it enters the amorphous semiconductor film 17 and only the irradiated portion is immediately melted in the thickness direction.

  The temperature rise of the amorphous semiconductor film 17 at this time is transmitted to the base protective film 16 and the cap insulating film 18 to be stored. This heat storage effect prevents the irradiated portion of the amorphous semiconductor film 17 from rapidly cooling when the pulsed laser beam is blocked. Therefore, it is possible to form a crystallized region having a large grain size. The amorphous semiconductor film 17 of the non-single crystal semiconductor film immediately melts in the thickness direction, and solidification (crystallization) starts from the reverse peak point where the fluence is minimized when the pulsed laser beam is interrupted, and the lateral Crystal grains grow in the direction, that is, the lateral direction (direction orthogonal to the thickness of the amorphous semiconductor film 17).

  In this crystal growth, since the lateral growth of crystal grains is promoted by the heat storage effect of the base protective film 16 and the cap insulating film 18, the size of the crystal grains after final solidification becomes large, and a wide range of single crystallization occurs in the irradiated portion. Is realized.

  In such a crystallization process, the excimer laser device 2 and the crystallized substrate 8 are relatively moved. For example, when the controller 11 moves the XYZθ stage 10, a predetermined region of the amorphous semiconductor film 17 is moved. It is carried out over the entire surface continuously or intermittently.

  In the lateral growth process using the ELA method, the purpose is to obtain a crystal having a large grain size, that is, a crystal having a long lateral growth length and a good electrical property. The factors affecting the directional growth length were examined.

First, a sample was prepared for examining how the impurities in the silicon film affect the lateral growth length. In order to accurately control the impurities in the sample and the concentration of the impurities, a sample was prepared using an ion implantation method. A sample structure having the same layer structure as that shown in FIG. 3 was prepared. Specifically, in order to prepare samples relating to various impurity concentrations, the sample was prepared through each step (steps 51 to 57) of the sample preparation step 50 shown in FIG. Each step of sample preparation will be described. First, a Si wafer with a 1 μm thermal oxide film is prepared as a base substrate (step 51). Further, an a-Si: H film is deposited as a non-single crystal semiconductor film, which is a crystallized film, with a thickness of, for example, 200 nm by plasma CVD using SiH ₄ as a source gas (step 52).

Further, a cap oxide film of 30 nm was deposited on the a-Si: H film using SiH ₄ and N ₂ O as source gases (step 53). This cap oxide film provides impurities in the a-Si: H film appropriately for ion implantation. After the ion implantation, the cap oxide film is removed (step 55), so that excess in the vicinity of the implantation surface is obtained. The amount of dose and excessive damage can be removed.

In step 54, carbon (C), nitrogen (N), oxygen (O), or fluorine (F) that are considered as impurities by ion implantation, and silicon (Si) as a reference for investigating the influence of ion implantation damage. Each injected sample was made. Thereafter, the cap oxide film for ion implantation was removed by wet etching by etching with dilute hydrofluoric acid or the like (step 55). Again, a cap oxide film of 300 nm was deposited using SiH ₄ and N ₂ O as source gases (step 57). As a reference, a sample was also prepared in which a cap SiO ₂ film was deposited on the amorphous semiconductor film 17 such as an a-Si: H film through the step 56 of creating a sample without opening to the atmosphere.

  The prepared sample was dehydrogenated by annealing at 580 ° C. in a nitrogen atmosphere, and then a single crystal of KrF or XeCl excimer laser light pulse was applied to the crystallized substrate 8 set at room temperature without heating in the air. Crystallized by shot. In other words, the substrate 8 to be crystallized at room temperature was irradiated with one pulse of laser light of pulse laser light. Then, the ultra-small structure (micro structure) of the film crystallized by being irradiated with one pulse of laser light was evaluated by a scanning electron microscope (SEM) after Secco-etching.

  Table 1 shows four types of samples prepared as samples 1 to 4 for examining the effects of opening to the atmosphere and the effects of ion implantation damage.

Samples 1 and 2 are samples prepared for examining the effects of the release to the atmosphere. By comparing the lateral growth lengths of Sample 1 and Sample 2, the effect of atmospheric release on the lateral growth length can be seen. Further, by examining the lateral growth lengths of Sample 2, Sample 3 and Sample 4, the influence of damage caused by ion implantation on the lateral growth length can be understood. The value of the implanted ions corresponds to the peak concentration value. The sample has a constant peak light intensity of 940 mJcm ⁻² in order to eliminate the influence of different light intensity / light intensity distribution, and the distance between the peak position and the valley position of the light intensity distribution, ie, Crystallization was performed with laser light having a light intensity distribution in which the slope width Ws (a value half the pitch interval PW of the light intensity distribution) was fixed at 15 μm. As will be described later, as a result of measuring the lateral growth length using various slope widths Ws for each sample structure, it has been found that the light intensity distribution of 15 μm gives the longest lateral growth length.

FIG. 6 shows the lateral growth length for four references. It can be seen that the lateral growth lengths of Sample 1 and Sample 2 in FIG. 6 are greatly different. That is, by exposing the interface between the cap film SiO ₂ and Si to the atmosphere, the lateral growth length is shortened from about 7 μm to about 4.7 μm. That is, in the laminated structure (sample 1) created without opening to the atmosphere, it is considered that contamination of the Si surface that occurs during exposure to the atmosphere greatly inhibits crystal growth because it has a longer crystal grain growth length. .

  On the other hand, the damage at the time of ion implantation is an important factor for the lateral crystal growth because there is almost no difference in the results of the lateral growth lengths of Sample 2, Sample 3 and Sample 4 shown in FIG. It turns out that it is not. The reason why the influence of ion implantation damage on the lateral growth is negligibly small is that the lateral crystallization melts Si by laser annealing. It seems to have been relaxed. Therefore, when discussing the dependency of the crystal growth length on the impurity concentration in the Si film, the evaluation can be made using a sample intentionally containing impurities by an ion implantation method. This is because ion damage during ion implantation hardly affects crystal growth.

FIG. 7 shows the depth profile of the carbon concentration in the laser annealed Si film. Two samples were used in this experiment. One sample has a laminated structure having an a-Si: H film having a thickness of 200 nm on a Si substrate with a 1 μm thermal oxide film and a capping SiO ₂ film having a thickness of 300 nm on the a-Si: H film. Another sample has a structure in which only a 200 nm a-Si: H film is deposited on the same 1 μm thermal oxide film substrate, that is, there is no capping film.

  It can be seen that carbon as an impurity of the Si film penetrates into a deeper portion of the sample every time the number of laser annealing shots is increased in the case of no capping film. On the other hand, in the case of a cap film with a thickness of 300 nm, it can be seen that the penetration of carbon is very small despite the 20 shots of the laser. It can be seen that the carbon penetration is due to contamination from the atmosphere, and the cap film plays an important role for contamination during the laser annealing process. From the results of FIGS. 6 and 7, it was found that a cap film for reducing contamination on the Si surface is necessary.

FIG. 8 shows SEM photographs of samples having various carbon concentrations. The carbon concentration was calculated by the dose amount of ion implantation. Regarding the irradiation intensity of the laser beam in the case of crystallization, the peak light intensity was fixed at 940 mJcm ⁻² . As the light intensity distribution, the experiment was conducted with the slope width Ws fixed at 15 μm. The lateral growth length becomes shorter as the carbon concentration increases. This is because the lateral growth stops at a position where vertical growth occurs spontaneously due to carbon. The probability of vertical growth is proportional to the probability of nucleation at the Si / SiO ₂ interface. Perhaps short lateral growth, which corresponds to the higher frequency of nucleation due to the carbon which is an impurity in the Si / SiO ₂ interface. This can also be seen from the second lateral growth produced by samples with higher carbon concentrations, for example 3 × 10 ¹⁸ cm ⁻³ or 1 × 10 ¹⁹ cm ⁻³ .

  From the above experiment, it has been found that impurities such as carbon in the amorphous semiconductor film inhibit the crystal growth. Then, the laser annealing without the cap film allows the carbon to penetrate into the semiconductor film. However, when the cap film is provided, the carbon penetration is reduced, so that the cap film is formed on the amorphous semiconductor film. It has been found that the provision of is effective for contamination of carbon and the like, and penetration into the film by laser annealing. Here, the cap film is shown as a cap film of only one layer, but a plurality of cap films may be used, and a silicon oxide film is used as the cap film, but a silicon oxynitride film or a silicon nitride film is used. May be.

  Next, the influence of the light intensity distribution of the irradiated laser beam on the lateral growth length was examined. FIG. 9 shows an expected light intensity distribution profile. The distance between the position of the peak portion and the position of the valley portion of the light intensity distribution of the inverted triangular peak pattern, that is, the slope width Ws is a value half the pitch interval PW of the light intensity distribution. It can be changed arbitrarily while maintaining the strength of the peak and valley portions.

  The lateral growth length for samples with several different carbon concentrations is shown in FIG. 10 as a function of the slope width Ws of the light intensity distribution. When the long slope width Ws, that is, the slope of the light intensity distribution is gentle, it is not always possible to obtain a long lateral growth length. As shown in FIG. 10, the lateral growth length has a relationship with a maximum value with respect to the slope width Ws, and there is an optimum slope width Ws corresponding to the longest lateral growth length.

Furthermore, it can be seen that as the carbon concentration increases, the lateral growth length decreases, and the optimum slope width Ws that gives the longest lateral growth length also moves from 15 μm to 10 μm. From this result, for impurities such as carbon, in the impurity concentration range of 1 × 10 ¹⁷ cm ^{−3 to} 1 × 10 ¹⁸ cm ⁻³ , the slope width Ws is 15 μm and the lateral growth length is the longest value, and 3 × 10 ¹⁸ cm ^{− In} the impurity concentration range of ³ or more, the slope width Ws is 10 μm and the lateral growth length is the longest value. Thereby, the lateral growth length is also affected by the light intensity distribution profile, and the optimum slope width Ws is 10 μm in the impurity concentration range of 10 ¹⁸ cm ⁻³ or more in the impurity concentration such as carbon, for example, It was found that the optimum slope width Ws was 15 μm in the impurity concentration range of 1 × 10 ¹⁷ cm ^{−3 to} 1 × 10 ¹⁸ cm ⁻³ .

First Embodiment FIG. 11 shows the relationship between the concentration of several impurity elements (C, N, O, F) and the lateral growth length. In the peak light intensity distribution, a distribution having a slope width Ws of 15 μm was selected so as to obtain the maximum lateral growth length. When the ion-implanted impurity is fluorine, the lateral growth length is hardly affected by the fluorine concentration. From the results shown in FIG. 11, it can be seen that the dependency between the impurity concentration and the lateral growth length decreases as the atomic number increases.

  This result suggests two possibilities. One possibility is thought to be related to the size of the impurity atoms. As the atomic number increases, the size of the atomic size also increases. The degree of size mismatch between Si atoms and impurity atoms is considered to correspond to the nucleation ratio. Another possibility relates to the number of impurity atom bonds. If the number of dangling bonds (unbonded hands) of impurity atoms is very small during the lateral growth compared to the total number of atoms, the bond flexibility increases. That is, it is considered that a long lateral growth length could be realized because there are few dangling bonds of impurity atoms that hinder the lateral growth of the Si crystal. This degree of flexibility directly corresponds to the nucleation rate.

  From the result of FIG. 11, it was found that even if fluorine is mixed in the a-Si: H film, the crystal growth is not affected. The current process requires a heat treatment (in a 580 ° C. nitrogen atmosphere) for dehydrogenation of a non-single crystal semiconductor film, for example, an a-Si: H film. This is because the a-Si: H film contains several to dozens of% of hydrogen, and when the ELA method is used, the temperature instantaneously rises to 1000 ° C. or higher. End up. Therefore, heat treatment is performed before ELA to reduce the hydrogen concentration in the a-Si: H film to about several percent or less. By this dehydrogenation treatment, excess hydrogen is released to prevent peeling during ELA, and a part of the remaining hydrogen works as termination of dangling bonds of silicon after crystallization, improving electrical characteristics. it can. However, it has also been found that the Si—H bond has a weak bonding force and is unstable. From the result of FIG. 11, it was found that the mixing of fluorine does not affect the crystal growth of the film. Using Si-F bonds, which have stronger bonding strength than Si-H bonds, and having silicon termination (termination), it is possible to obtain good crystal grains that have stable bonds of Si-F and have a longer lateral crystal growth. I was able to.

Second Embodiment Furthermore, the Si-F bond has a bond strength equal to or higher than that of the Si-H bond, and serves as a termination of a dangling bond of silicon instead of Si-H, thereby improving electrical characteristics. I thought I could do it. Therefore, the influence of fluorine content in the film on the electrical characteristics of the semiconductor was investigated.

FIG. 12 shows the dependence of field effect mobility on the fluorine concentration in the film. In this measurement, the gate insulating film thickness is 50 nm, the silicon film thickness is 50 nm, the channel concentration is 2 × 10 ¹⁶ cm ⁻³ , and channel implantation is performed at B. Compared with the normal use (the one with no fluorine at the left end (horizontal axis: no F)), the field effect mobility in the range of fluorine concentration of 1 × 10 ²⁰ cm ⁻³ or less is almost equal or higher Have. Furthermore, as indicated by the error bar, the normally used one has a large variation in field effect mobility, but the one containing fluorine has a small variation and stably provides a good mobility. I understand that. From the above results, in the implantation of fluorine into the film, if the fluorine concentration is in the range of 1 × 10 ²⁰ cm ⁻³ or less, it has better field-effect mobility than the normal use not containing fluorine. I understood.

Further, FIG. 13 shows the dependency of the threshold voltage shift on the fluorine concentration in the film. The TFT structure in measuring the threshold voltage shift is as follows: channel dose: B, channel concentration: 2 × 10 ¹⁶ cm ⁻³ , gate oxide film: 50 nm, gate electrode: Mo—W alloy, interlayer insulating film: SiO ₂ film 300 nm, S / D metal: Ti (barrier metal), Al—Cu—Si alloy 500 nm. The stress conditions are gate voltage: +15 V, source / drain voltage: 0 V, substrate temperature: 90 ° C., stress time: 10,000 seconds.

The threshold voltage shift in the fluorine concentration range of 1 × 10 ¹⁸ cm ⁻³ or higher is lower than that of the normal use (the one not containing fluorine at the left end (horizontal axis: no F)), and is indicated by an error bar. As can be seen, it provides a stable threshold voltage shift with very little variation. From the above results, in the implantation of fluorine into the film, if the fluorine concentration is in the range of 1 × 10 ¹⁸ cm ⁻³ or more, it has a good threshold voltage shift compared with the normal use not containing fluorine. understood.

As described above, from the results of the fluorine concentration dependence of the field effect mobility and the threshold voltage shift, it was possible to obtain good electrical characteristics when the fluorine concentration was 1 × 10 ¹⁸ cm ^{−3 to 1} × 10 ²¹ cm ⁻³ .

Third Embodiment In addition to the influence of fluorine concentration on field effect mobility, the influence of other elements was also examined. The field-effect mobility of a sample into which oxygen, nitrogen, and carbon were further implanted as an implanted element in addition to fluorine and a sample into which only fluorine was implanted as a reference were measured. FIG. 14 shows a sample in which only fluorine is injected (fluorine concentration: 1 × 10 ¹⁷ cm ^{−3 to 1} × 10 ²¹ cm ⁻³ ), a sample without injection (denoted as “no F” on the horizontal axis), and those with and without fluorine injection. The field-effect mobility of a sample in which an oxygen concentration of 1 × 10 ¹⁹ cm ⁻³ , a nitrogen concentration of 1 × 10 ¹⁹ cm ⁻³ , and a carbon concentration of 1 × 10 ¹⁸ cm ⁻³ is respectively injected into this sample is shown. The horizontal axis represents the implanted fluorine concentration, and the vertical axis represents the field effect mobility.

Samples injected with only fluorine show good field-effect mobility, but samples injected with oxygen concentration of 1 × 10 ¹⁹ cm ⁻³ , nitrogen concentration of 1 × 10 ¹⁹ cm ⁻³ and carbon concentration of 1 × 10 ¹⁸ cm ⁻³ each have a field effect. It can be seen that the mobility is lowered and the electrical characteristics are deteriorated. In the same experiment, no decrease in field-effect mobility was observed in samples implanted with an oxygen concentration of 1 × 10 ¹⁸ cm ⁻³ , a nitrogen concentration of 1 × 10 ¹⁸ cm ⁻³ , and a carbon concentration of 1 × 10 ¹⁷ cm ^−3, and the fluorine shown in FIG. Only the same value as the sample injected could be obtained (not shown). Therefore, in a state where the concentration of impurities other than fluorine is reduced, for example, oxygen is an oxygen concentration of 1 × 10 ¹⁸ cm ⁻³ or less, nitrogen is a nitrogen concentration of 1 × 10 ¹⁸ cm ⁻³ or less, and carbon is a carbon concentration of 1 × 10. Good electrical characteristics could be obtained by containing fluorine in a state of ¹⁷ cm ⁻³ or less.

  As a method of actually reducing the oxygen concentration, nitrogen concentration, and carbon concentration in the film, during the deposition of the non-single crystal semiconductor film, suppression of degassing from the chamber wall, pre-coating process before deposition, and oil from the exhaust device By dealing with backflow, the oxygen concentration, nitrogen concentration, and carbon concentration in the deposited film are very low, about 1/100 of the value of a standard film deposited by a standard plasma CVD apparatus. Could be realized.

Fourth Embodiment In the above embodiment, in order to quantitatively and accurately evaluate the dependence of lateral crystal growth and electrical properties on fluorine content in the film, a non-single crystal semiconductor film deposited by plasma CVD is used. A fluorine-containing non-single crystal semiconductor film formed by a fluorine ion implantation method was used. However, in an actual manufacturing process, the fluorine-containing non-single crystal semiconductor film must be deposited using a plasma CVD method. This is due to consistency with the current process, mass productivity, and the like. Therefore, in order to deposit a fluorine-containing non-single crystal semiconductor film by a plasma CVD method, in addition to a commonly used SiH ₄ gas, a fluorine-containing non-single crystal is obtained by a plasma CVD method using SiF ₄ as a raw material gas. An attempt was made to deposit a semiconductor film.

An example of the result is shown in FIG. In order to widely control the fluorine concentration in the film, 1000 ppm SiF ₄ (H ₂ base), 1% SiF ₄ (H ₂ base), and 100% SiF ₄ gas were used as the fluorine-containing gas. Film deposition conditions were as follows: the SiH ₄ flow rate was fixed at 100 sccm, the SiF ₄ gas flow rate was variable, and the pressure was 200 Pa and the substrate temperature was 280 ° C. The horizontal axis represents the actual flow rate ratio (%) of the actually introduced SiF ₄ gas with respect to the SiH ₄ gas flow rate, and the vertical axis represents the measured value of the fluorine concentration in the film deposited at that time. FIG. 15 shows that the fluorine content in the film can be widely controlled by changing the SiF ₄ / SiH ₄ flow rate ratio.

By this SiF ₄ / SiH ₄ gas flow rate ratio control method, a non-single crystal semiconductor film whose fluorine content is controlled is crystallized by the ELA method, and the lateral growth length and electrical characteristics of the crystallized film (electric field) When measuring the effective mobility, threshold voltage shift, etc.), it is possible to obtain results equivalent to the lateral growth length and electrical characteristics of the fluorine-containing non-single crystal semiconductor film prepared by the fluorine ion implantation method. did it.

That is, even in a film deposited using SiH ₄ and SiF ₄ as source gases, the fluorine concentration in the film is in the range of 1 × 10 ¹⁸ cm ^{−3 to 1} × 10 ²⁰ cm ⁻³ as shown in the previous embodiment. Then, good lateral growth length and electrical characteristics could be obtained. As shown in FIG. 15, the SiF ₄ / SiH ₄ flow rate ratio corresponding to the fluorine concentration in the film ranging from 1 × 10 ¹⁸ cm ^{−3 to 1} × 10 ²⁰ cm ⁻³ was 0.008 to 0.3%. Therefore, by setting the SiF ₄ and SiH ₄ gas flow rates within this flow ratio range, the fluorine concentration can be controlled in the range of 1 × 10 ¹⁸ cm ⁻³ to 1 × 10 ²⁰ cm ⁻³ , and a good lateral growth length can be obtained. And electrical characteristics can be obtained.

Fifth Embodiment In the fluorine-containing non-single-crystal semiconductor film deposited in the fourth embodiment, the effect of reducing the oxygen concentration, nitrogen concentration, and carbon concentration in the film shown in the third embodiment is achieved. As a method to realize this, an experiment was conducted assuming that the above-described cap film could be effective in preventing contamination from the outside. Attempts were made to reduce contamination from the air atmosphere by continuously depositing on the fluorine-containing non-single-crystal semiconductor film deposited by the method described in the fourth embodiment without opening to the air. The deposited fluorine-containing non-single crystal semiconductor film was crystallized by ELA method, and the lateral growth length and electrical characteristics (field effect mobility, threshold voltage shift, etc.) of the crystallized film were measured. Better results were obtained compared to a fluorine-containing non-single crystal semiconductor film with no film attached or with a cap film after being opened to the atmosphere.

  Here, continuous deposition refers to 1: after depositing a fluorine-containing non-single-crystal semiconductor film by plasma CVD, and continuously switching to the raw material gas for the cap film while maintaining the plasma without turning off the plasma. When depositing 2: After depositing a fluorine-containing non-single-crystal semiconductor film by plasma CVD method, when turning off the plasma once, switching to the raw material gas for the cap film in the same chamber, igniting the plasma again and depositing, 3: After depositing a fluorine-containing non-single-crystal semiconductor film by plasma CVD method, it is transferred to another chamber for cap film deposition without opening to the atmosphere (for example, transfer between chambers in a clustered multi-chamber apparatus, etc.) Therefore, a cap film may be deposited there.

Above, the cap film is not limited to the silicon oxide film using SiH ₄ and _{N 2} O as source gases (SiO _{2 film),} SiH _4, N 2 O, silicon oxynitride film and NH ₃ as raw material gases (SiON Film) or a silicon nitride film (SiN film) using SiH ₄ or NH ₃ as a source gas may be used. In particular, when the base film is a fluorine-containing non-single-crystal semiconductor film, it is effective to use a silicon oxynitride film or a silicon nitride film having high fluorine resistance and strong adhesion to the fluorine-containing film. Further, the cap film does not need to be a single film of each film, and may be a laminated film having a plurality of layers in which the respective films are laminated. For example, a fluorine-resistant silicon nitride film is deposited on a fluorine-containing non-single-crystal semiconductor film, and a silicon oxide film with low stress and high compatibility with other processes is laminated on the silicon nitride film. A laminated film having a two-layer structure may be used.

Sixth Embodiment Next, a structure of a thin film transistor (TFT) in a crystallized region and a manufacturing method thereof will be described with reference to FIG. A thin film transistor was manufactured using the substrate to be crystallized 8 having a semiconductor film large crystallized by the above crystallization method. The same parts as those in FIG. 3 are denoted by the same reference numerals, and detailed description thereof is omitted.

A base protective film 16 is formed on a substrate 15 made of an insulator or a semiconductor, for example, a low alkali glass substrate. The base protective film 16 is an insulating film containing a silicon oxide film (SiO ₂ film) or a silicon nitride film as a main component, for example, a silicon oxide film having a thickness of 300 nm. The base protective film 16 is preferably formed in close contact with the glass substrate. The base protective film 16 is a film that functions to prevent impurities from diffusing from the substrate 15, for example, a glass substrate, into the fluorine-containing amorphous semiconductor film 17.

  A fluorine-containing non-single crystal semiconductor film such as a fluorine-containing amorphous semiconductor film 17 is formed on the base protective film 16. The fluorine-containing amorphous semiconductor film 17 is, for example, a fluorine-containing amorphous silicon film. The amorphous silicon film is a hydrogenated amorphous silicon film (a-Si: H film) containing fluorine having a thickness of 200 nm formed by, for example, a plasma chemical vapor deposition method.

  A cap insulating film 18 is formed on the fluorine-containing amorphous silicon film to form the crystallized substrate 8. The crystallized substrate 8 has a plurality of inverted triangular peak patterns shown in FIG. 2 (c) as a result of the pulse laser beam homogenized by the homogenizer 5 shown in FIG. A pulse laser beam 12 having a light intensity distribution is formed, and the substrate 8 to be crystallized is irradiated with the laser beam 12 to complete the crystallization process.

  Next, after the crystallization process is completed, the cap insulating film 18 on the fluorine-containing amorphous semiconductor film 17 is removed by etching. Next, in alignment with the crystallized region of the exposed fluorine-containing amorphous semiconductor film 17, a semiconductor circuit, for example, the thin film transistor 35 shown in FIG. 16 is manufactured as follows. First, the fluorine-containing amorphous semiconductor film 17 is patterned using photolithography to define the shape of the active region, and substantially corresponds to each of the channel region 36, the source region 37, and the drain region 38 in a planar view. A predetermined predetermined pattern of Si islands (island-like Si regions) 39 is formed. At this time, at least the channel region 36 of the thin film transistor 35 is formed in the crystallized region. The Si islands 39 are all island-shaped crystallization regions or partial crystallization regions. In the island-like crystallization region, it is optimal that a channel region 36, a source region 37, and a drain region 38 are formed. At least a channel region 36 is formed in the island-shaped crystallization region.

Next, a gate insulating film 40 is formed on the channel region 36, the source region 37 and the drain region 38. The gate insulating film 40 is a material mainly composed of a silicon oxide film or a silicon oxynitride film (SiON film), and is a silicon oxide film having a thickness of 10 to 200 nm or a silicon oxynitride film having a thickness of 30 to 500 nm. In this embodiment, a silicon oxide film 41 having a thickness of 3 nm is formed on the side in contact with silicon, and SiH ₄ (silane gas), NH ₃ (ammonia gas), and N ₂ O (laughing gas) are used as raw materials on the upper side by plasma CVD. The silicon oxynitride film 42 (SiON film) thus formed was formed to a thickness of 50 nm to form a gate insulating film 40. The reason for the two layers is that the silicon oxide film 41 having a low interface state density is used on the silicon interface side, and the SiON film 42 having a high dielectric constant is used on the upper side in order to reduce the leakage current. The present invention is not limited, and only one layer does not depart from the spirit of the present invention. Although not shown in this embodiment, a film obtained by oxidizing the surface of the island-shaped Si island 39 with oxygen plasma or the like can be used as the silicon oxide film 41 below the gate insulating film 40.

  Next, a conductive layer was formed on the gate insulating film 40 in order to form the gate electrode 43. The conductive layer was formed using a material mainly composed of elements such as Ta, Ti, W, Mo, and Al by a known film formation method such as a sputtering method or a vacuum evaporation method. For example, a Mo—W alloy was used. The gate electrode metal layer was patterned using photolithography to form a gate electrode 43 having a predetermined pattern.

Next, the source region 37 and the drain region 38 were formed by implanting impurities using the gate electrode 43 as a mask. For example, when forming a p-channel TFT, a p-type impurity such as boron ion is implanted using an ion implantation method. The boron concentration in this region was set to, for example, 1.5 × 10 ^{20 to} 3 × 10 ²¹ cm ⁻³ . In this way, high-concentration p-type impurity regions constituting the source region 37 and the drain region 38 of the p-channel TFT are formed. At this time, it goes without saying that an n-channel TFT is formed if n-type impurities are implanted.

  Next, a heat treatment step is performed to activate the impurity element implanted by the ion implantation method. This step can be performed by methods such as furnace annealing, laser annealing, and rapid thermal annealing. In the present embodiment, the activation process is performed by furnace annealing. The heat treatment is desirably performed in a temperature range of 300 to 650 ° C. in a nitrogen atmosphere. In this example, heat treatment was performed at 500 ° C. for 4 hours.

  Next, an interlayer insulating film 44 was formed on the gate electrode 43 and the gate insulating film 40. The interlayer insulating film 44 may be formed of a silicon nitride film, a silicon oxide film, a silicon oxynitride film, or a laminated film combining them. The film thickness may be 200 to 600 nm, and is 400 nm in this embodiment.

  Next, a contact hole is opened at a predetermined position in the interlayer insulating film 44. Then, a conductive layer is formed inside the contact hole and on the surface of the interlayer insulating layer 44, and this conductive layer is patterned into a predetermined shape. In this example, the source / drain electrodes were formed as a laminated film having a three-layer structure in which a Ti film was formed with a thickness of 100 nm, an aluminum film containing Ti with a thickness of 300 nm, and a Ti film with a thickness of 150 nm. Thus, the thin film transistor 35 shown in FIG. 16 was formed.

Seventh Embodiment Hereinafter, an example in which a thin film transistor obtained in the above-described embodiment is actually applied to an active matrix liquid crystal display device will be described. FIG. 17 is a diagram illustrating an example of an active matrix display device using a thin film transistor. The display device 70 has a panel structure including a pair of insulating substrates 71 and 72 and an electro-optical material 73 held between the substrates. A liquid crystal material is widely used as the electro-optical material 73. A pixel array unit 74 and a drive circuit unit are integrated on the lower insulating substrate 71. The drive circuit section is divided into a vertical drive circuit 75 and a horizontal drive circuit 76.

  Further, a terminal portion 77 for external connection is formed at the upper end of the peripheral portion of the insulating substrate 71. The terminal portion 77 is connected to the vertical drive circuit 75 and the horizontal drive circuit 76 via the wiring 78. In the pixel array portion 74, row-shaped gate wirings 79 and column-shaped signal wirings 80 are formed. A pixel electrode 81 and a thin film transistor 82 for driving the pixel electrode 81 are formed at the intersection of both wirings. The gate electrode of the thin film transistor 82 is connected to the corresponding gate wiring 79, the drain region is connected to the corresponding pixel electrode 81, and the source region is connected to the corresponding signal wiring 80. The gate wiring 79 is connected to the vertical driving circuit 75, while the signal wiring 80 is connected to the horizontal driving circuit 76.

  The thin film transistor 82 for switching and driving the pixel electrode 81 and the thin film transistor included in the vertical drive circuit 75 and the horizontal drive circuit 76 are manufactured according to the present invention and have higher mobility than the conventional one. Therefore, not only the drive circuit but also a higher-performance processing circuit can be integrated.

  As described above, according to the above embodiment, the non-single crystal semiconductor film is irradiated with the pulsed laser light modulated so that the light intensity has a light intensity distribution with a pattern in which the light intensity monotonously increases and decreases monotonously. The vibration component is reduced, and the appearance of small grain crystals due to the vibration component is effectively suppressed. Since the temperature of the semiconductor film to be crystallized is raised due to the effect of heating the pseudo substrate, the lateral growth length of the crystal grains is greatly promoted, and large-sized crystal grains can be densely formed.

  Further, in a crystallization step, crystallization with a large particle size can be performed even at a low temperature, for example, a room temperature and a temperature in the vicinity thereof (for example, 5 to 50 ° C.). Furthermore, in this embodiment, for example, an XeCl excimer laser device that oscillates 308 nm laser light is used as the excimer laser device 2, but the same good effect can be obtained by a KrF excimer laser device that oscillates laser light having a wavelength of 248 nm. Is confirmed. The crystallization process using a laser beam having a wavelength of 300 nm or less is desirably used in a range where the focal point and the crystallization position are allowed to be displaced due to heat generation of the optical system due to light absorption. Therefore, as another embodiment of the present invention, a KrF excimer laser device that oscillates laser light having a wavelength of 248 nm may be used as the excimer laser device 2.

  In the above embodiment, examples of ion implantation or plasma CVD have been described as means for containing fluorine in the non-single-crystal semiconductor film, but fluorine may be contained by other means such as thermal diffusion.

It is a block diagram which shows the structure of the crystallization apparatus for demonstrating the crystallization method of this invention. It is a figure for demonstrating the structure of the phase shifter of FIG. 1, and the light intensity distribution modulated by this phase shifter, (a) is a top view of a phase shifter, (b) is sectional drawing of (a) figure, (C) is a figure which shows the light intensity distribution of the laser beam which passed the phase shifter and was phase-modulated, (d) is a figure which shows the light intensity distribution of the laser beam of (c) figure three-dimensionally. It is sectional drawing for demonstrating the structure of the to-be-crystallized substrate of FIG. It is a block diagram for demonstrating the optical system of the homogenizer of FIG. It is a flowchart which shows the sample preparation process by an ion implantation method. It is a figure which shows the horizontal direction growth length after ELA of each sample. It is a figure which shows the depth profile of the carbon concentration in Si film by which the depth profile of the carbon concentration was laser-annealed. It is a figure which shows the condition of crystal growth by the SEM photograph after ELA of the sample which has various carbon concentration. It is a figure which shows the profile of the light intensity distribution which has an estimated cross-section inverted triangular peak pattern. It is a figure which shows the dependence of the light intensity distribution slope width Ws of a horizontal direction growth length. It is a figure which shows the relationship between the density | concentration of an impurity element (C, N, O, F) and lateral growth length. It is a figure which shows the fluorine concentration dependence in a film | membrane of field effect mobility. It is a figure which shows the fluorine concentration dependence in a film | membrane of a threshold voltage shift. It is a figure which shows the fluorine concentration dependence in a film | membrane of the field effect mobility of the sample from which the impurity concentration in a film differs. It is a figure which shows the SiF4 / SiH4 flow rate ratio dependence of the fluorine concentration in a film | membrane. It is a cross-sectional schematic diagram for demonstrating the process of forming a thin-film transistor in the crystallization area | region crystallized by the crystallization apparatus of FIG. FIG. 17 is a perspective view of a display device for illustrating a step of forming a display device using the thin film transistor of FIG. 16.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Crystallization apparatus, 2 ... Excimer laser apparatus, 3, 4 ... Optical system, 5 ... Homogenizer, 6 ... Phase shifter, 7 ... Projection lens, 8 ... Substrate to be crystallized, 9 ... Mounting stand, 10 ... XYZtheta stage, DESCRIPTION OF SYMBOLS 11 ... Controller, 12 ... Pulse laser beam, 15 ... Substrate, 16 ... Base protective film, 17 ... Amorphous semiconductor film, 18 ... Cap insulating film, 25 ... First cylindrical lens, 26 ... First condenser optical system 27 ... second cylindrical lens, 28 ... second condenser optical system, 35 ... thin film transistor, 36 ... channel region, 37 ... source region, 38 ... drain region, 39 ... Si island, 40 ... gate insulating film, 41 ... Silicon oxide film, 42 ... SiON film, 43 ... Gate electrode, 44 ... Interlayer insulating film, 70 ... Display device 71, 72 ... Insulating substrate, 73 ... Electro-optical material, 74 ... Pixel array part, 75 ... Vertical drive circuit, 76, Horizontal drive circuit, 77 ... Terminal part, 78 ... Wiring, 79 ... Gate wiring, 80 ... Signal wiring 81 ... Pixel electrode, 82 ... Thin film transistor.

Claims (19)

A crystallization method for crystallizing a non-single crystal semiconductor film by irradiating a laser beam,
The laser beam is a laser beam having a light intensity distribution having a plurality of cross-section inverted triangular peak patterns,
The crystallization method, wherein the non-single crystal semiconductor film is a fluorine-containing non-single crystal semiconductor film. 2. The crystallization method according to claim 1, wherein the fluorine concentration in the fluorine-containing non-single-crystal semiconductor film is 10 ¹⁸ cm ^{−3 to} 10 ²⁰ cm ⁻³ . 3. The crystallization method according to claim 1, wherein an oxygen concentration in the film of the fluorine-containing non-single-crystal semiconductor film is 10 ¹⁸ cm ⁻³ or less. 4. The crystallization method according to claim 1, wherein a nitrogen concentration in the fluorine-containing non-single-crystal semiconductor film is 10 ¹⁸ cm ⁻³ or less. 5. The crystallization method according to claim 1, wherein the fluorine-containing non-single-crystal semiconductor film has a carbon concentration in the film of 10 ¹⁷ cm ⁻³ or less. The crystallization method according to claim 1, wherein the fluorine-containing non-single-crystal semiconductor film is deposited using SiH ₄ and SiF ₄ . The crystallization method according to claim 6, wherein a gas flow rate ratio of the SiF _{4 to} the SiH ₄ is in a range of 0.008% to 0.3%.   The crystallization method according to any one of claims 1 to 7, wherein the fluorine-containing non-single-crystal semiconductor film is provided with a cap film on a laser light incident surface.   The crystallization method according to claim 8, wherein the cap film is continuously deposited after the fluorine-containing non-single-crystal semiconductor film is formed by deposition.   The crystallization method according to claim 8 or 9, wherein the cap film is at least one of a silicon oxide film, a silicon nitride film, and a silicon oxynitride film.   11. The laser beam according to claim 1, wherein the laser beam is a pulsed laser beam modulated by using a phase shifter so that the light intensity has a repetitive pattern light intensity distribution that repeats monotonous increase and monotonous decrease. The crystallization method according to claim 1.   The crystallization method according to any one of claims 1 to 11, wherein the laser light is light that is made uniform with respect to an incident angle and light intensity. Depositing a fluorine-containing non-single crystal semiconductor film using SiH ₄ and SiF ₄ on a substrate;
Forming at least one cap film on the laser light incident surface of the fluorine-containing non-single-crystal semiconductor film;
The fluorine-containing non-single-crystal semiconductor film is irradiated with the homogenized pulsed laser light through the cap film, and the irradiated portion of the fluorine-containing non-single-crystal semiconductor film is melted. Forming a crystallization region in the non-single crystal semiconductor film containing,
And a step of forming a thin film transistor in alignment with the crystallization region.   A substrate to be crystallized, comprising: a fluorine-containing non-single-crystal semiconductor film provided on a substrate; and at least one cap film on a laser light incident surface of the fluorine-containing non-single-crystal semiconductor film.   13. A thin film transistor comprising a channel region, a source region, and a drain region formed in a crystallization region manufactured by the crystallization method according to claim 1.   13. A display device comprising a thin film transistor for switching pixels in a crystallization region manufactured by the crystallization method according to claim 1. A fluorine-containing non-single-crystal semiconductor film directly or indirectly provided on the substrate;
A semiconductor device comprising the fluorine-containing non-single-crystal semiconductor film and a crystallization region provided in an island shape. A fluorine-containing non-single-crystal semiconductor film directly or indirectly provided on the substrate;
A crystallization region provided in an island shape in the fluorine-containing non-single-crystal semiconductor film;
A thin film transistor comprising at least a channel region provided in the crystallization region. A fluorine-containing non-single-crystal semiconductor film directly or indirectly provided on the substrate;
A crystallization region provided in an island shape in the fluorine-containing non-single-crystal semiconductor film;
A thin film transistor having at least a channel region provided in the crystallization region constitutes at least one drive circuit among drive circuits such as a drive circuit for switching and driving a pixel electrode, a vertical line drive circuit, and a horizontal line drive circuit. A display device characterized by comprising:

Images