VEHICLE, DISPLAY DEVICE AND MANUFACTURING METHOD FOR A SEMICONDUCTOR DEVICE

Abstract

To provide a semiconductor device in which a layer to be peeled is attached to a base having a curved surface, and a method of manufacturing the same, and more particularly, a display having a curved surface, and more specifically a light-emitting device having a light emitting element attached to a base with a curved surface. A layer to be peeled, which contains a light emitting element furnished to a substrate using a laminate of a first material layer which is a metallic layer or nitride layer, and a second material layer which is an oxide layer, is transferred onto a film, and then the film and the layer to be peeled are curved, to thereby produce a display having a curved surface.

Claims

1. (canceled)

2. A method for manufacturing a semiconductor device comprising a transistor comprising a semiconductor layer, the method comprising: forming a separation layer over a substrate; forming a peeled layer comprising the transistor over the separation layer; and irradiating the separation layer with a laser light to separate the peeled layer from the substrate, wherein the peeled layer is capable of curving in a first direction, wherein a scan direction of the laser light and a direction of a channel length of the semiconductor layer are facing in a second direction, and wherein the first direction is perpendicular to the second direction.

3. The method according to claim 2, wherein the separation layer is formed of an amorphous silicon or a polysilicon.

4. The method according to claim 2, wherein the separation layer is formed by a sputtering method or a plasma CVD method so as to put a large amount of hydrogen into the separation layer.

5. The method according to claim 2, wherein the substrate is a glass substrate or a quartz substrate.

6. The method according to claim 2, wherein elements included in the peeled layer are formed at 410 C. or below.

7. The method according to claim 2, wherein a shape of the laser light is linear or rectangular.

8. The method according to claim 2, wherein the laser light is irradiated to a back surface of the separation layer through the substrate.

9. A method for manufacturing a semiconductor device comprising a transistor comprising a semiconductor layer, the method comprising: forming a separation layer over a substrate; forming a peeled layer comprising the transistor over the separation layer; irradiating the separation layer with a laser light to separate the peeled layer from the substrate; and transferring the peeled layer to an object, wherein the peeled layer is capable of curving in a first direction, wherein a scan direction of the laser light and a direction of a channel length of the semiconductor layer are facing in a second direction, and wherein the first direction is perpendicular to the second direction.

10. The method according to claim 9, wherein the separation layer is formed of an amorphous silicon or a polysilicon.

11. The method according to claim 9, wherein the separation layer is formed by a sputtering method or a plasma CVD method so as to put a large amount of hydrogen into the separation layer.

12. The method according to claim 9, wherein the substrate is a glass substrate or a quartz substrate.

13. The method according to claim 9, wherein elements included in the peeled layer are formed at 410 C. or below.

14. The method according to claim 9, wherein a shape of the laser light is linear or rectangular.

15. The method according to claim 9, wherein the laser light is irradiated to a back surface of the separation layer through the substrate.

16. The method according to claim 9, wherein the object has a curved surface.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In the accompanying drawings:

[0063] FIGS. 1A to 1C are diagrams of steps illustrating an embodiment mode of the present invention;

[0064] FIG. 2 is a diagram indicating directional orientations in an embodiment mode of the present invention;

[0065] FIG. 3 is a layout diagram showing a configuration of a laser radiation device, according to Embodiment 1 of the present invention;

[0066] FIG. 4 is another layout diagram showing a configuration of a laser radiation device, according to Embodiment 1 of the present invention;

[0067] FIG. 5 is a diagram for explaining a construction of a substrate provided with a TFT, and the relationship between an arrangement of a semiconductor area constituting the TFT and a scan direction of a laser beam;

[0068] FIG. 6A to 6D are diagrams for explaining the laser beam scan direction along a semiconductor film, and steps of manufacturing a top gate-type TFT;

[0069] FIG. 7A to 7D are diagrams for explaining the laser beam scan direction along the semiconductor film, and steps of manufacturing a bottom gate-type TFT;

[0070] FIGS. 8A to 8G are diagrams of steps illustrating Embodiment 3 of the present invention;

[0071] FIG. 9 is a diagram illustrating V-I characteristics of an n channel-type TFT after peeling;

[0072] FIG. 10 is a diagram illustrating V-I characteristics of a p channel-type TFT after peeling;

[0073] FIGS. 11A to 11F are diagrams of steps illustrating Embodiment 4 of the present invention;

[0074] FIGS. 12A and 12B are external views of a curved semiconductor device having a light emitting element in which a layer containing organic material serves as the light emitting layer, according to Embodiment 4 of the present invention;

[0075] FIG. 13 is a diagram showing the vicinity in front of a driver's seat in a car, according to Embodiment 5 of the present invention;

[0076] FIG. 14 is a diagram showing the rear portion behind the car, according to Embodiment 5 of the present invention; and

[0077] FIGS. 15A to 15D are diagrams of steps illustrating Embodiment 6 of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0078] An embodiment mode of the present invention will be described below.

[0079] Hereinafter, a typical manufacturing order according to the present invention will be briefly described using FIGS. 1A to 1C and 2.

[0080] In FIG. 1A, reference numeral 10 denotes a substrate, 11a denotes a layer to be peeled, 12 denotes a pixel portion provided to the layer to be peeled, 13a denotes a semiconductor layer provided in the pixel portion, 13b denotes a channel length direction of the semiconductor layer 13a, 14a denotes a laser light irradiation area, and 14b denotes a laser light irradiation direction.

[0081] FIG. 1A shows a manufacturing step indicating the course of a completion of the layer to be peeled and is a schematic view indicating processing for irradiating laser light to the semiconductor layer. Laser crystallization and laser annealing can be conducted by the laser light irradiation processing. An oscillation mode may be either continuous oscillation or pulse oscillation. In order to continuously produce crystal growth with a molten state of the semiconductor film, it is desirable that a continuous oscillation mode is selected.

[0082] In FIG. 1A, all channel length directions of a large number of semiconductor layers included in the layer to be peeled are aligned with the same direction. In addition, assume that the laser light irradiation direction, that is, a scanning direction is the same direction as the channel length directions. Thus, when the crystal growth direction is aligned with the channel length direction, the field effect mobility can be substantially increased. Note that the example in which linear laser light irradiation is conducted is shown in FIG. 1A. However, the present invention is not particularly limited to this. In addition, here, the laser light irradiation is conducted after patterning the semiconductor layer. The laser light irradiation may be conducted before the patterning.

[0083] Next, various elements (such as a thin film diode, a photoelectric conversion element which is made of silicon and has a PIN junction, and a silicon resistor element) represented by a TFT are produced by forming electrodes, wirings, an insulating film, and the like to complete the layer to be peeled 11b, and then the layer to be peeled 11b is peeled from the substrate 10.

[0084] Note that the peeling method is not particularly limited. Here, a peeling method utilizing film stresses of a metallic layer or a nitride layer and an oxide layer is used as a peeling method which is not limited by a heat treatment temperature and a kind of substrate. First, before the state shown in FIG. 1A is obtained, a nitride layer or a metallic layer (not shown) is formed on the substrate 10. A typical example of the nitride layer or the metallic layer includes a single layer made of an element selected from the group consisting of Ti, W, Al, Ta, Mo, Cu, Cr, Nd, Fe, Ni, Co, Ru, Rh, Pd, Os, Ir, and Pt, or an alloy material or a compound material which contains mainly the element, or a laminate of those. In addition, a single layer made of nitride containing the element, for example, titanium nitride, tungsten nitride, tantalum nitride, or molybdenum nitride, or a laminate of those may be used. Next, an oxide layer (not shown) is formed on the nitride layer or the metallic layer. For a typical example of the oxide layer, a silicon oxide material, a silicon oxynitride material, or a metallic oxide material may be used. Note that the oxide layer may be formed by any film formation method such as a sputtering method, a plasma CVD method, or an applying method. It is important that film stress of the oxide layer is different from that of the nitride layer or the metallic layer. The respective film thicknesses are preferably set as appropriate in a range of 1 nm to 100 nm, thereby adjusting respective film stresses. In addition, an insulating layer or a metallic layer may be provided between the substrate and the nitride layer or the metallic layer to improve the contact property to the substrate 10. Next, a semiconductor layer is preferably formed on the oxide layer to obtain the layer to be peeled 11a. Note that, according to the above peeling method, even if the film stress of the oxide layer is different from that of the nitride layer or the metallic layer, film peeling or the like is not caused by heat treatment in a manufacturing step for the layer to be peeled. In addition, according to the above peeling method, the film stress of the oxide layer is different from that of the nitride layer or the metallic layer. Thus, peeling can be produced by relatively small force. In addition, the example in which the layer to be peeled 11b having a sufficient mechanical strength is assumed is indicated here. When the mechanical strength of the layer to be peeled 11b is insufficient, it is preferable that peeling is conducted after a support member (not shown) for fixing the layer to be peeled 11b is bonded thereto. Note that, when the layer to be peeled 11b is peeled, it is important to prevent a bend of the layer to be peeled 11b so that a crack is not caused in the layer to be peeled.

[0085] Thus, the layer to be peeled 11b which is formed on the oxide layer can be separated from the substrate 10. A state obtained after peeling is shown in FIG. 1B. In a stage shown in FIG. 1B, not only the semiconductor layer but also electrodes, wirings, and the like are formed. However, for simplification, they are not shown here.

[0086] The peeled layer 11c can be bent. A state obtained after the bending is shown in FIG. 1C. The peeled layer 11c is bent in a bending direction 19. It is needless to say that the peeled layer can be bonded to a transfer body (not shown) having a curved surface.

[0087] In FIG. 1C, reference numeral 15 denotes a driver circuit (X-direction), 16a denotes a semiconductor layer provided in the driver circuit (X-direction), 16b denotes a channel length direction of the semiconductor layer 16a, 17 denotes a driver circuit (Y-direction), 18a denotes a semiconductor layer provided in the driver circuit (Y-direction), and 18b denotes a channel length direction of the semiconductor layer 18a.

[0088] Thus, the maximum feature of the present invention is that the laser light irradiation direction 14b and the channel length directions 13b, 16b, and 18b of all the semiconductor layers provided in the layer to be peeled are set to be the same direction, and these directions and a bending direction 19 are set to be orthogonal to each other.

[0089] Note that, in order to further clear a correlation among these directions, the case where a TFT is noted is shown in FIG. 2. FIG. 2 briefly shows a TFT having a semiconductor layer 20, a gate electrode 21, and electrodes (source electrode and drain electrode) 22 and 23. Note that the TFT can be produced as follows by using a known technique. First, a semiconductor film having an amorphous structure (made of amorphous silicon or the like) is crystallized by a known crystallization technique to produce a semiconductor film having a crystalline structure (made of polysilicon or the like), and then patterned into a predetermined shape to form the semiconductor layer 20. Next, the semiconductor layer 20 is covered with a gate insulating film (not shown) and then the gate electrode 21 is formed so as to partially overlap the semiconductor layer 20 through the insulating film interposed therebetween. After that, an impurity element for imparting an n-type or p-type conductivity is added to a portion of the semiconductor layer to produce a source region and a drain region, an interlayer insulating film (not shown) covering the gate electrode is formed, and the electrodes (source electrode and drain electrode) 22 and 23 electrically connected with the source region and the drain region are formed on the interlayer insulating film.

[0090] In the present invention, laser light whose scanning direction is a scanning direction 25 shown in FIG. 2 is used for manufacturing the TFT. In addition, a portion of the semiconductor layer 20 which is overlapped with the gate electrode 21 through the gate insulating film interposed therebetween serves as a channel. Thus, a channel length direction becomes a channel length direction 24 shown in FIG. 2. The scanning direction 25 of laser light becomes the same direction as the channel length direction 24. In addition, a channel width direction which is a direction orthogonal to the channel length direction 24 is the same direction as a bending direction. The bending direction becomes a bending direction 26 shown in FIG. 2. Note that the example of a top gate TFT is shown in FIG. 2. The present invention can be applied to, for example, a bottom gate (inverse staggered) TFT or a staggered TFT independent on the TFT structure.

[0091] Although a TFT in which a semiconductor layer containing silicon serves as an active layer is shown here, but it is not limited to such a TFT particularly. It is also possible to manufacture an organic TFT in which an active layer is made of an organic material. Materials for the active layer of an organic TFT can be a material having considerable amount of carbon when it is combined with other materials, or a material containing an isotope of carbon element except diamond. As representative materials of the active layer of an organic TFT, C.sub.60, C.sub.70, thiophene polymer, thiophene substitution derivatives, poly (thienylene vinylene) and the like can be exemplified.

[0092] Also, the present invention can be applied to various manufacturing methods of semiconductor device. Particularly, when a plastic substrate is used for the transfer body and the support member, weight reduction can be realized.

[0093] When a liquid crystal display device is manufactured, it is preferable that the support member is used as a counter substrate and bonded to the layer to be peeled using a sealing member as a bonding layer. In this case, the element provided to the layer to be peeled has a pixel electrode. A liquid crystal material is filled into a space between the pixel electrode and the counter substrate. In addition, an order for manufacturing the liquid crystal display device is not particularly limited. For example, the counter substrate as the support member is bonded to the layer to be peeled which is provided to the substrate, a liquid crystal material is injected therebetween, and then the substrate is peeled and the plastic substrate as the transfer body is bonded to the layer to be peeled. Alternatively, after the pixel electrode is formed, the substrate is peeled, the plastic substrate as a first transfer body is boned to the layer to be peeled, and then the counter substrate as a second transfer body is bonded thereto.

[0094] Also, when a light emitting device represented by a device having a light emitting element in which a layer containing an organic compound serves as a light emitting layer is manufactured, it is preferable that the support member is used as a sealing member. Thus, a light emitting element is completely shielded from external so as to prevent entrance of a substance such as moisture or oxygen for promoting deterioration of an organic compound layer from external. In addition, when the light emitting device represented by the device having a light emitting element in which a layer containing an organic compound serves as a light emitting layer is manufactured, as in the case of the support member, it is preferable that the transfer body sufficiently prevents entrance of a substance such as moisture or oxygen for promoting deterioration of an organic compound layer from external. In addition, an order for manufacturing the light emitting device is not particularly limited. For example, after the light emitting element is formed, a plastic substrate as the support member is bonded to the layer to be peeled which is provided to a substrate, the substrate is peeled, and a plastic substrate as the transfer body is bonded to the layer to be peeled. Alternatively, after the light emitting element is formed, the substrate is peeled, a plastic substrate as a first transfer body is boned to the layer to be peeled, and then a plastic substrate as a second transfer body is bonded thereto. In addition, when it is important to suppress the deterioration occurring due to transmission of moisture or oxygen, a thin film is formed in contact with the layer to be peeled after peeling to repair a crack caused at peeling. When a film having thermal conductivity, specifically, an aluminum nitride or an aluminum oxynitride is used as the thin film which is in contact with the layer to be peeled, in addition to an effect for radiating heat generated in the element to suppress the deterioration thereof, an effect for preventing deformation or degradation of, the transfer body, specifically, a plastic substrate can be obtained. In addition, the film having thermal conductivity has an effect for preventing mixing of an impurity such as moisture or oxygen from external.

[0095] The present invention made by the above constitutions will be described in more detail through the following embodiments.

Embodiment 1

[0096] Here, an example of laser processing apparatus applicable to the present invention will be described.

[0097] Crystallization of amorphous silicon by laser annealing is conducted through a melting-solidification process. More specifically, the case where it is divided into two stages, that is, a stage of generation of crystal nucleus and a stage of crystal growth from the nucleus is considered. However, in the case of laser annealing using a pulse laser beam, a generation position of crystal nucleus and a generation density thereof cannot be controlled but left to natural generation. Thus, a crystal grain is formed at an arbitrary position within the surface of a glass substrate and only a small size of about 0.2 m to 0.5 m is obtained. A large number of defects are caused in a crystal boundary. This is considered to be a factor limiting the field effect mobility of a TFT.

[0098] It is considered that a method of conducting crystallization with melting-solidification by continuous oscillation laser scanning is a method similar to a zone melting method. However, according to the method, a large beam size cannot be obtained. In addition, it is obvious that much time is required for achieving crystallization over the entire surface of a large area substrate.

[0099] In this embodiment, a laser processing apparatus for conducting laser beam irradiation with a state in which an irradiation position is substantially aligned with a position in which a TFT is produced, over the entire surface of a large area substrate for crystallization, so that a crystalline semiconductor film having a large grain size can be formed at high throughput will be described below.

[0100] As a laser irradiation apparatus of Embodiment 1, the following apparatus may be used. The laser irradiation apparatus includes a first movable mirror for deflecting a laser beam in a main scanning direction and a second movable mirror for receiving the laser beam deflected in the main scanning direction and conducting scanning in a sub scanning direction, which is a long shape. The second movable mirror has means for scanning a laser beam in the sub scanning direction at a rotation angle about the axis of the long shape direction as a center to irradiate the laser beam to an object to be processed which is placed on a stage.

[0101] Also, as another laser irradiation apparatus, the following apparatus may be used. That is, the laser irradiation apparatus includes a first laser beam scanning system and a second laser beam scanning system. The first laser beam scanning system has a first movable mirror for deflecting a laser beam in a first main scanning direction and a long second movable mirror for receiving the laser beam deflected in the first main scanning direction and conducting scanning in a first sub scanning direction. The second laser beam scanning system has a third movable mirror for deflecting a laser beam in a second main scanning direction and a long fourth movable mirror for receiving the laser beam deflected in the second main scanning direction and conducting scanning in a second sub scanning direction. The second movable mirror has means for scanning a laser beam in the first sub scanning direction at a rotation angle about the axis of the long shape direction as a center to irradiate the laser beam to an object to be processed which is placed on a stage. The fourth movable mirror has means for scanning a laser beam in the second sub scanning direction at a rotation angle about the axis of the long shape direction as a center to irradiate the laser beam to the object to be processed which is placed on the stage.

[0102] In the above configuration, a galvanomirror or a polygon mirror is applied to the first and second movable mirrors. A solid laser or a gas laser is preferably applied to a laser for providing the laser beam.

[0103] In the above configuration, a laser beam is scanned in the main scanning direction by the first movable mirror and scanned in the sub scanning direction by the second movable mirror. Thus, the laser beam can be irradiated in an arbitrary position onto the object to be processed. In addition, a plurality of such laser beam scanning means are provided and laser beams are irradiated to a surface to be formed in biaxial directions. Thus, a laser processing time can be shortened.

[0104] Hereinafter, a laser irradiation apparatus of this embodiment will be described with reference to the drawings.

[0105] FIG. 3 shows a desirable example of a laser processing apparatus of this embodiment. The shown laser processing apparatus includes a solid laser 101 capable of conducting continuous oscillation or pulse oscillation, a lens such 102 as a collimator lens or a cylindrical lens for condensing a laser beam, a fixed mirror 103 for changing an optical path of the laser beam, a galvanomirror 104 for radially scanning the laser beam in a two-dimensional direction, and a movable mirror 105 for receiving the laser beam by the galvanomirror 104 and irradiating the laser beam toward a surface to be irradiated of a stage 106. An optical axis of the galvanomirror 104 and that of the movable mirror 105 are intersected each other and rotated in arrow directions shown in FIG. 3, respectively. Thus, a laser beam can be scanned over the entire surface of a substrate 107 placed on the stage 106. When the movable mirror 105 is used as an f mirror to correct an optical path difference, a beam shape on a surface to be irradiated can be also adjusted.

[0106] FIG. 3 shows a system for scanning a laser beam in a uniaxial direction of the substrate 107 placed on the stage 106 by the galvanomirror 104 and the movable mirror 105. As a more preferable configuration, as shown in FIG. 4, a half mirror 108, a fixed mirror 109, a galvanomirror 110, and a movable mirror 111 is added to the configuration shown in FIG. 3, and laser beams may be simultaneously scanned in biaxial directions (X- and Y-directions). A processing time can be shortened by using such a configuration. Note that the galvanomirrors 104 and 110 may be replaced by polygon mirrors.

[0107] A solid laser is preferable as the laser, and a solid laser using crystal such as YAG, YVO.sub.4, YLF, or YAl.sub.5O.sub.12 which is doped with Nd, Tm, or Ho, or a semiconductor laser is preferably used. A fundamental wave of an oscillation wavelength is changed dependent on a doping material. An oscillation is produced at a wavelength of 1 m to 2 m. When a non-single crystalline semiconductor film is crystallized, in order to selectively absorb a laser beam by the semiconductor film, it is preferable that the second harmonic to the fourth harmonic of the oscillation wavelength is applied. Typically, in the case of crystallization of amorphous silicon, the second harmonic (532 nm) of an Nd:YAG laser (fundamental wave: 1064 nm) is used.

[0108] In addition, a gas laser such an argon laser, a krypton laser, or an excimer laser can be applied.

[0109] Also, an atmosphere at laser light irradiation may be an atmosphere containing oxygen, an atmosphere containing nitrogen, an inert atmosphere, or a vacuum and is preferably selected as appropriate according to a purpose.

[0110] An oscillation mode may be either pulse oscillation or continuous oscillation. In order to achieve continuous crystal growth with a molten state of the semiconductor film, it is desirable that a continuous oscillation mode is selected.

[0111] In the case where a TFT which is made from a semiconductor film crystallized by laser annealing is formed on a substrate, when a crystal growth direction is aligned with a carrier moving direction, high field effect mobility can be obtained. In other words, when the crystal growth direction is aligned with the channel length direction, the field effect mobility can be substantially increased.

[0112] When a continuous oscillating laser beam is irradiated to a non-single crystalline semiconductor film for crystallization, a solid-liquid interface is kept. Thus, a continuous crystal growth can be obtained in the scanning direction of the laser beam. As shown in FIG. 4, with respect to a TFT substrate (substrate to which TFTs are mainly formed) 112 used for manufacturing an active matrix liquid crystal display device which driver circuits are integrally formed, driver circuit portions 114 and 115 are provided in the vicinity of a pixel portion 113. FIG. 4 shows a configuration of a laser irradiation apparatus made in consideration of such a layout. As described above, in the case of the configuration in which laser beams are incident from the biaxial directions, laser beams can be synchronously or asynchronously irradiated in an X-direction and a Y-direction indicated by arrows in the drawing by a combination of the galvanomirrors 104 and 110 and the movable mirrors 105 and 111. In addition, it is possible that a location is designated according to the layout of TFTs and a laser beam is irradiated thereto.

[0113] FIG. 5 shows a relationship between the substrate 112 to which TFTs are provided and an irradiation direction of a laser beam in detail. Regions in which the pixel portion 113 and the driver circuits 114 and 115 are formed are indicated by dot lines on the substrate 112. In a stage of crystallization, a non-single crystalline semiconductor film is formed on the entire surface. Semiconductor regions for forming TFTs can be designated by alignment makers or the like formed in end portions of the substrate.

[0114] For example, the driver circuit portion 114 is a region for forming a scan line driver circuit. In its partially enlarged view 301, semiconductor regions 204 of TFTs and a scanning direction of a laser beam 201 are indicated. The semiconductor regions 204 having an arbitrary shape can be applied. In any case, the channel length direction is aligned with the scanning direction of the laser beam 201. In addition, the driver circuit portion 115 extended in a direction in which it intersects the driver circuit portion 114 is a region for forming a data line driver circuit, and an arrangement of semiconductor regions 205 is aligned with a scanning direction of a laser beam 202 (enlarged view 302). Similarly, in the case of the pixel portion 113, as shown in an enlarged view 303, an arrangement of no semiconductor regions 206 is aligned and a laser beam 202 is scanned in a channel length direction. The scanning direction of the laser beam is not limited to a single direction and round trip scanning may be conducted.

[0115] Next, steps of crystallizing a non-single crystalline semiconductor film and producing a TFT from the formed crystalline semiconductor film will be descried with reference to FIGS. 6A to 6D. FIG. 6B is a longitudinal cross sectional view. A non-single crystalline semiconductor film 403 is formed on a glass substrate 401. A typical example of the non-single crystalline semiconductor film 403 is an amorphous silicon film. In addition, an amorphous silicon germanium film or the like can be applied. The thickness of 10 nm to 20 nm can be applied and may be further increased in accordance with a wavelength of a laser beam and an energy density thereof. In addition, it is desirable to employ such a measure that a blocking layer 402 is provided between the glass substrate 401 and the non-single crystalline semiconductor film 403 so as not to diffuse an impurity such as alkali metal from the glass substrate into the semiconductor film. A silicon nitride film, a silicon oxynitride film, or the like is applied as the blocking layer 402.

[0116] Also, a laminate 409 of a metallic layer or a metallic nitride layer and an oxide layer is formed between the blocking layer 402 and the substrate 401 for peeling. As the metallic layer or the nitride layer, there is preferably used a nitride comprising a single layer made of an element selected from Ti, Al, Ta, W, Mo, Cu, Cr, Nd, Fe, Ni, Co, Ru, Rh, Pd, Os, Ir, and Pt, or an alloy material or a compound material which contains the above element as a main ingredient, or a laminate of those. For example, a single layer made of titanium nitride, tungsten nitride, tantalum nitride, or molybdenum nitride, or a laminate of those is preferably used. Here, a titanium nitride film having a film thickness of 100 nm which is formed by a sputtering method is used. Note that, when a contact property to the substrate is low, a buffer layer is preferably provided. A single tungsten layer and a tungsten nitride have a high contact property and are exemplified as preferable materials. In addition, as the oxide layer, a single layer made of a silicon oxide material or a metallic oxide material, or a laminate of those is preferably used. Here, a silicon oxide film having a film thickness of 200 nm which is formed by a sputtering method is used. Bonding force between the metallic nitride layer and the oxide layer has a sufficient strength to withstand heat treatment Thus, film peeling (which is also called peeling) or the like is not caused. However, peeling can be simply performed in an inner portion of the oxide layer or a boundary thereof by a physical means. Note that a glass substrate is used here. However, various substrates can be used in the above peeling method. As the substrate 401, a quartz substrate, a ceramic substrate, a silicon substrate, a metallic substrate, or a stainless steel substrate may be used.

[0117] Next, crystallization is conducted by irradiation of a laser beam 400. Thus, a crystalline semiconductor film 404 can be formed. As shown in FIG. 6A, the laser beam 400 is scanned to a position of a semiconductor region 405 where a TFT will be formed. A beam shape can be set to be an arbitrary shape such as a rectangular shape, a linear shape, or an elliptical shape. With respect to the laser beam condensed by an optical system, an energy intensity at a central region thereof is not necessarily equal to that at an edge region. Thus, it is desirable that the semiconductor region 405 is not overlapped with the edge region of the beam.

[0118] Scanning of the laser beam is not limited to scanning in only a single direction and round trip scanning may be conducted. In this case, a laser energy density is changed every time scanning is conducted. Thus, a stepwise crystal growth can be produced. The scanning can also serve as dehydrogenation processing which is often required in the case of crystallization of amorphous silicon. For example, first scanning is conducted at a low energy density to release hydrogen, and then second scanning is conducted at an increased energy density to complete the crystallization.

[0119] When continuous oscillation laser beam irradiation is conducted in such a laser beam irradiation method, the growth of crystal having a larger grain size is possible. Of course, in order to realize this, it is necessary to set parameters such as a scanning speed of a laser beam and an energy density thereof in detail as appropriate. When the scanning speed is set to 10 cm/sec to 80 cm/sec, the above crystal growth can be realized. It is said that a speed of crystal growth through melting-solidification using a pulse laser is 1 m/sec. If a laser beam is scanned at a speed lower than the crystal growth speed and slow cooling is conducted, continuous crystal growth in a solid-liquid interface is possible. Thus, an increase in a grain size of crystal can be realized.

[0120] According to the laser irradiation apparatus of this embodiment, under such a situation, it is possible that a position on the substrate is arbitrarily designated and laser beam irradiation is conducted for crystallization. When laser beam irradiation is conducted from biaxial directions, a throughput can be further improved.

[0121] Also, when laser beam irradiation is conducted, clean peeling from the substrate can be performed with smaller force. Thus, a layer to be peeled that has a large area can be peeled over the entire surface thereof.

[0122] In order to further promote peeling, a granular oxide (for example, ITO (alloy of indium oxide and tin oxide), an alloy of indium oxide and zinc oxide (In.sub.2O.sub.3ZnO), a zinc oxide (ZnO) or the like) may be provided in an interface between the nitride layer, the metallic layer, or the metallic nitride layer and the oxide layer.

[0123] Then, as shown in FIGS. 6C and 6D, the formed crystalline semiconductor film is etched to form a semiconductor region 405 in an island-shape. In the case of a top gate TFT, a gate insulating film 406, a gate electrode 407, and impurity regions 408 having a conductivity type are formed on the semiconductor region 405 to produce a TFT. Then, wirings, an interlayer insulating film, and the like are preferably formed as appropriate by a known technique to produce an element.

[0124] Thus, after obtaining the element having the TFT, the substrate 401 is peeled in accordance with the embodiment mode. In this embodiment, a resultant layer formed on the blocking layer 402 corresponds to the layer to be peeled 11b which is described in the embodiment mode. When the mechanical strength of the layer to be peeled is insufficient, it is preferable that the substrate is peeled after a support member (not shown) for fixing the layer to be peeled is bonded thereto.

[0125] The layer to be peeled which is formed on the oxide layer can be simply separated from the substrate by peeling. The peeled layer can be bent in a certain direction. It is needless to say that the layer to be peeled can be bonded to a transcriptional body (not shown) having a curved surface.

[0126] Also in this embodiment, according to the present invention, the irradiation direction (scanning direction) of the laser light and the channel length directions of all semiconductor layers 204 to 206 and 405 provided to the layer to be peeled are set to be the same direction, and these directions and the bending direction are set to be orthogonal to each other. Thus, a display having a curved surface can be realized.

[0127] Also, this embodiment can be freely combined with the embodiment mode.

Embodiment 2

[0128] The example of the top gate TFT is descried in Embodiment 1. Here, an example of a bottom gate TFT will be described. Also, the structure except for the TFT is the same one as Embodiment 1 and the description is thereof omitted here.

[0129] Next, steps of crystallizing a non-single crystalline semiconductor film and producing a TFT from the formed crystalline semiconductor film will be descried with reference to FIGS. 7A to 7D.

[0130] FIG. 7B is a longitudinal cross sectional view. A non-single crystalline semiconductor film 503 is formed on a gate insulating film 506 covering a gate electrode. A typical example of the non-single crystalline semiconductor film 503 is an amorphous silicon film. In addition, an amorphous silicon germanium film or the like can be applied. The thickness of 10 nm to 200 nm can be applied and may be further increased in accordance with a wavelength of a laser beam and an energy density thereof. In addition, it is desirable to employ such a measure that a blocking layer 502 is provided between the glass substrate 501 and the gate electrode so as not to diffuse an impurity such as alkali metal from the glass substrate into the semiconductor film. A silicon nitride film, a silicon oxynitride film, or the like is applied as the blocking layer 502.

[0131] Also, a laminate 509 of a metallic layer or a metallic nitride layer and an oxide layer is formed between the blocking layer 502 and the substrate 501 for peeling. As the metallic layer or the nitride layer, there is preferably used a nitride comprising a single layer made of an element selected from Ti, Al, Ta, W, Mo, Cu, Cr, Nd, Fe, Ni, Co, Ru, Rh, Pd, Os, Ir, and Pt, or an alloy material or a compound material which contains the above element as a main ingredient, or a laminate of those is preferably used. For example, a single layer made of titanium nitride, tungsten nitride, tantalum nitride, or molybdenum nitride, or a laminate of those is preferably used. Here, a titanium nitride film having a film thickness of 100 nm which is formed by a sputtering method is used. Note that, when a contact property to the substrate is low, a buffer layer is preferably provided. A single tungsten layer and a tungsten nitride have a high contact property and are exemplified as preferable materials. In addition, as the oxide layer, a single layer made of a silicon oxide material or a metallic oxide material, or a laminate of those is preferably used. Here, a silicon oxide film having a film thickness of 200 nm which is formed by a sputtering method is used. Bonding force between the metallic nitride layer and the oxide layer has a strength resistant to heat treatment. Thus, film peeling (which is also called peeling) or the like is not caused. However, peeling can be simply performed in an inner portion of the oxide layer or a boundary thereof by a physical means.

[0132] Next, crystallization is conducted by irradiation of a laser beam 500. Thus, a crystalline semiconductor film 504 can be formed. The laser beam is obtained from the laser processing apparatus described in Embodiment 1. As shown in FIG. 7A, the laser beam 500 is scanned to a position of a semiconductor region 505 where a TFT will be formed. A beam shape can be set to be an arbitrary shape such as a rectangular shape, a linear shape, or an elliptical shape. With respect to the laser beam condensed by an optical system, an energy intensity at a central region thereof is not necessarily equal to that at an edge region. Thus, it is desirable that the semiconductor region 505 is not overlapped with the edge region of the beam.

[0133] Scanning of the laser beam is not limited to scanning in only a single direction and round trip scanning may be conducted. In this case, a laser energy density is changed every time scanning is conducted. Thus, a stepwise crystal growth can be produced. The scanning can also serve as dehydrogenation processing which is often required in the case of crystallization of amorphous silicon. For example, first scanning is conducted at a low energy density to release hydrogen, and then second scanning is conducted at an increased energy density to complete the crystallization.

[0134] When continuous oscillation laser beam irradiation is conducted in such a laser beam irradiation method, the growth of crystal having a larger grain size is possible. Of course, in order to realize this, it is necessary to set parameters such as a scanning speed of a laser beam and an energy density thereof in detail as appropriate. When the scanning speed is set to 10 cm/sec to 80 cm/sec, the above crystal growth can be realized. It is said that a speed of crystal growth through melting-solidification using a pulse laser is 1 m/sec. If a laser beam is scanned at a speed lower than the crystal growth speed and slow cooling is conducted, continuous crystal growth in a solid-liquid interface is possible. Thus, an increase in a grain size of crystal can be realized.

[0135] Also, when laser beam irradiation is conducted, clean peeling from the substrate can be performed with smaller force. Thus, a layer to be peeled that has a large area can be peeled over the entire surface thereof.

[0136] In order to further promote peeling, a granular oxide (for example, ITO (alloy of indium oxide and tin oxide), an alloy of indium oxide and zinc oxide (In.sub.2O.sub.3ZnO), a zinc oxide (ZnO) or the like) may be provided in an interface between the nitride layer, the metallic layer, or the metallic nitride layer and the oxide layer.

[0137] Then, as shown in FIGS. 7C and 7D, the formed crystalline semiconductor film is etched to form a semiconductor region 505 in an island-shape. Here, an etching stopper is provided on a semiconductor region 505, and impurity regions 508 having one conductivity type are formed to produce a TFT. Then, wirings, an interlayer insulating film, and the like are preferably formed as appropriate by a known technique to produce an element.

[0138] Thus, after obtaining the element having the TFT, the substrate 501 is peeled in accordance with the embodiment mode. In this embodiment, a resultant layer formed on the blocking layer 502 corresponds to the layer to be peeled 11b which is described in the embodiment mode. When the mechanical strength of the layer to be peeled is insufficient, it is preferable that the substrate is peeled after a support member (not shown) for fixing the layer to be peeled is bonded thereto.

[0139] The layer to be peeled which is formed on the oxide layer can be simply separated from the substrate by peeling. The peeled layer can be bent in a certain direction. It is needless to say that the layer to be peeled can be bonded to a transfer body (not shown) having a curved surface.

[0140] Even in this embodiment, the irradiation direction (scanning direction) of the laser light and the channel length directions of the semiconductor layer 505 provided to the layer to be peeled are set to be the same direction, and these directions and the bending direction are set to be orthogonal to each other. Thus, a display having a curved surface can be realized.

[0141] Also, this embodiment can be freely combined with the embodiment mode.

Embodiment 3

[0142] In accordance with the present embodiment, FIG. 8 shows a technique for transferring a layer to be peeled containing TFT.

[0143] In FIGS. 8A to 8G, reference numeral 830 indicates a first substrate; reference numeral 831 indicates a first material layer composed of a nitride layer or a metallic layer; reference numeral 832 indicates a second material layer composed of an oxide layer; reference numeral 833 indicates a layer to be peeled; reference numeral 834 indicates a first adhesive; reference numeral 835 indicates a second substrate; reference numeral 836 indicates a second adhesive; and reference numeral 837 indicates a third substrate.

[0144] In accordance with the present embodiment, the first substrate 830 may be constituted by a glass substrate, a quartz substrate, a ceramic substrate or the like. Further, it is also possible to use a semiconductor substrate such as a silicon substrate, or a metallic substrate such as a stainless steel substrate. Here, a glass substrate (#1737) having a thickness of 0.7 mm is used.

[0145] First, as shown in FIG. 8A, on top of the substrate 830, the first material layer 831 is formed. The first material layer 831 may be a material which, immediately after the film is formed, exhibits one of compression stress and tension stress. However, it is important to use a material in which abnormalities such as peeling due to thermal processing and laser light radiation, in the forming of the layer to be peeled, do not occur, and which exhibits tension stress in a range of 1 to 110.sup.10 (Dyne/cm.sup.2) after the forming of the layer to be peeled. A representative example is a single layer constituted by an element selected from the group consisting of W, WN, TiN, and TiW, or by an alloy metal or compound material having the element as its main component, or a laminate thereof. Note that, the first material layer 831 may be formed using a sputtering method.

[0146] Next, the second material layer 832 is formed on top of the first material layer 831. In the second material layer 832, it is important to use a material in which abnormalities such as peeling caused by the thermal processing and the laser light radiation, in the forming of the layer to be peeled, do not occur, and which exhibits compression stress in a range of 1 to 110.sup.10 (Dyne/cm.sup.2) after the forming of the layer to be peeled. Representative examples of the second material layer include oxide silicon, oxide silicon nitride, oxide metallic material, and a laminate of these. Note that, the second material layer 832 may be formed using a sputtering method. In the case where the second material layer 832 is formed using the sputtering method, an inert gas such as argon gas is introduced into the chamber, to include a minute amount of Argon gas elements into the second material layer 832.

[0147] Regarding the first material layer 831 and the second material layer 832, the film widths of each of the layers is set as needed within a range of 1 nm to 1000 nm, to thereby adjust the internal stress of the first material layer 831 and the internal stress of the second material layer 832.

[0148] Further, in FIGS. 8A to 8G, in order to streamline the process, an example has been shown in which the first material layer 831 is formed in contact with the substrate 830. However, an insulating layer or metallic layer serving as a buffer may be provided in between the substrate 830 and the first material layer 831 to improve the adhesion with the substrate 830.

[0149] Next, the layer to be peeled is formed onto the second material layer 832. (See FIG. 8A). The layer to be peeled 833 may contain various elements (e.g., a film transistor, light emitting elements in which a layer containing organic compounds serves as a light emitting layer, elements containing liquid crystals, a memory element, a thin-film diode, a photoelectric conversion element formed by a silicon pin junction, or a silicon resistor element). However, in the case of elements containing liquid crystals, the layer to be peeled 833 must include a substrate which opposes it. Further, the process of forming the layer to be peeled 833 can be accomplished by thermal processing conducted within the temperature range that the first substrate 830 can withstand. Note that, even if the internal stress in the second material layer 832 and the internal stress in the first material layer 831 are different than each other, the thermal processing in the manufacturing of the layer to be peeled 833 will not cause peeling to occur.

[0150] Next, a process is performed for partially reducing the adhesion between the first material layer 831 and the second material layer 832. The processing for partially reducing the adhesion is a process in which a laser light is partially radiated on the first material layer or on the second material layer along the perimeter of the region to be peeled, or is a process in which localized pressure is applied from the outside along the perimeter of the region to be peeled to apply damage to a part of the inside or the surface of the second material layer. Specifically, a diamond or other such hard needle may be pressed perpendicularly and moved while applying pressure. Preferably, a scriber device is used and is pressed down by an amount of 0.1 mm to 2 mm, with pressure being applied as it is moved. In this way, before performing the peeling, it is important to create a portion where peeling can occur easily, which serves as a starter. By performing the preprocessing in which the selective (partial) reduction of the adhesion takes place, defective peelings are eliminated and yield is improved.

[0151] Next, the second substrate 835 and the layer to be peeled 833 are adhered to each other using the first adhesive 834. (See FIG. 8B). The first adhesive 834 may be a reactive-curing type adhesive, a thermal-curing type adhesive, an ultraviolet-curing type adhesive or other such photo-curing type adhesive, or may be an aerophobic-type adhesive, or other various types of curing adhesive. Moreover, these adhesives may be soluble such that they dissolve in a solvent, and/or may be photosensitive such that their adhesiveness decreases when irradiated with light. The composition of these adhesives may be, for example, epoxy-type, acrylic-type, silicon-type or anything else. The application of the adhesive may be carried out by a coating method, for example. Note that, the first adhesive is removed in subsequent steps. Here, a soluble adhesive material which can dissolve in a solvent is used as the fist adhesive.

[0152] Further, instead of the first adhesive 834, a tape having adhesive on one or both of its surfaces may be used. The tape may include on one or both of its surfaces an adhesive which is soluble so as to dissolve in a solvent, or is photosensitive so as to lose adhesiveness when irradiated with light.

[0153] The second substrate 835 may be constituted by a glass substrate, a quartz substrate, a ceramic substrate, a plastic substrate or the like. Further, it is also possible to use a semiconductor substrate such as a silicon substrate, or a metallic substrate such as a stainless steel substrate.

[0154] The present embodiment employs a highly rigid quartz substrate (thickness: 1.1 mm) for the second substrate 835, having a thickness larger than the first substrate 830. In the case where a plastic film is used for the second substrate, when the elements formed onto the first substrate 830 are transferred onto the plastic filmwhich is to say when the layer to be peeled 833 and the film are adhered to each other by the first adhesive 834 and the film is lifted upthere was a risk that the film will bend and cause cracks to form in the layer to be peeled 833. Therefore, after fixing the layer to be peeled 833 formed over the first substrate 830 to the rigid second substrate 835 with the first adhesive 834, the first substrate 830 is peeled. Then, after the plastic film (i.e., the third substrate 837) is fixed to the layer with the second adhesive 836, the second substrate 835 is removed. By following this procedure, it becomes difficult for cracks to occur.

[0155] Next, peeling is performed from the above-mentioned region where the adhesiveness has been reduced, and the first substrate 830 having the first material layer 831 is separated by using a physical means (FIG. 8C). Since the second material layer 832 exhibits compressional stress and the first material has tension stress, the separation can be achieved with relatively little force (such as the force of a human hand, or wind pressure of gas blown from a nozzle, or ultrasonic waves, etc.).

[0156] Thus, the layer to be peeled 833 formed onto the second material layer 832 can be separated from the first material layer 830. FIG. 8D shows the post-peeled state.

[0157] Next, the third substrate 837 and the second material layer 832 (and the peeled layer 833) are adhered together with the second adhesive 836. (See FIG. 8E). It is important that the second adhesive 836 has greater adhesive force than the first adhesive 834.

[0158] The second adhesive 836 may be a reactive-curing type adhesive, a thermal-curing type adhesive, an ultraviolet-curing type adhesive or other such photo-curing type adhesive, or may be an aerophobic-type adhesive, or other various types of curing adhesive. Moreover, these adhesives may be soluble such that they dissolve in a solvent, and/or may be photosensitive such that their adhesiveness decreases when irradiated with light. The composition of these adhesives may be, for example, epoxy-type, acrylic-type, silicon-type or anything else. The application of the adhesive may be carried out by a coating method, for example. Note that, the second adhesive becomes one support for the layer to be peeled at a subsequent step. For the second adhesive 836, a material is used which will achieve a high degree of adhesion between the third substrate and the second adhesive, and also between the second adhesive and the layer to be peeled. Here, an ultraviolet-curing type adhesive is used for the second adhesive 836.

[0159] Further, in the case where the second adhesive 836 is made of a material which is soluble so as to dissolve in a solvent, or is photosensitive such that it loses adhesive strength when exposed to light, it becomes possible to peel the third substrate at a later step, and it is possible for only the second adhesive to serve as the support. Further, instead of the second adhesive 836, it is possible to use a tape having adhesive on one or both of its surfaces. The surface or surfaces of this tape may have an adhesive which is soluble so as to dissolve in a solvent, or photosensitive such that its adhesive strength decreases when the tape is exposed to light.

[0160] A flexible substrate may be used for the third substrate 837. The present embodiment employs a plastic film for the third substrate 837.

[0161] Once the situation shown in FIG. 8E is achieved, it is then soaked in the solvent and only the second substrate 835 is removed. (See FIG. 8F). Since the first adhesive is a soluble adhesive material, the second substrate 835 is removed easily, thus separating the second substrate 835 and the layer to be peeled 833.

[0162] Further, an input/output terminal of the element contained in the layer to be peeled 833 is formed so as to be exposed from the topmost layer (i.e., the layer closest to the second substrate side) of the layer to be peeled. Therefore, after the step of separating the second substrate, it is preferable that the first adhesive is completely removed from the surface of the layer to be peeled so that the input/output terminal portion can be exposed.

[0163] Further, in the present embodiment, there is shown the example in which the soluble adhesive material is used for the first adhesive 834 such that it dissolves in a solvent and in which the second substrate is soaked in the solvent and removed. However, the invention is not particularly restricted to this configuration. For example, a thermal-curing type adhesive (which loses adhesive strength when irradiated with ultraviolet light) may be used for the first adhesive, and ultraviolet rays may be radiated to thereby remove the second substrate.

[0164] The steps described above enable the manufacture of a semiconductor device equipped with the layer to be peeled 833 which serves as a support for the second adhesive 836 and the third substrate 837. Then, by curving the device as shown in FIG. 8G, it thus becomes possible to achieve a semiconductor device in which the curved surface of the semiconductor device exhibits a radius of curvature of from 50 cm to 200 cm. When curving the device, it can be attached to the curved surface to which it is going to be mounted. Note that, between the second adhesive 836 and the layer to be peeled 833, there is the oxide layer 832 that is the second material layer. In the semiconductor device obtained as described above, the second material layer 832 is applied by a sputtering method and minute amounts of inert gas elements are included in the second material layer 832. Therefore, the semiconductor device as a whole can be made flexible.

[0165] Further, in accordance with the present embodiment, the device was curved after being attached to the third substrate. However, it is also possible to curve the device by attaching it directly to a base that has a curved surface, with the second adhesive 836.

[0166] Here, an example was shown in which the semiconductor device is built to completion according to the steps described above, but it is also possible to follow the above-mentioned steps to complete the semiconductor device only partially. For example, according to the above-mentioned steps, it is possible to form the layer to be peeled containing a TFT circuits, and then, after obtaining the layer to be peeled which has as a support therefor the second adhesive and the third substrate, steps of forming elements may be added to complete any of a variety of semiconductor devices, such as a light emitting device or a liquid crystal display device having a light emitting elements in which a layer containing an organic compound serves as a light emitting layer.

[0167] Further, it is also possible to make a light emitting device having a light emitting element in which a layer containing a passive organic element compound serves as a light emitting layer.

[0168] Further, in a case where, in order to reduce the adhesion between the third substrate and the second adhesive, a plastic film in which an AlN.sub.XO.sub.Y film is formed on the surface is formed as the third substrate 837, it becomes possible to separate the second substrate 835 and the third substrate 837. It thus becomes possible to manufacture a semiconductor device equipped with the layer 833 to be peeled having the second adhesive 836 as a support. Since such a semiconductor device has only the second adhesive as a support, it can be made thin, lightweight and flexible.

[0169] Further, by following the above-mentioned steps, the present inventors actually performed electrical measurement of the TFT formed onto the first substrate before peeling the first substrate, and after separating the first and the second substrate, they performed the electrical measurement of the TFT once again. There was hardly any change in the characteristics of the TFT before and after separation. FIG. 9 is a V-I characteristic graph for an n-channel type TFT with the post-separation channel length L/channel width W=50 m/50 m. Further, FIG. 10 is a V-I characteristic graph for a p-channel type TFT with the post-separation channel length L/channel width W=50 m/50 m.

[0170] Since there was hardly any change in the characteristics of the TFT before and after the separation, it can be said that, even when the transferring and application are performed according to the sequence described above, the above-mentioned steps do not affect the TFTs. Further, it is also possible to directly form the TFT onto the plastic substrate; however, since the substrate's heat resistance is low, it would be difficult to perform thermal processing at 300 C. or higher. Thus, it would be difficult to form the TFT with the excellent characteristics shown in FIG. 9 and FIG. 10. As demonstrated in the present embodiment, after the TFT is formed onto the heat-resistant substrate, the heat-resistant substrate is then peeled. Accordingly, it thus becomes possible to form the TFT exhibiting the excellent characteristics as shown in FIG. 9 and FIG. 10.

Embodiment 4

[0171] In this embodiment, in accordance with the technology described in Embodiment 3, manufacturing steps of a light emitting device having a light emitting element in which a layer having an organic compound serves as a light emitting layer will be described with reference to FIG. 11.

[0172] First, pixel portions (n-channel TFTs and p-channel TFTs) and driver circuits (n-channel TFTs and p-channel TFTs) provided in the vicinity of the pixel portions are manufactured simultaneously on one substrate, organic light emitting elements (also called organic light emitting device) are formed thereon.

[0173] A first material layer 931 made of nitride layer or a metal layer and a second material layer 932 made of an oxide layer are formed on a first substrate in accordance with Embodiment 3.

[0174] Next, a layer containing TFTs and wirings is formed on the second material layer 932 in accordance with the technology shown in Embodiment 1. After an insulation film for covering respective TFT is formed, a cathode or an anode electrically connected with TFTs provided in the pixel portion is formed. Further, an insulator called bank is formed to cover the ends of the cathode or the anode on both ends thereof. Moreover, if necessary, it is practicable to form a passivation film (protection film) to cover TFTs optionally. And, an EL layer (organic compound material layer) and the anode or the cathode of organic light emitting elements are formed on the cathode or the anode both ends of which are covered by bank. When the under layer of the EL layer is a cathode, an anode can be provided on the EL layer, on the contrary, when the under layer of the EL layer is an anode, a cathode can be provided on the EL layer.

[0175] As the EL layer, an EL layer (layer for light emitting and making carrier perform the migrate for it) may be formed by freely combining the light emitting layer, a charge injection layer or a charge implantation layer. For example, low molecular system organic EL material and high molecular system organic EL material may be employed. Moreover, as an EL layer, a thin film out of a light emitting material (singlet compound) which light-emits (fluorescence) due to singlet excitation, or a thin film out of a light emitting material (triplet compound) which emits (phosphorescence) due to triplet excitation can be used. Moreover, an inorganic material such as silicon carbide or the like is capable of being used as a charge transport layer and a charge injection layer. For these organic EL material and inorganic material, the known materials can be used. In addition, the EL layer is totally formed in a thickness around 100 nm as a thin film layer. For this reason, it is necessary to enhance the evenness of the surface of the cathode or the anode.

[0176] As a material used for a cathode, it is said that it is preferable to use a metal having a small work function (representatively, metal elements belonging to I group or II group of the periodic table) or an alloy containing these. Since the smaller the work function is, the more the luminous efficiency is enhanced, it is preferable that among these, as a material used for a cathode, an alloy containing Li (lithium), which is one of alkaline metals, is used.

[0177] As a conductive film using for the anode, a material having a bigger working function in comparison with the material of the cathode such as ITO (indium oxide-tin oxide alloy), indium oxide-zinc oxide alloy (In.sub.2O.sub.3ZnO), zinc oxide (ZnO) or the like may be used. Further, a material having a lower sheet resistance than ITO, specifically, platinum (Pt), chromium (Cr), tungsten (W), or nickel (Ni) may be used.

[0178] An organic light emitting layer is defined in this specification as an aggregate of layers formed between an anode and cathode of a light emitting element in which a layer containing an organic compound serves as a light emitting layer. Specifically, an organic light emitting layer includes a light emitting layer, a hole injecting layer, an electron injecting layer, a hole transporting layer, an electron transporting layer, etc. The basic structure of an organic light emitting element is a laminate of an anode, a light emitting layer, and a cathode layered in order. The basic structure may be modified into a laminate of an anode, a hole injecting layer, a light emitting layer, and a cathode layered in order, or a laminate of an anode, a hole injecting layer, a light emitting layer, an electron transporting layer, and a cathode layered in order.

[0179] A light emitting element in which a layer containing an organic compound serves as a light emitting layer has, in addition to an anode and a cathode, a layer containing an organic compound (light emitting material) that generates luminescence (electro luminescence) when an electric field is applied (the layer is hereinafter referred to a light emitting layer).

[0180] When a current flowing to the light emitting element is controlled by TFTs, there are two methods in a rough dividing way. Specifically, one method is controlling the current in a voltage region called saturation region, the other is controlling the current in a voltage region before reaching the saturation region. In this specification, a Vd region where a current value is substantial constant is referred to as a saturation region in Vd-Id curve. In addition, in the invention, there is no limitation putted on the driving methods of the light emitting element, that is to say, any driving methods can be used.

[0181] By the steps up through this point, the layer to be peeled is formed by laminating a layer 933b containing the light emitting element in which a layer containing an organic compound serves as a light emitting layer, and a layer 933a having TFTs and connected to the light emitting element. Since the light emitting element in which a layer containing an organic compound serves as a light emitting layer is weak against moisture and oxygen, immediately after the light emitting element in which a layer containing an organic compound serves as a light emitting layer is formed, using a substrate, a sealing can, and a sealant to seal it.

[0182] Next, a process is performed for partially reducing the adhesion between a first material layer 931 and a second material layer 932. The processing for partially reducing the adhesion is a process in which a laser light is partially radiated on the first material layer or on the second material layer along the perimeter of the region to be peeled, or is a process in which localized pressure is applied from the outside along the perimeter of the region to be peeled to apply damage to a part of the inside or the surface of the second material layer. Specifically, a diamond or other such hard needle may be pressed perpendicularly and moved while applying pressure. Preferably, a scriber device is used and is pressed down by an amount of 0.1 mm to 2 mm, with pressure being applied as it is moved. In this way, before performing the peeling, it is important to create a portion where peeling can occur easily, which serves as a starter. By performing the preprocessing in which the selective (partial) reduction of the adhesion takes place, defective peelings are eliminated and yield is improved.

[0183] Subsequently, a FPC 901 is attached to a terminal electrode provided at the end of an outgoing wiring to which TFT provided on the layer to be peeled 933 is connected.

[0184] Next, a second substrate 935 and the layers to be peeled 933a, 933b are adhered to each other using a first adhesive 934. (See FIG. 11B) A film 902 is adhered to the second substrate 935 with an adhesive 903 in advance. The adhesive 903 is desired to be weaker in adhesive force than that of the first adhesive 934, and it is also desired to be an adhesive which is soluble so as to dissolve in a solvent, or is photosensitive so as to lose adhesiveness when irradiated with light. However, the adhesive 903 is removed in subsequent steps. Further, instead of the first adhesive 934, a tape having adhesive on one or both of its surfaces may be used. The tape may include on one or both of its surfaces an adhesive which is soluble so as to dissolve in a solvent, or is photosensitive so as to lose adhesiveness when irradiated with light.

[0185] The first adhesive 934 may be a reactive-curing type adhesive, a thermal-curing type adhesive, an ultraviolet-curing type adhesive or other such photo-curing type adhesive, or may be an aerophobic-type adhesive, or other various types of curing adhesive. The composition of these adhesives may be, for example, epoxy-type, acrylic-type, silicon-type or anything else. However, since a light emitting element in which a layer containing an organic compound serves as a light emitting layer is weak against moisture and oxygen, a material with high barrier to moisture and oxygen is preferable. The application of such adhesives may be carried out by a coating method, for example. In this embodiment, for the first adhesive 934, used is a thermal-curing type adhesive.

[0186] The second substrate 935 may be constituted by a glass substrate, a quartz substrate, a ceramic substrate, a plastic substrate or the like. Further, it is also possible to use a semiconductor substrate such as a silicon substrate, or a metallic substrate such as a stainless steel substrate.

[0187] The present embodiment employs a highly rigid quartz substrate (thickness: 1.1 mm) for the second substrate 935, having a thickness larger than that of the first substrate 930. In the case where a plastic film is used for the second substrate, when the elements formed onto the first substrate 930 are transferred onto the plastic filmwhich is to say when the layer to be peeled 933 and the film are adhered to each other by the first adhesive 934 and the film is lifted upthere was a risk that the film will bend and cause cracks to form in the layer to be peeled 933. Therefore, after fixing the layer to be peeled 933 formed over the first substrate 930 to the rigid second substrate 935 with the first adhesive 934, the first substrate 930 is peeled. Then, after the plastic film (i.e., the third substrate 937) is fixed to the layer with the second adhesive 936, the second substrate 935 is removed. By following this procedure, it becomes difficult for cracks to occur.

[0188] Next, peeling is performed from the above-mentioned region where the adhesiveness has been reduced, and the first substrate 930 having the first material layer 931 is separated by using a physical means. (See FIG. 11C) Since the second material layer 932 exhibits compressional stress and the first material has tension stress, the separation can be achieved with relatively little force (such as the force of a human hand, or wind pressure of gas blown from a nozzle, or ultrasonic waves, etc.).

[0189] It thus becomes possible to separate the layers to be peeled 933a and 933b formed on the second material layer 932 from the first substrate 930.

[0190] Subsequently, the third substrate 937 and the second material layer 932 (and layers to be peeled 933a, 933b) are bonded together by the second adhesive 936 (FIG. 11D). It is essential that the adhesive 936 has greater adhesive force than that of the adhesive 903.

[0191] The second adhesive 936 may be a reactive-curing type adhesive, a thermal-curing type adhesive, an ultraviolet-curing type adhesive or other such photo-curing type adhesive, or may be an aerophobic-type adhesive, or other various types of curing adhesive. In this embodiment, for the second adhesive 936, used is a thermal-curing type adhesive. Further, in the case where the second adhesive 936 is made of a material which is soluble so as to dissolve in a solvent, or is photosensitive such that it loses adhesive strength when exposed to light, it becomes possible to peel the third substrate at a later step, and it is possible for only the film 902, the first adhesive and the second adhesive to serve as the supports.

[0192] A flexible substrate can be used for the third substrate 937. In this embodiment, the plastic film used for 902 also is used for the third substrate 937.

[0193] After the state shown in FIG. 11D is obtained, the adhesive 903 is irradiated by ultraviolet so that the adhesive force is weaken, therefore, only the second substrate 935 is separated (FIG. 11E). The second substrate 935 is easily peeled by irradiating ultraviolet, the second substrate 935 and the film 902 are separated thereby.

[0194] The steps described above enable the manufacture of a semiconductor device equipped with the layers to be peeled 933a, 933b which serves as a support for the second adhesive 936 and the third substrate 937. Then, by curving the device as shown in FIG. 11F, it thus becomes possible to achieve a semiconductor device in which the curved surface of the semiconductor device exhibits a radius of curvature of from 50 cm to 200 cm. When curving the device, it can be attached to the curved surface to which it is going to be mounted. Note that, between the second adhesive 936 and the layer to be peeled 933a, there is the oxide layer 932 that is the second material layer. In the semiconductor device obtained as described above, the second material layer 932 is applied by a sputtering method and minute amounts of inert gas elements are included in the second material layer 932. Therefore, the semiconductor device as a whole can be made flexible.

[0195] External views of a bent semiconductor device having a light emitting element in which a layer containing an organic compound serves as a light emitting layer obtained by above steps are shown in FIGS. 12A and 12B.

[0196] FIG. 12A and FIG. 12B respond to FIG. 1, the same reference symbols are used for the same portions. A semiconductor shown in FIG. 12A emits light in a direction of an arrow indicated in FIG. 12A, and the device is bent in a bending direction 19. Although not illustrated here, all channel length directions of a large number of semiconductor layers provided on a pixel portion 12 and a driver circuit 17 are aligned with the same direction. In addition, assume that the laser light irradiation direction, that is, a scanning direction is the same direction as the channel length directions. Thus, when the crystal growth direction is aligned with the channel length direction, the field effect mobility can be substantially increased.

[0197] Further, a semiconductor device shown in FIG. 12B emits light in an opposite direction to the one shown in FIG. 12A, and the device is bent in a bending direction 19. Note that the emitting direction can be defined at operator's discretion in accordance with manufacturing methods of a light emitting element in which a layer containing an organic compound serves as a light emitting layer and compositions of pixel circuits.

Embodiment 5

[0198] The present embodiment illustrates an example in which a display having the curved surface obtained by the technique presented in any one of Embodiments 1 through 4 is mounted in a vehicle. Here, an automobile is used as a representative example of a vehicle, but restriction is not made to an automobile. Rather, it goes without saying that the invention may be applied in an aircraft, a train, an electric train, or the like.

[0199] FIG. 13 is a diagram showing the vicinity around a driver's seat in an automobile. A dashboard portion is provided with sound playback systems, specifically including a car audio system and a navigating system. A main unit 2701 of the car audio system includes a display portion 2702 and operating switches 2703 and 2704. By executing the present invention in the display portion 2702, a thin and lightweight car audio system can be achieved. Further, by executing the present invention in the car navigation system, a thin and lightweight car navigation system can be achieved.

[0200] Further, near a steering wheel portion 2602, the dashboard portion 2601 is formed with a display portion 2603 in which digital displays of a speedometer and other such measuring instruments are made. By executing the present invention in the display portion 2702, thin and lightweight mechanical display instruments can be achieved.

[0201] Further, it is also possible to form a display portion 2602 attached onto the dashboard 2601 having a curved surface. By executing the present invention in the display portion 2602, a thin and lightweight mechanical display instrument or image display device can be achieved. Note that the display portion 2602 is curved in the direction shown by the arrows.

[0202] Further, it is also possible to form a display portion 2600 onto the front windshield 2604 that has a curved surface. In the case where the present invention is adapted for the display portion 2600, a permeable material may be used, so that a thin and lightweight mechanical display instrument or image display device can be achieved by means of the present invention. Note that, the display portion 2600 is curved in the direction shown by the arrows. Here, the display portion 2600 was applied in on the windshield, but it may also be provided to other window glass areas.

[0203] It is also possible to form the display portion 2902 attached onto a rear window 2900, for example. FIG. 14 is a diagram showing the vicinity surrounding rear seats in the automobile. Note that, FIG. 14 and FIG. 13 correspond to each other, and since the steering wheel portions are identical, the same reference numerals has been used as in FIG. 13.

[0204] Further, by applying a flexible display device according to the present invention onto the rear window 2900 and mounting onto the exterior of the car a camera which can capture the area behind the car, and then by connecting the display device and the camera, the driver can see places which are obstructed by the car and could not be seen otherwise. Note that, the display portion 2902 is curved in the direction shown by the arrows.

[0205] Further, if the automobile is driven from the right side as shown in FIG. 14, there is a blind spot on the rear-left side since there is a portion of the vehicle body 2906 there (i.e., the part between windows). However, by applying a flexible display device (display portion 2901) according to the present invention onto the part between the windows and mounting onto the exterior of the car a camera which can capture the blind spot, and by connecting the display device and the camera, the driver can check the blind spot. Note that, the display portion 2901 is curved in the direction shown by the arrows.

[0206] Further, it is also possible to provide a display portion 2905 onto a seat 2904. A person sitting in the rear seat can watch television and view the display of the car navigational system.

[0207] Further, although it is not shown in the diagrams, the ceiling of the car may serve as a base, and a display device having a light emitting element in which a layer containing an organic compound serves as a light emitting layer is curved along the curved surface of the ceiling and is attached thereto, whereby image display and illumination inside the vehicle can be performed.

[0208] As described above, the display having the curved surface according to the present invention can be mounted easily onto any curved surface in the vehicle having a radius of curvature of 50 cm to 200 cm.

[0209] Further, the present embodiment illustrated an on-board car audio system and car navigating system, but the present invention may be used on other vehicle display instruments and on free-standing audio and navigational systems.

[0210] Further, the present embodiment may be combined freely with any one of Embodiments 1 through 4.

Embodiment 6

[0211] In Embodiments 1 through 5, the peeling method utilized the film stress (stress deformation) between the two layers to perform the peeling, but restriction is not made to this method. It is possible to use a method in which a separation layer is formed between the layer to be peeled and the substrate, and an etchant is used to separate the separation layer and the substrate, and also a method in which a layer constituted of a amorphous silicon (or a polysilicon) is provided between the layer to be peeled and the substrate, and a laser light is radiated through the substrate to drive out hydrogen contained in the amorphous silicon, thereby creating gaps so as to make the layer to be peeled and the substrate separate, for example.

[0212] Here, FIGS. 15A to 15D show an example in which the amorphous silicon (or the polysilicon) containing a large amount of hydrogen as its separating layer is used and laser light is irradiated onto the separation layer to perform the peeling.

[0213] In FIG. 15A, reference numeral 600 indicates a substrate, reference numeral 601 indicates a separation layer, and reference numeral 602 indicates an layer to be peeled.

[0214] In FIG. 15A, a translucent substrate such as a glass substrate or a quartz substrate or the like is used for the substrate 600.

[0215] Then, the separation layer 601 is formed. Amorphous silicon or polysilicon is used for the separation layer 601. Note that, a sputtering method or a plasma CVD method, or other film application methods may be used as the separation layer 601 so as to put a large amount of hydrogen into it as needed.

[0216] Next, the layer to be peeled 602 is formed onto the separation layer 601. (See FIG. 15A). The layer to be peeled 602 may contain various elements, of which a TFT is a typical example (others include thin film diodes, and photoelectric conversion elements and silicon resistant elements made with silicon PIN-junctions). Further, thermal processing may be performed within the temperature range that the substrate 600 can withstand. However, the separation layer 601 is handled in such a way that peeling off of the film and other problems potentially caused by the thermal processing in the manufacturing of the layer to be peeled 602 do not occur. In the case such as the present embodiment where the laser light is used to perform the peeling, in order that the hydrogen does not escape before the peeling is performed, it is desirable that the thermal processing temperature is set at 410 C. or below when forming the elements included in the layer to be peeled.

[0217] Next, laser light is irradiated through the substrate 600 onto the separation layer. (See FIG. 15B). For the laser light, it is possible to use an excimer laser or other such gas laser, a YAG laser or other such solid-state laser, or a semiconductor laser. Further, the oscillation of the laser light may be continuous oscillation or pulse oscillation, and the shape of the laser beam may be linear or rectangular. In the present embodiment, the laser radiation device indicated in Embodiment 1 is used. By using the laser radiation device indicated in Embodiment 1, it is possible to radiate the laser beam across the entirety of a large surface area with good throughput. Further, the laser radiation device indicated in Embodiment 1 can be used not only for crystallizing or peeling, but also for performing various laser annealing.

[0218] When the above-mentioned laser light causes the release of the hydrogen contained in the separation layer 601, gaps are created and the layer to be peeled 602 and the substrate 600 separate from each other. (FIG. 15C). By using the laser radiation device indicated in Embodiment 1, it becomes possible to peel a layer to be peeled having a large area, across the entire surface thereof with good yield.

[0219] The post-peeling state is shown in FIG. 15D. Further, the example shown here assumes that the mechanical strength of the layer to be peeled 602 is sufficient. However, in the case where the mechanical strength of the layer to be peeled 602 is insufficient, the peeling should be performed after applying a support (not shown in the diagram) for anchoring the layer to be peeled 602.

[0220] Further, the peeled layer after peeling can be curved in a certain direction. It also goes without saying that the peeled layer can be applied and transferred onto an object having a curved surface.

[0221] In the present embodiment as well, the direction of the laser light radiation (i.e., scan direction) and the directions of the channel length of all the semiconductor layers provided to the layer to be peeled are facing in the same direction, and this direction is made perpendicular to the direction of the curvature. Accordingly, the display having the curved surface can be realized.

[0222] Further, the configuration of the present embodiment may be combined freely with Embodiments 1 through 5.

[0223] Note that, in the case where the present embodiment is combined with Embodiment 1, the separation layer 601 of the present embodiment may be used instead of the separation layer 409 of Embodiment 1, and the laser may be radiated from the back side to perform the peeling.

[0224] Similarly, in the case where the present embodiment is combined with Embodiment 2, the separation layer 601 of the present embodiment may be used instead of the separation layer 509 of Embodiment 2, and the laser may be radiated from the back side to perform the peeling.

[0225] According to the present invention, crystallization is performed by radiating a laser beam across the entire surface of a substrate having a broad surface area while directing it at the location of a semiconductor region which forms the TFTs, whereby a crystalline semiconductor layer having a large grain size can be formed and also improvement of the TFT characteristics is attained to realize a display having a curved surface.

[0226] According to the present invention, a display having a curved surface is realized. Thus, in the case where an imaging or measuring display is to be furnished in a limited space such as at the driver's seat in an automobile or aircraft or other such vehicle, the display can be mounted to various locations that have curved surfaces (such as the window, the ceiling, the door, the dashboard, etc.), thereby reducing the space occupied by the display.