JP2009071163A - Semiconductor manufacturing method, semiconductor manufacturing apparatus, and display device - Google Patents

Abstract

The mobility of a microcrystalline silicon film is increased.
A semiconductor manufacturing method for manufacturing at least one of an n-channel thin film transistor and a p-channel thin film transistor, wherein a microcrystalline silicon film 20 is grown in a crystal orientation array of at least (220) using high-density plasma. A first step of forming, and a second step of terminating the microcrystalline silicon film 20 with hydrogen by hydrogen-containing plasma. Accordingly, the microcrystalline silicon film 20 with few dangling bonds can be formed, and mobility can be increased.
[Selection] Figure 6

Description

  The present invention relates to a semiconductor manufacturing method for manufacturing a thin film transistor using a microcrystalline silicon film, a semiconductor manufacturing device for manufacturing a thin film transistor using the method, and a display device incorporating the thin film transistor manufactured by the device.

  A silicon oxide film functioning as an active layer of a thin film transistor includes an amorphous film (a-Si: amorphous silicon film), a polycrystalline film (Poly-Si: polysilicon film), and a microcrystalline film (μc-Si: microcrystal silicon film). ) To form a film. Since the amorphous film does not have a crystal structure, the film is crystallized by performing annealing treatment or laser annealing after the film formation.

  As shown in FIG. 3B, there are dangling bonds of silicon atoms at grain boundaries (grain boundaries) of the grains in the film having a crystal structure. The dangling bonds at the grain boundaries trap carriers and hinder their transportation, which causes a decrease in mobility. The grain boundary dangling bonds also affect the threshold voltage of the device. That is, when the gate voltage is applied, the dangling bond acts as a trap, so the on / off ratio (the current that flows when the gate voltage Vg is 0 V (off current) and the current value that flows when the gate voltage Vg is a predetermined voltage) (Ratio of (on current: saturation current)) becomes small, and a large current is required to turn on / off the current flowing through the circuit. On the other hand, when the gate voltage Vg is set extremely high, the gate insulating film is destroyed.

Therefore, conventionally, a dangling bond at a grain boundary is generated by exciting a gas containing H ₂ gas to generate plasma and reacting hydrogen (H) in the plasma with silicon (Si) in the silicon oxide film. Hydrogen plasma treatment is performed in which hydrogen is terminated with hydrogen (see, for example, Patent Documents 1 and 2). According to this, since the dangling bonds are terminated with hydrogen, it is possible to avoid trapping of carriers including electrons and holes at the grain boundaries of the grains. As a result, the mobility in the silicon oxide film can be increased.

Japanese Patent Laid-Open No. 228042 JP-A-2-232934

  However, in the high-temperature annealing process and thermal oxidation process for crystallization of an amorphous film, the maximum temperature of the process cannot be limited to 600 ° C. or lower, so that a quartz substrate is used to suppress thermal deformation of the substrate. Glass having a low strain point cannot be used.

  On the other hand, there is a method using laser annealing for crystallization of an amorphous film. For example, in laser annealing using a pulse oscillated from an excimer laser or the like, since the irradiation time is several tens of nsec, the amorphous film melts but heat is not transmitted to the glass substrate. In the case of a CW (continuous wave) laser, if a visible region is used like an argon (wavelength 488 nm) laser, only an amorphous film absorbs and melts visible light, and the glass substrate is transmitted and heat is not transmitted. For this reason, cheap glass can be used for the substrate. However, this method increases the number of laser annealing steps and reduces productivity.

  In order to solve the above problems, in the present invention, a microcrystalline silicon film having a crystal orientation array of at least (220) is formed using a large area, high density plasma, and hydrogen plasma treatment is performed on the formed microcrystalline silicon film. The semiconductor manufacturing method which performs this, the semiconductor manufacturing apparatus which manufactures a thin-film transistor using the method, and the display apparatus incorporating the thin-film transistor manufactured by the apparatus are provided.

  That is, in order to solve the above-described problem, according to an aspect of the present invention, there is provided a semiconductor manufacturing method for manufacturing at least one of an n-channel thin film transistor and a p-channel thin film transistor, and at least (220) using high-density plasma. There is provided a semiconductor manufacturing method comprising: a first step of forming a microcrystalline silicon film so that the microcrystalline silicon film is grown in a crystal orientation arrangement; and a second step of terminating the microcrystalline silicon film with hydrogen by hydrogen-containing plasma. Is done.

  As described above, dangling bonds (crystal defects) exist at the grain boundaries of the grains in the microcrystalline silicon film. The dangling bond at the grain boundary traps the carrier and prevents its transport, which causes a decrease in mobility. In addition, crystal defects affect the threshold voltage of the device.

  On the other hand, according to such a configuration, the microcrystalline silicon film is formed so as to grow to at least (220) crystal orientation array using high-density plasma. As shown in FIG. 6, the inventors found that the number of dangling bonds in the (220) orientation plane of the microcrystalline silicon film is one digit less than that in the (111) orientation plane due to ESR (Electron Spin Resonance) intensity. I found out.

  In this phenomenon, columnar crystals are observed in the (111) orientation, whereas in the (220) orientation, the crystal spreads in the lateral direction, and the grain boundary is not so clear, so the crystal grows in the (220) orientation. In the process, it is derived from the estimation that dangling bonds (crystal defects) may be small. According to this, the carrier moving on the (220) orientation plane of the microcrystalline silicon film has a lower probability of being trapped than the carrier moving on the (111) orientation plane. Therefore, it can be seen that the mobility and on / off ratio are higher in the (220) orientation plane of the microcrystalline silicon film than in the (111) orientation plane.

  In addition, in the present invention, hydrogen plasma treatment is performed on the microcrystalline silicon film. According to this, since the dangling bonds are terminated with hydrogen as shown in FIG. 3, carriers are not trapped, and the mobility and the on / off ratio can be further increased. As a result, a thin film transistor capable of high speed processing with low power consumption can be manufactured.

  In addition, since the microcrystalline silicon film does not need to be annealed, the maximum temperature during the process can be made lower than the strain point of the glass. Thereby, it can replace with expensive quartz and can use a comparatively cheap glass substrate. An inexpensive glass substrate can also be used by laser annealing the amorphous film, but the productivity is lowered because the number of annealing processes increases as described above.

  Based on FIG. 9, the inventors have found that there is a correlation between the crystal orientation of the microcrystalline silicon film, the temperature in the vicinity of the substrate during the process, and the hydrogen flow rate. Specifically, the inventors found that when the temperature in the vicinity of the substrate is 250 ° C., the growth ratio of the crystal orientation array (220) to the crystal orientation array (220) is not so great even if the flow rate of the supplied hydrogen gas is increased. Although it did not improve, when the temperature in the vicinity of the substrate was 300 or 350 ° C., it was found that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) with respect to the crystal orientation array (111) increased as the temperature increased. .

  Based on this result, the inventors set the temperature in the vicinity of the substrate to 300 to 300 so that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) to the crystal orientation array (220) is high in the first step. It was concluded that the temperature was preferably set within a range of 350 ° C., and it was concluded that the microcrystalline silicon film can be more effectively grown to the (220) orientation and the mobility can be further increased.

  Further, based on FIG. 10, the inventors have found that the higher the flow rate of hydrogen with respect to the total flow rate of the deposition gas for forming the microcrystalline silicon film (or the flow rate of argon gas at the time of film formation), It has been clarified that the growth ratio of the film to the crystal orientation array (220) with respect to the crystal orientation array (111) increases.

  Based on this result, the inventors have determined that the hydrogen gas is used as the deposition gas so that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) to the crystal orientation array (220) is high in the first step. Found to mix. Thereby, the microcrystalline silicon film can be more effectively grown to the (220) orientation with few defects. As a result, the mobility can be further increased.

  Further, as shown in FIG. 10 as the pretreatment, the inventors set the first crystal growth ratio so that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) with respect to the crystal orientation array (111) increases. It has been found that it is effective to introduce a predetermined gas (raw gas) containing hydrogen gas into the treatment chamber before the step. As a result, the microcrystalline silicon film can be more effectively grown to the (220) orientation, and the mobility can be further increased.

  The second step may be performed after forming the microcrystalline silicon film. Further, the second step may be performed before forming the passivation layer.

Note that high-density plasma refers to plasma with an electron density of 10 ¹¹ cm ⁻³ or higher, and can be generated using microwaves, ICP (Inductively Coupled Plasma), ECR (Electron Cyclotron Resonance), or the like.

  The high-density plasma is generated from the radial line slot antenna of the plasma processing apparatus (hereinafter also referred to as the RLSA plasma processing apparatus) in which a radial line slot antenna (RLSA) 345 shown in FIG. 5 is arranged. It can be generated by exciting a desired gas using the power of the microwave supplied to the inside.

  The high-density plasma is microwaves supplied from the plurality of dielectric plates 31 of the plasma processing apparatus shown in FIG. 13 (hereinafter also referred to as a CMEP (Cellular Microwave Excitation Plasma) plasma processing apparatus). It can also be generated by exciting a desired gas using the power of.

  In particular, in the CMEP plasma processing apparatus, tile-shaped dielectric plates 31 are provided in an array. Each dielectric 31 is supported by beams 26 formed in a lattice shape and fixed to the ceiling surface of the processing vessel. The beam 26 is formed of a nonmagnetic conductive member. The wavelength in the free space of the microwave of 2.45 GHz is about 120 mm. The microwave transmitted through each dielectric plate 31 propagates as a surface wave (traveling wave) between the lower surface of the dielectric 31 and the plasma, and when it reaches the beam 26, it is reflected and becomes a reflected wave. Usually, a standing wave is generated by interference between a traveling wave and a reflected wave. However, in the CMEP plasma processing apparatus, the dielectric plate 31 has a size of about 120 mm × 120 mm, which is at most about one wavelength of a standing wave in both the vertical and horizontal directions. It can be considered that the waves hardly occur. Since the standing wave hinders the stable generation of uniform plasma, according to the CMEP plasma processing apparatus, by providing a large number of dielectric plates 31 in an array at predetermined intervals, the plasma can be made uniform and uniform. As a result, it is possible to accurately perform a precise plasma process on a large-scale object to be processed using uniform plasma.

In the microwave plasma processing apparatus, plasma electron density n _e is lower than the cut-off density n _c is for microwave (electromagnetic wave) propagates through the plasma, the plasma of high electron temperature is generated, plasma The state also becomes unstable. On the other hand, the plasma electron density n _e is higher than the cut-off density n _c, microwaves, become surface waves propagating between the dielectric plate and the plasma. During propagation, part of the microwave is absorbed by the plasma as an evanescent wave and used to maintain the plasma, thereby stably generating a plasma with a low electron temperature.

The principle of the plasma generator, microwave plasma, the plasma electron density n _e is higher than the cut-off density n _c, as compared to the plasma generated by capacitively coupled and inductively coupled plasma processing apparatus higher plasma electron density n _e is, due to low electron temperature T _e, can be prepared in high-speed plasma processing (220) high orientation quality microcrystalline silicon film.

  During the process, the temperature in the vicinity of the substrate may be controlled to 600 ° C. or lower. According to this, a thin film transistor can be formed on a glass substrate that is relatively inexpensive compared to expensive quartz.

  According to the semiconductor manufacturing apparatus for manufacturing a thin film transistor using the semiconductor manufacturing method described above, the microcrystalline silicon film is formed so as to grow at least in the crystal orientation array of (220), and the microcrystal is formed by hydrogen plasma treatment. Grain boundary dangling bonds are terminated with hydrogen. Accordingly, a thin film transistor that can maintain high mobility and an on / off ratio and can perform high-speed processing with low power consumption can be manufactured.

  In particular, the above-described CMEP plasma processing apparatus can generate high-density plasma uniformly over a wide range so that it can cope with the manufacture of a thin film transistor whose area is expected to increase in the future. A thin film transistor having good characteristics and a large area can be manufactured by forming a microcrystalline silicon film using the large-area high-density plasma and then terminating with a hydrogen.

  In addition, by incorporating a thin film transistor manufactured by the semiconductor manufacturing apparatus into a display device, a display device capable of high-speed processing and low power consumption can be commercialized.

  As described above, according to one embodiment of the present invention, mobility of a thin film transistor can be increased.

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, a TFT (thin film transistor) process (semiconductor manufacturing method) according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings. Note that, in the following description and the accompanying drawings, the same reference numerals are given to components having the same configuration and function, and redundant description is omitted. In this specification, at 0 ° C. and 1 atm, 1 sccm is 10 ⁻⁶ / 60 (m ³ / sec) and 1 mTorr is 10 ⁻³ × 101325/760 (Pa).

  In the TFT process of this embodiment, a thin film transistor having a bottom gate structure in which a gate electrode is arranged in the same direction as a glass substrate as viewed from a microcrystalline silicon film formed as an active layer is manufactured. 1 and 2 show a bottom gate TFT process. In the figure, each step of the n-channel TFT process is shown. However, if the impurity to be doped is changed, each step of the p-channel TFT process is performed.

1. Formation of Gate Oxide Film In the bottom gate structure TFT process, first, as shown in FIG. 1A, a doped silicon film doped with phosphorus (P) (low resistance layer (p ⁺ )) is formed on a glass substrate S. A gate oxide (SiO ₂ ) film 10 is formed. The gate oxide film 10 generates plasma by exciting a mixed gas of silane (SiH ₄ ) and nitrogen oxide (N ₂ O) at a normal pressure and a substrate temperature of 400 ° C., and has a thickness of 100 nm. Filmed (atmospheric pressure CVD). The glass substrate S functions as a gate electrode, and the gate oxide film 10 functions as a gate insulating film.

  The formed gate oxide film 10 can be polycrystallized (Poly-Si) by laser annealing in consideration of the strain point of the glass substrate, for example. In addition, when a material having a high strain point such as quartz is used for the substrate, the substrate can be annealed at a high temperature. As an example, the gate oxide film 10 may be annealed at 800 ° C. for 30 seconds in an oxygen atmosphere by instantaneous lamp heating (RTA: Rapid Thermal Annealing).

2. Formation of Microcrystalline Silicon Film Next, as shown in FIG. 1B, a microcrystalline silicon (μc (micro crystal) -Si) film 20 is formed on the gate oxide film 10 by microwave plasma CVD (low pressure CVD (Low-CVD)). (Pressure Chemical Vapor Deposition) At this time, the microwave power density is 6.6 W / cm ² , the pressure in the processing vessel is in the range of 4 to 13.3 Pa, and the substrate temperature is in the range of 250 to 350 ° C. A mixed gas of silane and hydrogen is supplied at a flow rate of 0.48 sccm of silane gas and 10 sccm of hydrogen gas, and argon gas may be mixed as an example of an inert gas. Using plasma generated by exciting a mixed gas with the power of Forming a crystalline silicon film 20 to a thickness of 100 nm.

At this time, the electron density N _e is 3 × 10 ¹² (cm ⁻³ ), high-density plasma of 10 ¹¹ cm ⁻³ or more, and the electron temperature T _e is 1.5 to 2.0 eV. Thus, in the microwave plasma, as compared with capacitive coupling type or inductive coupling type plasma, high plasma electron density n _e is, due to low electron temperature T _e, we produce high quality products at high speed plasma processing can do.

  The microcrystalline silicon film 20 formed using high-density plasma has a small grain size, a low grain boundary barrier, and excellent crystallinity, so that an annealing process and a laser recrystallization process are not required. Therefore, when the microcrystalline silicon film 20 is used for the channel region of the TFT, higher carrier mobility and better operating characteristics can be realized as compared with the case where an amorphous silicon film is used, and the annealing process is omitted. Therefore, throughput can be improved and costs can be reduced. Note that the consideration on the characteristics of the microcrystalline silicon film 20 formed by high-density plasma will be described in detail.

3. Formation of Low Resistance Layer (n ⁺ ) Next, as shown in FIG. 1C, for example, the substrate temperature is set to 300 ° C., a mixed gas of silane and hydrogen is excited to generate plasma, and boron A doped silicon film (low resistance layer (n ⁺ )) 30 doped with (B) is formed to a thickness of 100 nm. The low resistance layer (n ⁺ ) 30 functions as a source region and a drain region.

4). Patterning of Microcrystalline Silicon Film and Low Resistance Layer (n Channel) After the formation of the low resistance layer 30, the microcrystalline silicon film 20 and the low resistance layer 30 are patterned in an island shape as shown in FIG.

5). Formation of Aluminum Wiring Film Next, as shown in FIG. 1E, an aluminum wiring film (Al layer) 40 is formed by sputtering. The aluminum wiring film 40 may be formed by vacuum deposition.

6). Channel Etching Next, as shown in FIG. 2A, the aluminum wiring film 40 and the low resistance layer 30 are etched to form an electrode pattern (channel etching). As a result, a gate electrode (glass substrate S) is formed at a position facing the microcrystalline silicon film 20 across the gate oxide film 10, and the source / drain electrodes 30 s are disposed at positions facing the microcrystalline silicon film 20 adjacent to each other. 30d is formed.

7). Next, as shown in FIG. 2B, after etching the back surface of the substrate S, an Al layer 50 is formed on the back surface of the substrate S by vapor deposition.

8). Hydrogen Plasma Treatment Next, as shown in FIG. 2C, hydrogen plasma treatment is performed on the microcrystalline silicon film 20 and the source / drain electrodes 30s and 30d by microwave plasma CVD. At this time, the pressure is set to 13.3 Pa, the substrate temperature is set to 300 ° C., and a bias voltage V _{pp of} 600 to 700 V is applied to the substrate with a high frequency power (RF) of 400 kHz. In this state, hydrogen plasma having a flow rate of 4000 sccm is excited by a microwave having a power of 1.8 W / cm ² to generate hydrogen plasma.

  Before describing the hydrogen plasma treatment, the state of the microcrystalline film (μc-Si) before the hydrogen plasma treatment will be described while clarifying the difference from the polycrystalline film (Poly-Si). The state of the polycrystalline film and the microcrystalline film before the hydrogen plasma treatment is shown in FIG.

The polycrystal shown in FIG. 14A has a large grain size of silicon grains, that is, grains (grains) in a film having a crystal structure. Since the polycrystalline grains are in a state almost the same as that of the single crystal, the carrier mobility μ (cm ² / Vsec) when moving through the grains is high. On the other hand, the barrier h becomes high at the grain boundary between the grains. Therefore, the carrier mobility μ decreases rapidly at the grain boundary. Accordingly, in the case of polycrystal, the difference between the carrier mobility μ when moving in the grain and the carrier mobility μ when moving in the grain boundary is large, and the operation of the thin film transistor (TFT) becomes unstable. .

  On the other hand, in the microcrystal shown in FIG. 14B, the grain size of grains is small. Therefore, the carrier mobility μ when moving in the grain is lower than that in the case of polycrystal. On the other hand, the grain boundary barrier h is not so high. Therefore, in the case of a microcrystal, the difference between the carrier mobility μ when moving in the grain and the carrier mobility μ when moving in the grain boundary is small, and the operation of the thin film transistor (TFT) is stabilized.

  3A shows the relationship between the mobility of the microcrystalline film before the hydrogen plasma treatment and the barrier (same as FIG. 13B), FIG. 3B shows the state of the microcrystal during the hydrogen plasma treatment, and FIG. (C) shows the relationship between the mobility of the microcrystalline film after the hydrogen plasma treatment and the barrier.

  As described above, the microcrystalline silicon film 20 is formed of grains having a grain size smaller than that of the polycrystalline film, and dangling bonds exist in the grain boundary. The grains are generally covalent bonds and there are few dangling bonds. In the grain boundary, there is a bond (dangling bond) in which the silicon atom loses the covalent bond partner and is occupied by electrons (unpaired electrons) that are not involved in the bond.

Grain boundary dangling bonds trap carriers made up of electrons and holes and hinder their transport, which reduces mobility. Therefore, in the present embodiment, as shown in FIG. 3B, a dangling bond existing in the grain boundary of the microcrystalline silicon film 20 is terminated with hydrogen by performing a hydrogen plasma process. Specifically, hydrogen ions (H ⁺ ) in the hydrogen plasma are combined with dangling bonds present at the grain boundary during acceleration toward the substrate by a bias voltage applied to the substrate. Thereby, defects at the grain interface are repaired and the characteristics of the microcrystalline silicon film 20 are improved.

  After the hydrogen plasma treatment, the grain boundary barrier h shown in FIG. 3C is the barrier of the grain boundary before the hydrogen plasma treatment shown in FIG. 3A because the dangling bonds are terminated by hydrogen. It is lower than h. As described above, according to the hydrogen plasma treatment, when the carrier moves over the grain boundary barrier h, the height of the barrier h is lowered, and the carrier mobility μ and the grain boundary when moving through the grain are reduced. The difference between the carrier mobility μ and the carrier mobility μ can be reduced. As a result, the average mobility of the entire microcrystalline silicon film 20 can be increased, and the operation of the thin film transistor (TFT) is extremely stabilized by reducing the difference in carrier mobility μ in the grain and the grain boundary. be able to. Note that the hydrogen plasma treatment may use microwave plasma or parallel plate remote plasma.

The inventors set the microwave power to 2000 W, the processing chamber pressure to 13.3 Pa, the substrate temperature to 300 ° C., and the TFT width / length to 20/5 μm. Under these conditions, the hydrogen plasma treatment was performed for 5 minutes. The change of the mobility μ (cm ² / Vsec) when applied was verified. According to this experiment, as shown in FIG. 11, the mobility is improved when the power of the bias voltage is 15 to 35 W, and in particular, the mobility is dramatically improved when the power of the bias voltage is 20 to 30 W. I found out that

9. Passivation formation / annealing Finally, in order to protect the TFTs laminated on the glass substrate S as described above, as shown in FIG. 2D, an insulating film such as a SiN film is formed on the passivation layer by plasma CVD. 60. The formed passivation layer 60 is heat-treated for 2 hours while flowing argon gas under conditions of a pressure of 40 Pa and a substrate temperature of 450 ° C.

  Note that the hydrogen plasma treatment may be performed while the passivation layer 60 is heat-treated. In this case, hydrogen diffuses inside the passivation layer 60 and reaches the channel region (30s, 30d, etc.), and dangling bonds in the channel region are terminated with hydrogen.

(Substrate processing system)
Next, a substrate processing system that performs the series of processes shown in FIGS. 1 and 2 will be described with reference to FIG. The substrate processing system 100 according to the present embodiment is a cluster type apparatus having a plurality of processing apparatuses.

(Substrate processing system 100)
The substrate processing system 100 includes a load lock chamber LLM, a transfer chamber TM (Transfer Module), a cleaning chamber CM (Cleaning Module), and four process modules PM (Process Module) 1 to 4.

  The load lock chamber LLM has a predetermined depressurization inside in order to transfer the glass substrate S transferred from the atmospheric system to the process module PM maintained at a predetermined vacuum degree via the transfer chamber TM in a reduced pressure state. This is a vacuum transfer chamber held in a state. The transfer chamber TM is provided with a multi-joint transfer arm Arm that can bend and stretch. The substrate processing system 100 first transports the glass substrate S from the load lock chamber LLM to the cleaning chamber CM using the transport arm Arm, cleans the substrate surface, transports it to the process module PM2, and performs gate oxide film by APCVD. After forming the film 10, it is transferred to PM3 and a microcrystalline silicon film 20 is formed by microwave plasma CVD.

  FIG. 5 schematically shows a longitudinal section of an RLSA plasma CVD apparatus (PM3) that performs microwave plasma processing. The RLSA plasma CVD apparatus has a cylindrical processing container 300 having an open ceiling surface. A shower plate 305 is fitted into the opening on the ceiling surface. The processing container 300 and the shower plate 305 are hermetically sealed by an O-ring 310 disposed between the step portion of the inner wall of the processing container 300 and the outer peripheral portion of the lower surface of the shower plate 305, whereby a processing chamber in which plasma processing is performed. U is formed. For example, the processing container 300 is made of a metal such as aluminum, and the shower plate 305 is made of a metal such as aluminum or a dielectric, and is electrically grounded.

A susceptor (mounting table) 315 on which the wafer W is mounted is installed on the bottom of the processing container 300 via an insulator 320. The susceptor 315 is connected to a high frequency power source 325b through the matching unit 325a, adapted to apply a predetermined bias voltage V _pp by high-frequency power output from the high-frequency power source 325b into the processing chamber 300 . The susceptor 315 is connected to a high voltage DC power supply 330b via a coil 330a, and the substrate S is electrostatically attracted by a DC voltage output from the high voltage DC power supply 330b. A cooling jacket 335 that supplies cooling water to cool the wafer W is provided inside the susceptor 315.

The shower plate 305 is covered with a cover plate 340 at the top thereof. A radial line slot antenna 345 is provided on the upper surface of the cover plate 340. The radial line slot antenna 345 is provided between a slot plate 345a on a disk in which a number of slots (not shown) are formed, an antenna body 345b on the disk holding the slot plate 345, a slot plate 345a, and an antenna body 345b. And a slow phase plate 345c formed of a dielectric such as alumina (Al ₂ O ₃ ). In the radial line slot antenna 345, a microwave generator 355 is installed outside via a coaxial waveguide 350.

  A vacuum pump (not shown) is attached to the processing container 300, and the processing chamber U is decompressed to a desired degree of vacuum by discharging the gas in the processing container 300 through the gas discharge pipe 360. It has become.

The gas supply source 365 includes a plurality of valves V, a plurality of mass flow controllers MFC, a hydrogen (H ₂ ) gas supply source 365a, an argon (Ar) gas supply source 365b, and a silane (SiH ₄ ) gas supply source 365c. . The gas supply source 365 supplies a gas having a desired concentration to the inside of the processing container 300 by controlling the opening / closing of each valve V and the opening of each mass flow controller MFC.

  Hydrogen gas and argon gas are supplied to the upper side of the processing chamber U from a gas introduction pipe 375 that passes through the shower plate 305 through the first flow path 370a, and silane gas is integrated through the second flow path 370b. The gas pipe 380 is supplied below the first gas. According to such a configuration, various gases are excited by the power of the microwave incident from the microwave generator 355 through the slot and the shower plate 305 into the processing chamber U to generate plasma. A microcrystalline silicon film 20 is formed by the generated plasma.

  After the microcrystalline silicon film 20 is formed, the low resistance layer 30 is formed by further supplying hydrogen gas and silane gas while doping boron (B) in the process module PM3.

  After forming the microcrystalline silicon film 20 in this way, the substrate processing system 100 transfers the substrate S to the process module PM1 shown in FIG. 4, and uses the desired pattern to form the microcrystalline silicon film 20 and the low resistance layer 30. Is etched into an island shape.

  Next, the substrate processing system 100 transports the substrate S to the process module PM4, forms the Al layer 40 by sputtering, transports the substrate S again to the process module PM1, and etches the Al layer 40 and the low resistance layer 30. To expose the channel portion. Further, the back surface of the substrate is etched by the module PM1, and an Al film 50 is formed on the back surface of the substrate by sputtering with a vapor deposition apparatus (not shown) or the process module PM4.

  Again, the substrate processing system 100 transports the substrate S to the process module PM3, generates hydrogen plasma by exciting the hydrogen gas supplied from the hydrogen gas supply source 365a with the power of the microwave, and generates hydrogen ions in the plasma. By accelerating toward the substrate by a bias voltage and terminating the grain boundary dangling bonds with hydrogen ions, the characteristics of the interface of the microcrystalline silicon film 20 are improved. Finally, a passivation layer 60 is formed in the module PM3, and the passivation layer 60 is annealed in a heat treatment chamber (not shown).

(Orientation of microcrystalline silicon film)
Among the steps described above, since the carriers (electrons or holes) move near the interface of the microcrystalline silicon film 20, the characteristics of the film are very important. Therefore, as described above, in the semiconductor manufacturing method according to the present embodiment, hydrogen is terminated to the grain boundary dangling bonds by hydrogen plasma treatment to improve the film characteristics.

  In addition, in the semiconductor manufacturing method according to the present embodiment, when the microcrystalline silicon film is formed, the process is controlled so that the crystal orientation of the microcrystalline silicon film easily grows to the (220) orientation.

  As described above, in the dangling bond, the silicon atom has lost the covalent bond partner. In other words, a dangling bond has a bond occupied by unpaired electrons. Therefore, by measuring the state of unpaired electrons at the interface of the microcrystalline silicon film 20 by ESR (Electron Spin Resonance), the state of dangling bonds at the grain boundaries of the grains can be clarified.

The results of ESR are shown in FIG. From this result, the density of dangling is about 6 × 10 ¹⁷ cm ⁻³ in the (111) orientation and about 7 × 10 ¹⁶ cm ⁻³ in the (220) orientation, and the (220) orientation has the (111) orientation. About an order of magnitude smaller than the orientation. This indicates that the (220) orientation has fewer dangling bonds than the (111) orientation. The reason for this is that columnar crystals are seen in the (111) orientation, whereas in the (220) orientation, the crystal spreads in the lateral direction, and the grain boundary is not very clear, so the crystal grows in the (220) orientation. In the process, it is estimated that there are few dangling bonds (crystal defects).

  Note that the microcrystalline silicon film 20 grown in the (111) orientation was formed by microwave plasma under high flow rate conditions: silane gas 3 sccm, hydrogen gas 35 sccm, pressure 13.3 Pa, substrate temperature 350 ° C. . On the other hand, the microcrystalline silicon film 20 grown in the (220) orientation was formed by microwave plasma under the conditions of silane gas 0.48 sccm, hydrogen gas 3 sccm, pressure 4 Pa, and substrate temperature 350 ° C. as low flow rate conditions. .

  The inventors created a TFT using this film formation condition and performed a hydrogen plasma treatment. Based on the results, the inventors first examined the film thickness dependence of the crystal structure of the microcrystalline silicon film 20. In FIG. 7, the horizontal axis represents the thickness (nm) of the microcrystalline silicon film 20, and the left vertical axis represents an X-ray diffraction (XRD) peak intensity ratio I (220) / I (111). And the right vertical axis represents RMS (Root Mean Square). RMS means irregularities on the film surface.

  The solid line in the graph indicates the film thickness dependence of the peak intensity ratio I (220) / I (111) based on the XRD diffraction results. For example, according to an example of the measurement result of the XRD intensity shown in FIG. 8, the peak intensity ratio I (220) / I (111) ≈1.14. On the other hand, the broken line in the graph is RMS. According to these results, the film can be stably grown in the direction in which the growth ratio from the film thickness of about 300 nm or more to the (220) orientation increases.

(Temperature dependence and hydrogen flow rate dependence of microcrystalline film)
Next, the inventors conducted experiments on the temperature dependency and hydrogen flow rate dependency of the microcrystalline silicon film 20. In this experiment, the flow rate of silane gas was set to 3 sccm, the total flow rate of hydrogen gas and argon gas was set to 35 sccm (total gas flow rate 38 sccm), the substrate temperature was set to three types of 250 ° C., 300 ° C., and 350 ° C. The flow rate of hydrogen gas was changed with respect to the total flow rate of the membrane gas. The result is shown in FIG.

  This shows that there is a correlation between the orientation in which the microcrystalline silicon film grows, the temperature in the vicinity of the substrate, and the flow rate of hydrogen. Specifically, when the substrate temperature was 250 ° C., the growth ratio from the (111) orientation to the (220) orientation did not improve so much. On the other hand, when the substrate temperature is 300 ° C., the growth ratio of the crystal orientation array (111) to the crystal orientation array (220) is higher than that when the substrate temperature is 250 ° C. Further, the substrate temperature is 350 ° C. In this case, the substrate temperature was higher than that in the case where the substrate temperature was 300 ° C. As a result, the inventors have found that the growth ratio of the crystal orientation array (220) to the crystal orientation array (220) can be increased by setting the substrate temperature in the range of 300 to 350 ° C.

  Further, it was found that at the same temperature, the higher the flow rate ratio of hydrogen gas to the total gas flow rate, the higher the growth ratio of the crystal orientation array (111) to the crystal orientation array (220). As a result, the inventors set the temperature in the vicinity of the substrate within a range of 300 to 350 ° C., and further increased the flow rate ratio of hydrogen gas to the total flow rate of the gas, thereby increasing the crystallites grown to (220) orientation. It has been found that the silicon film 20 can be formed.

(Dependence on hydrogen flow rate of microcrystalline film)
Next, the inventors conducted another experiment on the hydrogen flow rate dependency of the microcrystalline silicon film 20. In this experiment, as shown in FIG. 10, the flow rate of silane gas was set to 0.48 sccm, and the total flow rate of hydrogen gas and argon gas was changed to two patterns of 10 sccm (curve G) and 35 sccm (curve H). .

  As a result, when the microcrystalline silicon film 20 is formed, the inventors increase the content of hydrogen gas with respect to the total gas flow rate, so that the crystal with respect to the crystal orientation array (111) of the formed microcrystalline silicon film increases. It has been found that the growth ratio to the orientation array (220) increases.

Further, before the plasma irradiation, a raw gas composed of argon gas and hydrogen gas is introduced into the processing chamber (Ar / H ₂ pre-treatment), and then the above-mentioned various conditions are controlled while the total flow rate of hydrogen gas and argon gas is controlled to 10 sccm. When a microcrystalline silicon film was formed by plasma while introducing a gas, a microcrystalline silicon film having a very high growth ratio to the (220) orientation could be formed.

  Therefore, the inventor mixes hydrogen gas with the deposition gas when forming the microcrystalline silicon film 20, thereby increasing the growth ratio of the microcrystalline silicon film 20 to the crystal orientation array (220) with respect to the crystal orientation array (111). Furthermore, it is possible to form a microcrystalline silicon film with very few dangling bonds by flowing a raw gas containing hydrogen into the treatment chamber (pretreatment) before forming the microcrystalline film. Elucidated.

(Hydrogen plasma treatment of microcrystalline silicon film, RF bias dependence of microcrystalline silicon film)
As described above, the inventors, as shown in FIG. 11, have a microwave power of 2000 W, a processing chamber pressure of 13.3 Pa, a substrate temperature of 300 ° C., and a TFT width / length (W / L). ) Was set to 20/5 μm, and mobility μ (cm ² / Vsec) when hydrogen plasma treatment was performed for 5 minutes under these conditions was verified. As a result, when the microcrystalline silicon film 20 is formed, the inventors set the RF bias power in the range of 15 to 35 W, so that the mobility μ is increased with the bias voltage V _pp of about 500 to 800 V. It has been found that the mobility μ can be further improved by setting the RF bias power in the range of 20 to 30 W.

  FIG. 12 shows operating characteristics of the thin film transistor after the hydrogen plasma treatment. FIG. 12A shows mobility μ and drain current with respect to the gate voltage Vg of a thin film transistor not subjected to hydrogen plasma treatment, and FIG. 12B shows mobility μ and drain current when hydrogen plasma treatment is performed. . Note that the width / length of the TFT is 20/4 μm and the drain voltage is 0.1V.

According to this, a large difference was observed in the mobility of the microcrystalline silicon film 20. Specifically, when the hydrogen plasma treatment is not performed, the maximum value of the mobility μ of the microcrystalline silicon film 20 is less than 0.001 cm ² / Vsec, whereas when the hydrogen plasma treatment is performed. It was possible to achieve 1.4 cm ² / Vsec.

  In addition, a significant difference was caused in the change of the drain current Id with respect to the gate voltage Vg. In other words, the on / off ratio when the hydrogen plasma treatment was not performed was about two digits, whereas the on / off ratio when the hydrogen plasma treatment was performed was able to achieve five digits or more.

  As described above, the inventor forms the microcrystalline silicon film 20 by optimizing the process conditions so as to grow to at least (220) crystal orientation array using high-density plasma. The microcrystalline silicon film 20 was further terminated with hydrogen, thereby succeeding in forming a microcrystalline silicon film having very few defects, high mobility, and high on / off characteristics.

  In a microcrystalline silicon film having a diamond crystal structure, more atoms exist in the (111) orientation plane than in the (220) orientation plane. Therefore, in the case of a microcrystalline silicon film, the (111) orientation plane with higher atomic density is more likely to have a dangling bond because the silicon atom loses the covalent bond partner than the (220) orientation plane. Conceivable. Therefore, by controlling the formation process of the microcrystalline silicon film so that the film grows in the (220) orientation, a channel layer with less dangling bonds and high mobility and on / off ratio is formed. Can do.

  Further, by performing hydrogen plasma treatment on the formed microcrystalline silicon film 20, dangling bonds at grain boundaries of the grains constituting the microcrystalline silicon film 20 are terminated with hydrogen. As a result, mobility and on / off ratio can be further increased, and high-speed processing can be performed with low power consumption, and a thin transistor can be manufactured.

(Other configurations of microwave plasma processing equipment)
In addition, the semiconductor manufacturing apparatus which manufactures a thin-film transistor using the semiconductor manufacturing method concerning this embodiment is not restricted to the RLSA plasma apparatus of FIG. 5, For example, as shown in FIG. 13, the desired vacuum degree was maintained. A plasma processing apparatus for performing plasma processing on a substrate S placed on a susceptor 505 inside a processing container 500, wherein a plurality of tile-shaped dielectric plates 510 are evenly placed on the ceiling surface of the processing container 500. A CMEP (Cellular Microwave Excitation Plasma) plasma processing apparatus may be used.

  The microwave plasma processing apparatus includes a processing container 500 and a lid 510. The processing vessel 500 has a bottomed cubic shape with an upper portion opened. The processing container 500 and the lid 510 are sealed by an O-ring 32 disposed between the lower surface outer peripheral portion of the lid main body 21 and the upper surface outer peripheral portion of the processing container 500, whereby the processing chamber U in which plasma processing is performed. Is formed. The processing container 500 and the lid body 510 are made of a metal such as aluminum and are electrically grounded.

  The processing container 500 is provided with a susceptor 11 (mounting table) for mounting the substrate G therein. The susceptor 11 is made of, for example, aluminum nitride, and a power feeding unit 11a and a heater 11b are provided therein.

  A high frequency power source 12b is connected to the power supply unit 11a via a matching unit 12a (for example, a capacitor). In addition, a high-voltage DC power supply 13b is connected to the power supply unit 11a via a coil 13a. The matching unit 12a, the high frequency power source 12b, the coil 13a, and the high voltage DC power source 13b are provided outside the processing container 500. The high frequency power supply 12b and the high voltage DC power supply 13b are grounded.

  The power feeding unit 11a applies a predetermined bias voltage to the inside of the processing container 500 by the high frequency power output from the high frequency power source 12b. The power feeding unit 11a is configured to electrostatically attract the substrate G with a DC voltage output from the high-voltage DC power supply 13b.

  An AC power supply 14 provided outside the processing container 500 is connected to the heater 11b, and the substrate G is held at a predetermined temperature by the AC voltage output from the AC power supply 14.

  The bottom surface of the processing container 500 is opened in a cylindrical shape, and one end of the bellows 15 is attached to the outer peripheral edge thereof. The other end of the bellows 15 is fixed to the elevating plate 16. In this way, the opening at the bottom of the processing container 500 is sealed by the bellows 15 and the lifting plate 16.

  The susceptor 11 is supported by a cylinder 17 disposed on the elevating plate 16, and moves up and down integrally with the elevating plate 16 and the cylinder 17 so that the susceptor 11 has a height corresponding to the processing process. It is supposed to adjust to. A baffle plate 18 for controlling the gas flow in the processing chamber U to a preferable state is provided around the susceptor 11.

  A vacuum pump (not shown) provided outside the processing container 500 is provided at the bottom of the processing container 500. The vacuum pump discharges the gas in the processing container 500 through the gas discharge pipe 19 to reduce the pressure in the processing chamber U to a desired degree of vacuum.

The lid 510 is provided with a lid body 21, six rectangular waveguides 33, a slot antenna 38, and a dielectric (consisting of a plurality of dielectric plates 31). The six rectangular waveguides 33 have a rectangular cross-sectional shape and are arranged in parallel inside the lid body 21. The inside of each rectangular waveguide 33 is filled with a dielectric member 34 such as a fluororesin (for example, Teflon (registered trademark)), alumina (Al ₂ O ₃ ), quartz, etc., and λg ₁ = The in-tube wavelength λg ₁ of each rectangular waveguide 33 is controlled according to the equation of λc / (ε ₁ ) ^1/2 . Here, λc is the wavelength of free space, and ε ₁ is the dielectric constant of the dielectric member 34.

  Each rectangular waveguide 33 is opened at the top, and a movable portion 35 is inserted into the opening so as to be movable up and down. The movable portion 35 is made of a conductive material that is a non-magnetic material such as aluminum.

  An elevating mechanism 36 is provided outside the lid main body 21 and on the upper surface of each movable portion 35 to move the movable portion 35 up and down. With this configuration, the height of the rectangular waveguide 33 can be arbitrarily changed by moving the movable portion 35 up and down up to the upper surface of the dielectric member 34.

The slot antenna 38 is formed integrally with the lid body 21 below the lid body 21. The slot antenna 38 is made of a metal that is a nonmagnetic material such as aluminum. The slot antenna 38 is provided with the thirteen slots 37 (openings) shown in FIG. 2 arranged in series on the lower surface of each rectangular waveguide 33. Each slot 37 is filled with a dielectric member such as fluororesin, alumina (Al ₂ O ₃ ), quartz, and the like, and according to the equation: λg ₂ = λc / (ε ₂ ) ^1/2 The guide wavelength λg ₂ of each slot 37 is controlled. Here, λc is the wavelength of free space, and ε ₂ is the dielectric constant of the dielectric member inside the slot 37.

  Each dielectric plate 31 is attached so as to straddle two slots among 26 (= 13 × 2 rows) slots 37 provided on the lower surface of two rectangular waveguides 33 adjacent to each other. ing. With the above configuration, the plurality of dielectric plates 31 formed in a tile shape are attached to the lower surface of the slot antenna 38 in an array at equal intervals.

Each dielectric plate 31 is formed using a dielectric material such as quartz glass, AlN, Al ₂ O ₃ , sapphire, SiN, or ceramics. As shown in FIG. 13, each dielectric plate 31 is formed with irregularities on the surface facing the substrate G. Thus, by providing at least one of the concave portion or the convex portion on each dielectric plate 31, the loss of electric field energy when the surface wave propagates on the surface of each dielectric plate 31 is increased. Surface wave propagation can be suppressed. As a result, generation of standing waves can be suppressed and uniform plasma can be generated. The number of slots 37 formed on the lower surface of each rectangular waveguide 33 is arbitrary.

  On the lower surface of the slot antenna 38, a beam 26 formed in a lattice shape is provided to support the plurality of dielectric plates 31. Each dielectric plate 31 is supported by the beam 26 at the periphery thereof so that a step is provided between the dielectric plate 31 and the beam 26 (beams 26a to 26d). That is, the beam 26 is provided so that the beam 26 protrudes toward the substrate G at the periphery of each dielectric plate 31. The beam 26 is formed of a conductive material that is a non-magnetic material such as aluminum.

  A plurality of supports 27 (supports 27 a to 27 d) are provided on a part of the lower surface of the beam 26. Both ends of each gas pipe 28 (for example, a part as a unit constituting the lower gas shower head) are supported by a support 27. The gas pipe 28 is made of a dielectric material such as alumina.

  A cooling water supply source 45 disposed outside the microwave plasma processing apparatus 100 is connected to the cooling water pipe 44, and the cooling water supplied from the cooling water supply source 45 circulates in the cooling water pipe 44. By returning to the cooling water supply source 45, the lid body 21 is kept at a desired temperature.

  With the configuration described above, for example, 2.45 GHz × 3 microwave output from the microwave generator 40 propagates through each rectangular waveguide 33, passes through each slot 37, and passes through each dielectric plate 31. So as to enter the processing chamber U.

  The gas supply source 43 includes a plurality of valves (valves 43a1, 43a3, 43b1, 43b3, 43b5, 43b7), a plurality of mass flow controllers (mass flow controllers 43a2, 43b2, 43b6), an oxygen gas supply source 43a4, a silane gas supply source 43b4, and argon. It consists of a gas supply source 43b8.

  The gas supply source 43 supplies a desired concentration of hydrogen gas, silane gas, and argon gas into the processing vessel 500 by controlling the opening and closing of the valves V and the opening of the mass flow controllers MFC, respectively. Yes.

  The gas introduction pipe 29 (gas introduction pipes 29 a to 29 d) penetrates the beam 26. A hydrogen gas supply source 43a4 is connected to the gas introduction pipes 29a and 29c via the first flow path 42a. Further, a silane gas supply source 43b4 and an argon gas supply source 43b8 are connected to the gas introduction pipes 29b and 29d via the second flow path 42b.

  The hydrogen gas is introduced into the space between each dielectric plate 31 and each gas pipe 28 through the gas introduction pipes 29a and 29c, for example, and is converted into plasma by microwave electric field energy. On the other hand, the mixed gas of silane gas and argon gas is introduced to the substrate G side on the susceptor 11 from the gas supply holes provided in each gas pipe 28 through the gas introduction pipes 29b and 29d. A microcrystalline silicon film 20 is formed by the plasma thus generated.

  In particular, in the CMEP plasma processing apparatus, tile-shaped dielectric plates 31 are provided in an array. Each dielectric 31 is supported by beams 26 formed in a lattice shape and fixed to the ceiling surface of the processing vessel. The beam 26 is formed of a nonmagnetic conductive member. The wavelength in the free space of the microwave of 2.45 GHz is about 120 mm. The microwave transmitted through each dielectric plate 31 propagates as a surface wave (traveling wave) between the lower surface of the dielectric 31 and the plasma, and when it reaches the beam 26, it is reflected and becomes a reflected wave. Usually, a standing wave is generated by interference between a traveling wave and a reflected wave. However, in the CMEP plasma processing apparatus, the dielectric plate 31 has a size of about 120 mm × 120 mm, which is at most about one wavelength of a standing wave in both the vertical and horizontal directions. It can be considered that there is almost no standing wave. Since the standing wave hinders stable generation of uniform plasma, according to the CMEP plasma processing apparatus, by providing a large number of dielectric plates 31 in an array at predetermined intervals, uniform and stable A precise microcrystalline silicon film 20 can be formed on a glass substrate having a large area by using the plasma generated in step (b).

  According to the semiconductor manufacturing apparatus that manufactures a thin film transistor using the semiconductor manufacturing method described above, a thin film transistor that maintains high mobility and on / off ratio at the interface of the channel layer, enables high-speed processing, and has low power consumption. Can be manufactured. The semiconductor manufacturing method according to the present embodiment includes a method for manufacturing a semiconductor formed on a silicon wafer and a method for manufacturing a semiconductor formed on a flat panel display (FPD: Flat Panel Display).

  In addition, by using a microcrystalline silicon film for the channel layer, an annealing process is unnecessary, and thus a thin film transistor can be formed over an inexpensive glass substrate by keeping the temperature during the process at 600 ° C. or lower.

  In the above embodiment, the operations of the respective units are related to each other, and can be replaced as a series of operations in consideration of the relationship between each other. And by replacing in this way, an embodiment of a semiconductor manufacturing method for manufacturing a thin film transistor can be an embodiment of a semiconductor manufacturing apparatus for manufacturing a thin film transistor using the semiconductor manufacturing method.

  In the above embodiment, the microcrystalline silicon film is formed by plasma CVD using an RLSA plasma processing apparatus or a CMEP plasma processing apparatus. However, the microcrystalline silicon film can also be formed by a CVD process or a sputtering process using high-density plasma generated using a capacitively coupled (parallel plate type) plasma processing apparatus or an inductively coupled plasma processing apparatus.

  In addition, by incorporating a thin film transistor manufactured by the semiconductor manufacturing apparatus into a display device, a display device capable of high-speed processing and low power consumption can be commercialized. Examples of the display device include an organic EL (Electroluminescence) display, a plasma display, and a liquid crystal display (LCD).

  The size of the glass substrate processed by the semiconductor manufacturing apparatus is 730 mm × 920 mm or more. For example, the above-mentioned semiconductor manufacturing apparatus continuously has a G4.5 substrate size of 730 mm × 920 mm (dimension in the chamber, 1000 mm × 1190 mm) or a substrate of 1100 mm × 1300 mm (dimension in the chamber, 1470 mm × 1590 mm) or larger than the G5 substrate size. A film forming process can be performed.

  The object to be processed by the semiconductor manufacturing apparatus is not limited to a glass substrate, and may be a silicon wafer having a diameter of 200 mm or 300 mm.

  As mentioned above, although preferred embodiment of this invention was described referring an accompanying drawing, it cannot be overemphasized that this invention is not limited to the example which concerns. It will be apparent to those skilled in the art that various changes and modifications can be made within the scope of the claims, and these are naturally within the technical scope of the present invention. Understood.

  For example, in the above embodiment, a manufacturing process of a thin transistor having a bottom gate structure is given as an example. It can also be used in the manufacturing method.

  Moreover, the board | substrate used for this invention should just be a board | substrate used for an organic electroluminescent display, a plasma display, a liquid crystal display etc., for example, The semiconductor manufacturing apparatus concerning this invention is the above-mentioned manufacturing method to such a board | substrate. Any device capable of forming a thin film transistor may be used.

It is sectional drawing of the device which showed the process of the semiconductor manufacturing method concerning one Embodiment of this invention. FIG. 2 is a cross-sectional view of the device showing the process following FIG. 1 of the semiconductor manufacturing method according to the same embodiment. It is a figure for demonstrating the relationship between a microcrystal structure, a hydrogen plasma process, and a mobility. It is the figure which showed an example of the substrate processing system concerning the embodiment. It is a longitudinal cross-sectional view of a RLSA plasma processing apparatus. It is the figure which showed the result of ESR of the microcrystal silicon film in each process condition. It is the figure which showed the film thickness dependence in each process condition. It is a figure showing XRD intensity of a microcrystal silicon film. It is the figure which showed the temperature dependence and hydrogen flow rate dependence of a microcrystal silicon film. It is another figure which showed the hydrogen flow rate dependence of the microcrystal silicon film. It is the figure which showed the power dependence of the bias voltage of a microcrystal silicon film. 12A shows the mobility and on / off ratio (drain current) when hydrogen plasma treatment is not performed, and FIG. 12B shows the mobility and on / off ratio (drain current) when hydrogen plasma treatment is performed. It is a longitudinal cross-sectional view of a CMEP plasma processing apparatus. It is a figure for demonstrating the relationship between a polycrystalline-film structure, a microcrystal film structure, and a mobility.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Gate oxide film 20 Microcrystalline silicon film 30 Low resistance layer 40 Aluminum wiring film 50 Back surface Al layer 60 Passivation layer 100 Substrate processing system PM1, PM2, PM3, PM4 Process module S substrate

Claims (12)

A semiconductor manufacturing method for manufacturing at least one of an n-channel thin film transistor and a p-channel thin film transistor,
A first step of forming a microcrystalline silicon film to grow to at least (220) crystal orientation using high-density plasma;
And a second step of terminating the microcrystalline silicon film with hydrogen by hydrogen-containing plasma.   The high-density plasma is generated by exciting a desired gas using a microwave power supplied from the radial line slot antenna of a plasma processing apparatus in which a radial line slot antenna is disposed. The semiconductor manufacturing method described in 1.   The high-density plasma excites a desired gas by using microwave power supplied from a plurality of dielectric plates of a plasma processing apparatus in which a plurality of tile-shaped dielectric plates are arranged in an array. The semiconductor manufacturing method according to claim 1, which is generated by:   In the first step, the temperature in the vicinity of the object to be processed is within a range of 300 to 350 ° C. so that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) with respect to the crystal orientation array (111) increases. The semiconductor manufacturing method according to claim 1, wherein the semiconductor manufacturing method is set.   5. The hydrogen gas is mixed with the deposition gas so that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) with respect to the crystal orientation array (220) is increased in the first step. The semiconductor manufacturing method described in any one.   The semiconductor manufacturing method according to claim 1, wherein the second step is performed after the microcrystalline silicon film is formed.   The semiconductor manufacturing method according to claim 1, wherein the second step is performed before forming the passivation layer.   The high-density plasma is generated by exciting a desired gas using the power of a microwave transmitted through a plurality of dielectric plates formed in a tile shape and supplied into a processing chamber. The semiconductor manufacturing method described in any one of.   The semiconductor manufacturing method according to claim 1, wherein the temperature in the vicinity of the object to be processed is controlled to 600 ° C. or lower during the process.   A predetermined gas containing hydrogen gas is introduced into the processing chamber before the first step so that the growth ratio of the microcrystalline silicon film to the crystal orientation array (220) with respect to the crystal orientation array (111) increases. The semiconductor manufacturing method described in any one of 1-9.   The semiconductor manufacturing apparatus which manufactures a thin-film transistor using the semiconductor manufacturing method described in any one of Claims 1-10.   A display device incorporating a thin film transistor manufactured by the semiconductor manufacturing apparatus according to claim 11.

Images