JP4660743B1 - Thin film transistor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a thin film transistor having a metal silicon compound thin film as a channel region.
A transition metal-silicon compound, in which a transition metal-encapsulating silicon cluster surrounded by 7 to 16 silicon atoms surrounds a transition metal atom as a unit structure, and includes first and second transition metal atoms. A thin film transistor characterized in that a metal silicon compound thin film in which silicon is arranged in the adjacent atom is used as a channel region.
[Selection] Figure 1

Description

  The present invention relates to a thin film transistor having a metal silicon compound thin film as a channel region.

  Forming transistors using amorphous Si is widely used for TFTs for driving liquid crystal panels. However, amorphous silicon requires the termination of dangling bonds with hydrogen in order to maintain the properties of the material, and it is known that dehydrogenation due to hot electrons or light irradiation causes deterioration of the material. (Staebler-Wronski effect, see Non-Patent Document 1). In particular, fluctuations in threshold voltage and a decrease in carrier mobility significantly deteriorate transistor characteristics.

  Polycrystalline Si has a feature that it has higher mobility than amorphous Si and can form a transistor with excellent characteristics. However, since it is produced by high temperature CVD, heat treatment of amorphous Si, or crystallization by laser irradiation, the cost of the production process is high. In addition, it is difficult to make a very thin film, and the characteristics vary due to the presence of crystal grains, so that a fine transistor cannot be manufactured.

  In general, organic semiconductors have the disadvantage of low mobility. Organic materials such as low molecular weight pentacene are known to have higher mobility than amorphous Si, and are promising as materials for organic devices such as electroluminescence, field effect transistors, and solar cells (non-patent literature). 2). However, heat resistance and oxidation resistance are low compared to Si materials, and stable operation is difficult. In particular, the melting point is about 300 ° C., and it cannot be used for a device that becomes high temperature. Since it is a non-Si-based material, it is poorly matched to the Si LSI process.

It has been reported that high mobility exceeding 100 cm ² / Vs can be realized for an N-type transistor using an oxide semiconductor. There is a problem in forming a P-type semiconductor having the same mobility as that of N-type. In addition, since toxic metals such as Cd and rare metals such as In are used, there are problems in terms of environment and manufacturing costs. The carrier transport characteristics vary depending on the film composition, oxygen bonding state, and heat treatment conditions (see Non-Patent Document 3). Like organic semiconductors, it is a non-Si-based material, so it has poor matching with Si LSI processes.

The present inventors have previously _MSi n (M: transition metal, n = 7-16) proposes a metal silicon compound thin film according to (Patent Document 1). The outline of the metal-silicon compound thin film is as follows.
Among metallic silicide MSi ₂ is, Si clusters encapsulating the transition metal is not present. For example, in the crystal structure of WSi ₂ (C11b type), 10 Sis are coordinated around W, but other Ws are also arranged around 10 Si atoms. In other words, 10 Si atoms are shared by multiple Ws. Even if the closest atom for W is Si, the second neighboring atom is W.
In contrast, the previous metal silicide film (MSi _n film) according to the invention, by a transition metal-containing silicon cluster are bonded to each other Si-Si, that even a Si second neighbor atoms for the transition metal M Features. By having this structure, a semiconductor thin film having the following performance can be formed.

That is, based on the fact that transition metal inclusion silicon clusters are open semiconductors with the gap E _HL between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), by aggregating them, a finite band gap A semiconductor film having the following can be formed.
In particular, when the transition metal M is Mo or W and the number of Si atoms n = 10, 12 is deposited, the band gap of the semiconductor film exceeds the value of 0.8 eV required for room temperature operation of the transistor.
When the number of valence electrons contained transition metal M is odd, since the total number of valence electrons of a transition metal-containing silicon cluster becomes odd (odd electron system), E _HL becomes 0eV strictly. However, by aggregating transition metal-encapsulating silicon clusters, bonds are formed between the clusters, and by exchanging charges accompanied by structural relaxation, as in the case of aggregating even-electron transition metal-encapsulating silicon clusters It becomes a semiconductor film. That is, a semiconductor film can be formed by aggregating transition metal inclusion silicon clusters having a semi-occupied orbit (SUMO) and a LUMO gap E _SL opened instead of E _HL .

A transition metal-silicon compound according to an earlier application (Patent Document 1), wherein a transition metal-containing silicon cluster surrounded by 7 to 16 silicon atoms is formed as a unit structure around the transition metal atom. features of the metal silicon compound film silicon in the first and second neighbor atoms of atoms are arranged (MSi n (n ₌ 7-16) film) is as follows.
1. Suppression MSi _n film deterioration due to hydrogen desorption, because terminating a dangling bond of Si by M, since by hydrogen Unlike amorphous Si does not require termination of dangling bonds, the deterioration of performance due to dehydrogenation There is no.
2. Thin film deposited an electrically conductive control and high carrier mobility MSi _n clusters by field effect is to terminate dangling bonds defect Si by a unit structure of MSi n (n ₌ 7-16), electrical by field effect Control of conduction is possible. In the MoSi ₁₀ film, a p-type field effect mobility of 3.0 × 10 ⁻³ cm ² / Vsec was observed. Further, MSi _n film is the amorphous if you look at the macroscopic, With MSi n (n ₌ 7-16) a unit structure, locally aligned structure, higher carrier mobility than amorphous Si (See Patent Document 1 and Non-Patent Document 4). Further, the carrier type and density can be controlled by changing the type of M (see Patent Document 1 and Non-Patent Document 4). A film having a Si cluster containing M as a unit structure can form a channel region of a thin film transistor in place of amorphous Si.

Japanese Patent Application No. 2009-37261 (Japanese Patent Laid-Open No. 2010-93221)

Kazunobu Tanaka, Junichi Maruyama, Junichi Shimada, Hiroaki Okamoto, Japan Society of Applied Physics: Amorphous Silicon, Ohmsha, 1993, p201-206, p242 Kazuhiro Marumoto: Journal of the Physical Society of Japan November 2007 issue p851 Takahiro Ukai: All about Thin Film Transistor Technology, Industrial Research Committee, 2007, p93-100 J. Kanicki: Amorphous and Microcrystalline Semiconductor Devices: optoelectronic devices (Artech House, 1991) pp. 121-122

  An object of the present invention is to provide a thin film transistor in which a metal silicon compound thin film is used as a channel region.

In order to solve the above problems, the present invention provides the following means.
( 1 ) A transition metal / silicon compound, in which a transition metal-encapsulating silicon cluster surrounded by 7 to 16 silicon atoms surrounds the transition metal atom as a unit structure, and the first and second proximity of the transition metal atom A thin film transistor characterized in that a transition metal silicon compound thin film in which silicon is arranged in an atom is used as a channel region.
( 2 ) The transition metal atom is any of titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zirconium, niobium, molybdenum, ruthenium, rhodium, hafnium, tantalum, tungsten, rhenium, osmium, iridium. The thin film transistor according to (1), wherein the thin film transistor is provided.
( 3 ) A transition metal inclusion silicon cluster (MSi _z ) surrounded by z silicon atoms Si around a transition metal atom M is a unit structure, and the composition ratio of silicon and transition metal (= silicon / transition metal) is n. When M is zirconium, n = 12 to 14, or z is 11 to 12, when M is molybdenum, n = 7 to 13, or z is 8 to 10. The thin film transistor according to (1), wherein the thin film transistor is provided.

  According to the present invention, it is possible to obtain a thin film transistor in which deterioration due to hydrogen desorption is suppressed and the field effect mobility is equal to or higher than that of an amorphous Si thin film transistor.

(A) MSi _n film top gate type TFT using a channel, a bottom gate type TFT using a channel (b) MSi _n film Thermal desorption spectra from WSi ₁₄ H _x films. (A) Si ₁₄ H ₂ , (b) WSi ₁₄ H ₂ stable structure and hydrogen bond energy E (H ₂ ) Bottom gate type TFT using a MSi _n film in a channel V _g dependence of I _d -V _d characteristics of bottom gate TFT using MoSi ₁₀ film as channel Example of stable structure when M = Mo, n = 10 In Fig. 6, an example of the arrangement of Si coordinated to Mo Conceptual diagram explaining the calculation method of radial distribution function (RDF) Conceptual diagram explaining the radial (r) dependence of the radial distribution function (RDF) of Si around M Radial distribution function of Si around Ti in stable structure when M = Ti, n = 12 Radial distribution function of Si around V in a stable structure with M = V and n = 12 Radial distribution function of Si around Cr in stable structure when M = Cr, n = 12 Radial distribution function of Si around Mn in stable structure when M = Mn, n = 12 Radial distribution function of Si around Fe in stable structure when M = Fe, n = 12 Radial distribution function of Si around Co in a stable structure with M = Co and n = 12. Radial distribution function of Si around Ni in stable structure with M = Ni, n = 12 Radial distribution function of Si around Ni in a stable structure with M = Ni and n = 10 Radial distribution function of Si around Cu in stable structure when M = Cu, n = 12 Radial distribution function of Si around Zr in stable structure when M = Zr, n = 12 Radial distribution function of Si around Zr in stable structure when M = Zr, n = 14 Radial distribution function of Si around Nb in stable structure when M = Nb, n = 12 Radial distribution function of Si around Mo in a stable structure with M = Mo and n = 13 Radial distribution function of Si around Ru in stable structure when M = Ru, n = 12 Radial distribution function of Si around Rh in stable structure when M = Rh, n = 12 Radial distribution function of Si around Ta in stable structure when M = Ta, n = 12 Radial distribution function of Si around W in stable structure with M = W and n = 12 Radial distribution function of Si around Re in a stable structure with M = Re n = 12. Radial distribution function of Si around Ir in stable structure when M = Ir, n = 12 Radial distribution function of Si around Mo in stable structure with M = Mo, n = 11.5 Radial distribution function of Si around Mo in a stable structure with M = Mo and n = 10 Radial distribution function of Si around Mo in a stable structure with M = Mo and n = 9 Radial distribution function of Si around Mo in stable structure with M = Mo and n = 8 Radial distribution function of Si around Mo in stable structure when M = Mo, n = 7 The solid line shows the density of electronic states in the stable structure when M = Ti and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Ti from the stable structure. The solid line is the density of electronic states in the stable structure when M = V and n = 12, and the dotted line is the density of electronic states for the Si-only system excluding V from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Cr and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Cr from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Mn and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Mn from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Fe and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Fe from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Co and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Co from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Ni and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Ni from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Ni and n = 10, and the dotted line shows the density of electronic states for the Si-only system excluding Ni from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Cu and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Cu from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Zr and n = 14, and the dotted line shows the density of electronic states for the Si-only system excluding Zr from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Nb and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Nb from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Mo and n = 10, and the dotted line shows the density of electronic states for the Si-only system excluding Mo from the stable structure. The solid line shows the density of electronic states in the stable structure when M = Ta and n = 12, and the dotted line shows the density of electronic states for the Si-only system excluding Ta from the stable structure. The solid line shows the density of electronic states in the stable structure when M = W and n = 10. The dotted line shows the density of electronic states for the Si-only system excluding Ti from the stable structure. Comparison of vibrational density of states of MoSi ₁₀ film and hydrogenated amorphous Si by computer simulation Raman spectra of MoSi ₁₀ film and hydrogenated amorphous Si films prepared as a channel of a thin film transistor

Figure 1, MSi _n film is a metal silicide: shows a thin film transistor and (M a transition metal, n = 7 to 16) a channel region (TFT).
FIG. 1A shows a top gate type TFT, and FIG. 1B shows a bottom gate type TFT. In Figure 1, 1 denotes a gate electrode, 2 denotes a source electrode, 3 the MSi _n film channel, 4 denotes a gate insulating film, 5 denotes a drain electrode, 6 is a TFT substrate.
Hereinafter, a method for manufacturing each TFT will be described.

In manufacturing the top gate TFT, a method for manufacturing the top gate TFT using laser ablation for channel formation will be described in detail.
(1) Deposit MSi _n H _x film on glass substrate. MSi _n H _x films are synthesized and deposited by pulsed laser ablation of transition metal silicon compound targets such as M targets or MSi ₂ targets in a SiH ₄ atmosphere. At this time, the substrate temperature is room temperature.
(2) MSi _n H _x film desorbed and H was heated to 190 ° C. or higher MSi _n film formation. In other words, if using the M capable of synthesizing MSi _n H _x film to form a MSi _n film channel. MSi _n film, a metal (M) is vapor and monosilane gas (SiH ₄₎ was synthesized hydrogenated transition metal clusters MSi _n H _x in the raw material, since they are prepared by depositing on a solid substrate, hydrogen remains Yes. Therefore, the MSi _n film, for use as a channel material of a TFT containing no hydrogen, it is necessary to establish the hydrogen desorption process by heat treatment.

FIG. 2 shows a thermal desorption spectrum of the WSi ₁₄ film deposited on the quartz substrate. The horizontal axis indicates the temperature of the sample surface, and the vertical axis indicates the pressure in the vacuum chamber. A desorption peak was observed around the sample surface temperature of 140 ° C., and desorption was completed at 190 ° C. That is, hydrogen release from the WSi ₁₄ H _x film occurs at 190 ° C. or lower. This is lower than the hydrogen desorption temperature (320 ° C.) from hydrogenated amorphous Si, and shows that the WSi ₁₄ film has a lower hydrogen binding energy than a-Si: H. In fact, it has been found from the first principle calculation that hydrogen bonded to WSi ₁₄ has about 53% lower binding energy than hydrogen bonded to Si ₁₄ having the same Si cage structure. (See Figure 3)
(3) Form source and drain electrodes such as Al, Mo, MSi ₂ silicide.
(4) depositing a gate insulating film (SiO _2, SiN _x, etc.) by using CVD or sputtering.
(5) Form gate electrodes such as Al and Mo using vacuum deposition or sputtering.

In manufacturing the top gate TFT, a method for manufacturing the top gate TFT using sputtering for channel formation will be described.
(1) A sputtered film is deposited on a glass substrate using a target of M: Si = 1: n (Ar: 30 sccm, 3.0 × 10 ⁻³ Pa, DC power 30 W) to form a channel. At this time, the substrate temperature is room temperature.
(2) Heat the sputtered film to 500 ° C. or higher to improve the channel film quality.
(3) Form source and drain electrodes such as Al, Mo, MSi ₂ silicide.
(4) depositing a gate insulating film (SiO _2, SiN _x, etc.) by using CVD or sputtering.
(5) Form gate electrodes such as Al and Mo using vacuum deposition or sputtering.

Next, in manufacturing the bottom gate TFT, a manufacturing method of the bottom gate TFT using laser ablation for channel formation will be described.
(1) A gate electrode such as Al or Mo is formed on a glass substrate using vacuum deposition or sputtering.
(2) Deposit a gate insulating film (SiO ₂ , SiN _x, etc.) using CVD or sputtering.
(3) MSi _n H _x film deposition. MSi _n H _x clusters are synthesized and deposited by pulsed laser ablation of a transition metal silicon compound target such as an M target or an MSi ₂ target in a SiH ₄ atmosphere. At this time, the substrate temperature is room temperature.
(4) MSi _n H _x film desorbed and H was heated to 190 ° C. or higher MSi _n film formation. In other words, if using the M capable of synthesizing MSi _n H _x film to form a MSi _n film channel.
(5) Form source and drain electrodes such as Al, Mo, MSi ₂ silicide.

In manufacturing the bottom gate TFT, a method for manufacturing the bottom gate TFT using sputtering for channel formation will be described.
(1) A gate electrode such as Al or Mo is formed on a glass substrate by vacuum deposition or sputtering.
(2) Deposit a gate insulating film (SiO ₂ , SiN _x, etc.) using CVD or sputtering.
(3) Sputtered films are deposited using a target of M: Si = 1: n (Ar: 30 sccm, 3.0 × 10 ⁻³ Pa, DC power: 30 W) to form a channel. At this time, the substrate temperature is room temperature.
(4) Heat the sputtered film to 500 ° C. or higher to improve the channel film quality.
(5) Form source and drain electrodes such as Al, Mo, MSi ₂ silicide.

The field effect of the thin film transistor according to the present invention is measured.
As shown in FIG. 4, a TFT was fabricated using a MoSi ₁₀ film. Here, a bottom gate type TFT using the Si thermal oxidation substrate as the gate electrode 14 and the gate insulating film 13 was fabricated.
The Si substrate was an n-type 0.01 Ωcm substrate, and the thickness of the gate insulating film 13 made of an SiO ₂ oxide film was 200 nm.
A MoSi ₁₀ film was deposited to a thickness of 15 nm, and an Al electrode was formed thereon as a source / drain electrode 11 by vacuum evaporation. The width of the source / drain electrodes is 300 μm, and the channel width is 100 μm. Further, an Al electrode for contact was similarly formed on the Si substrate acting as the gate electrode 14 by vacuum deposition.
FIG. 5 shows the gate voltage (V _g ) dependency of the drain voltage (V _d ) characteristics of the drain current (I _d ). A p-channel enhancement type field effect characteristic was obtained, the threshold voltage V _th was −3 V, and the field effect mobility was estimated from the drain current value in the linear region to be 3 × 10 ⁻³ cm ² / Vsec.

  A transition metal and silicon compound, a transition metal-encapsulating silicon cluster surrounded by 7 to 16 silicon atoms surrounding the transition metal atom as a unit structure, and silicon in the first and second neighboring atoms of the transition metal atom The present invention has been described by exemplifying a thin film transistor having a metal silicon compound thin film in which a channel is disposed as a channel region. However, the present invention is not limited to this, and is a compound of transition metal and silicon, and a composition of silicon and transition metal. The present invention can also be applied to a thin film transistor in which a metal silicon compound thin film having a ratio (silicon / transition metal) of 7 to 16 is used as a channel region.

Next, an example of a channel using a transition metal silicon compound in which a cluster in which a silicon atom Si includes one transition metal atom M is a unit structure is provided by first-principles calculation.
The calculation is performed in two stages. In the first stage, a cluster MSi in which a transition metal silicon compound material in which the composition ratio of silicon and transition metal M (= silicon / transition metal) is n coordinates z (average Z) Si per M Indicates that _z is a unit structure. The second stage shows that M has an effect of terminating Si dangling bonds in the transition metal silicon compound material having the structural characteristics shown in the first stage.

First, the first stage will be described. Generate M and Si coordinates using random numbers. The total energy and the force felt by each atom are calculated by the first principle calculation, and the atomic coordinates are updated so that the total energy is reduced. The atomic coordinates are repeatedly updated until the force felt by all atoms falls below 0.01 electron volts per angstrom.
An example of a stable structure of a system in which the atomic ratio of molybdenum (M = Mo) and Si is 1:10 (n = 10) is shown in FIG. In FIG. 6, Si and Mo are represented by white and black spheres, respectively.
FIG. 7 shows an example of the structure of 8, 9, and 10 Si surrounding the Mo from FIG.

In order to identify the structure surrounding Si around M, the radial distribution function is calculated. A method for calculating the radial distribution function is shown in FIG. The number of Si (shaded circles) in a spherical shell with a thickness Δr between a sphere with a radius r (dotted line) and a sphere with a radius r + Δr (dotted line) around one M (black circle) is 4πr ^2. Let f (r) be divided by. For each transition metal atom, calculate f (r) for r from 0 angstroms to 10 angstroms. A radial distribution function is obtained by calculating f (r) as an average value per transition metal atom.

An example of the radial distribution function obtained by the above method is shown in FIG. “RDF” represents a radial distribution function, and “r” represents a distance (angstrom) from the position of M. The presence of Si surrounding M is identified by the first peak shown with diagonal lines. The first peak is separated from the second peak by a gap (FIG. 7). The gap position is defined as r giving the minimum value of the radial distribution function between the first peak position and r = 4 angstroms. The number of Si belonging to the first peak is the average number Z of the number z of Si surrounding M1. To calculate Z, the radial distribution function is multiplied by 4πr ² and r is integrated from the origin to the gap position.

Radial distribution function for M = Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zr, Nb, Mo, Ru, Rh, Ta, W, Re, Ir when n is 7 or more and 14 or less Is shown in FIGS. 10 to 33, the value of the radial distribution function (vertical axis) is normalized to 1 as the maximum value. The moving radius (horizontal axis) is in angstrom units.
It can be seen that all of the MSi _n shown in FIGS. 10 to 33 have a feature that an MSi _z cluster in which M contains z Si is included as a unit structure. The average Z of z varies from 6.6 (NiSi ₁₀ , FIG. 17) to 11.9 (ZrSi ₁₄ , FIG. 20) depending on M and n. The values of z and n need not match. For example, when n = 12 to 14 in the case of zirconium, z is 11 to 12, and in the case of molybdenum, z remains in the range of 8 to 10 even if n = 7 to 13.

As the second stage of the calculation, an example is given in which a dangling bond of Si is terminated by M being surrounded by Si. The electronic density of states for M = Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Zr, Nb, Mo, Ta, and W is shown in FIGS. 34 to 47 as a function of energy.
34 to 47, the solid line represents the electronic state density in the stable structure when M and n are selected, and the dotted line represents the electronic state density for the Si-only system excluding M from the stable structure. 34 to 47, the former is normalized so that the number of states integrated to the Fermi level matches that of the latter. The unit of the vertical axis is the number of states per electron volt, and zero on the horizontal axis represents the Fermi level.
The zero point of energy is an electron level called Fermi level. The whole system has an electronic state with an energy lower than the Fermi level in the ground state, and does not take an electronic state with an energy higher than the Fermi level.

  Here, it is described that the termination of dangling bonds can be determined by reducing the Fermi level and the density of electronic states in the vicinity thereof. For this purpose, the relationship between electronic levels and chemical bonds is outlined. Electronic states with energy below the Fermi level strengthen the bond between atoms, but electronic states with energy above the Fermi level weaken the bond. For this reason, the former may be referred to as a coupled state and the latter as an anti-coupled state.

  In other words, the Fermi level is an energy value that separates the bonded state and the anti-bonded state. The electronic state resulting from the dangling bond is a non-bonded state that does not belong to either the bonded state or the antibonded state, and its energy matches the Fermi level. Even if the energy is lower (or higher) than the Fermi level, an electronic state having an energy close to the Fermi level is weak in bonding (anti-bonding) properties and has non-bonding properties. The unbonded state is unstable, and when an electron (hole) is added to the system, the added electron (hole) is easily trapped in the unbonded orbital in order to cancel the unbonded state. Therefore, when there is a dangling bond in the channel of the transistor, a non-bonding state caused by the dangling bond is eliminated, so that carriers introduced into the channel are captured by the dangling bond and do not contribute to electrical conduction.

Based on the above, when the electronic state density (solid curve) of the channel using the metal silicon compound material shown in FIGS. 34 to 47 is viewed, the electronic state density is minimal at the Fermi level. I understand that The electron density of states for the system in which M is removed and only Si is left is displayed as a dotted line in each of FIGS. From the comparison of the density of electronic states indicated by the solid line and the dotted line, it can be seen that the dangling bond of Si is terminated by M in the channel using the metal silicon compound. FIG. 45 shows the density of electronic states of the channel material MoSi ₁₀ film of the thin film transistor shown in FIG. The transistor is operated by terminating the dangling bond of Si by Mo. Therefore, the transistor operates as long as the electronic density of states shown in FIGS.

FIG. 48 shows the result of calculating the vibrational state density for the MoSi ₁₀ film structure shown in FIG. 6 in comparison with the vibrational state density of hydrogenated amorphous Si. Than the optical phonon peak frequency 480 cm ^-1 of the hydrogenated amorphous Si in the region of the lower wave number side of 200-400cm ^-1, it can be seen that the vibrating state density of MoSi ₁₀ film becomes high.

FIG. 49 shows Raman spectra of the MoSi ₁₀ film synthesized as the channel of the thin film transistor shown in FIG. 5 and the hydrogenated amorphous Si film. As expected from the calculation results, it was observed that the vibrational state density of the MoSi ₁₀ film was higher on the lower wavenumber side than the optical phonon peak 480 cm ⁻¹ of hydrogenated amorphous Si. It is considered that the structure of the MoSi ₁₀ film constituting the channel of the thin film transistor and the structure of the film shown in the calculation simulation are the same.

DESCRIPTION OF SYMBOLS 1 Gate electrode 2 Source electrode 3 MSi _n film channel 4 Gate insulating film 5 Drain electrode 6 TFT substrate 11 Source / drain electrode 12 MSi _n film channel 13 Gate insulating film 14 Gate electrode (n-type Si substrate)

Claims (3)

A transition metal and silicon compound, a transition metal-encapsulating silicon cluster surrounded by 7 to 16 silicon atoms surrounding the transition metal atom as a unit structure, and the first and second adjacent atoms of the transition metal atom are silicon. A thin film transistor characterized in that a transition metal silicon compound thin film in which is disposed is used as a channel region. The transition metal atom is any of titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zirconium, niobium, molybdenum, ruthenium, rhodium, hafnium, tantalum, tungsten, rhenium, osmium, and iridium. The thin film transistor according to claim 1. When the transition metal inclusion silicon cluster (MSi _z ) surrounded by z silicon atoms Si around the transition metal atom M is a unit structure and the composition ratio of silicon and transition metal (= silicon / transition metal) is n, When M is zirconium, n = 12 to 14, or z is 11 to 12, when M is molybdenum, n = 7 to 13, or z is 8 to 10. The thin film transistor according to claim 1.