JPH05259458A - Manufacture of semiconductor device - Google Patents

Abstract

PURPOSE:To make a thin film semiconductor active element which is large in area and has high field effect mobility by covering both obverse and reverse of a semiconductor film with thick insulating films 5-10nm thick, with its upper limit of thickness 50nm. CONSTITUTION:A-Si; H films (semiconductor films) 2 being active layers and SiHx films (insulating films) 3 are made continuously thereon by the changeover of reactive gas using plasma CVD method, using a fused quartz fitted with RF sputter SiO2 as a substrate 1. Here, the SiNX 3, 5nm thick does not cause composition change to the heat process of 700 deg.C, and besides the a-Si: H film 2 caught between the SiNX films 3 does not composition change, either. Moreover, the a-Si: H film 2, 50nm thick can have about the same crystal property as the one 100nm thick, and it functions as a stabler element to the distortion from the substrate or the heat diffusion of impurities than the thermocrystallized one of the simple of the semiconductor layer.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for manufacturing a semiconductor device suitable for use in a display device such as a liquid crystal display.

[0002]

2. Description of the Related Art Conventional thin film semiconductor active devices (hereinafter referred to as TFTs) have been widely used in electric and electronic fields such as active devices for liquid crystal displays, active devices for image sensors, various measuring instruments and switch displays. There is. Of these, a TFT having a high electric field mobility is useful in that it can have a memory transmission and storage function in addition to a function as a switch.

By the way, with the recent development of information processing technology, various solid-state displays using liquid crystal displays and ferroelectrics have been developed as display devices to replace cathode ray tubes. A TFT having a high field effect mobility is indispensable as an active element. In addition, a new electro-optical element due to interaction or mutual conversion between an electric signal and an optical signal, or a TFT as an active element when two-dimensional or three-dimensional expansion is required in a recording material
is necessary. Particularly in recent years, there is an increasing demand for high pixel display in electronic display devices such as liquid crystal displays, electroluminescence displays, and plasma displays and fluorescent display devices. It has been proposed that a TFT having mobility be directly formed in a pixel portion, and at the same time, a driving circuit portion around the pixel be integrally formed.

[0004]

However, a TFT used in a display device such as a liquid crystal display is composed of an amorphous silicon thin film or a polysilicon thin film, and the field effect mobility of the TFT (hereinafter referred to as μF).
Write E. ) Depends on the TFT constituent material, but μFE of amorphous silicon is 0.1 to ¹ cm ² / Vsec.
The extent of its application is limited.

The reason why the TFT made of amorphous silicon has a small μFE is that dangling bonds (unpaired electron pairs of silicon atoms) are mainly present in the amorphous silicon in a large amount. Therefore, as long as the silicon thin film that constitutes the TFT is in the amorphous state, TF
It is considered difficult to improve the μFE of T. Generally, when an amorphous silicon thin film is formed on a substrate, a silicon-containing source gas such as silane gas (SiH ₄ ) or disilane gas (Si ₂ H ₆ ) is decomposed in a high frequency electric field to deposit silicon on the substrate. However, when silicon is rapidly cooled on the surface of the substrate and deposited as amorphous silicon, active species including unpaired electron pairs such as SiH and SiH ₂ are inevitably taken in. Therefore, it is impossible to completely remove the dangling bond from the amorphous silicon thin film, and as a result, μF of 1 cm ² / Vsec or more is obtained.
It was not possible to form an amorphous silicon thin film having E.

On the other hand, TET having a μFE of more than 40 cm ² / Vsec by the LPCVD method at 600 ° C. or lower.
It is reported that can be manufactured. However, the LPCVD method can be formed only on a substrate having a size of about 30 cm square, and cannot be used when a TFT for a large area and high pixel display device is manufactured due to the problem of shrinkage of the substrate. was there.

As a method for solving such a problem, it has been attempted to improve the μFE by crystallizing the surface of an amorphous silicon thin film capable of forming a large area by thermal annealing. This method is currently widely used when manufacturing a TFT at 600 ° C. or lower where a glass substrate or the like can be used. In this method,
There is a disadvantage that it is difficult to obtain a TFT having a high μFE. Further, since the TFT is distorted due to the difference in the coefficient of thermal expansion between the glass substrate and the silicon thin film, and cracks or peeling occur, it is impossible to manufacture a large-area TFT.

Therefore, the surface of the amorphous silicon thin film is locally heated by using an ultraviolet laser or a visible laser, and
00 cm ² / V. TFT with μFE exceeding sce
Have been proposed. However, this method can be obtained because it is difficult to uniformly irradiate the surface of the amorphous silicon thin film with laser light, and it is difficult to control the film forming conditions of the amorphous silicon thin film due to fluctuations in the laser light intensity. There is a problem in that the crystallized silicon film to be formed lacks uniformity.

The present invention has been made to solve the above problems, and has a large area and a high TF.
It is an object to provide a manufacturing method capable of supplying T.

[0010]

In order to achieve the above object, in the method for manufacturing a semiconductor device according to the present invention, it is a means of solving the problem that the semiconductor thin film has a step of coating both front and back surfaces with an insulating thin film. did.

The upper limit of the thickness of the semiconductor thin film is 50 nm,
The insulating thin film has a thickness of 5 to 10 nm.

[0012]

Since the method of manufacturing the semiconductor device of the present invention has the above-described structure, crystallization is effectively promoted, and T having a high mobility is obtained.
An FT is created.

Further, since the semiconductor thin film sandwiched by the insulating film from both sides is not exposed to the outside air directly, it is possible to fabricate a device that can withstand high temperature without being contaminated from the outside.

Further, since the front and back surfaces are covered with the same insulating thin film, the influence of the shrinkage and expansion due to heat of the glass and the base film to be the substrate can be alleviated.

[0015]

EXAMPLES In the method of manufacturing a semiconductor device according to the present invention, the method of manufacturing an amorphous silicon thin film is not particularly limited, and ordinary film forming means such as plasma CVD method, sputtering method and LPCVD method at 600 ° C. or lower are used. be able to. The silicon thin film obtained by the usual film forming means contains 35% or less of hydrogen in volume ratio, or does not contain hydrogen at all, and further necessarily contains an amorphous component. Plasma CV
To manufacture an amorphous silicon thin film using the D method, for example, silane (SiH ₄ ) or disilane (Si
₂ H ₆ ) or other silane-based gas at a direct current of 2.45 GH
A method of decomposing in a high electric field up to z can be used. When the sputtering method is used, a direct current or magnetron type RF applied stappering method can be used. At that time, a high-purity silicon single crystal, a high-purity silicon polycrystal, or the like can be used as a target, and the atmosphere thereof can be an argon atmosphere or an argon atmosphere of hydrogenation, or a 100% hydrogen atmosphere. As the LPCVD method, for example, a method of bringing a silane-based gas such as silane gas or dicine lan gas into contact with an inert gas onto a substrate heated under reduced pressure of 10 Torr or less can be used. In any of the above methods, the oxygen content in the atmosphere was set to 10 ²¹
/ Cm ³ or less, preferably 10 ¹⁹ / cm ³ .

The insulating thin film of the present invention is preferably a thin film that prevents diffusion of impurities, such as silicon nitride (SiN _x ). This SiN _x
A plasma CVD method is used for the manufacturing method.

The method and results of the experiment leading to the manufacturing method of the semiconductor device of the present invention are shown below. Example 1 FIG. 1 shows a cross-sectional structure of a multilayer film used in an experiment for confirming the temperature stability of a semiconductor device manufactured by the method of the present invention. Fused quartz with RF sputtered SiO ₂ was used as the substrate 1, and plasma CV was formed on it.
The a-Si: H film (active layer) and the SiN _x film (insulating film) were continuously formed by switching the reaction gas using the D method. The apparatus used for forming the multilayer film is a parallel plate plasma chamber.

The thickness of the SiN _x insulating layer of the multilayer film is fixed (5
nm) and the total film thickness of the a-Si: H layer is 100 n.
The film thickness of one layer was changed using m as a constant. Table 1 is a comparison table of the experiment No and the film thickness of each layer. These are 100 nm, 50 nm × 2, 20 nm × 5, 10
Five conditions of nm × 10 and 5 nm × 20 were set.

[Table 1]

Table 2 shows the conditions of a-Si: H and SiNx film. Film forming temperature, reaction pressure, and RF power are each 250
C., 4. It was Opa and 20W. During film formation, cleaning by H ₂ plasma after forming each layer was not performed.

[Table 2]

The laminated film of FIG. 1 was set in a vacuum heating furnace, and
Vacuum heating was carried out at 00 ° C. for 48 hours. The distribution of nitrogen atoms in the thickness direction was investigated in order to investigate the structural change of the multilayer film by heat treatment. FIG. 2 shows the distribution of nitrogen atoms before and after the heat treatment measured by SIMS. As dero (during film formation) and 700
Since the nitrogen atom distribution of the sample after ℃ thermal annealing is almost unchanged, the multilayer film structure is sufficiently stable and does not change the composition of nitrogen due to diffusion even if it is subjected to a high temperature treatment of at least 700 ° C. It was found that the film thickness was reduced due to the influence of.

The crystallinity of the multilayer film was examined by laser Raman spectroscopy. An Ar laser with a wavelength of 488 nm was used for the analysis. The lowest layer a-Si: H / SiN in this wavelength range
_{x The} multilayer film can absorb laser light from the surface to the bottom layer. In order to evaluate the crystallinity of the minute area, the size of the laser beam was narrowed to φ1 μm. At the time of measurement, steps were taken to minimize the laser output power in order to avoid crystallization due to laser irradiation. FIG. 3 shows one layer of a-S
i: Raman spectrum of samples having different H film thicknesses. The heat treatment conditions were 650 ° C. and 48 hrs. Although Raman peaks showing crystallinity appear in all the samples, as the thickness of one a-Si: H layer becomes smaller than that of the single-layer a-Si: H film, the Raman spectrum shows It has become gentle and the peak has shifted to the lower wavenumber side.

FIG. 4 shows one layer of a-Si having Raman shift and full width at half maximum (FWHM) obtained from the Raman spectrum of FIG.
H film thickness dependency is shown. For more than a thickness of 20nm, the film thickness dependency of the Raman shift and FWHM hardly, but were respectively 517 cm ^-1 and 8 cm ^-1, the film thickness is 20nm
Both parameters will start to change rapidly when
time of nm, respectively become 502cm ^-1 and 17cm ^-1.

The cause of the Raman shift depending on the film thickness of the a-Si: H active layer has not been clarified yet, but it is presumed to be related to the average grain size of crystalline Si in the active layer. According to the research results on the Raman spectrum of the conventional microcrystalline silicon film, the Raman peak shifts to the lower wave number side as the crystal grain size becomes smaller. If crystal nuclei are generated at the interface between the active layer and the insulating layer and crystal grows along a certain crystal orientation, it is possible to grow within one layer.
The maximum particle size of y-Si is determined by its film thickness. Therefore, if the theoretical active layer thickness is adjusted, crystalline Si
The particle size can be controlled.

Through the above experiments, the following conclusions were reached. (1) 5 nm the SiN _x film thickness of found to have no also composition change and no rise to compositional changes of the heat transfer process 700 ° C. active layer incorporated tried narrow its the SiN _x film (semiconductor thin film) It was This indicates that the insulating film having a thickness of 5 nm thermally crystallizes the encapsulated active layer stably with respect to heat and without chemical change. (2) Then, considering the mechanism in which the active layer actually sandwiched by the insulating film of SiN _x on both sides is thermally crystallized, it was found that the effect of the thickness is great. 2 from FIG.
The active layer with a thickness of 0 nm or less has a limit due to the thickness due to restrictions such as grain size, but with a thickness of 50 nm is 1
It can have the same degree of crystallinity as those with a thickness of 00 nm. Conversely, a thickness of 50 nm means that the semiconductor layer functions as a more stable element against strain from the substrate or thermal diffusion of impurities than thermal crystallization of the semiconductor layer alone.

[0025]

[Embodiment 2] FIG. 5 is a longitudinal sectional view showing an example of a structure of a TFT using a laminated structure of SiN _x / Si: H / SiN _x . In the figure, reference numeral 11 is an island polySi, SiN _x 5n
m, a-Si: H is a laminated structure of 50 nm and SiN _x 5 nm, and is made into poly-Si by thermal annealing at 550 ° C. for 24 hours.

N + source: n + (S), drain electrode:
n + (D) is formed by ion implantation of phosphorus atoms.
As a procedure of this process, the gate Si is
The steps of forming O ₂ 12 and etching the gate SiO ₂ are performed. At this time, Si on the poly-sourced drain is formed.
The N _x film is removed. This makes it possible to make smooth contact between the source / drain extraction electrodes and the source / drain in this part after the activation process by a laser after ion implantation. In the figure, 13 is a display electrode, 14 is an Al electrode, and 15 is a poly- electrode.
Si: n + (G) gate, 16 is an interlayer SiO ₂ .

The characteristics of the TFT manufactured as described above are shown in FIG. This makes the ON / OFF ratio 6 digits, V _{+ h}
= 1.3V (Vo = 10V), a TFT having a mobility exceeding 80 cm ² / Vs was obtained.

The structure of FIG. 5 is not special and is commonly used. The high mobility obtained by sandwiching the active layer with two thin film insulators in this way indicates that the stability of this structure and the strain from the substrate due to heat are reduced.

[0029]

As described above, in the method of manufacturing a semiconductor device according to the present invention, by covering both the front and back surfaces of a semiconductor thin film with an insulating film, crystallization is effectively promoted and a high mobility TFT is obtained. Is a manufacturing method.

[Brief description of drawings]

FIG. 1 is a schematic cross-sectional view of an a-Si: H / SiNx multilayer film.

FIG. 2 is a diagram showing the distribution of nitrogen content before and after heat treatment evaluated by SIMS.

FIG. 3 is a diagram showing Raman spectra of samples having different a-Si: H film thicknesses.

FIG. 4 is a diagram showing the dependence of Raman shift or full width at half maximum (FWHM) on the a-Si: H film thickness of one layer.

FIG. 5 is a schematic cross-sectional view of a TFT-LCD (liquid crystal display device) substrate.

FIG. 6 is a diagram showing characteristics of a TFT-LCD made according to Example 2.

[Explanation of symbols]

1 Substrate 2 a-Si: H film (active layer) 3 SiNx film (insulating film) 11 Island poly-Si (SiN _x 5 nm / active layer)
50 nm / SiN _x 5 nm) 12 gate SiO ₂ 13 display electrode 14 Al electrode 15 poly-Si, n + (G) gate 16 interlayer SiO ₂

Claims (2)

[Claims] 1. A method of manufacturing a semiconductor device, comprising a step of coating both surfaces of a semiconductor thin film with insulating thin films. 2. The semiconductor thin film and insulating thin film used in the above step are 50 nm or less and 5 n, respectively.
The method for manufacturing a semiconductor device according to claim 1, wherein the range is m to 10 nm.