JPS63173369A - Thin film semiconductor element - Google Patents

Abstract

PURPOSE:To make compatibility of a small OFF-current and a large ON current possible, which is mainly suitable for a thin film transistor to drive an active matrix panel, by forming a hole or a nick for a permeation path of hydrogen in a gate electrode in the plane pattern of the gate electrode. CONSTITUTION:The minimum dimension Lt over which a gate electrode 6 continuously extends on a channel region 2 is designed as a dimension which is highly effective for hydrogen substitution treatment, by providing a gate electrode 6 with a hole 11. According to this design, the defect density can be decreased all over the boundary surface between the channel region 2 and an insalative film 5, so that the ON-current is highly increased. In the case where the gate electrode is not continuously formed from the source side end-portion to the drain side end-portion on the channel region, and even one part in the above section is cut by a hole promoting hydrogen permeation, the crossectional structure of the part does not constitute an MOS type structure, so that a channel is not formed in this part, and the OFF-current is extraordinarily decreased. A thin film transistor of this element is obtained wherein the ON-current is sufficiently large and the OFF-current is small.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to the structure of a thin film semiconductor element formed of amorphous or polycrystalline material, and in particular realizes characteristics of sufficiently large on-current and small off-current. The present invention relates to a thin film MOS type transistor.

[Conventional technology]

In recent years, research on forming thin film transistors on insulating substrates has been actively conducted. This technology is used in many applications, including active-mapped IJ panels that realize thin-film displays using inexpensive insulating substrates, three-dimensional integrated circuits formed on ordinary semiconductor integrated circuits, and image sensors such as line sensors. Applications are expected.

In the following description, an example will be explained in which polycrystalline silicon is used as the material of a thin film transistor and is applied to a thin film display panel, but the present invention is equally applicable to other cases using thin film transistors. Furthermore, the semiconductor material is not limited to polycrystalline silicon, but can be similarly applied to other materials.

When polycrystalline silicon is used as a semiconductor material in thin film transistors, there are problems such as high threshold voltage, low field effect mobility, and large leakage current. is not obtained.

A possible cause of this is that many defect levels such as dangling bonds are present in the grain boundaries of polycrystalline silicon.

For example, the gate voltage applied to turn on the transistor is used to fill defect levels in polycrystalline silicon, increasing the threshold voltage.

Even if a channel is formed by applying a sufficiently high gate pressure, the carriers trapped at the grain boundaries will
It acts as a potential barrier for free carriers that contribute to electrical conduction, reducing field effect mobility.

Further, even in a junction structure formed by doping with phosphorus, arsenic, or anotymone, there is a problem in that the electric field is concentrated at the grain boundaries due to the crystal defects, and a leakage current tends to flow.

Various measures have been considered to improve such polycrystalline silicon grain boundaries. One of the known countermeasures is to replace hydrogen with dangling bonds at polycrystalline silicon grain boundaries to form silicon-hydrogen bonds (hereinafter referred to as 8i-H) and reduce the local level density. ing.

Specifically, as a means of replacing hydrogen, a plasma silicon nitride film containing a large amount of hydrogen (hereinafter referred to as P-8IN) is used as a passivation film, and this is (1) heat-treated at 300 to 500°C after deposition. ,(2
)) exposing a thin film transistor to high-frequency plasma hydrogen at an ambient temperature of 300 to 400[deg.] C.; and (3) implanting hydrogen ions to perform heat treatment.

For example, in a polycrystalline silicon thin film transistor, P −
The one using 8tN as a hydrogen method is the Journal of the Japan Society of Applied Physics, 25.1m2 (1986).

Pages 121 to 123 (JAPAF SE JOUR
NALOF APPLIED P)n'8Ics,V
ol, 2 5, 隘2 (1986) PP, L 12
1-r, 123) and discuss the effects of P-8IN.

However, the effects of these hydrogen replacement methods depend on the transistor size (channel length and channel width), structure, manufacturing process, materials, etc., and the above-mentioned structure is particularly effective for the hydrogen replacement methods. Nothing was made clear in the paper.

FIG. 2 shows a general structure of an N-channel thin film transistor using a semiconductor thin film, in which (,) is a plan view, and (b) is a cross-sectional view taken along a line from X in (,).

In the same figure, the material is an insulating material such as glass (transparent substrate,
2 is a channel region made of a semiconductor thin film such as polycrystalline silicon, 3 is a source region formed by doping impurities such as phosphorus or arsenic into the semiconductor, 4 is also a drain region,
5 is a gate insulation film, 6 is a gate electrode, 7 is an interlayer insulation film,
8 is an aluminum electrode. Further, in the figure, W is the channel width of the transistor, and L is the channel length.

[Problem that the invention seeks to solve]

FIG. 3 shows the characteristics when the semiconductor element having the conventional structure described above is exposed to high frequency plasma hydrogen.

The high-frequency plasma conditions were a temperature of 350° C., a chamber pressure of 0.75 Torr, and a time of 90 minutes.

This is the on-characteristic of the semiconductor element, that is, the drain 111
This is the condition when the first stream is most improved by the plasma hydrogen replacement method.

FIG. 3 shows the dependence of the drain current ID on the channel length when the channel width W is kept at a constant value of 15 μm and the channel length is varied from 5 μm to 75 μm.

Note that this data is the result obtained through experiments conducted by the inventors.

The horizontal axis of this graph is the ratio of W and L. In this case W is 1
Since the W/L ratio is kept constant at 5 μm, the smaller the value of the W/L ratio, the larger the value. The vertical axis is toru-in current ID.

The solid line C shows the drain current characteristics after the hydrogen replacement treatment, and the broken lines B and D show the drain current characteristics before the hydrogen replacement. Lines C and D are drain current characteristics when the thin film transistor is turned on, and lines C and D are drain current characteristics when the thin film transistor is turned off.

The problems to be solved by the present invention will be clarified from the relationship between the current characteristics required of the thin film transistors constituting the active matrix panel and the characteristics shown in FIG.

FIG. 4 is a matrix layout diagram of liquid crystal driving elements using thin film transistors, which constitutes the display area of the finished display. In the figure, 7 is a data signal fin, 8 is a data signal fin, and 8 is a data signal fin.
is line 1B in front of my feet.

Each liquid crystal driving element is composed of a thin film transistor 9 and a capacitor 10.

The thin film transistor 9 is generally a field effect conductor element having the structure shown in FIG. 2, and performs data switching. The capacitor 10 is used to hold data signals, and is mainly composed of the capacitance of the liquid crystal itself.

As can be seen from the above, the thin film transistor 9 is used to switch the data voltage applied to the liquid crystal.

At this time, the characteristics required of thin film transistors are broadly classified into the following two types.

(1) When the thin film transistor 9 is turned on, sufficient current can flow to charge the capacitor 10.

(2) No current flows when the thin film transistor is turned off.

The above (1) is related to the writing characteristics to the capacitor, and in order to be able to write data in a short time, it is necessary to have a large! , it is necessary to flow the flow. Further, the above (2) is required to hold data written in the capacitor, and if the leakage IM, K is large, the data will not be held correctly and for a desired time.

Let's examine the characteristics shown in Figure 3 from the above perspective.

From the point of view of the above-mentioned (1) on-current, the straight entry when hydrogen replacement treatment is performed and the 1R when it is not performed.
It can be seen that in both cases 1B, the on-current can be increased by decreasing the channel length.

In addition, especially as the channel length becomes smaller, the lines A and B
The difference becomes larger, and it can be seen that the effect of the water cumulative replacement method is higher as the channel length becomes smaller.

Next, from the perspective of the off-state current in (2) above, it can be seen that the off-state current increases as the channel length is reduced for both the straight line c when hydrogen replacement treatment is performed and the straight line iD when it is not performed. .

It can also be seen that the straight lines of C and D have approximately the same value.

From the above, it can be seen that if the channel length of a conventional semiconductor element is made smaller in order to increase the effectiveness of the hydrogen substitution method, the on current becomes larger, but the off current also becomes larger accordingly. In order to retain the data stored in the liquid crystal capacitor 1o, the off-state current needs to be suppressed to a certain value, for example 1O-11 (A) or less, so simply reducing the channel length is not sufficient. It is clear that this is not the case.

The purpose of the present invention is to increase the effect of hydrogen replacement treatment to the same extent as when the channel length is shortened to increase the on-current, and at the same time to reduce the off-current (leakage current) to the same level as when the channel length is lengthened. An object of the present invention is to provide a thin film transistor having a structure in which the value of 0 is reduced below a certain value.

[Means for solving problems]

In order to solve the above-mentioned problems, the present invention has been developed by forming one or more holes in the gate electrode plane pattern using exposure technology or by making the plane pattern structure of the gate electrode into a blind shape. By doing this, we can calculate the channel length of each part of the channel region where the gate/electrode extends over the channel region, that is, the distance from each end of the channel region to the outermost hole or notch, and the distance between adjacent holes or notches. The effect of the substitution method can be obtained (for example, 50
(μm or less).

As a result, it is possible to provide a thin film transistor that has a large on-current and can suppress an off-current to a predetermined value or less.

[Effect]

In the planar pattern of the gate insulating film, the gate electrode is
By forming a part (hole or notch) that serves as a hydrogen entry path, hydrogen combines with the dangling bonds of the channel-forming polycrystalline silicon directly below the gate, reducing the defect level and increasing the on-current. improves. In this way, the on-current is improved without shortening the channel length of the semiconductor element, so the off-current is also kept small.

The mechanism of the relationship between the channel length and the effect of hydrogen replacement treatment, which is the basis for demonstrating the effectiveness of the present invention, will be explained.

Characteristic C shown in FIG. 3 was obtained as a result of exposing the thin film transistor having the structure shown in FIG. 2 to hydrogen plasma. Changes in channel length and properties are related to the entry path of hydrogen in the plasma.

FIG. 5(,)(b) shows the entry paths of hydrogen in conventional thin film transistors having different channel lengths.

The relationship between the channel lengths La and Lb is La ) Lb . As shown by the arrow in the figure, hydrogen in a plasma state passes from the interlayer insulating film 7 made of, for example, a phosphor glass film, through the gate insulating film 5 made of, for example, 8i02, and then enters the polycrystalline silicon forming the channel. reaches the polycrystalline silicon near the interface between channel region 2 and gate insulating film 5.

This hydrogen combines with dangling bonds of polycrystalline silicon to form silicon-hydrogen bonds, thereby reducing the defect density at this interface. This increases the on-current.

In this case, since the hydrogen diffusion length is constant, Figure 5 (b
), hydrogen is substituted throughout the interface, and the on'lt flow is significantly improved.

On the other hand, in a device with a long channel length as shown in the same figure (,),
However, since hydrogen is not replaced and a region with high defect density remains, this impairs the potential barrier to electrons traveling from the source to the drain, so that the on-current flow is not improved.

That is, in a device with a long channel length, since the gate electrode 6 acts as a barrier to hydrogen entry, the on-current does not improve much even by hydrogen replacement treatment.

Generally, gate electrode 7 is formed of polycrystalline silicon, similar to channel region 2. However, when the present inventors analyzed using infrared analysis, only a few hundred 8j-)1 bonds were formed in the depth direction from the surface of polycrystalline silicon, and the thickness of the gate electrode 6 was reduced to 4000 mm. In the device used for measuring the characteristics shown in FIG. 3, in which the thickness of the channel region 2 was 4000 mm, it was also found that the on-current did not improve when the channel length was long. Further, with the current technology, it is a national problem to reduce the thickness of the gate electrode 6 to several hundred amps or less.

〔Example〕

An embodiment of the present invention will be described below with reference to FIG. 1(a) is a planar pattern of the present invention, and FIG. 1(b) is a sectional view taken along the line Y-Y' in FIG. Note that x −
The cross-sectional view along the x' line is the same as FIG. 2(b).

In FIG. 1, 1 is a substrate made of glass or quartz, 2 is a channel region made of polycrystalline silicon, and 3 is a substrate made of glass or quartz.
．． Reference numeral 4 denotes a source region and a drain region doped with phosphorus, antimony, or boron.

Further, 5 is a gate insulating film formed of StO, etc., 6 is a gate electrode formed of polycrystalline silicon doped with phosphorus, antimony, or boron, or metal;
7 is an interlayer insulating film formed of phosphor glass or the like, 8 is an aluminum electrode, and 11 is a hole made (for example, by exposure technology) in the gate electrode.

The feature of this embodiment lies in the planar pattern of the gate electrode 6, specifically in that a hole 11 is formed in the gate electrode 6.

The operation of the semiconductor device of this example will be explained below.

The voltage VG8 between the gate electrode 6 and the source region 3 is set to zero potential (
VO2-OV), and in the off state when the potential between the source region 3 and drain region 4 is set to a positive potential (Vsn, >ov), an off-state current (leakage current fi) that depends on vsD flows.

This off-iE current is inversely proportional to the channel length, as seen in FIG. This off-state current depends on the resistivity of the channel region. Therefore, even if the defect density of the polycrystalline silicon near the interface between the channel region and the gate insulating film 5 is reduced by the hydrogen substitution treatment, the off-state current flows through the entire channel region 2, so the off-state current does not become large.

That is, by designing the channel length to be longer than a specified off-state current based on, for example, FIG. 3, the off-state current can be suppressed to below a desired value.

Furthermore, considering the case of a mouth-type field effect transistor in which the source region 3 and drain region 4 are doped with phosphorus or antimony, a positive potential (YGs, >0) is applied to the gate electrode 6 with respect to the source region 3. When applied, the device turns on.

When the device is in the on state, electrons are induced at the interface between the channel region 2 and the gate insulating film 5, and an inversion layer channel having a thickness of 100 Å or less is formed. This lowers the surface resistance and allows a large on current to flow depending on the value of the gate voltage.

In the device having the structure shown in FIG. 1, by providing a hole 9 in the gate electrode 6, the minimum dimension Lt in which the gate electrode 6 extends continuously over the channel region 2 can be set as a dimension with which the hydrogen replacement treatment is highly effective. can be designed.

By designing in this way, the defect density can be reduced over the entire interface between the channel region 2 and the gate insulating film 5, so that the on-current can be greatly improved.

In this case, if the gate IK electrode is not formed continuously from the source side end to the drain side end on the channel region due to the hole that promotes hydrogen intrusion, and is even partially cut off in between, it is obvious that the Since the cross-sectional structure of is not an MO8 type structure, no channel is formed here, and the on-current is significantly reduced.

As described above, the present structural element can provide a thin film transistor with sufficiently large on-state current and small off-state current.

In the present invention, the shape (for example, circular, hexagonal) and size of the hole 9 of the gate electrode, the number of holes, etc. are not particularly specified, but they may be designed in accordance with the current value necessary for driving the liquid crystal. According to the experiments conducted by the present inventors, Li in Fig. 1 was changed to 5
The effect was obtained when the temperature was lower than 0 μmal.

Another embodiment of the invention is shown in FIG. In this embodiment, instead of making a hole on the gate electrode 6 as in the first embodiment, a cut is made in the gate electrode 6 to encourage hydrogen to enter the channel region 2 and increase the on-current. This is an example.

It will be easily understood from the above description that the characteristics can be improved by designing the widths r and t of the gate 141 pole 6 in the channel direction to dimensions that provide a hydrogen substitution effect (for example, 50 μm or less).

In the above description, no particular mention is made of the cross-sectional structure of the element, but the cross-sectional structure of the element can also be formed as shown in FIG. 7.

The device structure shown in FIG. 7 can also be formed, for example, by ion-implanting phosphorus, antimony, etc. into the gate electrode 6, source region 3, drain region 4, and buried IS region 12 in a self-aligned manner. . Thereafter, using the hole in the gate electrode 6, hydrogen replacement treatment as described above is performed.

In this case, if the impurity concentration is sufficiently high (and the resistance of the buried layer region 12 is lower than the resistance of the inversion layer when the transistor is turned on), the current between the source region 3 and the drain region 4 flows through the buried layer. Since it flows throughout the prefecture, the effective channel length is 4Li.

Therefore, if the elements have the same dimensions, the on-current can be increased, while if the on-current is the same, the element dimensions can be reduced.

〔Effect of the invention〕

According to the present invention, it is possible to realize a planar pattern structure of a gate electrode with a high hydrogen replacement treatment effect, and the thin film semiconductor has both a low off-current and a large off-current, and is particularly suitable as a thin film transistor for driving an active matrix panel. It becomes possible to obtain the element.

[Brief explanation of the drawing]

FIG. 1 (, Xb) is a plan view of an embodiment of the present invention and a sectional view taken along the line Y-Y', and FIG. 3 is a drain current characteristic diagram of a conventional thin film semiconductor element, FIG. 4 is a circuit diagram of an active matrix, and FIG. 5 (Xb) is a sectional view for explaining the hydrogen substitution effect. FIG. 6 is a plan view showing the gate electrode pattern of the second embodiment of the present invention, and FIGS. 7(a) and 7(b) are plan views and cross sections along the line Y-Y' of the third embodiment of the present invention. It is a diagram. 1... Insulating substrate, 2... Channel region, 3...
Source region, 4...Drain region, 5...Gate insulating film, 6...Gate electrode, 11...Hole on gate electrode, 12...Buried layer agent Patent attorney Michihito Hiraki Figure 1 (b) Figure 2 Figure 3 Figure 1O-1100101 W/L Figure 4 Figure 5 Figure 6

Claims (3)

[Claims] (1) In an amorphous or polycrystalline silicon semiconductor layer formed on an insulating substrate, a source region and a drain region, which are arranged opposite to each other with a channel region in between and have a first conductivity type; In a thin film semiconductor device in which a gate electrode is formed on a channel region between drain regions via a gate insulating film, in a planar pattern of the gate electrode, at least one hole or notch is provided in the gate electrode, and the above-mentioned A thin film semiconductor device, characterized in that the gate electrode is configured to be continuous from a source side end to a drain side end on a channel region. (2) The distance from each end of the channel region to the outermost hole or notch, and the interval between adjacent holes or notches, are selected to dimensions that allow the effectiveness of the hydrogen replacement method. A thin film semiconductor device according to claim 1. (3) Hydrogen is allowed to enter the channel region through the hole or notch to form a silicon-hydrogen bond.
The thin film semiconductor device described in .