JPS58180063A - Thin film transistor - Google Patents

Abstract

PURPOSE:To markedly reduce the OFF current of the titled transistor without an ON current by a method wherein a plurality of transistors are connected in series, the electrode located at both ends of said transistors are forme common as a source electrode and a drain electrode, and the specific transistors are euqally formed in shape. CONSTITUTION:The titled thin film transistor consists of a source S, a drain D, a gate G and N pieces of thin film transistors to be connected in series. The gates of the N pieces of thin film transistors are made into a gate by forming them common. Besides, the thin film transistor of i (i=1, 2...N) and (N-i+1) are equally formed in shape. According to this constitution, transistors can be arranged symmetrically with the source and the drain, thereby enabling the transistors to have their characteristics unchanged even when the source is replaced with the drain.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a semiconductor thin film transistor having a structure that reduces leakage current between a source and a drain.

In recent years, research on forming thin film transistors on insulating substrates has been actively conducted. This technology is used for active matrix panels that create thin displays using inexpensive insulating substrates, three-dimensional integrated circuits that form active elements such as transistors on ordinary semiconductor integrated circuits, and inexpensive, high-performance image sensors. It is expected to have many applications, such as , high-density memory, etc. Hereinafter, a case where a thin film transistor is applied to an active matrix panel will be described as an example, but the present invention can be similarly applied to other cases using thin film transistors. This is because the gist of the present invention is to reduce leakage current, which is an essential improvement in the characteristics of thin film transistors.

A liquid crystal display button material when a thin film transistor is applied to an active matrix panel is generally composed of an upper glass substrate, a lower thin film transistor substrate, and a liquid crystal sealed in the nR. By selecting the liquid crystal drive elements arranged in a matrix on the top by an external selection circuit and applying a pressure of 1 µm to the liquid crystal drive electrodes connected to the liquid crystal drive elements, any character, figure, or image can be displayed. This is what we do. A general circuit diagram of the thin film transistor substrate is shown in FIG. 1.

FIG. 1(α) is a matrix layout diagram of liquid crystal driving elements on a thin film transistor substrate. The area surrounded by 1 in the figure is the display area, and the liquid crystal drive elements 2 are arranged in a matrix within the area. 3 is a data signal line to the liquid crystal driving element 2, and 4 is a timing signal line to the liquid crystal driving element 2. The circuit diagram of the liquid crystal drive element 2 is shown in the first
Shown in Figure (b). A thin film transistor 5 performs data switching. 6 is a capacitor, which is used for holding data signals. The capacity of this capacitor includes the capacity of the liquid crystal itself and the capacity of an intentionally provided capacitor, but in some cases, it may be made up of only the capacity of the liquid crystal. 7 is a liquid crystal panel;
-1 is a liquid crystal drive electrode formed corresponding to each liquid crystal drive element, and 7-2 is an upper glass panel.

As can be seen from the above description, thin film transistors are used to switch voltage data applied to liquid crystals, and the characteristics required of thin film transistors at this time are broadly classified into the following two types.

(i) Sufficient current must be able to flow to charge the capacitor when the thin film transistor is turned on.

(2) When the thin film transistor is put into the 0IFF state,
Make sure that no current flows as much as possible.

(1) relates to the characteristics of writing data to the capacitor. Since the display of the liquid crystal is determined by the potential of the capacitor, it is not possible to completely write data in a short period of time. Must not be. At this time, the current is below C, O
It is called N current. ) is determined by the capacitance of the capacitor and the write time, and thin film transistors must be manufactured so as to clear the 0NTt current. The ON current that can flow through a thin film transistor largely depends on the size (channel length and channel width), structure, manufacturing process, gate voltage, drain voltage, etc. of the transistor.

(The second term relates to the retention characteristics of data written to a capacitor. Generally, the written data must be retained for a much longer time than the writing time. The capacitance of the capacitor is usually about 1 pF. Because this value is small, if even a small amount of leakage current (hereinafter referred to as OFF current) flows when the thin film transistor is in the 0IFF state, the drain potential (that is, the capacitor potential) will rapidly approach the source potential, and the written The data will no longer be retained correctly. Therefore, O
The FF current must be made as small as possible. 0
The mechanism of the flow will be described in detail later as it relates to the gist of the present invention.

As can be seen from the above description, reducing the 0FIF% flow of thin film transistors has very important significance. This is because if an attempt is made to obtain a sufficient ON current by decreasing the channel length and increasing the channel width, the 0FIF current also increases, which deteriorates the data retention characteristics. Therefore, reducing the 071 current is an urgent need for improving the characteristics of thin film transistors. This is exactly the same when thin film transistors are applied to uses other than active matrix panels. For example, when a normal logic circuit is constructed using thin film transistors, static current increases, and when a memory or image sensor rt#l is constructed, it causes malfunction.

The present invention aims to eliminate such drawbacks of conventional thin film transistors, and its purpose is to
An object of the present invention is to provide a thin film transistor having a structure that reduces current. Below, 0? ] After describing the mechanism of the IF current in detail, the present invention will be explained in detail based thereon.

FIG. 2 is a cross-sectional view showing the general structure of an N-channel thin film transistor using a semiconductor thin film.

8 is an insulating transparent substrate such as glass or quartz; 9 is a semiconductor thin film such as polycrystalline silicon; 1o is a source region formed by doping impurities such as phosphorus or arsenic into the semiconductor thin film;
1 is a drain region, 12 is a gate film, 13 is a gate electrode, 14 is an interlayer insulating film, 15 is a source electrode, and 16 is a drain electrode. Typical characteristics of a thin film transistor having this structure are shown in FIGS. 3 and 4.

Figure 1@3 shows channel length = 20μ, channel cap W = 1
2 is a graph showing the characteristics of a thin film transistor having a size of 0μ turtle. Note that this data is the result obtained from experiments conducted by the applicant.

The horizontal axis of this graph is the gate voltage yos relative to the source.
, and the vertical axis is the drain current.

The parameter is the drain voltage VD8 with respect to the source, and the curve A is VDl=IV and the curve B is yns==
47, the curve a corresponds to 7xz+=8V, respectively. As you can see, yos is lower than the drain current.
It takes a minimum value near =QV, and increases more than the drain current as the absolute value of yes increases. 7G11
An increase in the drain current in a region where is positive means that the transistor changes from the OIF]F state to the ON state, and it is desirable that the current increase rate be as large as possible. On the other hand, the fact that the drain current increases in the region where yon is negative means that the 071 current has gate voltage dependence, which is not desirable as a transistor characteristic. Furthermore, the drain current varies greatly depending on the drain voltage vD1. Especially 7Gfl
The drain current in the region where is negative, i.e. OFF
The current has a greater drain voltage dependence than the ON current.

FIG. 4 is a graph showing the channel length dependence of the characteristics of a thin film transistor with a channel width W of 10 μm. In addition,
This data is also the result of experiments conducted by the applicant. The drain voltage is constant at yfil=4v, and the parameter is the channel length. D song 1i1# (L=10
μ~, the curve of E is L=2Dμ poison, and the 1# line of A is L=
For the 4Ωμ bait, the G curve corresponds to L=100μ bait, respectively. As can be seen from this, in the positive region, 'VGJI is inversely proportional to the channel length rather than the drain current, and is
S Fl! This is consistent with the theory of iT). However, Vfg
In the region where is negative, as the absolute value of VOS increases, the dependence on the channel length decreases until L
dependence is completely eliminated. That is, VO2 is about -8v
In the following cases, the 0FIP current becomes constant for any L.

From the data shown in Figures 3 and 4, O'1!1! The current is thought to be generated by the following mechanism.

That is, yos = QVkl - o y ? im is determined by the specific resistance of the semiconductor thin film, but when VGll is negatively biased, the oyIF current is generated between the P-type layer induced on the surface of the semiconductor thin film and the N-type layer in the source and drain regions. is defined by the current flowing through the PM junction formed in . Generally, since many traps exist in a semiconductor thin film, this PH junction is incomplete, and therefore junction leakage current easily flows. The reason why the 0FIF current increases as the gate voltage is biased more negatively is because the carrier concentration of the 'P diffused layer formed on the surface of the semiconductor thin film increases and the width of the energy barrier of the PM junction becomes narrower.
This is because electric field concentration occurs and junction leakage current increases. Furthermore, the drain voltage dependence of the 0IFF current is also due to the same reason. Furthermore, the channel length dependence of the oyy current can also be explained by the junction leakage current. That is, as YES is biased more negatively, the 0IFF current is dominated by the junction leakage current near the drain, and the current flowing due to the specific resistance of the semiconductor thin film becomes negligible.

The mechanism of the OFF current is as described above, but in the past, few effective measures have been taken to actually reduce the 017 current. In particular, OF when the gate voltage is negatively biased? In order to reduce the current, it is necessary to reduce the junction leakage current, so little effort has been made to this end.

The present invention reduces the dependence of the 0IFIF current on the gate voltage, and even if VOW increases to a negative value, the OFF
The present invention provides an innovative thin film transistor having a characteristic that current hardly increases.

In order to achieve this, in the present invention, in a thin film transistor using a semiconductor thin film and having a source electrode, a drain electrode, and a gate electrode, N pieces (N22) of the above film transistors are connected in series, and the electrodes at both ends are connected to the source electrode. In addition to making the gate electrodes of the N thin film transistors common, the sixth thin film transistor (i
The present invention provides a thin film transistor characterized in that the shape of the (N-i+1)th thin film transistor is the same as that of the (N-i+1)th thin film transistor.

Hereinafter, the present invention will be described in detail with reference to the drawings.

The 51st Δ shows a general circuit diagram of the present invention. S indicates a source, D indicates a drain, and G indicates a gate. Further, N is an integer of 2 or more and represents the number of thin film transistors connected in series. The numbers in the diagram are
These are the numbers of thin film transistors assigned in order from the source side. Further, the gates of all the N thin film transistors are made common to one gate. Furthermore, the 6th (i =
1.

2...s') thin film transistor and (M-i+
1) The shapes (channel length and channel width W) of the th thin film transistor are equal. That is, the shapes of the first transistor and the Mth transistor are equal, and the shapes of the second transistor and the (N-1)th transistor are equal. The same applies hereafter. Expressed more generally, xri=r, ・N+Key+1 Wj=W・N+Key+1. L) e'l is the channel length and channel width of the first transistor, respectively.The gist of the present invention is:
By treating the thin film transistors as a plurality of thin film transistors having the above-mentioned (2U&'' configuration), extremely excellent transistor characteristics can be achieved.The reason for this will be explained with reference to FIG.

FIG. 6(a) is a schematic diagram when 1i=2 in FIG. 5. For the sake of simplicity, the present invention will be explained using the case where y,=2 as an example. In the figure, s, D.

The meaning of G is the same as in FIG. S, D*G
, 'e1st place tc in x vs, vp, v, respectively
a, 'Iix. - Also, the number in the diagram is 2
This is a number assigned to each thin film transistor, and each transistor has the same channel length and the same channel width W. Further, in FIG. 6Cb), the transistor (α) is isometrically replaced with one transistor, and its channel length is 2L and its channel width is W. Drain voltage FEVDsxs'f- of transistor 1
111 voltage VoaseRO: The drain voltage yngz and gate voltage VGg2 of the transistor 2 are given by the following equation.

VD81; VX-78 Vass=Va-Va VD82=VD-Vx Vos2=Va-'Vx The potential VX at point X is determined so that the current flowing through transistor 1 and the current flowing through transistor 2 are equal. At this time, va (v When a certain relationship is established between the channel length and the channel length, the sixth fjll (
Compared to A), the drain current increases due to the shorter channel length, and as a result, the transistors shown in Figure 6 (α) and Figure 6 (
The current values of the transistor b) are the same.

In fact, in the case of V o -vtr ) O, this relationship holds and the ON current does not change. In other words, the current value does not change even if the channel length is divided.

However, the situation is different when V a −V # (0. This is because, as shown in FIG. 4, when the gate voltage is negatively biased, the dependence of the drain current on the channel length disappears. In other words, as the gate voltage increases in the negative direction, the OFF current becomes independent of the channel length, so the 6th mCa) and (6)
With this, the effect of the difference in channel length disappears. Therefore, in (b), the O1'll current decreases by the amount that the drain voltage applied to each transistor decreases. This effect becomes more pronounced as the gate voltage is biased more negatively.

The above phenomenon can also be explained in terms of physical properties as follows. When a transistor is on, a channel is formed on the surface of the semiconductor thin film, and since a nearly uniform potential gradient (electric field) is generated from the source to the drain, it is difficult to determine how to divide the channel. However, the drain current does not change. On the other hand, when the transistor is off, most of the electric field is concentrated at the PN junction near the drain, so by dividing the transistor,
The electric field concentration applied to each Fli junction can be weakened, and the junction leakage current, that is, OIF? Current can be reduced.

Next 1. The shape of the ε-th transistor and (N −i+1
) The effect of making the shapes of the transistors the same (in the case of FIG. 6, the shapes of the two transistors are made the same) will be described. One of the characteristics of a typical field effect transistor is that it has symmetrical characteristics with respect to the source and drain. In other words, even if the source and drain are swapped, the transistor characteristics do not change.

This is particularly important in applications where the source and drain of the transistor are interchanged over time, such as when the transistor is used as a switching element in an active matrix panel. In such a case, if the transistor characteristics, especially the oyy characteristics, change due to the swapping of the source and drain, this will cause variations in the characteristics of each transistor, and the system as a whole will not be able to obtain sufficient characteristics. Therefore, it is required that the transistor characteristics do not change even if the source and drain are interchanged. The shape of the first transistor and (N − i −) 1
) The reason for making the shapes of the transistors the same is to satisfy such a requirement. That is, by adopting the above configuration, the transistor can be arranged symmetrically with respect to the source and drain, so that the transistor characteristics do not change even if the source and drain are interchanged.

Next, experimental data will be shown to demonstrate the effects of the present invention.

FIG. 7 is a graph showing the characteristics of the thin film transistor according to the present invention. In Figure 6 (α), L, = L, = 1
These are the transistor characteristics when 0μII, w, = w, = 10μ. This transistor is isometrically equivalent to the transistor shown in FIG.

Note that this data is also the result of experiments conducted by the applicant. The parameter is the drain voltage, the H curve becomes 'Vbs=IV, the engineering curve becomes Vna=4V,
The curves of J correspond to yBa=SV, respectively. As can be seen from this graph, the region where vGII is positive, that is, the ON current, almost matches the data in Figure 3, but the region where yos is negative, that is, the OFF current, is
This is significantly different from the figure, and the value remains almost constant at a low value. That is, while maintaining the same ON current as conventional thin film transistors, the 01P1P current is significantly reduced. The applicant also performed a combinatorial simulation based on the conventional transistor characteristics and calculated the OFF characteristics of the thin film transistor according to the present invention, and the results were very similar to the graph in FIG. Agreed.

In the above explanation, for simplicity, when N=2, that is, 2
Although the case where two thin film transistors are connected in series has been described, the same explanation can be given to the case where three or more thin film transistors are connected in series. When the number of thin film transistors connected in series is increased, the loading of the 011 current becomes noticeable when the drain voltage is high. This is because the drain voltage applied to each transistor decreases as the number of transistors increases. Therefore, the number N may be selected depending on the purpose of the thin film transistor and the required level of OFF current. When applied to an active matrix panel, N=2 to 3 is usually sufficient because the drain voltage is low (approximately 10 V or less). When configuring a logic circuit using thin film transistors, a high gate voltage is usually applied to obtain a sufficient ON current, but the drain voltage is also nearly as high, so N
Is it better to increase the value of OIF? It is effective in reducing the current.

8FA is a graph showing the effect of the present invention, and FIG.
In the thin film transistor configured as in α), the oyy characteristics are shown when the shapes of the two thin film transistors are changed in various ways. Figure 8 (a) may be the shape of transistor 1, where □ = 5μ, W = 10μ, transistor 2.
Shape of = 15μ, W, = 01 when IQμ
! 7 characteristics, and FIG. 8(b) shows that the shapes of the transistors 1 and 2 are the same, L, = L, = 10μ a "1 = W
It shows the oypy characteristics when R=10μ. In addition, the solid line graph in the figure shows the case where the source and drain positions are as shown in the 61st Te (α), and the broken line graph shows the case where the source and drain positions are swapped. It shows. As shown in FIG. 8(α), when two transistors have different shapes (channel lengths in this case), the transistor characteristics change significantly by replacing the source and drain. On the other hand, the 8th
As shown in FIG. ) is an exact match. These data demonstrate the effects of the present invention with respect to the channel lengths of the two transistors, but the effects of the present invention are similarly demonstrated with respect to the channel widths.

FIG. 9 shows the transistor characteristics when three thin film transistors are connected in series and the shape of each transistor is changed. Figure 9(a) shows that the shape of transistor 1 is L1 = 5μ, 1 = 1 (0μ), shape force L, = 20μ, W, = 10μ, and transistor 3. Shape, = 15μ, W, = 10
Figure 9 (h) shows the 011 characteristic at μ habituation, where the shape of transistor 1 is 10 μs g W 1 =
10 μs, shape of transistor 2 = 20 μs.

W, ==1Oμ, shape of transistor 3, =10
071 characteristics when μs, W1=10 μm. Further, the solid line graph and the broken line graph in the figure show the characteristics when the source and drain are replaced. 9th
As shown in Figure (cL), when transistors 1 and 3 have different shapes (channel lengths in this case), the transistor characteristics change significantly by replacing the source and drain. On the other hand, as shown in FIG. 9Cb), if transistors 1 and 3 have the same shape as in the present invention, even if the shape of transistor 2 is different, the source and drain can be interchanged. The transistor characteristics do not change, and the two songs # (
The solid line and the broken line 41) completely match. Of course, the same applies even if all three transistors have the same shape.

These data demonstrate the effects of the present invention with respect to the channel lengths of the six transistors, but the effects of the present invention are similarly demonstrated with respect to the channel widths.

Using FIG. 8 and FIG. 9, it is useful to make the shape of the first transistor and the shape of the (M − i +1)th transistor equal when two or three thin film transistors are connected in series. However, N
In the case of ≧4, its usefulness has been confirmed in exactly the same way.

As described above, the present invention is a breakthrough that has the excellent effect of significantly reducing the oyy current without reducing the ON current, and not causing any change in transistor characteristics even if the source and drain are replaced. The purpose of this invention is to provide a thin film transistor with a unique structure.

[Brief explanation of the drawing]

FIG. 1 is a general Iw1 diagram when a thin film transistor is applied to an active matrix panel. FIG. 2 is a cross-sectional view showing the general structure of a summer channel thin film transistor using a semiconductor thin film. FIGS. 3 and 4 are graphs showing the characteristics of conventional thin film transistors. FIG. 5 is a circuit diagram showing the general configuration of the present invention. FIG. 6 shows, as an example of the present invention, a circuit diagram in which two thin film transistors are connected in series, and a single thin film transistor equivalent thereto. FIG. 7 is a graph showing the characteristics of the thin film transistor according to the present invention shown in FIG. FIG. 8 is a graph showing the effects of the present invention when two thin film transistors having various shapes are connected in series. FIG. 9 is a graph showing the effects of the present invention when three thin film transistors having various shapes are connected in series. Applicant Suwa Seikosha Co., Ltd. Agent Patent Attorney Tsutomu Mogami Figure 1 Figure 3 Vt, s (vo It) VGS (volt) Figure 5%s (volt) Figure 8 VGs Cvo1f') ”-5 ( volt) %s (voIf)

Claims (1)

[Claims] In a thin film transistor using a semiconductor thin film and having a source electrode, a drain electrode, and a gate electrode, N (N≧2
) are connected in series, the electrodes at both ends thereof are used as a source electrode and a drain electrode, and all the N thin film transistors have a common gate electrode, and further, counting from the source electrode side or the drain electrode side. A thin film transistor characterized in that a t-th (i=1.2..., N) thin-film transistor and a (N-i+1)-th thin film transistor have the same shape.