JPS61234072A - Solid-state image sensor - Google Patents

Abstract

PURPOSE:To suppress the decrease in the sensitivity of each photodiode when coloring and the deteriorating in the spectral sensitivity characteristic by alternately varying the conductive types from the main surface of a semiconductor substrate in the depthwise direction of the substrate, and regularly arranging photodiode arrays of different regions. CONSTITUTION:Since an N-type region 27 for forming red (long wavelength light) diode is formed extremely thickly to 6 to 10mum, photoexcited most charge is detected in a diffused region to become a signal charge to remarkably reduce the charge absorbed to an N-type silicon substrate 1. In a green light photodiode, the width of an N-type region 17 is slightly smaller than a red light N-type region to the value (3 to 6mum) matched to the absorption coefficient. In other words, the light of an intermediate wavelength range is sufficiently detected, but long wavelength light range is excluded by the substrate 1. Thus, the spectral sensitivity characteristics of the photodiodes necessary to color can be controlled by regulating the width of the N-type diffused region layer.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a color solid-state imaging device that can be used in video cameras and the like.

(Prior Art) In recent years, the performance of solid-state image sensing devices has improved markedly, and they are reaching the stage of practical use, especially in home color video cameras.

FIG. 3 shows a schematic cross-sectional view of a pixel in a conventional solid-state image sensor. Each pixel corresponds to one that is sensitive to blue light, green light, and red light when colored. In FIG. 3, 1 is an n-type silicon substrate, 2 is a p-type region, 3 is a silicon dioxide film, and 4.14.24 is a p-type silicon substrate.
n-type region forming an n-junction photodiode, 5, 15
．． 25 is an n-type region forming a drain portion connected to an output line, and 6°16.26 is a polycrystalline silicon film forming a switching gate for signal readout. The p-type region 2 has an impurity concentration on the order of 10"am-" and is formed uniformly within the device with a junction depth of several micrometers.

The operation of the unit pixel of the pixel array configured as described above will be explained.

FIG. 4 shows the electron potential immediately after the signal is read out along the line AA' shown in FIG. Note that the circles in the figure indicate the charges generated by the incident light, and the arrows indicate the movement of the charges. A pn junction photodiode is formed in a state where a positive bias voltage Vmmh is applied to an n-type substrate 1.

The light incident through the n-type region 4 is photoelectrically converted and generates charges, which are separated into two directions depending on the position where the charges are generated. In other words, the charges generated on the surface side of the maximum electron potential in Figure 4 move as shown by the arrow (A) and are accumulated in the photodiode, becoming signal charges, but the charges generated deeper than that move as shown by the arrow (A). b), and move the n-type substrate 1 as shown in
It will be absorbed into the cell and will not become a signal charge. Arrow (a)
The charges moved as shown in the figure are accumulated in the pn junction capacitance for a certain period of time, and then read out to the drain section 5 by the switching gate 6 and become a signal current. Here, when used as a color element, each photodiode 4.14 and 2
4 detects optical signal currents each having a predetermined wavelength range.

(Problems to be Solved by the Invention) In the above configuration, the width of the active layer where photoelectric conversion is performed is the same for each photodiode, resulting in imbalance in sensitivity characteristics for light in each wavelength range. It had its drawbacks. This is because the optical absorption coefficient within silicon has wavelength dependence.

This is because in the case of long-wavelength light, photoelectric conversion occurs deep in the silicon substrate, so a portion of the charge is absorbed by the silicon substrate and does not contribute to the signal current.

SUMMARY OF THE INVENTION An object of the present invention is to provide a color solid-state image pickup device that can eliminate the conventional drawbacks and suppress a decrease in sensitivity of each photodiode and deterioration of spectral sensitivity characteristics when colorizing.

(Means for Solving the Problems) The solid-state image sensing device of the present invention alternately changes the conductivity type from the main surface of the semiconductor substrate in the depth direction of the semiconductor substrate, so that the conductivity types are sequentially changed from the main surface of the semiconductor substrate to the first region and the second region. the first region and the second region.
A photodiode is constituted by a region of , and photodiode rows of different regions are regularly arranged.

(Function) In the solid-state image sensor with the above configuration, most of the light components incident on the photodiodes assigned to each hue become signal charges, so this reduces the sensitivity of the solid-state image sensor as a whole and the deterioration of the spectral sensitivity characteristics. Significant improvements are possible.

(Example) An example of the present invention will be described based on FIGS. 1 and 2.

FIG. 1 shows a schematic cross-sectional view of three pixels of the solid-state image sensing device of the present invention. In this figure, the same parts as in FIG. 3 are given the same reference numerals, and their explanations will be omitted.

The thicknesses of n-type regions 4, 14 and 24 in FIG. 3 are adjusted to have the spectral characteristics assigned to each photodiode in FIG. 1, respectively.
27.

FIGS. 2(a) and 2(b) show potential distributions immediately after signals are read out along lines BB' and cc' shown in FIG. 1. The circles and arrows in the figure indicate the movement of charges as in the case of FIG. 4. Although the light incident through the n-type regions 7, 17, and 27 generates charges, only the charges generated in the p-type head region that move toward one surface become signal charges, as shown in Figure 3. This is the same as the conventional configuration shown in .

However, in this embodiment, the number of n-type regions 27 forming a diode for red light (long wavelength light) is 6 to 1.

Since it is formed extremely thick, .mu.m, it has a potential distribution as shown in FIG. 2(b). Therefore, most of the optically excited charges are detected within the diffusion region and become signal charges, so that the charges absorbed into the n-type silicon substrate 1 are significantly reduced. In the photodiode for green light, the width of the n-type region 17 is slightly smaller than that of the n-type region for red light, and has a value (3 to 6 μm) that matches the absorption coefficient. That is, although light in the intermediate wavelength range is sufficiently detected, long wavelength light components are excluded by the n-type silicon substrate 1. Also, B-B
The potential distribution of the short wavelength n-type region 7 indicated by
As shown in Figure (a), this is almost the same as the conventional configuration shown in Figure 4, in which a potential well is formed shallowly from the surface, and the charge excited by light with a long wavelength in the intermediate wavelength range or above is The shape is such that it is diffused and absorbed into the silicon substrate.

As described above, according to this embodiment, the photoelectric conversion characteristics of each photodiode necessary for colorization are maximized, and the overall device has high sensitivity and excellent spectral sensitivity characteristics.

In this example, the silicon substrate was of n type, but P
A similar effect can be obtained by using a pnp configuration. Further, although the photodiode is a pn junction formed in a silicon substrate, similar effects can be obtained with a structure made of amorphous silicon or a compound semiconductor.

Further, the n-type region whose thickness changes in the depth direction can be easily formed by an ion implantation method or an epitaxial method.

(Effects of the Invention) According to the present invention, the spectral sensitivity characteristics of each photodiode necessary for colorization can be controlled by adjusting the width of the n-type diffusion region layer, and the filter configuration for color elements is facilitated. There is. In extreme cases, the possibility of using it as a filterless color element can be fully expected.

[Brief explanation of the drawing]

Figure 1 is a schematic cross-sectional view of three pixels of a solid-state image sensor according to an embodiment of the present invention, Figures 2 (a) and (b) are potential distribution diagrams in the same depth direction, and Figure 3 is a diagram of the conventional FIG. 4 is a schematic cross-sectional view of a pixel row of a solid-state image sensor, and is a potential distribution diagram in the same depth direction. 1... N-type silicon substrate, 2... P-type head area, 3... Silicon film, 4, 14, 24.7,
17.27...n-type region, 5,15.25...
n-type drain region. 6.16.26...Polycrystalline silicon film. Patent applicant: Matsushita Electronics Co., Ltd. Figure 1 Figure 2 (a) (b) Figure 3 A' Figure 4

Claims (1)

[Claims] The semiconductor substrate has a first region and a second region whose conductivity types are alternately changed in the depth direction of the semiconductor substrate from the main surface of the semiconductor substrate, and the first region and the second region are photosensitive. A solid-state imaging device characterized by regularly arranging photodiode rows that constitute diodes and have regions with different thicknesses.