JPH04196589A - Optical semiconductor device - Google Patents

Abstract

PURPOSE:To reduce the occurrence of cross talk phenomena by providing areas in which an impurity of the same type as that of a semiconductor substrate is selectively diffused at high concentrations and another areas in which another impurity forming a conductivity type opposite to that of the substrate is selectively diffused to the surface of the substrate opposite to the photodiode forming surface of the substrate. CONSTITUTION:This optical semiconductor device 4 is provided with a P-type diffusion layer 3 formed by selectively diffusing boron as a P-type impurity at aimed positions on an N-type Si substrate 1 and a rear Au electrode 2 formed on the rear surface of the substrate 1. Most of the carriers generated in photodetector elements 4 irradiated with light flow to the electrode 2, since no N/N<+> interface exists in the element 4. As a result, no carrier flows to photodetector elements which are not irradiated with light and the occurrence of cross talk phenomena can be reduced. Since a P layer and N<+> layer coexist at an appropriate rate on the rear surface of the device 4 and a diffused carrier absorbing zone is formed on the rear surface of this optical semiconductor device 4, an abnormal increase in photocurrent can be prevented.

Description

Detailed Description of the Invention [Industrial Application Field] The present invention relates to optical semiconductors such as split photodiodes and phototransistor arrays, which are formed by arranging one or more light receiving elements on a semiconductor substrate. Regarding equipment.

[Prior art] Optical semiconductor devices such as photodiodes include those in which almost the entire surface of the substrate is used as an effective light-receiving surface, as well as those in special forms in which a part of the substrate is used as an effective light-receiving surface, and those in which the effective light-receiving surface is formed on a part of the substrate. Some have multiple photodiodes formed on the substrate. However, these special forms have special problems that the general forms do not have.

First, in the case where a part of the substrate is used as an effective light-receiving surface, carriers generated when light is irradiated onto a part other than the effective light-receiving surface that should originally be a dead zone diffuse and flow into the effective light-receiving surface. Therefore, there is a problem in that a photocurrent exceeding the amount of light received by the effective light receiving section is generated. In this case, there is a tendency that the larger the area ratio of the dead zone to the area of the effective light-receiving portion is, the larger the amount of increase in photocurrent is.

Second, when multiple photodiodes are formed on the same substrate, the output of one photodiode is left open between adjacent photodiodes, and the photocurrent of the other photodiode is reduced. When measured, carriers generated in an open photodiode diffuse and flow into the other photodiode, causing a crosstalk phenomenon in which an output is generated even in the light receiving element where no light is incident. .

As a technology to solve such problems,
There is a technique described in Japanese Patent No. 74350.

This technology selectively diffuses impurities that form a conductivity type opposite to that of the semiconductor substrate near the diffusion surface that constitutes the photodiode, and shorts the formed diffusion surface and the semiconductor substrate. In this technique, a diffused carrier absorption band is provided to reduce the flow of diffused carriers from other regions of the same substrate into the photodiode.

[Problems to be Solved by the Invention] However, in the above technology, the photocurrent due to diffused carriers flowing into the photodiode is reduced to 20 to 1 % of the original signal photocurrent.
It can only be suppressed to about 0% or less.

The reason for this will be explained next.

For example, when forming a diode array on an N-type semiconductor substrate, since the semiconductor substrate has a common electrode,
An N-type impurity (such as phosphorus) is diffused at a high concentration on the back surface of the N-type semiconductor substrate opposite to the diode formation surface to form an N-type semiconductor layer, which is in contact with the back electrode.

In such a case, since the N-type semiconductor layer exists on the back surface, carriers generated by light reception can diffuse over a longer distance due to the existence of the N/N interface, and as a result, the diffused carrier absorption band covers all the diffused This is because carriers cannot be absorbed, and the amount of light reaching the light-receiving element portion where no light is incident is thought to increase.

SUMMARY OF THE INVENTION An object of the present invention is to provide an optical semiconductor device that does not generate a photocurrent exceeding the amount of light received unique to an effective light receiving part and reduces crosstalk phenomena.

[Means for Solving the Problems] The above object is to provide a region on the surface opposite to the surface of the semiconductor substrate on which the photodiode is formed, in which impurities of the same type as that of the substrate are selectively diffused at a high concentration, and a region where impurities of the same type as the substrate are selectively diffused. The non-conductive region includes a region in which impurities forming a conductivity type opposite to that of the substrate is selectively diffused, and an electrode formed in contact with the high concentration selective diffusion region and the impurity selective diffusion region. This can be achieved by an optical semiconductor device comprising a thin film.

[Operation] As described above, by diffusing N-type impurities at a high concentration on the back surface of the N-type semiconductor substrate, which is the opposite side to the diode formation surface,
It is an N-type semiconductor layer and is in ohmic contact with the back electrode.

However, the inventor did not form a back N-type semiconductor layer,
We have experimentally found that crosstalk can be reduced by using an electrode that is alloyed with pure Au deposited directly on the substrate.

Further, in order to reduce crosstalk between light receiving elements, it is sufficient to promptly absorb unnecessary carriers that diffuse. For this purpose, the present inventor thought that the crosstalk phenomenon would be reduced by forming an unnecessary carrier absorption band on the back surface of the wafer.

However, since it is necessary to take out the common electrode from the back side,
The entire back surface cannot be used as a carrier absorption layer.

Therefore, it has been experimentally found that in order to satisfy both conditions, it is sufficient to mix the P layer and the N layer in an appropriate ratio.

When the P layer and the N layer are mixed in an appropriate ratio to form a diffused carrier absorption band on the back surface, carriers generated in the dead zone flow into this diffused carrier absorption band,
Flows to the board through a short circuit path. As a result, it is possible to prevent these carriers from flowing into the light receiving element portion and causing an abnormal increase in the original photocurrent.

[Example] Next, an example of the present invention will be described with reference to the drawings.

A first embodiment of the present invention will be described with reference to FIG.

This figure is an enlarged cross-sectional view of an optical semiconductor device according to one embodiment.

As shown in the figure, the optical semiconductor device 4 of this embodiment includes a P-type wave dispersion layer 3 formed by selectively diffusing boron as a P-type impurity at a target position on an N-type Si substrate 1. and a rear Au electrode 2 formed on the lower surface of the Si substrate.

A method of manufacturing this optical semiconductor device 4 will be explained.

First, a Si wafer with a specific resistance of 100 Ωl and a thickness of 400 μm is used as an N-type Si substrate.

A wet oxide film with a thickness of 3,000 wafers is formed on the surface of this Si wafer, and the surface oxide film is etched into a desired pattern.

Next, boron is thermally diffused to a depth of 1 μm to form a PN junction on the main sensitive surface.

Next, the oxide film on the back surface of the wafer is removed by etching.

Next, phosphorus is thermally diffused to a depth of approximately 1.5 μm over the entire back surface of the wafer to form an N layer. Thereafter, the formed back N layer is removed by etching using fluoronitric acid. Note that this back surface N layer does not have to be formed from the beginning.

Next, contact holes are formed at desired positions on the wafer surface.

Next, in order to form electrodes, high-purity gold is deposited on the entire surface of the wafer, and aluminum is deposited to a thickness of 1 μm on the front surface. Then, etch it into the desired pattern.

Next, an alloying treatment is performed at approximately 500° C. for 10 minutes in a N2 atmosphere.

Note that, after gold is deposited, nickel may be further deposited as the back electrode.

The operation of the optical semiconductor device 4 having the above configuration will be explained.

The carriers generated in the light receiving element 4 irradiated with light are N/N.
Since there is no interface, most of it flows to the back Au electrode 2. As a result, carriers do not flow to the light-receiving elements that are not irradiated with light, and crosstalk phenomena can be reduced.

Note that specific data regarding the extent to which the optical semiconductor device of this embodiment can reduce the losstalk phenomenon will be explained later together with data regarding the optical semiconductor device of the second embodiment, which will be explained in the fifth section.

Next, a second embodiment of the present invention will be described using FIG. 2.

This figure is an enlarged cross-sectional view of an optical semiconductor device according to one embodiment.

As shown in the figure, the optical semiconductor device 5 of this embodiment has a P-swelled diffusion layer 6 formed by selectively diffusing boron as a P-type impurity at a target position on the surface of an N-type Si substrate 7. , the above Si
A carrier absorption band 8 consisting of a P expansion diffusion layer formed at a target position on the back surface of the substrate, an 8 layer 9 formed adjacent to this carrier absorption band 8, and a back surface of this carrier absorption band 8 and 8 layer 9. and a back electrode 10 formed on the back surface electrode 10 .

The material of this back electrode 10 does not necessarily have to be gold.

Next, as in the above configuration, a carrier absorption band 8 is provided on the back surface,
The reason for the structure in which eight layers 9 are formed will be explained.

Even the optical semiconductor device of the first embodiment in which gold is directly formed on the back surface of the optical semiconductor device has a sufficient photocurrent separation effect.

Further, according to the inventor's experimental results, the diode constant was also good.

However, when conducting repeated experiments with many prototypes,
In all cases, there were times when ohmic contact could not be established. In addition, diode characteristics were sometimes observed between the gold electrode and the silicon substrate.

Therefore, in order to ensure that ohmic contact is established and to have a structure that absorbs unnecessary carriers, the structure of this embodiment was adopted.

Next, a method for manufacturing this optical semiconductor device 5 will be explained using FIG. 3.

First, a Si wafer 7 with a specific resistance of 100 Ω and a thickness of 400 μm is used as an N-type Si substrate.

On both the front and back sides of this Si wafer, a thickness of 3t)O0
Wet oxide films 11a and llb are formed (see the same figure).
1)).

Next, the surface oxide film is etched into a desired pattern, and a resist 12a is applied to protect the patterned surface oxide film. Also, photoresist 1 for etching the SiO2 on the back side into the desired pattern.
2b ((2) in the same figure).

Next, the backside oxide film is etched into the desired pattern.

Next, boron is thermally diffused to a depth of approximately 1 μm on the front and back surfaces to form a P layer, and a main sensitive surface 13 is formed on the front surface and a carrier absorption band 14 is formed on the back surface (see FIG. 3)).

Next, an oxide film 15a is formed on the front and back surfaces again.

15b is formed.

The portions of the oxide film formed on the back surface where boron has not been diffused are removed by etching.

Next, phosphorus is thermally diffused to a depth of about 1.5 μm on the back side.
There are 8 layers 16 ((4) in the same figure).

Next, a contact hole 17 is inserted into the target position of the surface oxide film.
At the same time, the oxide film on the back surface is completely removed ((5) in the same figure).

Next, high-purity gold is vapor-deposited on the back surface, and the back Au electrode 1,
19. In addition, the surface is coated with approximately 1 μm of aluminum.
is deposited to a thickness and etched into a desired pattern to form the front AI electrode 18.

Next, an alloying treatment is performed at approximately 500° C. for 10 minutes in a N2 atmosphere ((6) in the same figure).

Note that, after gold is deposited, nickel may be further deposited as the back electrode.

Next, the operation of the optical semiconductor device 5 having the above configuration will be explained.

On the back surface of the optical semiconductor device having the above configuration, the P layer and the N layer are mixed in an appropriate ratio to form a diffused carrier absorption band, so that the carriers generated in the dead zone are absorbed by this diffused carrier absorption band. band and flows to the substrate through a short circuit path. As a result, it is possible to prevent these carriers from flowing into the light receiving element portion and abnormally increasing the original photocurrent.

Next, regarding an example of the arrangement of the back carrier absorption layer, the first
This will be explained using Figures 1A, B, and C.

Figure 11A is a plan view showing the pattern on the back side;
The figure is a plan view showing the surface pattern, and Figure 11C is the figure 11B.
It is a sectional view taken along the line a-a in the figure.

As shown in the figure, when the carrier absorption layer 10 is provided on the back surface, P is applied to the back surface so as to match the surface pattern as shown in FIGS. 11A, B, and C.

An N diffusion layer may also be formed.

Further, crosstalk can be further reduced by providing a carrier absorption band on the surface between the main photosensitive surfaces as shown in the prior art.

Next, the results of a comparative experiment between the optical semiconductor devices of the first and second embodiments and a conventional optical semiconductor device will be shown.

FIG. 4 shows a plan view of three semiconductor devices used in this experiment.

One of the three semiconductor devices is the semiconductor device 40 shown in the figure.
On the substrate 41, four semiconductor light receiving elements 43, 4 are arranged.
4, 45, and 46 are arranged.

Plan views of the other two semiconductor devices are also similar to FIG. 4.

Note that FIG. 5 shows the structure of the back surface when the optical semiconductor device shown in the second embodiment is used.

As shown in the figure, this backside structure has phosphorus spread to a depth of about 1.5μ so as to surround the boron diffusion layer 51 with a side of −1.0μ.
It has a structure in which N is diffused into N layers.

The following experiments were conducted on these three semiconductor devices.

A white spot light having a diameter of 60 μm is moved in the direction of an arrow 47 shown in FIG. 4 in steps of 100 μm. At this time, I connected an ammeter to the ■ terminal and observed changes in the output. Note that the other electrodes W, U, and Z were left open without being connected to anything.

The results are shown in FIG.

FIG. 6 is a graph showing the photocurrent distribution appearing at the terminal V when the white spot light moves.

In the figure, the vertical axis shows the photocurrent, and the horizontal axis shows the moving distance of the white spot light. The unit of photocurrent is nA and is expressed in logarithm. In addition, the horizontal axis is referenced to one end of the semiconductor light receiving element 44 (Onm).
, and the moving direction of the white spot light is positive. Further, the unit of distance is I.

In the upper part of the figure, a schematic diagram of the semiconductor light receiving element is shown to show the positional relationship with the semiconductor light receiving element.

In the figure, #61 indicates the photocurrent distribution when the conventional optical semiconductor device is used, and the broken line 62 and the dashed-dotted line 63 indicate the photocurrent distribution when the optical semiconductor device of the first and second embodiments are used, respectively. The photocurrent distribution in the case of

As shown in the figure, when the optical semiconductor devices of the first and second embodiments are used, the crosstalk when the white spot light is on the main photosensitive surface of W is significantly reduced compared to the conventional product. I understand. It can be seen that the results are particularly good when the optical semiconductor device of the third embodiment in which a carrier absorption band is provided on the back surface is used.

However, it can be seen that there is almost no change in the output when the light is incident on the main photosensitive surface of the terminal V for all three.

As is clear from the results shown in FIG. 6, the structures shown in the first and second embodiments may be used to reduce crosstalk.

Next, an experiment in which the entire surface of a semiconductor device is irradiated with light will be described.

In this experiment, the output of the terminal (2) was measured when the entire surface of the semiconductor device 40 shown in FIG. 4 was irradiated with 1001 ux of light using a light source A.

Note that the same three samples used in the above spot light movement experiment were used.

The results of this experiment are shown in FIG.

FIG. 7 is an explanatory diagram showing how the photocurrent flowing through the terminal V when the entire surface of the semiconductor device 4o is irradiated under the above conditions changes depending on the connection of other main photosensitive surface terminals.

There are two types of connections for the other main photosensitive surface terminals: open and short to the common terminal.

As shown in the figure, if there is no influence from other main photosensitive surfaces, the output of terminal V will not change even if the connections of other main photosensitive surface terminals are open or shorted to the common terminal. As expected, it can be seen that there are changes depending on the structure of the back surface of the light receiving element.

In addition, the closer the open/short value shown in the figure is to 1.0, the better. From this point of view, compared to the semiconductor device using the conventional light receiving element, the first and second embodiments are better. Semiconductor devices using light-receiving elements are excellent.

Next, an experiment was conducted to measure the If-Vf characteristics of the three semiconductor devices described above, determine the diode constant n of the terminals, and examine the ohmic properties.

The results of this experiment are shown in Figures 8.9.10.

The vertical axis in Figure 8.9.10 represents the current value (A) logarithmically,
The horizontal axis represents the voltage value (mV).

Figure 8 shows the results when using the light receiving element of the prior art, Figure 9 shows the results when using the light receiving element of the first embodiment, and Figure 10 shows the results when using the light receiving element of the second embodiment. show.

The Dyde constant n was calculated from the slope of a straight line connecting the It"If values when Vr = 100 mV and 400 mV.

As a result, when using the conventional photodetector, n=1
．． 054, when the light-receiving element of the first example is used, n=1.030, and when the light-receiving element of the second example is used, n=1.048, all of which are good without much difference. Ta.

[Effects of the Invention] The present invention has the effect of realizing an optical semiconductor device that has a simple structure, does not generate a photocurrent exceeding the amount of light received by the effective light receiving part, and reduces crosstalk phenomena. be.

[Brief explanation of the drawing]

FIG. 1 is an enlarged sectional view of the optical semiconductor device of the first embodiment, and FIG.
The figure is an enlarged sectional view of the optical semiconductor device of the second embodiment, FIG. 3 is an explanatory diagram of the manufacturing method of the optical semiconductor device of the second embodiment, FIG. 4 is a plan view of the semiconductor device used in the experiment, The figure is a plan view of the backside structure when using the optical semiconductor device shown in the second embodiment, FIG. 6 is a graph showing the photocurrent distribution at terminal V when the white spot light moves, and FIG. 7 is the semiconductor device. Fig. 8 is a graph showing the ItVt characteristics of a semiconductor device using a conventional photodetector, and Fig. 9 is a graph showing the change in the photocurrent flowing through the terminal ■ when the entire surface is irradiated. A graph showing the I t −V characteristics of the semiconductor device when using the light receiving element of the example, FIG. 10 is a graph showing the ItVt characteristics of the semiconductor device when using the light receiving element of the second example,
FIG. 11A is a plan view showing the pattern on the back side, FIG. 11B is a plan view showing the front pattern, and FIG. 11C is a sectional view taken along the line aa in FIG. 11B.

Claims (1)

[Claims] 1. In an optical semiconductor device comprising a photodiode formed by selectively diffusing an impurity forming a conductivity type opposite to that of the substrate in a suitable position of a semiconductor substrate. , On the surface opposite to the semiconductor substrate surface on which the photodiode is formed, there is a region where impurities of the same type as the substrate are selectively diffused at a high concentration, and a region where this high concentration selective diffusion is not performed. A region for selectively diffusing impurities forming opposite conductivity types, and a thin film serving as an electrode formed in contact with the high concentration selective diffusion region and the impurity selective diffusion region. Characteristic optical semiconductor device. 2. Near the diffusion surface that constitutes the photodiode,
2. The light source according to claim 1, wherein an impurity having the same type as said impurity is selectively diffused, and a diffused carrier absorption band formed by short-circuiting a diffusion surface to be formed and said semiconductor substrate is provided. Semiconductor equipment. 3. The thickness of the part where the same type of impurity as the substrate is selectively diffused at a high concentration and the thickness of the part where the impurity forming the conductivity type opposite to the conductivity type of the substrate is selectively diffused are approximately the same thickness. The optical semiconductor device according to claim 1 or 2, characterized in that: 4. The optical semiconductor device according to claim 1, 2 or 3, wherein the semiconductor substrate is a Si substrate. 5. An optical semiconductor device comprising a photodiode formed by selectively diffusing an impurity forming a conductivity type opposite to that of the substrate in an appropriate position of a semiconductor substrate, the photodiode comprising: a diffused carrier absorption band formed by selectively diffusing an impurity forming a conductivity type opposite to that of the substrate in the vicinity of the diffusion surface, and short-circuiting the formed diffusion surface and the semiconductor substrate; An optical semiconductor device comprising: a thin film serving as an electrode formed in contact with the substrate on a surface opposite to the semiconductor substrate surface on which the photodiode is formed. 6. The optical semiconductor device according to claim 1, 2, 3, 4 or 5, wherein the thin film serving as the electrode is made of Au.