JPS5936437B2 - Semiconductor photodetector - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a semiconductor light receiving device suitable as a photodetecting element in, for example, optical communication.

Conventionally, photodiodes have been used as optical readers in electronic computers, various automatic control systems, and light detection elements in alarm devices.

As is well known, a photodiode is a device that captures carriers excited in a semiconductor by an optical signal using various barriers and extracts them as a photocurrent or photovoltage signal. When using a PN junction to which a bias voltage is applied, minority carriers that are photoexcited in the semiconductor move by diffusion, are absorbed in the depletion layer, and become a photocurrent.

By the way, a normal photodiode does not have an amplification function inside the element.

Therefore, avalanche photodiodes (hereinafter referred to as APDs), which amplify photocurrent by utilizing the avalanche multiplication effect in semiconductors,
) was developed. Since this APD has extremely high sensitivity and high speed compared to ordinary photodiodes, it is considered to be promising as a photodetecting element in optical communication. Now, generally speaking, when a photodiode is irradiated with light,
For example, when the wavelength of light is 6323, 4I from the semiconductor surface
For example, if the wavelength is as long as about 8,500 wavelengths, it will penetrate about 150 wavelengths from the semiconductor surface.

As described above, since a photodiode converts carriers photoexcited within a semiconductor into photocurrent in a depletion layer, the depletion layer must expand appropriately according to the wavelength of light. That is, the wavelength is short, for example, the 6328
If the depletion layer exists only in a shallow position, such as a normal photodiode, good operation can be expected even in a device with a shallow depletion layer. Ordinary photodiodes can no longer be used. This is exactly the same for APD. That is, in a typical APD, as shown in FIG. 1, an N-type layer 2 is formed by spreading on a portion of a P-type semiconductor substrate 1 that is to be a photosensitive region. And P between the substrate 1 and the N-type layer 2
A reverse bias voltage is applied to the N junction, and a depletion layer extending from the PN junction toward the P-type semiconductor substrate 1 is utilized. This depletion layer can be expanded as the value of the reverse bias voltage is increased. However, in the APD shown in FIG. 1, since the impurity concentration of the P-type semiconductor substrate 1 is high, the expansion of the depletion layer is small even though the value of the reverse bias voltage is increased. Moreover, increasing the value of the reverse bias voltage causes an increase in dark current, and the multiplication factor is limited by the current saturation effect, so the effect of enlarging the depletion layer is diminished. Therefore, the quantum efficiency in a normal APD cannot be increased very much. Therefore, a reach-through type APD with the structure shown in FIG. 2 was developed as an element to eliminate the drawbacks of the APD. In the figure, 21 is a P-type or π-type semiconductor substrate, 22 is a Pf-type contact layer, and 23 is a P-type region, 24
indicate N-shaped areas, respectively.

The impurity concentration distribution in the depth direction in the APD with this structure is as shown in FIG. When the voltage is applied, a depletion layer generated at the PN junction formed by the P-type region 23 and the N-type region 24 easily spreads into the substrate 21, and the contact layer 2
Reach up to 2.

Therefore, the depletion layer is expanded over a wide range of the P-type region 23 and the substrate 21 and becomes a photosensitive region, so that the quantum efficiency is greatly improved. The electric field distribution in the depth direction in such an APD is as seen in Figure 4, and the Y axis in the figure is P.
It represents the N junction surface.

As can be seen, the electric field in the P-type region 23 is high, and the electric field in the π-type semiconductor substrate 21 is low. It should be noted that the P-type region 23, ie, the high electric field region, is related to light multiplication, and the π-type semiconductor substrate 21, ie, the low electric field region, is related to light absorption, and therefore quantum efficiency. This reach-through type APD is usually (:1
Although it has superior properties compared to 5APD, it still has drawbacks.

That is, in order to form the PN junction of the APD shown in FIG. For example, it is not uniform due to abnormal diffusion phenomenon, etc., and therefore,
The PN junction surface is formed with unevenness. When direct pressure is applied to such an uneven PN junction surface, the P
The electric field distribution at the N-junction also becomes non-uniform. That is,
The light multiplication factor at the PN junction surface partially changes, and therefore, the photosensitivity of the semiconductor light receiving device partially changes. Currently, the number of pages of the N-shaped layer 24, which is the photosensitive area surface in a reach-through type APD, is, for example, 300.

φ, and on the other hand, the glass fiber used for optical communication is 20CIi)φ, so the conventional reach-through type A
In a PD, when a glass fiber is brought into contact with the N''' type layer 24, the sensitivity will differ depending on the location.In addition, in general, when impurities are diffused into a semiconductor substrate,
It is common knowledge that the diffusion depth varies slightly from rod to rod.

Therefore, in the above-mentioned APD, the electric field distribution in the PN junction is different for each rod, resulting in different breakdown voltages. The present invention, in a semiconductor photodetector, prevents the electric field at the PN junction surface from being immediately affected even if the depth of impurity diffusion is uneven, thereby preventing variations in sensitivity or breakdown voltage from occurring. A low impurity concentration semiconductor substrate (or semiconductor layer) of one conductivity type, a medium impurity concentration region of one conductivity type buried in the low impurity concentration semiconductor layer of the one conductivity type, and A medium impurity concentration semiconductor region of the opposite conductivity type formed in the depth direction from the surface of the low impurity concentration semiconductor layer and facing the buried medium impurity concentration semiconductor region of the one conductivity type at a predetermined distance in the depth direction and forming a photosensitive region. The present invention provides a semiconductor light receiving device characterized by comprising a high concentration impurity semiconductor layer, which will be described in detail below.

In the present invention, the conventional reach-through type APD (:
I) The knowledge that the sensitive fluctuation of the electric field distribution at the PN junction surface is caused by the high impurity concentration in the P-type region 23 is the basis.

That is, the electric field E at the PN junction surface is: Q: ionized charge amount per unit area of the depletion layer εs: relative dielectric constant of silicon εo: dielectric constant in vacuum.

It is clear that Q in this equation (1) is influenced by the impurity concentration of the P-type region 23, and therefore, the higher the impurity concentration of the P-type region 23, the greater the influence on the electric field E at the PN junction surface. Therefore, even a slight non-uniformity in the diffusion depth of the N+ body layer 24 forming the photosensitive area will cause a large variation.

Therefore, in order to eliminate the drawbacks of the conventional reach-through type APD, it is sufficient to prevent the N+ type layer and the highly doped P type region from coming into contact with each other. FIG. 5 is an explanatory diagram of main parts of an embodiment of the present invention. In the figure, 51 is a π-type, that is, a P-type semiconductor substrate (one conductivity type low impurity concentration semiconductor substrate or semiconductor layer), 52 is a P+ type contact layer, and 53 is a P type region (one conductivity type medium impurity concentration semiconductor area), 54 is the substrate 5
1 is a part of the P-type semiconductor layer, and 55 is an N+ type layer (high concentration impurity semiconductor layer of the opposite conductivity type) forming the photosensitive surface. The P-type region 53 in this configuration can be formed extremely easily by using an ion implantation method and appropriately adjusting the implantation energy.

Of course, other techniques, such as the epitaxial method by appropriately changing the steps, may also be employed. FIG. 6 shows the impurity concentration distribution of the embodiment shown in FIG. As can be seen from FIGS. 5 and 6, the N+ type layer 55 is a P type semiconductor layer 5 in which almost no ionized charges exist.
4, it faces the P-type region 53 through the P-type region 53, so even if the diffusion depth changes, Q in the above equation (1) does not change, so the electric field distribution is not affected at all and is almost uniform. be. In order to eliminate the adverse effects of variations in diffusion depth, the thickness of the P-type semiconductor layer 54 is preferably about 1,000 mm or more.
Such thickness control can be achieved with sufficient accuracy even if the P-type region 53 is formed by ion implantation. As can be seen from the above explanation, according to the present invention, even if the diffusion depth of the high impurity concentration semiconductor layer of the opposite conductivity type forming the photosensitive area is uneven, unevenness in electric field distribution does not occur. There is no local change in power, making it suitable for injecting optical signals from thin glass fibers for optical communications, and there is no difference in breakdown voltage from rod to rod.

[Brief explanation of the drawing]

Figure 1 is an explanatory diagram of the conventional example, Figure 2 is also an explanatory diagram of the conventional example, Figure 3 is a diagram showing the impurity concentration distribution of the conventional example in Figure 2, and Figure 4 is a diagram showing the impurity concentration distribution of the conventional example.
Figure 2 shows a diagram showing the electric field distribution in the conventional example, Figure 5 shows an explanatory diagram of the main part of an embodiment of the present invention, and Figure 6 shows a diagram showing the impurity concentration distribution in the embodiment of Figure 5. . In the figure, 51 is a P type semiconductor substrate, 52 is a P+ type contact layer, 53 is a P type region, 54 is a P− type semiconductor layer,
55 each indicate an N+ type layer.

Claims (1)

[Claims] 1. A low impurity concentration semiconductor substrate (or semiconductor layer) of one conductivity type, a medium impurity concentration semiconductor region of one conductivity type buried in the low impurity concentration semiconductor layer of the one conductivity type, and the low impurity concentration semiconductor region of the one conductivity type. A high concentration impurity of an opposite conductivity type formed in the depth direction from the surface of the concentrated semiconductor layer and facing the medium impurity concentration semiconductor region of one conductivity type buried at a predetermined distance in the depth direction and forming a photosensitive region. A semiconductor light receiving device comprising: a semiconductor layer;