JP7045005B2 - Semiconductor device - Google Patents

Description

In the field of power devices for power conversion, SiC, GaN, Ga ₂ O ₃ , diamond, etc. are attracting attention for practical use. However, these semiconductor devices have a problem in supply capacity.

On the other hand, there is a device called an Insulated Gate Bipolar Transistor (IGBT) using silicon (Si). The IGBT has a low on-voltage similar to that of a bipolar transistor, a voltage control function and a high-speed switching function of a metal oxide semiconductor (MOS) transistor, and has a feature of having a larger breakdown tolerance than a bipolar transistor. Due to these characteristics, the IGBT market is expanding rapidly.

Since Si-IGBTs have excellent characteristics, they support power electronics in a wide range of fields, from consumer equipment such as air conditioners, refrigerators, washing machines, and microwave ovens to large equipment such as general-purpose inverters, robots, and motor control equipment for trains. It is used as a typical semiconductor device.

Recently, by installing an IGBT, the control circuit can be converted into an inverter, which enables energy saving and a wide range of power control, so its application to hybrid vehicles and electric vehicles is expanding.

One of the elemental technologies that have greatly contributed to the high performance of IGBTs is the introduction of an n ⁺ field stop (FS) layer containing a high concentration of donors. Since the N ⁺ FS IGBT has an n ⁺ layer to stop the depletion layer, the thickness of the n ^- drift layer can be reduced, which makes it possible to reduce the on-voltage.

Further, since the thickness of the drift layer is thin, there is an advantage that excess carriers are small and therefore turn-off loss can be reduced.

Epitaxial substrates have been widely used for the fabrication of IGBTs. However, the manufacturing method using an epitaxial substrate is expensive. Currently, instead of the epitaxial substrate, an n ⁺ FS IGBT using a silicon substrate (FZ Si) manufactured by the floating zone melting method (FZ method) is used.

However, there is a problem that the n ⁺ FS layer cannot be formed in the deep region of the wafer by ion implantation from the back surface of a normal Group 5 donor.

Therefore, hydrogen ions (H ⁺ ) are injected from the back surface of the FZ Si wafer, hydrogen is introduced into deep regions that are difficult to implant with group 5 element ions, and the lattice defects generated by ion implantation are combined with hydrogen. As a result, a technique for forming a hydrogen-related donor and using it as an n ⁺ FS layer is drawing attention.

However, the current concentration of hydrogen-related donors is about 5 x ^{10 15} cm ^-3 , and it is necessary to develop a technology that enables the position and high concentration of the n ⁺ FS layer.

It relates to a technique for forming a high concentration of Ultra Shallow Thermal Donor (USTD) at the substituted carbon position in a self-aligned manner by the effect of the substituted carbon gettingtering hydrogen. As a result, the position of the n + FS layer can be controlled and the concentration can be increased at the same time.

The identity of hydrogen-related donors is in a chaotic state even at the research level, but since it has been reported that hydrogen-related donors are also formed when hydrogen is ion-implanted into FZ Si crystals, hydrogen was injected. It is believed that hydrogen bound to the lattice defects (such as vacancies) that were sometimes formed, resulting in a donor with a shallow energy level.

Therefore, it is considered that the more the lattice defects introduced during ion implantation and the hydrogen bonded to them, the more hydrogen-related donors are formed.

However, it is generally known that the distribution of lattice defects introduced by ion implantation and the concentration peak position of the injected hydrogen are different (Fig. 1), and the concentration of hydrogen-related donors is the peak value at the hydrogen concentration peak position. Is not shown (Fig. 2).

Generally, the hydrogen-related donor concentration is about 1/100 of the hydrogen concentration. This relationship holds in the tail region of the hydrogen concentration distribution formed during hydrogen ion implantation.

However, at the peak concentration of injected hydrogen, for example, where the hydrogen concentration is 1 x 10 ¹⁸ cm ^-3 , the donor concentration drops to about 1/20-1/300 of the hydrogen concentration. It is considered that this is because the position of the lattice defect introduced by ion implantation and the position of the concentration peak of the injected hydrogen are different (FIGS. 1 and 2).

The present invention has been made in view of the above problems, and relates to a technique for simultaneously controlling the peak position of a hydrogen-related donor and increasing the concentration.

As a result of diligent studies, the present inventor has come up with various aspects of the invention shown below.

When a hydrogen-related donor is used for the n ⁺ FS layer of the IGBT, it is necessary to ion-implant hydrogen from the back surface of the wafer. The identity of hydrogen-related donors is in a chaotic state even at the research level, but since it has been reported that hydrogen-related donors are also formed when hydrogen is ion-implanted into FZ Si crystals, hydrogen was injected. It is believed that hydrogen bound to the lattice defects (such as vacancies) that were sometimes formed, resulting in a donor with a shallow energy level.

Therefore, it is considered that the more the lattice defects introduced during ion implantation and the hydrogen bonded to them, the more hydrogen-related donors are formed. However, it is generally known that the distribution of lattice defects introduced by ion implantation and the concentration peak position of the injected hydrogen are different (Figs. 1 and 2), and the concentration of hydrogen-related donors at the hydrogen concentration peak position. Does not show a peak value (Fig. 2).

Since substituted carbon has a gettering effect on hydrogen, the spatial distribution of hydrogen can be controlled by controlling the spatial distribution of carbon by the following method. First, the peak position of the hydrogen concentration formed by hydrogen ion implantation is made to be close to the position of the substituted carbon (Fig. 3).

Subsequently, heat treatment is applied at a low temperature of 300 ° C. or lower to transfer hydrogen. Since hydrogen has a large diffusion coefficient, it diffuses and moves with a slight heat treatment.

The diffused hydrogen forms a peak of hydrogen concentration at the substituted carbon position in a self-aligned manner due to the effect of the substituted carbon getting hydrogen (Fig. 4). As a result, USTD is formed in high concentration at the substituted carbon position. As a result, it is possible to control the position of the USTD and increase the concentration at the same time (Fig. 4).

Furthermore, by controlling the distribution of oxygen impurities so as to overlap the spatial distribution of carbon, USTD is formed, and the spatial distribution and high concentration of donors are realized.

The absorption spectrum of such USTD is shown in FIG.

It is possible to form USTD at a high concentration at the substituted carbon position. As a result, it is possible to control the position of the USTD and increase the concentration at the same time.

Furthermore, by controlling the distribution of oxygen impurities so as to overlap the spatial distribution of carbon, USTD is formed, and the spatial distribution of donors is controlled and the concentration of USTD is further increased.

Hereinafter, the first embodiment will be described in detail. In this embodiment, an oxygen-free Si wafer (FZ Si) is used. Further, in this embodiment, ion implantation is performed from the back surface of the wafer.

As the first step, carbon is injected at a depth of about 1.2 μm from the back surface of the Si wafer at an acceleration energy of 600 KeV so that the peak concentration becomes 1x10 ¹⁸ cm ^-3 .

It attracts and stabilizes the injected carbon in the substituted position by heat treatment at 800 ° C.

Next, hydrogen ions are implanted from the back surface of the wafer. For hydrogen ion implantation, set the dose amount so that the peak concentration of hydrogen is 2 x 10 ¹⁸ cm ^-3 . The acceleration energy of ion implantation is set so that the peak concentration position is 2 μm deep from the back surface of the wafer.

Subsequently, heat treatment at 300 ° C. or lower is applied to diffuse hydrogen.

FIG. 6 shows the results of measuring the distribution of hydrogen before and after the heat treatment of hydrogen diffusion at 300 ° C. by secondary ion mass spectrometry (SIMS). Before the heat treatment, the hydrogen concentration peak exists at a position of about 2 μm from the back surface, but after the hydrogen diffusion heat treatment, the hydrogen concentration peak changes to a position of about 1.2 μm from the back surface.

This position coincides with the carbon concentration peak position. Therefore, it can be considered that hydrogen is diffused by heat treatment and accumulated at the position of substituted carbon due to the gettering effect of carbon.

Furthermore, the spread resistance was measured in order to observe the concentration distribution of the formed USTD (Fig. 7). As a result, it was confirmed that the resistivity was the smallest at the substituted carbon position. It can be considered that this phenomenon is caused by the shift of the peak position of the hydrogen concentration to the substituted carbon position and the shift of the peak position of the USTD to the substituted carbon position.

Hereinafter, the second embodiment will be described in detail. In this embodiment, an oxygen-free Si wafer (FZ Si) is used. Further, in this embodiment, ion implantation is performed from the back surface of the wafer.

Carbon is injected from the back surface of the Si wafer at an acceleration energy of 600 KeV so that the peak concentration is 1 x 10 ¹⁸ cm ^-3 .

It attracts and stabilizes the injected carbon in the substituted position by heat treatment at 800 ° C.

By ion-implanting oxygen impurities from the back surface of the wafer so that they overlap the spatial distribution of carbon and have a peak concentration of 2 x 10 ¹⁸ cm ^-3 , oxygen impurities are introduced at the carbon positions and crystallized by heat treatment at 800 ° C. Restore sex.

Subsequently, hydrogen ions are implanted from the back surface of the wafer. For hydrogen ion implantation, set the dose amount so that the peak concentration of hydrogen is 1 x 10 ¹⁸ cm ^-3 . The acceleration energy of ion implantation is set so that the peak position of the hydrogen concentration is 2 μm deep from the back surface of the wafer.

Subsequently, heat treatment at 300 ° C or lower is applied to diffuse hydrogen and at the same time form USTD.

By this step, it is possible to control the spatial distribution of USTD to the substituted carbon position and increase the concentration of the donor.

The result of resistivity measurement is shown in FIG. It can be confirmed that there is a peak concentration of the donor at the substituted carbon position. In addition, it can be confirmed that the resistance is reduced by ion implantation of oxygen.

Approximate distribution of ion-implanted hydrogen and lattice defects formed by ion implantation. Concentration distribution and hydrogen concentration distribution of hydrogen-related donors formed by ion-implanted hydrogen. When hydrogen is ion-implanted into an FZ Si crystal in which substituted carbon exists in a region where high concentration is expected. Distribution of hydrogen concentration formed after heat treatment for diffusing hydrogen (hydrogen diffusion heat treatment) and distribution of hydrogen-related donors (mainly USTD). Absorption spectra of Ultra Shallow Thermal Donor (USTD) and Hydrogen Shallow Thermal Donor [STD (H)] Hydrogen concentration distribution formed in FZ Si crystals with substituted carbon. Comparison of hydrogen concentration distribution before and after hydrogen diffusion heat treatment. Depth distribution of resistance measured by spread resistance. Depth distribution of resistance measured by spread resistance. Depth distribution of resistance with and without oxygen ion implantation.

1. 1. FZ Si wafer 2. Lattice defect regions formed by hydrogen injection
3. 3. Substituted carbon region

Claims (3)

By controlling the spatial distribution of substituted carbon for hydrogen-related donors formed when hydrogen is ion-implanted, the spatial distribution of ultra-shallow-thermal donor (USTD) is controlled to the substituted carbon position in a self-aligned manner. A method of forming a donor in silicon, characterized in that. Claim 1 is characterized in that the spatial distribution of USTD is controlled by controlling the distribution of oxygen impurities so as to overlap the spatial distribution of substituted carbon, and the concentration of donors is increased. Claim 1 or claim 2, wherein the gettering effect of the substituted carbon on hydrogen is utilized.

Images