JPS62293611A - Epitaxial growth of silicon - Google Patents

Abstract

PURPOSE:To restrict the epitaxial growth temperature within a range lower than or equal to 900 deg.C, and decrease the autodoping of substrate impurity, by performing a growth wherein disilicon hexahydride or trisilicon octahydride is applied as a silicon source and a gas pressure is reduced to a range lower than or equal to 60 Torr. CONSTITUTION:On a single crystal silicon substrate 3 of high impurity density, an Si epitaxial layer is grown. In this process, disilicon hexahydride or trisilicon octahydride is applied as a silicon source, and hydrogen is used as a carrier gas. A growth is performed under a condition of reduced pressure lower than or equal to 60 Torr. Then the epitaxial growth temperature is restricted within a range lower than or equal to 900 deg.C, so that the autodoping of substrate impurity into the epitaxial layer is decreased.

Description

Detailed Description of the Invention 3. Detailed Description of the Invention C Overview] The epitaxial growth temperature is raised to 900' by using 6 disilicon hydride or 8 trisilicon hydride as a silicon source and growing under reduced pressure of 60 Torr or less. A silicon epitaxial growth method that suppresses autodoping and solid phase diffusion of substrate impurities into an epitaxial layer by suppressing the temperature to below C or lower.

[Industrial application field]

For example, bipolar transistors using epitaxial silicon (Si) layers are increasingly formed epitaxially in order to achieve miniaturization and speeding up.

This is because when the base region is formed shallowly in order to suppress the horizontal expansion of the base region and achieve miniaturization, and to increase the speed by narrowing the base width, the buried high impurity concentration region at the bottom of the epitaxial layer and This is to reduce the collector resistance by reducing the distance to the bottom surface of the base region as much as possible, thereby reducing the series resistance and improving the operating speed.

However, when the epitaxial layer is grown thinly, autodoping and solid-phase diffusion of impurities from the buried high impurity concentration region at the bottom results in a high impurity concentration reaching the region near the surface of the epitaxial layer where the base is formed. A problem arises in that the collector breakdown voltage decreases and the driving ability of the bipolar transistor decreases.

Therefore, as devices become smaller and faster, there is a need for a Si epitaxial growth method that minimizes autodoping and solid-phase diffusion of impurities.

[Conventional technology]

Methods for epitaxially growing St include a normal pressure growth method and a reduced pressure growth method. The normal pressure growth method has the advantage of fast growth rate, and the reduced pressure growth method has the advantage of good step coverage.

Conventionally, in the atmospheric pressure epitaxial growth of Si, monosilane, that is, silicon tetrahydride (S i I! 4) is used as the Si source.
), dichlorosilane (SilhClz), and tetrachlorosilane (SiC1a) were used, and the growth temperatures were 950° C. or higher, 1io0° C., and 1150° C., respectively.

Further, 5 iHzC1z was used for the reduced pressure epitaxial growth, and the growth temperature was 1100°C.

As mentioned above, in the conventional Si epitaxial growth method, Si is grown at a high temperature of about 950 to 1150°C, regardless of whether it is under normal pressure or reduced pressure. The depth of the impurity introduced region, that is, the autodoped region formed at the bottom of the epitaxial layer (the side in contact with the substrate) reached 1 μm or more.

Therefore, by the conventional St epitaxial growth method,
It has been extremely difficult to manufacture high-speed bipolar transistors and the like with good performance using epitaxial layers as thin as about 1 to 3 μm.

[Problem that the invention seeks to solve]

The problem to be solved by the present invention is that in the conventional St epitaxial growth method, the depth of the autodoped region formed by impurities introduced from the substrate cannot be formed to be shallow.

[Means for solving problems]

The above problem can be solved by growing on a single-crystal silicon substrate at a temperature of 900°C or less under reduced pressure of 60 Torr or less using disilicon hexahydride or trisilicon 8hydride as a silicon source and hydrogen as a carrier gas. This problem is solved by the silicon epitaxial growth method according to the present invention.

[For production]

That is, the present invention uses silicon hydride as a silicon source, which is advantageous in reducing the amount of autodoping because it does not have etching properties.

As a growth method, a low pressure epitaxial growth method is used, which has the effect of reducing the amount of autodoping by forming a thin growth gas stagnation layer on the growth target surface during growth.

Furthermore, in order to limit the depth of the autodoped region to a value of about 0.2 μm or less, which does not hinder the formation of high-speed devices,
The growth temperature is kept below 900'C.

Among the silicon hydrides, 6-hydride 2 is sufficiently epitaxialized, that is, single crystallized, at the above-mentioned temperature.
Silicon or trisilicon octahydride is used as the silicon source to ensure that the epitaxial growth is sufficiently uniform and that the epitaxial growth is carried out at the lowest possible temperature.

To explain in more detail, silicon (Si) according to the present invention
In the epitaxial growth method, hydrogen (
A growth gas is used in which H2) is used as a carrier gas and a Si source is mixed in a predetermined ratio with the H2 carrier gas.

Since it contains chlorine (C1), it has etching properties, and therefore impurities in the Si substrate to be grown are eluted into the growth gas, thereby increasing the impurity concentration in the growth gas and increasing the amount of autodoping. Chlorosilane (SiHzClz), h-dichlorosilane (Si
Avoid using chlorosilane-based Si sources such as lC1,), and use silicon hydride that has no etching or notching properties and does not cause an increase in the amount of autodoping due to the above-mentioned etching and notching effects.
Use as a source.

The growth method is to thin the stagnant layer of the growth gas formed on the Sii substrate to be grown, so that impurity atoms diffused outward from the Si substrate at the growth temperature are ejected through the stagnant layer. In order to suppress the increase in impurity concentration within the wafer and thereby reduce the amount of autodoping, the method was limited to a reduced pressure growth method.

On the other hand, in the above-mentioned reduced pressure growth, the depth of the auto-doped region is approximately determined by the impurity diffusion length which depends on the growth temperature, and the relationship as shown in FIG. 6 is obtained as a result of actual measurements. Therefore, in the present invention, based on this actual measurement result, the growth temperature was limited to 900° C. or less so that the depth of the autodoped region was limited to at least about 0.2 μm or less.

Note that the autodoping region depth is expressed as the depth from the epitaxial layer's interface with the substrate (where the carrier concentration is almost the same as the base) to the region of the epitaxial layer's original carrier concentration.Usually, this In the case of a bipolar transistor, the concentration difference is about 4 degrees.

Further, as the Si source, from among the silicon hydrides,
Disilane, that is, 6
Disilicon hydride (SizIs) and trisilane, ie, 8-trisilicon hydride (StJe), were limited.

Furthermore, regarding the pressure at which the growth gas is reduced, the above Si,
11. It has been confirmed from the experimental results that there is a relationship as shown in Figure 7, and it has been experimentally confirmed that uniform epitaxial formation is achieved at the growth temperature of 900°C or less. limited.

In order to make the depth of the autodoped region shallower, it is desirable to grow at a lower pressure so that a complete epitaxial layer can be obtained at a lower temperature, but in this case, a gradual decrease in the growth rate is a practical problem.

Figure 8 shows the results of an experiment investigating the relationship between growth gas pressure and growth rate. In order to obtain a sufficiently practical growth rate of about 0.07 μm/min, the gas pressure must be below 60 Torr, which is limited by the present invention. , 60 to ITOrr is desirable.

As a result, the depth of the autodoped region formed in the epitaxial layer can be significantly reduced compared to the conventional method.

〔Example〕

The present invention will be specifically explained below with reference to illustrated embodiments.

Fig. 1 is a schematic diagram of the growth apparatus used in the Si epitaxial growth method of the present invention, Fig. 2 is a profile diagram of the growth process in the example, and Fig. 3 is the arsenic (As) of the Si epitaxial layer in the first example. Concentration profile diagram,
FIG. 4 is a profile diagram of boron (B) concentration in the Si epitaxial layer in the second embodiment, and FIG. 5 is a profile diagram of the boron (B) concentration in the Si epitaxial layer in the third embodiment.
? It is a profile diagram of m degrees.

For the epitaxial growth of St according to the present invention, a single-wafer epitaxial growth apparatus, such as that shown in FIG. 1, which is advantageous in reducing the pressure to a low level, is used.

In the figure, ■ is a Pelger serving as a growth container, 2 is a graphite heater that heats the growth substrate placed thereon, and 3 is a graphite heater that heats the growth substrate placed thereon.
is a Si substrate to be grown, 4 is a growth gas introduction pipe, 5 is a flowmeter for introducing replacement nitrogen (N2) gas, 6 is a flowmeter for introducing N2 carrier gas, and 7 is for introducing Si source gas such as 5i2116. 8 is a vacuum exhaust pipe, 9 is a mechanical booster pump for reducing the pressure inside the Pelger, 10 is also a rotary pump, 11 is power wiring, and 12 is a heating power source for a graphite heater.

The epitaxial growth is performed, for example, according to the procedure shown in the process profile diagram shown in FIG. 2G.

In the figure, N2 indicates a nitrogen atmosphere, 1 (2) indicates a hydrogen atmosphere, and (SizH6+ ++,) indicates a growth gas atmosphere.

In the first embodiment, the above apparatus is used, and the inside of the Pel Jar 1 is first replaced with N2, and then the inside of the Pel Jar 1 is replaced with N2, for example, 4
A Si substrate 1 to be grown having an arsenic (8S) concentration of approximately SiJ, (Si source) and 1
The atmosphere is replaced with a growth gas atmosphere mixed with 11□ (carrier gas) at a flow rate of 0.02/min, and vacuum evacuation is performed to adjust the growth gas pressure to, for example, 3.4 Torr.
After growing for a minute, switch the growth gas to hydrogen gas,
The temperature of the substrate 7 is lowered to room temperature, and the inside of the Pelger is filled with N2.
to complete the growth.

FIG. 3 is a profile diagram of As95 degrees in the depth direction of the Si epitaxial layer grown according to the above example.

As shown in the figure, according to the above embodiment, the concentration of A S ? is higher than the original concentration of the epitaxial layer. The depth of the north autodoped region AD having a degree, that is, the depth DA from the surface S in contact with the St substrate to the surface E having the original epitaxial layer concentration.
D is formed extremely shallowly at about 0°1 μm.

Note that when 5iJa was used as the Si source, the same autodoped region depth DAD as in the above embodiment could be obtained by supplying 5t31(a) at a rate of 3 cc/min.

Further, the above growth was possible even if the growth temperature was lowered to about 700°C.

FIG. 4 shows the results of the second embodiment, that is, the B in the depth direction in a Si epitaxial layer grown under the same conditions as the first embodiment on a Si substrate having a boron (B) concentration of about 4 XIO"cm-'. FIG. 3 is a concentration profile diagram.

Further, FIG. 5 shows the third embodiment, that is, l XIQ19cm-
FIG. 3 is a profile diagram of the Sb degree in the depth direction of an Si epitaxial layer grown under the same conditions as in the first embodiment on a Si substrate having an antimony (Sb) degree of about 100 mm.

The results shown in FIGS. 4 and 5 show that even when Si is epitaxially grown on a Si substrate containing a high concentration of B and sb, the depth of the main dove region formed in the Si epitaxial layer is , it is shown that it can be suppressed to a shallow depth of about 0.1 to 0.2 μm.

In the above embodiments, the present invention was explained with reference to an example in which an epitaxial layer is grown on a Si substrate with a high impurity concentration. For example, in the manufacturing process of bipolar transistors, it is particularly effectively applied when growing an Si epitaxial layer with a low carrier concentration to serve as a collector region on a Si substrate with a buried diffusion region formed on the surface. Ru.

In addition, 501 (Silicon on In5ulato
r) Of course, it also applies to structures.

〔Effect of the invention〕

As explained above, according to the St epitaxial growth method of the present invention, when a Si epitaxial layer is grown on a Si substrate having a high impurity concentration, the depth of the autodoped region of impurities formed at the bottom of the epitaxial layer is can be kept extremely shallow to about 0.1 to 0.2 μm.

Therefore, the present invention is effective in improving the performance of high-speed semiconductor devices such as high-speed bipolar transistors constructed using thin epitaxial layers.

[Brief explanation of the drawing]

Figure 1 is a schematic diagram of the growth apparatus used in the Si epitaxial growth method of the present invention, Figure 2 is a profile diagram of the growth process in the example, and Figure 3 is the arsenic (As) of the Si epitaxial layer in the first example. tfi degree profile diagram, Figure 4 is the second
Boron (B) in the Si epitaxial layer in the example of
(8 degree profile diagram, Figure 5 is an antimony (sb) - m degree profile diagram of the Si epitaxial layer in the third example, and Figure 6 shows the relationship between the depth of the autodoped region and the growth temperature. Figure 7 is a diagram showing the relationship between epitaxialization temperature and growth gas pressure, and Figure 8 is a diagram showing the relationship between growth gas pressure and growth rate. In the figure, 1 is a Pelger, 2 is a graphite heater, 3 is a graphite heater. is the Si substrate to be grown, 4 is the growth gas introduction pipe, 5.6.7 is the flow meter, 8 is the vacuum exhaust pipe, 9 is the mechanical booster pump, 10 is the low-pressure pump, 1 is the power wiring, 12 is the heating power supply, AD is the autodoped region, S is the surface in contact with the Si substrate, E is the original epitaxial N fat surface, and DAD is the depth of the autodoped region.

Claims (1)

[Scope of Claims] Grown on a single crystal silicon substrate at a temperature of 900°C or less under a reduced pressure of 60 Torr or less, using disilicon hexahydride or trisilicon 8hydride as a silicon source and hydrogen as a carrier gas. A silicon epitaxial growth method characterized by performing the following steps.