JP2012174962A - Method of forming delta-doped structure - Google Patents

Abstract

PROBLEM TO BE SOLVED: To form a delta-doped layer of boron easily in a crystal layer of silicon or the like.SOLUTION: In a state where the surface of a first semiconductor layer is irradiated with exciting light (step S101), a boron introduction layer is formed by introducing first source gas containing boron atoms onto the first semiconductor layer (step S102). Subsequently, a second semiconductor layer is formed by introducing second source gas onto the boron introduction layer which is irradiated with the exciting light (step S103).

Description

  The present invention relates to a method for forming a delta-doped structure in which boron is shallowly doped into a semiconductor layer of silicon, germanium, or silicon-germanium mixed crystal.

  With the recent high density integration of semiconductor devices, the importance of doping technology is increasing. For example, in a well-known CMOS device, by thinning the impurity doping layer to form a shallow junction, it becomes possible to form a CMOS element by laminating it in the thickness direction (vertical direction) of the substrate. . In this way, by stacking CMOS elements, for example, the degree of integration can be further improved.

Recently, a highly accurate doping method has been developed for the purpose of forming such a shallow junction. A typical example is plasma doping. Speaking of p-type doping with boron, shallow doping becomes possible when a hydrogenated boron gas such as B ₂ H ₆ or BF ₃ is excited with plasma and irradiated onto the substrate surface with relatively low energy. Further, when a cluster ion beam is used as a boron source and accelerated by an electric field, the kinetic energy per nuclide is reduced, so that the penetration depth into the substrate is small and shallow doping can be obtained.

  In addition to the doping technique performed from such a surface to a shallow place, there is a technique called delta doping. The purpose of delta doping is to perform doping at a high concentration with a thickness of several atomic layers and confine a large number of carriers in the doped several atomic layer region. This is the ultimate local doping method. When the above-mentioned CMOS devices are stacked in the vertical direction, local doping by delta doping is an essential technique. Delta doping can also be applied to quantum devices.

T. Tatsumi, et al., "Activation efficiency of a B√3 × √3 / Si (111) structure covered with molecular beam deposited amorphous SI or SiOx", Appl. Phys. Lett., Vol.57, no.1 , pp.73-75, 1990. Z. Zhang et al., "B / Si (100) surface: Atomic structure and epitaxial Si overgrowth", J. Vac. Sci. Technol. B, vol.14, no.4, pp.2684-2689, 1996. T. Tatsumi et al., "Surface Segregation at Boron Planar Doping in Silicon Molecular Beam Epitaxy", Japanese Journal of Applied Physics, vol.27, no.6, pp.L945-L956, 1988. H. Jorke and H. Kibbel, "Boron delta doping in Si and Si0.8Ge0.2 layers", Appl. Phys. Lett., Vol.57, no.17, pp.1763-1765, 1990. D.J.Godbey and M.G.Ancona, "Ge segregation during the growth of a SiGe buried layer by molecular beam epitaxy", J.Vac.Sci.Technol. B, vol.11, no.3, pp1120-1123, 1993.

  For example, in a silicon-based device, in order to form the above-described delta doping structure, ions can be implanted and distributed from a surface to a certain depth and annealed. For example, a p-type shallow junction can be formed by forming a delta doping structure in a silicon layer with boron which is a trivalent atom. However, since the energy of ions is high, the distribution of implanted boron ions has a considerable spatial spread in the depth direction. Furthermore, in order to recover the crystallinity of the silicon layer damaged by the ion implantation, a post-annealing process is indispensable, and a considerable amount of boron atoms diffuses up and down during this heating.

  In addition to the above-described ion implantation, there is a technique for forming a delta doping structure of boron under the formed silicon layer by introducing boron onto the surface of the silicon substrate and forming a silicon layer thereon. In this technique, a solid source MBE (Molecular Beam Epitaxy) has been mainly used for formation (growth) of a boron introduction layer (boron delta doped layer) and a silicon layer. In the formation of the delta doped structure by this solid source MBE, it is important how to form the silicon layer on the boron layer.

  For example, a technique has been reported in which the silicon layer is formed at a low temperature after the boron delta doped layer is formed, and this silicon layer remains amorphous (see Non-Patent Document 1). A technique for crystallizing an amorphous silicon layer formed on a boron delta doped layer by annealing has also been reported (see Non-Patent Document 2). In addition, a technique has been reported in which after a boron delta doped layer is formed, a silicon layer is crystal-grown at a substrate temperature of 700 ° C. or higher (see Non-Patent Document 3).

  However, it is difficult to quantitatively supply boron atoms from a solid source with good controllability, which is not suitable for production. Further, it has been reported that a considerable amount of boron atoms diffuses when the substrate temperature is increased in order to make the silicon layer crystalline. When many boron atoms are diffused, the delta doped structure is not obtained.

  As described above, a method using a gas source can be considered for forming the boron delta doped layer and the silicon layer thereon. In order to effectively decompose this gas source, it is common to use plasma. However, in CVD using plasma, hydrogen atoms generated when gas molecules are decomposed are often taken into the film, and in many cases, only an amorphous film can be obtained. Even in this case, hydrogen can be removed by raising the substrate temperature to a temperature at which hydrogen atoms are desorbed. However, at such a high temperature, boron atoms diffuse and a delta-doped structure cannot be obtained.

  As described above, conventionally, there has been a problem that it is not easy to form a delta-doped layer of boron in a crystal layer such as silicon.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to facilitate the formation of a boron delta-doped layer in a crystal layer such as silicon.

  In the method for forming a delta doped structure according to the present invention, the surface of the first semiconductor layer made of a semiconductor in which excitation light selected from ultraviolet light, vacuum ultraviolet light, and soft X-rays is selected from silicon, germanium, and silicon-germanium mixed crystals. A first step of introducing a first source gas containing boron atoms on the first semiconductor layer and forming a boron introduction layer on the first semiconductor layer, and the introduction of the first source gas. Subsequently, a second step of introducing a second source gas containing atoms constituting the first semiconductor layer onto the boron introduction layer irradiated with excitation light and forming the second semiconductor layer on the boron introduction layer; At least.

  In the method for forming the delta doped structure, heat treatment for electrically activating boron in the boron introduction layer may be performed after the second step.

  In the delta-doped structure forming method, in the first step, a point in time at which the boron film starts to grow is predicted from a change in the amount of boron atoms introduced to the surface of the first semiconductor layer measured by a spectroscopic ellipsometer, and The introduction of the first source gas may be stopped before.

  As described above, according to the present invention, the boron source layer is formed by introducing the first source gas containing boron atoms onto the first semiconductor layer in a state where the surface of the first semiconductor layer is irradiated with the excitation light. Then, the second source gas is introduced onto the boron introduction layer irradiated with the excitation light to form the second semiconductor layer. Therefore, in the crystal layer such as silicon, An excellent effect that the delta doped layer can be easily formed is obtained.

FIG. 1 is a flowchart for explaining a method of forming a delta doped structure in an embodiment of the present invention. FIG. 2 is a characteristic diagram showing the Ψ-Δ locus of boron film growth by photoexcited decomposition of diborane gas for each light energy. FIG. 3 is an enlarged characteristic diagram showing the Ψ-Δ locus in the initial stage of growth at the optical energies of 2.3 eV and 3.4 eV shown in FIG. FIG. 4 is a characteristic diagram showing the Ψ-Δ locus for each light energy when the growth of the upper layer of silicon by continuous introduction of diborane gas and photoexcitation decomposition of disilane gas is performed at a growth temperature of 300 ° C. FIG. 5 is a characteristic diagram showing pseudo dielectric response functions at room temperature of the silicon buffer layer (dotted line) and the Si / B / Si delta doped structure (solid line) for the delta doped structure described with reference to FIG. FIG. 6 is a characteristic diagram showing the ψ-Δ locus for each light energy when a silicon upper layer is grown at a growth temperature of 420 ° C. after diborane gas is introduced at a growth temperature of 250 ° C. to form a boron introduction layer. is there. FIG. 7 is a characteristic diagram showing pseudo dielectric response functions at room temperature of the silicon buffer layer (dotted line) and the Si / B / Si delta doped structure (solid line) for the delta doped structure described with reference to FIG. FIG. 8 is a characteristic diagram showing the ψ-Δ locus for each light energy when a silicon upper layer is grown at a growth temperature of 390 ° C. after diborane gas is introduced at a growth temperature of 250 ° C. to form a boron introduction layer. is there. FIG. 9 is a characteristic diagram showing pseudo dielectric response functions at room temperature of the silicon buffer layer (dotted line) and the Si / B / Si delta doped structure (solid line) for the delta doped structure described with reference to FIG. FIG. 10 is a distribution diagram showing the result of examining the diffusion state of boron atoms from the boron introduction layer described with reference to FIGS. 6 and 8 by secondary ion mass spectrometry (SIMS).

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a flowchart for explaining a method of forming a delta doped structure in an embodiment of the present invention. In this method of forming the delta-doped structure, first, in step S101, a first semiconductor made of a semiconductor in which excitation light selected from ultraviolet light, vacuum ultraviolet light, and soft X-rays is selected from silicon, germanium, and silicon-germanium mixed crystal. Irradiate the surface of the layer.

  Next, in step S102, a first source gas containing boron atoms is introduced onto the first semiconductor layer irradiated with the excitation light, and a boron introduction layer (boron delta doped layer) is formed on the first semiconductor layer. Form.

  Next, in step S103, following the introduction of the first source gas, a second source gas containing atoms constituting the first semiconductor layer is introduced onto the boron introduction layer irradiated with the excitation light, and the boron introduction layer is introduced. A second semiconductor layer is formed thereon. For example, when the first source gas is introduced at a predetermined introduction amount for a predetermined time, the introduction of the first source gas may be stopped and then the introduction of the second source gas may be started.

Here, as the first source gas containing boron atoms, diborane (B ₂ H ₆ ), which is a hydride of boron, is representative, but decaborane (B ₁₀ H ₁₄ ) can also be used. As the second source gas containing atoms constituting the first semiconductor layer, for example, when the first semiconductor layer is silicon, the atoms constituting the first semiconductor layer is silicon, and silane (SiH ₄₎ which is a hydride of silicon. ), Disilane (Si ₂ H ₆ ), trisilane (Si ₃ H ₇ ), and the like.

  The source gas described above can be efficiently decomposed by excitation light selected from ultraviolet light, vacuum ultraviolet light, and soft X-rays. Photolysis by ultraviolet light is valence electron excitation, and photolysis by vacuum ultraviolet light and soft X-rays is by inner shell excitation. Among these, decomposition by inner shell excitation has a higher quantum yield of the reaction, and the region deposited from the source gas is excited by the secondary electrons emitted when the first semiconductor layer is excited. There exists an advantage which can be limited only to an irradiation part. In addition, each substance on the surface of the irradiated layer is also excited by the above-described excitation light irradiation. Therefore, in the layer irradiated with the excitation light, hydrogen that is a decomposition product of each source gas described above. Is also excited, so that this hydrogen is prevented from being taken into the irradiated layer.

  By the way, for example, in the vapor phase growth of boron on the first semiconductor layer made of silicon, at the stage where the surface of the first semiconductor layer at the initial stage of growth is not completely covered with the boron layer, the deposition rate of boron atoms is Low. From the point where the amount of boron exceeds a certain value, the deposition rate increases rapidly, and the formation of a boron film begins. In the growth of such a boron film, if the boron atoms to be deposited become a boron film having a certain thickness, an epitaxial layer of silicon cannot be formed on the boron film.

  For this reason, in order to control the amount of boron atoms to be introduced, it is important to precisely control the amount (introduction time) of the boron source gas introduced onto the first semiconductor layer. In such introduction control of the boron source gas (first source gas), the surface state during vapor phase growth of boron may be measured (monitored) by spectroscopic ellipsometry.

  Spectroscopic ellipsometry, which performs optical measurement, has the advantage that measurement can be performed non-destructively, non-eroding, and non-contact. From the measurement data of spectroscopic ellipsometry, predict the point when the boron film starts to grow continuously, and stop introducing the boron source gas before this point, so that the amount of boron atoms introduced to the surface is below a certain level. The formation of a boron-introducing layer having a thickness greater than necessary can be suppressed.

  Next, when the second semiconductor layer is grown, the surface of the first semiconductor layer is already covered with high-density boron atoms (a boron introduction layer is formed), so that it is a harsh environment for epitaxial growth. There is no doubt. For example, when the growth temperature of the second semiconductor layer is low, the boron introduction layer prevents the crystal state information on the surface of the first semiconductor layer (silicon crystal) from being transmitted, and the crystallinity of the second semiconductor layer is polycrystalline or non-crystalline. It becomes a crystalline layer.

  On the other hand, the crystal grows by raising the growth temperature. However, at such a high temperature state, boron atoms in the boron introduction layer are thermally diffused, and a delta doped structure cannot be formed. For example, in order to perform the growth of the second semiconductor layer by the gas source MBE, it is necessary to raise the growth temperature to about 600 ° C. If such a high temperature is continued for a long time, boron atoms are easily diffused. End up.

  In order to solve the above-described problem, as in the formation of the boron introduction layer, the second source containing atoms constituting the first semiconductor layer using excitation light selected from ultraviolet light, vacuum ultraviolet light, and soft X-rays The second semiconductor layer is grown (vapor phase growth) at a low temperature by photoexcitation decomposition of the gas. However, even in the formation of the second semiconductor layer by photoexcitation decomposition, if the growth temperature is too low, epitaxial growth does not occur. Conversely, if the growth temperature is too high, boron diffusion becomes significant. Therefore, in the formation of the second semiconductor layer by photoexcitation decomposition, it is desirable to set the growth temperature to the lowest temperature at which epitaxial growth is possible or a temperature slightly above the lowest temperature. Further, if a process of annealing for a short time is added after the second semiconductor layer is grown to produce the delta-doped structure, the boron-doped layer can be electrically activated sufficiently while suppressing the diffusion of boron.

[Example]
Hereinafter, it demonstrates in detail using an Example. Hereinafter, the case where silicon is used as the semiconductor will be described, and the first semiconductor layer is a silicon substrate. Further, since the light absorption of boron atoms in the ultraviolet region is small, white light having an energy of 10 to 1000 eV from the vacuum ultraviolet region to the soft X-ray region was used as an excitation source (excitation light). The growth was performed in an ultra-high vacuum chamber having an ultimate vacuum of 10 ⁻⁸ Pa. For introduction of boron atoms, diborane diluted to 1% with helium gas was used. Disilane was used as the silicon source.

  First, the silicon substrate was treated with dilute hydrofluoric acid to remove the oxide layer on the surface, thereby terminating the hydrogen. The silicon substrate thus treated was carried into a vacuum chamber, the substrate temperature was raised to a predetermined temperature, disilane gas was introduced, and a silicon buffer layer was grown by the gas source MBE. The purpose of this process is to provide an epitaxial substrate that is flatter and cleaner than the original silicon substrate. It can be said that the silicon buffer layer is also a first semiconductor layer.

  Next, the substrate temperature was lowered to a predetermined temperature, and diborane gas was introduced onto the substrate for a predetermined time in a state where the substrate was irradiated with radiation light (excitation light). The diborane introduced by the emitted light is decomposed, and a high-concentration borane layer (boron introduction layer) is deposited on the silicon substrate (silicon buffer layer). If the substrate temperature is high, the thermal decomposition of diborane is rapidly accelerated, so the substrate temperature must be kept at 300 ° C. or lower.

FIG. 2 is an example, but after the introduction of diborane gas at a substrate temperature of 300.degree. It is a thing. The diborane gas partial pressure was 3.1 × 10 ⁻³ Pa. Here, the (Ψ, Δ) angle is given by R _p / R _s = tan Ψ exp (iΔ) as the ratio of the complex reflectivity R _p and R _s of p-polarized light and s-polarized light. The (ψ, Δ) angles were simultaneously monitored at light energies “1.5 eV”, “2.3 eV”, “3.4 eV”, and “4.3 eV”.

  As shown in FIG. 2, reflecting that the optical constants are greatly different between silicon and boron, the (Ψ, Δ) point, which is the point of the (Ψ, Δ) angle, grows along the locus O → A → B. → Go away from the starting point O like C → D → E.

  FIG. 3 is a graph in which the transition in the initial stage of growth is enlarged and plotted for 2.3 eV and 3.4 eV. In FIG. 3, data points are acquired at 10 second intervals. First, when introduction of diborane gas is started, the point (Ψ, Δ) is shifted to the left along the locus O → A. This reflects the fact that diborane is first physically adsorbed on the surface when the substrate temperature is low. While the number of boron atoms deposited on the surface is still less than one atomic layer, the (Ψ, Δ) point moves very slowly in the low Δ angle direction.

  Also, corresponding to point B after about 300 seconds, when the silicon surface is completely covered with boron atoms, the decomposition of diborane on the layer covered with boron atoms is now under the activation energy low. As it progresses, the deposition rate of boron increases at an accelerated rate, and the locus rapidly extends in the low Δ angle direction. Accordingly, it can be seen from these results that the introduction of diborane gas should be stopped when the (Ψ, Δ) point is slightly shifted from the origin.

  Next, an example will be described in which the introduction of diborane gas and the growth of the silicon upper layer (second semiconductor layer) by photoexcited decomposition of disilane gas are performed only by switching the gas in the middle at the same substrate temperature. FIG. 4 is a characteristic diagram showing the monitoring result of the (Ψ, Δ) angle when the silicon upper layer is formed on the boron introduction layer only by gas switching.

In this example, the substrate temperature is 300 ° C., and boron deposition when a diborane gas having a partial pressure of 1.5 × 10 ⁻³ Pa is introduced for 240 seconds corresponds to the locus O → A → B. The introduction of diborane gas was stopped between points B and C, and subsequently, disilane gas having a partial pressure of 0.16 Pa was introduced. In 3.4 eV and 4.3 eV, the trajectory C → D → E is directed upward so as to draw a loop from the C point. 3.4 eV and 4.3 eV are energy corresponding to transitions between E1 and E2 bands of the silicon crystal, respectively, and the point (Ψ, Δ) returns from the C point to the vicinity of the start point O. It means that it is an epitaxial layer.

  On the other hand, 1.5eV and 2.3eV reflect the fact that Si / B / Si has a multi-layer structure because the light penetration depth is larger on the longer wavelength side than the band edge, and the silicon upper layer grows. Accordingly, a locus is drawn monotonously toward the low Δ angle side.

The pseudo dielectric response function (<ε> = <ε ₁ > + i <ε ₂ >) at room temperature of the Si / B / Si delta doped structure thus obtained is compared with that at the stage where the silicon buffer layer is formed. FIG. There is no great difference between them, and the dielectric response function of the upper silicon layer is almost the same as that of the silicon buffer layer (first silicon layer), indicating that the upper silicon layer is an epitaxial layer.

  When the supply (introduction) rate of disilane is increased, the state of growth becomes close to CVD from the gas source MBE, the supply of silicon hydride is superior to the desorption of hydrogen, and the crystallinity of the upper layer of silicon is deteriorated. To confirm this, the substrate temperature at the time of introduction of diborane gas was set to 250 ° C., a boron introduction layer was formed, the substrate temperature was raised to 420 ° C. or 390 ° C., and disilane was introduced at a pressure of 0.31 Pa. The upper layer of silicon is formed by photoexcitation differentiation. In this confirmation, the boron introduction layer is formed thicker than in the case described with reference to FIG. Therefore, in the following example, the condition is such that the silicon upper layer is less likely to undergo crystal growth (epitaxial growth) as compared with the case described with reference to FIG.

  First, when the substrate temperature is 420 ° C. shown in FIG. 6, the trajectory O → A of the (Ψ, Δ) point for each light energy corresponds to the time when diborane is introduced. When the supply of diborane gas was stopped and the substrate temperature was raised to 420 ° C., the dielectric response of the system changed and changed along the locus A → B. Next, when disilane is introduced, at the initial stage of introduction, a locus B → C → D is drawn and remains in a region having substantially the same (Ψ, Δ) angle. This indicates that silicon is epitaxially grown. The reason why the starting point O and the point D are different is that the substrate temperature is different. However, as the growth of the upper silicon layer continued further, the crystallinity deteriorated a little, and the (Ψ, Δ) point shifted along the trajectory D → E, albeit slowly.

Next, the pseudo dielectric response function after growth is shown in FIG. Similar to the case described with reference to FIG. 5, the dielectric response function of the upper layer of silicon is not limited to that of silicon crystal, indicating that the boron-introduced layer is sandwiched therebetween, but epitaxially grown. However, since the height of the E ₂ transition peak in the <ε ₂ > spectrum is slightly lower than that in FIG. 5, the crystallinity of the upper silicon layer is slightly inferior to that in the condition described with reference to FIG.

  FIG. 8 shows the result of setting the growth temperature of the silicon upper layer as low as 390.degree. Other conditions are the same as those described with reference to FIG. In this case, after the loop-type locus in the initial growth stage of the silicon upper layer, the locus D → E extends fairly rapidly toward the low Δ angle side. Further, the direction of the trajectory changes from point E, and the change continues along the trajectory E → F to the low Δ angle region. The trajectory D → E corresponds to the formation of microcrystalline silicon, and the trajectory E → F corresponds to the formation of amorphous silicon. This result means that when the growth temperature was lowered to 390 ° C., the crystallinity of the upper silicon layer deteriorated as the growth progressed and changed from a microcrystal to an amorphous state. This result is caused by growing on the boron-introduced layer, resulting in poor interface conditions and low-temperature growth by photoexcitation decomposition. The pseudo dielectric response function after growth is shown in FIG. The pseudodielectric response function is very different from that of a silicon substrate, which represents the formation of microcrystalline silicon.

  Next, the results of examining the diffusion state of boron atoms from the obtained delta doping layer (boron introduction layer) by secondary ion mass spectrometry (SIMS) will be described with reference to FIG. 10A shows the depth distribution of boron atoms in the boron-introduced layer described with reference to FIG. 6, and FIG. 10B shows the boron atoms in the boron-introduced layer described with reference to FIG. Depth distribution is shown. In both cases, the location where the concentration peak is formed is the surface (S) of the silicon upper layer, the inside of the boron introduction layer (M), and the interface (I) between the silicon buffer layer and the silicon substrate.

  First, the peak of the surface (S) of the silicon upper layer will be described. This is called “Surface peak” (see Non-Patent Document 5). Thus, it is well known that the dopant is segregated on the surface.

  Next, the peak of the interface (I) will be described. At the interface between the silicon buffer layer and the silicon substrate, slight crystallinity is likely to be disturbed, and boron atoms are selectively trapped by such defects. Such a trapped state appears as a peak of the interface (I). However, the boron atom concentration in the defect portion is not so high.

  Next, the peak in the boron introduction layer (M) will be described. The boron concentration distribution in the boron-introduced layer decreases in the silicon buffer layer (silicon substrate) side, which is a state called “Leading edge” (see Non-Patent Document 5). This decrease in concentration is considered to be due to the diffusion of a part of boron on the silicon substrate side while depositing boron atoms. This diffusion of boron atoms is thought to be promoted by synchrotron radiation excitation.

  Further, the tailing toward the upper layer side of silicon is called “Trailing edge” (see Non-Patent Document 5). When the substrate temperature is 390 ° C. in FIG. 10B, the boron distribution is sharply cut off at the interface between the boron introduction layer and the silicon upper layer. Since the silicon upper layer is formed by photoexcitation decomposition of the source gas, the diffusion of boron atoms is suppressed by the “surfactant” effect of hydrogen atoms, and in this case, “trailing edge” is hardly seen. It is thought that.

  However, when the silicon upper layer is epitaxially grown under the condition of the substrate temperature of 420 ° C. in FIG. 10A, a tailing distribution toward the silicon upper layer side is seen, and at the same time, it swells to a depth of about 230 nm. A hump-like structure exists, suggesting diffusion. However, this state is not so remarkable, and both the fact that the upper layer of silicon is epitaxial and that boron atoms are confined in the boron introduction layer at a high concentration are compatible results. From this, it is clear that a silicon upper layer having excellent crystallinity can be formed on the boron-introduced layer (layer thickness of about 50 nm) produced in this example by performing a process by photoexcitation differentiation at a substrate temperature of 420 ° C. or higher. .

  In the example described with reference to FIG. 4, the silicon upper layer can be crystal-grown at a growth temperature of 300 ° C. This is a case where the boron introduction layer is thinner (the amount of boron deposited is small). . On the other hand, when a thicker boron-introduced layer (layer thickness of about 50 nm) is formed, the silicon upper layer can be made an epitaxially grown layer by raising the growth temperature to 420 ° C. as described above.

  As described above, according to the present invention, since the second semiconductor layer is formed by switching the source gas in the same apparatus following the formation of the boron introduction layer, the second semiconductor layer is formed without exposure to the atmosphere. A boron introduction layer having interface characteristics can be formed continuously. Further, since the boron introduction layer and the second semiconductor layer are formed by deposition that decomposes the source gas with excitation light, a low-temperature process that can suppress thermal diffusion of boron atoms is possible, and at the same time, the “Surfactant of hydrogen atoms” By the effect, the diffusion of boron atoms can be suppressed.

  Furthermore, by monitoring the formation state of the boron introduction layer by spectroscopic ellipsometry, the amount of boron atoms introduced into the surface of the first semiconductor layer can be known during growth. By monitoring in this way, even when conditions such as the production temperature and the gas pressure change, it becomes possible to determine the time point at which the introduction of the first source gas should be stopped for each delta dope structure being manufactured.

The present invention is not limited to the embodiment described above, and many modifications and combinations can be implemented by those having ordinary knowledge in the art within the technical idea of the present invention. It is obvious. For example, although the case where silicon is mainly used as a semiconductor has been described in the above-described embodiment, the present invention is not limited to this, and the same applies to germanium or silicon-germanium mixed crystals. If germanium (GeH ₄ ) is used as a source gas for germanium, the same process can be performed. However, the temperature at which boron atoms begin to diffuse or the temperature at which germanium or silicon germanium is deposited is different. In general, in a silicon germanium mixed crystal, the crystallization temperature decreases as the germanium concentration increases.

Claims (3)

The first semiconductor layer is irradiated with excitation light selected from ultraviolet light, vacuum ultraviolet light, and soft X-rays on the surface of the first semiconductor layer made of a semiconductor selected from silicon, germanium, and silicon-germanium mixed crystal. A first step of introducing a first source gas containing boron atoms on the first semiconductor layer and forming a boron introduction layer on the first semiconductor layer;
Subsequent to the introduction of the first source gas, a second source gas containing atoms constituting the first semiconductor layer is introduced onto the boron introduction layer irradiated with the excitation light, and the top of the boron introduction layer is introduced. And at least a second step of forming a second semiconductor layer. A method for forming a delta-doped structure. The method for forming a delta-doped structure according to claim 1,
After the second step, a heat treatment for electrically activating boron in the boron introduction layer is performed. In the formation method of the delta dope structure of Claim 1 or 2,
In the first step, a time point at which a boron film starts to grow is predicted from a change in the amount of boron atoms introduced to the surface of the first semiconductor layer measured by a spectroscopic ellipsometer, and the first source is predicted before the predicted time point. A method of forming a delta-doped structure, characterized by stopping the introduction of gas.

Images