JP6367434B2 - Semiconductor device and manufacturing method thereof - Google Patents

Description

  Embodiments described herein relate generally to a semiconductor device and a method for manufacturing the same.

  SiC (silicon carbide) is expected as a material for next-generation power semiconductor devices. Compared with Si (silicon), SiC has excellent physical properties such as a band gap of 3 times, a breakdown electric field strength of about 10 times, and a thermal conductivity of about 3 times. By utilizing this characteristic, a power semiconductor device capable of operating at high temperature with low loss can be realized.

  On the other hand, it is known that when a MOS (Metal Oxide Semiconductor) structure is formed on the C-plane of 4H—SiC, the gate leakage current becomes larger than that of the Si-plane.

JP 2010-27962 A

  The problem to be solved by the present invention is to provide a semiconductor device capable of suppressing the leakage current of the gate insulating film provided on the SiC layer and realizing high carrier mobility, and a method for manufacturing the same.

The semiconductor device of the embodiment includes a SiC layer and a gate insulating film provided on the surface of the SiC layer, and includes an oxide film or an oxynitride film in contact with the surface of the SiC layer, and the oxide film or oxynitride film is Mg , Containing at least one element selected from Ca, Sr, and Ba, the peak of the concentration of the element in the gate insulating film is on the SiC layer side of the gate insulating film, and the peak of the concentration of the element is an oxide film or oxynitride A gate electrode provided on the gate insulating film, the gate insulating film having a region having a concentration of the element of 1 × 10 ¹⁶ cm ⁻³ or less on a side opposite to the SiC layer having the peak of the element concentration in the film; The gate insulating film contains C (carbon), and the peak of C (carbon) concentration in the gate insulating film is on the SiC side of the gate insulating film, and the C (carbon) Concentration of Click is in the oxide film or oxynitride film.

1 is a schematic cross-sectional view illustrating a semiconductor device according to a first embodiment. It is a figure which shows the enlarged view and element distribution of the gate insulating film of 1st Embodiment, and a gate electrode part. It is explanatory drawing of the effect | action and effect of 1st Embodiment. It is explanatory drawing of the effect | action and effect of 1st Embodiment. It is explanatory drawing of the effect | action and effect of 1st Embodiment. It is explanatory drawing of the effect | action and effect of 1st Embodiment. It is explanatory drawing of the effect | action and effect of 1st Embodiment. It is explanatory drawing of the effect | action and effect of 1st Embodiment. It is a process flow figure showing a manufacturing method of a semiconductor device of a 1st embodiment. It is a schematic cross section which shows the 1st manufacturing method of the semiconductor device of 1st Embodiment. It is a schematic cross section which shows the 1st manufacturing method of the semiconductor device of 1st Embodiment. It is a schematic cross section which shows the 1st manufacturing method of the semiconductor device of 1st Embodiment. It is a schematic cross section which shows the 1st manufacturing method of the semiconductor device of 1st Embodiment. It is a schematic cross section which shows the 1st manufacturing method of the semiconductor device of 1st Embodiment. It is a schematic cross section which shows the 1st manufacturing method of the semiconductor device of 1st Embodiment. It is a figure which shows element distribution of the gate insulating film of 1st Embodiment and a comparison form. It is a schematic cross section showing a semiconductor device of a second embodiment.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In the following description, the same members and the like are denoted by the same reference numerals, and the description of the members and the like once described is omitted as appropriate.

In the following description, the notations n ⁺ , n, n ⁻ and p ⁺ , p, p ⁻ represent the relative level of impurity concentration in each conductivity type. That is, n ⁺ indicates that the n-type impurity concentration is relatively higher than n, and n ⁻ indicates that the n-type impurity concentration is relatively lower than n. Further, p ⁺ indicates that the p-type impurity concentration is relatively higher than p, and p ⁻ indicates that the p-type impurity concentration is relatively lower than p. In some cases, n ⁺ type and n ⁻ type are simply referred to as n type, p ⁺ type and p ⁻ type as simply p type.

(First embodiment)
The semiconductor device of this embodiment includes a SiC layer and a gate insulating film provided on the surface of the SiC layer, and includes an oxide film or an oxynitride film in contact with the surface of the SiC layer, and the oxide film or oxynitride film is B ( Including at least one element selected from boron), Al (aluminum), Ga (gallium), In (indium), Sc (scandium), Y (yttrium), and La (lanthanum). The peak is on the SiC layer side of the gate insulating film, the peak of the element is in the oxide film or the oxynitride film, and the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the side opposite to the SiC layer of the peak A gate insulating film having a region; and a gate electrode provided over the gate insulating film.

In particular, the SiC substrate having the first and second surfaces, the first conductivity type SiC layer provided on the first surface side of the SiC substrate, and the second conductivity type second provided on the surface of the SiC layer. 1 SiC region, a first conductivity type second SiC region provided on the surface of the first SiC region, a SiC layer, and a gate insulating film continuously provided on the surface of the first SiC region. An oxide film or an oxynitride film in contact with the surface of the SiC layer, and the oxide film or the oxynitride film is B (boron), Al (aluminum), Ga (gallium), In (indium), Sc (scandium), And containing at least one element selected from Y (yttrium) and La (lanthanum), the peak of the element in the gate insulating film is on the first SiC region side of the gate insulating film, and the peak of the element is an oxide film or In the oxynitride film and above A gate insulating film in which the concentration of the element on the opposite side of the SiC layer of a peak having a 1 × 10 ¹⁶ cm ^-3 or less in the region, a gate electrode formed on the gate insulating film, formed on the second SiC region on And a second electrode formed on the second surface side of the SiC substrate.

  FIG. 1 is a schematic cross-sectional view showing a configuration of a MOSFET that is a semiconductor device of the present embodiment. This MOSFET (Metal Oxide Semiconductor Field Effect Transistor) 100 is a Double Implantation MOSFET (DIMOSFET) in which a p-well and a source region are formed by ion implantation.

  MOSFET 100 is an n-channel transistor using electrons as carriers. The MOSFET 100 is a vertical transistor that moves carriers between a source electrode on the front surface side of the semiconductor substrate and a drain electrode on the back surface side.

MOSFET 100 includes an SiC substrate (silicon carbide substrate) 12 having first and second surfaces. In FIG. 1, the first surface is the upper surface of the drawing, and the second surface is the lower surface of the drawing. The SiC substrate 12 is, for example, a 4H—SiC SiC substrate (n substrate) having an impurity concentration of 1 × 10 ¹⁸ cm ⁻³ or more and 1 × 10 ¹⁹ cm ⁻³ or less and containing, for example, N (nitrogen) as an n-type impurity. It is.

  The first surface is the C surface, that is, the (000-1) surface. The first surface may be offset from the C surface within a range of 8 degrees or less, for example.

On the first surface of the SiC substrate 12, for example, an n-type SiC layer (n ^- SiC layer) 14 having an n-type impurity impurity concentration of 5 × 10 ¹⁵ or more and 2 × 10 ¹⁶ cm ⁻³ or less is formed. ing. The film thickness of the n ⁻ SiC layer 14 is, for example, about 5 to 20 μm.

On a partial surface of the n ⁻ SiC layer 14, a p-type first SiC region (p well with an impurity concentration of p-type impurities of, for example, 5 × 10 ¹⁵ cm ⁻³ or more and 1 × 10 ¹⁷ cm ⁻³ or less is used. Region) 16 is formed. The depth of the p well region 16 is, for example, about 0.6 μm. The p well region 16 functions as a channel region of the MOSFET 100.

On the partial surface of the first SiC region (p-well region) 16, the n ⁺ -type first impurity concentration of the n-type impurity is, for example, not less than 1 × 10 ¹⁸ cm ^{−3 and not} more than 1 × 10 ²² cm ⁻³ . Two SiC regions (source regions) 18 are formed. The depth of the source region 18 is shallower than the depth of the first SiC region (p well region) 16 and is, for example, about 0.3 μm.

Further, an impurity concentration of, for example, 1 × 10 p-type impurity is formed on a part of the surface of the first SiC region (p-well region) 16 and on the side of the n ⁺ -type second SiC region (source region) 18. A p ⁺ -type third SiC region (p-well contact region) 20 of ¹⁸ or more and 1 × 10 ²² cm ⁻³ or less is formed. The depth of the p well contact region 20 is shallower than the depth of the first SiC region (p well region) 16 and is, for example, about 0.3 μm.

A gate insulating film 28 is formed on the surface of the n ⁻ SiC layer 14 and the first SiC region (p well region) 16 so as to straddle these regions and layers. A gate electrode 30 is formed on the gate insulating film 28. For example, polysilicon or the like can be applied to the gate electrode 30.

  FIG. 2 is an enlarged view of the gate insulating film and the gate electrode portion and the element distribution in the present embodiment. FIG. 2A is an enlarged view of the gate insulating film 28 and the gate electrode 30, and FIG. 2B is a diagram showing element distribution.

  Gate insulating film 28 includes an oxide film or an oxynitride film in contact with the surface of first SiC region 16, and the oxide film or oxynitride film is B (boron), Al (aluminum), Ga (gallium), or In (indium). ), Sc (scandium), Y (yttrium), and La (lanthanum). The peak of the element in the gate insulating film 28 is on the first SiC region 16 side of the gate insulating film 28, and the peak of the element is in the oxide film or the oxynitride film.

  Hereinafter, the case where the element is B (boron), the oxide film or the oxynitride film is a silicon oxide film, and the entire gate insulating film 28 is also a silicon oxide film will be described as an example.

As shown in FIG. 2A, the gate insulating film 28 is in contact with the surface of SiC, that is, the surface of the first SiC region 16, and the interface doped region 28a having a high B concentration and the B on the gate electrode 30 side are completely separated. The undoped region 28b is not doped. Here, in the undoped region 28b which is not doped at all, the concentration of B becomes below the measurement limit by various physical analyzes (SIMS, atom probe, XPS, etc.). Therefore, in the B concentration distribution in the gate insulating film 28, as shown in FIG. 2B, the peak of B exists on the first SiC region 16 side in the gate insulating film 28. Further, the gate insulating film 28 has a region where the B concentration is 1 × 10 ¹⁶ cm ⁻³ or less on the side opposite to the first SiC region (SiC layer) 16 of the B peak.

  For example, when B or the like is diffused from the gate electrode side, or when B is introduced into the entire insulating film, it is difficult to form an insulating film having the structure of the undoped region 28b / interface doped region 28a. In order to obtain an insulating film having this configuration, some way of making such as a lamination of a B-doped film and a B-undoped film is required as will be described later.

On the gate electrode 30, an interlayer insulating film 32 made of, for example, a SiO ₂ film is formed. The first SiC region 16 sandwiched between the second SiC region (source region) 18 and the n ⁻ SiC layer 14 under the gate electrode functions as a channel region of the MOSFET 100.

  Then, a conductive first electrode (source / p-well common electrode) 24 electrically connected to the second SiC region (source region) 18 and the third SiC region (p-well contact region) 20 is formed. I have. The first electrode (source / p-well common electrode) 24 includes, for example, a Ni (nickel) barrier metal layer 24a and an Al metal layer 24b on the barrier metal layer 24a. The Ni barrier metal layer 24a and the Al metal layer 24b may form an alloy by reaction.

  In addition, a conductive second electrode (drain electrode) 36 is formed on the second surface side of the SiC substrate 12. The second electrode (drain electrode) 36 is, for example, Ni.

  In the present embodiment, for example, N (nitrogen) or P (phosphorus) is preferable as the n-type impurity, but As (arsenic) or the like can also be applied. For example, Al (aluminum) is preferable as the p-type impurity, but B (boron), Ga (gallium), In (indium), or the like can also be applied.

  Hereinafter, the operation and effect of the present embodiment will be described in detail. 3 to 8 are explanatory diagrams of the operation and effect of the present embodiment.

FIG. 3 is a diagram showing a band structure of SiC and a SiO ₂ gate insulating film 28 formed on the SiC. The left figure shows the case of 4H-SiC Si (silicon) surface, that is, the (0001) surface. The right figure shows the case of 4H—SiC in the C (carbon) plane, that is, the (000-1) plane.

As shown in the figure, when the Si surface and the C surface are compared, the energy difference (hereinafter also referred to as a band offset (ΔEc)) between the conduction bands of SiC and SiO ₂ is 0.4 eV to 0 for the C surface. About 6 eV. For this reason, when the gate complete film 28 and the gate electrode 30 are formed on SiC, there is a problem that the gate leakage current is larger on the C surface.

  On the other hand, the carrier mobility is higher in the C plane than in the Si plane.

  In order to clarify the difference between the band offset and the mobility between the Si surface and the C surface, the inventor conducted a study based on the first principle calculation. Hereinafter, based on the examination result, the operation and effect of the present embodiment will be described in detail.

In the vicinity of the SiC / SiO ₂ interface, a C (carbon) defect is generated on the SiC side due to oxidation when the gate insulating film 28 is formed. The generation energy of C defects is low near the surface of SiC. Therefore, if there is atomic oxygen, it is easily released as CO.

Furthermore, carbon atoms are released in the process of oxidizing the vicinity of the SiC surface. At this time, since the bond between carbon atoms (CC bond) is strong, a C ₂ dimer structure is formed on the SiO ₂ side. This C ₂ dimer structure is a structure in which O (oxygen) of SiO ₂ is substituted with C (carbon). Hereinafter, the C ₂ dimer structure is referred to as a CoCo structure.

  FIG. 4 is an explanatory diagram of the CoCo structure. 4A shows the second adjacent oxygen substitution structure, and FIG. 4B shows the first adjacent oxygen substitution structure.

There are several variations of the CoCo structure. According to the first principle calculation, in a hexagonal structure formed by Si—O—Si bonds, carbon is substituted for the second adjacent oxygen as shown in FIG. The structure in which C is connected to (referred to as a second adjacent oxygen substitution structure) was the most stable. At this time, the carbons have a structure in which Si and a bond are combined instead of the hydrogen atom of ethylene (H ₂ C═CH ₂ ). This structure appears because C can stably take a three-coordinate structure, and does not appear in Si.

  FIG. 5 is an explanatory diagram of the electronic state of the CoCo structure. FIG. 5A shows the second adjacent oxygen substitution structure, and FIG. 5B shows the first adjacent oxygen substitution structure.

  As for the electronic state of the second adjacent oxygen substitution structure, one empty state and one buried state exist in the band gap. The state where electrons are empty is at the lower end of the conduction band of 4H—SiC, and this state is considered to cause mobility degradation. That is, it is considered that the mobility is lowered by the electrons moving through the channel being trapped in the empty state.

  On the other hand, the state where the electrons are buried is deeper than the upper end of the valence band and does not have a great influence. Since the second adjacent oxygen substitution structure is mainly formed on the Si surface, the carrier mobility is greatly deteriorated.

  Among the CoCo structures, the second most stable structure is a structure in which carbon is substituted for the first adjacent oxygen as shown in FIG. 4B, and C is connected to the remaining Si bond (first adjacent oxygen). (Referred to as a substitution structure). As shown in FIG. 5 (b), the electronic state of the first adjacent oxygen substitution structure is such that the vacant state is about 1.9 eV above the lower end of the conduction band of 4H—SiC. This state does not cause mobility degradation. On the other hand, the state where the electrons are buried is about 0.6 eV higher than the VB of 4H—SiC.

FIG. 6 is a diagram showing the CoCo structure of the Si face and the C face. On the C plane, an interface state exists at the upper end of the valence band of 4H—SiC. For this reason, in the first adjacent oxygen substitution structure, electrons can be taken into the interface state of the C plane. Therefore, electrons are taken from the first adjacent oxygen substitution structure in the SiO ₂ film, and the entire system is stabilized. By this stabilization, conversion from the second adjacent oxygen substitution structure to the first adjacent oxygen substitution structure occurs on the C plane. Therefore, the first adjacent oxygen substitution structure is mainly formed in the vicinity of the interface of the C plane.

At this time, as shown on the right in FIG. 6, the SiC / SiO ₂ interface, a fixed dipole direction to lower the SiC / SiO ₂ interface ΔEc occurs. Therefore, ΔEc is reduced by about 0.4 to 0.6 eV in the C plane compared to the Si plane. In addition, since the first adjacent oxygen substitution structure is mainly formed and the ratio of the second adjacent oxygen substitution structure is low, mobility deterioration on the C plane is smaller than that on the Si plane.

  As described above, when the Si surface and the C surface are compared, the second adjacent oxygen substitution structure becomes more stable on the Si surface, and the first adjacent oxygen substitution structure becomes more stable on the C surface. Therefore, the mobility is relatively small and ΔEc is large on the Si surface. In other words, the mobility is large on the C plane, but ΔEc is small.

In the present embodiment, B (boron), Al (aluminum), Ga (gallium), In (indium), Sc (scandium), Y (yttrium), La, and the like in the interface doped region 28 a in contact with SiC of the gate insulating film 28. It contains at least one element selected from (lanthanum). The _{CO 2 O} 3 _CO structure in SiO ₂ is an electron trap by introducing trivalent or divalent elements such as B, Al, Ga, In, Sc, Y, La, Mg, Ca, Sr, and Ba. Can be converted to Hereinafter, although the calculation result is shown in the case of B, the same applies to other elements.

When in SiO ₂ is B, carbon, in SiO _2, to form a stable CsOB structure than CoCo structure. The CsOB structure is a structure in which C (carbon) substituted with an Si atom, O (oxygen) in contact with the C, and B (boron) substituted with an Si atom in contact with the O on the opposite side are bonded. That is, the Si—O—Si bond portion is replaced with a C—O—B bond. Here, B may be Al, Ga, In, Sc, Y, La, Mg, Ca, Sr, or Ba.

In particular, if B is present at the stage where carbon diffusion into SiO ₂ occurs, the CsOB structure is formed before the CoCo structure is formed. Therefore, the CsOB structure is surely formed. According to the first principle calculation, the CsOB structure has a gain of at least 1.2 eV per CsOB structure and is more stable in energy than the CoCo structure.

  FIG. 7 is an explanatory diagram of the electronic state of the CsOB structure. FIG. 8 is an explanatory diagram of a fixed dipole formed with a CsOB structure.

As shown in FIG. 7, when the electronic state of the CsOB structure is calculated, an electron trap level is formed immediately above the valence band of SiO ₂ . This makes it possible to move electrons from the interface state of the SiC / SiO ₂ interface to this electron trap level. Therefore, due to the movement of electrons to the SiO ₂ side, a fixed dipole having a positive SiC side and a negative SiO ₂ side can be formed as shown in FIG.

When the CsOB structure is formed, the CoCo structure in SiO ₂ disappears on both the Si surface and the C surface. Therefore, deterioration of mobility is suppressed.

Furthermore, ΔEc is improved by the fixed dipole at the interface. Since electrons move from the interface state of the SiC / SiO ₂ interface to this electron trap level, a fixed dipole is formed regardless of the plane orientation (for example, Si plane, C plane, A plane, M plane are representative). Is done. Therefore, ΔEc is improved by about 0.2 eV to 0.6 eV from the original position regardless of the plane orientation. Therefore, the gate leakage current is suppressed as compared to before the CsOB structure is introduced. For example, if a graph of gate leakage and gate applied voltage is written, the voltage at which gate leakage starts is shifted to the larger one.

  For example, in the C plane, ΔEc was reduced by about 0.4 eV compared to the Si plane (ΔEc was reduced due to charge trapping in the CoCo structure), but improved by about 0.5 eV (to CsOB). The introduction of a charge trap makes a dipole in the opposite direction). Therefore, the difference is expected to be about 0.9 to 1.0 eV.

Further, since it is considered that the second adjacent oxygen substitution structure remains even in the C plane, the mobility of the C plane is also improved. Therefore, the mobility is expected to reach 50 cm ² / Vs or higher regardless of the plane orientation.

In the present embodiment, the B peak exists on the first SiC region 16 side in the gate insulating film 28. Since B is present at a high concentration in the interface dope region 28a near the SiC / SiO ₂ interface, a CsOB structure is effectively formed.

In the undoped region 28b on the gate electrode 30 side away from the SiC / SiO ₂ interface, B is not doped as much as possible. It is possible to prevent deterioration of reliability due to the presence of CsOB structure and B between lattices. That is, it must be excluded that an extra electron trap level is generated inside the film. Although it is desirable that the undoped region 28b does not contain B at all, it does not exclude that a certain amount of B is contained. B concentration of undoped region 28b is smaller that (1 × ¹⁰ 16 ^{cm -3} or less extent) are desirable as much as possible, as long as at least the measurement limit or less (eg SIMS, degree 1 × ¹⁰ 16 ^{cm -3} or less, the atom probe If it is, it is desirable that it is about 1 × 10 ¹⁵ cm ⁻³ or less. Here, if it is 1 × 10 ¹⁶ cm ⁻³ or less, it is considered that the mobility and reliability are not greatly affected. Can be considered.

CsOB structure in SiO _2, as is possible from the interface state of the SiC / SiO ₂ interface moving electrons, it is preferable to present the SiC / SiO ₂ interface area. Therefore, it is desirable that the peak of B is in the range of 5 nm or less from the interface between SiC (first SiC region 16) and the gate insulating film 28. If the distance is longer than this distance, it becomes difficult to move electrons from the interface state, and there is a possibility that a fixed dipole is not formed. Furthermore, in order to ensure electron transfer, it is preferable to make it less than 1 nm. The wave function has a tail to about 2.5 nm, but is almost localized within 1 nm. Therefore, in order to ensure a sufficient overlap, it is desirable that it is in the range of 5 nm or less, and it is certain if it can be made less than 1 nm.

The B concentration at the peak of B is preferably 5.6 × 10 ¹⁷ cm ⁻³ or more and 5 × 10 ²⁰ cm ⁻³ or less. If it is below the above range, the amount of the fixed dipole is not sufficient, and there is a possibility that sufficient ΔEc cannot be improved. It is desirable to ensure ΔEc> 0.01V.

The shift X (volt) due to the fixed polarization amount, which is composed of the insulating film and the SiC interface, is
X = (charge) × (area density) × (polarization length) / dielectric constant. In more detail,
X (volt) = (charge) × (surface density) × (polarization length) / dielectric constant
= (Charge 2 × 1.602 × 10 ⁻¹⁹ Coulomb) × (Area density cm ⁻² ) ×
(Polarization length × 10 ⁻⁸ cm) / (relative permittivity ε) / [8.854 × 10 ⁻¹² (fard
/ M)]
= 1.81 × 10 ⁻¹⁴ (several surface density cm ⁻² unit) × (polarization length Å unit) / (relative permittivity)
It becomes.

Here, the electric charge of the dangling bond at the SiC interface is 1, the surface density is about 10 ¹³ cm ^{−2 at the} maximum, and the dielectric constant is 10. Since the lower limit of the number density is considered, the maximum polarization length is 10 nm, that is, about 100 Å (angstrom). Therefore,
X = 1.81 × 10 ⁻¹⁴ × 1 × 10 ¹³ × ^100/10 = 1.8 (V)
It becomes. Since it is desired to ensure 0.01 V, 0.0056 × 10 ¹³ cm ⁻² or more is necessary. Below this, the required shift amount may not be obtained. It is desirable that this is simply converted into a density with a uniform film thickness of 1 nm to be 5.6 × 10 ¹⁷ cm ³ or more.

In addition, when the above range is exceeded, excess B forms an electron trap, which may cause a decrease in mobility and a deterioration in the reliability of the gate insulating film. The amount of carbon that is supposed to come out of the substrate is, for example, 5 × 10 ²⁰ cm ⁻³ or less when a SiC / SiO ₂ (thermal oxide film) structure is formed and measured. Therefore, if more B is introduced, a CsOB structure cannot be formed, and there is a high possibility that B will diffuse to the electrode side of the insulating film. Therefore, 5 × 10 ²⁰ cm ⁻³ or less is desirable.

Also, from the viewpoint of moving electrons from the interface state existing at the SiC / SiO ₂ interface to the CsOB structure without excess or deficiency, the amount of B is preferably equal to the amount of the interface state. From this viewpoint, it is more desirable that the peak B concentration is 5 × 10 ¹⁸ cm ⁻³ or more and 5 × 10 ¹⁹ cm ⁻³ or less. The surface density of the interface state can be reduced to about 5 × 10 ¹² cm ⁻² by introducing nitrogen or the like. Therefore, in order to see the upper limit, it is desirable that the width is 1 × 10 ¹⁹ cm ⁻³ or less as the 1 nm width. The lower limit is preferably 5 × 10 ¹⁸ cm ⁻³ or more as a 10 nm width.

From the above, it is desirable that the B concentration at the peak of B is 5.6 × 10 ¹⁷ cm ⁻³ or more and 5 × 10 ²⁰ cm ⁻³ or less. Furthermore, 5 × 10 ¹⁸ cm ⁻³ or more and 5 × 10 ¹⁹ cm ⁻³ or less are desirable. The target value of the B amount is, for example, about 3 × 10 ¹⁹ cm ⁻³ .

Further, in order to form the CsOB structure, it is necessary that the B interface doped region 28a of the gate insulating film 28 contains C (carbon) as a B pair. On the other hand, in the undoped region 28b on the gate electrode 30 side away from the SiC / SiO ₂ interface, the concentration of C (carbon) forming an electronic state serving as an electron trap should be as low as possible from the viewpoint of not deteriorating reliability. desirable. In other words, it is preferable to employ a process that does not contain C as much as possible so that C is not contained at least in the initial film formation stage.

Therefore, it is desirable that a part of the gate insulating film 28 contains C (carbon), and the peak of C (carbon) in the gate insulating film 28 is on the SiC (first SiC region 16) side of the gate insulating film 28. . Furthermore, it is desirable that the peak of C (carbon) is within a range of 10 nm or less from the interface between SiC (first SiC region 16) and the gate insulating film 28. Although it is desirable that the undoped region 28b does not contain C at all, it does not exclude that a certain amount of C is contained. Carbon concentration of the non-doped region 28b is smaller that (1 × ¹⁰ 16 ^{cm -3} or less extent) are desirable as much as possible, as long as at least the measurement limit or less (eg SIMS, degree 1 × ¹⁰ 16 ^{cm -3} or less, the atom probe If it is, it is desirable that it is about 1 × 10 ¹⁵ cm ⁻³ or less. Here, if it is 1 × 10 ¹⁶ cm ⁻³ or less, it is considered that the mobility and reliability are not greatly affected. Can be considered.

In order to form the CsOB structure, it is desirable that the amount ratio of C (carbon) and B (boron) in the vicinity of the SiC / SiO ₂ interface of the gate insulating film 28 is close to 1: 1. Therefore, it is desirable that the concentration of B at the peak of B is 80% or more and 120% or less of the concentration of C (carbon) at the peak of C (carbon). The C (carbon) peak concentration of C (carbon) is preferably 4.5 × 10 ¹⁷ cm ⁻³ or more and 6 × 10 ²⁰ cm ⁻³ or less, and 5.6 × 10 ¹⁷ cm ^−3. It is desirable that it is 5 × 10 ²⁰ cm ⁻³ or less. Furthermore, it is preferably 4 × 10 ¹⁸ cm ⁻³ or more and 6 × 10 ¹⁹ cm ⁻³ or less, and more preferably 5 × 10 ¹⁸ cm ⁻³ or more and 5 × 10 ¹⁹ cm ⁻³ or less.

  Further, it is desirable that the distributions of the C concentration and the B concentration in the film thickness direction coincide within a range of 80% to 120%. For example, if the process shown in the second manufacturing method of the following embodiment is used, the B concentration can be brought close to the C concentration by adjusting the B concentration. At this time, the distributions of the C concentration and the B concentration in the film thickness direction are in good agreement, and are approximately 1: 1. In the gate insulating film by the second manufacturing method, B forms a Cs and CsOB structure, and excess B remains on the top of the film. The remaining B is removed by HF etching.

  The concentration and concentration distribution of B (boron) and C (carbon) in the gate insulating film 28 can be measured using, for example, a secondary ion mass spectrometry (SIMS) or an atom probe (Atom Probe).

  In the present embodiment, a fixed dipole with the SiC (first SiC region 16) side as positive and the gate insulating film 28 side as negative at the interface between SiC (first SiC region 16) and the gate insulating film 28 is used. Exists. The presence / absence and polarity of the fixed dipole can be determined, for example, by measuring the film thickness dependence of the CV characteristics of the gate insulating film 28.

  The effective work function of the gate electrode is measured by changing the thickness of the gate insulating film. The flat band shift (Vfb) of the CV curve and the film thickness dependence of the gate insulating film are measured, and the effective work function (φeff) is calculated by extrapolation to the point where the insulating film is zero, that is, by the value of the intercept of the Vfb axis. Can be requested.

For example, the work function of phosphorus-doped polysilicon is about 4.1 eV, but when SiO ₂ is formed on the C surface of the SiC substrate, the effective work function is about 4.6 eV. This indicates that a dipole that causes a shift of 0.5 eV, in which the insulating film side is positive and the substrate side is negative, is formed at the interface.

  Next, for example, when measured in the case of the present embodiment manufactured by the following second manufacturing method, the effective work function is about 3.6 eV. This indicates that a dipole that causes a shift of 0.5 eV, which is negative on the insulating film side and positive on the substrate side, is formed at the interface. Thus, the characteristics of the fixed dipole can be measured.

  Note that the case where the element contained in the gate insulating film 28 is B (boron), the oxide film or the oxynitride film is a silicon oxide film, and the entire gate insulating film 28 is also a silicon oxide film has been described as an example. Other configurations can also be applied.

  Even in the case of an oxynitride film other than the oxide film, it is possible to form a CsOB structure in the film and form a fixed dipole at the interface by containing O (oxygen).

  As described above, elements contained in the oxide film or the oxynitride film include, in addition to B (boron), Al (aluminum), Ga (gallium), In (indium), Sc (scandium), Y (yttrium), An element selected from La (lanthanum) is applicable. As an element contained in the oxide film or the oxynitride film, it is desirable to apply B (boron) or Al (aluminum) because it can be easily replaced with Si and can easily form a CsOB structure. In addition, it is more preferable to use B (boron) because it is less likely to form oxides and silicides than Al and it is easier to form a CsOB structure.

As the oxide film or oxynitride film, for example, a hafnium oxide film (HfO ₂ ), a hafnium oxynitride film (HfON), a hafnium silicate oxide film (HfO ₂ ), in addition to the silicon oxide film (SiO ₂ ) and the silicon oxynitride film (SiON). HfSiO), hafnium silicate oxynitride film (HfSiON), zirconium oxide film (ZrO ₂ ), zirconium oxynitride film (ZrON), zirconium silicate oxide film (ZrSiO), zirconium silicate oxynitride film (ZrSiON) alumina (Al ₂ O ₃₎ It is also possible to apply a high dielectric film such as alumina nitride (AlON).

However, considering the height of the barrier, a SiO ₂ film or a SiON film is suitable near the interface. For example, at least part of the vicinity of the SiC side interface is formed of a SiO ₂ film or a SiON film, and from the substrate interface side, B-doped SiON (1 nm) / undoped SiO ₂ (HTO film, 9 nm) / undoped HfSiON (30 nm) For example, the structure may be substantially three or more layers. At this time, it seems that there is some B diffusion, but according to the diffusion, a part of the first layer and the second layer on the SiC side is regarded as an interfacial doped region, and the remainder of the second layer (with less diffusion) It is also possible to regard the upper layer as an undoped region.

Here, attention should be paid to Al in Al ₂ O ₃ . It is important that Al in this embodiment is substituted at the Si position in SiO ₂ to form a CsOAl structure. On the other hand, Al ₂ O ₃ is not a structure in which Si in SiO ₂ is replaced with Al but a structure peculiar to the Al ₂ O ₃ material.

  The gate insulating film 28 includes an oxide film or an oxynitride film having a high concentration of an element such as B, and a film having a concentration of an element such as B sufficiently lower than that of the oxide film or the oxynitride film, that is, a measurement limit or less. It may be a laminated film with a film in which only an amount of B is detected. For example, the oxide film or the oxynitride film may be a silicon oxide film having a high B concentration, and a hafnium oxide film having a sufficiently low B concentration may be stacked thereon. Alternatively, it may be a laminated film of a silicon oxide film having a high B concentration and a silicon nitride film having a sufficiently low B concentration. Alternatively, a stacked film of a silicon oxide film having a high B concentration and a C concentration and an HTO (High Temperature Oxide) film having a sufficiently low B concentration and a carbon concentration may be used. Of course, it is better that the undoped region is not doped with B. If there is, it only degrades the characteristics.

  Next, a first manufacturing method of the semiconductor device of this embodiment will be described.

The first manufacturing method of the semiconductor device of this embodiment is as follows: B (boron), Al (aluminum), Ga (gallium), In (indium), Sc (scandium), Y (yttrium), La on the SiC layer. A first insulating film made of an oxide film or an oxynitride film containing at least one element selected from (lanthanum) is formed, and an insulation having a sufficiently lower element concentration than the first insulating film is formed on the first insulating film. A film is formed, and a gate electrode is formed over the insulating film. Here, sufficiently low means no doping, but actually means below the measurement limit. For example, it is 1 × 10 ¹⁶ cm ⁻³ or less.

  FIG. 9 is a process flow diagram illustrating the first method for manufacturing a semiconductor device of this embodiment. 10 to 15 are schematic cross-sectional views showing a first manufacturing method of the semiconductor device of this embodiment.

As shown in FIG. 9, the first manufacturing method of the semiconductor device includes n ⁻ SiC layer formation (step S100), p-type impurity ion implantation (step S102), n-type impurity ion implantation (step S104), and p-type impurity. Ion implantation (step S106), annealing (step S108), first insulating film formation (step S110), second insulating film formation (step S112), gate electrode formation (step S114), interlayer film formation (step S116) First electrode formation (step S118), second electrode formation (step S120), and annealing (step S122) are provided.

First, an n-type SiC substrate 12 containing P (phosphorus) or N (nitrogen) as an n-type impurity at an impurity concentration of about 5 × 10 ¹⁸ cm ⁻³ and having a thickness of 300 μm and a low resistance of 4H—SiC, for example. prepare.

In step S100, a high-resistance n ^- SiC layer 14 containing, for example, N as an n-type impurity at an impurity concentration of about 1 × 10 ¹⁶ cm ⁻³ and having a thickness of about 10 μm on the C surface of the SiC substrate 12 by epitaxial growth. Is epitaxially grown.

Thereafter, a first mask material 42 made of, for example, SiO ₂ is formed by patterning using photolithography and etching. In step S102, using the first mask material 42 as an ion implantation mask, Al, which is a p-type impurity, is ion-implanted into the n ⁻ SiC layer 14 to form a first SiC region (p well region) 16. (FIG. 10).

Thereafter, a second mask material 44 made of, for example, SiO ₂ is formed by patterning using photolithography and etching. In step S104, using this second mask material 44 as an ion implantation mask, N (or P), which is an n-type impurity, is ion-implanted into the n ⁻ SiC layer 14 to form a second SiC region (source region) 18. Is formed (FIG. 11).

Thereafter, a third mask material 46 made of, for example, SiO ₂ is formed by patterning using photolithography and etching. In step S106, using the third mask material 46 as an ion implantation mask, Al, which is a p-type impurity, is ion-implanted into the n ⁻ SiC layer 14 to form a third SiC region (p-well contact region) 20. (FIG. 12).

  In step S108, annealing for activating the p-type impurity and the n-type impurity is performed. For this annealing, for example, argon (Ar) gas is used as an atmosphere gas, and a condition such as a heating temperature of 1600 ° C. and a heating time of 30 minutes is used. At this time, activation of impurities introduced into the SiC can be realized, but diffusion is slight.

  Through step S110 and step S112, the gate insulating film 28 is formed (FIG. 13). The first insulating film formed in step S110 corresponds to the interface doped region 28a shown in FIG. The second insulating film formed in step S112 corresponds to the undoped region 28b shown in FIG.

  In step S110, an oxide film containing at least one element selected from B (boron), Al (aluminum), Ga (gallium), In (indium), Sc (scandium), Y (yttrium), and La (lanthanum) A first insulating film made of an oxynitride film is formed.

  For example, the first insulating film is formed by the following process. First, a silicon film containing B (boron) is deposited by, for example, a CVD (Chemical Vapor Deposition) method. Then, this silicon film is thermally oxidized to form a silicon oxide film containing B at a high concentration.

The film thickness of the silicon film containing B (boron) is, for example, not less than 1 nm and not more than 5 nm. The film thickness of the silicon oxide film formed by thermal oxidation is, for example, not less than 2 nm and not more than 10 nm. In this embodiment, approximately 3 nm of B-doped polysilicon is formed, and approximately 6 nm of B-doped SiO ₂ is obtained. B is concentrated in the vicinity of the SiC / SiO ₂ interface due to the snow-plowing effect (B in Si is pushed into the Si side when oxidized). Furthermore, stabilization due to electron transfer to CsOB is also a gain that occurs only when the CsOB structure is in the vicinity of the interface, so the CsOB structure is concentrated at the interface. Thus, due to the two effects, B and C are concentrated at the interface of SiC / SiO ₂ and have a peak.

In this example, before the SiC surface is oxidized and C (carbon) is released, a film containing B is deposited on the SiC surface. Therefore, when the silicon oxide film formed, _{before C} O _{C O} structure is formed on SiC / SiO ₂ interface vicinity of the silicon oxide film, can be formed CsOB structure. Therefore, the CsOB structure can be formed by a relatively low temperature process. A high temperature (for example, 1200 ° C.) is required to break the CoCo structure once formed into a CsOB structure, but here, oxidation is performed at 700 ° C. to 900 ° C. A temperature is selected at which Si oxidizes but SiC hardly oxidizes. This temperature is 600 ° C. or more and 1,000 ° C. or less, and it is preferable that the temperature is low in consideration of interface deterioration, but it is preferable that the temperature is high in consideration of throughput. Considering the plane orientation dependency of SiC oxidation, it is appropriate that the temperature is 700 ° C. for the C plane, 800 ° C. for the A and M planes, and 900 ° C. for the Si plane.

  The first insulating film may be formed by oxynitriding the silicon film containing B (boron). Further, the first insulating film may be formed, for example, by forming a silicon oxide film containing B (boron) by, for example, a CVD method.

In step S112, a second insulating film having a lower concentration of an element such as B (boron) than the first insulating film is formed on the first insulating film. For example, a film not containing an element such as B is formed. The second insulating film is, for example, an HTO film formed by a CVD method. The HTO film is a desirable film as the second insulating film because of its low C (carbon) content. After step S112, annealing in a non-oxidizing atmosphere (for example, 1,000 ° C. for 30 minutes in N ₂ ) may be performed for densification of the HTO film.

The film thickness of the second insulating film is, for example, 30 nm or more and 100 nm or less. The thinner, the better the drive performance, but the greater the leak. In the thinnest case, the film thickness is such that dielectric breakdown is avoided, and is considered to be about 30 nm. To increase the insulation resistance, a thicker film is better, but the drive performance is too low. The limit is about 100 nm. Actually, about 50 nm is usually used in terms of the SiO ₂ film thickness. This is because when ΔEc is low, the leak cannot be stopped unless it is about 50 nm. However, in this embodiment, ΔEc is improved, so that it is possible to aim for a thinner region. In this embodiment, the target is around 40 nm. Here, an approximately 40 nm HTO film is used.

In step S114, for example, a polysilicon gate electrode 30 is formed on the gate insulating film 28 (FIG. 14). In step S116, an interlayer insulating film 32 of, eg, a SiO ₂ film is formed on the gate electrode 30 (FIG. 15).

  Thereafter, in step S118, the conductive first electrode (source / p well) electrically connected to the second SiC region (source region) 18 and the third SiC region (p well contact region) 20 is provided. Common electrode) 24 is formed. The first electrode (source / p-well common electrode) 24 is formed by sputtering of Ni (nickel) and Al, for example.

In step S <b> 120, a conductive second electrode (drain electrode) 36 is formed on the second surface side of the n ⁻ SiC substrate 12. The second electrode (drain electrode) 36 is formed by sputtering of Ni, for example.

  In step S122, it is desirable to perform annealing at as low a temperature as possible in order to reduce the contact resistance between the first electrode 24 and the second electrode. Annealing is performed at 400 ° C. to 1000 ° C. in an argon gas atmosphere, for example. Here, it is 800 degreeC.

  The MOSFET 100 shown in FIG. 1 is formed by the above manufacturing method.

The MOSFET 100 of this embodiment includes a CsOB structure and a fixed dipole on the SiC side of the gate insulating film 28. This increases the band offset (ΔEc) and suppresses the leakage current of the gate insulating film formed on the C plane. In addition, since there is no C 2 _{O 2 O} 3 _{O CO} structure on the SiC side of the gate insulating film 28 (C in the film is converted to a CsOB structure), the carrier mobility is further improved. According to the method for manufacturing the MOSFET 100 of the present embodiment, the MOSFET 100 having the above-described excellent effects can be formed.

When the semiconductor device is manufactured by the first manufacturing method, ΔEc is increased by about 0.9 eV from the initial value. Thus, for example, even in the case of a thin film of 46 nm, gate leakage is almost eliminated. Moreover, the mobility was high and was about 100 cm ² / Vs. This value of 100 cm ² / Vs is sufficient for using the MOSFET. Since there is no trap in the vicinity of the interface, the reliability is also improved, and the threshold value shift is less than 0.1 V when 1000 hours have passed after applying ± 20 V. Thus, (1) excellent drive performance (effect of thinning), (2) no gate leakage, (3) high mobility (effect in which CsOB is concentrated on the number of interfaces of nm), and (4) high reliability A SiC MOSFET is obtained.

Looking at the SIMS data of the gate insulating film manufactured by the first manufacturing method, B and C have a peak at a position of about 0.8 nm inside SiO ₂ from the interface, and about 2 × 10 ¹⁹ / cm ³ ( The surface density is about 2 × 10 ¹³ / cm ² ), and the positions are in good agreement. Thereby, it is considered that ΔEc is increased by about 0.9 eV from the initial value. Further, at a position 5 nm or more away from the peak, the concentrations of B and C are below the measurement limit.

  Furthermore, a CsOB structure is also formed at the interface between the n-SiC layer 14 and the gate insulating film 28 (hereinafter referred to as the JFET region interface). Since ΔEc is also high here, the resistance against dielectric breakdown at this portion is also improved.

  Next, a second manufacturing method of the semiconductor device of this embodiment will be described. The second manufacturing method is different from the first manufacturing method in the method for manufacturing the gate insulating film.

For example, the first insulating film is formed by the following process. First, a silicon film containing B (boron) is deposited by, for example, a CVD (Chemical Vapor Deposition) method. Then, this silicon film is thermally oxidized to form a silicon oxide film containing B at a high concentration. The film thickness of the silicon film containing B (boron) is, for example, about 3 nm, and a B-doped SiO ₂ film of about 6 nm is formed by thermal oxidation.

In the process of thermal oxidation, (1) B concentrates on the SiC side (extruded to the Si side during oxidation: snow plow effect), (2) C is released from the substrate, (3) CsOB structure formation is formed, As a result, an interface dipole is formed. B and C form a pair on the SiC side in the SiO ₂ insulating film, cause electron transfer from the SiC interface state, form a dipole, and stabilize. Thus, the CsOB structure is concentrated in the vicinity of the SiC layer.

Here, HF treatment is added to the surface of the B-doped SiO ₂ . Since B and C in a region away from the substrate become traps, they need to be reduced as much as possible. Therefore, it is effective to reduce the film thickness to 10 nm or less by an etching process, for example, HF process. The ideal thickness is 1 nm or less. For example, the thickness is reduced to 1 nm. As a result, a first insulating film is formed. This 1 nm region is the interface region, and the CsOB structure is confined in this region.

  Next, on the first insulating film, an HTO (High Temperature Oxide) film that basically does not contain B or C is formed as a second insulating film, for example, by 40 nm. When the interface is thinned to 1 nm, the equivalent film thickness of the entire gate insulating film is about 41 nm.

In the second manufacturing method, a gate insulating film having a first insulating film of 1 nm and a second insulating film of 40 nm is formed. In this case, B and C are not introduced at all in the HTO separated from the substrate by 1 nm or more. When confirmed by the atom probe method, peaks of B and C are observed at a position less than 1 nm from the interface in SiO ₂ , and both are approximately 2 × 10 ¹⁹ / cm ³ , and the amounts and distributions are in good agreement. .

On the other hand, in the undoped region above it, the amount of B and C is below the measurement limit (1 × 10 ¹⁶ / cm ^3, but in reality it is hardly included). Since B or C in SiO ₂ forms a trap, it is desired that it be introduced only at the SiC / SiO ₂ just interface. According to the second manufacturing method, B or C forms a stable structure called CsOB at the interface of 1 nm, forms a dipole, and does not become a new trap. C and B are confined in 1 nm and hardly diffuse into the HTO film formed at a position 1 nm or more away from the interface.

As a result, in the gate insulating film manufactured by the second manufacturing method, ΔEc is increased by about 1.0 eV from the initial value. Thus, the gate leakage is almost eliminated despite the thin film of 41 nm. Moreover, the mobility is high, approximately 120 cm ² / Vs. Since there is no trap in the vicinity of the interface, the reliability is also improved, and the threshold value shift is less than 0.1 V when 1000 hours have passed after applying ± 20 V.

  Thus, (1) excellent drive performance (effect of thinning), (2) no gate leak, (3) high mobility ((effect in which CsOB is concentrated at 1 nm interface)), (4) high reliability A SiC MOSFET was obtained, and the point that the breakdown resistance at the interface of the JFET region was improved was the same as that produced by the first production method.

Looking at the SIMS data, B and C have a peak at a position of about 1 nm inside SiO ₂ from the interface and are about 3 × 10 ¹⁹ / cm ³ (area density is about 3 × 10 ¹³ / cm ² ). , The position is well matched. Thereby, it is considered that ΔEc is increased by about 1.0 eV compared to the initial value. Further, at a position 2 nm or more away from the peak, the concentrations of B and C are below the measurement limit. An ultra-thin interface dope layer is formed.

  Here, substances other than B can be formed in the same manner. Several nanometers of silicon containing Al (aluminum), Ga (gallium), In (indium), Sc (scandium), Y (yttrium), and La (lanthanum) are formed. For example, it may be formed by sputtering using a silicide pellet of these substances as a target. Alternatively, these materials and silicon may be alternately formed by CVD. Thus, amorphous silicon containing these substances may be formed and oxidized. If the oxide film thus formed is subjected to HF treatment and subjected to a process according to the second manufacturing method, a MOS interface having similar characteristics can be formed.

Next, a third manufacturing method of the semiconductor device of this embodiment will be described. The third manufacturing method is different from the first manufacturing method in the method for manufacturing the gate insulating film. Instead of thermal oxidation in the first manufacturing method, NO oxynitridation (the same applies when N ₂ O is used) was performed.

For example, the first insulating film is formed by the following process. First, a silicon film containing B (boron) is deposited by, for example, a CVD (Chemical Vapor Deposition) method. Then, the silicon film is subjected to NO oxynitridation to form a silicon oxynitride film containing B at a high concentration. The film thickness of the silicon film containing B (boron) is, for example, about 3 nm, and a B-doped SiO ₂ film of about 6 nm serving as the first insulating film is formed by NO oxynitridation.

In the process of NO oxynitriding, (1) B concentrates on the SiC side (it is pushed out to the Si side during oxidation: snow shoveling effect), (2) C is released from the substrate, (3) CsOB structure formation is formed, As a result, an interface dipole is formed. B and C form a pair on the SiC side in the SiO ₂ insulating film, cause electron transfer from the SiC interface state, form a dipole, and stabilize. Thus, the CsOB structure is concentrated in the vicinity of the SiC layer. (4) In NO oxynitriding, N is further introduced into the interface, and the remaining interface defects are repaired.

If there is a dangling bond at the interface, it will be in a high energy state, but if nitrogen comes, it will convert it to a low energy state. That is, substitution of some surface Si in a high energy state with N and introduction of nitrogen into carbon Vc in the substrate occur. Here, the point of adding HF treatment to the surface of the thin film SiO ₂ is the same.

  Next, a second insulating film having a lower concentration of an element such as B (boron) than the first insulating film is formed over the first insulating film. For example, a film not containing an element such as B is formed. The second insulating film is, for example, an HTO film formed by a CVD method.

In the gate insulating film formed by the third manufacturing method, B and C are not introduced at all in the HTO separated from the substrate by 1 nm or more. When confirmed by the atom probe method, peaks of B and C are observed at a position less than 1 nm from the interface in SiO ₂ , and both are approximately 2 × 10 ¹⁹ / cm ³ , and the amounts and distributions are in good agreement. .

  On the other hand, in the undoped region above it, the amounts of B and C are below the measurement limit. According to the third manufacturing method, B and C form a stable structure called CsOB at the interface of 1 nm, form a dipole, and do not become a new trap. C and B are confined in 1 nm and hardly diffuse into the HTO film formed at a position 1 nm or more away from the interface. Furthermore, nitrogen is introduced into the just interface, and mobility degradation is suppressed.

As a result, in the gate insulating film manufactured by the third manufacturing method, ΔEc is increased by about 1.0 eV from the initial value. Thus, the gate leakage is almost eliminated despite the thin film of 41 nm. Moreover, the mobility is high and is approximately 250 cm ² / Vs. Since the dangling bonds at the just interface with nitrogen and the carbon defects in the substrate are added, high mobility can be obtained. Since there is no trap in the vicinity of the interface, the reliability is improved, and the threshold shift is less than 0.1 V when 1000 hours have passed after applying ± 20 V.

  Thus, (1) excellent drive performance (thinning effect), (2) no gate leakage, (3) high mobility ((effect of concentrating CsOB at 1 nm interface and nitrogen termination effect), (4 A high-reliability SiC MOSFET is obtained, and the dielectric breakdown resistance at the interface of the JFET region is improved as in the case of the first or second manufacturing method.

Looking at the SIMS data, B and C have a peak at a position of about 1 nm inside SiO ₂ from the interface and are about 3 × 10 ¹⁹ / cm ³ (area density is about 3 × 10 ¹³ / cm ² ). , The position is well matched. Thereby, it is considered that ΔEc is increased by about 1.0 eV compared to the initial value. Further, at a position 2 nm or more away from the peak, the concentrations of B and C are below the measurement limit. An ultra-thin interface dope layer is formed. Further, nitrogen has a peak at the just interface, and a tail is also drawn on the SiC substrate side. Such high mobility is considered to be an overspec for the MOSFET, but is optimal when a bipolar device such as an IGBT requires a high energy density MOS interface.

As a comparative manufacturing method, consider a manufacturing method in which a thick film SiO ₂ (about 55 nm) is formed as a gate insulating film and B is diffused later. FIG. 16 is a diagram showing an element (B) distribution in the gate insulating film of this embodiment and the comparative embodiment. As shown in FIG. 16, in the comparative embodiment, an insulating film in which B or C is not doped is formed on the electrode side of the insulating film, and sufficient B is concentrated on the interface on the SiC side. It is extremely difficult to make the element distribution shown.

  As a result, it becomes difficult to improve ΔEc, and the gate leakage becomes large when the insulating film thickness is made thin. Therefore, it is impossible to make the gate insulating film thinner, and improvement in drive characteristics cannot be expected. In addition, since a CoCo structure is formed in the insulating film, a portion that is not converted to the CsOB structure remains, which causes mobility degradation. At the same time, since B and C diffused throughout the insulating film form traps, a large threshold fluctuation occurs.

Looking at the SIMS data of the gate insulating film manufactured in the comparative form, B has a broad peak at a position of about 6 nm inside SiO ₂ from the interface, but about 1 × 10 ¹⁷ / cm ³ (in terms of surface density) 1 × 10 ¹⁰ / cm ² ). Further, since the distribution of C is only uniformly distributed and has no peak, it is considered that the CsOB structure is not formed.

As a result, there is no change in ΔEc in the gate insulating film manufactured in the comparative form. In addition, it can be seen that the concentration of B and C is not lower than the measurement limit even at a position 10 nm or more away from the interface in the insulating film, and charge traps are distributed. Thus, the mobility is low and is about 10 cm ² / Vs. The threshold fluctuation is also large, and a fluctuation of 2 V or more is observed in the measurement at ± 20 V and 1000 hours.

(Second Embodiment)
In the present embodiment, the first embodiment is a MOSFET formed on the C plane, that is, the (000-1) plane, except that it is formed on the Si plane, that is, the (0001) plane. This is the same as in the first embodiment. Therefore, description of the contents overlapping with those of the first embodiment is omitted.

  FIG. 17 is a schematic cross-sectional view showing the configuration of a MOSFET that is a semiconductor device of this embodiment. MOSFET 200 is a vertical transistor that moves carriers between a source electrode on the front surface side of the semiconductor substrate and a drain electrode on the back surface side.

This MOSFET 200 includes an SiC substrate 12 having first and second surfaces. In FIG. 1, the first surface is the upper surface of the drawing, and the second surface is the lower surface of the drawing. The SiC substrate 12 is, for example, a 4H—SiC SiC substrate (n substrate) having an impurity concentration of 1 × 10 ¹⁸ cm ⁻³ or more and 1 × 10 ¹⁹ cm ⁻³ or less and containing, for example, N (nitrogen) as an n-type impurity. It is.

  The first surface is the Si surface, that is, the (0001) surface. The first surface may be offset from the Si surface within a range of, for example, 8 degrees or less.

The structure and manufacturing method are the same as those in the first embodiment except that the surface of SiC formed on the SiC substrate 12 is an Si surface. However, when performing the process of depositing silicon containing B and oxidizing it, it is preferable to consider the plane orientation dependence of the oxidation of SiC. 700 ° C. is appropriate for the C surface, 800 ° C. for the A and M surfaces, and 900 ° C. for the Si surface.

  Similar to the MOSFET 100 of the first embodiment, the MOSFET 200 of the present embodiment also includes a CsOB structure and a fixed dipole on the SiC side of the gate insulating film 28. For this reason, the band offset (ΔEc) is increased, and the leakage current of the gate insulating film formed on the Si surface is further suppressed. Further, since there is no CoCo structure on the SiC side of the gate insulating film 28, carrier mobility is improved.

  As a result, ΔEc is increased by about 0.5 eV compared to the initial value. Thus, the gate leak was almost eliminated despite the thin film of 46 nm. Moreover, the mobility was high and was about 80 cm 2 / Vs. Since there is no trap in the vicinity of the interface, the reliability is improved, and the threshold shift is less than 0.1 V when 1000 hours have passed after applying ± 20 V.

  Thus, (1) excellent drive performance (effect of thinning), (2) no gate leak, (3) high mobility (effect in which CsOB is concentrated near the interface), and (4) highly reliable SiC MOSFET is obtained. Further, the dielectric breakdown resistance at the JFET region interface is improved as in the first embodiment.

  As described above, in the embodiment, the C plane and the Si plane of 4H—SiC have been described as examples. However, the effects of the CsOB structure and the fixed dipole are also exhibited in other plane orientations such as the A plane and the M plane. Further, although the case of 4H—SiC has been described as an example of the crystal structure of silicon carbide, the present invention can also be applied to silicon carbide having other crystal structures such as 6H—SiC and 3C—SiC.

  In the embodiment, the vertical MOSFET in which the gate insulating film and the gate electrode are formed on the upper surface of the SiC has been described as an example. However, the vertical type in which the gate insulating film and the gate electrode are formed on the inner surface of the trench formed in the SiC. It is also possible to apply the present invention to a MOSFET.

  In the embodiments, an n-channel transistor using electrons as carriers has been described as an example. However, the present invention can also be applied to a p-channel transistor using holes as carriers.

  The present invention can also be applied to other devices having a structure in which a gate insulating film and a gate electrode are formed on the SiC surface, for example, a vertical IGBT. In particular, in the IGBT, as shown in the second manufacturing method of the first embodiment, it is significant to increase mobility by introducing nitrogen. In the bipolar device, a current having a density of 100 times or more flows as compared with the unipolar device. Therefore, it is necessary to make the MOS interface as low as possible. Here is the maximum meaning of introducing nitrogen (maximum meaning of introducing oxynitride film). Similar to the first embodiment, the dielectric breakdown resistance at the JFET region interface of the vertical IGBT is improved. Since the vertical IGBT is desired to have a higher dielectric breakdown resistance than the vertical MOSFET, the improvement of the dielectric breakdown resistance at the interface of the JFET region is very effective.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the scope of the invention. For example, a component in one embodiment may be replaced or changed with a component in another embodiment. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

12 SiC substrate 14 n-type SiC layer (n ^- SiC layer)
16 First SiC region (p-well region)
18 Second SiC region (source region)
20 Third SiC region (p-well contact region)
24 First electrode (source / p-well common electrode)
28 Gate insulating film 28a Interface doped region 28b Undoped region 30 Gate electrode 32 Interlayer insulating film 36 Second electrode (drain electrode)
100 MOSFET
200 MOSFET

Claims (21)

A SiC layer;
A gate insulating film provided on the surface of the SiC layer, comprising an oxide film or oxynitride film in contact with the surface of the SiC layer, wherein the oxide film or oxynitride film is at least selected from Mg, Ca, Sr, Ba One element, the peak of the concentration of the element in the gate insulating film is on the SiC layer side of the gate insulating film, the peak of the concentration of the element is in the oxide film or oxynitride film, The gate insulating film having a region where the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the opposite side of the SiC layer from the peak of the concentration of the element;
A gate electrode provided on the gate insulating film;
With
The gate insulating film contains C (carbon), the peak of C (carbon) concentration in the gate insulating film is on the SiC layer side of the gate insulating film, and the peak of C (carbon) concentration is A semiconductor device in the oxide film or oxynitride film.   2. The semiconductor device according to claim 1, wherein a peak of the C (carbon) concentration is in a range of 5 nm or less from an interface between the SiC layer and the gate insulating film.   3. The semiconductor device according to claim 2, wherein the concentration of the element at the peak of the element concentration is 80% to 120% of the concentration of C (carbon) at the peak of the C (carbon) concentration. .   4. The semiconductor device according to claim 3, wherein the distribution of the element in the film thickness direction and the distribution of the C (carbon) in the film thickness direction are in the range of 80% to 120%. A SiC layer;
A gate insulating film provided on the surface of the SiC layer, comprising an oxide film or oxynitride film in contact with the surface of the SiC layer, wherein the oxide film or oxynitride film is at least selected from Mg, Ca, Sr, Ba One element, the peak of the concentration of the element in the gate insulating film is on the SiC layer side of the gate insulating film, the peak of the concentration of the element is in the oxide film or oxynitride film, The gate insulating film having a region where the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the opposite side of the SiC layer from the peak of the concentration of the element;
A gate electrode provided on the gate insulating film;
With
A concentration of the element at a peak of the element concentration is 5.6 × 10 ¹⁷ cm ⁻³ or more and 5 × 10 ²⁰ cm ⁻³ or less. A SiC layer;
A gate insulating film provided on the surface of the SiC layer, comprising an oxide film or oxynitride film in contact with the surface of the SiC layer, wherein the oxide film or oxynitride film is at least selected from Mg, Ca, Sr, Ba One element, the peak of the concentration of the element in the gate insulating film is on the SiC layer side of the gate insulating film, the peak of the concentration of the element is in the oxide film or oxynitride film, The gate insulating film having a region where the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the opposite side of the SiC layer from the peak of the concentration of the element;
A gate electrode provided on the gate insulating film;
With
A semiconductor device, wherein a fixed dipole having a positive side on the SiC layer side and a negative side on the gate insulating film side exists at an interface between the SiC layer and the gate insulating film. A SiC layer;
A gate insulating film provided on the surface of the SiC layer, comprising an oxide film or oxynitride film in contact with the surface of the SiC layer, wherein the oxide film or oxynitride film is at least selected from Mg, Ca, Sr, Ba One element, the peak of the concentration of the element in the gate insulating film is on the SiC layer side of the gate insulating film, the peak of the concentration of the element is in the oxide film or oxynitride film, The gate insulating film having a region where the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the opposite side of the SiC layer from the peak of the concentration of the element;
A gate electrode provided on the gate insulating film;
With
A semiconductor device, wherein the oxide film or oxynitride film is a silicon oxide film or a silicon oxynitride film. The gate insulating film contains C (carbon), the peak of C (carbon) concentration in the gate insulating film is on the SiC layer side of the gate insulating film, and the peak of C (carbon) concentration is In the oxide film or oxynitride film,
The concentration of the element at the peak of the element concentration is 80% or more and 120% or less of the concentration of C (carbon) at the peak of the concentration of C (carbon),
The semiconductor device according to claim 7, wherein the concentration of the element at the peak of the element concentration is 5 × 10 ¹⁸ cm ⁻³ or more and 5 × 10 ²⁰ cm ⁻³ or less.   9. The semiconductor device according to claim 1, wherein a peak of the concentration of the element is in a range of 5 nm or less from an interface between the SiC layer and the gate insulating film.   10. The gate insulating film according to claim 1, wherein the gate insulating film is a stacked film of the oxide film or oxynitride film and a film having a lower concentration of the element than the oxide film or oxynitride film. A semiconductor device according to claim 1. 11. The semiconductor device according to claim 10, wherein a concentration of the element in the film having a lower concentration of the element than the oxide film or the oxynitride film is 1 × 10 ¹⁶ cm ⁻³ or less. A SiC substrate having first and second surfaces;
A first conductivity type SiC layer provided on the first surface side of the SiC substrate;
A first SiC region of a second conductivity type provided on the surface of the SiC layer;
A second SiC region of a first conductivity type provided on a surface of the first SiC region;
A gate insulating film continuously provided on a surface of the SiC layer and the first SiC region, and including a silicon oxide film or a silicon oxynitride film in contact with the surface of the first SiC region; The film or the silicon oxynitride film contains at least one element selected from Mg, Ca, Sr, and Ba, and the peak of the concentration of the element in the gate insulating film is on the first SiC region side of the gate insulating film. The concentration peak of the element is in the silicon oxide film or the silicon oxynitride film, and the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the side opposite to the SiC layer of the concentration peak of the element The gate insulating film having a region;
A gate electrode formed on the gate insulating film;
A first electrode formed on the second SiC region;
A second electrode formed on the second surface side of the SiC substrate;
A semiconductor device comprising:   13. The semiconductor device according to claim 12, wherein a peak of the concentration of the element is in a range of 5 nm or less from an interface between the first SiC region and the gate insulating film. A SiC substrate having first and second surfaces;
A first conductivity type SiC layer provided on the first surface side of the SiC substrate;
A first SiC region of a second conductivity type provided on the surface of the SiC layer;
A second SiC region of a first conductivity type provided on a surface of the first SiC region;
A gate insulating film continuously provided on a surface of the SiC layer and the first SiC region, the oxide layer or the oxynitride film being in contact with the surface of the first SiC region; The nitride film contains at least one element selected from Mg, Ca, Sr, and Ba, the peak of the element in the gate insulating film is on the first SiC region side of the gate insulating film, and the concentration of the element The gate insulating film having a region in which the concentration of the element is 1 × 10 ¹⁶ cm ⁻³ or less on the opposite side of the SiC layer from the peak of the concentration of the element in the oxide film or the oxynitride film ,
A gate electrode formed on the gate insulating film;
A first electrode formed on the second SiC region;
A second electrode formed on the second surface side of the SiC substrate;
With
The gate insulating film contains C (carbon), and the peak of C (carbon) concentration in the gate insulating film is on the first SiC region side of the gate insulating film, and the concentration of C (carbon) The semiconductor device is characterized in that the peak of is in the oxide film or oxynitride film.   The semiconductor device according to claim 14, wherein a peak of the C (carbon) concentration is in a range of 5 nm or less from an interface between the first SiC region and the gate insulating film.   The semiconductor device according to claim 12, wherein the SiC substrate is of a first conductivity type. Forming a first insulating film made of a silicon oxide film or a silicon oxynitride film containing at least one element selected from Mg, Ca, Sr, and Ba on the SiC layer;
Forming a second insulating film having a lower concentration of the element than the first insulating film on the first insulating film;
Forming a gate electrode on the second insulating film;
The method for manufacturing a semiconductor device, wherein the concentration of the element in the second insulating film is 1 × 10 ¹⁶ cm ⁻³ or less . Forming a first insulating film made of a silicon oxide film or a silicon oxynitride film containing at least one element selected from Mg, Ca, Sr, and Ba on the SiC layer;
Forming a second insulating film having a lower concentration of the element than the first insulating film on the first insulating film;
Forming a gate electrode on the second insulating film;
The formation of the first insulating film, forming a silicon film containing the element, the manufacturing method of the silicon film semiconductors devices you characterized oxide or a oxynitride to Turkey the. The first insulating film is formed by forming a silicon film containing the element, oxidizing or oxynitriding the silicon film, and thinning the oxidized or oxynitrided silicon film by etching. 19. A method of manufacturing a semiconductor device according to claim 17 or claim 18 . Forming a first insulating film made of an oxide film or an oxynitride film containing at least one element selected from Mg, Ca, Sr, and Ba on the SiC layer;
Forming a second insulating film having a lower concentration of the element than the first insulating film on the first insulating film;
Forming a gate electrode on the second insulating film;
The second insulating film, the concentration of C (carbon) than the first insulating film is rather low,
Method of manufacturing a semi-conductor device, wherein the concentration of said element in said second insulating film is 1 × 10 ¹⁶ cm ^-3 or less. The first insulating film is formed by forming a silicon film containing the element, oxidizing or oxynitriding the silicon film, and thinning the oxidized or oxynitrided silicon film by etching. the method according to claim 20, wherein you said.

Images