INTEGRATED ELECTRONIC DEVICE AND METHOD FOR MANUFACTURING THEREOF

[0001] This application is related to the U.S. patent application Ser. No.: ______ entitled VERTICAL-CONDUCTION INTEGRATED ELECTRONIC DEVICE AND METHOD FOR MANUFACTURING THEREOF, (Attorney Docket No.: 2110-505-03) filed ______, and application Ser. No.: ______ entitled INTEGRATED ELECTRONIC DEVICE WITH EDGE-TERMINATION STRUCTURE AND MANUFACTURING METHOD THEREOF, (Attorney Docket No.: 2110-507-03) filed ______ and which are incorporated herein by reference in their entireties.

[0002] The instant application claims priority to Italian Patent Application No. TO2010A000722, filed Aug. 30, 2010, which application is incorporated herein by reference in its entirety.

[0003] An embodiment relates to an integrated electronic device and to the method for manufacturing thereof.

[0004] As is known, there are today available numerous electronic devices made at least in part of silicon carbide (SiC).

[0005] For example, there are today available metal-oxide semiconductor field-effect transistors (MOSFETs) made at least in part of silicon carbide, which is characterized by a bandgap that is wider than the bandgap of silicon. Consequently, given the same doping level, the critical electrical field of silicon carbide is higher than the critical electrical field of silicon. For example, in the case where silicon carbide has a so-called 3C polytype, the critical electrical field is approximately equal to four times the critical electrical field of silicon; in the cases where silicon carbide has instead a 6H or a 4H polytype, the critical electrical field is, respectively, approximately eight times and ten times the critical electrical field of silicon.

[0006] Thanks to its high critical electrical field, silicon carbide enables provision of junctions having breakdown voltages higher than what may be obtained using silicon. Furthermore, exploiting the high critical electrical field, it may be possible to provide transistors with drift regions having thicknesses smaller than the drift regions of traditional silicon transistors; said transistors are hence characterized by low on-resistances (R_on).

[0007] On the other hand, silicon carbide has a low diffusiveness of the dopant species, even at high temperatures. In addition, as compared to silicon, silicon carbide is characterized by a reduced mobility μ of the carriers. In fact, in silicon carbide, the mobility μ of the carriers is typically of the order of some hundreds of cm²/Vs, whereas, in silicon, the mobility μ of the carriers can exceed even thousands of cm²/Vs. In particular, in the case of MOSFETs made of 4H-polytype silicon carbide, the mobility μ of the carriers in the respective channel regions is limited to approximately 50 cm²/Vs, on account of the generation of states at the oxide-semiconductor interfaces.

[0008] In greater detail, there are today available electronic devices formed starting from a silicon-carbide substrate. However, the technology today available does not enable provision of silicon-carbide wafers with diameters larger than four inches; consequently, the manufacture of said electronic devices is generally more costly and technologically complex than the manufacture of electronic devices starting from silicon substrates.

[0009] In order to combine the advantages of silicon and silicon carbide, electronic devices have moreover been proposed formed starting from a silicon substrate and comprising one or more silicon-carbide epitaxial layers. For example, U.S. Pat. No. 5,877,515, which is incorporated by reference, describes a semiconductor device, and in particular a MOSFET, having a silicon layer, which is deposited on a silicon-carbide layer, which in turn is deposited on a silicon substrate.

[0010] Operatively, the silicon-carbide layer enables a concentration of charge to be obtained that is higher than what may be obtained in the case of a silicon layer, given the same breakdown voltage. However, it may be possible that in certain conditions, and in particular in the case where the semiconductor device is biased so as to work in the region of inhibition, a non-negligible electrical field is generated within the silicon substrate. In said conditions, it is the silicon itself that limits, with its own critical electrical field, the breakdown voltage of the semiconductor device.

[0011] In order to prevent generation of a non-negligible electrical field within the silicon substrate, it may possible to increase the thickness of the silicon-carbide layer; however, said operation, in addition to being technologically complex, entails an increase of the on-resistance of the semiconductor device.

[0012] An embodiment is an integrated electronic device and a manufacturing method that enable drawbacks of the known art to be overcome at least in part.

[0021] FIG. 1 shows an embodiment of the present integrated electronic device, which forms in the case in point a trench MOSFET 1.

[0022] In detail, the trench MOSFET 1 is formed by a body 2 of semiconductor material, which is made up of a silicon substrate 4, of an N++ type (for example, doped with arsenic) and has a bottom surface S₄; the silicon may have a crystallographic structure of the so-called <100> type. Furthermore, the body 2 includes a first buffer layer 6, a second buffer layer 8, and a drift layer 10.

[0023] In detail, the substrate 4 has a thickness h₄ of approximately between 500 μm and 1500 μm, and moreover has a doping level higher than approximately 1·10¹⁹ cm⁻³, in such a way that the resistivity ρ is lower than approximately 6 mΩ·cm.

[0024] The first buffer layer 6 is set on top of the substrate 4, with which it is in direct contact, is made of silicon carbide and is of an N++ type (for example, doped with nitrogen or phosphorus). In particular, the first buffer layer 6 is made of silicon carbide of the 3C polytype, which, when grown on silicon of the <100> type, has a crystalline lattice with a low concentration of imperfections. In addition, the first buffer layer 6 has a thickness h₆ lower than approximately 0.5 μm, and moreover has a doping level of approximately between 5·10¹⁶ cm⁻³ and 1·10¹⁷ cm⁻³.

[0025] The second buffer layer 8 is arranged on top of the first buffer layer 6, with which it is in direct contact, is made of silicon carbide and is of an N+ type (for example, doped with nitrogen or phosphorus). In particular, also the second buffer layer 8 is made of silicon carbide of the 3C polytype. In addition, the second buffer layer 8 has a thickness h₈ lower than approximately 0.5 μm, and moreover has a doping level of approximately between 1·10¹⁶ cm⁻³ and 5·10¹⁶ cm⁻³.

[0026] The drift layer 10 is arranged on top of the second buffer layer 8, with which it is in direct contact, is made of silicon carbide, and is of an N type (for example, doped with nitrogen or phosphorus). In particular, also the drift layer 10 is made of silicon carbide of the 3C polytype. In addition, the drift layer 10 has a thickness h₁₀ of approximately between 1 μm and 10 μm, and moreover has a doping level of approximately between 1·10¹⁵ cm⁻³ and 1·10¹⁶ cm⁻³. Furthermore, the drift layer 10 defines an intermediate surface S₁₀.

[0027] The body 2 of the trench MOSFET 1 further includes a top layer 12, which is made of silicon and is of an N type (for example, doped with phosphorus).

[0028] In detail, the top layer 12 defines a top surface S₁₂ and extends over the intermediate surface S₁₀, in direct contact with the drift layer 10. Furthermore, the top layer 12 has a thickness h₁₂ in the range approximately between 1 μm and 3 μm; again, the top layer 12 has a doping level of approximately between 1·10¹⁵ cm⁻³ and 1·10¹⁶ cm⁻³.

[0029] In practice, the body 2 of semiconductor material is delimited by the top surface S₁₂ and by the bottom surface S₄. Furthermore, extending underneath the intermediate surface S₁₀ are a first semiconductor region 14a and a second semiconductor region 14b, both of a P+ type.

[0030] In detail, the first and second semiconductor regions 14a, 14b are arranged at a distance apart laterally so as to define an internal region 16, arranged between the first and second semiconductor regions 14a, 14b. Furthermore, the first and second semiconductor regions 14a, 14b extend from the intermediate surface S₁₀, with a depth h₁₄ smaller than the thickness h₁₀, for example, smaller than approximately 1 μm. Once again, the first and second semiconductor regions 14a, 14b have a doping peak level of approximately between 1·10¹⁷ cm⁻³ and 1·10¹⁹ cm⁻³.

[0031] Operatively, the first and second semiconductor regions 14a, 14b function, respectively, as first and second deep body regions 14a, 14b.

[0032] The trench MOSFET 1 further includes a trench 20, which extends from the top surface S₁₂ and has a thickness h₂₀. As illustrated by way of example in FIG. 1, the thickness h₂₀ may be such that h₂₀≧h₁₂. In other words, the trench 20 may extend through the top layer 12 until it partially penetrates within the drift layer 10; for example, it may be possible that h₂₀=h₁₂+Δ, with Δ approximately between 0.5 μm and 1 μm. Furthermore, the trench 20 is arranged at a distance apart laterally with respect to the first and second deep body regions 14a, 14b, so as to extend within the internal region 16 without contacting the first and second deep body regions 14a, 14b, which are approximately equidistant from the trench 20 itself.

[0033] In detail, the trench 20 is delimited by a wall 22, and the first and second deep body regions 14a, 14b are both at a distance of approximately d from the wall 22. In practice, if we designate by w₂₀ the width of the trench 20 and by w₁₆ the width of the internal region 16, we have w₁₆≈w₂₀+2d.

[0034] In greater detail, the wall 22 is coated internally with a first oxide layer 24. Furthermore, present within the trench 20 is a first gate region 26, which is in direct contact with the first oxide layer 24 and is made of polysilicon.

[0035] The trench MOSFET 1 further includes a first oxide region 30, which is deposited on the top surface S₁₂, is vertically aligned with the trench 20, and is in direct contact with the first oxide layer 24 and with the first gate region 26. The first oxide region may have a thickness, for example, approximately between 20 nm and 80 nm.

[0036] The trench MOSFET 1 further includes a first top body region 32a and a second top body region 32b, both of a P type (for example, doped with boron) and with a peak doping level of approximately between 1·10¹⁶ cm⁻³ and 1·10¹⁸ cm⁻³. In detail, the first and second top body regions 32a, 32b are arranged laterally at a distance apart and extend through the top layer 12 starting from the top surface S₁₂, until they contact, respectively, the first and second deep body regions 14a, 14b. Furthermore, the first and second top body regions 32a, 32b extend laterally until they contact the wall 22 of the trench 20, which separates the first and second top body regions 32a, 32b.

[0037] The trench MOSFET 1 further includes a first source region 34 formed by a first source subregion 34a and a second source subregion 34b, both of an N+ type (for example, doped with phosphorus) and with a peak doping level of approximately 1·10²⁰ cm⁻³. In detail, the first and second source subregions 34a, 34b extend, starting from the top surface S₁₂, within the top layer 12. In greater detail, the first and second source subregions 34a, 34b are arranged laterally at a distance apart, in such a way that the trench 20 extends between the first and second source subregions 34a, 34b, with which it is in direct contact. In particular, the first and second source subregions 34a, 34b are in direct contact with the first oxide layer 24; moreover, the first and second source subregions 34a, 34b are in direct contact with the first oxide region 30.

[0038] In practice, the first source region 34 and the trench 20 are housed within a operative region O, a bottom portion of which is formed by the internal region 16 and is hence delimited by the first and second deep body regions 14a, 14b.

[0039] The trench MOSFET 1 further includes a first dielectric region 35, which overlies, in direct contact, the first oxide region 30, and is in direct contact with the first and second source subregions 34a, 34b.

[0040] The trench MOSFET 1 further includes a first top metallization 36 and a bottom metallization 38, as well as a first gate metallization, the latter (not shown) contacting, in a way in itself known, the first gate region 26.

[0041] In detail, the first top metallization 36 extends over the top surface S₁₂, surrounding at the top and laterally the first dielectric region 35, and contacting the first and second source subregions 34a, 34b, as well as the first and second top body regions 32a, 32b.

[0042] The bottom metallization 38 extends, instead, underneath the bottom surface S₄ of the substrate 4, with which it is in direct contact. In practice, the substrate 4, the first and second buffer layers 6, 8, and the drift layer 10 function as a drain region.

[0043] Operatively, the first top metallization 36 functions as a source metallization, whilst the bottom metallization 38 functions as a drain metallization. Furthermore, the first gate region 26, the first oxide layer 24, and the top layer 12 (and in particular, the first and second top body regions 32a, 32b) form a junction of the metal-oxide-semiconductor type. Hence, by biasing in a way in itself known the first top metallization 36 and the first gate metallization, it is possible to form a channel of an N type within the top layer 12, in particular within a region of the top layer 12 arranged in direct contact with the first oxide layer 24. Furthermore, by biasing in a way in itself known the first top metallization 36 and the bottom metallization 38 with a voltage V_DS, it is possible to generate a current I_DS.

[0044] The current I_DS flows between the first top metallization 36 and the bottom metallization 38; hence, it has a vertical direction and flows both through the silicon and through the silicon carbide. In particular, the current I_DS flows within the channel of an N type, traversing the top layer 12, as well as the drift layer 10 and the first and second buffer layers 6, 8.

[0045] In the case where the voltage V_DS is such as to reversely bias the PN junctions present between the drift layer 10 and the first and second deep body regions 14a, 14b, the magnitude of the voltage V_DS cannot exceed a maximum voltage V_max; otherwise, a breakdown phenomenon within the trench MOSFET 1 would be triggered.

[0046] In particular, the maximum voltage V_max is particularly high thanks to the presence of the first and second buffer layers 6, 8, and to the respective doping levels, which prevent the generation of significant electrical fields within the silicon substrate 4.

[0047] Likewise, the first and second deep body regions 14a, 14b prevent generation of significant electrical fields within the top silicon layer 12. In other words, the voltage V_DS drops substantially within the first and second deep body regions 14a, 14b, as well as within the drift layer 10 and the first and second buffer layers 6, 8. Instead, within the top layer 12 and the substrate 4, there is little to no voltage drop, because the electrical field is, to a first approximation, negligible. Consequently, the maximum voltage V_max is limited at the top, instead of by the critical electrical field of silicon, by the critical electrical field of silicon carbide, which, as has been said, is higher than the critical electrical field of silicon.

[0048] Furthermore, with regard to the top layer 12, the electrical field present therein is negligible not only in the proximity of the first and second deep body regions 14a, 14b, but also in the proximity of the internal region 16, i.e., where the top layer 12 is not in direct contact with the first and second deep body regions 14a, 14b. In fact, as illustrated qualitatively in FIG. 1, underneath the first and second deep body regions 14a, 14b, the equipotential lines L that are generated in use within the trench MOSFET 1 are approximately parallel to the first and second deep body regions 14a, 14b. Instead, in an area corresponding to the internal region 16, the equipotential lines L bend on account of the presence of the trench 20, and in particular of the first oxide layer 24.

[0049] FIG. 2 illustrates a different embodiment of the present integrated electronic device, which in the case in point forms a planar-cell MOSFET 50, described in what follows. Elements already present in the trench MOSFET 1 illustrated in FIG. 1 are designated by the same reference numbers; moreover, the ensuing description is limited to the differences of the planar-cell MOSFET 50 with respect to the trench MOSFET 1 illustrated in FIG. 1.

[0050] In detail, the planar-cell MOSFET 50 is without the trench 20, and consequently also without the first oxide layer 24 and the first gate region 26. Furthermore, arranged between the first oxide region 30 and the first dielectric region 35 is a second gate region 52, made of polysilicon, which overlies the first oxide region 30, with which it is in direct contact. In particular, the second gate region 52 has a thickness equal, for example, to approximately 600 nm.

[0051] In addition, the first and second top body regions 32a, 32b surround laterally, respectively, the first and second source subregions 34a, 34b. Furthermore, in a way similar to what is illustrated in FIG. 1, also the planar-cell MOSFET 50 includes the first top metallization 36, the bottom metallization 38, and the first gate metallization, the latter (not shown) contacting in a way in itself known the second gate region 52.

[0052] Operatively, the second gate region 52, the first oxide region 30, and the top layer 12 form a junction of the metal-oxide-semiconductor type; consequently, by biasing in a way in itself known the first source region 34 and the second gate region 52, it may be possible to form, underneath the first oxide region 30, a channel of an N type. In particular, the channel of an N type extends within portions of the first and of the second top body regions 32a, 32b in contact with the first oxide region 30 and between the first and second source subregions 34a, 34b.

[0053] Furthermore, by biasing in a way in itself known the first top metallization 36 and the bottom metallization 38 with the voltage V_DS, it may be possible to generate the current I_DS, which flows vertically, traversing the internal region 16, in a way similar to what has been described previously.

[0054] FIG. 3 illustrates a different embodiment of the present integrated electronic device, which in the case in point forms an enhanced trench MOSFET 60, described in what follows. Elements already present in the trench MOSFET 1 illustrated in FIG. 1 are designated by the same reference numbers; moreover, the ensuing description is limited to the differences of the enhanced trench MOSFET 60 with respect to the trench MOSFET 1 illustrated in FIG. 1.

[0055] In detail, the enhanced trench MOSFET 60 includes one or more intermediate layers of an N type (for example, doped with nitrogen or phosphorus), which are arranged between the drift layer 10 and the top layer 12, and are made of silicon carbide of the 3C polytype. Furthermore, the intermediate layers may have approximately one and the same thickness h₆₂ and one and the same doping level, for example equal to the doping level of the drift layer 10.

[0056] By way of example, the embodiment illustrated in FIG. 3 has a first intermediate layer and a second intermediate layer, designated by 62a and 62b, respectively. In particular, the first intermediate layer 62a overlies the drift layer 10, with which it is in direct contact. Instead, the second intermediate layer 62b overlies the first intermediate layer 62a, with which it is in direct contact, and is overlaid by the top layer 12, with which it is in direct contact. The second intermediate layer 62b defines moreover a contact surface S₆₂, which contacts the top layer 12. In addition, in said embodiment, the trench 20 may extend, for example, up to the contact surface S₆₂, i.e., with h₂₀=h₁₀ (or approximately equal to h₁₀).

[0057] The enhanced trench MOSFET 60 further includes a first intermediate body region 64 and a second intermediate body region 66, both of a P type (for example, doped with boron) and with peak doping levels of approximately between 1·10¹⁶ cm⁻³ and 1·10¹⁸ cm⁻³. In detail, the first and second intermediate body regions 64, 66 are arranged laterally at a distance apart and extend through the first and second intermediate layers 62a, 62b, starting from the contact surface S₆₂, until they contact, respectively, the first and second deep body regions 14a, 14b. Furthermore, the first and second intermediate body regions 64, 66 are, respectively, in contact with the first and second top body regions 32a, 32b.

[0058] Operatively, defining as a confinement region 68 the region formed by the portions of the first and second intermediate layers 62a, 62b not occupied either by the first intermediate body region 64 or by the second intermediate body region 66, it is possible to verify that, in use, the electrical field assumes, in this confinement region 68, a direction to a first approximation horizontal, i.e., parallel to the intermediate surface S₁₀. In this way, the possibility of formation within the top layer 12 of an appreciable electrical field is further reduced.

[0059] As illustrated in FIG. 4, a different embodiment of the present integrated electronic device is possible, which forms a modified planar-cell MOSFET 70, which in what follows will be referred to as “enhanced planar-cell MOSFET 70”, described in what follows. Elements already present in the planar-cell MOSFET 50 illustrated in FIG. 2 are designated by the same reference numbers.

[0060] In detail, with respect to the planar-cell MOSFET 50, the enhanced planar-cell MOSFET 70 includes the already described one or more intermediate layers, as well as the first and second intermediate body regions 64, 66. In particular, the embodiment illustrated in FIG. 4 has the already mentioned first and second intermediate layers 62a, 62b, which enable the aforementioned reduction of the electrical field to be obtained within the top layer 12.

[0061] According to further embodiments of the present integrated electronic device, it may form, moreover, an insulated-gate bipolar junction transistor (IGBT), and in particular a trench IGBT 80 (FIG. 5), or else a planar-cell IGBT 90 (FIG. 6).

[0062] As illustrated in FIG. 5, where elements already illustrated in FIG. 1 are designated by the same reference numbers, the trench IGBT 80 differs from the trench MOSFET 1 in that it is formed starting from an alternative substrate 4b, instead of from the substrate 4 described previously.

[0063] In particular, the alternative substrate 4b is made of silicon with crystallographic structure of the <100> type, and is of a P++ type (for example, doped with boron); moreover, the alternative substrate 4b has a thickness h_4b of approximately between 500 μm and 1500 μm, and has a doping level higher than approximately 1·10¹⁹ cm⁻³, in such a way that the resistivity ρ is lower than approximately 6 mΩ·cm.

[0064] Operatively, the first and second deep body regions 14a, 14b function, together with the first and the second top body regions 32a, 32b, as a collector of an additional bipolar junction transistor of the PNP type, the base of which is formed by the first and second buffer layers 6, 8 and by the drift layer 10, and the emitter of which is formed by the alternative substrate 4b.

[0065] Likewise, as illustrated in FIG. 6, where elements already illustrated in FIG. 2 are designated by the same reference numbers, the planar-cell IGBT 90 differs from the planar-cell MOSFET 50 in that it is formed starting from the alternative substrate 4b, instead of from the substrate 4.

[0066] According to yet a further embodiment, the present integrated electronic device may moreover form a junction field-effect transistor 100 (JFET), illustrated in FIG. 7, where elements already illustrated in FIG. 1 are designated by the same reference numbers.

[0067] In detail, the JFET 100 includes the substrate 4, the first and second buffer layers 6, 8, and the drift layer 10, as well as the first and second semiconductor regions 14a, 14b, which function, respectively, as first and second gate subregions 14a, 14b. Instead, the JFET 100 is without the top layer 12.

[0068] The JFET 100 further includes a second source region 102, made of silicon and of an N++ type (for example doped with phosphorus). In particular, the second source region 102 extends over the intermediate surface S₁₀, in direct contact with the internal region 16, and has a doping level approximately equal to 1·10¹⁹ cm⁻³ and a thickness h₁₀₂ lower than approximately 2 μm.

[0069] The JFET 100 further includes, instead of the first top metallization 36, a second top metallization 104, which overlies the second source region 102, with which it is in direct contact, and performs the function of source metallization. Furthermore, the JFET 100 includes a second dielectric region 106 (formed, for example, of borophosphosilicate glass, BPSG), which extends over the intermediate surface S₁₀ and surrounds laterally the second source region 102 and the second top metallization 104, with which it is in direct contact. Once again, in addition to the bottom metallization 38, the JFET 100 includes a second gate metallization 108, which extends over the intermediate surface S₁₀ and surrounds the second dielectric region 106, with which it is in direct contact. Furthermore, the second gate metallization 108 contacts the first and second gate subregions 14a, 14b.

[0070] In use, in the absence of voltage between the second top metallization 104 and the second gate metallization 108, at the interface between the drift layer 10 and the first gate subregion 14a a first depleted region D_a is formed, and at the interface between the drift layer 10 and the second body subregion 14b a second depleted region D_b is formed. These first and second depleted regions D_a, D_b extend within the drift layer 10 with a depth of approximately w_db. Furthermore, in this embodiment, the width w₁₆ of the internal region 16, i.e., the lateral distance between the first and second gate subregions 14a, 14b, and the doping levels of the drift layer 10 and the first and the second gate subregions 14a, 14b are such that w_db≈w₁₆/2. In other words, in the absence of voltage between the second top metallization 104 and the second gate metallization 108, the first and second depleted regions D_a, D_b contact one another, insulating the second source region 102 from the substrate 4 and from the first and second buffer layers 6, 8, which function as drain regions. Consequently, the JFET 100 is a vertical-current-conduction device of the normally off type.

[0071] According to a further embodiment, the present integrated electronic device may form a bipolar junction transistor 120, illustrated in FIG. 8, where elements already illustrated in FIG. 1 are designated by the same reference numbers.

[0072] In detail, the bipolar junction transistor 120 includes the substrate 4, the first and second buffer layers 6, 8, the drift layer 10, and the top layer 12 described previously. Furthermore, the bipolar junction transistor 120 includes an emitter region 122, of an N+ type (for example, doped with phosphorus), which extends, starting from the top surface S₁₂, within the top layer 12, with a doping level approximately equal to 1·10²⁰ cm⁻³.

[0073] In addition, the bipolar junction transistor 120 includes, instead of the first and second semiconductor regions 14a, 14b, a base region 124 of a P type (for example, doped with boron), and moreover includes an edge region 126 of a P-type (for example, doped with boron).

[0074] In detail, the base region 124 has a peak doping level of approximately between 1·10¹⁶ cm⁻³ and 1·10¹⁸ cm⁻³; moreover, the base region 124 faces the top surface S₁₂ and extends within the top layer 12 so as to surround laterally and at the bottom the emitter region 122, to which it is vertically aligned. In particular, the base region 124 traverses the entire top layer 12 and extends partially within the drift layer 10, underneath the intermediate surface S₁₀.

[0075] The edge region 126 has a peak doping level of approximately between 1·10¹⁴ cm⁻³ and 1·10¹⁶ cm⁻³; moreover, the edge region 126 faces the top surface S₁₂ and extends within the top layer 12, externally with respect to the base region 124, with which it is in direct contact. In particular, the edge region 126 traverses the entire top layer 12 and extends partially within the drift layer 10, underneath the intermediate surface S₁₀, laterally surrounding the base region 124.

[0076] The bipolar junction transistor 120 further includes a channel-stopper region 128 of an N+ type (for example, doped with phosphorus) and with a doping level approximately equal to 1·10²⁰ cm⁻³, which extends in the top layer 12, externally and at a distance from the edge region 126.

[0077] The bipolar junction transistor 120 further includes an emitter metallization 130, a base metallization 132, and an equipotential ring 134 (EQR), made of metal material, in addition to the bottom metallization 38, which functions as collector metallization and is coupled, in a way in itself known, to the equipotential ring 134.

[0078] In detail, the emitter metallization 130, the base metallization 132, and the equipotential ring 134 extend over the top surface S₁₂ and are in direct contact, respectively, with the emitter region 122, the base region 124, and the channel-stopper region 128.

[0079] The bipolar junction transistor 120 finally includes an insulating region 140, made of dielectric material, deposited on the top surface S₁₂, with which it is in direct contact, but for the contacts between the emitter metallization 130, the base metallization 132, the equipotential ring 134, and, respectively, the emitter region 122, the base region 124 and the channel-stopper region 128.

[0080] In use, the drift layer 10, part of the top layer 12, and the first and second buffer layers 6, 8 function as collector region of the bipolar junction transistor 120. Furthermore, the edge region 126 prevents concentration of the equipotential lines, and hence generation of a high electrical field, in the proximity of the PN junctions, which would be formed, in the absence of said edge region 126, between the base region 124 and the top layer 12 and the drift layer 10. In this way, a reduction of the maximum voltage V_max at which the bipolar junction transistor 120 can operate without incurring in the breakdown phenomenon is avoided. As explained previously, also in the case of the bipolar junction transistor 120, said maximum voltage V_max depends to a first approximation upon the critical electrical field of silicon carbide, thanks to the fact that the base region 124 extends also underneath the intermediate surface S₁₀.

[0081] As regards, instead, the channel-stopper region 128, it performs, in a way in itself known, the function of preventing formation of channels between the base region 124 and the outermost portions of the bipolar junction transistor 120.

[0082] An embodiment of he present electronic device may be obtained using the manufacturing method described in what follows and represented in FIGS. 9-24. In particular, the ensuing description regards, without this implying any loss of generality, manufacture of the trench MOSFET 1, of the enhanced trench MOSFET 60, of the JFET 100, and of the bipolar junction transistor 120; however, an embodiment of the manufacturing technique may be used to manufacture the devices 50, 70, 80, and 90 as well.

[0083] As illustrated in FIG. 9, to obtain the trench MOSFET 1 the substrate 4 is provided, and subsequently the first buffer layer 6, the second buffer layer 8, and the drift layer 10 are formed by epitaxial growth.

[0084] Next (FIG. 10), using a first resist mask 200, a first implantation of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 202, so as to localize the dopant species in a first thin layer 14a′ and a second thin layer 14b′, both of a P+ type, which are arranged underneath the intermediate surface S₁₀ and are to form, respectively, the first and second deep body regions 14a, 14b, once appropriate annealing processes are terminated, as explained hereinafter. In particular, the first implant is executed with a hot process, i.e., at a temperature of approximately between 500° C. and 850° C., in order to limit the defects introduced during the implant itself within the silicon-carbide crystalline lattice; moreover, the first implant may be made with a dosage in a range approximately between 1·10¹³ cm⁻² and 1·10¹⁵ cm⁻² and an energy in the range approximately between 30 keV and 2 MeV. Furthermore, in a way in itself known, and hence not illustrated, the first implant may be made after providing, on the intermediate surface S₁₀, a temporary silicon layer, which is removed at the end of the first implant; in this way, the lattice defects introduced during the first implant within the silicon-carbide crystalline lattice are further limited.

[0085] Next (FIG. 11), the first resist mask 200 is removed, and an annealing at a temperature of approximately between 1250° C. and 1365° C., and of the duration, for example, of approximately thirty minutes, is performed in order to reduce the sites of the drift layer 10 damaged following upon the first implant, and activate the dopant species. During said annealing, the thickness of the first and second thin layers 14a′, 14b′ increases, and the first and second thin layers 14a′, 14b′ form, respectively, the first and second deep body regions 14a, 14b.

[0086] Next (FIG. 12), the top layer 12, made of silicon, is formed, by means of hetero-epitaxy.

[0087] In a way in itself known, and consequently not illustrated, by means of chemical etching, the trench 20 is formed, and subsequently formed in succession, once again in a way in itself known, are the first oxide layer 24, the first gate region 26, the first and second top body regions 32a, 32b, and the source region 34.

[0088] Then formed, in a way in itself known, are the first oxide region 30 and the first dielectric region 35, and finally the first top metallization 36 and the bottom metallization 38.

[0089] As regards the enhanced trench MOSFET 60, to produce one may perform the operations described in what follows, which refer, by way of example, to the embodiment illustrated in FIG. 3.

[0090] In detail, the operations illustrated in FIGS. 9-10 are performed. Next (FIG. 13), hence after performing the first implant, the first resist mask 200 is removed and, by means of epitaxial growth, the first intermediate layer 62a is formed.

[0091] Next, by means of a second resist mask 210, a second implant of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 212, so as to localize the dopant species in a third thin layer 64′ and a fourth thin layer 66′, both of a P+ type. In particular, if one designates by S_62a the top surface of the first intermediate layer 62a, the third and the fourth thin layers 64′ and 66′ are arranged underneath the surface S_62a of the first intermediate layer 62a, and are to form, respectively, the first and second intermediate body regions 64, 66, once appropriate annealing processes are terminated. Also this second implant is executed with a hot process, i.e., at a temperature of approximately between 500° C. and 850° C., and with a dosage comprised in the range approximately between 1·10¹³ cm⁻² and 1·10¹⁵ cm⁻² and an energy approximately between 30 keV and 2 MeV. Furthermore, in a way in itself known, and hence not illustrated, also the second implant may be made after providing, on the surface S_62a of the first intermediate layer 62a, a temporary silicon layer, which is removed at the end of the second implant.

[0092] Next (FIG. 14), the second resist mask 210 is removed and, by means of epitaxial growth, the second intermediate layer 62b is formed, and subsequently, by means of a third resist mask 220, a third implant of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 222, so as to localize the dopant species in a fifth thin layer 64″ and a sixth thin layer 66″, both of a P+ type. In particular, the fifth and sixth thin layers 64″ and 66″ are arranged underneath the contact surface S₆₂ and are to form, respectively, the first and second intermediate body regions 64, 66, once appropriate annealing processes are terminated. Also this third implant is executed with a hot process, i.e., at a temperature of approximately between 500° C. and 850° C., and with a dosage in the range approximately between 1·10¹³ cm⁻³ and 1·10¹⁵ cm⁻³ and an energy approximately between 30 keV and 2 MeV. Furthermore, in a way in itself known, and hence not illustrated, also the third implant may be made after providing, on the contact surface S₆₂, a temporary silicon layer, which is removed at the end of the third implant.

[0093] Next (FIG. 15), the third resist mask 220 is removed, and an annealing at a temperature of approximately between 1250° C. and 1365° C., and of the duration, for example, of thirty minutes is performed. In this way, the first and second thin layers 14a′, 14b′ form, respectively, the first and second deep body regions 14a, 14b. Likewise, the third and fifth thin layers 64′, 64″ increase in thickness and form the first intermediate body region 64; moreover, the fourth and sixth thin layers 66′, 66″ increase in thickness and form the second intermediate body region 66.

[0094] It may be noted that, even though FIGS. 13 and 14 show an embodiment in which the third and fourth thin layers 64′ and 66′ have a thickness h_l lower than the thickness h₆₂ of the first intermediate layer 62a, and in which the fifth and sixth thin layer 64″ and 66″ have a thickness h₁₁ lower than the thickness h₆₂ of the second intermediate layer 62b, it may be likewise possible to carry out the second and third implants in such a way that h_l≧h₆₂ and h_ll≧h₆₂.

[0095] Next, the top silicon layer 12 is formed, by means of hetero-epitaxy, and then, in a way in itself known, the first and second top body regions 32a, 32b, the trench 20, the first gate region 26, the first oxide layer 24, the first source region 34, the first oxide region 30, the first dielectric region 35, as well as the first top metallization 36, the bottom metallization 38, and the first gate metallization are provided.

[0096] To manufacture the JFET 100 it is, instead, possible to carry out the operations illustrated in FIGS. 9-11.

[0097] Next (FIG. 16), hence, after performing the annealing, an additional layer 232, which is made of silicon and is of an N++ type (for example, doped with phosphorus), and which is to form the second source region 102, is formed, by means of hetero-epitaxy.

[0098] Next (FIG. 17), by means of a chemical etch, portions of the additional layer 232 are selectively removed so as to form the second source region 102. Next, an insulating layer 236 is deposited, which is made, for example, of borophosphosilicate glass and is to form the second dielectric region 106.

[0099] In a way not illustrated, by means of chemical etching, portions of the insulating layer 236 are selectively removed so as to form the second dielectric region 106, and the second top metallization 104 and the second gate metallization 108 are provided.

[0100] Finally, to manufacture the bipolar junction transistor 120, it is possible to proceed as described in what follows.

[0101] In detail, the operations illustrated in FIG. 9 are performed. Next (FIG. 18) by means of a fourth resist mask 250, a fourth implant of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 252, so as to localize the dopant species in a seventh thin layer 126a′ of a P type, which is localized underneath the intermediate surface S₁₀ and is to form the base region 124. Also this fourth implant is executed with a hot process, i.e., at a temperature of approximately between 500° C. and 850° C., and with a dosage in the range approximately between 1·10¹⁴ cm⁻² and 1·10¹⁵ cm⁻² and an energy approximately between 30 keV and 2 MeV. Once again, also the fourth implant may be made after providing, on the intermediate surface S₁₀, a temporary silicon layer, which is removed at the end of the fourth implant.

[0102] Next (FIG. 19) the fourth resist mask 250 is removed, and by means of a fifth resist mask 260, a fifth implant of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 262, so as to localize the dopant species in an eighth thin layer 126a′ of a P-type, which is localized underneath the intermediate surface S₁₀ and is to form the edge region 126. Also this fifth implant is executed with a hot process, i.e., at a temperature of approximately between 500° C. and 850° C., and with a dosage in the range approximately between 1·10¹² cm⁻² and 1·10¹³ cm⁻² and an energy approximately between 30 keV and 2 MeV. Once again, also the fifth implant may be carried out after providing, on the intermediate surface S₁₀, a temporary silicon layer, which is removed at the end of the fifth implant.

[0103] Next (FIG. 20) the fifth resist mask 260 is removed, and an annealing, at a temperature of approximately between 1250° C. and 1365° C., and of the duration, for example, of approximately thirty minutes is performed. In this way, the seventh and eighth thin layers 124a′, 126a′ form, respectively, a first portion of base region 124a and a first portion of edge region 126a, which extend facing the intermediate surface S₁₀, the first portion of edge region 126a surrounding the first portion of base region 124a.

[0104] Next (FIG. 21), by means of epitaxial growth, the top layer 12 is formed, arranged above the intermediate surface S₁₀.

[0105] Next (FIG. 22), by means of a sixth resist mask 270, a sixth implant of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 272, so as to localize the dopant species in a ninth thin layer 124b′ of a P type, which is localized underneath the top surface S₁₂ and is to form the base region 124.

[0106] Next (FIG. 23), the sixth resist mask 270 is removed and, by means of a seventh resist mask 280, a seventh implant of dopant species of a P type (for example, aluminium or boron atoms) is performed, represented by the arrows 282, so as to localize the dopant species in a tenth thin layer 126b′ of a P-type, localized underneath the top surface S₁₂, which surrounds the ninth thin layer 124b′ and is to form the edge region 126.

[0107] Next (FIG. 24), the seventh resist mask 280 is removed and an annealing at a temperature of approximately between 1100° C. and 1150° C., and of the duration of approximately one hundred and twenty minutes is performed, in such a way that the ninth and tenth thin layers 124b′, 126b′ form, respectively, a second portion of base region 124b and a second portion of edge region 126b, extending facing the top surface S₁₂, the second portion of edge region 126b surrounding the second portion of base region 124b. Furthermore, the second portion of base region 124b and the second portion of edge region 126b traverse the entire thickness of the top layer 12, until they connect, respectively, to the first portion of base region 124a and to the first portion of edge region 126a so as to form the base region 124 and the edge region 126.

[0108] In a way in itself known, and consequently not illustrated, the emitter region 122, the channel-stopper region 128, the insulating region 140, the emitter metallization 130, the base metallization 132, the equipotential ring 134, and the bottom metallization 38 are then provided.

[0109] Advantages that embodiments of the present integrated electronic device and the present manufacturing method afford emerge clearly from the foregoing discussion.

[0110] In detail, an embodiment of the present electronic device is formed starting from a silicon substrate, with consequent benefits in terms of reduced costs and availability of wafers of large dimensions. In addition, the present integrated electronic device uses the properties of silicon carbide (wide bandgap) to confine the electrical field within the silicon-carbide epitaxial layers, preventing generation of the breakdown phenomenon within the silicon substrate. In particular, the use of the first and second buffer layers 6, 8 prevent occurrence of breakdown within the substrate 4. In addition, the doping profiles and the thicknesses h₆, h₈ of the first and second buffer layers 6, 8 may be modulated so that the electrical field that is generated therein has a desired profile.

[0111] Furthermore, according to an embodiment of the present manufacturing method, the first and second deep body regions 14a, 14b, as well as, in the case of the bipolar junction transistor 120, the first portion of base region 124a, are obtained by ion implantation, with consequent greater control of the corresponding thicknesses, which may be larger than what may be obtained by diffusion.

[0112] Finally, it is evident that modifications and variations may be made to the present integrated electronic device and manufacturing method, without thereby departing from the scope of the present disclosure.

[0113] For example, it may be possible for there to be present just one of the first buffer layer 6 and the second buffer layer 8. In addition, it may be possible to reverse all the conductivity types of the semiconductor elements described, and/or use different semiconductor materials.

[0114] Furthermore, an embodiment of the present electronic device may be disposed on a first integrated circuit, which may be coupled to one or more second integrated circuits to form a system. One or more of the first and second integrated circuits may include a controller such as a processor.

[0115] From the foregoing it will be appreciated that, although specific embodiments have been described herein for purposes of illustration, various modifications may be made without deviating from the spirit and scope of the disclosure. Furthermore, where an alternative is disclosed for a particular embodiment, this alternative may also apply to other embodiments even if not specifically stated.