HEMT POWER DEVICE WITH REDUCED GATE OSCILLATION AND MANUFACTURING PROCESS THEREOF

Abstract

A heterojunction power device includes: a substrate containing semiconductor material; a first active area and a second active area, arranged on the substrate symmetrically opposite with respect to an axis of symmetry and accommodating respective heterostructures; a separation region, extending along the axis of symmetry between the first active area and the second active area. The power device further includes: a first conductive bus configured to distribute a first electric potential of the power device in parallel to the first and the second active areas; a second conductive bus configured to distribute a second electric potential of the power device, different from the first electric potential, in parallel to the first and the second active areas. The first and the second conductive buses extend along the axis of symmetry above the separation region and the second conductive bus overlies the first conductive bus.

Claims

1. A heterojunction power device, comprising: a substrate including a semiconductor material; a first active area and a second active area, arranged on the substrate symmetrically opposite with respect to an axis of symmetry and accommodating respective heterostructures; a separation region, extending along the axis of symmetry between the first active area and the second active area; a first conductive bus configured to distribute a first electric potential of the power device in parallel to the first and the second active areas; and a second conductive bus configured to distribute a second electric potential of the power device, different from the first electric potential, in parallel to the first and the second active areas, wherein the first and the second conductive buses extend along the axis of symmetry above the separation region, the second conductive bus overlying the first conductive bus.

2. The power device according to claim 1, wherein the first conductive bus is a gate bus and the first electric potential is a control signal of the power device.

3. The power device according to claim 1, wherein the second conductive bus is a source bus and the second electric potential is a source electric potential of the power device.

4. The power device according to claim 1, comprising a first level of metallization, wherein the first conductive bus is a portion of the first level of metallization; the first level of metallization further including: first-level drain electrodes and first-level source electrodes associated with the first active area, extending transversely to the first conductive bus and superimposed on the first active area; and first-level drain electrodes and first-level source electrodes associated with the second active area, extending transversely to the first conductive bus, superimposed on the second active area and separated, at the separation region, from the first-level drain electrodes and from the first-level source electrodes associated with the first active area.

5. The power device according to claim 4, wherein the first-level drain electrodes and the first-level source electrodes associated with the first active area are arranged in a symmetrical manner, with respect to the axis of symmetry, to respective first-level drain electrodes and first-level source electrodes associated with the second active area; and wherein in the first active area and in the second active area the first-level drain electrodes are arranged alternating, along the axis of symmetry, with the first-level source electrodes.

6. The power device according to claim 5, comprising a second level of metallization superimposed on the first level of metallization, wherein the second conductive bus is a portion of the second level of metallization; the second level of metallization further including: second-level drain electrodes and second-level source electrodes associated with the first active area, extending transversely to the second conductive bus and superimposed on the first active area; second-level drain electrodes and second-level source electrodes associated with the second active area, extending transversely to the second conductive bus and superimposed on the second active area; the second-level source electrodes extending from the second conductive bus in opposite directions with respect to the axis of symmetry; and the second-level drain electrodes associated with the first active area being separated, at the separation region, from the second-level drain electrodes associated with the second active area.

7. The power device according to claim 6, wherein the second-level drain electrodes and the second-level source electrodes associated with the first active area are arranged, with respect to the axis of symmetry, in a manner symmetrical, to the second-level drain electrodes and to the second-level source electrodes associated with the second active area; and the second-level drain electrodes being arranged alternating, along the axis of symmetry, with the second-level source electrodes.

8. The power device according to claim 7, comprising a third level of metallization, superimposed on the second level of metallization, provided with: a drain metallization region, superimposed on the first active area; a source metallization region, superimposed on the second active area; third-level drain electrodes and third-level source electrodes, extending transversely to the axis of symmetry starting from the drain metallization region and from the source metallization region, respectively; and the third-level drain electrodes being arranged alternating, along the axis of symmetry, with the third-level source electrodes.

9. The power device according to claim 8, comprising second-level conductive vias between the first level of metallization and the second level of metallization and third-level conductive vias between the second level of metallization and the third level of metallization.

10. The power device according to claim 1, wherein the first and the second active areas each include: a respective channel layer and a respective barrier layer defining respective heterojunctions; a plurality of respective active elements, extending transversely to the axis of symmetry and arranged periodically along the axis of symmetry, wherein the active elements include respective portions of the channel layers and of the barrier layers and respective gate regions superimposed, in contact, on the barrier layer of the respective active area; the active elements each further including a drain terminal, a source terminal and a gate terminal; and the gate terminals including gate electrodes, superimposed, in contact, on respective gate regions and extending transversely to the axis of symmetry.

11. The power device according to claim 10, when dependent on claim 8, wherein: the third-level drain electrodes, the second-level drain electrodes, the first-level drain electrodes and the drain terminals of the active elements are superimposed on each other and all electrically connected to each other in a direct manner; the third-level source electrodes, the second-level source electrodes, the second conductive bus, the first-level source electrodes and the source terminals of the active elements are superimposed on each other and all electrically connected to each other in a direct manner; and the first-level drain electrodes and the first-level source electrodes have respective contact portions which extend transverse to the axis of symmetry and form the drain terminals and the source terminals.

12. The power device according to claim 10, wherein the first conductive bus, the gate electrodes and the gate terminals are all electrically connected to each other in a direct manner.

13. The power device according to claim 10, wherein the active elements are HEMT-type transistors, electrically connected to each other in parallel and formed in gallium nitride technology.

14. A process for manufacturing a heterojunction power device, comprising: on a substrate including semiconductor material, forming a first active area and a second active area symmetrically opposite with respect to an axis of symmetry, accommodating respective heterostructures and separated by a separation region, extending along the axis of symmetry; forming, on the separation region, a first conductive bus configured to distribute a first electric potential of the power device in parallel to the first and the second active areas; and forming, on the separation region, a second conductive bus configured to distribute a second electric potential of the power device, different from the first electric potential, in parallel to the first and the second active areas, wherein the first and the second conductive buses extend along the axis of symmetry, the second conductive bus overlying the first conductive bus.

15. The manufacturing process according to claim 14, comprising: forming a first level of metallization; and from the first level of metallization forming: the first conductive bus; first-level drain and source electrodes, extending transversely to the first conductive bus, configured to provide, respectively, drain electric potentials and source electric potentials of the power device to the first active area and to the second active area; forming a second level of metallization above first level of metallization; and from the second level of metallization forming: the second conductive bus; second-level source electrodes, extending from the second conductive bus in opposite directions with respect to the axis of symmetry and transversely to the second conductive bus, electrically connected to the first-level source electrodes; and second-level drain electrodes, extending transversely to the second conductive bus and electrically connected to the first-level drain electrodes.

16. A method, comprising: distributing a first electric potential from a first conductive bus in parallel to first transistor terminals of first active area and to first transistor terminals of a second active area, the first active area and a second active area arranged on the semiconductor substrate symmetrically opposite with respect to an axis of symmetry and accommodating respective heterostructures, the first conductive bus being positioned on a separation region extending along the axis of symmetry between the first active area and the second active area; and distributing a second electric potential from a second conductive bus in parallel to second transistor terminals of the first and the second active areas, wherein the first and the second conductive buses are stacked and extend along the axis of symmetry above the separation region.

17. The method of claim 16, wherein the first transistor terminals are gate terminals, wherein the first conductive bus is a gate bus and the first electric potential is a control signal of the power device.

18. The method of claim 17, wherein the second transistor terminals are source terminals, wherein the second conductive bus is a source bus and the second electric potential is a source electric potential.

19. The method of claim 18, wherein the first conductive bus and the gate terminals are all electrically connected to each other in a direct manner.

20. The method of claim 16, wherein the first and second transistor terminals are terminals of HEMT-type transistors electrically connected to each other in parallel and formed in gallium nitride technology.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0012] For a better understanding of the present disclosure, preferred embodiments are provided, by way of non-limiting example, with reference to the attached drawings, wherein:

[0013] FIG. 1 is a top-plan view of a power device in accordance with an embodiment of the present disclosure;

[0014] FIG. 2 is a perspective view of an internal portion of the power device of FIG. 1, in accordance with an embodiment of the present disclosure;

[0015] FIG. 3 is a cross-section of an internal portion of the power device of FIG. 1, in accordance with an embodiment of the present disclosure;

[0016] FIG. 4 is an electric diagram relating to the power device of FIG. 1, in accordance with an embodiment of the present disclosure; and

[0017] FIGS. 5-7 are top-plan views of the power device of FIG. 1 in subsequent steps of a manufacturing process thereof, in accordance with an embodiment of the present disclosure.

DETAILED DESCRIPTION

[0018] With reference to FIG. 1, a power device in accordance with an embodiment of the present disclosure is indicated, as a whole, with the numeral 1 and includes a die, or chip, 2, and a frame 3 having the die 2 accommodated therein. The power device 1 further includes: a first terminal 4, a second terminal 6 and a third terminal 8, which respectively define in use the external drain terminal, the external source terminal and the external gate terminal (also said control terminal) of the power device 1; a fourth terminal 10, also called sensing terminal or Kelvin terminal, which defines a reference for the control signal applied to the external gate terminal 8, and therefore for switching the power device 1, and is typically at the same potential as the external source terminal 6.

[0019] The terminals 4, 6, 8, 10 of the power device 1 may be mechanically coupled to the frame 3 and are electrically connected to relative pads of the die 2 (see below) by bonding wires 12.

[0020] The power device 1 may also comprise an insulating body of a resin or other polymeric material (not shown in the attached Figures) from which the terminals 4, 6, 8, 10 may protrude through a plurality of protrusions (at least one) for respective terminal which remain exposed to be accessible from the outside.

[0021] FIG. 2 illustrates an internal portion of the die 2 in a triaxial reference system of axes orthogonal to each other X, Y, Z. The die 2 includes an active region 20, which extends in a planar manner in the directions of the X and Y axes above a substrate 21 and is divided, along the Y axis, into a first active area 20a and a second active area 20b identical to each other.

[0022] The first and the second active areas 20a, 20b extend in an identical manner to each other inside the die 2, i.e., they have the same dimensions along the X axis and the Y axis, and are arranged symmetrically opposite with respect to an axis of symmetry H parallel to the X axis. The first and the second active areas 20a, 20b are also separated by a separation region 23 of the same semiconductor material as the substrate 21, which is arranged centrally to the die 2 and extends parallel to the axis of symmetry H. The first and the second active areas 20a, 20b are therefore at a distance from each other and the separation region 23 is interposed therebetween. The first and the second active areas 20a, 20b define two identical and separate power modules of the power device 1.

[0023] The substrate 21 may comprise, for example, a silicon (Si) layer or a silicon carbide (SiC) layer and a gallium nitride (GaN) buffer layer, not shown separately in the accompanying Figures.

[0024] The first and the second active areas 20a, 20b comprise a plurality of active elements 22, identical to each other and distributed in a specular manner. The active elements 22, as also shown in FIG. 3, are equal to each other, are arranged side by side in a periodic manner along the X axis and extend parallel to the Y axis.

[0025] The active elements 22 may be transistors of the HEMT type. In this case, the first and the second active areas 20a, 20b each comprise a channel layer 25 and a barrier layer 27. The channel layers 25 of the first and the second active areas 20a, 20b extend on the substrate 21 and are formed by a first semiconductor material, for example, of gallium nitride (GaN). The barrier layers 27 are superimposed, in contact, each on a respective channel layer 25 and are formed by a second semiconductor material, different from the first, for example, of aluminum gallium nitride (AlGaN). The channel layers 25 and the barrier layers 27 may have, for example, N-type conductivities and define a heterostructure. Interfaces between the channel layers 25 and the barrier layers 27 define heterojunctions 31.

[0026] The active elements 22 are provided with respective drain terminals 24, source terminals 26 and gate terminals 28 and comprise respective portions of the channel layers 25 and of the barrier layers 27 and respective gate regions 29.

[0027] Each gate region 29 extends above and in contact with the respective barrier layer 27 and acts as a control region of each active element 22, in particular as a channel modulation region; the gate region 29 is formed by a conductive material, typically a third doped semiconductor material, such as for example P-type conductivity gallium nitride (pGaN). As known, the gate region 29 operates, on the basis of the voltage applied thereon, so as to modulate the thickness of the conductive channel which forms at the interface between the channel layer 25 and the barrier layer 27; the gate region 29, moreover, ensures the operation in the so-called normally-off mode to an active element 22 of the HEMT type.

[0028] The power device 1 of the present embodiment may therefore be a HEMT power device of GaN-type which exploits, as is typical in the sector, a plurality of elementary cells (namely, the active elements 22) electrically connected in parallel in order to increase the conductive properties thereof (in terms of maximum current that can be supplied) to the external drain and source terminals 4, 6. The die 2 of the power device 1 may have maximum dimensions, for example, equal to 7 mm along the X axis and 23 mm along the Y axis, and each power module defined by the first and the second active areas 20a, 20b may contain, for example, 50200 active elements 22. The first and the second active areas 20a, 20b may extend along the Y axis with typical lengths equal to 0.51.5 mm and the relative separation region 23 may extend along the Y axis for example for 2060 m.

[0029] The drain, source and gate terminals 24, 26, 28 of each active element 22 have substantially the shape of a strip and extend parallel to the Y axis according to the topology shown in FIG. 2.

[0030] In detail, the gate terminal 28 of each active element 22, for example a conductive multilayer of TiN/AlCu/TiN, extends above and in direct electrical contact with the respective gate region 29 and is formed through a common metallization step for all the active elements 22 of both the first and the second active areas 20a, 20b. In particular, the gate terminals 28 are defined by respective gate electrodes or fingers 30 which extend without interruptions along the Y axis throughout all the active region 20, including the separation region 23, and connect gate regions 29 symmetrically opposite with respect to the axis of symmetry H.

[0031] In other words, the gate electrodes 30 extend throughout the entire length of the die 2 parallel to the Y axis and also run above the separation region 23. The gate electrodes 30 are obtained following deposition and lithographic definition steps, using techniques that are typical of the sector (etching).

[0032] The drain and source terminals 24, 26 of each active element 22 are defined, for the first active area 20a, by respective drain electrodes 40a and by respective source electrodes 50a, while, for the second active area 20b, by respective drain electrodes 40b and by respective source electrodes 50b. The drain electrodes 40a, 40b have equal extension along the Y axis and the source electrodes 50a, 50b have equal extension along the Y axis. The position of the drain electrodes 40a and of the source electrodes 50a is symmetrical with respect to the H axis to the position of the drain electrodes 40b and the source electrodes 50b, respectively. Furthermore, the drain electrodes 40a, 40b are separated from each other at the separation region 23, as well as the source electrodes 50a, 50b. The drain electrodes 40a, 40b and the source electrodes 50a, 50b are formed starting from a same first level of metallization M.sub.1, over and through a first intermediate insulating layer 100, of dielectric material, which covers the active region 20. The drain electrodes 40a, 40b and the source electrodes 50a, 50b are hereinafter referred to as first-level drain electrodes and first-level source electrodes.

[0033] Considering the first active area 20a (but the same considerations also apply to the second active area 20b) and referring to FIGS. 2 and 3, the first-level drain electrodes 40a and the first-level source electrodes 50a extend above the first active area 20a alternating to each other, forming a repetitive pattern, along the X axis. The first-level drain electrodes 40a and the first-level source electrodes 50a are in direct contact with the heterostructure of the first active area 20a, defining the respective drain and source terminals 24, 26 of the relative active elements 22. More precisely, the first-level drain electrodes 40a and the first-level source electrodes 50a have respective contact portions 45a, 46a (contact portions 45b, 46b refer, in an identical manner, to the second active area 20b) which extend through the first intermediate insulating layer 100 and run along the Y axis, forming the drain and source terminals 24, 26.

[0034] In detail, a first-level source electrode 50a is shared by two adjacent active elements 22, which therefore share the same source terminal 26. The first-level source electrodes 50a extend above pairs of consecutive gate electrodes 30 and contact the first active area 20a in intermediate positions between these consecutive gate electrodes 30.

[0035] The alternate pattern of the first-level drain electrodes 40a and the first-level source electrodes 50a is such that, observing in section two adjacent active elements 22 (FIG. 3), two first-level drain electrodes 40a placed laterally and symmetrically with respect to the first-level source electrode 50a correspond to a central first-level source electrode 50a. Furthermore, the first-level source electrodes 50a extend above pairs of respective gate electrodes 30 so that a projection of these gate electrodes 30, along the Z axis, is completely contained in these source electrodes 50a; an arrangement of this type typically ensures electrical shielding functions to the corresponding gate regions 29.

[0036] The first level of metallization M.sub.1 also includes a main electrode or gate bus 35 which extends in a transversal manner to the gate electrodes 30, i.e., along the X axis. In detail, the gate bus 35 extends above the gate electrodes 30 and centrally to the die 2: the gate bus 35 is in fact interposed between the first and the second active areas 20a, 20b at the separation region 23, whereon it is superimposed; the H axis is therefore an axis of symmetry of the gate bus 35. In other words, the gate bus 35 extends above a region of the die 2 wherein the active elements 22 are absent.

[0037] The gate bus 35 is electrically connected to the gate electrodes 30 by vias of conductive material which extend along the Z axis (not shown in the attached Figures) and distributes to the gate electrodes 30, and therefore to the active elements 22, the external control signal. The central arrangement of the gate bus 35 allows a symmetrical and balanced distribution of this control signal, in terms of parasitic components of the electrical paths, to both power modules defined by the first and the second active areas 20a, 20b: the relative gate regions 29 therefore switch almost simultaneously and in a balanced manner for all the active elements 22 in parallel; the driving of the two power modules is therefore coherent.

[0038] The gate bus 35 may have a length close to the length along the X axis of the die 2 and a width (along the Y axis) normally between 10 and 50 m.

[0039] The first level of metallization M.sub.1 is separated from the active region 20, along the Z axis, by the first intermediate insulating layer 100. The electrical connections between the first-level drain and source electrodes 40a, 40b, 50a, 50b and the drain and source terminals 24, 26 of the active elements 22 are ensured by the contact portions 45a, 46a, 45b, 46b through the first intermediate insulating layer 100.

[0040] The first-level drain and source electrodes 40a, 40b, 50a, 50b and the gate bus 35 are all obtained starting from the same deposition and definition processes of first level of metallization M.sub.1 and may be formed for example with a multilayer of TiN/AlCu/TiN.

[0041] The power device 20 further includes a second insulating layer 101 on the first intermediate insulating layer 100 and a second level of metallization M.sub.2 of the power device 20 on the second intermediate insulating layer 101. The second level of metallization M.sub.2 overlies the first level of metallization M.sub.1 and includes: a source main electrode (bus) 55; second-level drain electrodes 60a and second-level source electrodes 70a on the first active area 20a; second-level drain electrodes 60b and second-level source electrodes 70b on the second active area 20b. The source bus 55 extends along the X axis while the second-level drain electrodes 60a, 60b and the second-level source electrodes 70a, 70b extend in a transversal manner to the source bus 55, i.e., along the Y axis. The second-level source electrodes 70a, 70b are in direct electrical and structural connection with the source bus 55. More precisely, the second-level source electrodes 70a, 70b extend monolithic from the source bus 55 in opposite directions with respect to the H axis towards the respective first and second active areas 20a, 20b and represent fingers of the source bus 55.

[0042] In detail, the source bus 55 extends along the H axis, which is also an axis of symmetry of the source bus 55, and completely overlies the gate bus 35, shielding it electrically: the shielding function given by the structure of the source connections of second level of metallization M.sub.2 is explained in more detail below.

[0043] The source bus 55 may have a length close to the size of the die 2 along the X axis and a width at least equal to the width of the gate bus 35 along the Y axis and with a value normally between 10 and 50 m.

[0044] Considering the first active area 20a (but the same considerations also apply to the second active area 20b) and referring to FIGS. 2 and 3, the second-level drain electrodes 60a and the second-level source electrodes 70a are arranged alternating to each other, forming a repetitive pattern, along the X axis. In particular, the second-level drain electrodes 60a are interdigitated between two consecutive second-level source electrodes 70a. Furthermore, the second-level drain electrodes 60a and the second-level source electrodes 70a at least partially overlie respective first-level drain electrodes 40a and first-level source electrodes 50a. The second-level drain electrodes 60a and the second-level source electrodes 70a are in electrical connection with the respective first-level drain electrodes 40a and first-level source electrodes 50a by second-level vias 65, of conductive material, that extend along the Z axis.

[0045] The structure of the source connections of second level of metallization M.sub.2 is therefore such that a pair of second-level source electrodes 70a, 70b extends, starting from the source bus 55, in a symmetrical opposite manner with respect to the H axis towards, respectively, the first and the second active areas 20a, 20b and overlying, respectively, a pair of first-level source electrodes 50a, 50b.

[0046] The second-level drain electrodes 60a, 60b are instead separated from each other at the separation region 23.

[0047] The second level of metallization M.sub.2 is separated from the first level of metallization M.sub.1, along the Z axis, by the second intermediate insulating layer 101. The electrical connections between the second-level drain and source electrodes 60a, 60b, 70a, 70b and the first-level drain and source electrodes 40a, 40b, 50a, 50b are ensured by the second-level vias 65 through the second intermediate insulating layer 101.

[0048] The second-level drain and source electrodes 60a, 60b, 70a, 70b and the source bus 55 are all obtained starting from the same deposition and definition processes of second level of metallization M.sub.2 and may be formed for example with a multilayer of TiN/AlCu/TiN.

[0049] The power device 20 includes a third intermediate insulating layer 102 on the second intermediate insulating layer 101 and a third level of metallization M.sub.3 on the third intermediate insulating layer 102. The third level of metallization M.sub.3 overlies the second level of metallization M.sub.2 and includes: a drain metallization region 85 and a source metallization region 95, identical and symmetrical with respect to the H axis, third-level drain electrodes 80 which extend from the drain metallization region 85 and third-level source electrodes 90 which extend from the source metallization region 95. The drain and source regions 85, 95 respectively overlie the first and the second active areas 20a, 20b. The third-level drain and source electrodes 80, 90 extend along the Y axis monolithic from the respective drain and source regions 85, 95. The third-level drain electrodes 80 and the source electrodes 90 extend above the second level of metallization M.sub.2, starting from the drain regions 85 and the source regions 95, respectively, in such a way that, along the X axis, third-level drain electrodes 80 alternate with third-level source electrodes 90, i.e., they are interdigitated with each other (see also FIG. 7).

[0050] As a result, the third-level drain and source electrodes 80, 90 form fingers of the respective drain and source regions 85, 95. Furthermore, each third-level drain electrode 80 overlies (at least partially) and is electrically connected to a pair of second-level drain electrodes 60a, 60b symmetrically opposite with respect to the H axis. Similarly, each third-level source electrode 90 overlies (at least partially) and is electrically connected to a pair of second-level source electrodes 70a, 70b symmetrically opposite with respect to the H axis. These electrical connections between the third level of metallization M.sub.3 and the second level of metallization M.sub.2 are formed by third-level vias 75, of conductive material, which extend along the Z axis through the third intermediate insulating layer 102.

[0051] The third-level drain and source electrodes 80, 90 and the drain and source regions 85, 95 are all obtained starting from the same deposition and definition processes of third level of metallization M.sub.3 and may be formed for example with a multilayer of TiN/AlCu/TiN.

[0052] The third level of metallization M.sub.3 may also be provided, in this type of power devices, with pads which allow the electrical connection to different external levels of electric potential as previously described.

[0053] In particular, referring to FIGS. 1, 2 and 7, a drain pad 86 extends above the drain metallization region 85, wherewith it is in direct electrical connection, while a source pad 96 extends above the source metallization region 95, wherewith it is in direct electrical connection. Furthermore, the third level of metallization M.sub.3 includes a gate pad 36, typically placed at a corner of the die 2 and arranged side by side to the source metallization region 95, wherefrom it is electrically insulated. The gate pad 36 is electrically connected to the gate bus 35 by at least one via of conductive material which extends along the Z axis (not shown in the attached Figures) through the intermediate insulating layers 100, 101, 102 up to the first level of metallization M.sub.1; the gate pad 36 then provides the external control signal to the gate bus 35 to be distributed to the active elements 22.

[0054] The drain, source and gate pads 86, 96, 36 are accessible for coupling with the bonding wires 12 in the definition of the lead frame of the power device 1. In particular, the drain pad 86 is connected to the external drain terminal 4, the source pad 96 is connected to the external source terminal 6 and the gate pad 36 is connected to the external gate terminal 8. The sensing terminal 10, which is typically at the same potential as the external source terminal 6, is therefore connected to the source pad 96.

[0055] The first, the second and the third levels of metallization M.sub.1, M.sub.2, M.sub.3 therefore ensure, in the embodiment presented, that the drain, source and gate electrical paths reach the respective correct electric potentials relating to the voltages externally applied, allowing the flow of the currents thereof through the structure of the power device 1.

[0056] It should be noted that in FIG. 2 portions of the first, the second and the third levels of metallization M.sub.1, M.sub.2, M.sub.3 and of the gate electrodes 30 have been intentionally omitted for clarity of understanding.

[0057] The identical dimensions of the first and the second active areas 20a, 20b allow a balance in terms of conductive paths for the current flowing between the external drain and source terminals 4, 6; in fact, these terminals, as well as the respective drain and source pads 86, 96 of the present topology, are shared by both the first and the second active areas 20a, 20b. Concurrently, the arrangement of the gate bus 35 above the separation region 23 allows the insulation of the distribution grid of the control signal from the potential difference, in use, between the drain and source electrodes of the power device 1, which may assume values, for example, two orders of magnitude greater than the voltage levels of the control signal.

[0058] The first and the second levels of metallization M.sub.1, M.sub.2 of the power device 1 are configured to minimize the inductive parasitic components of the electrical path of the control signal; any unwanted oscillations of the control signal during the switching steps are consequently reduced, decreasing the probability of damage to the gate regions 29.

[0059] In detail, phenomena of cancellation of control signal current flows flowing in opposite directions contribute to minimizing the inductive parasitic components. This cancellation is made possible by the topology described above, and, in more detail, by the gate loop created by the arrangement of the first and the second levels of metallization M.sub.1, M.sub.2.

[0060] With reference to FIGS. 2 and 4, when a control signal T is applied between the external gate terminal 8 and the sensing terminal 10, a forward current log of the control signal flows in the gate bus 35 in the negative direction of the X direction and a corresponding return current I.sub.SK flows in the source bus 55 in the positive direction of the X direction, i.e., in the direction opposite to that of the forward current I.sub.GG. The return current I.sub.SK in fact shares the same conductive path as the source metallizations since the sensing terminal 10 is connected, as described, to the source pad 96; the source bus 55 therefore represents a return path of the control signal loop, the sensing terminal 10 being the reference terminal of the voltage applied to the external gate terminal 8.

[0061] The previously described superimposition of the source bus 55 on the gate bus 35 allows, ultimately, the inductive effects of the forward current I.sub.GG and of the return current I.sub.SK to be cancelled since they flow in opposite directions; the source bus 55 therefore has the function of electrical shielding of the gate bus 35. Furthermore, since the source bus 55 and the gate bus 35 run on a central region of the power device 1 along the axis of symmetry H, the synchronization of the switchings of the active elements 22 is ensured.

[0062] FIG. 4 shows in a circuit diagram the mechanism for canceling the current flows of the control signal. In detail, the circuit diagram has two active elements 22 of HEMT-type described above parallel connected to each other. Through the external gate terminal 8 and the sensing terminal 10, the control signal (graphically represented by a square wave generator T) is applied between the gate terminals 28 and the source terminals 26 of the active elements 22, giving rise to the forward and return currents I.sub.GG, I.sub.SK. Inductances L.sub.GG and L.sub.SK represent the parasitic components of the power device 1 in the gate loop, respectively in the forward and return path; inductances L.sub.DBOND and L.sub.SBOND instead represent the inductive losses generated by the bonding wires 12 (called bonding inductances). Capacitances C.sub.DB and C.sub.SB are present in this type of devices active respectively between the drain terminals 24 and the substrate 21 and between the source terminals 26 and the substrate 21.

[0063] As a result, the minimization of the inductive parasitic components L.sub.GG and L.sub.SK due to the cancellation of the current flows significantly reduces the possibility of forming LC-type resonant circuits with the substrate capacitances C.sub.DB and C.sub.SB; these resonant circuits are in fact one of the causes of uncontrolled oscillations and peaks at the gate terminals 28 during the switching phases.

[0064] The described power device 1 may be obtained by the following simplified manufacturing process, according to which the first, the second and the third levels of metallization M.sub.1, M.sub.2, M.sub.3 of the die 2 are deposited and defined as shown in FIGS. 5-7.

[0065] In particular, the die 2 is provided being part of a semiconductor wafer not illustrated, and already provided with the substrate 21, the active region 20 and the gate electrodes 30; the substrate 21, the active region 20 and the gate electrodes 30 are protected by the first intermediate insulating layer 100, of dielectric material. A first deposition of metal layer (FIG. 5) and subsequent lithographic definition, using etching procedures typical of the sector, give rise to the first level of metallization M.sub.1. In detail, the first-level drain and source electrodes 40a, 40b, 50a, 50b and the gate bus 35 are formed. As may be observed in FIG. 5, the gate bus 35 is L-shaped outside the active region 20 so that it may be connected through relative vias of conductive material (not shown in the attached Figures) to the gate pad 36 of the subsequent third level of metallization M.sub.3.

[0066] The contact portions 45a, 46a, 45b, 46b are also defined, at relative openings already formed in the first intermediate insulating layer 100, in order to ensure the electrical connections (reference be made to FIG. 3) between the first level of metallization M.sub.1 and the underlying gate electrodes 30 and active region 20.

[0067] The second intermediate insulating layer 101 is then formed, which covers the first level of metallization M.sub.1. Subsequently, a second deposition of metal layer (FIG. 6) and subsequent lithographic definition, with techniques which may be identical to the first deposition, give rise to the second level of metallization M.sub.2. In detail, the second-level drain and source electrodes 60a, 60b, 70a, 70b and the source bus 55 are formed. As may be observed in FIG. 6, the source bus 55 is also (L) shaped so as to shield the underlying gate bus 35 throughout its extension.

[0068] The second-level vias 65 are also defined, at relative openings formed in the second intermediate insulating layer 101, in order to ensure the electrical connections (reference be made to FIG. 3) between the second level of metallization M.sub.2 and the first level of metallization M.sub.1.

[0069] The third intermediate insulating layer 102 is then formed, which covers the second level of metallization M.sub.2. Subsequently, a third deposition of metal layer (FIG. 7) and subsequent lithographic definition, with techniques which may be identical to the first and the second depositions, give rise to the third level of metallization M.sub.3. In detail, the third-level drain and source electrodes 80, 90 and the drain and source regions 85, 95 are formed; subsequently the drain, source and gate pads 86, 96, 36 are also defined.

[0070] The third-level vias 75 are also defined, at relative openings formed in the third intermediate insulating layer 102, in order to ensure the electrical connections (reference be made to FIG. 3) between the third level of metallization M.sub.3 and the second level of metallization M.sub.2.

[0071] Finally, in a manner not shown in the attached Figures, the final steps of the present manufacturing process follow, including depositing any dielectric passivation layers above the third level of metallization M.sub.3, dicing the wafer to obtain the die 2, positioning the die 2 inside the frame 3 and coupling the pads of the die 2 to the relative external terminals by the wire bondings 12, and therefore the steps of co-molding the insulating body, thus obtaining the power device 1 of FIG. 1.

[0072] Finally, it is clear that modifications and variations may be made to what has been described and illustrated herein without thereby departing from the scope of the present disclosure, as defined in the attached claims.

[0073] For example, the active elements 22 may be of a type different from the GaN HEMT type, for example MOSFET (Metal Oxide Semiconductor Field-Effect Transistor) transistors, superjunction MOSFETs, IGBTs (Insulated-Gate Bipolar Transistor) and the like, capable of operating at high voltages (up to hundreds of Volts) and with currents that have rapid switching time trends.

[0074] The active region 20 may for example comprise further layers inside the relative heterostructures defined by the first and the second active areas 20a, 20b, in addition to the channel layer 25, the barrier layer 27 and the gate region 29 described above.

[0075] The die 2 may for example be provided with further levels of metallization, and relative intermediate insulating layers, in addition to the first, the second and the third levels of metallization M.sub.1, M.sub.2, M.sub.3 described above.

[0076] A heterojunction power device, may be summarized as including: a substrate (21) containing semiconductor material; a first active area (20a) and a second active area (20b), arranged on the substrate (21) symmetrically opposite with respect to an axis of symmetry (H) and accommodating respective heterostructures (25, 27, 29); a separation region (23), extending along the axis of symmetry (H) between the first active area (20a) and the second active area (20b); a first conductive bus (35) configured to distribute a first electric potential (V.sub.G) of the power device (1) in parallel to the first and the second active areas (20a, 20b); and a second conductive bus (55) configured to distribute a second electric potential (V.sub.S) of the power device (1), different from the first electric potential (V.sub.G), in parallel to the first and the second active areas (20a, 20b), wherein the first and the second conductive buses (35, 55) extend along the axis of symmetry (H) above the separation region (23), the second conductive bus (55) overlying the first conductive bus (35).

[0077] The first conductive bus (35) may be a gate bus and the first electric potential (V.sub.G) may be a control signal of the power device (1).

[0078] The second conductive bus (55) may be a source bus and the second electric potential (V.sub.S) may be a source electric potential of the power device (1).

[0079] The power device may include a first level of metallization (M.sub.1), wherein the first conductive bus (35) may be a portion of the first level of metallization (M.sub.1); the first level of metallization (M.sub.1) further including: first-level drain electrodes (40a) and first-level source electrodes (50a) associated with the first active area (20a), extending transversely to the first conductive bus (35) and superimposed on the first active area (20a); and first-level drain electrodes (40b) and first-level source electrodes (50b) associated with the second active area (20b), extending transversely to the first conductive bus (35), superimposed on the second active area (20b) and separated, at the separation region (23), from the first-level drain electrodes (40a) and from the first-level source electrodes (50a) associated with the first active area (20a).

[0080] The first-level drain electrodes (40a) and the first-level source electrodes (50a) associated with the first active area (20a) may be arranged in a symmetrical manner, with respect to the axis of symmetry (H), to respective first-level drain electrodes (40b) and first-level source electrodes (50b) associated with the second active area (20b); and wherein in the first active area (20a) and in the second active area (20b) the first-level drain electrodes (40a, 40b) may be arranged alternating, along the axis of symmetry (H), with the first-level source electrodes (50a, 50b).

[0081] The power device may include a second level of metallization (M.sub.2) superimposed on the first level of metallization (M.sub.1), wherein the second conductive bus (55) may be a portion of the second level of metallization (M.sub.2); the second level of metallization (M.sub.2) further including: second-level drain electrodes (60a) and second-level source electrodes (70a) associated with the first active area (20a), extending transversely to the second conductive bus (55) and superimposed on the first active area (20a); second-level drain electrodes (60b) and second-level source electrodes (70b) associated with the second active area (20b), extending transversely to the second conductive bus (55) and superimposed on the second active area (20b); the second-level source electrodes (70a, 70b) extending from the second conductive bus (55) in opposite directions with respect to the axis of symmetry (H); and the second-level drain electrodes (60a) associated with the first active area (20a) being separated, at the separation region (23), from the second-level drain electrodes (60b) associated with the second active area (20b).

[0082] The second-level drain electrodes (60a) and the second-level source electrodes (70a) associated with the first active area (20a) may be arranged, with respect to the axis of symmetry (H), in a manner symmetrical, to the second-level drain electrodes (60b) and to the second-level source electrodes (70b) associated with the second active area (20b); and the second-level drain electrodes (60a, 60b) may be arranged alternating, along the axis of symmetry (H), with the second-level source electrodes (70a, 70b).

[0083] The power device may include a third level of metallization (M.sub.3), superimposed on the second level of metallization (M.sub.2), provided with: a drain metallization region (85), superimposed on the first active area (20a); a source metallization region (95), superimposed on the second active area (20b); third-level drain electrodes (80) and third-level source electrodes (90), extending transversely to the axis of symmetry (H) starting from the drain metallization region (85) and from the source metallization region (95), respectively; and the third-level drain electrodes (80) being arranged alternating, along the axis of symmetry (H), with the third-level source electrodes (90).

[0084] The power device may include second-level conductive vias (65) between the first level of metallization (M.sub.1) and the second level of metallization (M.sub.2) and third-level conductive vias (75) between the second level of metallization (M.sub.2) and the third level of metallization (M.sub.3).

[0085] The first and the second active areas (20a, 20b) each may include: a respective channel layer (25) and a respective barrier layer (27) defining respective heterojunctions (31); a plurality of respective active elements (22), extending transversely to the axis of symmetry (H) and arranged periodically along the axis of symmetry (H), wherein the active elements (22) include respective portions of the channel layers (25) and of the barrier layers (27) and respective gate regions (29) superimposed, in contact, on the barrier layer (27) of the respective active area (20a, 20b); the active elements (22) each further including a drain terminal (24), a source terminal (26) and a gate terminal (28); and the gate terminals (28) including gate electrodes (30), superimposed, in contact, on respective gate regions (29) and extending transversely to the axis of symmetry (H).

[0086] The third-level drain electrodes (80), the second-level drain electrodes (60a, 60b), the first-level drain electrodes (40a, 40b) and the drain terminals (24) of the active elements (22) may be superimposed on each other and all electrically connected to each other in a direct manner; the third-level source electrodes (90), the second-level source electrodes (70a, 70b), the second conductive bus (55), the first-level source electrodes (50a, 50b) and the source terminals (26) of the active elements (22) may be superimposed on each other and all electrically connected to each other in a direct manner; and the first-level drain electrodes (40a, 40b) and the first-level source electrodes (50a, 50b) may have respective contact portions (45a, 46a, 45b, 46b) which extend transverse to the axis of symmetry (H) and form the drain terminals (24) and the source terminals (26).

[0087] The first conductive bus (35), the gate electrodes (30) and the gate terminals (28) may be all electrically connected to each other in a direct manner.

[0088] The active elements (22) may be HEMT-type transistors, electrically connected to each other in parallel and formed in gallium nitride technology.

[0089] A process for manufacturing a heterojunction power device, may be summarized as including: on a substrate (21) containing semiconductor material, forming a first active area (20a) and a second active area (20b) symmetrically opposite with respect to an axis of symmetry (H), accommodating respective heterostructures (25, 27, 29) and separated by a separation region (23), extending along the axis of symmetry (H); forming, on the separation region (23), a first conductive bus (35) configured to distribute a first electric potential (V.sub.G) of the power device (1) in parallel to the first and the second active areas (20a, 20b); and forming, on the separation region (23), a second conductive bus (55) configured to distribute a second electric potential (V.sub.S) of the power device (1), different from the first electric potential (V.sub.G), in parallel to the first and the second active areas (20a, 20b), wherein the first and the second conductive buses (35, 55) extend along the axis of symmetry (H), the second conductive bus (55) overlying the first conductive bus (35).

[0090] The manufacturing process may include: forming a first level of metallization (M.sub.1); and from the first level of metallization (M.sub.1) forming: the first conductive bus (35); first-level drain and source electrodes (40a, 40b, 50a, 50b), extending transversely to the first conductive bus (35), configured to provide, respectively, drain electric potentials and source electric potentials of the power device (1) to the first active area (20a) and to the second active area (20b); forming a second level of metallization (M.sub.2) above first level of metallization (M.sub.1); and from the second level of metallization (M.sub.2) forming: the second conductive bus (55); second-level source electrodes (70a, 70b), extending from the second conductive bus (55) in opposite directions with respect to the axis of symmetry (H) and transversely to the second conductive bus (55), electrically connected to the first-level source electrodes (50a, 50b); and second-level drain electrodes (60a, 60b), extending transversely to the second conductive bus (55) and electrically connected to the first-level drain electrodes (40a, 40b).

[0091] These and other changes can be made to the embodiments in light of the above-detailed description. In general, in the following claims, the terms used should not be construed to limit the claims to the specific embodiments disclosed in the specification and the claims, but should be construed to include all possible embodiments along with the full scope of equivalents to which such claims are entitled. Accordingly, the claims are not limited by the disclosure.