Bidirectional transistor having a low resistance heterojunction in an on state

10692997 ยท 2020-06-23

Assignee

Inventors

Cpc classification

International classification

Abstract

A bidirectional heterojunction transistor includes first and second conduction electrodes, first and second gates between the conduction electrodes, and first and second reference electrodes between the gates. The transistor further includes a superposition of semiconductor layers, including channel zones that are vertically in line with the gates, a first conduction zone between the first conduction electrode and the first channel zone, and a second conduction zone between the second conduction electrode and the second channel zone. The superposition of semiconductor layers also includes a third conduction zone that is separated from the first and second conduction zones by the first and second channel zones, respectively, and a first electrical connection that is connected to the third conduction zone and to the first reference electrode.

Claims

1. A device including a bidirectional heterojunction field-effect transistor, wherein the bidirectional transistor comprises: first and second conduction electrodes, first and second gates that are arranged between the first and second conduction electrodes, and first and second reference electrodes that are arranged between the first and second gates; a superposition of a first semiconductor layer and of a second semiconductive layer so as to form an electron gas layer, the electron gas layer comprising: a first channel zone that is arranged vertically in line with the first gate; a second channel zone that is arranged vertically in line with the second gate; a first conduction zone running between the first conduction electrode and the first channel zone; a second conduction zone running between the second conduction electrode and the second channel zone; a third conduction zone that is separated from the first and second conduction zones by the first and second channel zones, respectively; a first electrical connection that is connected to the third conduction zone in proximity to the first channel zone and connected to the first reference electrode; a second electrical connection that is connected to the third conduction zone in proximity to the second channel zone and connected to the second reference electrode; the device also comprising a first control circuit that is connected to the first reference electrode, the first control circuit being programmed to generate a control voltage for the transistor from a voltage of the first reference electrode and programmed to apply said control voltage to the first gate; the device also comprising a second control circuit that is connected to the second reference electrode, the second control circuit being programmed to generate a control voltage for the transistor from a voltage of the second reference electrode and programmed to apply said control voltage to the second gate; wherein said first and second gates extend over the electron gas layer from one end to the other in a first direction that is perpendicular to a direction of conduction between the first and second conduction electrodes, the transistor including respective contacts for the first gate, for the second gate, for the first conduction electrode and for the second conduction electrode, the third conduction zone being arranged between the contacts of the first gate and of the first conduction electrode on one side and the contacts of the second gate and of the second conduction electrode on the other side; wherein said first and second connections are electron gas strips that protrude with respect to the third conduction zone in said first direction.

2. The device according to claim 1, wherein a distance between the first and second conduction zones is at least equal to 3 m.

3. The device according to claim 2, wherein the distance between the first and second conduction zones is at least equal to 6 m.

4. The device according to claim 1, wherein a distance between said first connection and said first channel zone is at most equal to 1.5*Lg, where Lg is a gate length.

5. The device according to claim 1, wherein a distance between said first connection and said first channel zone is at most equal to 0.15*D, where D is the distance between the first and second channel zones.

6. The device according to claim 1, wherein a distance between said first connection and said first channel zone is at most equal to 2 m.

7. The device according to claim 1, wherein a distance between said first and second conduction zones is at least equal to 10 m.

8. The device according to claim 1, wherein the first and second control circuits are configured to apply, simultaneously, control voltages to the first and second gates, according to combinations selected from the group consisting of: a control voltage to the first gate that is higher than a threshold voltage of the first gate and a control voltage to the second gate that is higher than a threshold voltage of the second gate; a control voltage to the first gate that is lower than a threshold voltage of the first gate and a control voltage to the second gate that is higher than a threshold voltage of the second gate; a control voltage to the first gate that is lower than a threshold voltage of the first gate and a control voltage to the second gate that is lower than a threshold voltage of the second gate.

9. The device according to claim 8, wherein said first and second control circuits are configured to apply, simultaneously, control voltages to the first and second gates with a control voltage to the first gate that is higher than the threshold voltage of the first gate and a control voltage to the second gate that is higher than the threshold voltage of the second gate, a difference between the control voltage for the first gate and the threshold voltage of the first gate being greater than a difference between the control voltage for the second gate and the threshold voltage of the second gate.

10. The device according to claim 1, wherein the first and second reference electrodes include respective contacts with said first and second connections, the third conduction zone being arranged between the contacts of the first and second reference electrodes.

11. The device according to claim 1, wherein said first and second control circuits are configured to apply control voltages of different amplitudes to the first and second gates, respectively.

12. The device according to claim 1, wherein the first and second semiconductor layers are GaN alloys.

Description

(1) Other features and advantages of the invention will become clearly apparent from the description of it that is given below, by way of nonlimiting indication, with reference to the appended drawings, in which:

(2) FIG. 1 is a cross-sectional view of an HEMT transistor according to one embodiment of the invention;

(3) FIG. 2 is a top view of the transistor of FIG. 1;

(4) FIG. 3 is a cross-sectional view of the transistor of FIG. 1 in the on state;

(5) FIG. 4 is a graph showing the voltage/current between the conduction electrodes of the transistor in the on state;

(6) FIG. 5 is a cross-sectional view of an HEMT transistor according to one embodiment of the invention, in the off state;

(7) FIG. 6 is a top view of the transistor of FIG. 5;

(8) FIG. 7 is a graph showing the voltage/current between the conduction electrodes of the transistor in the off state;

(9) FIG. 8 is a cross-sectional view of an HEMT transistor according to one embodiment of the invention, in the asymmetric on state;

(10) FIG. 9 is a top view of the transistor of FIG. 8;

(11) FIG. 10 is a graph showing the voltage/current between the conduction electrodes of the transistor in an asymmetric on state;

(12) FIG. 11 is a graph showing the voltage/current between the conduction electrodes of the transistor in an asymmetric off state;

(13) FIG. 12 illustrates one step of a process for fabricating a transistor according to one embodiment of the invention;

(14) FIG. 13 illustrates another step of a process for fabricating a transistor according to one embodiment of the invention;

(15) FIG. 14 illustrates another step of a process for fabricating a transistor according to one embodiment of the invention;

(16) FIG. 15 illustrates another step of a process for fabricating a transistor according to one embodiment of the invention;

(17) FIG. 16 illustrates another step of a process for fabricating a transistor according to one embodiment of the invention.

(18) FIG. 1 is a schematic figure showing an exemplary HEMT transistor 2, i.e. a high-electron-mobility field-effect transistor, according to one embodiment of the invention. FIG. 2 is a top view of this transistor 2. The transistor 2 includes a silicon substrate 6, which is typically slightly p-doped. A layer of semiconductor material 7 made of a binary alloy of a group III element nitride (in this instance GaN) is formed above the substrate 6. What is termed a barrier layer of semiconductor material, made of a ternary alloy of a group III element nitride (in this instance AlGaN), 9 is formed above the layer 7. An electron gas layer 8 is intrinsically formed at the interface between the layer 7 and the layer 9.

(19) The transistor 2 includes a first conduction electrode 11 and a second conduction electrode 12, between which a current must selectively be established or interrupted.

(20) The transistor 2 further includes a first control gate 21 and a second control gate 22. The control gates 21 and 22 are arranged between the conduction electrodes 11 and 12 so that each of the gates 21 and 22 is able to interrupt the current between the conduction electrodes 11 and 12. The control gates 21 and 22 thus extend over the electron gas layer 8 from one end to the other in a transverse direction, i.e. a direction that is perpendicular to the direction of conduction between the conduction electrodes 11 and 12.

(21) The transistor 2 also includes a first reference electrode 31 and a second reference electrode 32. The reference electrodes 31 and 32 are electrically connected to the electron gas layer 8, in proximity to the respective control electrodes 21 and 22. The reference electrodes 31 and 32 are electrically connected to the electron gas layer via respective electrical connections 311 and 321. The reference electrodes 31 and 32 are here arranged between the control electrodes 21 and 22. In particular, the reference electrodes 31 and 32 are arranged in proximity to the gates 21 and 22, respectively.

(22) The electrodes 11, 12, 31 and 32 and the gates 21 and 22 are here formed on top of the layer 9.

(23) The layer 8 includes a channel zone 81, vertically in line with the gate 21. The channel zone 81 is selectively made conductive or nonconductive according to the voltage applied to the gate 21. The layer 8 includes a channel zone 82, vertically in line with the gate 22. The channel zone 82 is selectively made conductive or nonconductive according to the voltage applied to the gate 22. Depending on the design of the transistor 2, the zone 81 turns the left-hand portion of this transistor 2 off or on for a zero difference in potential between the reference electrode 31 and the gate 21. The transistor 2 is here a normally on transistor. Depending on the design of the transistor 2, the zone 82 turns the right-hand portion of this transistor 2 off or on for a zero difference in potential between the reference electrode 32 and the gate 22. The transistor 2 is here a normally off transistor.

(24) The layer 8 also includes a zone 83, running between the electrode 11 and the zone 81. The zone 83 forms an electrical connection between the electrode 11 and the channel zone 81. The electrode 11 is electrically connected in a manner known per se to the zone 83, for example by locally doping the layer 9 below the electrode 11. The layer 8 also includes a zone 84, running between the electrode 12 and the zone 82. The zone 84 forms an electrical connection between the electrode 12 and the channel zone 82. The electrode 12 is electrically connected in a manner known per se to the zone 84, for example by locally doping the layer 9 below the electrode 12. The distance between the zones 83 and 84 is advantageously at least equal to 3 m, preferably at least equal to 6 m or, lastly, preferably at least equal to 10 m with a view to guaranteeing a high breakdown voltage for the transistor 2. The layer 8 also includes a central zone 85, located between the zones 81 and 82. The conduction zone 85 is separated from the conduction zones 83 and 84 by the channel zones 81 and 82, respectively. The electrical connection 311 is connected to the conduction zone 85 in proximity to the channel zone 81. The electrical connection 311 thus allows the potential of the conduction zone 85 to be applied to the reference electrode 31 as close as possible to the channel zone 81. The electrical connection 311 is here formed from an extension or lateral protrusion of the electron gas layer 8 (in particular of the zone 85) to vertically in line with the reference electrode 31. The electrical connection 321 is connected to the conduction zone 85 in proximity to the channel zone 81. The electrical connection 321 thus allows the potential of the conduction zone 85 to be applied to the reference electrode 32 as close as possible to the channel zone 82. The electrical connection 321 is here formed from a lateral extension or protrusion of the electron gas layer 8 (in particular of the zone 85) to vertically in line with the reference electrode 32.

(25) To make the transistor 2 more compact, the zone 85 is advantageously arranged between a group of contacts for the gate 21 and for the electrode 11 and a group of contacts for the gate 22 and for the electrode 12. These groups of contacts are thus arranged on either side of the zone 85 in the transverse direction. Advantageously, the zone 85 is advantageously arranged between a group of contacts for the gate 21, for the electrode 31 and for the electrode 11 and a group of contacts for the gate 22, for the electrode 31 and for the electrode 12.

(26) The layer 9 includes a zone 41, vertically in line with the electrode 31. The zone 41 connects the electrode 31 to the electrical connection 311, for example by the layer 9 being locally doped vertically in line with the electrode 31. The distance between the connection 311 and the zone 81 (i.e. the distance between the closest points of the connection 311 and of the zone 81) is advantageously at most equal to 1.5 times the length of the gate 21. The distance between the connection 311 and the zone 81 is advantageously at most equal to 0.15 times the width of the zone 85. The distance between the connection 311 and the zone 81 is advantageously preferably at most equal to 2 m. The layer 9 also includes a zone 42, vertically in line with the electrode 32. The zone 42 connects the electrode 32 to the electrical connection 321, for example by the layer 9 being locally doped vertically in line with the electrode 32. The distance between the connection 321 and the zone 82 (i.e. the distance between the closest points of the connection 321 and of the zone 82) is advantageously at most equal to 1.5 times the length of the gate 22. The distance between the connection 321 and the zone 82 is advantageously at most equal to 0.15 times the width of the zone 85. The distance between the connection 321 and the zone 82 is advantageously preferably at most equal to 2 m. These parameters promote the retrieval of a reference potential in proximity to the channel whose conduction state it is desired to control.

(27) A control circuit 51 is configured to apply a bias voltage to the gate 21 which is defined relative to the voltage of the reference electrode 31. Controlling relative to the potential of a reference electrode in this way makes it possible to avoid having to reference the potential of a gate relative to the potential of its conduction electrode, which is advantageous for a bidirectional transistor for which a driver circuit for driving the gates is easy to size because the difference in potential between the reference electrodes 31 and 32 is smaller. Similarly, a control circuit 52 is configured to apply a bias voltage to the gate 22 which is defined relative to the voltage of the reference electrode 32. The transistor 2 is here insulated from other electronic components by trench isolations (not illustrated).

(28) Such a transistor 2 behaves as a bidirectional switch that can be driven, for each direction, by a voltage applied between a gate and a potential reference, but the footprint and the resistance in the on state of which are comparable to those of a unidirectional transistor. Specifically, when the transistor 2 is in the off state, the breakdown voltage is here ensured by the distance all the way between the gates 21 and 22 (corresponding to the zone 85). Furthermore, this breakdown voltage is obtained without negatively affecting the on-state resistance of the transistor 2, since the conduction distance through the electron gas 8 is substantially equal to the distance providing the breakdown voltage of the transistor 2 in the off state, contrary to the teaching of the document EP2736078.

(29) The following notation will be used: V11: voltage on the electrode 11; V12: voltage on the electrode 12; V51: the potential difference applied between the gate 21 and the reference electrode 31; V52: the potential difference applied between the gate 21 and the reference electrode 32; V1211 the potential difference between the conduction electrodes 11 and 12.

(30) In the present example, the transistor 2 is a normally off transistor, with a positive threshold voltage Vth.

(31) When the voltages V51 and V52 equal one another and are higher than Vth, the current flows through the layer 8 between the electrodes 11 and 12, from the electrode 11 to the electrode 12. The transistor is then in a first operating mode: symmetric on state. This behaviour is illustrated by FIG. 3 and the voltage/current graph of FIG. 4. The electron gas layer 8 runs here continuously until making electrical contact with the conduction electrodes 11 and 12. The transistor 2 thus exhibits similar behaviour with the current in the opposite direction, when V11 is higher than V12 or when V12 is higher than V11. In this operating mode, the transistor 2 is used in current-modulation mode via the amplitude of the potential differences V51 and V52. The measurement was taken here with a voltage V1211 from 3 V to 3 V and potential differences V51 and V52 of 6 V (with Vth=2 V).

(32) However, when V51 and V52 are lower than Vth, whatever the voltages V11 and V12, the transistor 2 is off in both directions. This is a second operating mode: the symmetric off state. This behaviour is illustrated by FIGS. 5 and 6 and the voltage/current graph of FIG. 7. The electron gas layer 8 is here suppressed between the contacts with the electrodes 11 and 12. In this example, the potential differences V51 and V52 are equal to 6 V.

(33) When V51>V52>Vth (and reciprocally, if V52>V51>Vth), the transistor 2 has a current that is regulated by the gate voltages, in particular V52, which is the smallest: the gate 51 allows more electrons into the electron gas layer 802 than the gate 52. This is a third operating mode: the asymmetric on state. The transistor 2 behaves as a diode for conduction from the electrode 12 to the electrode 11. The current from the electrode 11 to the electrode 12 is relatively small. This behaviour is illustrated by FIGS. 8 and 9 and the voltage/current graph of FIG. 10. In this example, Vth=2 V and the potential differences V51 and V52 are 6 V and 3 V, respectively.

(34) When V51>Vth>V52, asymmetric operation in the off state is obtained: the transistor 2 again behaves as a diode for conduction from the electrode 12 to the electrode 11. Current from the electrode 11 to the electrode 12 is blocked, as illustrated in FIG. 11. In this example, Vth=2 V and the potential differences V51 and V52 are 6 V and 1 V, respectively.

(35) FIGS. 12 to 16 are sectional views of an HEMT transistor 2 at various steps in an exemplary fabrication process. FIGS. 12 to 16 illustrate the right-hand portion of the transistor 2 at the conduction electrode 12, the gate 22 and the reference electrode 32. The left-hand portion of the transistor 2 exhibits symmetry with respect to this right-hand portion, with which it is connected via an intermediate portion including the conduction zone 85.

(36) For the transistor 2 in the process of being formed as illustrated in FIG. 12, a silicon substrate 6, which is typically slightly p-doped, with a thickness of 1 mm is provided here, on which a semiconductor layer 7 made of a binary alloy of a group III element nitride (in this instance GaN) is formed. A semiconductor barrier layer 9, made of a ternary alloy of a group III element nitride (for example AlGaN, for example with an aluminium content ranging from 20 to 25%, for a layer 7 made of GaN), with a thickness of 24 nm for example, is formed above the layer 7. An electron gas layer (not shown here) is intrinsically formed at the interface between the layer 7 and the layer 9. The electron gas layer may be delimited according to the configuration illustrated as a top view in FIG. 12. A passivation layer 10 is formed over the layer 9. The layer 10 may for example be formed in situ, straight after the deposition of the layer 9. The layer 10 may for example be made of Si.sub.3N.sub.4 and have a thickness of 10 nm. The layer 10 is covered with an insulating layer 11, formed for example of Si.sub.3N.sub.4. A groove 100 is in particular formed at the site of the gate to be formed, by anisotropically etching the layers 11, 10 and 9. The groove 100 extends here partly into the layer 7. For example, the width of the groove 100 is between 0.25 and 0.5 m, and the depth thereof is between 225 and 325 nm. The groove 100 runs through the electron gas layer from one end to the other, separating it into two disjunct areas.

(37) Next, as illustrated in FIG. 13, the contact for the gate 22 is formed by successively depositing the following layers. The layer 220, made of a material such as Al.sub.2O.sub.3/SiO.sub.2, is first deposited, for example using a wafer-scale atomic layer deposition (ALD) process, and then potentially subjected to a post-deposition annealing operation. Next, the layer 221, including for example a material such as TiN, is deposited, for example using a wafer-scale physical vapour deposition (PVD) process. Lastly, the layer 222, including a material such as W or AlCu, is deposited, for example using a wafer-scale chemical vapour deposition (CVD) process. Next, after etching, only the gate stack, typically exhibiting a zone protruding for example by 25 m at least on either side of the groove 100, shown in FIG. 12, is retained from the assembly of layers 220, 221 and 222. The insulating layer 11 is thus accessible over the rest of the component.

(38) Next, as illustrated in FIG. 14, the layer 301 is formed by means of a wafer-scale deposition of a material, typically SiO.sub.2, over a thickness ranging for example from 100 to 200 nm. Next, on either side of the stack of layers 220, 221 and 222, the layer 301 is etched so as to provide access to the insulating layer 11 over a length for example ranging from 1 to 1.5 m.

(39) FIG. 15 illustrates the opening, by means of a lithography process, of an access 400 for the reference contact, of an access 401 for the gate contact and of an access 402 for the contact for the conduction electrode.

(40) FIG. 16 illustrates the deposition of the metal for the contact for the reference electrode 32, of the metal for the gate contact 22 and of the metal for the contact for the conduction electrode 12. These deposits will advantageously be made of TiAlCu alloy, deposited in a wafer-scale process, before being subjected to a lithography etching process so as to form contacts that are independent of one another. It should be noted that the reference contact 32, located towards the median portion of the component, and the contact for the conduction electrode 12, located at the end of the component, rest on the layer 7, whereas the gate contact 22 rests on the layer 222.

(41) The above examples describe reference electrodes that are electrically connected to the zone 85 of the electron gas layer via a lateral extension of this electron gas layer. It is also possible to envisage forming the electrical connection for the reference electrodes by forming a metal deposit running from the zone 85 of the electron gas layer to a remote lateral zone, where a respective reference electrode is connected.