Patent Application
20230117946
The present disclosure relates to semiconductor devices. Particularly, but not exclusively, the disclosure relates to hetero-structure AlGaN/GaN high electron mobility transistors or rectifiers.
Gallium Nitride (GaN) is a wide band gap material with properties that make it a suitable candidate for use in several fields of application (e.g. radio-frequency electronics, opto-electronics, power electronics) which require solid-state devices.
GaN technology allows transistors with high electron mobility and high saturation velocity to be designed. These properties of GaN have made it a good candidate for high-power and high-temperature microwave applications, for example radar and cellular communications systems. As systems expand in subscribers and desired capacity, interest in increasing their operating frequency and power has grown correspondingly. Higher frequency signals can carry more information (bandwidth) and allow for smaller antennas with very high gain.
Additionally, GaN with its wide bandgap offers the potential for emitting light at higher frequencies for example the green, blue, violet, and ultraviolet portions of the electromagnetic spectrum.
In the last decade, Gallium Nitride (GaN) has increasingly been considered as a very promising material for use in the field of power devices. The application areas range from portable consumer electronics, solar power inverters, electric vehicles, and power supplies. The wide band gap of the material (Eg=3.39 eV) results in high critical electric field (Ec=3.3 MV/cm) which can lead to the design of devices with a shorter drift region, and therefore lower on-state resistance if compared to a silicon-based device with the same breakdown voltage.
The use of an Aluminium Gallium Nitride (AlGaN)/GaN heterostructure also allows the formation of a two-dimensional electron gas (2DEG) at the hetero-interface where carriers can reach very high mobility (μ=2000 cm2/(Vs)) values. In addition, the piezopolarization charge present at the AlGaN/GaN heterostructure, results in a high electron density in the 2DEG layer (e.g. 1×1013cm−2). These properties allow the development of High Electron Mobility Transistors (HEMTs) and Schottky barrier diodes with very competitive performance parameters. An extensive amount of research has focused on the development of power devices using AlGaN/GaN heterostructures.
However, the 2DEG which inherently exists at the AlGaN/GaN hetero-interface has been found to be unstable and lose electron charge under a prior high voltage stress. This phenomenon is known as dynamic Ron or current collapse. When the device is under high voltage stress in the off-state (a high potential is present on the high voltage drain terminal with respect to the source terminal and the gate voltage is at a potential below the threshold voltage to ensure that the device is off), part of the electrons in the 2DEG are lost by trapping mechanisms in either traps in the bulk of the GaN, the transition layer (i.e. the layer placed between the substrate and the GaN buffer) or at the surface of the device. Hot carrier injection in the passivation layer, next to the gate (the control terminal) also may play a role, though this is not believed to be the major contribution to the dynamic Ron phenomenon. This loss of charge from the 2DEG layer to the surrounding traps leads to a subsequent decrease in the conductivity during on-state and thus an increase in the on-state resistance. The effect could be seen also during the switching or when the device is in operation in a real system.
The hot carrier injection and the trapping effect of the electrons in the bulk of the GaN, the transition layer (i.e. the layer placed between the substrate and the GaN buffer) or at the surface of the device, in the vicinity of the gate leads not only to a loss in the 2DEG layer conductivity but is also associated with a shift in the threshold voltage. The trapping and the hot carrier injection are significantly more prominent at higher electric fields. One of the peaks of the electric field occurs at the drain side of the edge of the gate structure.
The use of metal field plates has also only been partially successful in reducing the electric field peak. At the same time, if the extension of the field plate is too long towards the drain terminal, it reduces the effective distance of the drift region where the field is accommodated, resulting in a loss of the breakdown voltage.
Another way to alleviate the increase in the on-state resistance is to inject holes during the off-state or during the on-state operation to passivate the bulk or surface traps and prevent or slow down the trapping of electrons in the vicinity of the 2DEG layer (either in the bulk or at the surface).
There are therefore two possible ways to reduce or minimise the effect of the dynamic increase in the on-state resistance, when an off-state stress was a priori present: (i) to design the device in order to reduce the field peaks in the structure, in particular around the gate structure, to reduce or minimise the effect of electron trapping from the 2DEG and/or hot carrier injection, and (ii) and/or to inject holes into the bulk and actively passivate the bulk or surface traps.
There are several theories describing the effect of the injected holes. One possible mechanism is that the injected holes change the electric field distribution from the 2DEG to the substrate, by moving the high fields away from the 2DEG vicinity, towards the substrate. As a result the 2DEG is exposed to lower electric fields which slows down the mechanism of electron trapping in the vicinity of the 2DEG as shown By D. Pagano et al in [“Suppression of substrate coupling in GaN high electron mobility transistors (HEMTs) by hole injection from the p-GaN gate”, Appl. Phys. Lett. 115, 203502 (2019), https://doi.org/10.1063/1.5121637]. Note that in the presence of injected holes, the 2DEG still depletes at high voltages during off-state and at very high voltages (closer to nominal breakdown) the whole of the 2DEG is depleted, but the peak of the electric field is moved away from the 2DEG towards the substrate.
IKOSHI et al., US 2011/0215379 A1 patent application describes a p-type injector placed at the surface of the semiconductor which is connected to a hole injector electrode having virtually the same potential as that of the drain. The patent describes embodiments where the hole injector is connected via metallization to the drain electrode. Other embodiments, all based on a hardwired injection electrode placed on a p-type injector region, describe placing the p-injector region inside the drift region, between the gate and the high voltage terminal (drain), or outside the drift region, in the close proximity of the high voltage terminal (drain).
KINZER et al., US 2019/0326427 A1 patent application describes a p-type injector placed between the gate and the drain terminals, which is not in direct electrical contact to the drain electrode but capacitively connected to it. The p-injector can inject holes via displacement currents flowing through the dielectric layer (acting as the dielectric of a capacitor). The device injects holes only during the transient, but not under steady-state during the on-state or off-state conditions. The amount of injection is also limited by the transient dV/dt current and the capacitance of the dielectric layer.
When the p-injector is connected to an electrode, as described in IKOSHI et al., US 2011/0215379 A1 patent application, it leads to undesirable off-state vertical and lateral leakage in the device. This is particularly the case at very high off-state voltages. Vertical leakage is known to be present at high voltages and to dominate the total leakage current. Often the breakdown in such lateral HEMTs is not set by an avalanche condition but by the high leakage current, especially at high temperatures. Vertical leakage current is increased by the presence of the hole current, particularly for high off-state voltages.
On the other hand, by capacitively connecting the p-injectors, as described in KINZER et al., US 2019/0326427 A1, the device can only inject holes during fast transient conditions and the current is limited by the capacitance value. As such, the hole current injected may be insufficient to passivate the charge, unless a very large area for the capacitive injector is used, which adds cost and leads to degradation in the overall performance.
[K. J. Chen et al., “GaN-on-Si Power Technology: Devices and Applications,” in IEEE Transactions on Electron Devices, vol. 64, no. 3, pp. 779-795, March 2017.]
[M. J. Uren et al., ““Leaky Dielectric” Model for the Suppression of Dynamic Ron in Carbon-Doped AlGaN/GaN HEMTs,” in IEEE Transactions on Electron Devices, vol. 64, no. 7, pp. 2826-2834, July 2017.]
While the configuration described above could solve the dynamic Ron, the use of a four terminal device and the control of the p-injector electrode by an additional circuit could complicate the driving circuit of the transistor and therefore be less attractive from a system perspective.
It is an object of this disclosure to provide a p-type injector in a different structure/configuration setup that operates as a hole injector both in steady-state and during the high voltage transients of the drain terminal, but it is not hardwired or physically wired to an external hole injector electrode. The aim is to prevent the depletion of the 2DEG in the on-state when a previous high voltage stress condition was applied, but limiting and/or preventing the injection of holes at larger off-state drain voltages, which could otherwise significantly increase the leakage currents and/or limit the maximum voltage allowed in the off-state (equivalent to a breakdown voltage).
Another object of this disclosure is to provide one or several layers of p-type floating structures in order to reduce or minimise the electric field around the gate structure. These layers of p-type floating structures could be contacted each by metal layers, which remain floating, not connected to a terminal. Such p-type floating structures could be provided as discontinuous regions in the third dimension. They optionally could be connected by a continuous floating metal layer. Such p-type floating structures could also be used in conjunction with metal field plates to enhance the protection of the gate structure against high electric fields.
We disclose herein a III-nitride semiconductor based heterojunction device, comprising a substrate; a III-nitride semiconductor region located longitudinally above or otherwise over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas; a first terminal operatively connected to the III-nitride semiconductor region; a second terminal operatively connected to the III-nitride semiconductor region and laterally spaced from the first terminal; a gate structure located above or otherwise longitudinally over the III-nitride semiconductor region and laterally spaced between the first and second terminals, the gate structure being a first conductivity type; a control gate terminal operatively connected to the gate structure, wherein a potential applied to the control gate terminal modulates and controls a current flow through the two-dimensional carrier gas between the first and second terminals, the at least one two-dimensional carrier gas being a second conductivity type; and at least one floating structure of the first conductivity type located over the III-nitride semiconductor region and laterally spaced from the second terminal.
Aspects and preferred features are set out in the accompanying claims.
A floating structure as described herein may comprise one or more of a region of highly doped III-V semiconductor, a floating contact layer, an injector of carriers, an electron to hole converter, a floating island, and/or a field ring.
In general, references to longitudinal and lateral positions and/or distances are references to positions and distances in the longitudinal and lateral dimensions respectively. As described herein, a floating contact layer is not electrically connected to an external electrode or biased externally during the on-state and/or off-state operation of the device.
The skilled person will understand that the at least one floating structure being laterally spaced from the second terminal includes the at least one floating structure being located between the second terminal and the control gate terminal, and/or the at least one floating structure being located away from the second terminal and away from the control gate terminal.
Optionally, a lateral spacing between the at least one injector and the second terminal may be less than a lateral spacing between the at least one injector and the first terminal. In other words, the one or more injectors may be positioned closer to the second terminal than the first terminal in the lateral dimension, or otherwise positioned proximal to or in the vicinity of the second terminal.
Alternatively, the one or more injectors may be positioned closer to the first terminal than the second terminal in the lateral dimension, or otherwise positioned proximal to or in the vicinity of the first terminal, or the gate terminal. This is particularly advantageous in the case the dynamic Ron stress is shown to have maximum effect at relatively low voltage off-state stress (e.g. 100 to 200 V) applied to the terminal. In this case the injection of holes is triggered at lower off-state voltages, protecting the device against the 2DEG depletion.
By placing the floating contact closer to the first or gate terminal, the blocking capability may be affected. Several solutions could be employed to reduce or minimise this effect. Optionally the floating contact could be placed under a metal field plate connected to either the gate terminal or the first terminal. Alternatively, the at least one injector of carriers of the first conductivity type could be positioned in the lateral dimension in the proximity of the floating contact and between the floating contact and the second terminal to facilitate the extension of the depletion region (in the blocking mode) beyond the injector layer towards the second terminal.
One or several p-type layers, similar to those used for the gate terminal or the hole injector could be positioned laterally between the floating contact and the second terminal to facilitate the extension of the depletion region (in the blocking mode), beyond the injector layer towards the second terminal. Such p-type layers may be floating and not connected to any terminal or the floating contact. The p-type layers could comprise p-type doped GaN or AlGaN. The p-type layers may be continuous or discontinuous in the third dimension (perpendicular to the lateral and vertical dimensions). In cases where the p-type layers form a discontinuous layer, the pitch between such layers in the third dimension should be small enough to ensure full depletion between them at significantly lower voltages than the breakdown voltage, but large enough, not to impact on the static on-state resistance.
The skilled person will understand the term “pitch” as used herein to refer to a separation between features. The terms “pitch” and “separation” are used herein interchangeably.
We also disclose herein a III-nitride semiconductor based heterojunction device, comprising: a substrate; a III-nitride semiconductor region formed over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas of the first conductivity type; a first terminal operatively connected to the III-nitride semiconductor region; a second terminal operatively connected to the III-nitride semiconductor region and laterally spaced from the first terminal , a gate structure with a control gate terminal placed above the III-nitride semiconductor region, and laterally spaced between the first and the second terminal where the potential applied to the control terminal modulates and controls the current flow through the two-dimensional carrier gas between the first and the second terminal; a floating contact layer operatively connected to the III-nitride semiconductor region, placed above the two dimensional carrier gas and laterally spaced away from the second terminal; at least one injector of carriers of the second type conductivity placed in the vicinity of the second terminal, and laterally spaced away from the second terminal, where the at least one said injector is connected electrically to the floating contact.
Optionally, at least one portion of the at least one two-dimensional carrier gas of the second conductivity type may be present between the second terminal and the floating contact. The portion of the at least one two-dimensional carrier gas may provide a conductive path made of carriers of the second conductivity type between the second terminal and the floating contact layer. In other words, this two-dimensional carrier gas, when present (i.e. when not depleted), provides a conductive path made of carriers of the second conductivity type between the second terminal and the floating contact (of the floating contact layer).
In some embodiments, when in operation the at least one portion of the at least one two-dimensional carrier gas type may at least partially deplete upon application of a potential to the second terminal significantly greater than a potential to the first terminal. In other words, if the difference in applied potentials is high enough, the portion of the two-dimensional carrier gas which connects the contact of the second terminal to the contact of the floating contact layer may be completely depleted.
In one embodiment, the device may be a high electron mobility transistor (HEMT); one or more of the at least one two-dimensional carrier gas may be a two-dimensional electron gas (2DEG); the first terminal may be the source terminal; and the second terminal may be the drain terminal. One or more of the at least one injector of carriers of the second type conductivity may be a hole injector, also known as a p-injector of holes. The 2DEG is formed at the interface of the heterojunction comprised for example of two layers, GaN and AlGaN. Other heterojunction layers comprising at least two material with different bandgaps are possible.
The p-injector could be formed of any III-Nitride layer such as Gallium Nitride (GaN) material or Aluminium Gallium Nitride (AlGAN) material. The p-type dopant may be, for example, Magnesium, but the skilled person will understand that other p-type dopants may be utilized in addition to or instead of a Magnesium p-type dopant. The p-injector could be formed within one of the existing layer as part of the existing heterojunction (GaN or AlGaN) by doping with p-dopant a particular region of that layer or could be grown or regrown or formed by deposition as an additional III-nitride layer. The p-injector could be for example made of a p-type doped GaN material grown selectively on top of the AlGaN layer.
In some embodiments, one or more of the floating contact layer, the first terminal and the second terminal may be configured to be operatively connected to at least a portion of the at least one two-dimensional carrier gas. Optionally, some or all of these operative connections to the two-dimensional carrier gas may be electrical connections.
The gate structure may be formed by p-type region (recessed or non-recessed, grown or regrown), as known in state-of-the-art p-Gate HEMTs and connected to the gate terminal through an ohmic or Schottky metallization. Advantageously, in this embodiment a normally-off HEMT may be realized with a positive threshold voltage. The p-type dopant may be, for example, Magnesium, but the skilled person will understand that other p-type dopants may be utilized in addition to or instead of a Magnesium p-type dopant.
The gate structure may be comprised of a p-type gate and the electrical contact to the p injector of the floating contact layer may be similar in nature to the electrical contact made by the gate control terminal to the p-type gate. The p-type gate region may be made or formed in the same process step as the p-injector region. Alternatively, the p-type gate region and the p-injector region may be formed in separate process steps. In embodiments with a plurality of p-type gate regions and/or a plurality of p-injector regions, any or all of these regions may be formed in the same or separate process steps.
Alternatively, the gate structure may be made or formed of recessed or non-recessed insulating layer above which the gate terminal is placed. This is known state-of-the-art MISFETs (Metal Insulating FETs). MISFETs may be either normally-off or normally-on transistors.
Alternatively, the gate structure could be made of a Schottky metal directly placed on the III-Nitride region. In this case a normally-on transistor (also known as a depletion mode transistor) is formed with a negative threshold voltage.
By being operatively connected to the III-nitride region, the first terminal, the second terminal and the floating contact layer make an electrical contact to the 2DEG layer or portions of the 2DEG. The contacts of the first terminal and the second terminal to the 2DEG is preferably ohmic, as known in prior art. The floating contact layer may make an ohmic contact to the two-dimensional carrier gas placed below, as well as, or alternatively, making an ohmic contact to the hole injector. Alternatively, a Schottky contact is also possible to either or both the two-dimensional carrier gas placed below and the hole injector.
As a result, in a further embodiment, the floating contact layer may have or be an ohmic contact or a Schottky contact with one or more of the at least one two-dimensional carrier gas and the at least one injector. In other words, the floating contact layer may make an ohmic contact to the two-dimensional carrier gas placed below and/or make an ohmic contact to the hole injector. Alternatively or in addition to this ohmic contact, the floating contact layer may make a Schottky contact to the two-dimensional carrier gas placed below and/or the hole injector.
The floating contact layer may comprise at least one isolated metal layer that is not operatively connected to any electrodes or terminals. More generally, the floating contact layer may be an isolated metal layer or split in several isolated metal layers which are not connected to any other electrodes or terminals. The isolated metal layer or layers may be made or formed in the same process step as one or more of the other metal layers in the process, such as those used for the source, drain or gate terminals. Alternatively, any or all of the isolated metal layers may be made or formed in a separate process step.
Alternatively, the floating contact layers may be made out of very thin, very conductive layers, such as metal silicides or other materials similar to metals.
The contact to the 2DEG of the floating contact layer may be similar to the contact made by source or drain to the 2DEG. Such contact may be established in the same process step and using the same mask as those done for the source and the drain. For example, in some embodiments a trench or a recess may be provided for the metallization of one or more of the first and second terminals and the floating contact layer, such that a direct connection is established between the one or more of the first and second terminals and the floating contact layer and their respective portions of the at least one two-dimensional carrier gas.
The contact to the p-injector of the floating contact layer may be similar to the contact made by the p-type gate (in case of p gate HEMT). Such contact may be established in the same process step and using the same mask as that done for the p-type gate.
The injector of holes need not be connected physically to any hole injector electrode (or the drain terminal), and in some embodiments may only be electrically connected to the floating contact layer that also connects part of the 2DEG. A portion of the 2DEG connects the second terminal (drain) to the floating contact. When the second terminal (drain) is positively biased with respect to the first terminal, a current made of electrons flows from the second terminal through the portion of the 2DEG that connects the second terminal (drain) to the floating contact. Note that electrons have negative charge and the actual electron carriers move in the opposite direction to the current. The floating contact layer (i.e. metallization region that is not connected to any external electrodes) converts the electron carriers into hole carriers. In other words, the current made of electrons is now converted into an equal or substantially equal current made of holes with the p-injector acting as a source of holes (i.e. provider/injector of holes). The holes are then injected back into the III-Nitride region from the p-injector region. This injection is particularly prominent during the off-state stress or during the transient while the drain voltage raises. However, at very high gate voltages, the 2DEG portion between the second terminal (drain) and the floating contact gets depleted. As a result, there are no more electrons flowing into the floating contact which would otherwise be converted into holes and re-injected as leakage into the structure. The outcome is a lower leakage current at high voltages than in the case where the p-injector would have been hardwired to a hole injector electrode (e.g., drain).
The floating contact layer (in combination with the p-injector) is configured to convert electrons from the 2DEG into holes from the p-injector and therefore acts as an electron to hole converter. In some embodiments, this means that a current of the holes injected by the at least one p-injector is approximately, substantially or exactly equal to a current of electrons between the drain terminal and the floating contact layer. The current of electrons between the drain terminal and the floating contact layer may flow via a portion of the one or more 2DEG. In general, the floating contact layer making an electrical and/or physical connection to the 2DEG and the p-injector may act as an electron to hole converter. The hole injection provided through the floating electrode converter by the p-injector (source of holes) is very effective in passivating the traps in the III-Nitride region or moving the depletion region away from the 2DEG placed between the source and the drain during the off-state stress. Thus, as a consequence of hole injection, the device does not suffer or suffer less from the dynamic Ron (or current collapse) phenomena. Moreover, as the floating electrode is not hard-wired to any external electrode or the drain terminal, at very high voltages, as the 2DEG portion between the floating electrode and the drain terminal may get depleted by the strong vertical electric field (created during the application of the high potential on the drain with respect to the substrate) and the hole injection becomes weaker and eventually stops. This has the extra benefit of reducing the overall leakage during very high voltages or increasing the breakdown of the device (defined here by when the leakage at the maximum junction temperature reaches a maximum acceptable leakage current).
In one embodiment, an injection of carriers of the first conductivity type (holes) may occur when an off-state voltage is applied to the second terminal with respect to the first terminal or during the transient while the voltage of the second terminal rises. In other words, the hole injector may inject holes during the off-state, when high voltage stress is provided (high voltage on the second terminal). This is done by converting the electrons flowing through a portion of the 2DEG into holes, with the help of the floating contact metal which acts as a carrier converter. Alternatively or in addition, an injection of carriers of the first conductivity type may occur during an on-state of the device. This means that hole injection may occur in the on-state in addition to or instead of hole injection in the off-state. This may occur as the second terminal (drain) is at slightly higher potential than the source. This injection will be however limited as the potential difference between the terminals is significantly smaller than in the off-state. This on-state hole injection may however be significant in the scenario where trapped electrons, and therefore a region of negative charge, persist in the GaN buffer after a period of off-state bias. Injection of holes can also happen during the transient regime (i.e. turn-off) as the potential on the second terminal rises in time.
Recombination of holes with electrons in the 2DEG can occur. Tunnelling or thermionic emission or a combination of both tunnelling or thermionic emission of holes through the AlGaN layer (e.g. the top layer of the III-Nitride region) and eventually the 2DEG layer to reach traps placed below the 2DEG layer can occur. The injection of holes is stronger when the 2DEG layer is depleted. 2DEG gets depleted when the potential on the second terminal rises with respect to the first terminal and it is completely depleted at very high potentials.
Optionally, the at least one injector is recessed below the surface of the III-nitride semiconductor region. In general, a recess may be provided for the p injector, so that the p injector is closer to the 2DEG or below the 2DEG to facilitate easier injection of holes. The p injector could be made of a GaN material, which could be grown or regrown after a recess is carried out.
Optionally, the operative connection of the floating contact layer to the III-nitride semiconductor region may be laterally spaced between the first and second terminals. Alternatively, the operative connection of the floating contact layer to the III-nitride semiconductor region may be laterally spaced outside of a region between the first and the second terminals. In other words, the actual contact of the floating contact layer may be placed within the drift region or outside the drift region, in the close vicinity of the drain.
In addition to or alternatively from the above, any or all of the at least one injectors may be laterally spaced between the first and second terminals. Alternatively or in addition, any or all of the at least one injectors may be laterally spaced outside of a region between the first and the second terminals. In other words, the p-injector region could be placed within the drift region (laterally spaced between the first and the second terminal) or outside the drift region (laterally spaced outside the region between the first terminal and the second terminal), in the close vicinity of the drain.
The p injector and the floating contact can be inserted anywhere laterally between the first and the second terminal or outside the region between the first and the second terminal but in the proximity of the second terminal. The injection of holes becomes higher when the electric field reaches the p-injector. The portion of the 2DEG under the p-injector may deplete first (or may be already be depleted due to the p-region), before the portion of the 2DEG which connects the contact of the floating contact layer and the drain depletes at higher voltages. This facilitates the injection of holes. Such holes are provided by the electron to hole converter (i.e. floating contact layer). The electrons which convert to holes are provided through the portion of the 2DEG which connects the contact of the floating contact layer and the drain.
Placing the p-injector and the floating contact layer laterally between the first and the second terminal closer to the first terminal, would allow the hole injection to become stronger at lower potentials applied to the drain. Several such p injector/floating contact structures can be placed laterally between the first and the second terminal. The downside of placing one or more of such p-injector/floating contact structures inside the drift region is that right under the p-injector regions the 2DEG locally is weakened and this may increase statically the on-state resistance of the device. Placing islands of such structures in the third dimension help to alleviate this effect. Alternatively, such structures may be placed just outside the drift region under the drain metallization (but as described in this invention not having a physical connection to the drain metallization). Alternatively, to alleviate the detrimental effect of placing p-injectors (or p-injector islands) inside the drift region, between source and drain on the 2DEG charge strength between the source and drain, regions of n type (grown or doped with n-type dopants) could be placed under such p-islands, to increase the strength of the 2DEG under the p-injectors or p-injector islands. Alternatively, AlGaN regions with higher mole fraction or higher thickness could be grown or regrown under such p-injector regions. The higher mole fraction or the higher the thickness of the AlGaN, the higher the strength of the 2DEG in that region.
In some embodiments, the device may include a plurality of injector islands in a third dimension, wherein said plurality of injector islands are either (i) all connected to a single floating contact layer or (ii) each connected to at least one of a plurality of single floating contact layers. This means that each of the two or more injectors may be displaced from one another along (or proximate to/in vicinity of) an axis normal to the longitudinal and lateral dimensions (i.e. an axis of the third dimension). Multiple p-injector islands in the third dimension could be displaced within the drift region or outside the drift region. In some embodiments, some or all of the p-injector islands may be proximate to or in the close vicinity of the drain terminal or otherwise closer to the drain terminal than the source terminal in the lateral dimension. Such regions may be connected to multiple floating contact layers or to a single floating contact layer.
The 2DEG portion between the drain terminal and the contact of the floating contact layer may be within the drift region or outside the drift region, in the close vicinity of the drain.
In some embodiments, the floating contact layer and a portion of the at least one two-dimensional carrier gas between the operative contact of the floating contact layer to the III-nitride semiconductor region and the second terminal are in a second active area of the device, wherein the second active area of the device is isolated from a first active area of the device comprising the first and second terminals; optionally either (i) the portion of the at least one two-dimensional carrier gas in the second active area is in contact with or a continuation of the at least one two-dimensional carrier gas in the first active area, or (ii) the portion of the at least one two-dimensional carrier gas in the second active area is isolated or otherwise separated from the at least one two-dimensional carrier gas in the first active area. In other words, one or more p type injector region may be placed outside the main active area of the device, inside its own isolated active area. In this case, one or more of the floating contact layer and the 2DEG portion between the drain terminal and the contact of the floating contact metal a may also be placed outside the main active area in another, second, isolated active area of the device. Such isolated areas could be placed, for example, at the edge of the device or alternatively in the vicinity of or even right under the drain pad or terminal.
This means that the 2DEG formed between the contact of the floating contact layer and the drain may be in contact with or a continuation of the 2DEG formed between the source and the drain terminals. Alternatively, the 2DEG formed between the contact of the floating contact layer and the drain may be separated, isolated or not in direct contact to the 2DEG formed between the source and the drain terminals.
In a further embodiment, one or more of the at least one injector, the floating contact layer and a portion of the at least one two-dimensional carrier gas between the operative contact of the floating contact layer to the III-nitride semiconductor region and the second terminal is located below a drain pad.
In one embodiment, one or more second floating p-type region may be laterally spaced between the second terminal and the contact of the floating contact layer, or otherwise formed between the first and second terminals in the lateral dimension. A second floating p-type region may be inserted laterally between the drain and the contact of the floating contact metal. This second floating region weakens the charge of the portion of the 2DEG between the drain metal and the actual contact of the floating contact layer. This in turn would facilitate the depletion of this portion of the 2DEG when a very high voltage is present on the drain terminal.
Optionally, the first terminal and the gate control terminal may be operatively connected. In other words, the first terminal and the control terminal could be connected together. In this case a diode could be formed. Such diode could be a separate device or integrated with the HEMT monolithically. The p-injector according to this invention in this case would allow to reduce or minimise the on-state loss of conductivity when the diode is forward-biased and when previously was exposed to a high voltage stress during reverse-bias.
The device described above according to this invention can operate in the low side or high side. In the high side, both the first terminal and the second terminal could operate at elevated potentials. A half bridge based on a low side device in series with a high side device according to this invention could be monolithically integrated.
According to a second aspect of the present disclosure, there is provided a III-nitride semiconductor based heterojunction device, comprising a substrate; a III-nitride semiconductor region located over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas; a first terminal operatively connected to the III-nitride semiconductor region and forming a Schottky or an ohmic contact to the at least one two-dimensional carrier gas; a second terminal operatively connected to the III-nitride semiconductor region and laterally spaced from the first terminal, forming a Schottky or an ohmic contact to the at least one two-dimensional carrier gas; at least one injector of carriers of a first conductivity type laterally spaced away from the second terminal; and a floating contact layer located over the two dimensional carrier gas and operatively connected to the at least one injector and the III-nitride semiconductor region, wherein the two-dimensional carrier gas is of a second conductivity type.
We further describe a III-nitride semiconductor based heterojunction device, comprising: a substrate; a III-nitride semiconductor region formed over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas of the first conductivity type; a first terminal operatively connected to the III-nitride semiconductor region and forming a Schottky or an ohmic contact to the two-dimensional carrier gas; a second terminal operatively connected to the III-nitride semiconductor region laterally spaced from the first terminal, forming an ohmic contact to the two-dimensional carrier gas; a floating contact layer operatively connected to the III-nitride semiconductor region placed above the two dimensional carrier gas and laterally spaced away from the second terminal, an injector of carriers of the second type conductivity placed in the vicinity of the second terminal, and laterally spaced away from the second terminal, where the at least one said injector is connected electrically to the floating contact layer.
In an embodiment of the second aspect, the device may be a diode or a high voltage diode (when the first terminal makes a Schottky to the 2DEG) or a simple resistor. The p-injector according to the invention in this case may allow the on-state loss of conductivity to be reduced or minimised when the diode is forward-biased and when previously exposed to a high voltage stress during reverse-bias.
Preferably, the starting substrate in either aspect of the invention may be silicon. However, any other substrate combining silicon with another semiconducting material compatible with state-of-the-art fabrication processes may be used. Employment of a silicon substrate facilitates etching techniques, low cost, high reproducibility, and wide availability of foundries supporting the process. Alternative substrate materials may include sapphire, silicon carbide (SiC), gallium nitride (GaN), and gallium arsenide (GaAs). Other alternatives may also be possible. Also, this disclosure covers the possibility of growing the III-nitride semiconductor region comprising the heterojunction (or heterostructure) on a first substrate suitable for the purpose and then transferring the heterostructure on a second substrate. The second substrate might be any of the previously mentioned substrates or a different one (e.g. a polydimethylsiloxane, PDMS, flexible substrate, diamond). Silicon as substrate material will be used for illustrative purposes in the embodiments hereafter and in the accompanying drawings.
The semiconductor layers forming any of the heterostructures included in this disclosure may be formed using any one of Metal-Organic Chemical Vapour Deposition (MOCVD) and Molecular Beam Epitaxy (MBE) growth techniques. Other alternative growth techniques may also be possible.
The heterojunction of the III-nitride semiconductor region may comprises: a first III-nitride semiconductor layer having a first band gap formed over the substrate; a second III-nitride semiconductor layer having a second bandgap different from the first band gap disposed on the first III-nitride semiconductor layer; and at least one two-dimensional carrier gas formed at the interface between the first and second III-nitride semiconductor layers to provide a channel.
The first III-nitride semiconductor layer may comprise any one of gallium nitride (GaN), aluminium gallium nitride (AlGaN) and indium aluminium gallium nitride (InAlGaN); and the second III-nitride semiconductor layer may comprise any one of gallium nitride (GaN), aluminium gallium nitride (AlGaN) and indium aluminium gallium nitride (InAlGaN). The p injector may comprise any one of gallium nitride (GaN), aluminium gallium nitride (AlGaN) and indium aluminium gallium nitride (InAlGaN) and may be doped with Magnesium.
The heterostructure may comprise elemental semiconductors (e.g. Si, Ge), binary compounds from Group III and Group IV (e.g. GaAs), or Group II and Group VI (e.g. ZnS), or Group III and Group V (e.g. GaN), binary alloy semiconductors (e.g. SiGe), ternary alloys (e.g. AlGaAs), quaternary alloys (e.g. InGaAsP) or even pentary materials (e.g. GaInPSbAs). Some examples of possible heterostructures thus follow: AlGaAs/GaAs, InGaP/GaAs, InP/InGaAs, InAlAs/InGaAs, SiGe/Si, AlGaAs/InGaAs, InAlAs/InGaAs and InAlN/GaN, AlGaN/GaN. Preferably, the heterostructure will be an AlGaN/GaN heterostructure comprising a two dimensional electron gas (2D carrier gas); this will be used for illustrative purposes in the embodiments hereafter and in the accompanying drawings. GaN and AlGaN technology allows transistors with high electron mobility and high saturation velocity to be designed.
The two-dimensional carrier gas may be a two-dimensional electron gas (2DEG) or a two dimensional hole gas (2DHG). The 2DEG has a considerably higher carrier mobility and is therefore preferred to a 2DHG.
The first terminal, the second terminal and the floating contact layer may be in contact with the III-nitride semiconductor region. Preferably the contact will be at the surface of the semiconductor and in physical touch with the second III-nitride semiconductor layer of the III-nitride semiconductor region. Alternatively, a trench or a recess could be provided such that a direct physical touch of either of the first terminal, second terminal or the floating contact layer could be established to their respective portions of the 2DEG. This would facilitate a better ohmic contact to such portions of the 2DEG layer. As already mentioned, the p-injector could be recessed as well to provide better injection access to the traps under the main 2DEG.
The device may further comprise a transition layer formed between the substrate and the III-nitride semiconductor region. The transition layer may be present between the heterostructure and the bulk substrate to reduce or minimise the lattice mismatch or accommodate the mechanical stress in the heterostructure. The transition layer may comprise a nucleation layer. The nucleation layer may be made of Aluminium Nitride (AlN) or any other suitable material, and may be placed on the substrate. The nucleation layer may form the first sub-layer of a transition layer made of different sub-layers containing the same materials as the first and second semiconductor layers of the heterostructure device (for example AlGaN or GaN materials). This helps to release the mechanical stress and accommodate the lattice mismatch between the substrate (for example silicon) and the heterostructure formed on top.
In embodiments of the disclosure where the substrate is silicon and the heterostructure is based on III-Nitride compounds such as GaN, the transition layer may comprise a single layer of AlxGaN1−xN (x being the aluminium mole fraction of the compound) or any composition of AlxGaN1−xN/AlyGaN1−yN (y being the aluminium mole fraction of the compound) creating a multilayer stack, also known as superlattice.
The device may be a normally-off device, a normally-on device (threshold voltage is negative) or a diode.
According to a further aspect of the present disclosure, there is provided a method of manufacturing a III-nitride semiconductor based heterojunction device, the method comprising forming a substrate; forming a III-nitride semiconductor region located over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas; forming a first terminal operatively connected to the III-nitride semiconductor region; forming a second terminal laterally spaced from the first terminal in a first dimension and operatively connected to the III-nitride semiconductor region; forming a gate terminal on a gate structure, said gate structure being positioned above the III-nitride semiconductor region; forming a floating contact layer over the at least one two dimensional carrier gas, wherein the floating contact layer is operatively connected to the III-nitride semiconductor region and laterally spaced away from the second terminal; forming at least one injector of carriers of a first conductivity type laterally spaced away from the second terminal; and forming at least one operative connection between the at least one said injector and the floating contact layer.
We further describe a method of manufacturing a III-nitride semiconductor based heterojunction device, the method comprising: forming a substrate; forming a III-nitride semiconductor region formed over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas of the second conductivity type; forming a first terminal operatively connected to the III-nitride semiconductor region; forming a second terminal laterally spaced from the first terminal in a first dimension and operatively connected to the III-nitride semiconductor region; forming a gate terminal on a gate structure placed above the III-nitride semiconductor region; forming a floating contact layer operatively connected to the III-nitride semiconductor region, placed above the two dimensional carrier gas and laterally spaced away from the second terminal; forming an injector of carriers of the second type conductivity placed in the vicinity of the second terminal, and laterally spaced away from the second terminal, where the at least one said injector is connected electrically to the floating contact layer.
Optionally, the gate structure is formed of p type material and the p injector is formed in the same process step as the gate structure.
Optionally, the metal layer or layers used for the floating contact layer can be done in the same process step as one of the other metal layers in the process, such as those used for the first terminal, second terminal or the gate control terminal.
We also describe herein a III-nitride semiconductor based heterojunction device, comprising a substrate; a III-nitride semiconductor region located longitudinally above or otherwise over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas; a first terminal operatively connected to the III-nitride semiconductor region; a second terminal operatively connected to the III-nitride semiconductor region and laterally spaced from the first terminal; a gate structure located above or otherwise longitudinally over the III-nitride semiconductor region and laterally spaced between the first and second terminals, the gate structure being a first conductivity type; a control gate terminal operatively connected to the gate structure, wherein a potential applied to the control gate terminal modulates and controls a current flow through the two-dimensional carrier gas between the first and second terminals, the at least one two-dimensional carrier gas being a second conductivity type; at least one other floating structure of first conductivity type over the III-nitride semiconductor region and laterally spaced between the gate and second terminals, in close proximity to the gate terminal.
In a GaN HEMT device based on a 2DEG layer, the floating structure of the first conductivity type is of p-type and its composition and dimensions may be the same as those of the p-type injector described above. The floating structure of the first conductivity type can also be further connected to a floating metal, as the p-type injector as described above. However if the floating p-type structure is not connected via a metal and contact layer to the 2DEG, it does not inject holes and its sole purpose if that of shaping the electric field in such a way to reduce the electric field peak around the gate structure. In other words, when used as a field ring and not as a hole injector, a connection between the 2DEG and the floating contact is omitted or absent.
The device may comprise a plurality of floating structures of the first conductivity type laterally displaced from each other in the third dimension. The floating structures may be p-type structures, and the plurality of floating structures may comprise a plurality of floating regions of first conductivity type’, wherein said plurality of injector islands are either (i) all connected to a single floating metal layer or (ii) there are rows of such islands each connected to at least one of a plurality of single floating metal layers. The distance between such islands should be optimised. A small distance provides good protection against the electric field. Nevertheless, the presence of the p type structures on top of the AlGaN layer tend to weaken the conductivity of the 2DEG directly underneath such regions. Therefore to reduce or minimise an increase in the static on-state resistance, the distance between the islands should be relatively high. Such pitch between the islands could be between 0.2-5 microns. The length of the islands in the third dimension could be from 0.4 to 20 microns. Different rows of islands could have different pitches between the islands. For example the row closest to the gate terminal could have the smallest pitch to protect the drain edge of the gate structure. The second and subsequent rows could have an increase pitch to alleviate the increase in the static on-state resistance.
This p-type floating islands may be displaced from one another along (or proximate to/in vicinity of) an axis normal to the longitudinal and lateral dimensions (i.e. an axis of the third dimension). Multiple p-type floating islands in the third dimension could be displaced within the drift region, in the proximity of the gate structure. Such regions may be connected to multiple floating metal layers. Existing metal layers such as the metal layer used for the gate (but not connected to gate) or the metal used for the source (but not connected to source) could be used for this.
The floating p-type layers may comprise the same material and be formed in the same process step as the p-type structure for the gate or p-type structure for the hole injector.
Optionally the floating p-type structures could be formed under the source or gate metal field plates to provide further protection to the gate structure.
Alternatively, the floating p-type structures may be formed under floating metal structures. Such floating metal structures could be optionally connected to the floating p-type structures.
According to a further aspect of the present disclosure, there is provided a method of manufacturing a III-nitride semiconductor based heterojunction device, the method comprising forming a substrate; forming a III-nitride semiconductor region located over the substrate, wherein the III-nitride semiconductor region comprises a heterojunction comprising at least one two-dimensional carrier gas; forming a first terminal operatively connected to the III-nitride semiconductor region; forming a second terminal laterally spaced from the first terminal in a first dimension and operatively connected to the III-nitride semiconductor region; forming a gate terminal on a gate structure, said gate structure being positioned above the III-nitride semiconductor region; forming at least one floating structure of the first conductivity type over the III-nitride semiconductor region and laterally spaced between the gate and second terminals, in close proximity to the gate terminal.
Optionally, the p-type floating structure is formed of p type material and is formed in the same process step as the gate structure.
Optionally, the floating metal layer or layers used for connecting at least one of the floating p-type structures may be formed in the same process step as one of the other metal layers in the process, such as those metals from which the first terminal, second terminal and/or the gate control terminal are formed.
The present disclosure will be understood more fully from the accompanying drawings, which however, should not be taken to limit the disclosure to the specific embodiments shown, but are for explanation and understanding only.
A fourth region of highly p-doped III-V semiconductor 11 is formed in contact with the third region 1. This has the function of reducing the 2DEG carrier concentration when the device is unbiased and is pGaN material (preferably made of GaN and doped with Magnesium) in this embodiment. A gate control terminal 10 is configured over the fourth region 11 in order to control the carrier density of the 2DEG at the interface of the second 2 and third region 1. A high voltage drain terminal 9 is arranged in physical contact with the third region 1. The high voltage drain terminal forms an ohmic contact to the 2DEG. A low voltage source terminal 8 is also arranged in physical contact with the third region 1 and also forms an ohmic contact to the 2DEG.
A first portion of surface passivation dielectric 7 is formed on top of the fourth region 11 and between the drain terminal 9 and source terminal 8. A p-injector region 101, preferably made using the same material as the pGaN gate 11, is also formed in contact with third region 1 possibly in the same step as fourth region 11. A floating contact layer 102 is formed in contact with both the p-injector region 101 and third region 1. A second portion of surface passivation dielectric 7 is formed on top of the p-injector region 101, and may be separate from or connected with the first portion of surface passivation dielectric. A layer of SiO2 or nitride or another oxide-nitride playing the role of a passivation layer 6 is formed above the surface passivation dielectric 7 and source and drain terminals 8, 9.
The injector of holes 101 is not connected physically to any hole injector electrode (or the drain terminal 9) but only physically (electrically) connected to the floating contact 102 that also connects a part or a portion of the 2DEG. A portion of the 2DEG connects the second terminal (drain) 9 to the floating contact 102. This is illustrated in
Note that the substrate terminal 5 in this example is at the same potential as the source terminal 8, that is an electrical connection might exist between the two either at package level or possibly circuit level. This is commonly the case in discrete GaN-on-Si or GaN-on-SiC devices. It might however not be the case in GaN-on-Sapphire case or for example, a device on the high side of a monolithically integrated GaN-on-Si half bridge.
When the drain terminal 9 is positively biased with respect to the source terminal 8, a current flows from the drain terminal 9 through the portion of the 2DEG that connects the drain terminal 9 to the floating contact layer 102. This current is formed by a flow of electrons from the floating contact layer 102 to the drain terminal 9 as illustrated in
However, at very high gate voltages, the 2DEG portion between the drain terminal 9 to the floating contact 102 gets depleted as illustrated in
Hole injection can happen also in the on-state, as the drain terminal 9 is at slightly higher potential than the source 8. This injection is generally limited as the potential difference between the terminals is significantly smaller than in the off-state. This on-state hole injection may however be significant in the scenario where trapped electrons, and therefore a region of negative charge 120, persist in the GaN buffer 2 after a period of off-state bias. This is illustrated in
The good electrical connection 115 via the 2DEG between the drain terminal 9 and the floating contact 102 at zero bias (or moderate levels of off-state bias) is shown clearly in
The weakening or non-existent electrical connection 116 via the 2DEG between the drain terminal 9 and the floating contact 102 at high off-state bias is clearly shown in
Alternatively, a rectifier where the gate and source terminal of
The above-described effect may be related to the electric field peak observed at edge of the gate terminal (on the drain side of the gate).
In
It may be useful to optimise the device described in this invention, so that the p-injector region 101 starts emitting a significant number of holes at off-state bias voltages around 100V-200V or where the Dynamic Ron effect is most severe in other examples. The off-state voltage at which significant hole injection from p-injector region 101 is observed might be adjusted by decreasing the separation of the p-injector region 101 and the floating contact layer 102 from the gate terminal 10, 11. The smaller the distance between the gate terminal and the floating contact layer 102 the lower the off-state voltage at which the hole injection from the p-injector region might be observed.
An example where the distance between the gate terminal 11 and floating contact layer 102 is reduced, compared to the embodiment in
Additionally, a corresponding increase in the distance between the floating contact layer 102 and the drain contact layer 9 (with the gate-drain distance remaining unchanged in
The formation of the floating contact layer 104 and p-injector region 101 as islands rather than a continuous strip can have two advantages:
An increase in the distance between the floating contact layer 104 and the drain contact layer 9 (with the gate-drain distance remaining unchanged in
In
In
In an additional embodiment, the floating contact layer 102 and the p-injector region 101 are separated along the gate-drain axis and connected with metal interconnect 125 as illustrated in
Similarly, in other embodiments, which feature p-injector region islands (such as the embodiment illustrated in
In some embodiments, the distance between the floating contact 102, 104 and gate terminal 10 may be smaller than the distance between the p-injector region island 101 and the gate terminal 10.
In other embodiments, the distance between the p-injector region island 101 and gate terminal 10 may be smaller than the distance between the floating contact 102, 104 and the gate terminal 10.
In other embodiments, the floating contact 102, 104 may be placed between the gate terminal 10 and the p-injector region island 101 (along the z-axis).
In other embodiments, the p-injector region island 101 may be placed between the gate terminal 10 and the floating contact 102, 104 (along the z-axis).
One example of the embodiments described here is illustrated in
In the example illustrated in
In this example (c)>(b)>(a). These dimensions may be an important design parameter which shapes the electrostatic potential in the device in off-state conditions and therefore can be optimized in different device designs.
In addition to the embodiment in
Generally speaking, in this disclosure, the first conductivity type refers to p-type conductivity and the second conductivity type generally refers to the n-type conductivity. However, it will be understood that the conductivity type can be readily reversed. The skilled person will understand that in the preceding description and appended claims, positional terms such as ‘top’, ‘above’, ‘overlap’, ‘under’, ‘lateral’, etc. are made with reference to conceptual illustrations of a device, such as those showing standard cross-sectional perspectives and those shown in the appended drawings. These terms are used for ease of reference but are not intended to be of limiting nature. These terms are therefore to be understood as referring to a device when in an orientation as shown in the accompanying drawings.
Although the disclosure has been described in terms of preferred embodiments as set forth above, it should be understood that these embodiments are illustrative only and that the claims are not limited to those embodiments. Those skilled in the art will be able to make modifications and alternatives in view of the disclosure which are contemplated as falling within the scope of the appended claims. Each feature disclosed or illustrated in the present specification may be incorporated in the disclosure, whether alone or in any appropriate combination with any other feature disclosed or illustrated herein.
Many other effective alternatives will occur to the person skilled in the art. It will be understood that the disclosure is not limited to the described embodiments, but encompasses all the modifications which fall within the spirit and scope of the disclosure.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2021/050561 | 1/13/2021 | WO |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 16740996 | Jan 2020 | US |
| Child | 17792302 | US |
