PEAK TRANSCONDUCTANCE WHILE MAINTAINING ON-STATE RESISTANCE

AND RELATED DEVICES

CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority from U.S. Patent Application Serial No. 18/310,684, filed May 2, 2023, which is a continuation-in-part of U.S. Patent Application Serial No. 17/834,144, filed June 7, 2022, with the United States Patent and Trademark Office, the disclosures of which are incorporated by reference herein in their entireties.

FIELD

[0002] The inventive concepts described herein relate to semiconductor devices, and more particularly, to transistor devices and related fabrication methods.

BACKGROUND

[0003] Power semiconductor devices are widely used to carry large currents, support high voltages and/or operate at high frequencies such as radio frequencies. A wide variety of power semiconductor devices are known in the art including, for example, power switching devices and power amplifiers. Many power semiconductor devices are implemented using various types of field effect transistor (FET) devices including MOSFETs (metal-oxide semiconductor field-effect transistors), DMOS (double-diffused metal-oxide semiconductor) transistors, HEMTs (high electron mobility transistors), MESFETs (metal-semiconductor field-effect transistors), LDMOS (laterally-diffused metal-oxide semiconductor) transistors, etc.

[0004] Modern power semiconductor devices are generally fabricated from wide bandgap semiconductor materials (i.e., having a band-gap greater than 1.40 eV). For example, power HEMTs may be fabricated from gallium nitride (GaN) or other Group III nitride based material systems that are formed on a silicon carbide (SiC) substrate. As used herein, the term “Group III nitride” refers to those semiconducting compounds formed between nitrogen and the elements in Group III of the periodic table, usually aluminum (Al), gallium (Ga), and/or indium (In). These compounds have empirical formulas in which one mole of nitrogen is combined with a total of one mole of the Group III elements. For high power, high temperature, and/or high frequency applications, devices formed in wide bandgap semiconductor materials such as silicon carbide (2.996 eV bandgap for alpha silicon carbide at room temperature) and the Group III nitrides (e.g., 3.36 eV bandgap for gallium nitride at room temperature) may provide higher electric field breakdown strengths and higher electron saturation velocities as compared to gallium arsenide (GaAs) and silicon (Si) based devices. [0005] In order to increase the output power and current handling capabilities, power semiconductor devices may be implemented in a “unit cell” configuration in which a large number of individual unit cell transistor structures of the active region are electrically connected (e.g., in parallel) to function as a single power semiconductor device. In high power applications, such a power semiconductor device may include thousands or tens of thousands of unit cells implemented in a single chip or “die.” A die or chip may include a small block of semiconducting material or other substrate in which electronic circuit elements are fabricated.

[0006] Field effect transistors such as HEMTs and MOSFETs may be classified into depletion mode and enhancement mode types, corresponding to whether the transistor is in an ON-state or an OFF-state at a gate-source voltage of zero. In enhancement mode devices, the devices are OFF at zero gate-source voltage, whereas in depletion mode devices, the device is ON at zero gate-source voltage. Conventional high performance Group III nitride-based HEMTs may typically be implemented as depletion mode (normally-on) devices, in that they are conductive at a gate-source bias of zero due to the polarization-induced charge at the interface of the barrier and channel layers of the device.

[0007] When a HEMT device is in an ON-state, a two-dimensional electron gas (2DEG) is formed at the heterojunction of two semiconductor materials with different bandgap energies, where the smaller bandgap material has a higher electron affinity. The 2DEG is an accumulation layer in the undoped (“unintentionally doped”) smaller bandgap material and can contain a very high sheet electron concentration. Additionally, electrons that originate in the wider-bandgap semiconductor material transfer to the 2DEG layer, allowing high electron mobility due to reduced ionized impurity scattering. This combination of high carrier concentration and high carrier mobility can give the HEMT a very large transconductance (which may refer to the relationship between output current and input voltage) and may provide a strong performance advantage over MOSFETs for high-frequency applications. [0008] High electron mobility transistors fabricated in Group Ill-nitride based material systems also have the potential to generate large amounts of radio frequency (RF) power because of the combination of material characteristics that includes the aforementioned high breakdown fields, wide bandgaps, large conduction band offset, and/or high saturated electron drift velocity. As such, Group III-Nitride based HEMTs may be promising candidates for high power RF applications, as well as for low frequency high power switching applications. However, problems may arise as device integration increases. For example, a HEMT may be subject to short channel effects (SCE) as gate length is reduced.

SUMMARY

[0009] According to some embodiments, a transistor device includes a semiconductor structure comprising a channel layer and a barrier layer, source and drain contacts on the semiconductor structure, and a conductive element in a recess in the barrier layer between the source and drain contacts. The barrier layer has a first thickness adjacent the source or drain contact, a second thickness at floor of the recess between the conductive element and the channel layer, and the first thickness is about 1.2 times to 4 times greater than the second thickness.

[0010] In some embodiments, the conductive element is a gate, and first thickness is about 1.5 times to 4 times greater than the second thickness.

[0011] In some embodiments, the first thickness is about 150 A to 300 A, and the second thickness is about 30 A to 150 A.

[0012] In some embodiments, a gate is provided on the semiconductor structure between the source and drain contacts. The recess in the barrier layer is between the gate and the source or drain contact, and the conductive element is a field plate.

[0013] In some embodiments, the first thickness is about 150 A to 300 A, and the second thickness is about 20 A to 150 A.

[0014] In some embodiments, the recess is a second recess in the barrier layer, and wherein the gate is in a first recess in the barrier layer between the second recess and the source or drain contact.

[0015] In some embodiments, the second thickness of the barrier layer at the floor of the second recess between the field plate and channel layer is less than or equal to a third thickness of the barrier layer at a floor of the first recess between the gate and the channel layer.

[0016] In some embodiments, the second thickness of the barrier layer is less than a third thickness of the barrier layer at corner portions of the recess adjacent opposing sidewalls thereof.

[0017] In some embodiments, the barrier layer and the channel layer respectively comprise gallium-based materials, and the barrier layer has a higher bandgap than that of the channel layer. [0018] In some embodiments, the second thickness is greater than about 20 Angstroms (A). [0019] In some embodiments, a surface roughness along a floor of the recess is less than about 3 Angstroms (A).

[0020] According to some embodiments, a transistor device includes a semiconductor structure comprising a channel layer and a barrier layer, and source and drain contacts and a gate therebetween on the semiconductor structure. The barrier layer comprises at least one recess therein between the source and drain contacts. The at least one recess comprises opposing sidewalls and a floor therebetween, and a first thickness of the barrier layer at edges of the at least one recess adjacent the opposing sidewalls is greater than a second thickness of the barrier layer at the floor of the recess therebetween.

[0021] In some embodiments, the edges comprise inclined surfaces, and the floor comprises a substantially planar surface that is between the inclined surfaces.

[0022] In some embodiments, the opposing sidewalls respectively comprise a first portion defining an angle of about 85 degrees to about 95 degrees relative to the floor, and a second portion comprising the inclined surfaces having an angle of about 40 degrees to 50 degrees relative to the floor.

[0023] In some embodiments, the at least one recess comprises a gate recess having the gate therein. A thickness of a portion of the barrier layer extending between the source or drain contact and the gate is about 1.5 times to 4 times greater than the second thickness.

[0024] In some embodiments, at least one recess comprises a field plate recess between the gate and the source or drain contact. A field plate is provided in the field plate recess, and a thickness of a portion of the barrier layer extending between the source or drain contact and the field plate is about 1.2 times to 4 times greater than the second thickness.

[0025] In some embodiments, the at least one recess further comprises a gate recess in the barrier layer between the field plate recess and the source or drain contact, and the gate is in the gate recess.

[0026] In some embodiments, the second thickness at the floor of the field plate recess between the field plate and channel layer is less than or equal to a thickness of the barrier layer at a floor of the gate recess between the gate and the channel layer.

[0027] In some embodiments, the barrier layer and the channel layer respectively comprise gallium-based materials, and the barrier layer has a higher bandgap than that of the channel layer.

[0028] In some embodiments, the second thickness is greater than about 20 Angstroms (A). [0029] According to some embodiments, a transistor device includes a semiconductor structure comprising a channel layer and a barrier layer, source and drain contacts on the semiconductor structure, a gate in a first recess in the barrier layer between the source and drain contacts, and a field plate in a second recess in the barrier layer between the gate and the source or drain contact. A surface roughness along a floor of each of the first recess and the second recess is less than about 3 Angstroms (A).

[0030] In some embodiments, the barrier layer has a first thickness at the floor of the first recess between the gate and the channel layer, a second thickness at the floor of the second recess between the field plate and the channel layer, and a third thickness adjacent the source or drain contact, and the third thickness is about 1.2 times to 4 times greater than the first thickness or the second thickness.

[0031] In some embodiments, the second thickness of the barrier layer at the floor of the second recess between the field plate and channel layer is less than or equal to the first thickness of the barrier layer at the floor of the first recess between the gate and the channel layer.

[0032] According to some embodiments, a method of fabricating a transistor device includes providing a semiconductor structure comprising a channel layer and a barrier layer, performing a looped recess process to define a recess in the barrier layer, and forming a conductive element in the recess.

[0033] In some embodiments, the looped recess process comprises a plurality of loops that are configured to remove, layer-by-layer, a portion of the barrier layer to define the recess therein.

[0034] In some embodiments, the looped recess process is a plasma etch process, and each of the loops is configured to remove a monolayer of the barrier layer.

[0035] In some embodiments, each of the loops comprises an adsorption process and a desorption process with a purge process therebetween.

[0036] In some embodiments, the adsorption process is chlorine-based, the desorption process is argon-based, and the purge process is helium- or nitrogen-based.

[0037] In some embodiments, the channel layer and the barrier layer comprise a gallium- based material, and the barrier layer has a higher bandgap than that of the channel layer. [0038] In some embodiments, the plasma etch process is performed using an inductively- coupled plasma tool with a peak-to-peak voltage of about 70 V to 110 V. [0039] In some embodiments, prior to performing the looped recess process, the method further includes forming a passivation layer on the semiconductor structure, and forming an opening in the passivation layer that exposes a surface of the barrier layer.

[0040] In some embodiments, the method further includes forming spacers on opposing sidewalls of the opening in the passivation layer prior to performing the looped recess process.

[0041] In some embodiments, the method further includes forming spacers on opposing sidewalls of the opening in the passivation layer after performing the looped recess process. [0042] In some embodiments, responsive to performing the looped recess process, a surface roughness along a floor of the recess is less than about 3 A.

[0043] In some embodiments, the method further includes forming source and drain contacts on the semiconductor structure, where the recess is between the source and drain contacts, and the conductive element is a gate.

[0044] In some embodiments, the method further includes forming source and drain contacts on the semiconductor structure, and forming a gate on the semiconductor structure between the source and drain contacts, where the recess is between the gate and the source or drain contact, and the conductive element is a field plate.

[0045] In some embodiments, the looped recess process is a second looped recess process and the recess is a second recess in the barrier layer, and the method further includes performing a first looped recess process to define a first recess in the barrier layer, and forming the gate in the first recess in the barrier layer.

[0046] In some embodiments, responsive to performing the first and second looped recess processes, a thickness at a floor of the second recess between the field plate and channel layer is less than or equal to a thickness at a floor of the first recess between the gate and the channel layer.

[0047] In some embodiments, responsive to performing the looped recess process, a first thickness of the barrier layer at corner portions of the recess adjacent opposing sidewalls thereof is greater than a second thickness of the barrier layer at a floor of the recess.

[0048] Other devices, apparatus, and/or methods according to some embodiments will become apparent to one with skill in the art upon review of the following drawings and detailed description. It is intended that all such additional embodiments, in addition to any and all combinations of the above embodiments, be included within this description, be within the scope of the invention, and be protected by the accompanying claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] FIG. 1 A is a schematic cross-sectional view illustrating a transistor device according to some embodiments of the present invention.

[0050] FIG. IB is an enlarged cross-sectional view illustrating the gate recess of FIG. 1 A.

[0051] FIG. 1C is an electron microscope image of a cross-section of a gate recess of a transistor device according to some embodiments of the present invention.

[0052] FIG. ID is a schematic cross-sectional view illustrating a transistor device according to further embodiments of the present invention.

[0053] FIG. IE is a schematic cross-sectional view illustrating a transistor device according to still further embodiments of the present invention.

[0054] FIG. IF is an enlarged cross-sectional view illustrating the gate recess and the field plate recess of FIG. IE.

[0055] FIGS. 2A and 2B are schematic diagrams illustrating methods of forming a gate recess for a transistor device using a looped recess process according to some embodiments of the present invention.

[0056] FIG. 3 is a graph illustrating example parameters for a looped recess process according to some embodiments of the present invention.

[0057] FIGS. 4 A, 4B, 4C, and 4D are schematic cross-sectional views illustrating methods of forming a gate recess for a transistor device using a looped recess process according to some embodiments of the present invention.

[0058] FIGS. 5A, 5B, 5C, and 5D are schematic cross-sectional views illustrating methods of forming a gate recess for a transistor device using a looped recess process according to further embodiments of the present invention.

[0059] FIGS. 6 A, 6B, 6C, and 6D are schematic cross-sectional views illustrating methods of forming a gate recess for a transistor device using a looped recess process according to yet further embodiments of the present invention.

[0060] FIG. 7 is a schematic plan view of a Group III nitride-based transistor die according to embodiments of the present invention that illustrates metallization on a surface of the semiconductor structure thereof.

[0061] FIGS. 8A, 8B, and 8C are schematic cross-sectional views illustrating example packages including transistor devices according to embodiments of the present invention to provide packaged transistor amplifiers.

DETAILED DESCRIPTION OF EMBODIMENTS [0062] Power transistor devices as described herein may include both discrete devices that are coupled with other circuitry such as, for example, impedance matching networks, or as MMIC devices (e.g., a multi-stage HEMT amplifier with built-in impedance matching networks). FIGS. 1 A, ID, and IE are schematic cross-sectional views illustrating a transistor devices according to some embodiments of the present invention, illustrated by way of example with reference to unit cell transistor structures 100, 100’, 100” (also referred to herein as a transistor structure or transistor cell; collectively 100) of a transistor device, such as a HEMT. In particular, FIG. 1 A illustrates a transistor structure 100 including a gate 110 in a recess 126R1 in barrier layer 126, FIG. ID illustrates a transistor structure 100’ including a field plate 120 in a recess 126R2 in barrier layer 126, and FIG. IE illustrates a transistor structure 100” including the gate 110 and the field plate 120 in respective recesses 126R1 and 126R2 in barrier layer 126.

[0063] As shown in FIGS. 1 A, ID, and IE, a transistor structure 100 is formed on a substrate 122 such as, for example, a silicon carbide substrate. Hundreds or thousands of unit cell transistor structures 100 may be formed on the semiconductor substrate 122, and may be electrically connected (e.g., in parallel) to provide the HEMT device. The substrate 122 may be a semi-insulating silicon carbide substrate that may be, for example, the 4H polytype of silicon carbide. Other silicon carbide candidate polytypes may include the 3C, 6H, and 15R polytypes. Although silicon carbide may be used as a substrate material, embodiments of the present disclosure may utilize any suitable substrate, such as sapphire (AI2O3), aluminum nitride (AIN), aluminum gallium nitride (AlGaN), gallium nitride (GaN), silicon (Si), GaAs, LGO, zinc oxide (ZnO), LAO, indium phosphide (InP), and the like. The substrate 122 may be a SiC wafer, and the HEMT device may be formed, at least in part, via wafer-level processing. The wafer may then be diced or otherwise singulated to provide a multiple dies, where each die includes a plurality of the unit cell transistor structures 100.

[0064] The transistor structure 100 includes a channel layer 124 on the substrate 122, and a barrier layer 126 on the channel layer 124 opposite the substrate 122. Source and drain electrodes (also referred to herein as source and drain contacts) 115 and 105 are formed laterally (e.g., along the X-direction) spaced apart from each other. The source contact 115 and the drain contact 105 may form ohmic contact to the barrier layer 126 or to the channel layer 124.

[0065] One or more insulator layers (for example, one or more passivation layers) 150 are formed on the barrier layer 126, and a gate contact (or simply “gate”) 110 is formed on the barrier layer 126 between the source and drain contacts 115 and 105. The gate 110 may be formed closer to the source contact 115, such that the gate-to-source length LGS may be smaller than the gate-to-drain length LGD in some embodiments. Depending on configuration, one or more of the insulator layers 150 may be formed before and/or after formation of the gate 110. As further illustrated in FIGS. ID and IE, an additional insulator layer 151 (for example, an inter-field plate passivation layer IFPP) may be formed on the gate 110 and the passivation layer 150, and a field plate 120 may be formed on the IFPP 151.

[0066] The channel layer 124 may have a bandgap that is less than the bandgap of the barrier layer 126 and the channel layer 124 may also have a larger electron affinity than the barrier layer 126. The channel layer 124 and barrier layer 126 may together define a semiconductor structure 190, with the source contact 115, the drain contact 105, and the gate 110 formed on the semiconductor structure 190. In the illustrated examples, the semiconductor structure 190 may be a semiconductor layer structure including one or more layers formed by epitaxial growth, and thus include one or more epitaxial layers 124, 126. Techniques for epitaxial growth of Group III nitrides have been described in, for example, U.S. Pat. No. 5,210,051, U.S. Pat. No. 5,393,993, and U.S. Pat. No. 5,523,589, the disclosures of which are also incorporated by reference herein in their entireties.

[0067] In the illustrated HEMT device 100, the semiconductor layer structure 190 may be Group-Ill nitride based, although other material systems can also be used. As noted above, Group III nitrides may refer to semiconductor compounds formed between nitrogen and the elements in the Group III of the periodic table, such as aluminum (Al), gallium (Ga), and/or indium (In), and may form binary (e.g., GaN), ternary (e.g., AlGaN, AllnN), and quaternary (e.g., AlInGaN) compounds. Accordingly, formulas such as AkGai-xN, where 0 < x < 1, may be used to describe these compounds. One or both of the channel layer 124 and the barrier layer 126 may include sub-layers including doped or undoped (i.e., “unintentionally doped”) layers of Group Ill-nitride materials, including material compositions which may be stepwise or continuously graded. For example, the channel layer 124 may be a multi-layer structure, such as a superlattice or combinations of GaN, AlGaN or the like. The channel layer 124 may be under compressive strain in some embodiments. The barrier layer 126 may comprise a single layer or may be a multi-layer structure. In particular embodiments of the present disclosure, the barrier layer 126 may be thick enough and may have a high enough aluminum (Al) composition and doping to induce a significant carrier concentration at the interface between the channel layer 124 and the barrier layer 126 through polarization effects when the barrier layer 126 is buried under ohmic contact metal. [0068] While semiconductor structure 190 is shown with reference to one or more epitaxial layers 124, 126 for purposes of illustration, the semiconductor structure 190 may include additional layers/structures/elements such as isolation layer(s), buffer and/or nucleation layer(s) on or between substrate 122 and the one or more epitaxial layers 124, and/or a cap layer on an upper surface of the epitaxial layer 126. For example, an AIN buffer layer may be formed on the upper surface of the substrate 122 to provide an appropriate crystal structure transition between the silicon carbide substrate 122 and the remainder of the layers of the semiconductor structure 190. Additionally, strain balancing transition layer(s) may also and/or alternatively be provided. The optional buffer/nucleation/transition layers, as well as the channel layer 124 and/or the barrier layer 126, may be deposited, for example, by metalorganic chemical vapor deposition (MOCVD), molecular beam epitaxy (MBE), and/or hydride vapor phase epitaxy (HVPE .

[0069] The material of the gate 110 may be chosen based on the composition of the semiconductor structure 190, and may, in some embodiments, be a Schottky contact. Some materials capable of making a Schottky contact to a Group III nitride based semiconductor material that may be used as the gate 110 may include, for example, nickel (Ni), platinum (Pt), nickel silicide (NiSix), copper (Cu), palladium (Pd), chromium (Cr), tungsten (W) and/or tungsten silicon nitride (WSiN).

[0070] The source contact 115 and/or the drain contact 105 may include a metal that can form an ohmic contact to the semiconductor material of the structure 190. For example, a conductive metal material may be deposited and annealed (e.g., at a temperature of about 600° C to 1050° C) to form the ohmic contacts. Suitable metals may include refractory metals, such as Ti, W, titanium tungsten (TiW), silicon (Si), titanium tungsten nitride (TiWN), tungsten silicide (WSi), rhenium (Re), Niobium (Nb), Ni, gold (Au), aluminum (Al), tantalum (Ta), molybdenum (Mo), NiSix, titanium silicide (Ti Si), titanium nitride (TiN), WSiN, Pt and the like. Thus, the source contact 115 and/or the drain contact 105 may contain an ohmic contact portion in direct contact with the layer 124 and/or 126. In some embodiments, the source contact 115 and/or the drain contact 105 may be formed of a plurality of layers to form an ohmic contact that may be provided as described, for example, in commonly assigned U.S. Pat. No. 8,563,372 and U.S. Pat. No. 9,214,352, the disclosures of which are hereby incorporated herein in their entirety by reference.

[0071] In operation, a 2DEG layer 40 may be formed at a junction between the channel layer 124 and the barrier layer 126 when the HEMT device is biased to be in its conducting or “on” state. The 2DEG layer 40 acts as a highly conductive channel that allows current to flow between the source and drain regions that are beneath the source contact 115 and the drain contact 105, respectively. In particular, the channel layer 124 and the barrier layer 126 of the semiconductor structure 190 may be formed of materials having different bandgaps, such that a heterojunction is defined at an interface between the channel layer 124 and the barrier layer 126. The 2DEG conduction channel 40 can be induced at the heterointerface between the channel layer 124 and the barrier layer 126. The channel layer 124, 2DEG conduction channel 40, and barrier layer 126 can generally form the active region of the HEMT device. It should be noted that while described herein primarily with reference to fabrication and structures of HEMT devices, the elements and concepts of embodiments described herein can be applied to many different types of transistor structures.

[0072] In some embodiments, the transistor device 100 includes a gallium-based semiconductor structure 190 that is formed on the substrate 122 (e.g., a silicon carbide substrate). As used herein, the term “gallium-based” refers to semiconducting compounds that include at least gallium, for example, GaN or GaAs. In some embodiments, the gallium- based semiconductor structure 190 may include, for example, a gallium nitride-based channel layer 124 (e.g., a GaN channel layer) and a gallium nitride based-barrier layer 126 (e.g., an AlGaN barrier layer) that is formed on the gallium nitride-based channel layer 124 opposite the substrate 122. Gallium nitride-based semiconductor structures 190 may include at least gallium and nitrogen, including gallium nitride (GaN) as well as ternary and quaternary compounds such as aluminum gallium nitride (AlGaN) and aluminum indium gallium nitride (AlInGaN).

[0073] A plurality of enhancement and/or depletion mode transistor devices 100 may be formed on and in the gallium-based semiconductor structure 190. The source and drain contacts 115 and 105 may be formed on the semiconductor structure 190, and the gate may be formed on the barrier layer 126 between the contacts 115 and 105. For example, the source and the drain contacts 115 and 105 may be formed as ohmic contacts (e.g., on implanted regions 215 and 205 in the channel layer 124) such that an electric current flows between the source and drain contacts 115 and 105 via the 2DEG channel region 40 induced at the heterointerface between the channel layer 124 and barrier layer 126 when the gate 110 is biased at an appropriate level.

[0074] In some devices, a portion 126B of the gallium-based barrier layer 126 that is between the source and drain contacts 115 and 105 may be at least partially etched to form at least one opening 126R (also referred to herein as a recess 126R) in the barrier layer 126. A conductive element 110, 120 (such as a gate or field plate) may be formed in the recess 126R. As such, the portion 126B of the barrier layer 126 under the conductive element 110, 120 may have a thickness T2 that is less than a thickness T1 of portions 126A of the barrier layer 126 adjacent the source contact 115 or the drain contact 105. The portion 126B of the barrier layer 126 that extends under the gate 110 (i.e., between the gate 110 and the channel layer 124) may be referred to as the gate region 126B1, while the portion 126B of the barrier layer 126 under the field plate 120 (i.e., between the field plate 120 and the channel layer 124 may be referred to as the field plate region 126B2. The region between the gate 110 and the source contact 115 (and the region between the gate 110 and the drain contact 105) may be referred to herein as access regions, or source/drain access regions 126A. The gate region 126B1 may include a gate recess 126R1 in the barrier layer 126, while the field plate region 126B2 may include a field plate recess 126R2 in the barrier layer 126.

[0075] The gate recess 126R1 may extend partially through (e.g., in a depletion mode device) or completely through (e.g., in an enhancement mode device) the barrier layer 126. In a GaN HEMT including a GaN channel layer and an AlGaN barrier layer, conventional methods for recessing the barrier layer in enhancement mode and/or in depletion mode devices may use a single etch step to remove a portion of the AlGaN barrier layer to form the gate recess. This approach may have limited control over the etch depth, and/or may result in undesirable surface morphology effects on the barrier layer, including increased surface roughness, surface state changes, etc. Such surface morphology effects may cause variability in performance metrics, including but not limited to channel sheet resistance and threshold voltage, as well as potential reliability risks.

[0076] Some embodiments of the present invention may arise from realization that, as device integration increases, improved performance can be achieved by fabrication methods that allow more highly controlled removal of the barrier layer (e.g., a Ga-based material, including but not limited to GaN-based and GaAs-based materials) to form one or more recesses in the barrier layer (e.g., the gate recess or a field plate recess), with limited or without substantial impact to the final surface states of the barrier layer. Some embodiments described herein may provide fabrication methods for forming (and transistor devices including) barrier layers that are thicker in the source/drain access regions 126A and/or thinner in the gate region 126B1 and/or field plate region 126B2 than some conventional devices. In particular, embodiments of the present invention include a looped recess process for improved control over the depth of the gate recess 126R1 and/or field plate recess 126R2, which may in some embodiments be comparable or equivalent to monolayer removal or atomic layer etching (ALE). The thickness T1 may be used herein to refer to the thickness of the barrier layer 126 outside the recess(es) 126R described herein (e.g., T1 may refer to the barrier layer thickness in in the source/drain access region), while the thickness T2 may be used herein to refer to the barrier layer thickness at the floor 126F of the recess(es) 126R. For example, the thickness T2 may refer to the barrier layer thickness T2G at the floor 126F1 under the gate 110, or the barrier layer thickness T2FP at the floor 126F2 under the field plate 120.

[0077] Benefits of maintaining the barrier layer thickness in the source and drain access regions 126A while selectively removing a portion of the barrier layer (e.g., in a depletion mode device) or an entirety of the barrier layer (in an enhancement mode device) in the gate region 126B1 to define the gate recess 126R1 may include (but are not limited to) maintaining sub-threshold gate control and increasing peak transconductance (gm) without sacrificing on-state resistance (Ron), which may be advantageous for sub- 100 nm gate length (LG) technology development. In particular, as gate length LG is reduced without an increase in input voltage, a HEMT may be subject to short channel effects. In depletion mode devices, this may be alleviated by also reducing the thickness of the barrier layer in the channel region and/or increasing the thickness of the barrier layer in the source/drain access region; however, reducing the thickness of the barrier layer may negatively affect carrier concentration and conduction in the 2DEG channel region.

[0078] As such, by locally reducing the thickness T2G (see FIG. IB) of the barrier layer in the gate region 126B1 (e.g., layer-by-layer, using a looped recess process as described herein), while maintaining or increasing the thickness T1 of the barrier layer 126 in the source/drain access regions 126 A, embodiments of the present invention may allow for reduction of the gate length LG (for example, below 100 nm or below 90 nm), while maintaining on-resistance Ron and sub-threshold gate control and increasing transconductance gm. That is, embodiments of the present invention may provide performance benefits by increasing transconductance gm without sacrificing on-resistance. Some embodiments may achieve a peak transconductance gm of greater than about 400 mS/mm (e.g., about 450 mS/mm to 800 mS/mm, about 500 mS/mm to 600 mS/mm, or about 510 to 550 mS/mm), which may be at least 15% to 20% greater than that of some conventional HEMT devices.

[0079] Also, by locally reducing the thickness T2FP of the barrier layer in the field plate region 126B2 (e.g., layer-by-layer, using a looped recess process as described herein), while maintaining or increasing the thickness T1 of the barrier layer 126 in the source/drain access regions 126 A, embodiments of the present invention may allow for reduction in parasitic sources, such as parasitic capacitance between the gate and drain (C _g d) and/or between the gate and source (C _gs ). In some embodiments, an additional passivation layer (e.g., IFFP 151) may be deposited in the field plate recess 126R2 before forming the field plate 120 therein. [0080] FIG. IB is an enlarged cross-sectional view of a section of the active area of the HEMT structure 100, 100’, 100” of FIGS. 1A, ID, or IE illustrating the recess 126R (e.g., the gate recess 126R1 or the field plate recess 126R2). In particular, FIG. IB illustrates the selective removal of a portion of the barrier layer 126 (e.g., using a looped recess process as described herein) in region 126B to define the recess 126R. FIG. 1C is an electron microscope image of a cross-section of a gate recess of a transistor device according to some embodiments of the present invention.

[0081] As shown in FIGS. IB and 1C, a first portion 126A of the barrier layer 126 adjacent the source 115 or drain contact 105 (for example in the source/drain access region 126A between the source contact 115 or drain contact 105 and the gate 110) has a thickness Tl, and a second portion 126B of the barrier layer between the conductive element (e.g., the gate 110 or the field plate 120) and the channel layer 124 (i.e., along a floor 126F of the recess 126R) has a thickness T2 (which may represent the barrier layer thickness T2G under the gate 110 or the barrier layer thickness T2FP under the field plate 120). In some embodiments, the thickness Tl of the barrier layer 126 in the source/drain access region 126A may be greater than about 150 Angstroms (A) (e.g., about 150 A to 350 A, or about 200 A to 250 A), while the thickness T2 of the barrier layer 126 under the conductive element 110, 120 may be less than about 150 A (e.g., about 30 A to 150 A, or about 50 A to 100 A). The first thickness Tl of the source/drain access portion 126A of the barrier layer 126 may be greater than the second thickness T2 of the recess portion 126B by about 250 A or less or about 150 A or less, for example, by about 100 A.

[0082] The looped recess process as described herein may allow for controlled, layer-by- layer removal of the portion 126B of the barrier layer 126 to form the recess 126R (for either the gate recess 126R1 or the field plate recess 126R2) to a desired depth D (as measured from a surface of the barrier layer 126 opposite the channel layer 124), in some embodiments independent of the thickness of the barrier layer 126 in the source/drain access region 126A, and with limited effect on the surface states of the etched portion 126B of the barrier layer 126. For example, the floor 126F of the recess 126R may have a surface roughness of less than about 3 A, for example, about 1 A to 2 A, e.g., about 1.5 A, when formed in accordance with embodiments described herein. The surface roughness of less than about 3 A may be present on one or more or all of the recessed surfaces, including the floor 126F, the corners 126C, and the sidewalls 126S of the gate recess 126R. The surface roughness may refer to a root mean square (RMS) surface roughness (Rq) or an average surface roughness (Ra) per unit area of the floor 126F. In some embodiments, the edges or corner portions 126C of the recess 126R (which may include inclined surfaces in some embodiments) may have a similar or same surface roughness as the substantially planar portions of the floor 126F therebetween. That is, the recess 126R may have the desired surface roughness from sidewall 126S to sidewall 126S, along an entirety of the portion 126B of the barrier layer 126 between the sidewalls 126S (including the floor 126F and the edges 126C). In some embodiments, the edges 126C of the recess 126R may be coplanar with the substantially planar portions of the floor 126F extending therebetween (such that the recess 126R includes a substantially planar surface extending from sidewall 126S to sidewall 126S), while in other embodiments, the edges 126C may be inclined or sloped relative to the substantially planar portions of the floor 126F, as described greater detail below.

[0083] In some embodiments, the depth D of the recess 126R may be less than 250 A, for example, about 50 A to 150 A, e.g., about 100 A. That is, the depth D of the recess 126R may be precisely controlled such that the first thickness T1 of the source/drain access portion 126A of the barrier layer 126 may be about 1.2 times to about 4 times greater (e.g., about 1.5 to about 4 times greater) than the second thickness T2 of the recess portion 126B (e.g., T1 may be about 2 times thicker or about 3 times thicker than T2). In some embodiments, the relative thickness T2 of the recess portion 126B may be expressed as a percentage of the thickness T1 of the source/drain access portion 126A. For example, T2 may be about 25% to 83% (e.g., about 25% to 66%) of T1. For example, for a barrier layer thickness T1 of about 30 nm, the thickness T2 in the recess portion 126B of the barrier layer 126 may be about 7.5 nm to about 25 nm.

[0084] In some embodiments, the length of the gate 110 along the floor 126F of the gate recess 126R1 (referred to herein as the gate length LG) may be less than about 100 nanometers (nm). For example, the gate length LG may be about 90 nm (or less), about 70 nm (or less), about 60 nm (or less), or about 50 nm (or less). Looped recess processes in accordance with some embodiments of the present invention may be used to precisely scale the thickness T2G of the portion 126B of the barrier layer 126 in the gate recess 126R1 with the gate length LG, such that a desired ratio of the gate length LG to the barrier layer thickness T2 in the gate recess 126R1 may be maintained with decreasing gate length.

[0085] Controlling the relative thicknesses T1 and T2 of the barrier layer 126 (as well as the ratio of the barrier layer thickness T2G relative to the gate length LG) in the ranges described herein may be critical to achieving some of the improved electrical performance characteristics described herein. For example, providing the thicknesses T1 and T2G within the ranges described herein and/or maintaining a desired ratio of the gate length LG to the barrier layer thickness T2G in the gate recess 126R1 can maintain sub-threshold control with decreasing gate length LG, and thus, may contribute to increased peak transconductance gm max. In some embodiments, for a gate length LG of about 150 nm, the barrier layer 126 may be recessed to a thickness T2G of about 150 A (15 nm) at the floor 126F of the gate recess 126R1. As another example, for a gate length LG of about 90 nm, the barrier layer 126 may be recessed to a thickness T2G of about 90 A (9 nm) at the floor 126F of the gate recess 126R1. In another example, for a gate length LG of less than 90 nm, the barrier layer 126 may be recessed to a thickness T2G of less than 90 A (9 nm) at the floor 126F of the gate recess 126R1. In some embodiments, the barrier layer thickness T2omay be no less than 50 A (5 nm) at the floor 126F of the gate recess 126R1.

[0086] That is, embodiments of the present invention may provide or maintain a ratio of gate length LG (in nm) to recessed barrier layer thickness T2G (in A) of about 10: 1 or less, i.e., such that the recessed barrier layer thickness T2G is about one tenth of the gate length LG (or less). In contrast, some conventional transistor devices with a gate length of about 350 nm may have a barrier layer thickness of about 250A (25 nm) in the gate recess (or a gate length of about 250 nm and a barrier layer thickness of about 180A (18 nm) in the gate recess), that is, with a ratio of gate length to barrier layer thickness of about 13 : 1 or 14: 1 or more.

[0087] Some embodiments of the present invention may provide transistor devices 100 with a peak transconductance gm of greater than about 400 mS/mm, for example, about 450 mS/mm to 800 mS/mm, about 500 mS/mm to 600 mS/mm, or about 510 to 550 mS/mm (as compared to a gm of up to about 400 mS/mm for some conventional devices) and an on-resistance Ron of about 1.5 Q-mm to 3 Q-mm, for example, about 1.6 Q-mm, with a cut-off frequency (fr) of about 15 to 40 GHz. Some conventional transistor devices may thus exhibit a significantly lower peak transconductance gm than transistor devices having the barrier layer thicknesses Tl, T2, and/or gate length-to-b airier layer thickness ratios LG: T2G described herein.

[0088] Likewise, controlling the relative thicknesses Tl and T2 of the barrier layer 126 in the ranges described herein may be critical to achieving improved electrical performance characteristics with respect to field plates 120 described herein, independent of or in combination with the performance improvements associated with controlling the barrier layer thickness T2G in the gate recess 126R1. For example, providing the thicknesses Tl and T2FP within the ranges described herein may allow for independent control of field plate-to- channel spacing (i.e., via the barrier layer thickness T2FP in the field plate region 126B2), field plate length LFP, and/or surface states in the field plate recess 126R2, which may affect parasitic capacitance between the gate and drain (C _g d) and/or between the gate and source (Cgs).

[0089] Also, the thickness T2G at the floor 126F1 of the recessed barrier layer 126 in the gate region 126B1 and the thickness T2FP at the floor 126F2 of the recessed barrier layer 126 in the field plate region 126B2 may be independently controlled, e.g., by using respective looped recess processes as described herein. As such, the thickness T2G at the floor 126F1 of the recessed barrier layer 126 in the gate region 126B1 and the thickness T2FP at the floor 126F2 of the recessed barrier layer 126 in the field plate region 126B2 may be formed to have the same thickness (e.g., when formed using the same recess process) or different thicknesses (e.g., when formed using respective recess processes).

[0090] FIG. IF is an enlarged cross-sectional view of a section of the active area of the HEMT structure 100” of FIG. IE illustrating an example gate recess 126R1 and field plate recess 126R2 in greater detail. In particular, FIG. IE illustrates the selective removal of respective portions of the barrier layer 126 in regions 126B1 and 126B2 (e.g., using multiple sequential looped recess processes as described herein) to define recesses 126R1 and 126R2 of different depths. For example, the gate recess 126R1 may be formed using a first looped recess process 200, 200’, while the field plate recess 126R2 may be formed using a second looped recess process 200, 200’ that is performed before or after the first looped recess process. In some embodiments, the respective recess processes may be selectively performed in one of the regions 126B1 or 126B2 and then the other (e.g., by forming a protective layer or mask on one of the regions 126B1 or 126B2 while recessing the other). In some embodiments, the first recess process may be used to recess the barrier layer 126 to a same depth or by a same amount in both regions 126B1 and 126B2, and the second recess process may selectively recess the barrier layer 126 in region 126B2 without further recessing the barrier layer 126 in region 126B1 (e.g., by forming a protective layer or mask on region 126B1 while further recessing region 126B2), or vice versa.

[0091] In some embodiments, the thickness T2FP at the floor of the recess 126R2 in the field plate region 126B2 may be less than or equal to (e.g., about 50% to 100% of) the thickness T2G at the floor of the recess 126R1 in the gate region 126B1. For example, the thickness T2G at the floor of the recess 126R1 in the gate region 126B1 may be about 30 A (3 nm) to 150 A (15 nm), e.g., about 50 A (5 nm), and the thickness T2FP at the floor of the recess 126R2 in the field plate region 126B2 may be about 20 A (2 nm) to about 150 A (15 nm), e.g., about 25 A (2.5 nm). In some embodiments, the thickness T2FP of the barrier layer 126 at the floor of the field plate recess 126R2 may be no less than 50 A (5 nm), or no less than 20 A (2 nm).

[0092] The lengths Loand LFP of the gate 110 and field plate 120 in the respective recesses 126R1 and 126R2 can likewise be independently controlled, e.g., by respective photolithography processes. In some embodiments, the thickness T2G of the recessed barrier layer 126 in the gate region 126B1 and the thickness T2FP of the recessed barrier layer 126 in the field plate region 126B2 may be independently controlled based on the distance or lateral offset (as desired) between the gate 110 in region 126B1 and the field plate 120 in region 126B2. The chemistry of the looped recess process can also be adjusted to coordinate or independently control the surface states in the regions 126B1 and 126B2. That is, embodiments of the present disclosure may utilize looped recess processes and/or photolithography processes as described herein to independently control recess depth(s), recess length(s), and/or surface state(s) in the respective recesses 126R1 and/or 126R2, in various combinations to improve electrical performance characteristics as described herein. [0093] In addition, the looped recess process as described herein may include one or more parameters that can be controlled to provide the sidewalls 126S of the gate recess 126R1 and/or the field plate recess 126R2 (collectively, 126R) with one or more desired profiles or slope angles, as shown in FIGS. IB and IF. In particular, when formed using a looped recess process according to some embodiments of the present invention, the recess 126R in the barrier layer 126 may include opposing sidewalls 126S defining one or more angles of about 45 degrees to about 90 degrees relative to (or relative to a plane parallel to) the floor 126F extending between the sidewalls 126S. For example, the sidewall 126S may include a first portion SI having a slope or angle al of about 90 degrees (e.g., about 85 degrees to about 95 degrees, with greater controllability and less dependence on crystal structure than a conventional gate etch process) and a second portion S2 having a slope or angle a2 of about 45 degrees (e.g., about 40 degrees to about 50 degrees). The angle a2 of the sidewall portion S2 may thus define an intermediate thickness (between T1 and T2) such that the barrier layer 126 has a thickness T3 at the edge or corner portions 126C (between the sidewall SI and the floor 126F) of the recess 126R that is greater than the thickness T2 at the substantially planar floor 126F therebetween.

[0094] That is, a thickness T3 at edges 126C of the recess 126R may be greater than a thickness T2 at substantially planar portions of the floor 126F therebetween. The sidewall portion S2 at the angle a2 (defining the intermediate thickness T3) may define an inclined surface that provides a more gradual transition between the floor 126F and the more vertical sidewall portion SI of the recess 126R, which may improve carrier mobility and/or provide a less abrupt change in electric field distribution (or 2DEG charge density) in the barrier layer 126 in comparison to a recess formed by some conventional gate or field plate etch processes (e.g., which may provide sharper corners between the floor and sidewalls thereof). As such, the sidewall profile SI, S2 may reduce peak electric field concentration at corners 126C of the gate recess 126R1, which may improve device reliability, for example, with respect to electric field breakdown.

[0095] However, it will be understood that the inclined surfaces at the edges 126C may not be present in some embodiments (i.e., the angle a2 of portion S2 may be zero), such that the substantially planar floor 126F may continuously extend between the opposing sidewall portions SI with a substantially uniform thickness T2 (i.e., T2 may be equal to T3). For example, the portions S2 of the sidewalls 126S may have a different angle a2 in the gate recess 126R1 than in the field plate recess 126R2. Also, while illustrated with reference to a surface or portion S2 inclined at a single angle a2 by way of example, in some embodiments the surface S2 may include multiple inclined surfaces, each inclined at a respective angle between zero degrees and the angle al of the sidewall portion SI (inclusive), or may include a surface with a curvature between the surface of the floor 126F and the sidewall portion SI. In embodiments where the angle a2 is zero degrees, the sidewall SI may effectively continuously extend to the floor 126F (i.e., the substantially vertical sidewall portion SI may intersect the substantially horizontal floor 126F).

[0096] As shown in the image of FIG. 1C, the transistor device 100 includes a recess 126R (which may represent a gate recess 126R1 and/or a field plate recess 126R2) in an AlGaN barrier 126, with sidewalls 126S having the etch profile angles al, a2 and thicknesses Tl, T2, T3 shown in FIG. IB or IF. While illustrated with reference to two sidewall portions SI and S2, it will be understood that the sidewalls 126S of the recess 126R may include fewer or more portions by varying one or more parameters the looped recess process as described herein.

[0097] FIGS. 2A and 2B are schematic diagrams illustrating methods of forming a gate recess for a transistor device using a looped recess process according to some embodiments of the present invention. FIG. 3 is a graph illustrating example parameters for a looped recess process according to some embodiments of the present invention.

[0098] As shown in FIGS. 2A and 2B, embodiments of the present invention provide a looped recess process 200, 200’ that allows for layer-by-layer removal of a portion of a barrier layer 126 to define one or more recesses 126R (e.g., a gate recess 126R1 and/or a field plate recess 126R2) as described herein. The looped recess process 200, 200’ may be a plasma etch process, where each of the loops is configured to remove a monolayer of the barrier layer 126. A monolayer as described herein may have a thickness of about 5 Angstroms (A) or less, for example, about 0.1 A to 5 A, or about 0.25 A to 4 A. The monolayer thickness may correspond to the lattice constant of the material being recessed. For example, for an AlGaN barrier layer, the monolayer thickness may be about 3.186 A. [0099] FIG. 2A illustrates that each loop of the plasma etch process 200 includes a reactive (e.g., chlorine (Cl)-based) adsorption operation 205 and a low power (e.g., argon (Ar)-based) desorption operation 215, with purge operations 210, 220 (e.g., helium (He)- or nitrogen (N)- based) therebetween. The operations 205, 210, 215, and 220 may be repeated or “looped” to provide layer-by-layer removal of the barrier layer 126 until a desired depth D of the recess 126R is achieved.

[00100] FIG. 2B illustrates example parameters for a plasma etch process 200’ in greater detail. In the operations of FIG. 2B, a Ga-based semiconductor wafer (e.g., the substrate 122 including a GaN channel layer 124 and an AlGaN barrier layer 126) may be mounted on a chuck provided inside a chamber of an inductively coupled plasma (ICP) etching apparatus. The chamber may include a platen and may be surrounded by a coil. Power applied to the coil may generate plasma from a gas introduced into the chamber, while power applied to the platen may affect ion flux and speed at which the ions are accelerated toward the surface to be etched (e.g., the exposed surface of the AlGaN barrier layer 126).

[00101] In particular, in operation 205’, a reactant gas (e.g., Cl) is introduced into the chamber (e.g., at a pressure of about 6 mT), for example, with a flow rate of about 50 seem. High frequency (RF) power (e.g., at about 200 W) is applied to a coil of the ICP etching apparatus, which induces an electromagnetic current that acts on the reactant gas to generate a plasma for the adsorption process. For example, operation 205’ may include chlorinating the surface of AlGaN barrier layer (e.g., with a Cl-based gas, such as Ch or boron trichloride (BCh)). The chlorinating may be performed using coil power only (i.e., without ion driving/platen power), which may promote primarily chemical reactive adsorption of Cl on the exposed Ga-based surface. In operation 210’, a purge process is performed by introducing a purge gas (e.g., He or N) into the chamber to evacuate the reactant gas.

[00102] In operation 215’, an etchant gas (e.g., Ar) is introduced into the chamber (e.g., at a pressure of about 10 mT), for example, with a flow rate of about 50 seem. High frequency (RF) power (e.g., at about 200 W) is applied to a coil of the ICP etching apparatus to induce an electromagnetic current that acts on the etchant gas to generate a plasma for the desorption process. For example, operation 215’ may include an argon-based ion milling operation that results in the desorption of the volatile etch by-product. The Ar-based ion milling may be performed at a relatively low platen power (for example, about 4W to 10W, e.g., 6W), with a controlled peak-to-peak voltage so as to produce substantially uniform incident ion energy. The peak-to-peak voltage may refer to the magnitude of the RF voltage at the surface of the platen.

[00103] In operation 220’, another purge process is performed by introducing a purge gas (e.g., He or N) into the chamber to evacuate the etchant gas. The operations 205’, 210’, 215’, and 220’ are looped until the desired recess depth in the Ga-based barrier layer is achieved by the looped recess process 200’.

[00104] While described herein with respect to a looped recess process 200’ including a Cibased adsorption process 205’ and an Ar-based desorption process 215’, it will be understood that adsorption 205 and desorption 215 processes as described herein are not limited to these elements. For example, 02-based adsorption and BCh-based desorption loops may also be used. More generally, for example, the ALE process 200 may be terminated with any plasma chemistry, including (but not limited to) O2, Cl, He, F.

[00105] In some embodiments, one or more parameters of the looped recess process 200, 200’ may be defined to ensure monolayer removal, e.g., to account for ICP tool variation. FIG. 3 illustrates variation in etch rate (in A per loop) with peak-to-peak voltage in the looped recess process 200, 200’. As shown in the graph of FIG. 3, the plasma etch process is performed using an ICP tool and a peak-to-peak voltage of less than about 120V, for example, about 70V to about 110V. This peak-to-peak voltage range results in an etch rate of less than about 5 A per loop, effectively providing atomic layer etching 325 (i.e., monolayer removal). In contrast, at peak-to-peak voltages of greater than 120V, etching rate may be greater than about 5 A per loop, which may be similar or equivalent to a sputter etch regime or process 350.

[00106] FIGS. 2A, 2B and 3 are described herein with reference to methods of removing portions of an AlGaN barrier layer using a looped plasma process to effectively achieve atomic layer etching using an ICP tool, but it will be understood that embodiments of the present invention are not limited to the particular materials and/or tools described herein, and may be similarly applied to other materials and/or tools. For example, layer-by-layer removal of other Ga-based layers (e.g., an AlGaAs layer) may be performed using a looped Cl adsorption and Ar desorption process and parameters as shown in FIGS. 2A, 2B and 3. [00107] The looped recess process 200, 200’ described herein can thus be used in HEMT device fabrication to selectively etch the recess 126R in the barrier layer 126 (e.g., the gate recess 126R1 and/or the field plate recess 126R2), in both enhancement mode and depletion mode devices, with greater control of the recess depth D and improved surface characteristics at the floor 126F of the recess 126R than some conventional plasma-based etching. For example, in some embodiments, the looped recess process 200, 200’ may provide a surface roughness of less than about 3 A (e.g., about 1 A to 2 A) along the floor 126F of the recess 126R.

[00108] The looped recess process 200, 200’ may also allow for etching of the recess 126R in the barrier layer 126 with a higher degree of directionality than some conventional plasmabased etching. For example, as discussed above with reference to FIG. IB and IF, the recess 126R may have sidewalls 126S with a slope SI, S2 of about 90 degrees, or about 45 degrees to 90 degrees. In some embodiments, the sidewalls 126S of the recess 126R may include a lower portion S2 with a slope of about 45 degrees adjacent the floor 126F of the recess 126R, and an upper portion SI with a slope of about 90 degrees. More generally, one or more parameters of the looped recess process 200, 200’ may be varied to provide a desired sidewall slope 126S of up to about 90 degrees, as desired (and which may differ) for gate recesses 126R1 and field plate recesses 126R2.

[00109] FIGS. 4A-4D, 5A-5D, and 6A-6D are schematic cross-sectional views illustrating methods of forming a recess 126R (e.g., a gate recess 126R1 and/or a field plate recess 126R2) for a transistor device 100 using looped recess processes according to various embodiments of the present invention. In some embodiments, the operations of FIGS. 4A- 4D, 5 A-5D, and 6A-6D may be implemented after fabrication of the source and drain ohmic contacts 115 and 105, to avoid damaging the gate 110 in the high temperature ohmic process. [00110] FIGS. 4A-4D illustrate methods of forming a gate recess using a looped recess process according to some embodiments. As shown in FIG. 4A, one or more passivation layer(s) 150 are formed on the semiconductor structure 190 including the channel layer 124 and the barrier layer 126. In FIG. 4B, an etch mask 155 (e.g., a patterned photoresist mask) is formed on the passivation layer(s) 150. The etch mask 155 includes an opening 155op therein (e.g., as formed by electron beam lithography on the photoresist material). As shown in FIG. 4C, the passivation layer(s) 150 are etched using the mask 155 to expose a surface of the barrier layer 126 in the gate region 126B, thereby forming an initial opening or recess 150op. The surface of the barrier layer 126 exposed by the initial opening 150op in the mask 155 may define the desired gate length LG for a gate recess 126R1 and/or field plate length LFP for a field plate recess 126R2. The looped recess process 200, 200’ is then performed to remove, layer-by-layer, the portion of the barrier layer 126 in the gate region 126B exposed by the initial opening 150op to form the recess 126R, as shown in FIG. 4D. The recess 126R may include one or more of the characteristics (e.g., surface roughness, sidewall slope(s) al, a2, relative thicknesses Tl, T2, T3, gate length-to-recessed barrier layer thickness ratios LG:T2, transconductance gm, on-resistance Ron) described herein.

[00111] That is, prior to performing the looped recess process 200, 200’ to define the recess 126R with the desired characteristics, a passivation layer 150 is formed on the semiconductor structure 190, and an opening 150op is formed in the passivation layer 150 to expose a surface of the barrier layer 126. Although illustrated with reference to forming a single recess 126R, it will be understood that multiple recesses (e.g., multiple gate recesses 126R1 and/or multiple field plate recesses 126R2) may be simultaneously or sequentially formed by the looped recess process 200, 200’ in portions of the barrier layer 126 that are exposed by the initial opening 150op in different regions of the substrate 122. An additional passivation layer (e.g., IFPP 151) may be deposited in the field plate recess 126R2.

[00112] In the example of FIGS. 4A-4D, the source and drain contacts 115 and 105 and the source/drain access regions 126A are protected (e.g., by the etch mask 155) during the looped plasma etch process 200, 200’. Similarly, when multiple recesses 126R of different depths (e.g., a gate recess 126R1 with a first depth and a field plate recess 126R2 with a second depth that is different than the first depth) are formed, first and second looped recess processes 200, 200’ may be performed sequentially, where one of the recesses 126R may be protected during the looped plasma etch process 200, 200’ to form the other of the recesses 126R.

[00113] FIGS. 5A-5D illustrate methods of forming a gate recess using a looped recess process according to further embodiments. As shown in FIG. 5 A, one or more passivation layer(s) 150 are formed on the semiconductor structure 190 including the channel layer 124 and the barrier layer 126, and the passivation layer(s) 150 are patterned to define an opening 150op therein in the gate region 126B. In FIG. 5B, a spacer layer 160 is formed on the surface of the passivation layer(s) and in the opening 150op. As shown in FIG. 5C, the spacer layer 160 is patterned to form spacers 160S on opposing sidewalls of the passivation layer(s) 150 and to expose a surface of the barrier layer 126 in the gate region 126B. The surface of the barrier layer 126 exposed by the spacers 160S may define the desired gate length LG and/or field plate length LFP. The looped recess process 200, 200’ is then performed to remove, layer-by-layer, the portion of the barrier layer 126 in the gate region 126B exposed by the spacers 160S to form the recess 126R, as shown in FIG. 5D. The recess 126R may include one or more of the characteristics described herein.

[00114] That is, prior to performing the looped recess process 200, 200’ to define the recess 126R with the desired characteristics, a passivation layer 150 is formed on the semiconductor structure 190, an opening 150op is formed in the passivation layer 150 to expose a surface of the barrier layer 126, and spacers 160S are defined on opposing sidewalls of the opening 150op to define the desired gate length LG for the gate recess 126R1 and/or field plate length LFP for the field plate recess 126R2. An additional passivation layer (e.g., IFPP 151) may be deposited in the field plate recess 126R2. Although illustrated with reference to forming a single recess 126R, it will be understood that multiple recesses (e.g., 126R1, 126R2) may be simultaneously or sequentially formed by the looped recess process 200, 200’ in portions of the barrier layer 126 that are exposed by the opening 150op and/or the spacers 160S in different regions of the substrate 122.

[00115] In the example of FIGS. 5A-5D, the source and drain contacts 115 and 105 and the source/drain access regions 126A may be protected (e.g., by portions of the spacer layer 160) during the looped plasma etch process 200, 200’ . Similarly, when multiple recesses 126R of different depths (e.g., 126R1 and 126R2) are formed, first and second looped recess processes 200, 200’ may be performed sequentially, where one of the recesses 126R may be protected during the looped plasma etch process 200, 200’ to form the other of the recesses 126R.

[00116] FIGS. 6A-6D illustrate methods of forming a gate recess using a looped recess process according to yet further embodiments. As shown in FIG. 6A, one or more passivation layer(s) 150 are formed on the semiconductor structure 190 including the channel layer 124 and the barrier layer 126, and the passivation layer(s) 150 are patterned (e.g., using a mask 155) to define an opening 150op therein exposing a surface of the barrier layer 126 in the gate region 126B. In FIG. 6B, the looped recess process 200, 200’ is performed to remove, layer-by-layer, the portion of the barrier layer 126 in the gate region 126B exposed by the opening 150op to define an initial recess 126R’. The initial recess 126R’ may include one or more of the characteristics of the recess 126R described herein. In FIG. 6C, a spacer layer 160 is formed on the surface of the passivation layer(s) and in the initial recess 126R’. As shown in FIG. 6D, the spacer layer 160 is patterned to form spacers 160S on opposing sidewalls of the passivation layer(s) 150 and to expose a surface of the barrier layer 126 in the gate region 126B, defining the recess 126R. The surface of the barrier layer 126 exposed by the mask 155 and/or spacers 160S may define the desired gate length LG and/or field plate length LFP.

[00117] That is, a passivation layer 150 is formed on the semiconductor structure 190 and an opening is formed in the passivation layer 150 to expose a surface of the barrier layer 126 prior to performing the looped recess process 200, 200’ to define the initial recess 126R’ with the desired characteristics, while the spacers 160S are formed on opposing sidewalls of the passivation layer 150 and the recess 126R’ after performing the looped recess process 200, 200’. Although illustrated with reference to forming a single recess 126R, it will be understood that multiple recesses (e.g., 126R1, 126R2) may be simultaneously or sequentially formed by the looped recess process 200, 200’ in portions of the barrier layer 126 that are exposed by the opening 150op and/or spacers 160S in different regions of the substrate 122. An additional passivation layer (e.g., IFPP 151) may be deposited in the field plate recess 126R2.

[00118] In the example of FIGS. 6A-6D, the source and drain contacts 115 and 105 and the source/drain access regions 126A may be protected (e.g., by the etch mask 155) during the looped plasma etch process 200, 200’. Similarly, when multiple recesses 126R of different depths (e.g., 126R1 and 126R2) are formed, first and second looped recess processes 200, 200’ may be performed sequentially, where one of the recesses 126R may be protected during the looped plasma etch process 200, 200’ to form the other of the recesses 126R. [00119] Embodiments of the present invention including a gate recess 126R1 with characteristics as described herein may provide improved performance, for instance, in RF and power switching devices. For example, embodiments of the present invention may be used in applications with operating frequencies that range from less than about 4 GHz to Ka- Band designs (e.g., 26 - 40 GHz). Embodiments of the present invention including a field plate recess 126R2 with characteristics as described herein may also provide improved performance in RF and power switching devices. For example, by allowing for more precise control of the field plate-to-channel spacing, embodiments of the present invention may a significant reduction in parasitic capacitances between the gate and drain (C _g d) and/or between the gate and source (C _gs ). More generally, embodiments of the present invention may allow for greater control over a distance between a conductive element 110, 120 and channel 124 (and between the conductive element 110, 120 and the 2DEG 40), in particular, by recessing a barrier layer 126 using looped recess processes 200, 200’ as described herein. [00120] FIG. 7 is a schematic plan view of a Group III nitride-based transistor die according to embodiments of the present invention that illustrates metallization on a surface of the semiconductor structure thereof. As shown in FIG. 7, a transistor device or die 1000 may include multiple transistor structures 100 connected in parallel to device terminals or electrodes (e.g., an input terminal, an output terminal, and a ground terminal). For example, each of the gate 110, drain 105, and source 115 contacts may extend in a first direction (e.g., the Y-direction) to define gate, drain, and/or source 'fingers', which may be connected by one or more respective buses (e.g., by a gate bus and a drain bus on an upper surface of the semiconductor structure 190.

[00121] In FIG. 7, the gate fingers 110, drain fingers 105 and source fingers 115 may extend in parallel to each other, with the gate fingers 110 extending from the gate bus 112 in a first direction and the drain fingers 105 extending from the drain bus 114 in a direction opposite the first direction. Each gate finger 110 may be positioned between a drain finger 105 and a source finger 115 to define a unit cell 100. The gate fingers 110, drain fingers 105, and source fingers 315 (and connecting buses) may define part of gate-, drain-, and source- connected electrodes of the device, respectively, as defined by a top or frontside metallization structure. Dielectric layers that isolate the various conductive elements of the frontside metallization structure from each other are not shown in FIG. 7 to simplify the drawing. Since the gate fingers 110 are electrically connected to a common gate bus 112, the drain fingers 105 are electrically connected to a common drain bus 114, and the source fingers 115 are electrically connected together (e.g., through respective via openings 146), it can be seen that the unit cell transistors 100 are electrically connected together in parallel.

[00122] One of the terminals of the device (e.g., a source terminal connected to the source contact(s) 115) may be configured to be coupled to a reference signal such as, for example, an electrical ground. In some embodiments, a conductive through substrate via connection or structure (e.g., a backside via opening) may extend through the substrate 122 and epitaxial layer(s) 124, 126 to expose a portion of one of the contacts 105, 115, so as to allow for contact pads or terminals on the back side of the substrate (e.g., to couple the source contact 115 to ground). In other embodiments, a ground connection to one of the terminals device (e.g., the source terminal) may be provided outside the active area, e.g., in a peripheral area. In some embodiments, a backmetal layer on the back side of the substrate 122 may provide a backside ground plane, for example, in applications where proximity to ground may be desired.

[00123] While embodiments of the present invention have been described herein with reference to particular HEMT structures, the present invention should not be construed as limited to such structures, and may be applied to formation of many different transistor structures, such as pHEMTs (including GaAs/AlGaAs pHEMTs) and/or GaN MESFETs. Also, additional layers may be included in transistor structures while still benefiting from the teachings of the present invention. Such additional layers may include GaN cap layers, as described for example U.S. Patent No. 6,548,333 to Smith. In some embodiments, insulating layers such as SiNx, or relatively high quality AIN may be deposited for making a MISHEMT and/or passivating the surface. The additional layers may also include a compositionally graded transition layer or layers. In addition, the barrier layer 126 and/or channel layer 124 described above may include multiple layers. Thus, embodiments of the present invention should not be construed as limiting these layers to a single layer but may include, for example, barrier layers having combinations of GaN, AlGaN and/or AIN layers.

[00124] FIGS. 8A, 8B, and 8C are schematic cross-sectional views illustrating several example ways that that the RF transistor amplifier dies according to embodiments of the present invention may be packaged to provide packaged RF transistor amplifiers 800A, 800B, 800C. FIGS. 8A-8C show packaging of the transistor device 1000, which may include any of the transistor structures described herein.

[00125] FIG. 8A is a schematic side view of a packaged Group III nitride-based RF transistor amplifier 800A. As shown in FIG. 8A, packaged RF transistor amplifier 800A includes the RF transistor amplifier die 1000 packaged in an open cavity package structure 810A. The package structure 810A includes metal gate leads 822 A, metal drain leads 824 A, a metal submount 830, sidewalls 840 and a lid 842.

[00126] The submount 830 may include materials configured to assist with the thermal management of the package 800 A. For example, the submount 830 may include copper and/or molybdenum. In some embodiments, the submount 830 may be composed of multiple layers and/or contain vias/interconnects. In an example embodiment, the submount 830 may be a multilayer copper/molybdenum/copper metal flange that comprises a core molybdenum layer with copper cladding layers on either major surface thereof. In some embodiments, the submount 830 may include a metal heat sink that is part of a lead frame or metal slug. The sidewalls 840 and/or lid 842 may be formed of or include an insulating material in some embodiments. For example, the sidewalls 840 and/or lid 842 may be formed of or include ceramic materials. In some embodiments, the sidewalls 840 and/or lid 842 may be formed of, for example, AI2O3. The lid 842 may be glued to the sidewalls 840 using an epoxy glue. The sidewalls 840 may be attached to the submount 830 via, for example, brazing. The gate lead 822A and the drain lead 824A may be configured to extend through the sidewalls 842, though embodiments of the present invention are not limited thereto. [00127] The RF transistor amplifier die 1000 is mounted on the upper surface of the metal submount 830 in an air-filled cavity 812 defined by the metal submount 830, the ceramic sidewalls 840 and the ceramic lid 842. Gate and drain terminals 132, 134 of RF transistor amplifier die 1000 are on the top side of the semiconductor structure 190, while the source terminal 136 is on the bottom side of the semiconductor structure 190. The source terminal 136 may be mounted on the metal submount 830 using, for example, a conductive die attach material (not shown). The metal submount 830 may provide the electrical connection to the source terminal 136 and may also serve as a heat dissipation structure that dissipates heat that is generated in the RF transistor amplifier die 1000.

[00128] Input matching circuits 850 and/or output matching circuits 852 may also be mounted within the package 800A. The matching circuits 850, 252 may include impedance matching and/or harmonic termination circuits. The impedance matching circuits may be used to match the impedance of the fundamental component of RF signals that are input to or output from the RF transistor amplifier to the impedance at the input or output of the RF transistor amplifier die 1000, respectively. The harmonic termination circuits may be used to ground harmonics of the fundamental RF signal that may be present at the input or output of the RF transistor amplifier die 1000. More than one input matching circuit 850 and/or output matching circuit 852 may be provided. As schematically shown in FIG. 8A, the input and output matching circuits 850, 852 may be mounted on the metal submount 830. The gate lead 822A may be connected to the input matching circuit 850 by one or more bond wires 854, and the input matching circuit 850 may be connected to the gate terminal 132 of RF transistor amplifier die 1000 by one or more additional bond wires 854. Similarly, the drain lead 824A may be connected to the output matching circuit 852 by one or more bond wires 854, and the output matching circuit 852 may be connected to the drain terminal 134 of RF transistor amplifier die 1000 by one or more additional bond wires 854. The bond wires 854, which are inductive elements, may form part of the input and/or output matching circuits.

[00129] FIG. 8B is a schematic side view of a packaged Group III nitride-based RF transistor amplifier 800B that includes the transistor device 1000 packaged in a printed circuit board based package structure 810B. The packaged RF transistor amplifier 800B is very similar to the packaged RF transistor amplifier 800A of FIG. 8 A, except that the gate and drain leads 822 A, 824 A of package structure 810A are replaced with printed circuit board based leads 822B, 824B in package structure 810B.

[00130] The package structure 810B includes a submount 830, ceramic sidewalls 840, a ceramic lid 842, each of which may be substantially identical to the like numbered elements of package structure 810A discussed above. The package structure 810B further includes a printed circuit board 820. Conductive traces on the printed circuit board 820 form a metal gate lead 822B and a metal drain lead 824B. The printed circuit board 820 may be attached to the submount 830 via, for example, a conductive glue. The printed circuit board 820 includes a central opening and the RF transistor amplifier die 1000 is mounted within this opening on the submount 830. Other components of RF transistor amplifier 800B may be the same as the like-numbered components of RF transistor amplifier 800A, and hence further description thereof will be omitted.

[00131] FIG. 8C is a schematic side view of another packaged Group III nitride-based RF transistor amplifier 800C. RF transistor amplifier 800C differs from RF transistor amplifier 800A in that it includes a different package structure 810C. The package structure 810C includes a metal submount 830 (which may be similar or identical to the submount 830 of package structure 810A), as well as metal gate and drain leads 822C, 824C. RF transistor amplifier 800C also includes a plastic overmold 860 that at least partially surrounds the RF transistor amplifier die 1000, the leads 822C, 824C, and the metal submount 830. Other components of RF transistor amplifier 800C may be the same as the like-numbered components of RF transistor amplifier 800A and hence further description thereof will be omitted.

[00132] The present invention is described with reference to the accompanying drawings, in which embodiments of the invention are shown. However, this invention should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In the drawings, the thickness of layers and regions are exaggerated for clarity. Like numbers refer to like elements throughout.

[00133] It will be understood that when an element such as a layer, region or substrate is referred to as being "on" or extending "onto" another element, it can be directly on or extend directly onto the other element or intervening elements may also be present. In contrast, when an element is referred to as being "directly on" or extending "directly onto" another element, there are no intervening elements present. It will also be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may be present. In contrast, when an element is referred to as being "directly connected" or "directly coupled" to another element, there are no intervening elements present. [00134] It will also be understood that, although the terms first, second, etc. may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the present invention.

[00135] Furthermore, relative terms, such as "lower" or "bottom" and "upper" or "top," may be used herein to describe one element's relationship to another element as illustrated in the Figures. It will be understood that relative terms are intended to encompass different orientations of the device in addition to the orientation depicted in the Figures. For example, if the device in one of the figures is turned over, elements described as being on the "lower" side of other elements would then be oriented on "upper" sides of the other elements. The exemplary term "lower", can therefore, encompasses both an orientation of "lower" and "upper," depending on the particular orientation of the figure. Similarly, if the device in one of the figures is turned over, elements described as "below" or "beneath" other elements would then be oriented "above" the other elements. The exemplary terms "below" or "beneath" can, therefore, encompass both an orientation of above and below.

[00136] The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the description of the invention and the appended claims, the singular forms “a,” “an ” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. [00137] Embodiments of the invention are described herein with reference to cross-section illustrations that are schematic illustrations of idealized embodiments (and intermediate structures) of the invention. As such, variations from the shapes of the illustrations as a result, for example, of manufacturing techniques and/or tolerances, are to be expected. Thus, embodiments of the invention should not be construed as limited to the particular shapes of regions illustrated herein but are to include deviations in shapes that result, for example, from manufacturing. For example, an implanted region illustrated as a rectangle will, typically, have rounded or curved features and/or a gradient of implant concentration at its edges rather than a binary change from implanted to non-implanted region. Likewise, a buried region formed by implantation may result in some implantation in the region between the buried region and the surface through which the implantation takes place. Thus, the regions illustrated in the figures are schematic in nature and their shapes are not intended to illustrate the actual shape of a region of a device and are not intended to limit the scope of the invention.

[00138] Unless otherwise defined, all terms used in disclosing embodiments of the invention, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, and are not necessarily limited to the specific definitions known at the time of the present invention being described. Accordingly, these terms can include equivalent terms that are created after such time. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the present specification and in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety.

[00139] In the drawings and specification, there have been disclosed typical embodiments of the invention, and, although specific terms have been employed, they have been used in a generic and descriptive sense only and not for purposes of limitation.