Front end integrated circuits incorporating differing silicon-on-insulator technologies

JP2022078000A5Inactive Publication Date: 2024-11-15SKYWORKS SOLUTIONS INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2021183828
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2020-11-12
Filing Date
2021-11-11
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing CMOS technologies face challenges in integrating high power PA devices with switches and LNAs due to difficulties in fabricating small devices using bulk or thick-film SOI technology, leading to high costs and complex integration issues.

Method used

A fully integrated FEIC is developed using SOI technology, where PDSOI power amplifiers are integrated in thick film regions and FDSOI switches and LNAs are integrated in thin film regions, eliminating the need to remove the BOX layer, and utilizing selective epitaxial growth or local thinning to achieve optimal device placement.

Benefits of technology

This approach reduces parasitics, enhances active device performance, and lowers costs by integrating PDSOI and FDSOI devices efficiently, resulting in high-performance FEICs with robust active and passive components.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

To provide a front end integrated circuit and a manufacturing method therefor that achieve both advantages of an FDSOI device and a PDSOI device.SOLUTION: An SOI-based technology platform provides a fully integrated front end integrated circuit (FEIC) 100a that includes a switch device, a low-noise amplifier (LNA), and a power amplifier (PA). The PA are built in a thick film region of the integrated circuit, resulting in a partially depleted silicon-on-insulator (PDSOIPA) 110, and the switch device and LNA are built in a thin film region of the integrated circuit, resulting in a fully depleted silicon-on-insulator (FDSOI) switch device 130 and an FDSOILNA device 120. The resulting fully integrated FEIC includes PDSOIPAs with FDSOI switches and LNAs. Passive components can be built in the thick film region, the thin film region, or both of the regions.SELECTED DRAWING: Figure 1A
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims priority to U.S. Provisional Application No. 63 / 112,951, entitled "Front - End Integrated Circuit Incorporating Different Silicon - on - Insulator Technologies", filed on November 12, 2020, the entire disclosure of which is hereby expressly incorporated by reference.

[0002] This disclosure generally relates to front - end integrated circuits for radio - frequency applications.

Background Art

[0003] A front-end module (FEM) is an embedded module that integrates various functional components used in the wireless front-end circuitry of a wireless device. Front-end modules can be configured to process radio frequency (RF) signals compatible with various wireless protocols such as broadband cellular network technologies (e.g., 3G, 4G, 5G, Long-Term Evolution (LTE), etc.), wireless network technologies (e.g., Wi-Fi), short-range wireless technologies (e.g., Bluetooth®), and Global Positioning System (GPS) technologies. A front-end module typically includes circuitry and electrical components between an antenna and a digital baseband system sufficient for transmitting and receiving radio frequency signals. A given FEM may include all the filters, low-noise amplifiers (LNAs), and down-conversion mixers necessary to process the modulated signal received by the antenna into a signal suitable for input to a baseband analog-to-digital converter (ADC). The FEM may also include a transmitter power amplifier (PA) and other circuitry that processes the signal for transmission via the antenna. FEMs may be surface-mount technology (SMT) modules, multi-chip modules (MCMs), etc. A FEM may include PA blocks, LNA blocks, input / output matching, MIPI standard digital control blocks, filters, duplexers, multiplexers, antenna switches, band selector switches, etc. A front-end integrated circuit (FEIC) is a single semiconductor die that contains the functions of the FEM. [Overview of the Initiative]

[0004] According to a certain number of implementation examples, the present disclosure relates to a front-end integrated circuit including a substrate, an insulator layer on the top of the substrate, and a semiconductor layer on the top of the insulator, wherein the semiconductor layer forms a thin-film region and a thick-film region, the thin-film region including one or more fully depleted silicon-on-insulator (FDSOI) low-noise amplifier (LNA) devices and one or more FDSOI switch devices, and the thick-film region including one or more partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) devices.

[0005] In some embodiments, the insulator layer is at least 100 nm thick. In some embodiments, the semiconductor layer in the thin film region is at least 5 nm thick and 50 nm or less thick. In some further embodiments, the semiconductor layer in the thick film region is at least about 50 nm thick and 180 nm or less thick.

[0006] In some embodiments, the insulator layer is an embedded oxide layer. In some embodiments, the semiconductor layer in the thin-film region is 1 / 4 the gate length of one or more FDSOILNA devices. In some embodiments, the front-end integrated circuit further includes one or more passive devices incorporated in the thin-film region of the semiconductor layer. In some embodiments, the front-end integrated circuit further includes one or more passive devices incorporated in the thick-film region of the semiconductor layer.

[0007] In some embodiments, thin-film regions of the semiconductor layer are formed using local thinning. In some embodiments, thick-film regions of the semiconductor layer are formed using selective epitaxial growth.

[0008] Based on a certain number of implementation examples, this disclosure relates to a method for manufacturing a front-end integrated circuit. The method includes forming an insulator layer on the top of a substrate. The method also includes forming a semiconductor layer on the top of the insulator layer. The method also includes incorporating a fully depleted silicon-on-insulator (FDSOI) low-noise amplifier (LNA) device into the semiconductor layer. The method also includes incorporating an FDSOI switch device into the semiconductor layer. The method also includes increasing the thickness of a portion of the semiconductor layer to form a thick-film region of the semiconductor layer. The method also includes incorporating a PDSOOIPA device into a thick-film region of the semiconductor layer such that the FDSOILNA device and the FDSOI switch device are located in a thin-film region of the semiconductor layer, and the partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) device is located in a thick-film region.

[0009] In some embodiments, the insulator layer is at least about 100 nm thick. In some embodiments, the thin-film region of the semiconductor layer is at least about 5 nm thick and less than or equal to about 50 nm thick. In some further embodiments, the thick-film region of the semiconductor layer is at least about 50 nm thick and less than or equal to 180 nm thick.

[0010] In some embodiments, the thin-film region of the semiconductor layer is 1 / 4 the gate length of the FDSOILNA device. In some embodiments, the method further includes incorporating one or more passive devices into the thin-film region of the semiconductor layer. In some embodiments, the method further includes incorporating one or more passive devices into the thick-film region of the semiconductor layer. In some embodiments, increasing the thickness includes using selective epitaxial growth.

[0011] Based on a certain number of implementation examples, this disclosure relates to a method for manufacturing a front-end integrated circuit. The method includes forming an insulator layer on the top of a substrate. The method also includes forming a semiconductor layer on the top of the insulator layer. The method also includes incorporating a partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) device into the semiconductor layer. The method also includes reducing the thickness of a portion of the semiconductor layer to form a thin-film region of the semiconductor layer. The method also includes incorporating a fully depleted silicon-on-insulator (FDSOI) low-noise amplifier (LNA) device into the thin-film region of the semiconductor layer. The method also includes incorporating an FDSOI switch device into the thin-film region of the semiconductor layer such that the PDSOOPA device is located in the thick-film region of the semiconductor layer, and the FDSOI LNA device and FDSOI switch device are located in the thin-film region of the semiconductor layer.

[0012] In some embodiments, the insulator layer is at least about 100 nm thick. In some embodiments, the thin-film region of the semiconductor layer is at least about 5 nm thick and less than or equal to about 50 nm thick. In some further embodiments, the thick-film region of the semiconductor layer is at least about 50 nm thick and less than or equal to 180 nm thick.

[0013] In some embodiments, the thin-film region of the semiconductor layer is 1 / 4 the gate length of the FDSOILNA device. In some embodiments, the method further includes incorporating one or more passive devices into the thin-film region of the semiconductor layer. In some embodiments, the method further includes incorporating one or more passive devices into the thick-film region of the semiconductor layer. In some embodiments, reducing the thickness includes using localized thinning.

[0014] Based on a certain number of implementation examples, this disclosure relates to a method for manufacturing a front-end integrated circuit. The method includes forming an insulator layer on the top of a substrate. The method also includes forming a semiconductor layer having a first thickness on the top of the insulator layer. The method also includes forming a thick-film region of the semiconductor layer by increasing the thickness of a portion of the semiconductor layer such that the other portion of the semiconductor layer having the first thickness becomes a thin-film region. The method also includes incorporating a voltage analog circuit into the thick-film region. The method also includes incorporating a low-voltage analog circuit into the thin-film region.

[0015] In some embodiments, the insulator layer is at least about 100 nm thick. In some embodiments, the thin-film region of the semiconductor layer is at least about 5 nm thick and less than or equal to about 50 nm thick. In some further embodiments, the thick-film region of the semiconductor layer is at least about 50 nm thick and less than or equal to 180 nm thick.

[0016] In some embodiments, the high-voltage analog circuit includes a low-dropout regulator. In some embodiments, the high-voltage analog circuit includes a high-voltage power amplifier. In some embodiments, the method further includes integrating a digital circuit into a thin-film region.

[0017] Based on a certain number of implementation examples, this disclosure relates to a method for manufacturing a front-end integrated circuit. The method includes forming an insulator layer on the top of a substrate. The method also includes forming a semiconductor layer having a first thickness on the top of the insulator layer. The method also includes forming a thin-film region of the semiconductor layer by reducing the thickness of a portion of the semiconductor layer such that the other portion of the semiconductor layer having the first thickness becomes a thick-film region. The method also includes incorporating a radio frequency (RF) device into the thick-film region. The method also includes incorporating an analog or digital circuit into the thin-film region.

[0018] In some embodiments, the insulator layer is at least about 100 nm thick. In some embodiments, the thin-film region of the semiconductor layer is at least about 5 nm thick and less than or equal to about 50 nm thick. In some further embodiments, the thick-film region of the semiconductor layer is at least about 50 nm thick and less than or equal to 180 nm thick.

[0019] In some embodiments, the RF device in the thick film region includes a power amplifier (PA) device. In some further embodiments, the PA device includes a partially depleted silicon-on-insulator (PDSOI) PA device.

[0020] In some embodiments, the digital circuit includes logic gates.

[0021] For the purpose of summarizing the present disclosure, certain aspects, advantages, and novel features are described herein. It should be understood that not all of such advantages are necessarily achieved in accordance with any particular embodiment. That is, the disclosed embodiments can be implemented in a manner that achieves or optimizes one or a group of the advantages taught herein without necessarily achieving other advantages that may be taught or suggested herein.

Brief Description of the Drawings

[0022] [Figure 1A] Shows a fully integrated front-end integrated circuit (FEIC) fabricated using silicon-on-insulator (SOI) processing technology. [Figure 1B] Shows other fully integrated FEICs fabricated using SOI processing technology. [Figure 2A] Shows exemplary FEICs at different stages in the manufacturing process. [Figure 2B] Shows exemplary FEICs at different stages in the manufacturing process. [Figure 2C] Shows exemplary FEICs at different stages in the manufacturing process. [Figure 2D] Shows exemplary FEICs at different stages in the manufacturing process. [Figure 3A] Shows variations of the integration process of the FEIC described in relation to FIGS. 2A-2D. [Figure 3B] Shows variations of the integration process of the FEIC described in relation to FIGS. 2A-2D. [Figure 3C] Shows variations of the integration process of the FEIC described in relation to FIGS. 2A-2D. [Figure 3D] Shows variations of the integration process of the FEIC described in relation to FIGS. 2A-2D. [Figure 4A] Shows other exemplary FEICs at different stages in the manufacturing process. [Figure 4B] Shows other exemplary FEICs at different stages in the manufacturing process. [Figure 4C] This shows other exemplary FEICs at different stages in the manufacturing process. [Figure 4D] This shows other exemplary FEICs at different stages in the manufacturing process. [Figure 5A] Figures 4A to 4D show modified versions of the FEIC integration process. [Figure 5B] Figures 4A to 4D show modified versions of the FEIC integration process. [Figure 5C] Figures 4A to 4D show modified versions of the FEIC integration process. [Figure 5D] Figures 4A to 4D show modified versions of the FEIC integration process. [Figure 6A] This document describes how to incorporate an integrated FEIC comprising a partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) device, a fully depleted silicon-on-insulator (FDSOI) low-noise amplifier (LNA) device, and an FDSOI switch device. [Figure 6B] This document describes how to incorporate an integrated FEIC comprising a partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) device, a fully depleted silicon-on-insulator (FDSOI) low-noise amplifier (LNA) device, and an FDSOI switch device. [Figure 7A] An additional method for incorporating an integrated FEIC comprising a PDSOIPA device, an FDSOILNA device, and an FDSOI switch device is shown. [Figure 7B] An additional method for incorporating an integrated FEIC comprising a PDSOIPA device, an FDSOILNA device, and an FDSOI switch device is shown. [Modes for carrying out the invention]

[0023] The headings provided herein are for convenience only, if any, and do not necessarily affect the scope or meaning of the claimed subject matter. overview

[0024] A semiconductor front-end integrated circuit (FEIC) is a single semiconductor die that incorporates the functions of a front-end module (FEM). It is desirable to advance the technological foundation of FEICs to meet the ever-increasing demands for higher performance, miniaturization, and lower costs. Analog circuits typically drive the performance of chips used in FEICs. Analog circuits include elements such as low-noise amplifiers (LNAs), switches, power amplifiers (PAs), passive devices, analog circuits (e.g., level shifters, adders, current mirrors, etc.), digital circuits (e.g., logic gates), regulators (e.g., low-dropout modifiers), and charge pumps. A fully integrated radio frequency (RF) FEIC, as the terminology used herein suggests, incorporates elements for both transmission and reception on a single die. These elements may include PAs, LNAs, and switches, as well as appropriate passive devices, analog and digital circuits, regulators, etc.

[0025] A promising means for creating fully integrated RFFEICs includes silicon-on-insulator (SOI) process technology. SOI involves fabricating silicon semiconductor devices on a layered silicon-insulator-silicon substrate, improving performance by reducing parasitic capacitance within the device. SOI process technology enables improved RFFEIC characteristics (e.g., high bandwidth, low noise figure (NF) LNA performance, high linearity, power efficiency, small package footprint, low insertion loss, etc.). SOI may differ from bulk process technology, which involves extending doped silicon wells deep into the substrate and not stopping at an insulator layer above the substrate layer. Bulk transistors or devices include devices in which the source and drain are incorporated into a silicon substrate and dopants are added to the substrate to adjust its conductivity. As device sizes decrease (e.g., below approximately 28 nm), bulk transistors become increasingly complex, making it advantageous to construct transistors using SOI technology.

[0026] SOI structures include a silicon film (e.g., crystalline silicon) separated from a bulk substrate by a thin layer of an insulator, or insulator (e.g., embedded oxide, or BOX). The BOX layer is configured to improve isolation, reduce short-channel effects, reduce leakage current, and improve switching speed, at least partially due to a smaller drain-body capacitance. In SOI wafers, the insulator is typically a thermal silicon oxide (SiO2) layer, and the substrate is a silicon wafer. Depending on the application, the thickness of the silicon film is variable (e.g., from less than approximately 50 nm to several tens of micrometers). Similarly, the thickness of the BOX is also variable depending on the application (e.g., from several tens of nanometers to several micrometers). SOI manufacturing technologies include oxygen-injected separation (SIMOX), bond-etch-back SOI (BESOI), epitaxial layer transfer (ELTRAN®), NANOCLEAVE®, SMARTCUT®, etc.

[0027] SOI technology can be implemented using complementary metal-oxide-semiconductor (CMOS) technology. SOICMOS involves constructing a metal-oxide-semiconductor field-effect transistor (MOSFET) on a thin layer of semiconductor (e.g., silicon or germanium). The semiconductor thin layer is isolated from the substrate by an insulator layer (e.g., a buried oxide), electrically insulating the device from the underlying semiconductor substrate and from each other. The thickness of the insulator layer of an SOI device can be any thickness from about 5 nm to about 400 nm, and the thickness of the semiconductor film can be any thickness from about 5 nm to about 24 nm.

[0028] MOSFETs on an SOI wafer include a channel depletion layer between the source and drain. SOI devices can be classified into two types depending on the degree of channel depletion compared to the thickness of the silicon film: partially depleted SOI (PDSOI) devices and fully depleted SOI (FDSOI) devices. PDSOI devices include devices in which the silicon film is thicker than the maximum gate depletion layer width, and such devices exhibit a floating body effect. FDSOI devices include devices in which the silicon film is thin enough that the entire film is depleted before reaching a threshold condition.

[0029] FDSOI devices include an ultrathin layer of insulator (embedded oxide, or BOX) positioned on the top of the substrate, with an ultrathin silicon film used to form the transistor channel. FDSOI devices typically use undoped or lightly doped channels. The thin silicon layer is typically about 5 nm to 50 nm thick, or about 1 / 4 of the gate length. In addition, the insulating BOX layer may be thick (e.g., about 100 nm to 400 nm) or ultrathin (e.g., about 5 nm to 50 nm). In the case of FDSOI devices, the silicon layer beneath the gate insulator is thin enough to completely deplete the moving charge carriers, so such devices are "fully depleted." In other words, the depletion region reaches the embedded oxide during the switching of the FDSOI device from the off state to the on state.

[0030] In FDSOI devices, the semiconductor film is made extremely thin so that the depletion region covers the entire film. In FDSOI devices, the depletion charge supported by the gate oxide (GOX) is less than in bulk devices, resulting in an increase in inversion charge and thus a faster switching speed. The depletion charge limit imposed by the BOX suppresses depletion capacitance, which in turn leads to a substantial reduction in fluctuations below the threshold. As a result, FDSOIMOSFETs can operate with low gate biases, leading to low-power operation.

[0031] PDSOI devices include a thicker silicon layer on top of the BOX layer than FDSOI devices. The top silicon layer is typically about 50 nm to 180 nm thick. The silicon beneath the channel partially depletes the mobile charge carriers, resulting in a "partially depleted" device. The BOX layer is typically about 100 nm to 400 nm thick.

[0032] Due to its high integration capability and low cost, CMOS technology is one of the most promising candidates for fully integrated FEICs. Typically, the most difficult devices to construct using CMOS technology are high-power PA devices, which traditionally use lateral diffusion MOSFETs (LDMOS) or extended drain MOSFETs (EDMOS) for high-power applications at low frequencies. LDMOS and EDMOS PAs are typically easier to construct using bulk or thick-film SOI technology. However, difficulties arise when the device size is small (e.g., approximately 28 nm or less). For example, because the silicon film is very thin, manufacturing typically requires etching through the BOX layer on the wafer. As a result, the resulting PA becomes more comparable to a thick-film or bulk device than an FDSOI device. Existing CMOS technologies, whether bulk CMOS, thick-film SOICMOS, or thin-film SOICMOS, are all disadvantages in integrating switches and LNAs with LDMOS PAs, or in the high cost of integrating these elements.

[0033] Therefore, to address these and other issues, a SOI-based technology platform is described herein that provides a fully integrated FEIC including switches, LNAs, and PAs. The PA is incorporated into the thick-film region of the integrated circuit to form a PDSOOIPA, and the switches and LNAs are incorporated into the thin-film region of the integrated circuit to form FDSOI switches and LNAs. The resulting fully integrated FEIC includes a PDSOOIPA with FDSOI switches and LNAs. Passive components can be incorporated into the thick-film region, the thin-film region, or both. In some implementation examples, the FEIC includes one or more power amplifiers in the thick-film region and RF circuits in the thin-film region. In some implementation examples, the FEIC includes high-voltage analog circuits in the thick-film region and low-power analog circuits in the thin-film region. In a given implementation example, the FEIC includes RF devices mounted in the thick-film region and analog and / or digital circuits mounted in the thin-film region.

[0034] Attempts to construct a fully integrated FEIC included constructing each device (PA, LNA, switch) using bulk technology, or constructing each device using PDSOI technology. Attempts also included constructing a bulk LDMOSPA with FDSOILNA and switch (e.g., by removing the BOX layer). Attempts also included constructing a PDSOILDMOSPA with LDMOSLNA in a thick-film region and the switch in a local thin-film region.

[0035] In contrast to these attempts, disclosed herein is a fully integrated FEIC comprising an FDSOI switch and an LNA in addition to the PDSOOPA. In the disclosed embodiment, in contrast to the attempts described above, there is no removal of the BOX layer. Instead, the switch and LNA are incorporated into a thin-film region, and the PA is incorporated into a thick-film region. This can be achieved by starting with a thin film, constructing the switch and LNA, building up a thick-film region (e.g., using selective epitaxial growth or SEG), and incorporating the PA into the built-up thick-film region. This can also be achieved by starting with a thick film, constructing the PA into the thick-film region, obtaining a thin-film region using local thinning, and constructing the switch and LNA into the formed thin-film region. This can also be achieved by preparing the BOX layer and thin-film layer, obtaining a thick-film region by increasing the thickness of a portion of the thin-film region, and then constructing the FDSOINA device and switch device into the thin-film region and the PDSOOPA device into the thick-film region. This can also be achieved by preparing a BOX layer and a thick film layer, reducing the thickness of a portion of the thick film region to create a thin film region, and then constructing the PDSOIPA device in the thick film region and the FDSOILNA device and switch device in the thin film region.

[0036] In other words, the disclosed FEIC is a fully integrated CMOS front-end integrated circuit having improved performance and low cost. Some embodiments of the resulting structure of the disclosed FEIC include a thin-film region having one or more FDSOI switches and one or more FDSOILNAs, and a thick-film region having one or more PDSOIPAs (LDMOSPA or EDMOSPAs). In the disclosed FEIC, passive components can be incorporated into the thick-film region (with PAs), the thin-film region (with switches and LNAs), or both. In some embodiments, the disclosed FEIC includes one or more devices in the thick-film region forming a high-voltage analog circuit which may include a high-voltage power amplifier and / or a low-dropout regulator. In some embodiments, the disclosed FEIC includes an RF device (e.g., PA) in the thick-film region and a low-voltage analog circuit and / or digital circuit in the thin-film region.

[0037] Advantageously, the disclosed FEIC reduces parasitic effects on the PA compared to bulk packaging. This results in the FEIC's high active device performance and other ultimate performance advantages. Another advantage of the disclosed FEIC is that there is no need to remove a portion of the embedded oxide layer to construct a device (e.g., PA) in the thick-film region. Therefore, the integrated circuit in the thick-film region may have a thicker silicon film or layer compared to a given packaging example of an integrated circuit with FDSOIPA or PA in the thin-film region. As a result, robust active device performance and high passive device performance are present in the thick-film region. This improves the overall performance of the disclosed FEIC.

[0038] Front-end integrated circuit structure

[0039] Figure 1A shows a fully integrated front-end integrated circuit (FEIC) 100a fabricated using silicon-on-insulator (SOI) processing technology. The FEIC 100a includes a substrate 102 (e.g., a handle wafer), an insulator layer 104 (e.g., a buried oxide (BOX)) on the substrate 102, and an active device layer or silicon layer 106 (e.g., a silicon film such as crystalline silicon) on top of the insulator layer 104. In some embodiments, the BOX layer 104 may have a thickness of about 100 nm to about 400 nm. The silicon layer 106 forms a thick-film region 108a and a thin-film region 108b. In some embodiments, the thick-film region 108a may have a thickness of about 50 nm to about 180 nm. In some embodiments, the thin-film region 108b may have a thickness of about 5 nm to about 50 nm. In a given implementation example, the thick film region 108a is at least about twice as thick as the thin film region 108b, at least about 2.5 times thick and / or about 20 times thick or less, or at least about 5 times thick and / or about 15 times thick or less.

[0040] A partially depleted SOI (PDSOI) power amplifier (PA) device 110 is formed in the thick film region 108a. A fully depleted SOI (FDSOI) low-noise amplifier (LNA) device 120 and an FDSOI switch device 130 are formed in the thin film region 108b. The PDSOIPA device 110, the FDSOILNA device 120, and the FDSOI switch device 130 may each include an n-MOSFET and / or a p-MOSFET.

[0041] The PDSOIPA device 110 includes a MOSFET having a gate structure comprising a gate conductor 112 (e.g., polysilicon) and a gate insulator 119 (e.g., oxide). One or more spacers may also be used. The PDSOIPA device 110 includes a source diffusion section 114 and a drain diffusion section 116. In a given implementation example, the source diffusion section 114 and / or the drain diffusion section 116 may penetrate deep into the silicon layer 106 to reach or nearly reach the insulator layer 104. In some embodiments, the drain diffusion section 116 can be configured such that the PDSOIPA device 110 is a lateral diffusion MOSFET (LDMOS) or an extended drain MOSFET (EDMOS). The PDSOIPA device 110 includes a channel 118 beneath the gate insulator 119 between the source diffusion section 114 and the drain diffusion section 116. The channel 118 can be doped to achieve target performance characteristics. The channel 118 can be configured to have a thickness such that the depletion layer partially covers the space beneath the gate insulator 119 when the PDSOOPA device 110 is in the ON state. Thus, the PDSOOPA device 110 becomes partially depleted, at least in part, due to being incorporated into the thick film region 108a.

[0042] The FDSOILNA device 120 is similar to the PDSOOPA device 110 in that it includes a gate comprising a gate conductor 122 and a gate insulator 129, a source diffusion section 124, a drain diffusion section 126, and a channel 128 beneath the gate insulator 129. In the FDSOILNA device 120, the channel 128 can be configured to have a thickness such that the depletion layer covers the space beneath the gate insulator 129 when the FDSOILNA device 120 is in the ON state. Thus, the FDSOILNA device 120 becomes fully depleted, at least in part, due to being incorporated into the thin-film region 108b. In some embodiments, the channel 128 is either undoped or lightly doped.

[0043] The FDSOI switch device 130 is similar to the FDSOLNA device 120 in that it includes a gate comprising a gate conductor 132 and a gate insulator 139, a source diffusion section 134, a drain diffusion section 136, and a channel 138 beneath the gate insulator 139. The channel 138 can be configured to have a thickness such that the depletion layer covers the space beneath the gate insulator 139 when the FDSOI switch device 130 is in the ON state. Thus, the FDSOI switch device 130 becomes fully depleted, at least in part, due to being incorporated into the thin-film region 108b. In some embodiments, the channel 138 is either undoped or lightly doped.

[0044] FEIC100a includes a substrate 102, an insulator layer 104 on the top of the substrate, and a semiconductor layer 106 on the top of the insulator layer 104. The semiconductor layer 106 forms a thin-film region 108b and a thick-film region 108a. The thin-film region 108b includes one or more FDSOILNA devices 120 and one or more FDSOI switch devices 130. The thick-film region 108a includes one or more PDSOIPA devices 110. That is, the semiconductor layer 106 is thinner in the LNA device 120 and the switch device 130 than in the PA device 110.

[0045] Figure 1B shows another fully integrated FEIC100b fabricated using SOI processing technology. FEIC100b has the same structure as FEIC100a, including a substrate 102, an insulator layer 104 on the top of the substrate, and a semiconductor layer 106 on the top of the insulator layer 104. The semiconductor layer 106 forms a thin-film region 108b and a thick-film region 108a. The thick-film region 108a includes various thick-film devices 140. In some embodiments, the thick-film devices 140 may include analog circuits (e.g., high-voltage PAs, LDOs, high-voltage breakdown units, ESD protection units, charge pumps, high-power switches, power control units, etc.). The thin-film region 108b includes various thin-film devices 150. In some embodiments, the thin-film devices 150 may include analog circuits (e.g., low-power analog circuits, level shifters, adders, current mirrors, etc.) and / or digital circuits (e.g., logic gates). In other words, the semiconductor layer 106 is further thinner in the thin-film device 150 region than in the thick-film device 140 region. In some embodiments, the thick-film region 108a of FEIC 100b is used for high-voltage analog circuits. In such embodiments, the thin-film region 108b may be used for RF circuits. In a given embodiment, the thick-film region 108a of FEIC 100b is used for RF circuits (e.g., PAs), and the thin-film region 108b is used for analog circuits (e.g., low-power analog circuits) and / or digital circuits (e.g., logic gates).

[0046] Figures 2A, 2B, 2C, and 2D show exemplary FEIC200 at different stages in the manufacturing process. FEIC200 is similar to FEIC100a and 100b in that it includes a substrate 102 and an insulator layer 104 on the substrate 102. The substrate 102 may be a silicon support wafer or a handle wafer. The insulator layer 104 may be a buried oxide such as silicon dioxide. In some embodiments, the insulator layer 104 may have a thickness of about 100 nm to about 400 nm.

[0047] In Figure 2A, the FEIC200 includes an active layer or silicon layer 206 of substantially uniform thickness. The silicon layer 206 may be a silicon film. The thickness of the silicon layer 206 may be suitable for constructing an FDSOI device. For example, the thickness of the silicon layer 206 may be about 5 nm to about 50 nm, or about 1 / 4 of the gate length of the FDSOI device incorporated into the silicon layer 206.

[0048] In Figure 2B, FEIC200 includes an FDSOILNA device 220 and an FDSOI switch device 230. The FDSOILNA device 220 is similar to the FDSOILNA device 120 in Figure 1A in that it includes a gate structure with a gate conductor 222 on a gate insulator 229, a source diffusion section 224, a drain diffusion section 226, and a channel 228. The FDSOI switch device 230 is similar to the FDSOI switch device 130 in Figure 1A in that it includes a gate structure with a gate conductor 232 on a gate insulator 239, a source diffusion section 234, a drain diffusion section 236, and a channel 238. It should be understood that although a single FDSOILNA device 220 is shown, multiple FDSOILNA devices may be incorporated into the silicon layer 206. Similarly, although a single FDSOI switch device 230 is shown, multiple FDSOI switch devices may be incorporated into the silicon layer 206.

[0049] In Figure 2C, the FEIC200 includes a build-up region 207 that increases the thickness of the silicon film 206 in the target area. The build-up region 207 can be constructed on the silicon film 206 using any suitable process. One example of a process for building up a silicon film is selective epitaxial growth (SEG). For example, SEG can be used to build up a portion of the silicon layer 206 that does not contain any FDSOI devices.

[0050] This build-up portion 207 provides a thick-film region 208a of the silicon layer 206 that has a greater thickness than the thin-film region 208b of the silicon layer 206. The thickness of the resulting thick-film region 208a can be suitable for constructing a PDSOI device. For example, the thickness of the silicon layer 206 in the thick-film region 208a may be approximately 50 nm to 180 nm. This build-up portion 207 is shown as a shaded area on the top of the silicon layer 206, but it should be understood that the resulting increase in the thickness of the silicon film 206 does not necessarily result in an additional layer on the top of the silicon layer 206. Rather, the additional thickness of the build-up portion 207 represents an increase in the thickness of the silicon layer 206 itself.

[0051] In Figure 2D, FEIC200 includes a PDSOOPA device 210 incorporated in the thick-film region 208a of the silicon layer 206. The PDSOOPA device 210 is similar to the PDSOOPA device 110 in Figure 1A in that it includes a gate structure with a gate conductor 212 on a gate insulator 219, a source diffusion region 214, a drain diffusion region 216, and a well 218. The PDSOOPA device 210 may be an LDMOSPA device or an EDMOSPA device. It should be understood that although a single PDSOOPA device 210 is shown, multiple PDSOOPA devices may be incorporated into the silicon layer 206. FEIC200 may include passive devices in either or both of the thick-film region 208a and the thin-film region 208b.

[0052] For example, a suitable method for manufacturing FEIC200 includes forming an insulator layer 104 on the top of a substrate 102. The method includes forming a semiconductor layer 206 on the top of the insulator layer 104. The method includes constructing an FDSOILNA device 220 on the semiconductor layer 206. The method includes constructing an FDSOI switch device 230 on the semiconductor layer 206. The method includes increasing the thickness of a portion of the semiconductor layer 206 to form a thick film region 208b of the semiconductor layer 206. The method includes incorporating the PDSOIPA device 210 into the thick film region 208a of the semiconductor layer 206 such that the FDSOILNA device 220 and the FDSOI switch device 230 are located in the thin film region 208b of the semiconductor layer 206, and the PDSOIPA device 210 is located in the thick film region 208a of the semiconductor layer 206.

[0053] Figures 3A, 3B, 3C, and 3D show modified versions of the process for constructing the FEIC200 described in relation to Figures 2A-2D. In these modifications, the FEIC200 begins with a thin silicon layer 206, as shown in Figure 3A. In Figure 3B, the build-up region 207 of the FEIC is formed before the device is formed on the thin-film region 208b. The build-up portion 207 can be constructed on the silicon film 206 using any suitable process such as SEG. This build-up portion 207 results in a thick-film region 208a of the silicon layer 206 that has a greater thickness than the thin-film region 208b of the silicon layer 206. The resulting thickness of the thick-film region 208a may be suitable for constructing PDSOI devices of, for example, about 50 nm to about 180 nm. This build-up portion 207 is shown as a shaded portion on the top of the silicon layer 206, but it should be understood that the resulting increase in the thickness of the silicon film 206 does not necessarily result in an additional layer on the top of the silicon layer 206. Rather, the additional thickness of the build-up portion 207 represents an increase in the thickness of the silicon layer 206 itself. Once the thick film region 208a is formed, devices 210, 220, and 230 can be incorporated into the thick film region 208a and the thin film region 208b, as described herein with reference to Figures 2B and 2D. Figure 3C shows that source diffusion regions 214, 224, and 234, and drain diffusion regions 216, 226, and 236 can be formed together with channels 218, 228, and 238. Figure 3D shows that, in some embodiments, after these are formed, gate insulators 219, 229, and 239 can be formed between the corresponding source diffusion regions and drain diffusion regions. Above these corresponding gate insulators, gate conductors 212, 222, and 232 are formed. In a given implementation example, the mask used to construct the gate insulator 219 may be shared with the mask used for the gate insulators 229 and 239. Similarly, the mask used to form the gate conductor 212 may be the same as the mask used for the gate conductors 222 and 232.

[0054] Figures 4A, 4B, 4C, and 4D show other exemplary FEIC300 at different stages in the manufacturing process. FEIC300 is similar to FEIC100a in that it includes a substrate 102 and an insulator layer 104 on the substrate 102. The substrate 102 may be a silicon support wafer or a handle wafer. The insulator layer 104 may be a buried oxide such as silicon dioxide. In some embodiments, the insulator layer 104 may have a thickness of about 100 nm to about 400 nm.

[0055] In Figure 4A, the FEIC300 includes an active layer or silicon layer 306 of substantially uniform thickness. The silicon layer 306 may be a silicon film. The thickness of the silicon layer 306 may be appropriate for constructing a PDSOI device. For example, the thickness of the silicon layer 306 may be approximately 50 nm to approximately 180 nm.

[0056] In Figure 4B, FEIC300 includes a PDSOOPA device 310 incorporated into the silicon layer 306. The PDSOOPA device 310 is similar to the PDSOOPA device 110 in Figure 1A in that it includes a gate structure with a gate conductor 312 on a gate insulator 319, a source diffusion section 314, a drain diffusion section 316, and a well 318. The PDSOOPA device 310 may be an LDMOSPA device or an EDMOSPA device. It should be understood that although a single PDSOOPA device 310 is shown, multiple PDSOOPA devices may be incorporated into the silicon layer 306.

[0057] In Figure 4C, FEIC300 includes a region 307 representing a portion of the silicon layer 306 that is being removed to reduce the thickness of the silicon layer 306 in the target area. The removal portion 307 can be removed from the silicon film 306 using any suitable process. One example of a process for removing a portion of the silicon film is local thinning. For example, local thinning can be used to remove a portion of the silicon layer 306 that does not contain any PDSOI devices.

[0058] This removal section 307 yields a thick film region 308a of the silicon layer 306, which has a greater thickness than the thin film region 308b of the silicon layer 306. The thickness of the resulting thin film region 308b can be suitable for constructing an FDSOI device. For example, the thickness of the silicon layer 306 in the thin film region 308b may be about 5 nm to about 50 nm, and may be about 1 / 4 of the gate length of the FDSOI device incorporated into the thin film region 308b.

[0059] In Figure 4B, the FEIC 300 includes an FDSOILNA device 320 and an FDSOI switch device 330. The FDSOILNA device 320 is similar to the FDSOILNA device 120 in Figure 1A in that it includes a gate structure with a gate conductor 322 on a gate insulator 329, a source diffusion section 324, a drain diffusion section 326, and a channel 328. The FDSOI switch device 330 is similar to the FDSOI switch device 130 in Figure 1A in that it includes a gate structure with a gate conductor 332 on a gate insulator 339, a source diffusion section 334, a drain diffusion section 336, and a channel 338. It should be understood that although a single FDSOILNA device 320 is shown, multiple FDSOILNA devices may be incorporated into the silicon layer 306. Similarly, although a single FDSOI switch device 330 is shown, multiple FDSOI switch devices may be incorporated into the silicon layer 306. FEIC300 may include passive devices in either or both of the thick film region 308a and the thin film region 308b.

[0060] For example, a suitable method for manufacturing FEIC300 includes forming an insulator layer 104 on the top of a substrate 102. The method includes forming a semiconductor layer 306 on the top of the insulator layer 104. The method also includes constructing a PDSOOPA device 310 on the semiconductor layer 306. The method also includes reducing the thickness of a portion of the semiconductor layer 306 to form a thin-film region 308b of the semiconductor layer 306. The method also includes constructing an FDSOILNA device 320 on the thin-film region 308b of the semiconductor layer 306. The method also includes constructing an FDSOI switch device 330 on the thin-film region 308b of the semiconductor layer 306. The PDSOOPA device 310 is located in the thick-film region 308a of the semiconductor layer 306, while the FDSOILNA device 320 and the FDSOI switch device 330 are located in the thin-film region 308b of the semiconductor layer 306.

[0061] Figures 5A, 5B, 5C, and 5D show modified versions of the process for constructing the FEIC300 described in relation to Figures 4A-4D. In this modification, the FEIC300 begins with a thin silicon layer 306, as shown in Figure 5A. In Figure 5B, the FEIC300 includes a region 307 representing a portion of the silicon layer 306 that has been removed to reduce the thickness of the silicon layer 306 in a target area. The removed portion 307 can be removed from the silicon film 306 using any suitable process, such as local thinning. This removed portion 307 results in a thicker region 308a of the silicon layer 306 having a greater thickness than the thin-film region 308b of the silicon layer 306. The thickness of the resulting thin-film region 308b may be suitable for constructing an FDSOI device. For example, it may be about 5 nm to about 50 nm, or about 1 / 4 of the gate length of the FDSOI device incorporated into the thin-film region 308b. Once the thin-film region 308b is formed, devices 310, 320, and 330 can be incorporated into the thick-film region 308a and the thin-film region 308b, as described herein with reference to Figures 4B and 4D. Figure 5C shows that source diffusion regions 314, 324, and 334, and drain diffusion regions 316, 326, and 336 can be formed together with channels 318, 328, and 338. Figure 5D shows that, in some embodiments, after these are formed, gate insulators 319, 329, and 339 can be formed between the corresponding source diffusion regions and drain diffusion regions. Above these corresponding gate insulators, gate conductors 312, 322, and 332 are formed. In a given implementation example, the mask used to construct gate insulator 319 may be shared with the mask used for gate insulators 329 and 339. Similarly, the mask used to form gate conductor 312 may be shared with the mask used for gate conductors 322 and 332.

[0062] Manufacturing of front-end integrated circuits

[0063] Figure 6A shows a method 600 for constructing an integrated front-end integrated circuit (FEIC) comprising a partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) device, a fully depleted silicon-on-insulator (FDSOI) low-noise amplifier (LNA) device, and an FDSOI switch device. Figures 2A to 2D show examples of FEICs fabricated in accordance with the steps of method 600.

[0064] In block 605, a substrate comprising a buried oxide (BOX) layer and a thin-film silicon layer is prepared. The substrate comprising the BOX layer and thin-film silicon may take the form of, for example, a silicon-on-insulator (SOI) wafer. The preparation of the SOI wafer may be carried out in a separate process such that the steps in block 605 involve receiving or providing an SOI wafer rather than requiring the manufacture of the SOI wafer. The preparation may also include any steps taken to prepare the structure so as to enable the construction of an active device on the thin-film silicon layer. Figure 2A shows an example of FEIC in block 605.

[0065] The substrate may be a handle wafer. The BOX layer may be any suitable insulator, such as thermal silicon oxide (SiO2). The thin-film silicon layer may be a thin film deposited on top of the BOX, such as crystalline silicon. The thickness of the BOX layer may be any thickness from about 5 nm to about 400 nm, and at least about 100 nm and / or about 200 nm or less. This is sometimes referred to as a thick BOX layer in comparison to a thin or ultrathin BOX layer, which is typically about 5 nm to about 50 nm. The thickness of the thin-film silicon layer may be about 5 nm to about 50 nm, or about 1 / 4 of the gate length of the active device incorporated into the thin-film silicon layer. SOI wafers can be prepared using any suitable process, including oxygen-induced separation (SIMOX), bond-etch-back SOI (BESOI), epitaxial layer transfer (ELTRAN®), NANOCLEAVE®, SMARTCUT®, etc.

[0066] In block 610, one or more FDSOILNA devices and one or more FDSOI switch devices are incorporated into a thin silicon layer. These active devices may be constructed without doping channels between their source diffusion and drain diffusion portions, or with lightly doped channels. The thickness of the thin silicon layer may be such that the channels are completely depleted when the active devices are in the ON state. Figure 2B shows an example of FEIC in block 610.

[0067] In block 615, the thickness of the thin-film silicon region is increased, resulting in a thick-film region that does not contain any FDSOI active devices. Therefore, the thin-film silicon region that has not undergone the process of thickening the silicon film layer may be referred to as the thin-film region of the silicon layer. This thin-film region contains one or more FDSOILNA devices and one or more FDSOI switch devices. The thickness of the resulting thick-film region may be approximately 50 nm to approximately 180 nm. Figure 2C shows an example of FEIC in block 615.

[0068] To increase the thickness of a silicon layer, epitaxial deposition processes or epitaxy can be used. These processes can be used to grow a silicon film (e.g., crystalline silicon) or a layer of silicon (e.g., crystalline silicon) on a substrate. Selective epitaxial growth (SEG) is one example of a process that can be used to grow silicon on exposed silicon regions of a silicon film. Regions where silicon growth is undesirable can be masked by a dielectric film, typically silicon dioxide or silicon nitride. Epitaxial growth may involve the condensation of a liquid or gas precursor to form a film on the substrate. The gas precursor can be obtained, for example, by chemical vapor deposition and / or laser ablation.

[0069] In block 620, one or more PDSOOPA devices are incorporated into the thick film region. The one or more PDSOOPA devices may be LDMOS and / or EDMOSPA devices. The thickness of the thick silicon film may be such that the channels are partially depleted when the active device is ON. Figure 2D shows an example of FEIC in block 620. Optionally, in block 625, passive devices may be constructed in the thin film region, the thick film region, or both the thin film and thick film regions.

[0070] Method 600 offers several advantages. For example, associated parasitic contamination is less in LDMOSPA and EDMOSPA devices on SOI compared to implementation examples using bulk technology. This improves the performance of the active device. As another example, the embedded oxide layer may be a thick BOX layer (rather than an ultrathin layer) to achieve the desired performance characteristics. This results in robust active device performance, high power capacity, and improved passive device performance. As yet another example, the resulting FEIC has the advantages of both FDSOI active devices (e.g., switch devices and LNA devices) and PDSOI active devices (e.g., PA devices).

[0071] Figure 6B shows Method 650 for constructing an integrated FEIC comprising a PDSOOPA device, an FDSOILNA device, and an FDSOI switch device. Figures 3A to 3D show examples of FEICs manufactured according to the steps of Method 650. It should be noted that Method 650 is substantially the same as Method 600, which was described with reference to Figure 6A, and therefore shares the advantages of Method 600. The difference between Method 600 and Method 650 lies in the order of the steps of the method. Method 650 prepares the substrate before constructing the device. More specifically, Method 650 reverses the order of constructing the device in a thin-film region and increasing the thickness of a portion of the thin-film region to obtain a thick-film region. Therefore, the description of Method 650 is simplified in order to give details of Method 650 by relying on the description of Method 600.

[0072] In block 655, a substrate comprising a BOX layer and a thin-film silicon layer is prepared, which may take the form of, for example, an SOI wafer. Figure 3A shows an example of FEIC in block 655.

[0073] In block 660, the thickness of the thin-film silicon region is increased, resulting in a thick-film region. Therefore, the thin-film silicon region that has not undergone the process of thickening the silicon film layer may be referred to as the thin-film region of the silicon layer. The thickness of the resulting thick-film region may be approximately 50 nm to 180 nm. Figure 3B shows an example of FEIC in block 660.

[0074] In block 665, one or more FDSOILNA devices and one or more FDSOI switch devices are incorporated into a thin film region of the silicon layer. These active devices may be constructed without doping channels between their source diffusion and drain diffusion regions, or with lightly doped channels. The thickness of the thin film silicon may be such that the channels are fully depleted when the active devices are ON. In block 670, one or more PDSOOPA devices are incorporated into a thick film region. One or more PDSOOPA devices may be LDMOS and / or EDMOSPA devices. The thickness of the thick film silicon may be such that the channels are partially depleted when the active devices are ON. Optionally, in block 675, passive devices may be incorporated into the thin film region, the thick film region, or both the thin film region and the thick film region.

[0075] Figures 3C and 3D show examples of FEIC in blocks 665 and 670. Although blocks 665 and 670 show that the device is incorporated into the thick film region after being incorporated into the thin film region, it should be understood that other implementations of Method 650 include partially incorporating the device into the thin film region and the thick film region (as shown in Figure 3C, for example) and then completing the device in both regions (as shown in Figure 3D, for example). In some implementations, a shared mask may be used to complete the device, as described herein. This can also be applied to Method 600, which is completed in a separate step after the device has been partially incorporated into the thin film region and the thick film region, in which a shared mask may be used in some examples.

[0076] Figure 7A shows Method 700 for constructing an integrated FEIC comprising a PDSOOPA device, an FDSOILNA device, and an FDSOI switch device. Figures 4A to 4D show examples of FEICs manufactured corresponding to the steps of Method 500.

[0077] In block 505, a substrate having a buried oxide (BOX) layer and a thick silicon layer is prepared. The substrate having the BOX layer and thick silicon may be in the form of, for example, a silicon-on-insulator (SOI) wafer. The preparation of the SOI wafer may be carried out in a separate process such that the steps in block 505 involve receiving or providing an SOI wafer rather than requiring the manufacture of an SOI wafer. The preparation may also include any steps taken to prepare the structure so that an active device can be constructed on the thick silicon layer. Figure 4A shows an example of an FEIC in block 505.

[0078] The substrate may be a handle wafer. The BOX layer may be any suitable insulator, such as thermal silicon oxide (SiO2). The thin-film silicon layer may be a thin film deposited on top of the BOX, such as crystalline silicon. The thickness of the BOX layer may be any thickness from about 5 nm to about 400 nm, and at least about 100 nm and / or about 200 nm or less. This is sometimes referred to as a thick BOX layer in comparison to a thin or ultrathin BOX layer, which is typically about 5 nm to about 50 nm. The thickness of the thick-film silicon layer may be about 50 nm to about 180 nm. SOI wafers can be prepared using any suitable process, including separation by injected oxygen (SIMOX), bond etch-back SOI (BESOI), epitaxial layer transfer (ELTRAN®), NANOCLEAVE®, SMARTCUT®, etc.

[0079] In block 510, one or more PDSOOPA devices are incorporated into a thick silicon layer. The one or more PDSOOPA devices may be LDMOS and / or EDMOSPA devices. The thickness of the thick silicon layer may be such that the channel is partially depleted when the active device is ON. Figure 4B shows an example of FEIC in block 510.

[0080] In block 515, the thickness of the thin-film silicon region is reduced, resulting in a thin-film region that does not contain any PDSOI active devices. Therefore, the thick-film silicon region that has not undergone the process of thinning the silicon film layer may be referred to as the thick-film region of the silicon layer. This thick-film region contains one or more PDSOIPA devices. The thickness of the resulting thin-film region may be approximately 5 nm to 50 nm, or typically, approximately 1 / 4 of the gate length of the active device incorporated into the thin-film region. Figure 4C shows an example of FEIC in block 515.

[0081] Thinning of a thick silicon film may involve any suitable process for localized thinning. For example, thinning may include mechanical grinding, chemical mechanical planarization, wet etching, atmospheric downstream plasma dry chemical etching (ADPDCE), etc.

[0082] In block 510, one or more FDSOILNA devices and one or more FDSOI switch devices are incorporated into the thin-film region. These active devices may be constructed without doping channels between their source and drain diffusion regions, or with lightly doped channels. The thickness of the thin-film region may be such that the channels are completely depleted when the active devices are in the ON state. Figure 4D shows an example of FEIC in block 520. Optionally, in block 525, passive devices may be incorporated into the thin-film region, the thick-film region, or both the thin-film and thick-film regions.

[0083] Method 500 offers several advantages. For example, associated parasitic contamination is less pronounced in LDMOSPA and EDMOSPA devices on SOI compared to implementations using bulk technology. This improves the performance of active devices. As another example, the embedded oxide layer may be a thick BOX layer (rather than an ultrathin layer) to achieve the desired performance characteristics. This results in robust active device performance, high power capacity, and improved passive device performance. As yet another example, the resulting FEIC has the advantages of both FDSOI active devices (e.g., switch devices and LNA devices) and PDSOI active devices (e.g., PA devices). Furthermore, Method 500 may be less expensive than Method 400, because thin-film deposition is typically a less expensive process than epitaxial growth or deposition.

[0084] Figure 7B shows Method 750 for constructing an integrated FEIC comprising a PDSOOPA device, an FDSOILNA device, and an FDSOI switch device. Figures 5A to 5D show examples of FEICs manufactured corresponding to the steps of Method 750. It should be noted that Method 750 is substantially the same as Method 700, which was described with reference to Figure 7A, and therefore shares the advantages of Method 700. The difference between Method 700 and Method 750 lies in the order of the steps of the method. Method 750 prepares the substrate before constructing the device. More specifically, Method 750 reverses the order of constructing the device in a thick film region and then reducing the thickness of a portion of the thick film region to obtain a thin film region. Therefore, the description of Method 750 is simplified in order to give details of Method 750 by relying on the description of Method 700.

[0085] In block 755, a substrate comprising a BOX layer and a thick silicon layer is prepared, which may take the form of, for example, an SOI wafer. Figure 5A shows an example of FEIC in block 755.

[0086] In block 760, the thickness of the thick silicon region is reduced, resulting in a thin-film region. Therefore, the region of thick silicon that has not undergone the process of thinning the silicon film layer may be referred to as the thick-film region of the silicon layer. The thickness of the resulting thin-film region may be approximately 5 nm to 50 nm. Figure 5B shows an example of FEIC in block 760.

[0087] In block 765, one or more PDSOOPA devices are incorporated into a thick film region. The one or more PDSOOPA devices may be LDMOS and / or EDMOSPA devices. The thickness of the thick silicon may be such that the channels are partially depleted when the active devices are ON. In block 770, one or more FDSOILNA devices and one or more FDSOI switch devices are incorporated into a thin film region of the silicon layer. These active devices may be constructed without doping channels between their source diffusion and drain diffusion regions, or with lightly doped channels. The thickness of the thin silicon may be such that the channels are fully depleted when the active devices are ON. Optionally, in block 775, passive devices may be incorporated into the thin film region, the thick film region, or both the thin film region and the thick film region.

[0088] Figures 5C and 5D show examples of FEIC in blocks 765 and 770. Although blocks 765 and 770 show that the device is incorporated into the thick film region after being incorporated into the thin film region, it should be understood that other implementations of Method 750 include partially incorporating the device into both the thin film and thick film regions (as shown, e.g., in Figure 5C) and then completing the device in both regions (as shown, e.g., in Figure 5D). In some implementations, a shared mask may be used to complete the device, as described herein. This can also be applied to Method 600, which is completed in a separate step after the device has been partially incorporated into the thin film and thick film regions, in which a shared mask may be used in some examples.

[0089] In addition, it should be understood that the FEIC100b shown in Figure 1B can be prepared using methods 600, 650, 700, and 750. Modifications of methods 600, 650, 700, and 750 involve replacing the steps for preparing specific PDSOOPA devices, FDSOILNA devices, and / or FDSOI switches with those of other circuits described in relation to Figure 1B.

[0090] Additional embodiments and terms

[0091] This disclosure describes a variety of features, and no single feature alone is solely responsible for the benefits described herein. It should be understood that the various features described herein may be combined, modified, or omitted in a manner that is obvious to those skilled in the art. Other combinations and subcombinations not specifically described herein are obvious to those skilled in the art and are intended to form part of this disclosure. Various methods are described herein in relation to various flowchart steps and / or phases. It should be understood that in many cases, given steps and / or phases may be combined together, resulting in multiple steps and / or phases shown in a flowchart being performed as a single step and / or phase. Also, given steps and / or phases may be divided into additional subcomponents that are performed separately. In some examples, the order of steps and / or phases may be rearranged, and given steps and / or phases may be omitted entirely. Furthermore, the methods described herein should be understood to be open-ended, so that additional steps and / or phases may also be performed beyond those illustrated and described herein.

[0092] Unless the context explicitly requires otherwise, throughout the specification and claims, terms such as “includes,” “equipment,” and so on should be interpreted in a comprehensive sense, the opposite of an exclusive or exhaustive sense, i.e., “includes but not limited to.” The term “combined,” as used here, refers to two or more elements that may be directly connected or connected via one or more intermediate elements. Additionally, where used in this application, the terms “here,” “above,” “below,” and similar terms refer to the entire application and not to any particular part of it. Where the context allows, terms in the above detailed description that use singular or plural numbers may also include plural or singular numbers. The terms “or” and “or,” referring to a list of two or more items, cover all of the following interpretations of the term: any of the items in the list, all of the items in the list, and any combination of the items in the list. The term “exemplary” is used here exclusively in the sense of “serving as an example, illustration, or example.” None of the implementation examples described herein, provided as "exemplary," should necessarily be interpreted as being preferable or advantageous to other implementation examples.

[0093] This disclosure is not intended to be limited to the implementations shown herein. Various modifications to the implementations described herein may be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other implementations without departing from the spirit or scope of this disclosure. The teachings of the present invention given herein may be applied to other methods and systems, and are not limited to those described herein. Further embodiments may also be given by combining elements and functions of the various embodiments described herein. Thus, the novel methods and systems described herein may be embodied in various other forms. Furthermore, various omissions, substitutions, and modifications in the forms of the methods and systems described herein may be made without departing from the spirit of this disclosure. The appended claims and equivalents are intended to cover forms or modifications that fall within the scope and spirit of this disclosure.

Claims

1. 1. A front-end integrated circuit comprising: A substrate; an insulator layer on top of the substrate; a semiconductor layer on top of the insulator layer; Including, the semiconductor layer forms a thin film region and a thick film region; the thin film region includes one or more fully depleted silicon-on-insulator (FDSOI) low noise amplifier (LNA) devices and one or more FDSOI switch devices; the thick film region includes one or more partially depleted silicon-on-insulator (PDSOI) power amplifier (PA) devices; the semiconductor layer in the thin film region is at least 5 nm thick and no more than 50 nm thick; The semiconductor layer in the thick film region is at least 50 nm thick and no more than 180 nm thick.

2. The front-end integrated circuit of claim 1 , wherein the insulator layer is at least 100 nm thick.

3. 2. The front-end integrated circuit of claim 1, wherein said insulator layer is a buried oxide layer.

4. 2. The front-end integrated circuit of claim 1, wherein the semiconductor layer in the thin film region is 1 / 4 of a gate length of the one or more FDSOILNA devices.

5. The front-end integrated circuit of claim 1 , wherein the thin-film regions of the semiconductor layer are formed using localized thinning.

6. 2. The front-end integrated circuit of claim 1, wherein the thick region of the semiconductor layer is formed using selective epitaxial growth.