RF switch on high resistive substrate

Abstract

A device includes a semiconductor substrate of a first conductivity type, and a deep well region in the semiconductor substrate, wherein the deep well region is of a second conductivity type opposite to the first conductivity type. The device further includes a well region of the first conductivity type over the deep well region. The semiconductor substrate has a top portion overlying the well region, and a bottom portion underlying the deep well region, wherein the top portion and the bottom portion are of the first conductivity type, and have a high resistivity. A gate dielectric is over the semiconductor substrate. A gate electrode is over the gate dielectric. A source region and a drain region extend into the top portion of the semiconductor substrate. The source region, the drain region, the gate dielectric, and the gate electrode form a Radio Frequency (RF) switch.

Claims

1. A method comprising: performing a first implantation to implant a semiconductor substrate and to form a deep well region, wherein the semiconductor substrate is of a first conductivity type, and wherein in the first implantation, an impurity of a second conductivity type opposite to the first conductivity type is implanted; performing a second implantation to implant the semiconductor substrate, wherein a well region of the first conductivity type is formed over the deep well region, and wherein after the first and the second implantations, the semiconductor substrate comprises: a top portion overlying the well region; and a bottom portion underlying the deep well region, wherein the top portion and the bottom portion are substantially un-implanted in the first and the second implantations; forming a gate dielectric over the top portion of the semiconductor substrate; forming a gate electrode over the gate dielectric; performing a third implantation to implant the top portion of the semiconductor substrate and to form a source region and a drain region, wherein the source region and the drain region are spaced apart from the well region by a remaining part of the top portion of the semiconductor substrate, and wherein the source region, the drain region, the gate dielectric, and the gate electrode form a Radio Frequency (RF) switch configured to operate in a RF frequency range; forming a gate spacer on a sidewall of the gate electrode, wherein the gate spacer has an outer edge in physical contact with a top surface of one of the source region and the drain region; and forming a dielectric layer comprising a portion over and contacting the one of the source region and the drain region, wherein the dielectric layer has a substantially flat top surface extending to the outer edge of the one of the source region and the drain region.

2. The method of claim 1, wherein the forming the RF switch is free from any channel doping implantation.

3. The method of claim 1, wherein the third implantation is performed using an energy between about 2 KeV and about 10 KeV.

4. The method of claim 1 further comprising after the dielectric layer is formed, forming a source/drain silicide having an edge aligned to an edge of the dielectric layer.

5. The method of claim 1 further comprising implanting the top portion of the semiconductor substrate to form a lightly doped drain/source region using an implantation energy between about 2 KeV and about 10 KeV.

6. A method comprising: implanting a semiconductor substrate to form a deep n-well region in the semiconductor substrate; implanting the semiconductor substrate to form a p-well region, wherein the p-well region is over and contacting the deep n-well region; forming a Radio Frequency (RF) switch comprising: a gate dielectric directly over a top portion of the semiconductor substrate; a gate electrode over the gate dielectric; and a source region and a drain region extending into the top portion of the semiconductor substrate, wherein the source region and the drain region are spaced apart from the p-well region by the top portion of the semiconductor substrate; forming an RF passive device overlying and electrically coupled to the RF switch; forming a gate spacer on a sidewall of the gate electrode, wherein the gate spacer comprises an outer edge contacting a top surface of one of the source region and the drain region; forming a dielectric layer comprising a portion over and contacting one of the source region and the drain region, wherein the dielectric layer has an inner edge in physical contact with the outer edge of the gate spacer; and forming an inter-layer dielectric over the dielectric layer.

7. The method of claim 6, wherein the top portion of the semiconductor substrate is overlying the p-well region, and the semiconductor substrate further comprises a bottom portion underlying the deep n-well region, wherein the top portion and the bottom portion are of p-type, and both the top portion and the bottom portion have the resistivity greater than about 5,000 ohm-cm.

8. The method of claim 6, wherein the RF passive device is electrically connected to the gate electrode.

9. The method of claim 6, wherein the forming the RF switch is free from channel doping implantations.

10. The method of claim 6, wherein the source region and the drain region are formed through an implantation using an energy between about 2 KeV and about 10 KeV.

11. The method of claim 6 further comprising, after the forming the dielectric layer, forming a source/drain silicide region having an edge aligned to an edge of the dielectric layer.

12. A method comprising: implanting a semiconductor substrate to form a well region; forming a gate stack overlapping a top portion of the semiconductor substrate, wherein the top portion spaces the gate stack apart from the well region; forming a gate spacer on a sidewall of the gate stack; implanting the top portion of the semiconductor substrate to form a source/drain region using the gate spacer and the gate stack as an implantation mask; depositing a blanket dielectric layer over the gate stack and the source/drain region; patterning the blanket dielectric layer to form a dielectric layer overlapping the source/drain region, wherein the dielectric layer comprises an inner edge contacting an outer edge of the gate spacer; and siliciding a top portion of the source/drain region to form a silicide region over and contacting the source/drain region, wherein an inner edge of the silicide region contacts an outer edge of the dielectric layer.

13. The method of claim 12 further comprising forming a Radio Frequency (RF) passive device, wherein the gate stack is connected to an end of the RF passive device.

14. The method of claim 12, wherein the siliciding is performed using the dielectric layer as a mask to define a location of the inner edge of the silicide region.

15. The method of claim 12, wherein the dielectric layer has a substantially flat top surface, and each of the inner edge and the outer edge of the dielectric layer has a bottom contacting the source/drain region, and the substantially flat top surface of the dielectric layer extends to the outer edge of the dielectric layer.

16. The method of claim 12 further comprising implanting the semiconductor substrate to form a deep well region in the semiconductor substrate, with the deep well region being of a conductivity type opposite to a conductivity type of the well region, wherein a top surface of the deep well region is in contact with a bottom surface of the well region.

17. The method of claim 4, wherein the substantially flat top surface of the dielectric layer is in physical contact with the outer edge of the one of the gate spacer.

18. The method of claim 6, wherein the forming the dielectric layer comprises: forming a blanket dielectric layer over the gate spacer, the gate electrode, the source region, and the drain region; and etching the blanket dielectric layer.

19. The method of claim 1 further comprising electrically connecting a Radio Frequency (RF) passive device to the gate electrode, with an end of the RF device configured to have a same voltage as the gate electrode, wherein the RF passive device comprises a waveguide.

20. The method of claim 6, wherein the RF passive device comprises a waveguide.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) For a more complete understanding of the embodiments, and the advantages thereof, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:

(2) FIGS. 1 through 8 are cross-sectional views of intermediate stages in the manufacturing of a Radio Frequency (RF) switch in accordance with some exemplary embodiments.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

(3) The making and using of the embodiments of the disclosure are discussed in detail below. It should be appreciated, however, that the embodiments provide many applicable concepts that can be embodied in a wide variety of specific contexts. The specific embodiments discussed are illustrative, and do not limit the scope of the disclosure.

(4) A Radio-Frequency (RF) switch formed of a Metal-Oxide-Semiconductor (MOS) transistor and the method of forming the same are provided in accordance with various exemplary embodiments. The intermediate stages of forming the RF switch are illustrated. The variations of the embodiments are discussed. Throughout the various views and illustrative embodiments, like reference numbers are used to designate like elements.

(5) Referring to FIG. 1, substrate 20 is provided. Substrate 20 may comprise a semiconductor material such as silicon, although other semiconductor materials such as silicon germanium may be used. Substrate 20 may be lightly doped with a p-type impurity, for example, boron, indium, or combinations thereof. Substrate 20 has a high resistivity, which is higher than about 3,000 ohm-cm, or higher than about 5,000 ohm-cm. The resistivity may also be between about 3,000 ohm-cm and about 250,000 ohm-cm, and may be between about 5,000 ohm-cm and about 20,000 ohm-cm. The resistivity of substrate 20 is significantly higher than the resistivity of typical silicon substrates, which may have resistivity values close to about 10 ohm-cm. Accordingly, substrate 20 is referred to as a high-resistive substrate hereinafter. The respective p-type doping concentration may be between about 510.sup.6/cm.sup.3 and about 510.sup.10/cm.sup.3 in accordance with some embodiments. The p-type doping concentration of substrate 20 is also significantly lower than the p-type doping concentration of typical silicon substrates, which may have p-type doping concentrations between about 110.sup.12/cm.sup.3 and about 110.sup.15/cm.sup.3.

(6) Also referring to FIG. 1, an n-type implantation is performed to implant an n-type impurity into semiconductor substrate 20. The implantation is shown by arrows 21. As a result, Deep N-Well (DNW) 22 is formed at an intermediate level in substrate 20, with a un-implanted substrate portion 20A overlying DNW 22, and a un-implanted portion of substrate 20 underlying DNW 22. The n-type impurity may be selected from phosphorous, arsenic, antimony, or combinations thereof. The implantation energy may be between about 1,000 KeV and about 2,500 KeV. The dosage of the n-type impurity may be between about 110.sup.13/cm.sup.2 and about 110.sup.14/cm.sup.2. The resulting doping concentration of DNW 22 may be between about 110.sup.13/cm.sup.3 and about 110.sup.14/cm.sup.3.

(7) As shown in FIG. 1, isolation regions 24 are formed to extend from the top surface of semiconductor substrate 20 into semiconductor substrate 20. In some embodiments, isolation regions 24 are Shallow Trench Isolation (STI) regions. The depth of STI regions 24 may be between about 3,500 and about 4,000 in some embodiments. In alternative embodiments, isolation regions 24 are oxide regions formed by local oxidation of substrate 20. Isolation regions 24 may comprise silicon oxide in some embodiments. Isolation regions 24 define a portion of substrate 20 as an active region for forming a Metal-Oxide-Semiconductor (CMOS) transistor.

(8) Referring to FIG. 2, a p-type implantation is performed to implant a p-type impurity into semiconductor substrate 20, and hence p-well region 26 is formed. The implantation is shown as arrows 25. In some embodiments, p-well region 26 is over, and may contact, DNW 22. The top surface of p-well region 26 is lower than the top surface of substrate 20. Accordingly, portion 20A of substrate 20 over p-well region 26 remains to be un-implanted with the p-type impurity. It is appreciated that in the implantations for forming DNW 22 and p-well region 26, the implanted impurities may have some ions left in substrate portion 20A. Substrate portion 20A, however, is not intentionally implanted, and the ions left therein will be a small amount. The p-type impurity introduced by implantation 25 may be selected from indium, boron, or the like. In some embodiments, indium is implanted, and the implantation energy is between about 100 KeV and about 130 KeV. The dosage of the p-type impurity may be between about 110.sup.12/cm.sup.2 and about 110.sup.14/cm.sup.2. The resulting doping concentration of p-well region 26 may be between about 110.sup.11/cm.sup.3 and about 110.sup.14/cm.sup.3.

(9) In accordance with the embodiments of the present disclosure, the order for forming STI regions 24, DNW 22, and p-well region 26 may be different from that are in the illustrated embodiments, and may be changed to different orders. For example, DNW 22 and p-well region 26 may be formed before the formation of STI regions 24. Also, DNW 22 may be formed after the formation of p-well region 26 in some embodiments.

(10) FIG. 3 illustrates the formation of a gate stack, which includes gate dielectric 30 and gate electrode 32. Gate dielectric 30 may be formed of a material selected from silicon oxide, silicon nitride, silicon carbide, silicon oxynitride, combinations thereof, and multi-layers thereof. High-k dielectric materials such as hafnium-based oxides, aluminum-based oxides, lanthanum-based oxides, and combinations thereof, may also be comprised in gate dielectric 30. Gate electrode 32 may be formed of doped polysilicon. Alternatively, metals, metal nitrides, metal silicides, and/or other conductive materials can be used to form gate electrode 32. The formation of gate dielectric 30 and gate electrode 32 includes forming a blanket gate dielectric layer and a blanket gate electrode layer over the blanket gate dielectric layer, and then patterning the blanket gate dielectric layer and the blanket gate electrode layer. In accordance with some embodiments, gate length Lg of gate electrode 32 is smaller than about 0.3 m.

(11) In accordance with the embodiments of the present disclosure, no channel doping is performed. It is appreciated that in the formation of conventional MOS transistors, channel dopings may be performed to increase the doping concentrations in the channels of the respective MOS transistors. In the conventional channel doping for forming NMOS transistors, p-type channel doping may be performed through implanting a p-type impurity into the channels of the NMOS transistors. For forming PMOS transistors, n-type channel doping may be performed through implanting an n-type impurity into the channels of the PMOS transistors. In the embodiments of the present disclosure, no channel doping implantation is performed. This results in the channel doping concentration, which is in the channel 29 that is overlapped by gate electrode 32, to be very low. In accordance with some embodiments, channel region 29 has a p-type doping concentration lower than about 10.sup.13/cm.sup.3. The threshold voltage of the resulting transistor is thus low. Hence, the MOS transistors formed in accordance with the embodiments have very fast switching time, and are suitable for being used as RF switches.

(12) Referring to FIG. 4, tilt implantations 34 are performed to form Lightly Doped Drain/Source (LDD) regions 36, wherein the tilt implantation are tilted from the opposite sides of gate electrode 32. The implanted impurity may include an n-type impurity such as phosphorous, arsenic, or combinations thereof. The tilt implantation may be performed at tilt angle , which may be smaller than about 15 degrees, for example. Due to the tilt implantation, LDD regions 36 extend underlying gate dielectric 30 and gate electrode 32, with a portion of each of LDD regions 36 overlapped by gate dielectric 30 and gate electrode 32. The energy for implanting the n-type impurity, for example, arsenic, may be between about 2 KeV and about 10 KeV. Hence, LDD regions 36 are very shallow, and depth D1 of LDD regions 36 may be smaller than about 50 nm.

(13) Referring to FIG. 5, gate spacers 38 and heavily doped source and drain regions 40 (referred to as source/drain regions hereinafter) are formed. In accordance with some embodiments, gate spacers 38 are formed first, followed by performing an implantation to form source/drain regions 40. Hence, source/drain regions 40 have inner edges aligned to the outer edges of gate spacers 38. In alternative embodiments, an implantation is performed to form source/drain regions 40, followed by the formation of gate spacers 38. Hence, source/drain regions 40 have inner edges aligned to the edges of gate electrode 32. During the formation of source/drain regions 40, an n-type impurity such as arsenic, phosphorous, or the like, is implanted. In the embodiments wherein arsenic is implanted, the implantation energy may be between about 2 KeV and about 10 KeV. Hence, source/drain regions 40 are also very shallow, and depth D2 of source/drain regions 40 may be smaller than about 50 nm. In some embodiments, depth D1 of LDD regions 36 and depth D2 of source/drain regions 40 are substantially equal to each other. In certain embodiments, a difference between depths D1 and D2 may be smaller than about 5 percent, and may be between about 3 percent and about 5 percent.

(14) The bottoms of LDD regions 36 and source/drain regions 40 are spaced apart from the top surface of p-well region 26 by substrate portion 20A. Hence, LDD regions 36 and source/drain regions 40 form junctions with substrate portion 20A, which have a very low p-type impurity concentration.

(15) FIG. 6 illustrates the formation of dielectric layers 42, which are formed to overlap and contact source/drain regions 40. In accordance with some embodiments, each of dielectric layers 42 is in contact with one of gate spacers 38, and extends away from the respective gate spacers 38. The material of dielectric layers 42 may include oxide, nitride, carbide, oxynitride, and/or the like, and may be formed using Chemical Vapor deposition (CVD) methods such as Plasma Enhanced Chemical Vapor Deposition (PECVD), Atomic Layer Deposition (ALD), or the like. Thickness T1 of dielectric layers 42 may be between about 10 nm and about 50 nm. The top surface of dielectric layers 42 may be lower than the top surface of gate electrode 32 in some embodiments. The formation of dielectric layers 42 may include forming a blanket dielectric layer (not shown), forming a photo resist (not shown) over the blanket dielectric layer and patterning the photo resist, and then etching the blanket dielectric layer to form dielectric layers 42. Hence, dielectric layers 42, unlike gate spacers 38 that have sloped top surfaces, may have some top surfaces that are substantially flat.

(16) Next, referring to FIG. 7, gate silicide 44 and source/drain silicide regions 46 are formed. The formation of gate silicide 44 and source/drain silicide regions 46 may include forming a metal layer (not shown), which may include nickel, cobalt, or the like, on the structure shown in FIG. 6. An annealing is then performed, during which the exposed silicon reacts with the metal layer to form silicide regions 44 and 46. The portions of the metal layer unreacted with the metal layer are removed, leaving gate silicide 44 and source/drain silicide regions 46. Due to the formation of dielectric layers 42, source/drain silicide regions 46 are pushed farther away from gate electrode 32, and hence the breakdown voltage between, for example, gate electrode 32 and drain region 40 is increased. MOS transistor 100 is thus formed. MOS transistor 100 may be used as an RF switch.

(17) FIG. 8 illustrates the Back End of Line (BEOL) processes, in which gate contact plug 48, source/drain contact plugs 50, and Inter-Layer Dielectric (ILD) 52 are formed. Gate contact plug 48 and source/drain contact plugs 50 are electrically connected to gate electrode 32 and source/drain regions 40, respectively. Next, interconnect structure 54 is formed. Interconnect structure 54 may include a plurality of dielectric layers 56. In some embodiments, dielectric layers 56 are low-k dielectric layers having dielectric constants (k values) smaller than about 3.5, for example. The k values of low-k dielectric layers 56 may also be lower than about 3.0. Metal lines 58 and vias 60 are formed in dielectric layers 56. Some metal lines 58 and vias 60 are electrically coupled to MOS transistor 100. The metal lines in a same layer are collectively referred to as a metal layer.

(18) As also shown in FIG. 8, RF passive device 62 is formed in the BEOL process. RF passive device 62 may be a capacitor, an inductor, a transformer, a transmission line, a waveguide, or the like, whose characteristics (such as the frequency response and the Q factor) are suitable for being operated in RF frequency range (about 500 MHz or higher). FIG. 8 illustrates the cross-sectional view of a part of RF passive device 62. In some embodiments, RF passive device 62 extends into one or more of the metal layers in interconnect structure 54, wherein thickness T2 of each of the metal layers may be, for example, greater than about 1 m, and may be between about 1 m and about 5 m. Passive device 62 may also extend into an aluminum-containing layer (such as an aluminum copper layer) 64. Through the stacking of the plurality of metal layers and the aluminum-containing layer, the thickness of RF passive device 62 is great enough, and hence RF passive device 62 may be operated under high frequencies.

(19) Although the previously discussed embodiments provide a method of forming an RF switch comprising an NMOS device, the teaching provided in the present disclosure is readily available for the formation of RF switches comprising PMOS devices, with the conductivity types of the respective substrate, well regions, LDD regions, and/or source/drain regions inverted.

(20) In accordance with the embodiments of the present disclosure, MOS transistor 100 acts as an RF switch, which may be operated by the signals that are provided by RF passive device 62. Since RF switch 100 is formed based on substrate 20 that has a high resistivity, the insertion loss of the RF switch is very low. Simulation results revealed that the insertion loss of the RF switch formed in accordance with the embodiments of the present disclosure is about 0.34 dB, which is significantly lower than the specification requirement (lower than 1 dB). Furthermore, the switching time of the RF switch formed in accordance with the embodiments of the present disclosure is about 60 nanoseconds, which is significantly lower than the specification requirement (about 500 nanoseconds). Hence, the RF switch formed in accordance with the embodiments of the present disclosure may meet the specification requirement with a significant margin.

(21) In accordance with some embodiments, a device includes a semiconductor substrate of a first conductivity type, and a deep well region in the semiconductor substrate, wherein the deep well region is of a second conductivity type opposite to the first conductivity type. The device further includes a well region of the first conductivity type over the deep well region. The semiconductor substrate has a top portion overlying the well region, and a bottom portion underlying the deep well region, wherein the top portion and the bottom portion are of the first conductivity type, and have a high resistivity. A gate dielectric is over the top portion of the semiconductor substrate. A gate electrode is over the gate dielectric. A source region and a drain region extend into the top portion of the semiconductor substrate, wherein the source region and the drain region are spaced apart from the well region by the top portion of the semiconductor substrate. The source region, the drain region, the gate dielectric, and the gate electrode form an RF switch configured to operate in a RF frequency range.

(22) In accordance with other embodiments, a device includes a semiconductor substrate, a deep n-well region in the semiconductor substrate, and a p-well region over and contacting the deep n-well region. The semiconductor substrate includes a top portion overlying the p-well region and a bottom portion underlying the deep n-well region. The top portion and the bottom portion are of p-type. The bottom portion has a resistivity greater than about 5,000 ohm-cm. The device further includes an RF switch, which includes a gate dielectric over the top portion of the semiconductor substrate, a gate electrode over the gate dielectric, and a source region and a drain region extending into the top portion of the semiconductor substrate. The source region and the drain region are spaced apart from the p-well region by the top portion of the semiconductor substrate. An RF passive device is overlying and electrically coupled to the RF switch.

(23) In accordance with yet other embodiments, a method includes performing a first implantation to implant a semiconductor substrate to form a deep well region, wherein the semiconductor substrate is of a first conductivity type, and has a resistivity higher than about 5,000 ohm-cm. In the first implantation, an impurity of a second conductivity type opposite to the first conductivity type is implanted. A second implantation is performed to implant the semiconductor substrate, wherein a well region of the first conductivity type is formed over the deep well region. After the first and the second implantations, the semiconductor substrate includes a top portion overlying the well region and a bottom portion underlying the deep well region, wherein the top portion and the bottom portion are substantially un-implanted in the first and the second implantations. The method further includes forming a gate dielectric over the top portion of the semiconductor substrate, forming a gate electrode over the gate dielectric, and performing a third implantation to implant the top portion of the semiconductor substrate to form a source region and a drain region. The source region and the drain region are spaced apart from the well region by a remaining top portion of the semiconductor substrate. The source region, the drain region, the gate dielectric, and the gate electrode form an RF switch configured to operate in a RF frequency range.

(24) Although the embodiments and their advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the embodiments as defined by the appended claims. Moreover, the scope of the present application is not intended to be limited to the particular embodiments of the process, machine, manufacture, and composition of matter, means, methods and steps described in the specification. As one of ordinary skill in the art will readily appreciate from the disclosure, processes, machines, manufacture, compositions of matter, means, methods, or steps, presently existing or later to be developed, that perform substantially the same function or achieve substantially the same result as the corresponding embodiments described herein may be utilized according to the disclosure. Accordingly, the appended claims are intended to include within their scope such processes, machines, manufacture, compositions of matter, means, methods, or steps. In addition, each claim constitutes a separate embodiment, and the combination of various claims and embodiments are within the scope of the disclosure.