SOI (silicon on insulator) substrate improvements

Applicant claims the benefit of Provisional Application, Ser. No. 61/159,833, SOI (Silicon on Insulator) Substrate Improvements, filed on Mar. 13, 2009.

The present invention relates generally to SOI (silicon on insulator) substrates and more particularly to SOI substrate improvements.

The operation of an active device (e.g., a transistor) formed on a conventional SOI (silicon on insulator) substrate (including a buried oxide layer sandwiched between a top Si layer and a bottom Si layer) is usually affected detrimentally by the free mobile carriers near surface interface between the bottom Si layer and the buried oxide layer. Therefore, there is a need for structures (and methods for forming the same) in which the device operation is improved.

Methods of the present invention suppress free mobile carriers with the use of electrically neutral species implanted near the top surface of the bottom Si layer. Access for these implants is provided by trenches etched in the buried oxide in non-device areas. Trenches and device structures are mutually exclusive. Maximum effectiveness requires high trench density; but this is in conflict with the need to maintain uniform pattern density of active silicon and polysilicon features to permit wafer processing such as etching and chemical-mechanical polishing. Interleaving trench features and active or dummy device structures to achieve uniform density does not create an optimized structure because lithographic alignment tolerances require unused space between trenches and device or fill structures. In the present invention, the dummy structures are used as masks for creating the trenches, thereby avoiding the above drawbacks.

FIG. 1 shows an SOI (silicon on insulator) substrate, in accordance with embodiments of the present invention.

FIG. 2 shows an STI (shallow trench isolation) region formed in the structure of FIG. 1 , in accordance with embodiments of the present invention.

FIG. 3 shows a gate dielectric layer formed on the structure of FIG. 2 , in accordance with embodiments of the present invention.

FIG. 4 shows gate electrode regions formed on the structure of FIG. 3 , in accordance with embodiments of the present invention.

FIG. 5 shows gate dielectric spacers formed on side walls of the gate electrode regions of FIG. 4 , in accordance with embodiments of the present invention.

FIG. 6 shows source/drains regions formed on the structure of FIG. 5 , in accordance with embodiments of the present invention.

FIG. 7 shows silicide regions formed on the source/drains regions of FIG. 6 , in accordance with embodiments of the present invention.

FIG. 8 shows a protection layer formed on the structure of FIG. 7 , in accordance with embodiments of the present invention.

FIG. 9 shows a trench formed in the structure of FIG. 8 , in accordance with embodiments of the present invention.

FIG. 10 shows a passivation layer formed on the structure of FIG. 9 , in accordance with embodiments of the present invention.

FIGS. 1-10 illustrate a fabrication process for forming a semiconductor structure 100 in accordance with embodiments of the present invention. More specifically, with reference to FIG. 1 , the fabrication process can start with an SOI (silicon on insulator) substrate 110 + 120 + 130 . In one embodiment, the SOI substrate 110 + 120 + 130 can comprise (i) a bottom semiconductor (e.g., silicon) layer/substrate 110 , (ii) a buried dielectric (e.g., oxide) layer (BOX) layer 120 on the bottom silicon layer 110 , and (iii) a top semiconductor (e.g., silicon) layer 130 on the BOX layer 120 . The BOX layer 120 can comprise silicon dioxide (SiO2). The SOI substrate 110 + 120 + 130 can be formed using any conventional process.

Next, with reference to FIG. 2 , in one embodiment, portions of the top silicon layer 130 are removed and replaced by a dielectric material such as silicon dioxide resulting in shallow trench isolation (STI) region 132 . The (STI) region 132 can be formed by any conventional process. As a result of the formation of the STI region 132 , the top silicon layer 130 is divided into active Si region 130 a , and dummy Si regions 130 b and 130 c . In general, there can be any numbers of active and dummy Si regions. In one embodiment, the pattern density of all the active and dummy Si regions of the structure 100 (including the Si regions 130 a , 130 b , and 130 c ) is essentially uniform across the structure 100 . In other words, the pattern density of all the active and dummy Si regions of the structure 100 remains essentially the same for any two different areas on the structure 100 . In general, the pattern density of multiple regions for an area of the structure 100 is the percentage of that area which is occupied by the multiple regions. For example, if all the active and dummy Si regions of the structure 100 occupy 40 μm 2 of a 100 μm 2 area of the structure 100 , then the density of all the active and dummy Si regions of the structure 100 in that 100 μm 2 area is 40%.

A region is called active if it is a part of an active device (such as transistor, capacitor, resistor, diode, etc.) of the complete semiconductor structure/circuit 100 . A device is considered active if the device plays a role in the normal operation of the complete structure/circuit 100 . As will be clear later, the Si region 130 a is active because it is a part (i.e., body/channel region) of an active transistor of the complete semiconductor structure/circuit 100 . It should be noted that a Si region does not have to be part of an active transistor of the complete semiconductor structure/circuit 100 to be active. For example, the Si region can also be an active Si region if the Si region is part of a non-transistor active device of the complete semiconductor structure/circuit 100 such as a p-n junction diode.

In contrast, a region is called dummy if it is not a part of any active device of the complete semiconductor structure 100 . The purpose of a dummy region (like the Si regions 130 b and 130 c ) is to provide a uniform pattern density across the semiconductor structure 100 so as to facilitate subsequent fabrication steps (such as chemical mechanical polishing—CMP).

Next, with reference to FIG. 3 , in one embodiment, a gate dielectric layer 140 is formed on top of the semiconductor structure 100 of FIG. 2 . More specifically, the gate dielectric layer 140 can be formed by depositing a dielectric material (such as silicon dioxide—SiO2) on top of the semiconductor structure 100 of FIG. 2 . If SiO2 is used, the gate dielectric layer 140 can be formed by CVD (chemical vapor deposition) of SiO2 on top of the semiconductor structure 100 of FIG. 2 . Alternatively, the SiO2 gate dielectric layer 140 can be formed by thermal oxidation of a top silicon surface 134 ( FIG. 2 ) of the top silicon layer 130 .

Next, with reference to FIG. 4 , in one embodiment, active electrically conductive region 150 a and dummy electrically conductive regions 150 b and 150 c are formed on the semiconductor structure 100 of FIG. 3 . In one embodiment, the electrically conductive regions 150 a , 150 b , and 150 c comprise an electrically conductive material such as polysilicon. The polysilicon regions 150 a , 150 b , and 150 c can be simultaneously formed using conventional lithographic and etching processes. In general, there can be any numbers of active and dummy polysilicon regions (like the polysilicon regions 150 a , 150 b , and 150 c ). In one embodiment, the pattern density of all the active and dummy polysilicon regions of the structure 100 is essentially uniform across the structure 100 . As will be clear later, the polysilicon region 150 a is active because it is a part (i.e., a gate electrode region) of an active transistor of the complete semiconductor structure/circuit 100 . It should be noted that a polysilicon region does not have to be part of an active transistor of the complete semiconductor structure/circuit 100 to be active. For example, the polysilicon region can also be an active polysilicon region if the polysilicon region is part of a non-transistor active device of the complete semiconductor structure/circuit 100 such as a polysilicon resistor. It should be also noted that in general, a dummy polysilicon region does not necessarily overlap a dummy Si region, although in FIG. 4 , the polysilicon region 150 b overlaps dummy Si region 130 b , and the polysilicon region 150 c overlaps dummy Si region 130 c.

Next, with reference to FIG. 5 , in one embodiment, after the polysilicon regions 150 a , 150 b , and 150 c are formed, active gate dielectric spacers 160 a 1 and 160 a 2 are formed on side walls of the active gate electrode region 150 a , whereas dummy gate dielectric spacers 160 b 1 , 160 b 2 , 160 c 1 , and 160 c 2 are formed on side walls of the dummy gate electrode regions 150 b and 150 c . The gate dielectric spacers 160 a 1 , 160 a 2 , 160 b 1 , 160 b 2 , 160 c 1 , and 160 c 2 (or in short, the gate dielectric spacers 160 ) can comprise silicon nitride.

In one embodiment, all the gate dielectric spacers 160 can be simultaneously formed by (i) forming a conformal dielectric spacer layer (not shown) on top of the structure 100 of FIG. 4 , and then (ii) vertically etching back the dielectric spacer layer resulting in the dielectric spacers 160 of FIG. 5 . This vertical etching may over-etch the gate dielectric layer 140 ( FIG. 4 ) resulting in active gate dielectric region 140 a and dummy gate dielectric regions 140 b and 140 c ( FIG. 5 ) beneath the active gate electrode region 150 a and the dummy gate electrode regions 150 b and 150 c , respectively. In general, there can be any numbers of active and dummy gate dielectric spacers. In one embodiment, the pattern density of all the active and dummy gate dielectric spacers of the structure 100 is essentially uniform across the structure 100 . As will be clear later, the gate dielectric spacers 160 a 1 and 160 a 2 are active because they are parts of an active transistor of the complete semiconductor structure/circuit 100 .

Next, with reference to FIG. 6 , in one embodiment, active source/drain regions 170 a 1 and 170 a 2 are formed in the active silicon region 130 a . More specifically, the active source/drain regions 170 a 1 and 170 a 2 can be formed by implanting dopants (e.g., boron atoms if a p-channel device is to be formed) in the active silicon region 130 a using an ion implantation process with the active gate electrode region 150 a and the active spacer regions 160 a 1 and 160 a 2 serving as blocking masks.

It should be noted that as a by-product of the implantation process that forms the active source/drain regions 170 a 1 and 170 a 2 , dummy source/drain regions 170 b 1 , 170 b 2 , 170 c 1 , and 170 c 2 are also formed in the dummy Si regions 130 b and 130 c , respectively.

Next, with reference to FIG. 7 , in one embodiment, active silicide regions 172 a 1 and 172 a 2 are formed on and in direct physical contact with the active source/drain regions 170 a 1 and 170 a 2 , respectively. The active silicide regions 172 a 1 and 172 a 2 can comprise silicide nickel or another metal silicide such as cobalt silicide or titanium silicide. The active silicide regions 172 a 1 and 172 a 2 can be formed by a conventional process. For example, the active silicide regions 172 a 1 and 172 a 2 can be formed by (i) depositing a conformal nickel layer (not shown) on top of the structure 100 of FIG. 6 , then (ii) heating the structure 100 resulting in nickel of the conformal nickel layer chemically reacting with silicon of the active source/drain regions 170 a 1 and 170 a 2 , resulting in the active silicide nickel regions 172 a 1 and 172 a 2 , and then (iii) removing the unreacted nickel, resulting in the structure 100 of FIG. 7 .

It should be noted that as a by-product of the silicidation process that forms the active silicide regions 172 a 1 and 172 a 2 , dummy silicide regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 are also formed as shown in FIG. 7 . In an alternative embodiment, the dummy silicide regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 are not created as a result of the formation of the active silicide regions 172 a 1 and 172 a 2 . More specifically, in this alternative embodiment, first, a conformal silicon nitride layer (not shown) is formed on top of the structure 100 of FIG. 6 by illustratively CVD. Next, portions of the conformal silicon nitride layer above the active regions (i.e., the active gate electrode region 150 a , the active gate dielectric spacers 160 a 1 and 160 a 2 , and the active silicon region 130 a ) are removed. Next, the active silicide regions 172 a 1 and 172 a 2 are formed as described above (i.e., as described in the main embodiment). It should be noted that in this alternative embodiment, because of the silicon nitride layer, the dummy Si regions 130 b and 130 c are not exposed to the silicidation process that forms the active silicide regions 172 a 1 and 172 a 2 . As a result, the dummy silicide regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 ( FIG. 7 ) are not created in this alternative embodiment. Next, after the active silicide regions 172 a 1 and 172 a 2 are formed, the silicon nitride layer is removed resulting in the structure 100 of FIG. 7 (without the dummy silicide regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 ).

With reference to FIG. 7 , it should be noted that an active transistor (an active device) has just been formed having a channel region 130 a ′ in the active silicon region 130 a . The channel region 130 a ′ is disposed between the active source/drain regions 170 a 1 and 170 a 2 . The active transistor further comprises the active gate electrode region 150 a and the active gate dielectric region 140 a which electrically insulates the gate electrode region 150 a from the channel region 130 a ′. The active transistor further comprises gate dielectric spacers 160 a 1 and 160 a 2 . Hereafter the transistor can be referred to as the transistor 130 a ′+ 140 a + 150 a.

Next, after the silicide regions 172 a 1 and 172 a 2 are formed (and after the silicon nitride layer is removed if the alternative embodiment described above is applicable), with reference to FIG. 8 , in one embodiment, a patterned protection layer 180 (comprising a photoresist material for example) is formed on the structure 100 of FIG. 7 such that the active transistor 130 a ′+ 140 a + 150 a are covered by the patterned photoresist layer 180 whereas other areas of the structure 100 having no active devices are not covered by the patterned photoresist layer 180 . In one embodiment, the patterned photoresist layer 180 can be formed by a spin-on process followed by a lithographic process.

Next, the STI regions 132 and then the BOX layer 120 are etched vertically with the patterned photoresist layer 180 and the dummy regions (i.e., the dummy gate electrode regions 150 b and 150 c , the dummy gate dielectric spacers 160 b 1 , 160 b 2 , 160 c 1 , and 160 c 2 , the dummy silicide regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 , and the dummy Si regions 130 b and 130 c ) as blocking masks.

Also, in one embodiment, the etching of the STI region 132 and the BOX layer 120 is stopped at the top surface 112 of the bottom silicon layer 110 (as shown in FIG. 9 ) resulting in a trench 190 in the STI region 132 and the BOX layer 120 . As a result, some areas of the top surface 112 of the bottom silicon layer 110 are exposed to the surrounding ambient through the trench 190 . Also as a result, the side walls of the trench 190 are aligned with the patterned photoresist layer 180 and the dummy regions. It should be noted that in the alternative embodiment described above (in which the dummy silicide regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 of FIG. 7 are not created), the formation of the trench 190 ( FIG. 9 ) does not expose any silicide metal material to the surrounding ambient. This is not the case in the main embodiment, because the formation of the trench 190 exposes the dummy silicide metal regions 172 b 1 , 172 b 2 , 172 c 1 , and 172 c 2 to the surrounding ambient as shown in FIG. 9 .

Next, with reference to FIG. 9 , in one embodiment, the areas of the top surface 112 that are exposed to the surrounding ambient through the trench 190 are bombarded with ions such as argon (Ar) ions, helium (He) ions, neon (Ne) ions, xenon (Xe) ions, nitrogen (N) ions, germanium (Ge) ions, and carbon (C) ions resulting in damage in these areas of the top surface 112 . More specifically, the bombardment of these areas of the top surface 112 (represented by arrow 900 which indicates the direction of bombardment) can be performed using an ion implantation process. In one embodiment, the direction 900 is perpendicular to the top surface 112 of the bottom Si layer 110 . It should be noted that the transistor 130 a ′+ 140 a + 150 a is not affected (i.e., is not hit by any ion) by the bombardment 900 because of the protection of the patterned photoresist layer 180 .

Next, in one embodiment, the patterned photoresist layer 180 is removed, for example, by a wet etching process.

Next, with reference to FIG. 10 , in one embodiment, a dielectric material (such as BPSG—boron-phosphorous silicate glass) is deposited on the structure 100 of FIG. 9 , followed by a CMP process, resulting in a BPSG passivation layer 910 of the structure 100 of FIG. 10 . More specifically, in one embodiment, the BPSG deposition can be performed such that the trench 190 is completely filled by the BPSG layer 910 .

Next, in one embodiment, other conventional steps are performed so as to provide electric paths to the transistor 130 a ′+ 140 a + 150 a and so as to interconnect the transistor 130 a ′+ 140 a + 150 a and other devices (not shown) of the structure 100 . For example, electric paths to the transistor 130 a ′+ 140 a + 150 a can be created as follows. First, holes (not shown) can be created in the BPSG layer 910 such that the silicide regions 172 a and 172 b are exposed to the surrounding ambient through the created holes in the BPSG layer 910 . Next, the holes can be filled with an electrically conductive material (e.g., tungsten) resulting in contact regions in the holes. These contact regions provide electric paths to the source/drain regions 170 a and 170 b of the active transistor 130 a ′+ 140 a + 150 a (through the silicide regions 172 a and 172 b , respectively).

In one embodiment, no contact region is formed in the BPSG layer 910 to provide electric paths to the dummy source/drain regions 170 b 1 , 170 b 2 , 170 c 1 , and 170 c 2 .

In summary, the fabrication process of the present invention is carried out as follows. First, active devices (like the active transistor 130 a ′+ 140 a + 150 a of FIG. 7 ) and dummy regions (like the dummy regions 130 b , 130 c , 150 b , 150 c , 160 b 1 , 160 b 2 , 160 c 1 , and 160 c 2 , etc.) are formed on the SOI substrate 110 + 120 + 130 ( FIG. 7 ). Next, the patterned photoresist layer 180 ( FIG. 8 ) is formed so as to cover and protect the active devices (but not cover the dummy regions). Next, the STI region 132 and the BOX layer 120 are etched with the patterned photoresist layer 180 and the dummy regions serving as blocking masks resulting in some areas of the top surface 112 of the bottom Si layer 110 being exposed to the surrounding ambient. Next, the exposed areas of the top surface 112 of the bottom Si layer 110 are bombarded with heavy ions (such as Ar ions) resulting damages/defects to the top surface 112 ( FIG. 9 ). It should be noted that the active devices are not affected by this bombardment because of the protection by the patterned photoresist layer 180 . Next, the patterned photoresist layer 180 is removed, and then the dielectric layer 910 are formed, and then electrical paths to the active devices are created in the dielectric layer 910 ( FIG. 10 ).

As a result of the damages to the top surface 112 of the bottom Si layer 110 caused by the bombardment 900 ( FIG. 9 ) and also as a result of the presence of Ar atoms at the top surface 112 , the detrimental effects to the operation of the active devices (like the transistor 130 a ′+ 140 a + 150 a ) caused by free mobile carriers near the surface 112 of the bottom Si layer 110 are mitigated.

While particular embodiments of the present invention have been described herein for purposes of illustration, many modifications and changes will become apparent to those skilled in the art. Accordingly, the appended claims are intended to encompass all such modifications and changes as fall within the true spirit and scope of this invention.