US 20060286739 A1
A shallow trench isolation structure having a negative taper angle. A graded doped sacrificial layer is formed over a semiconductor substrate and etched to form a first trench therein having trench sidewalls that present a negative taper angle. The substrate is also etched to form a second trench therein overlying the first trench. Silicon dioxide fills both the first and the second trenches to form the shallow trench isolation structure, with the silicon dioxide in the first trench exhibiting a negative taper angle to avoid formation of polysilicon stringers during a gate polysilicon deposition.
1. A method for forming a shallow trench isolation structure in an integrated circuit, the method comprising:
providing a silicon substrate;
forming a sacrificial layer having a graded doped concentration overlying the substrate;
etching the sacrificial layer to form an opening therein, wherein sidewalls of the opening present a negative taper angle;
etching the substrate in a region below the opening to form a trench therein; and
forming the shallow trench isolation structure by forming an insulating material in the opening and the trench, wherein the insulating material in the opening exhibits a negative taper angle.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
7. The method of
8. The method of
9. The method of
10. The method of
11. The method of
12. The method of
13. The method of
14. The method of
15. A method for forming a shallow trench isolation structure in an integrated circuit, the method comprising:
providing a silicon substrate;
forming a pad oxide layer overlying the silicon substrate;
forming a sacrificial layer having a graded doped concentration overlying the pad oxide layer;
forming a silicon nitride layer overlying the sacrificial layer;
forming a photoresist layer overlying the silicon nitride layer;
forming a first opening in the photoresist layer;
using the first opening as a pattern, etching the sacrificial layer and the pad oxide layer to form a second opening therein, wherein sidewalls of the second opening present a negative taper angle;
etching the substrate in a region below the second opening to form a trench therein; and
forming an insulating material in the second opening and the trench, wherein the insulating material in the second opening exhibits a negative taper angle.
16. The method of
17. The method of
18. The method of
19. The method of
20. The method of
21. The method of
22. The method of
23. The method of
24. The method of
25. The method of
The present invention relates generally to integrated circuits and more particularly to a method for forming a shallow trench isolation (STI) structure within an integrated circuit, wherein the shallow trench isolation structure comprises a graded doped silicon dioxide material, and further to a shallow trench isolation structure comprising graded doped silicon dioxide material.
Integrated circuits (or chips) typically comprise a silicon substrate and semiconductor devices, such as transistors, formed from doped regions within the substrate. Interconnect structures disposed in parallel-like layers overlying the semiconductor substrate provide electrical connection between doped regions to form electrical devices and circuits. A conventional interconnect system comprises a plurality of substantially vertical conductive vias or plugs interconnecting substantially horizontal conductive traces, with a dielectric layer disposed between two vertically adjacent horizontal conductive traces.
Within an integrated circuit substrate it may be necessary to isolate certain doped regions to avoid the effects of parasitic devices that are formed by interaction of doped regions. For example, in a CMOSFET device (complimentary metal-oxide semiconductor field effect device) comprising n-channel and p-channel metal-oxide semiconductor field effect transistors (MOSFETS) formed in oppositely doped adjacent wells, a parasitic bipolar structure, i.e., a p-n-p-n thyristor, is inadvertently formed. Although the thyristor is off under normal operating conditions, if the isolation between doped regions is not sufficient, under certain bias conditions the p-n-p regions can supply base current to the n-p-n regions, causing a large current leakage between adjacent MOSFETS that can latch-up the CMOS device.
A thick oxide region, whether formed according to a local oxidation of silicon (LOCOS) process or an STI process as described below, electrically isolates adjacent transistors and other devices to minimize current therebetween and reduce these parasitic effects. The local oxidation of silicon process forms recessed LOCOS isolation regions in a non-active area (also referred to as a field region) of the semiconductor substrate. For example, LOCOS regions are formed in an area between the p-channel and n-channel devices of a CMOS device. According to a one LOCOS process, a layer of silicon nitride is deposited over the substrate and patterned according to conventional masking and etching steps to form openings that expose underlying substrate silicon. The isolating LOCOS regions are formed by oxidizing the exposed silicon through the openings. No oxidation occurs in the masked regions.
Shallow trench isolation enjoys certain advantages over the LOCOS method. An STI structure comprises a dielectric-filled substrate trench that electrically isolates doped regions of active devices, including CMOS devices, MOSFETS and bipolar transistors. STI is an important technology for device sizes below about 0.25 microns, as the trenches consume a smaller surface area and exhibit a flatter upper surface topology than the LOCOS structures. Since the shallow trench isolation structure consumes less substrate surface than the LOCOS structure, more transistors per unit area can be fabricated in an integrated circuit employing STI isolation techniques. STI structures also provide superior isolation because sharp corners at the bottom of the STI trench form voltage barriers that tend to block leakage currents between adjacent doped regions. LOCOS regions generally present rounded corners and thus may permit some leakage current. STI trenches are typically about 3000 Angstroms deep.
As is known, a plurality of integrated circuits are fabricated in a semiconductor wafer, each integrated circuit comprising doped regions formed in a wafer substrate, with dielectric layers and conductive interconnect layers formed over an upper surface of the substrate.
A semiconductor substrate 30 in
A photoresist layer 40 is deposited, exposed and developed according to known processes to form an opening 41 therein.
Using the photoresist layer 40 as an etch mask, an opening 46 (see
The photoresist layer 40 is removed and the wafer is cleaned. As shown in
A trench silicon dioxide liner 50 is then formed or deposited in the trench 48. See
As illustrated in
To complete formation of the STI structure 55, the wafer is cleaned, the silicon nitride layer 38 and the pad oxide layer 36 are removed and finally the wafer is cleaned again.
As can be seen from
According to subsequent processing steps a gate polysilicon layer is deposited on the surface 37. During this polysilicon deposition, undesired polysilicon stringers can form around the sidewalls 68 and 69 and the left and right corners 64 and 66, creating short circuits that defeat the STI isolation function, thereby degrading performance of the integrated circuit. Thus, a process technology to avoid formation of the polysilicon stringers is highly desired.
According to one technique, to reduce the probability of polysilicon short circuits, it is known to minimize the oxide undercut by reducing a duration of the clean step (typically a hydrofluoric acid clean) that follows removal of the silicon nitride layer 38 and the pad oxide layer 36. However, it is also known that the shorter cleaning step causes material residues and contaminants to remain on the substrate, possibly causing undesirable short circuits or leakage current within the STI structure 55 and the substrate 30.
According to one embodiment, the present invention comprises a method for forming a shallow trench isolation structure in an integrated circuit. The method comprises providing a silicon substrate, forming a sacrificial layer having a graded doped concentration overlying the substrate, etching the sacrificial layer to form an opening therein, wherein sidewalls of the opening present a negative taper angle, etching the substrate in a region below the opening to form a trench therein and forming the shallow trench isolation structure by forming an insulating material in the opening and the trench, wherein the insulating material in the opening exhibits a negative taper angle.
According to another embodiment of the invention, a semiconductor device, comprises a substrate, a device isolation structure formed on the substrate, the device isolation structure further comprising a trench formed within the substrate and an insulating film filling the trench, the insulating film extending above an upper surface of the substrate and presenting a negative taper angle with respect to the substrate
The present invention can be more easily understood and the advantages and uses thereof more readily apparent when the following detailed description of the present invention is read in conjunction with the figures wherein:
In accordance with common practice, the various described features are not drawn to scale, but are drawn to emphasize specific features relevant to the invention. Reference characters denote like elements throughout the figures and text.
Before describing in detail the particular method and apparatus according to the present invention, it should be observed that the present invention resides primarily in a novel and non-obvious combination of elements and process steps. Accordingly, so as not to obscure the disclosure with details that will be readily apparent to those skilled in the art having the benefit of the description herein, in the description that follows, certain elements and process steps have been described with lesser detail, while the drawings and specification describe in greater detail other elements and steps pertinent to understanding the invention. The following embodiments are not intended to define limits as to the structure or use of the invention, but only to provide exemplary constructions. The embodiments are permissive rather than mandatory and illustrative rather than exhaustive.
The process steps and structures described below do not comprise a complete process flow for fabricating integrated circuit devices or a completed device. Only the process steps and structures necessary to understand the present invention are described.
The present invention comprises a technique for forming an STI structure having a negative taper angle above the substrate surface 37, in contrast to the positive taper angle formed according to the prior art processes. A preferred methodology for forming the negative taper angle employs a graded doped sacrificial silicon dioxide layer 70 (see
In one embodiment, the graded doped silicon dioxide layer 70 comprises a layer about 1000 Angstroms thick comprising boron and phosphorous-doped oxide film (e.g., TEOS, plasma-deposited silane, LPCVD-deposited silane). In another embodiment, the layer 70 comprises a layer doped with either boron or phosphorous dopants. According to yet another embodiment, the graded doped oxide layer 70 comprises a plurality of sublayers of different dopant concentrations and/or dopant types.
The graded doped silicon dioxide layer 70 can be doped in situ by introducing a gas or liquid containing the desired dopant into the atmosphere of the deposition chamber. Depending on the desired doping concentration, it may also be possible to dope the layer 70 using an implant doping process.
In an embodiment wherein the graded doped silicon dioxide layer 70 or one or more of the doped sublayers thereof comprises a phosphoborosilicate glass (BPSG) material layer. According to one embodiment, the BPSG material layer can be formed by a low-pressure chemical vapor deposition (LPCVD) process, a plasma CVD process or a sub-atmospheric CVD (SACVD) process. An exemplary SACVD process for depositing the BPSG material layer is performed at about 480 degrees C. and about 200 Torr. The constituent gas flow rates are ozone (O3) at about 4000 sccm and helium at about 6000 sccm. The remaining liquid chemicals and their flow rates are: TEOS (tetraethyl orthosilicate) at about 550 mg/minute, triethylborate (TEB) at about 178 mg/minute and triethylphosphate (TEPO) at about 32 mg/minute. As is known by those skilled in the art, these are merely exemplary sources and recipe parameters, where the latter can be varied by at least ±10%. Further, these parameters can be varied to achieve different doping densities and in response to process variants.
In an embodiment wherein the graded doped silicon dioxide layer 70 or one or more of the doped sublayers thereof comprises a phosphosilicate glass (PSG) material layer. According to one embodiment, the PSG material layer can be formed by a plasma-enhanced chemical vapor deposition (PECVD) process. An exemplary PECVD process for depositing the PSG material layer is performed at about 480 degrees C. and about 2 Torr with 500 W applied to the plasma chamber. The constituent gas flow rates are oxygen (O) at about 1400 sccm and helium at about 800 sccm. The liquid chemicals and their flow rates are: TEOS (tetraethyl orthosilicate) at about 800 mg/minute and triethylphosphate (TEPO) at about 18 mg/minute. As is known by those skilled in the art, these are merely exemplary material source and recipe parameters, where the recipe parameters can be varied by at least ±10% with acceptable results.
A doping concentration of the graded doped silicon dioxide layer 70 decreases from a first concentration (e.g., a maximum dopant concentration) at a surface 72 proximate the pad oxide layer 36, to a second concentration, e.g., a minimum concentration, at a surface 76 proximate the silicon nitride layer 71. A representative doping profile is illustrated in
According to a preferred embodiment, both the boron and phosphorous doping the concentrations decrease in a direction away from the surface 72 of the layer 70, but the boron and phosphorous dopant concentrations may differ with respect to a slope of the doping concentration curve (the rate at which the doping concentration changes) and a magnitude of the doping concentration at any depth within the layer 70. Both the rate of change of the concentration and the concentration magnitudes can be selected to achieve a desired etch profile for the STI structure. Also, in an embodiment wherein the graded doped silicon dioxide layer 70 comprises a plurality of sublayers, the shape of the doping profile may vary from the representative doping profile of
It is known that a phosphorous-doped silicon dioxide layer etches faster than a boron-doped silicon dioxide layer. Further, a phosphorous-doped plus boron-doped silicon-dioxide layer etches faster than a solely phosphorous-doped silicon dioxide layer. Thus the STI etch rate and the STI sidewall profile can be controlled by varying the relative dopant concentrations of phosphorous and boron, the rate at which the concentrations change, the concentration magnitudes within the layer 70 and the thickness of the layer 70.
According to a preferred embodiment, the dopant concentration ranges from about 5% to about 1% for boron and about 5% to about 1% for phosphorous within the graded doped silicon dioxide layer 70. A combined dopant concentration, i.e., boron plus phosphorous, of up to about 10% is permitted, but it may not be desirable for either concentration to exceed about 5%. Thus, for example, a boron concentration of 3% plus a phosphorous concentration of 4% is acceptable, but a boron concentration of 3% plus a phosphorous concentration of 6% may not be desired since the phosphorous concentration exceeds 5%.
As is known in the art, other doping profiles can be created by adjusting the dopant gas concentrations during formation of the graded doped silicon dioxide layer 70. Such other dopant concentrations vill affect the etch rate of the layer 70 and thus the taper angle of the STI structure. In particular, if the layer 70 comprises a plurality of differently-doped sublayers having different doping concentrations (and in one embodiment different dopants) each sublayer exhibits a different etch rate, resulting in a different sidewall profile corresponding to each sublayer.
Using the photoresist layer 40 as an etch mask, the opening 46 (see
The photoresist layer 40 is removed and using the opening 46 as an etch mask, an opening 80 having negative taper angle sidewalls 82 is formed in the graded doped silicon dioxide layer 70 and the pad oxide layer 36, preferably using a wet etching process. In another embodiment the layer 70 is etched using a plasma etching process.
During etching of the graded doped silicon dioxide layer 70 the sidewall profiles are substantially controlled by the doping of the layer 70. It is known that an etch rate of a doped silicon dioxide material is a direct function of the doping concentration, i.e., the etch rate increases with increasing doping concentration. Thus because the doping concentration increases in a direction toward the pad silicon dioxide layer 36, the etch rate increases with etch duration, forming the sidewalls 82 with a negative taper angle.
As shown in
The silicon dioxide liner 50 is then formed or deposited in the trench 88. Typically, the liner 50 is formed by oxidizing exposed silicon in the trench 88 to form a high quality silicon dioxide material layer. However, it may be possible to form a suitably high-quality liner 50 by depositing silicon dioxide to form the liner 50.
In another embodiment of the invention, the pad silicon dioxide layer 36 is not etched during etching of the graded doped silicon dioxide layer 70. According to this embodiment the pad silicon dioxide layer 36 is etched during substrate etching.
As illustrated in
The wafer is cleaned, the silicon nitride layer 71 and the pad silicon dioxide layer 36 are removed and the wafer is cleaned again. See the final structures in
However, as can be seen in
In yet another embodiment (not illustrated) to avoid a possible positive taper angle in the region of the STI structure 92 adjacent the silicon nitride layer 71, the layer 71 is removed prior to forming the STI silicon dioxide in the trench 80, see
The STI structure 92 formed according to the teachings of the present invention does not exhibit the disadvantageous re-entrant profiles and weak notch regions that are known to occur according to the prior art processes. Therefore, the STI structure 92 avoids the formation of polysilicon stringers and shorts when the polysilicon gate is formed later. By drastically reducing or eliminating the polysilicon shorts and leakage paths, fabrication processes employing the present invention enjoy a higher, more consistent yield and improved device performance.
By controlling the dopant concentration during formation of the graded doped sacrificial silicon dioxide layer 70, different sidewall profiles can be achieved. See for example, sidewalls 110, 111 and 112 of
The silicon nitride layer 71 preferably comprises a thinner material layer than the silicon nitride layer 38 of the prior art. According to the present invention, a thickness of the silicon nitride layer 71 can be reduced substantially from about 1600 Angstroms to about 500 to about 800 Angstroms since the graded doped silicon dioxide layer 70 can serve as a material stop layer for the chemical mechanical polishing step that removes silicon from the surface 94. A thinner silicon nitride layer imposes less stress on the silicon substrate compared with the thicker silicon nitride layer that is required according to the prior art process. Improved wafer flatness, due to this stress reduction, results in better STI planarization.
An architecture and process have been described as useful for forming STI structure in a semiconductor substrate. Specific applications and exemplary embodiments of the invention have been illustrated and discussed, which provide a basis for practicing the invention in a variety of ways and in a variety of circuit structures. Numerous variations are possible within the scope of the invention. Features and elements associated with one or more of the described embodiments are not to be construed as required elements for all embodiments. The invention is limited only by the claims that follow.