RELATED APPLICATIONS This application is a division of U.S. patent application Ser. No. 10/711,771 filed on Oct. 4, 2004.
FIELD OF THE INVENTION The present invention relates to the field of semiconductor fabrication; more specifically, it relates a method of fabricating doped polysilicon lines and complementary metal-oxide-silicon (CMOS) doped polysilicon gates.
BACKGROUND OF THE INVENTION Advanced CMOS devices utilize doped polysilicon lines and gates with metal silicide layers as a method of improving and matching the performance of N-channel field effect transistors (NFETs) and P-channel field effect transistors (PFETs). However, controlling the width and sheet resistance of oppositely doped polysilicon lines and gates has become more important and difficult as the widths of polysilicon lines and gates have decreased. Therefore, there is a need for a method of fabricating doped polysilicon lines and gates with improved linewidth control.
SUMMARY OF THE INVENTION A first aspect of the present invention is a method of fabricating a semiconductor structure, forming a dielectric layer on a top surface of a substrate; forming a polysilicon layer on a top surface of the dielectric layer; implanting the polysilicon layer with N-dopant species, the N-dopant species essentially contained within the polysilicon layer; implanting the polysilicon layer with a nitrogen containing species, the nitrogen containing species essentially contained within the polysilicon layer.
BRIEF DESCRIPTION OF DRAWINGS The features of the invention are set forth in the appended claims. The invention itself, however, will be best understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying drawings, wherein:
FIGS. 1A through 1D are partial cross-sectional views illustrating initial steps for fabricating doped polysilicon lines and gates according to a first embodiment of the present invention;
FIGS. 2A through 2D are partial cross-sectional views illustrating initial steps for fabricating doped polysilicon lines and gates according to a second embodiment of the present invention;
FIGS. 3A through 3D are partial cross-sectional views illustrating initial steps for fabricating doped polysilicon lines and gates according to a third embodiment of the present invention;
FIG. 4 is a plot of concentration of implanted species versus distance from a top surface of a doped polysilicon layer according to the present invention;
FIGS. 5A and 5B are partial cross-sectional views illustrating common intermediate steps for fabricating doped polysilicon lines and gates according to the present invention;
FIG. 6 is a partial cross-sectional view of a problem solved by the present invention; and
FIGS. 7A through 7E are partial cross-sectional views illustrating common last steps for fabricating doped polysilicon lines and gates according to the present invention.
DETAILED DESCRIPTION OF THE INVENTION The present invention will be described using fabrication of doped polysilicon gates as exemplary of the fabrication process of the present invention. A doped polysilicon gate should be considered a doped polysilicon line used for a specific purpose.
FIGS. 1A through 1D are partial cross-sectional views illustrating initial steps for fabricating doped polysilicon lines and gates according to a first embodiment of the present invention. In FIG. 1A, formed in a silicon substrate 100 are an N-well 105, a P-well 110 and shallow trench isolation (STI) 115. STI 115 may be formed by etching a trench into substrate 100, depositing a dielectric layer on a surface 120 of the substrate of sufficient thickness to fill the trench, and then performing a chemical-mechanical-polishing step to remove excess dielectric layer. However, formation of STI 115 is optional, and STI 115 need not be present. Formed on top surface 120 of substrate 100 is a gate dielectric layer 125. Formed on a top surface 130 of gate dielectric layer 125 is a polysilicon layer 135. In one example, gate dielectric layer 125 is thermal silicon oxide having a thickness of between about 0.8 nm to about 4 nm. In one example, polysilicon layer 135 is undoped polysilicon having a thickness of between about 40 nm to about 200 nm.
In FIG. 1B, a photoresist layer 140 is formed on a top surface 145 of polysilicon layer 135. Photoresist layer 140 is then removed from over P-well 110 by one of by one of any number of photolithographic methods known in the art. Then a phosphorus ion implantation is performed. Photoresist layer 140 is of sufficient thickness to about block phosphorus ion implantation into polysilicon layer 135 over N-well 105. The ion implantation is performed to place the peak (the maximum) of the implanted phosphorus distribution concentration (in atm/cm3) proximate to top surface 145 of polysilicon layer 135. Proximate is defined herein as within about 0 nm to about a value of one fourth of the thickness of polysilicon layer 135. (See also FIG. 4, distances D1A and D1B). The ion implantation is further performed so that the concentration distribution profile of implanted phosphorus is such as to not significantly affect the overall P dopant level of P-well 110. The phosphorus ion implant concentration distribution profile is illustrated in FIG. 4 and described infra. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, phosphorus is implanted at a dose of about 5E14 atm/cm2 to about 5E16 atm/cm2 at an energy of about 30 KeV or less. Arsenic may be substituted for phosphorus and the arsenic. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, arsenic is implanted at a dose of about 5E14 atm/cm2 to about 5E16 atm/cm2 at an energy of about 60 KeV or less. Photoresist layer 140 is then removed. In other examples, the phosphorus and arsenic doses and energies should be scaled proportionally to the thickness of polysilicon layer 135.
In FIG. 1C, a photoresist layer 150 is formed on top surface 145 of polysilicon layer 135. Photoresist layer 150 is then removed from over N-well 105 by one of any number of photolithographic methods known in the art. Then a boron ion implantation is performed. Photoresist layer 150 is of sufficient thickness to about block boron ion implantation into polysilicon layer 135 over P-well 110. The ion implantation is performed to place the peak of the implanted phosphorus distribution concentration profile proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration distribution profile of implanted boron is such as to not significantly affect the overall N dopant level of N-well 105. Photoresist layer 150 is then removed.
In FIG. 1D, a nitrogen containing species ion implantation is performed. The ion implantation is performed to place the peak of the implanted nitrogen species distribution concentration profile proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration of implanted nitrogen penetrating into either gate dielectric layer 125, N-well 105 and P-well 110 is not significant. The nitrogen ion implant concentration distribution profile is illustrated in FIG. 4 and described infra. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, nitrogen (as N) is implanted at a dose of about 1E14 atm/cm2 to about 4E15 atm/cm2 at an energy of about 20 KeV or less. Other suitable nitrogen species include but is not limited to N, N2, NO, NF3, N2O and NH3. In other examples, the nitrogen dose and energy should be scaled proportionally to the thickness of polysilicon layer 135. The steps illustrated in FIGS. 5A and 5B are next performed.
For the first, as well as the second and third embodiments of the present invention, one intent of the phosphorus (or arsenic), boron and nitrogen containing species ion implantations is to keep a maximum amount as possible of implanted species contained within the polysilicon layer at the time of ion implantation as well as after various later heat cycles and to keep a minimum amount as possible of implanted species from penetrating through the polysilicon layer into the underlying layers or into the substrate. Thus, the ion implantations are shallow (low energy) with concentration peaks close to the surface of the polysilicon and concentration tails that fall off to very low concentrations while still within the polysilicon. Thus the implanted species is essentially contained within the polysilicon layer. Less than about 2E12 atm/cm2 of any of the ion implanted species is intended to penetrate into substrate in the case of a polysilicon line or into the gate dielectric layer or N-well or P-well in the gate of polysilicon gates.
FIGS. 2A through 2D are partial cross-sectional views illustrating initial steps for fabricating doped polysilicon lines and gates according to a second embodiment of the present invention. FIG. 2A is identical to FIG. 1A. In FIG. 2B, a nitrogen containing species ion implantation is performed. The ion implantation is performed to place the peak of the implanted nitrogen species distribution concentration profile proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration of implanted nitrogen penetrating into either gate dielectric layer 125, N-well 105 and P-well 110 is not significant. The nitrogen ion implant concentration distribution profile is illustrated in FIG. 4 and described infra. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, nitrogen (as N) is implanted at a dose of about 1E14 atm/cm2 to about 4E14 atm/cm2 at an energy of about 20 KeV or less. Other suitable nitrogen species include but is not limited to N2, NO, NF3, N2O and NH3. In other examples, the nitrogen dose and energy should be scaled proportionally to the thickness of polysilicon layer 135.
In FIG. 2C, a photoresist layer 155 is formed on top surface 145 of polysilicon layer 135. Photoresist layer 155 is then removed from over P-well 110 by one of any number of photolithographic methods known in the art. Then a phosphorus ion implantation is performed. Photoresist layer 155 is of sufficient thickness to about block phosphorus ion implantation into polysilicon layer 135 over N-well 105. The ion implantation is performed to place the peak of the implanted phosphorus distribution concentration proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration distribution profile of implanted phosphorus is such as to not significantly affect the overall P dopant level of P-well 110. The phosphorus ion implant concentration distribution profile is illustrated in FIG. 4 and described infra. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, phosphorus is implanted at a dose of about 5E14 atm/cm2 to about 5E16 atm/cm2 at an energy of about 30 KeV or less. Arsenic may be substituted for phosphorus. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, arsenic is implanted at a dose of about 5E14 atm/cm2 to about 5E 16 atm/cm2 at an energy of about 60 KeV or less. In other examples, the phosphorus and arsenic doses and energies should be scaled proportionally to the thickness of polysilicon layer 135. Photoresist layer 155 is then removed.
In FIG. 2D, a photoresist layer 160 is formed on top surface 145 of polysilicon layer 135. Photoresist layer 160 is then removed from over N-well 105 by one of any number of photolithographic methods known in the art. Then a boron ion implantation is performed. Photoresist layer 160 is of sufficient thickness to about block boron ion implantation into polysilicon layer 135 over P-well 110. The ion implantation is performed to place the peak of the implanted phosphorus distribution concentration profile proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration distribution profile of implanted boron is such as to not significantly affect the overall N dopant level of N-well 105. Photoresist layer 160 is then removed. The steps illustrated in FIGS. 5A and 5B are next performed.
FIGS. 3A through 3D are partial cross-sectional views illustrating initial steps for fabricating doped polysilicon lines and gates according to a third embodiment of the present invention. FIG. 3A is identical to FIG. 1A. In FIG. 3B, a photoresist layer 165 is formed on top surface 145 of polysilicon layer 135. Photoresist layer 165 is then removed from over P-well 110 by one of any number of photolithographic methods known in the art. Then a phosphorus ion implantation is performed. Photoresist layer 165 is of sufficient thickness to about block phosphorus ion implantation into polysilicon layer 135 over N-well 105. The ion implantation is performed to place the peak of the implanted phosphorus distribution concentration proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration distribution profile of implanted phosphorus is such as to not significantly affect the overall P dopant level of P-well 110. The phosphorus ion implant concentration distribution profile is illustrated in FIG. 4 and described infra. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, phosphorus is implanted at a dose of about 5E14 atm/cm2 to about 5E16 atm/cm2 at an energy of about 30 KeV or less. Arsenic may be substituted for phosphorus. In one example, with polysilicon layer 135 having a thickness of about 0.15 nm, arsenic is implanted at a dose of about 5E 14 atm/cm2 to about 5E16 atm/cm2 at an energy of about 60 KeV or less. In other examples, the phosphorus and arsenic doses and energies should be scaled proportionally to the thickness of polysilicon layer 135.
In FIG. 3C, a nitrogen containing species ion implantation is performed. Photoresist layer 165 is of sufficient thickness to about block nitrogen species ion implantation into polysilicon layer 135 over N-well 105. The ion implantation is performed to place the peak of the implanted nitrogen species distribution concentration profile proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration of implanted nitrogen penetrating into either gate dielectric layer 125 and P-well 110 is not significant. The nitrogen ion implant concentration distribution profile is illustrated in FIG. 4 and described infra. In one example, nitrogen (as N) is implanted at a dose of about 1E14 atm/cm2 to about 4E15 atm/cm2 at an energy of about 20 KeV or less. In other examples, the nitrogen dose and energy should be scaled proportionally to the thickness of polysilicon layer 135. Other suitable nitrogen species include but is not limited to N2, NO, NF3, N2O and NH3. Photoresist layer 165 is then removed.
In FIG. 3D, a photoresist layer 170 is formed on top surface 145 of polysilicon layer 135. Photoresist layer 170 is then removed from over N-well 105 by one of any number of photolithographic methods known in the art. Then a boron ion implantation is performed. Photoresist layer 170 is of sufficient thickness to about block boron ion implantation into polysilicon layer 135 over P-well 110. The ion implantation is performed to place the peak of the implanted phosphorus distribution concentration profile proximate to top surface 145 of polysilicon layer 135. The ion implantation is further performed so that the concentration distribution profile of implanted boron is such as to not significantly affect the overall N dopant level of N-well 105. Photoresist layer 170 is then removed. The steps illustrated in FIGS. 5A and 5B are next performed.
The present invention may be practiced by (1) fully matching ion implantation concentration profiles (concentration vs. ion implanted distance) of N-dopant (i.e. phosphorus or arsenic) and nitrogen species at the same distance into the polysilicon, by (2) matching ion implantation concentration profiles of N-dopant and nitrogen species, within a predetermined concentration range, at the same distances into the polysilicon, by (3) matching, within a predetermined concentration range, the surface concentrations of N-dopant and nitrogen in the polysilicon, or by (4) by matching, within a predetermined concentration range, peak concentrations of N-dopant and nitrogen at the same distance into the polysilicon.
FIG. 4 is a plot of concentration of implanted species versus distance from a top surface of a doped polysilicon layer according to the present invention. In FIG. 4, curve 175 (N-dopant) and 180 (nitrogen species) are illustrated using option (2), matching ion implantation concentration profiles of N-dopant and nitrogen species, within a predetermined concentration range, at the same distances into the polysilicon. That is, an equation defining curve 175 and an equation defining curve 180 would yield, for the same distance from the top surface of the polysilicon, a concentration of implanted species within predetermined range of concentration of each other. In a full ion implantation profiles match, option (1) curves 175 and 180 would overlay.
In FIG. 4, the N-dopant (phosphorus or arsenic) ion implantation concentration distribution profile is indicated by curve 175 and the nitrogen species ion implantation concentration distribution profile is indicated by curve 180. While curve 180 is illustrated above curve 175, curve 175 could be above 180. Also curve 175 and curve 180 could cross at one or more points. The exact relationship between curves 175 and 180 is determined by the specific ion implant dose and energy or the N dopant and the specific ion implant dose and energy or the nitrogen. The surface distribution concentration C2A of curve 175 and C2B of curve 180 occur respectively at distance 0 into the polysilicon layer. In one example, C2A is between about 1E18 atm/cm3 and about 1E21 atm/cm3 and concentration C2B is about 1E18 atm/cm3 and about 1E22 atm/cm3. The ranges of values for C2A and C2B may overlap.
The peak distribution concentration C3A of curve 175 and C3B of curve 180 occur respectively proximate to the surface of the polysilicon at distance D1A and D1B into the polysilicon layer. In one example, C3A is between about 1E18 atm/cm3 and about 1E22 atm/cm3 and concentration C3B is about 1E18 atm/cm3 and about 1E21 atm/cm3. The ranges of values for C3A and C3B may overlap.
In one example D1A is between about 0 nm and about ⅓ the thickness of the polysilicon and depth D1B is about 0 nm to about ⅔ the thickness of the polysilicon. The ranges for values for D1A and D1B may overlap.
A concentration C1 is defined in FIG. 4 for curve 175 at a distance D2A and for curve 180 at a distance D2B into the polysilicon layer. D2A is between about 10 nm and the thickness of the polysilicon and D2B is between about 50% and about 150% of D2A. Concentration C1 is a concentration at which an insignificant amount to none of the ion implanted species exists hence essentially the implanted N dopant species and implanted nitrogen containing species are contained with the polysilicon layer. An insignificant amount of implanted species is defined as an amount of implanted species, that if present, would not significantly effect chemical processes or electrical parameters of the polysilicon layer (or gate dielectric layer or P-well) in which the implanted species is present.
Table I summarizes the relationship between curve 175 (N Dopant) and curve 180 (Nitrogen species). TABLE I
Minimum Maximum
Value Value
N Dopant Surface Concentration (C2A) about 1E18 atm/cm3 about 1E22 atm/cm3
Nitrogen Species Surface about 1E18 atm/cm3 about 1E21 atm/cm3
Concentration(C2B—
N Dopant Peak Concentration (C3A) about 1E18 atm/cm3 about 1E22 atm/cm3
Nitrogen Species Peak about 1E18 atm/cm3 about 1E22 atm/cm3
Concentration(C3B)
N Dopant Peak Depth (D1A) about 0 nm about equal to ⅓ the
polysilicon thickness
Nitrogen Species Peak Depth (D1B) about 0 nm about equal to ⅔ the
polysilicon thickness
N Dopant and Nitrogen Species Not Applicable about 1E15 atm/cm3
Insignificant Concentration (C1)
N Dopant Insignificant Concentration 10 nm about equal to the full
Depth (D2A) thickness of the
polysilicon
Nitrogen Species Insignificant about 150% of D1A about 150% of D2A
Concentration Depth (D2B)
Also in FIG. 4, the gate dielectric layer occurs between a distance D3 and D4. Distance D3 is the same as the thickness of the polysilicon layer discussed supra in reference to FIG. 1A and (D4-D3) is the thickness of the gate dielectric layer discussed supra in reference to FIG. 1A.
FIGS. 5A and 5B are partial cross-sectional views illustrating common intermediate steps for fabricating doped polysilicon lines and gates according to the present invention. In FIG. 5A, polysilicon layer 135 (see FIG. 1D, 2D or 3D} is etched into gate electrodes 185A and 185B. Formation of gate electrodes 185A and 185B may be accomplished by one of any number of plasma etch processes selective to etch polysilicon over oxide well known in the art.
In FIG. 5B, an oxidation is performed to simultaneously grow a thermal oxide layer 190A over sidewalls 195A and a top surface 200A of gate electrode 185A and a thermal oxide layer 190B over sidewalls 195B and a top surface 200B of gate electrode 185B. The width of gate electrode 185A at top surface 200A and the width of gate electrode 185B at top surface 200B are both about equal to W1. Gate electrode 185A is doped P type and gate electrode 185B is doped N type. Gate electrode 185B (and possibly gate electrode 185A depending upon which embodiment of the present invention is used prior to the thermal oxidation step) has also been nitrogenated by the nitrogen ion species ion implantation described supra. This reduces (retards) the thermal oxidation rate of N-doped polysilicon. In one example, the thermal oxidation rate of N-doped polysilicon is retarded to be about the same as the thermal oxidation rate of P-doped polysilicon. An example of a thermal oxidation is a furnace oxidation performed for in about a 97% O2 and about 3% HCL generating gas a temperature of about 750° C. for 35 minutes which will grow about 40 angstroms of Si)2 on <100> single-crystal silicon. The steps illustrated in FIGS. 7A through 7E are next performed.
FIG. 6 is a partial cross-sectional view of a problem solved by the present invention. In FIG. 6, the situation that would otherwise prevail if the nitrogen species ion implantation had not been performed. After thermal oxidation, gate electrode 190C has a width W2 at a top surface 200C (where W2 is less than W1) because N-doped polysilicon oxidizes at a faster rate than P-doped polysilicon. The situation wherein the N-dopant concentration is higher near a top surface 200C of gate electrode 185C is illustrated.
Table II illustrates the effect of nitrogen species ion implantation: TABLE II
Thermal Oxide Thickness Polysilicon Width at Top
Nitrogen Implant N-doped P-doped N-doped P-doped
Energy and Dose Polysilicon Polysilicon Polysilicon Polysilicon
NONE 149 Å 61 Å 13 nm 28 nm
6.3 KeV, 5E15 atm/cm2 64 Å 58 Å 23 nm 27 nm
6.3 KeV, 1E16 atm/cm2 52 Å 58 Å 23 nm 30 nm
FIGS. 7A through 7E are partial cross-sectional views illustrating common last steps for fabricating doped polysilicon lines and gates according to the present invention. FIG. 7A is identical to FIG. 5B. In FIG. 7B, dielectric spacers 205A and 205B are formed over thermal oxide layers 190A and 190B on sidewalls 195A and 195B of gate electrodes 185A and 185B respectively. Spacers 205A and 205B may be formed by deposition of a conformal material (for example, silicon nitride) followed by a reactive ion etch (RIE) to remove the conformal material from surfaces perpendicular to the direction of the ion flux. Next, well known in the art extension and/or halo and source drain ion implants are performed to P+ source drains 210 in N-well 105 and N+ source drains 215 in P-well 115. Additional spacers may be formed between various extension, halo and source/drain ion implants.
In FIG. 7C, gate dielectric layer 125 is removed wherever the gate dielectric layer is not protected by gate electrodes 185A and 185B and by spacers 105A and 205B. (The gate dielectric on the sidewalls of the gate electrodes also protects the underlying gate dielectric layer.) Also, thermal oxide layer 190A and 190B on top surfaces 200A and 200B of gate electrodes 185A and 185B respectively is removed. Gate dielectric layer 125 and thermal oxide layer 190A and 190B removal may be accomplished, for example, using a dilute aqueous HF containing solution.
In FIG. 7D a metal layer 220 is deposited. Metal layer 220 may be nickel, titanium, platinum or cobalt. In FIG. 7E, a portion of metal layer 220 in contact with gate electrodes 185A and 185B and with P+ source drains 210 and N+ source/drains 215 is converted to a metal silicide 225 by annealing and removing unreacted metal layer 220 by methods well known in the art. Fabrication of a PFET 230 and an NFET 235 having similar gate electrode linewidths and resistivity is now complete.
Thus, the present invention provides a method of fabricating doped polysilicon lines and gates with improved linewidth control.
The description of the embodiments of the present invention is given above for the understanding of the present invention. It will be understood that the invention is not limited to the particular embodiments described herein, but is capable of various modifications, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention.
