SHALLOW-JUNCTION SEMICONDUCTOR DEVICES
Background of the Invention
This invention relates to shallow-junction semiconductor devices.
In various semiconductor devices, such as the metal-oxide-semiconductor field-effect transistor (MOSFET), it is desirable that a p-n junction of the device be as close to a substrate surface as possible. The present invention provides a means for obtaining p-n junctions which are significantly closer to a semiconductor body surface than was heretofore obtainable. Summary of the Invention
A neutral species is initially implanted into a surface region of a semiconductor body. In one example, the neutral species is implanted to form a layer whose maximum concentration occurs at a depth greater than that of a p-n junction to be subsequently formed. (In another example, the junction is subsequently established at a depth that is greater than the depth of the peak concentration of the neutral-species layer.) A dopant species is then implanted into the surface region at a depth less than the depth of the maximum-concentration of the previously implanted neutral species. Annealing to activate the dopant species is then carried out.
It is theorized that during this annealing step, the neutral-species layer serves to getter point defects in the body of the device. Additionally, this layer serves as a physical barrier to diffusion of dopant species. As a result, the diffusivity of the dopant species in the body is significantly lowered relative to the case in which no neutral-species layer is provided. In any event, the result is that a p-n junction is formed in the body of the device at an extremely shallow depth. Brief Description of the Drawing
FIGS. 1 through 4 are schematic representations of a portion of a MOSFET device at successive stages of a
fabrication sequence that embodies the principles of the present invention. Detailed Description
In accordance with this invention, shallow p-n junctions can be formed in a variety of semiconductor devices. These devices include, for example, p-n diodes, bipolar transistors and MOSFET devices. By way of example, the invention is described in connection with the provision of shallow p-n junctions in a MOSFET device. A portion of such a MOSFET device at an intermediate stage of its fabrication cycle is shown in FIG. 1, such portion comprising a known gate-and-source-and -drain (GASAD) structure. The structure comprises a silicon body 10 having field-oxide (silicon dioxide) portions 12, 14 thereon.
The structure further includes a gate-oxide (silicon dioxide) layer 16, a doped polysilicon layer 18, and a metallic suicide (e.g. tantalum disilicide) layer 20. Also, the structure includes additional silicon dioxide layers 22,24. Openings 25, 26 are defined by the oxide layers 12, 22 and 14, 24. Source and drain regions are later formed in the body 10 in approximate alignment with these openings.
In accordance with the invention, a so-called neutral species is implanted into regions of body 10 defined by the openings 25, 26. Known ion implantation techniques can be used.
The term "neutral species" means ion species that do not produce active carriers in the semiconductor body and that are effective to limit the diffusivity of active species in the body. Such neutral species include carbon, oxygen, argon or any other inert gas. Group IV elements such as silicon, germanium and tin, and nitrogen (minor activity) . ions directed at the FIG. 2 structure are represented by arrows 28, the ions reaching the body 10 only through the openings 25, 26. In FIG. 2, the depth of
O
the peak or maximum concentration of the approximately Gaussian-shaped distribution of the neutral-species implant in the body 10 is schematically depicted by lines 30, 32 formed with x's. Illustratively, the dosage of the neutral-species implant represented in FIG. 2 is selected to provide approximately one or two monolayers of the neutral species at the peak-concentration depth. Additionally, for the illustrative MOSFET device shown, the energy of the incident ions is selected such that the peak concentration of the implanted neutral species occurs approximately 2000 A below the surface of the body 10. In some devices, the peak concentration of the neutral-species implant is selected to occur at a depth greater than that of the p-n junction(s) to be subsequently formed in the body 10. In such devices, the depth of the subsequently formed p-n junction(s) is, for example, approximately one-tenth to three-quarters that of the depth of the peak concentration of the neutral species. (In other devices, described below, the depth of the p-n junction(s) is greater than the depth of the peak concentration of the neutral-species implant.) Various dosages and energies can be used. One set of dosage and energy values for the aforelisted neutral species is as follows: carbon, 5 x 10 15 ions per square centimeter (i/cm2), 80 kilo-electron-volts (keV); oxygen, 5 x 10 15 i/cm946, 80 keV; silicon, 5 x
1015 i/cm2, 150 keV; germanium, 5 x 1015 i/cm2, 300 keV; tin, 5 x 1015 i/cm2, 400 keV; argon, 5 x 10 15 i/cm946, 180 keV; and nitrogen, 5 x 10 15 i/cm*6, 80 keV. For these values, the respective peak-concentration depth of each of the neutral species is approximately 2000 A below the surface of the body 10 shown in FIG. 2.
In some, but not necessarily all cases, the device structure represented in FIG. 2 is next subjected to an annealing step. (For a subsequently introduced active species such as arsenic, it may actually be advantageous
_ O
• - wip
^ .N'
not to anneal the implanted neutral species at this point in the fabrication procedure.) In this annealing step, - damage to the crystalline structure of the body 10 caused by the neutral-species implant is reduced. Further, the implanted species is stabilized and in effect locked in place in the body 10. Also, some gettering of defects and impurities in the body 10 typically occurs during this annealing step.
The annealing is done, for example, at a temperature in the range 700-to-900 degrees Celsius in an inert ambient for about one-half hour. During annealing, no substantial vertical or lateral movement of the implanted neutral species occurs. Nor does any substantial movement occur later during the so-called activation annealing step described below.
Next, an active species is introduced into the structure by any of various known means, e.g., by ion implantation, as indicated by the arrows 34 in FIG. 3. The implanted active species comprises, for example, a pentavalent n-type impurity such as arsenic, phosphorus or antimony, or a trivalent p-type impurity such as boron or gallium.
The depth of the peak or maximum concentration of the approximately Gaussian-shaped distribution of the active-species implant in the body 10 is schematically represented in FIG. 3 by "lines 36, 38 formed with dots.
The peak concentration of the implanted active species is selected to occur relatively close to the top surface of the body 10, e.g., at a depth of approximately 200-to-1000 A.
More specifically, for a carbon neutral species implant having a peak-concentration depth 30, 32 (FIG. 3) of about 2000 A, an arsenic implant having a peak- concentration depth 36, 38 of approximately 200 A is achieved by implanting 4 x 10 i/cπr at 30 keV. For a carbon or nitrogen implant having a peak-concentration depth 30 , 32 ( FIG . 3 ) of about 2000 A , a boron impl ant
OMPI ■Λ ~ Wl?θ
having a peak-concentration depth 36, 38 of approximately 1000 A is achieved by implanting 4 x 1015 i/cm2 at 30 keV.
Subsequently, standard activation annealing of the second-implanted or active dopant species is carried out. During this step, it is believed that gettering of point defects and metallic impurities occurs in the body 10. It appears also that, because of this gettering action, the thermal diffusivity of the active species is substantially reduced relative to what it would have been in the absence of the neutral-species implant. Additionally, it appears that the priorly formed neutral-species implant serves as a physical barrier against diffusion of the active species. However, regardless of the physical phenomena involved, the result of the described process is that the p-n junction formed by the active-species implant occurs at a depth far less than it would have been if the neutral-species implant had not been present. Extremely shallow p-n junctions are thereby formed. For an active-species implant of arsenic, activation annealing is carried out at, for example, about 1000 degrees Celsius for approximately three hours in a standard mildly dry oxidizing atmosphere. The resulting p-n junction is at a depth of approximately 1400 A. without the priorly implanted neutral-species, but with all other processing conditions approximately the same, the p-n junction is at a depth of about 3700 A.
For an active-species implant of boron, activation annealing is carried out at, for example, about 900 degrees Celsius for approximately five hours in a standard mildly dry oxidizing atmosphere. The p-n junction occurs at approximately 3300 A. without the presence of the neutral-species implant, but with all other processing conditions approximately the same, the p-n junction occurs at a depth of about 6700 A.
In the example above in which the active species comprises arsenic, the p-n junction is at a depth less than
the depth of the peak concentration of an implanted neutral species layer. Conversely, in the example above in which the active species comprises boron, the p-n junction is at a depth greater than the depth of the peak concentration of the implanted neutral-species layer. In general, it is feasible to form p-n junctions at a depth in the range of approximately one-tenth to two times the depth of the peak concentration of the implanted neutral-species layer.
It is possible to form the p-n junction at a depth less than the depth of the peak concentration of the implanted neutral-species layer or in a number of ways. For example, the active impurity species can be initially introduced into the device structure at a shallower depth than specified above for boron and/or by activation annealing the structure at a lower temperature than specified above. Or the peak concentration of the neutral- species layer can be initially formed sufficiently deep that, after annealing, the junction is established at a depth less than the depth of the peak concentration of the neutral-species layer.