Graded Extended Drain Concept for Reduced Hot Electron Effect

The present invention relates to field effect transistor devices and, more particularly, to such devices containing differentially-doped subsurface structures which provide reduced electric field strength levels at the interface boundary between the gate oxide and the semiconductor body.

One of the most useful and advantageous electrical devices is the field effect transistor (FET). Very large scale integrated (VLSI) circuits have been created using metal-oxide-semiconductor (MOS) field effect devices. These so-called MOSFET devices take their name from the structure of the device in the vicinity of the gate. The gate electrode typically comprises a metal layer insulated by a silicon oxide layer from a doped semiconductor layer beneath the oxide. In the present invention, the problem addressed is the trapping of hot or energetic electrons in the gate oxide region, particularly in those regions adjacent to the device drain.

Decreased cost and improved performance provide an impetus for miniaturizing metal oxide semiconductor devices even further. Continuing advances in lithography and etching will permit even greater reductions in device size. Unfortunately, for reasons such as the compatibility with existing systems, the power supply voltages used in such devices usually fail to decrease with increasing circuit density. As a result of this violation of ideal scaling, electric field strength levels within the device increase as geometric distances shrink. High field effects within MOS devices include carrier mobility reduction and channel hot electron instability in MOSFETs, particularly n-channel MOSFETS, or NMOSFETs. Electrons flowing from the source to drain regions of an n-channel FET gain energy from the lateral electric field component. This component of electric field is parallel to the interface between the silicon semiconductor and the silicon oxide insulating layer beneath the gate electrode. Energetic electrons tend to surmount the silicon/silicon-oxide energy barrier and are trapped within the oxide gate insulator or generate undesirable interface states by mechanisms which are not yet fully understood.

Hot or energetic electrons from the channel current have a tendency to acquire sufficient energy to surmount the barrier between the semiconductor and the gate oxide and as a result, are injected into the oxide. Subsequent trapping of the injected electrons tends to cause device instabilities such as threshold-voltage shift and transconductance degradation. Accordingly, it is highly desirable that hot-electron resistant MOSFETs be devised to prevent such instabilities. These instabilities impose undesirably severe limitations on the further miniaturization of MOSFET devices in VLSI circuits.

While improvements in hot electron effect problems can be partially effected by increasing the thickness of the gate oxide, this is an undesirable option because it reduces gate control effects and results in slower operating devices.

It is seen that the hot electron problem is certainly undesirable and may eventually limit the minimum gate length for MOSFETs. Although improving the insulator quality with respect to hot electron trapping effects may be possible, most efforts at improving channel hot electron reliability have focused on reducing the lateral electric field strength within the FET channel. Such attempts include graded source/drain junctions, double-diffused source/drain regions, and an oxide sidewall spacer adjacent to the gate electrode and oxide. All of these methods reduce the lateral field strength by making the n.sup.+ -p drain-channel transition less abrupt. The instant inventors have found that the oxide sidewall spacer can increase the maximum safe operating voltage by as much as 2 volts. A maximum gain of 1.5 volts per 1,000 angstroms of spacer length is possible. It is additionally noted that any method or structure which is directed at relieving the hot electron effect problem should also be compatible with present VLSI semiconductor processing steps.

In accordance with a preferred embodiment of the present invention, a buried spacer is employed so as to move high density current flow away from the silicon/silicon-oxide interface boundary. In particular, a lightly doped region is developed by ion implantation in source or drain regions (preferably both) on either side of the gate electrode by means of a high voltage field. This lightly doped region lies below and spaced apart from the silicon/silicon-oxide boundary. This lightly doped region is first formed by implantation at high voltage with a more diffusive dopant such as phosphorous. A second, light or low concentration dopant such as arsenic, is also implanted in this same buried region. However, because of the greater capability of the phosphorous for diffusion, a more gradually graded channel structure is produced. Silicon oxide spacers on either side of the gate electrode are then formed by a process such as reactive ion etching, which is anisotropic. A second ion implantation operation is performed at an increased dopant concentration, but at a lower ion implant field strength. However, the higher dopant concentrations result in deeper semiconductor penetration. As a result of this process, more lightly doped and graded regions extend from the source and drain regions beneath the semiconductor surface. it is between these two source and drain extensions that the highest current density is most likely achieved. This results in a reduction of hot electron production in the vicinity of the gate electrode. The lateral electric field strength, particularly the field strength immediately beneath the gate oxide, is significantly reduced.

Accordingly, it is an object of the present invention to provide a method for the manufacture of MOSFET devices exhibiting greater immunity to hot electron effects.

It is also an object of the present invention to provide improved NMOSFET manufacturing processes.

It is yet another object of the present invention to provide a means for reducing the lateral electric field strength beneath the gate oxide in a MOSFET device.

It is a still further object of the present invention to provide a method for decreasing MOSFET circuit size without exacerbating hot electron effects.

Lastly, but not limited hereto, it is an object of the present invention to provide MOSFET, and in particular NMOSFET devices, in which hot electron effects are significantly mitigated.

The subject matter of which is regarded as the invention is particularly pointed out and distinctly claimed in the concluding portion of the specification. The invention, however, both as to organization and method of practice, together with further objects and advantages thereof, may best be understood by reference to the following description taken in connection with the accompanying drawing in which:

The FIGURE is a cross-sectional side elevation view of a MOSFET device produced in accordance with the present invention;

The FIGURE illustrates an NMOSFET device in accordance with the present invention. Except for the specific processing employed to carry out the purposes of the present invention, device fabrication is accomplished in a conventional fashion. FIG. 1 particularly indicates an underlying n-doped substrate 99. In accordance with conventional processes, field oxide layer 115 is grown on substrate 99 by exposing the substrate to oxygen at a high temperature, so as to form an insulating layer of silicon oxide 115. This layer is etched in a patterned fashion so as to define active areas on the chip or wafer. A p-doped well 100 is typically thereafter formed by ion implantation through the active area openings in the field oxide. A thin silicon oxide layer is then grown over the entire substrate. This oxide layer ultimately forms gate oxide 116 which comprises the structure which is subject to the hot electron trapping effect discussed above. Over this thin oxide layer, a layer of polycrystalline silicon is deposited and heavily doped with an n-type dopant so as to provide a material exhibiting high electrical conductivity. This layer eventually forms gate electrode 130. Alternatively, metal rather than doped polycrystalline silicon may be employed for the gate electrode material. In any event, at this stage in the process, a mask is generally employed to produce gate electrode and electrode interconnection patterns where desired. These patterns are created by selective removal of the polycrystalline silicon or metal gate electrode material which typically leaves a thin oxide layer in the active area.

At this stage in the semiconductor manufacturing process, special process steps are employed to effect the formation of the structure illustrated in FIG. 1. More particularly, at this point in the process, a light concentration of n-type dopant such as phosphorous is deposited in a buried layer. This deposition occurs by means of ion implantation at a voltage of approximately 90 keV. This results in the formation of lightly doped regions 103 and 104 (between the dotted lines) at a depth of approximately 1,000 angstroms below the silicon/oxide interface. Because of the diffusivity properties of phosphorous, lightly doped n.sup.- regions 103 and 104 extend inwardly toward the gate region to form diffused protrusions 105 and 106. It is these protrusions which provide for superior grading characteristics in the buried channel device of the present invention. In accordance with the present invention, a second light doping of arsenic is performed at approximately a potential of 200 keV. Thus, regions 103 and 104 are doped with a light concentration of both arsenic and phosphorous. However, because the phosphorous has greater diffusion mobility than does the arsenic dopant, diffused grading structures 105 and 106 are present. It should be noted that regions 103, 104, 105, and 106 extend from the field oxide 115 inwardly to the left and right edges of gate electrode 130. This doping is performed at an implant dosage of between approximately 10.sup.12 and 10.sup.14 dopant atoms per square centimeter of source or drain region. In contrast, prior art doping process steps are not carried out so as to provide such a graded structure.

At this stage of the process, another silicon oxide layer is deposited on the surface of the wafer or chip. However, in contrast to other methods for depositing an oxide layer, in this process step chemical vapor deposition is typically employed since it is desired to cover the entire surface of the device. A reactive ion etching step is then carried out, typically in an atmosphere of trifluoromethane (CHF.sub.3), to anisotropically etch away most of this silicon layer except in those areas exhibiting sharp vertical transitions such as on either side of gate electrode 130. The anisotropic ethcing is advantageous in that it leaves oxide spacers 117' on either side of gate electrodes 130. These spacers provide an important function in mitigating the effects of hot electrons generated near the oxide/semiconductor interface below the gate and more particularly in the vicinity between the gate and the device drain. Additionally, spacers 117' also provide a form of self aligned mask for the next process step which is the ion implantation or diffusion of the source and drain regions. In accordance with the present invention, a heavier, n.sup.+, doping is applied to the source and drain regions of the active area. Typically, this second doping process step is performed at an implant dosage of about 8.times.10.sup.15 dopant atoms per square centimeter. This results in the formation of regions 101 and 102 in FIG. 1 these regions typically extending to a depth of between about 1,500 and 4,000 angstroms. Because of the masking effect produced by spacers 117', only a small n.sup.- doped region of the source and drain is left to extend inwardly from regions 101 and 102. These are the regions 105 and 106 which form a part of layers 103 and 104 as shown. Typically each gate side spacer 117' is between about 0.1 and 0.2 microns in width. Accordingly, the lightly doped portions of the source and drain region extend inwardly toward the gate for a corresponding distance of between approximately 0.1 and 0.2 microns.

In accordance with the present invention, it is also possible to employ gate spacers 117' which comprise polysilicon material. In this embodiment, lightly doped regions 103, 104, 105, and 106 can be implanted at more conventional voltage levels, say for example, 90 keV. In this embodiment, buried channels are formed as the doped polysilicon or metal spacer is biased to a positive voltage via a doped polysilicon gate. In the event that it is desired to employ doped polycrystalline silicon spacers, reactive ion etching is employed to effect removal of the polycrystalline silicon material in the gate masking portion of the process.

In the present invention, it is noted that the phosphorous implant and subsequent diffusion may be performed before or after the arsenic implant or during the source/drain implant. The present invention has the advantage of having a better controlled graded-junction because the phosphorous concentration required to grade the lightly doped n.sup.- extended drain is significantly lower than that required for grading the heavily doped n.sup.+ drain.

Accordingly, it is seen from the above that the methods and structure of the present invention provide a viable solution to the problem of hot electron entrapment in the oxide layer of a MOSFEt device. It is also seen that the present invention provides a simplified process for producing the desired structural implant modifications without departing from conventional processing methodologies. It is also seen that the present invention permits the shrinkage of MOSFET circuits to densities which are even greater than those currently achievable without the necessity of scaling down supply voltages. Accordingly, devices manufactured in accordance with the present invention are much more readily incorporated in currently existing systems. It is also seen that the devices of the present invention are much better able to undergo voltage stress without significant device characteristic impairment. It is additionally noted that, while it is generally preferred that both the source and drain regions possess the buried channel structure of the present invention, it is possible to employ such a structure in only one of these regions. However, in transistors in which the current flow is unidirectional, it is generally preferred to employ such structures only in the drain region.

While the invention has been described in detail herein in accord with certain preferred embodiments thereof, many modifications and changes therein may be effected by those skilled in the art. Accordingly, it is intended by the appended claims to cover all such modifications and changes as fall within the true spirit and scope of the invention.