Patent Application
20070114604
The present invention relates generally to semiconductor devices, and more particularly to a method for implanting multiple extensions on a metal oxide semiconductor field effect transistor.
A MOSFET typically has three or four terminals. A three-terminal MOSFET includes a source, a drain, and a gate. A voltage applied to the gate terminal controls MOSFET channel resistance. The gate region, which is not a junction, is a metal oxide “sandwich” running the length of the channel surface. On the other hand, MOSFET source and drain regions are junctions and not simply contacts.
Taking an n-channel MOSFET for example, the source and drain are shallow n-type regions in an n-channel MOSFET. The gate material, such as polycrystalline silicon, is typically placed over a channel, but separated from the channel by a thin layer of insulating silicon dioxide. The channel is of a p-type material. With no bias voltage applied to the gate, the resistance of the path between the source and drain is high, and there will be no current flowing between the two regions. When an appropriate voltage is applied between the gate and source terminals, the electric field generated penetrates through the oxide and creates a so-called “inversion channel” in the channel underneath. Since the inversion channel is of the same type, i.e. P-type or N-type, as the source and drain, it provides a conduit (or the “channel”) through which current can pass. By varying the voltage between the gate and body, conductivity can be modulated.
Semiconductor device fabrication is a multiple-step sequence to create chips used in everyday electrical and electronic devices. Fabrication of a MOSFET normally includes a silicon substrate wafer as a starting base. A gate oxide is grown and polycrystalline silicon is added to create a gate. Light pattern is projected to photoresist and a chemical washes away material that is illuminated or shaded from the stencil depending on positive or negative exposure; this process is called developing. Ion implantation deposits N-type or P-type material to create a source and drain. Additionally the process can include many other steps such as phosphorus diffusion, annealing, and deposition of material to give the MOSFET a unique bias, resistive characteristics and properties.
However, conventional fabrication methods produce MOSFETs having not only a high resistance between the source, drain and diffusion channel, but also geometry issues in depth or spread related to diffusion during annealing. Furthermore, control issues may also be problematic.
As such, desirable in the art of semiconductor designs are additional MOSFET fabrication methods that provide more robust, more configurable MOSFETs.
In view of the foregoing, the following provides a method to enhance the source and drain layer resistance by using multiple lightly-doped extensions on the source and drain, and by making part of the heavily-doped extensions deep enough to provide more input and output conductivity.
In various embodiments, a MOS transistor structure is disclosed. A gate electrode is disposed on a semiconductor substrate. A first extension of a predetermined impurity type is substantially aligned with the gate electrode in the substrate. A second extension of the predetermined impurity type overlaps with the first extension in the substrate. The first extension has at least one lateral boundary line closer to the gate electrode than that of the second extension. Source and drain regions of the predetermined polarity type overlaps with the first and second extensions in the substrate. The second extension has at least one lateral boundary line closer to the gate electrode than that of the source and drain regions. The source and drain regions are deeper than the second extension, which is deeper than the first extension, so that they collectively reduce lateral abruptness of the source and drain, while maintaining a reduced extension resistance.
The construction and method of operation of the invention, however, together with additional objectives and advantages thereof will be best understood from the following description of specific embodiments when read in connection with the accompanying drawings.
The present disclosure provides methods for form double extension to improve the performance of transistors which will have less input and output resistance and a smoother source and drain layer transition. The following provides a detailed description of two methods for fabricating an improved MOSFET structure.
A source 108 and a drain 110 of the N-MOSFET are created by adding a lightly doped N-type material to the substrate 102 through implantation, which is typically performed by using an atomic accelerator. The implantation process deposits material near the surface and in a uniform process. The process also slightly damages the substrate 102. To fix the damage, the wafer is annealed, or heat treated, in order to modify the material's properties. However, a side effect of the annealing process is diffusion, which is caused by the natural migration of atoms from a more concentrated region to a less concentrated region. Generally, lateral diffusion is undesirable, however sometimes unavoidable, in semiconductor device fabrication as it causes lateral distortion of the device geometry and can create shorts or parasitic capacitance. A length 112 illustrates diffusion of the implanted material. The small overlap of length 112 of the N-type material and the poly 106 form a parasitic capacitance that changes the transistor's speed properties.
In semiconductor fabrication, the addition of layers of materials normally creates a parasitic capacitance between layers. This parasitic capacitance typically works against the design because it causes the transistor to turn on slowly as the parasitic capacitance charges up. A way to minimize parasitic capacitance resulting from lateral diffusion is to use various oxide offset spacers 114 which can be added to the vertical sides of the gate as an option, prior to a specific implant.
The half-completed transistor is then covered with an oxide material. The oxide is chemically etched away in such a way that only sidewall spacers 116 remain. This sidewall spacer 116 is used as a mask for additional selective implantations. Finally, a heavily doped N-type material is implanted and diffused to the substrate 102 to create a deep extension 118 at the source 108 and the drain 110.
The above steps explain the basic steps for making an N-MOSFET. Various ways of manufacturing a MOSFET exist and each depends on the desired properties of the transistor needed. A clear disadvantage from this conventional method is the lateral abruptness degradation with an increasing length 120. The surface area and volume taken by the lightly-doped material is not sufficient in many cases to form a good conductive medium for a required current density. This is due to a poor transition of the resistive properties of the materials used. Additionally, a depth 122 at any given length 112 may not be sufficient for the current flow when the gate is turned on. Another disadvantage is that the ratio between depth 122 and the length 120 ratio remains static. In addition, several annealing processes might encourage lateral diffusion along with vertical diffusion.
In step 214, oxide offset spacers 216 are added to the vertical sidewalls of the gate, which have a width ranging from 50 to 400 angstroms. The oxide offset spacers 216 function as a mask for placing a second extension 218. The second extension 218 of the same type (i.e., using the same type of impurities) is implanted and further diffused through a heating process. This time, the depth of the second extension is deeper than the first extension. This depth is an important factor that affects the resistance of the source and drain. This second extension overlapping the first extension creates a gradual increase in conductivity to the increasing geometric depth and the concentration created by extensions 218 and 220. The extension 220 diffuses under the oxide offset spacers 216, thereby minimizing a parasitic capacitance from being created.
In step 224, the top surface of the transistor-in-process is covered with an oxide. The oxide is etched in such a way where only sidewall spacers 226 remain. Like the oxide offset spacers, the sidewall spacers 226 are formed to function as a mask to self align the source and drain regions 228. The source and drain 228 are implanted and diffused through a heating process. The source and drain are heavily doped with an N-type impurity material. As it is seen, from the first extension to the third extension, each succeeded extension is implanted at a higher energy in order to create a deeper layer.
Advantages of using this fabrication technique arise from the gradual geometric size and impurity material concentration of the extensions, from small to large. This process benefits from rapid thermal processes or the use of high temperature sub-melt laser annealing which can contribute to diffusion-less profiles. Additionally, the fabrication of a transistor with a minimal diffusion distance 230 caused by lateral diffusion and an increase in an extension depth 232 can substantially reduce input and output resistance related to the source and the drain. It is understood that the abruptness of the extension of the source and drain is measured by the lateral diffusion distance 230. As it can be seen on the drawings, the overlapping area between the first and second extension that extends out and underneath the gate structure and the spacers provides a “two-step” decline in depth instead of one step as in the conventional practice. It is said that the two-step decline reduces the overall abruptness of the extension.
In step 316, the oxide offset spacers 310 are thinned by etching, thereby forming thinned spacers 318, which have widths narrower than those of the original “thick” spacers 310. The thinned spacers 318 is intended to help align a shallow implantation of impurities. The etching process creates a new shadow line, as depicted at a corner 320 between the edge of the oxide offset spacer 318 and the surface of the substrate 304 that will become the implant edge for the next extension. A thin layer of lightly-doped N-type impurity material is then implanted and diffused, thereby creating a thin extension 322. At this point, the extension for the source and drain is appropriately formed by the overlapping first and second extension, both formed by having the same type of impurities. As it can be seen from
In step 324, the surface of the transistor is covered with an oxide. The oxide is etched in such a way where only sidewall spacers 326 remain. The source and drain 328 are then implanted and diffused. The source and drain 328 includes a heavily-doped N-type material. The source and drain are implanted at a higher energy in order to create a deeper layer.
The need to implant the thin extension 322 only after implanting the first extension 314 is due to a concern with the thermal budget, or the exposure to fabrication heat cycles. This method can reduce transient enhanced diffusion or damage caused to the crystal during diffusion under the high energies of implantation. The reduction in exposure reduces diffusion of dopant in the source and drain 328, as well as the gradual diffusion of the source and drain layers due to annealing into the inversion channel area, which can cause a punch through if “smeared” long enough.
In various embodiments, the fabrication method specifically mentions N-MOSFET. While examples are given for N-MOSFET, it is understood by those skilled in the art that a P-MOSFET can also be fabricated using the same methods. Moreover, while only the basic steps to MOSFET fabrication are illustrated, additional steps are contemplated for changing a MOSFET's properties to match a given design requirement.
This invention proposes a novel method that uses multiple extensions of different geometries and concentrations. The transistor created by the present invention provides an (source/drain) extension profile that has a minimum lateral diffusion, while achieving minimal extension resistance by having deep vertical diffusion. Advantages of the invention include less diffusion with special low heat anneal processes and deeper source and drain geometries at the outer extremities, thereby leading to more conductivity or less resistance. Multiple extensions may be essential to achieve both short channel control and low extension resistance and thus high drive current.
The above illustration provides many different embodiments or embodiments for implementing different features of the invention. Specific embodiments of components and processes are described to help clarify the invention. These are, of course, merely embodiments and are not intended to limit the invention from that described in the claims.
Although the invention is illustrated and described herein as embodied in one or more specific examples, it is nevertheless not intended to be limited to the details shown, since various modifications and structural changes may be made therein without departing from the spirit of the invention and within the scope and range of equivalents of the claims. Accordingly, it is appropriate that the appended claims be construed broadly and in a manner consistent with the scope of the invention, as set forth in the following claims.
