Technique to obtain high mobility channels in MOS transistors by forming a strain layer on an underside of a channel

[0002] The above-referenced application is related to U.S. patent application Ser. No. 09/752,335, entitled “Technique to Produce Isolated Junctions by Forming an Insulation Layer”, which has a filing date of Dec. 29, 2000.

[0003] A method of obtaining a high mobility channel in a metal-oxide-semiconductor (MOS) transistor is described. More specifically, the present invention describes a method for forming a strain layer on an underside of a channel in the MOS transistor in order to produce a mechanical stress in the channel.

[0004] A conventional metal-oxide-semiconductor (MOS) transistor generally includes a semiconductor substrate, such as silicon, having a source, a drain, and a channel positioned between the source and drain. A gate stack comprised of a conductive material (a gate conductor), an oxide layer (a gate oxide), and sidewall spacers, is typically located above the channel. The gate oxide is typically located directly above the channel, while the gate conductor, generally comprised of polycrystalline silicon (polysilicon) material, is located above the gate oxide. The sidewall spacers protect the sidewalls of the gate conductor.

[0005] Generally, for a given electric field across the channel of an MOS transistor, the amount of current that flows through the channel is directly proportional to a mobility of carriers in the channel. Thus the higher the mobility of the carriers in the channel, the more current can flow and the faster a circuit can perform when using high mobility MOS transistors.

[0006] One way to increase the mobility of the carriers in the channel of an MOS transistor is to produce a mechanical stress in the channel. A current method for stressing the channel includes depositing a strain layer on an upper surface of the channel. Using this method, after depositing the strain layer, a gate stack is fabricated and the transistor is realized by implanting dopants into the semiconductor substrate forming a source on one side of the gate stack and a drain of the other side of the gate stack. After the dopants are implanted, a high temperature anneal is required to place the dopants on a crystallographic site in the silicon. A disadvantage of this method is that the high temperature anneal tends to relax the strain in the channel to a lower strain level, usually resulting in dislocations and/or diffusion in the channel crystal, decreasing the desired mechanical stress in the channel. As a result, it is difficult to produce a desirable increase in the mobility of the carriers in the channel using the prior art method.

[0007] The present invention addresses some of the shortcomings noted above.

[0008] The present invention is illustrated by way of example and is not limited by the accompanying figures:

[0009] FIG. 1 is a side cross-sectional view of a partially fabricated MOS transistor showing a first removal of a portion of a semiconductor substrate and a formation of a set of sidewall spacers.

[0010] FIG. 2 is a side cross-sectional view of a partially fabricated MOS transistor showing a second removal of a portion of a semiconductor substrate to a structure shown in FIG. 1.

[0011] FIG. 3 is a side cross-sectional view of a partially fabricated MOS transistor showing a third removal of a portion of a semiconductor to a structure shown in FIG. 2.

[0012] FIG. 4 is a side cross-sectional view of a partially fabricated MOS transistor showing a formation of a strain layer on a portion of an underside of a channel to a structure shown in FIG. 3.

[0013] FIG. 5 is a side cross-sectional view of a partially fabricated MOS transistor showing a removal of a set of sidewall spacers and a formation of a source tip and a drain tip adjacent to the sidewalls of the channel to a structure shown in FIG. 4.

[0014] FIG. 6 is a side cross-sectional view of an MOS transistor showing a formation of a source and a drain, and a formation of salicide adjacent to the source, the drain and a gate conductor to a structure shown in FIG. 5.

[0015] FIG. 7 is an isometric sectional view of an embodiment of an MOS transistor of the present invention.

[0016] FIG. 8 is a side cross-sectional view of a partially fabricated MOS transistor showing a semiconductor substrate formed adjacent to an insulator. Also shown is a first removal of a portion of the semiconductor substrate and a formation of a set of sidewall spacers.

[0017] FIG. 9 is a side cross-sectional view of a partially fabricated MOS transistor showing a second removal of a portion of a semiconductor substrate to a structure shown in FIG. 8.

[0018] FIG. 10 is a side cross-sectional view of a partially fabricated MOS transistor showing a third removal of a portion of a semiconductor to a structure shown in FIG. 9.

[0019] FIG. 11 is a side cross-sectional view of a partially fabricated MOS transistor showing a formation of a strain layer on a portion of an underside of a channel to a structure shown in FIG. 10.

[0020] FIG. 12 is a side cross-sectional view of a partially fabricated MOS transistor showing a removal of a set of sidewall spacers and a formation of a source tip and a drain tip adjacent to the sidewalls of the channel to a structure shown in FIG. 11.

[0021] FIG. 13 is a side cross-sectional view of an MOS transistor showing a formation of a source and a drain, and a formation of salicide adjacent to the source, the drain and a gate conductor to a structure shown in FIG. 12.

[0022] The description discloses a method for forming a strain layer on an underside of a channel in an MOS transistor in order to produce a mechanical stress in the channel, increasing a mobility of carriers in the channel and an apparatus produced from such a method. The following detailed description contains numerous specific details in order to provide a more thorough understanding of the present invention. However, it will be apparent to those skilled in the art to which this invention pertains that the present invention may be practiced without these specific details. In other instances, well-known devices, methods, procedures, and individual components have not been described in detail in order to avoid obscuring the present invention.

[0023] FIG. 1 shows a side cross-sectional view of a partially fabricated MOS transistor 10. The fabrication of the transistor 10 generally begins on a semiconductor substrate 16. The semiconductor substrate 16 is typically comprised of silicon and is typically doped at selected locations to form doped regions. In many cases the semiconductor substrate 16 is doped with boron, producing a p-type silicon region, or arsenic or phosphorous, producing an n-type silicon region, although other dopants may be used.

[0024] In many cases, the transistor 10 is placed in proximity to a large number of transistors or other devices. To help isolate the transistor 10 from these devices, isolation regions 18, typically comprised of an insulator such as an oxide, are typically formed adjacent to the semiconductor substrate 16. A variety of techniques, including known techniques, can be used to form the isolation regions 18. Examples include, but are not limited to, growing a thermal oxide, depositing an oxide or depositing a nitride.

[0025] A gate stack 13 is fabricated atop the semiconductor substrate 16. A variety of techniques, including known techniques, can be used to fabricate the gate stack 13. For example, one gate stack fabrication technique includes depositing a gate oxide layer 14 above the semiconductor substrate 16, and depositing a gate conductor 12, such as polysilicon, above the gate oxide layer 14. Chemical vapor deposition (CVD) is one available technique to form the gate oxide 14 and gate conductor 12. Lithography techniques known in the art can be used to pattern the various gate stack 13 layers. Accordingly, in one embodiment, the gate oxide 14 and the gate conductor 12 are dry etched. Then a first set of sidewall spacers 22, typically comprised of an oxide, is formed adjacent to the gate conductor 12. The gate oxide 14, the gate conductor 12, and the first sidewall spacers 22 are collectively referred to as the gate stack 13. It is to be noted that other techniques can be used to form the gate stack 13.

[0026] FIG. 1 also shows a first removal of a portion of the semiconductor substrate 16 adjacent to a channel 28 of the semiconductor substrate 16, wherein the channel 28 underlies the gate stack 13. In one embodiment, the first removal is accomplished by a self-aligned, vertical, anisotropic etch. The anisotropic etch forms open regions 24aand 24badjacent to the channel 28 in the semiconductor substrate 16. The area remaining under the gate stack 13 and between the open regions 24aand 24bdefines the channel 28. The anisotropic etch continues vertically downward until a desired depth 24dis reached.

[0027] The depth 24dand therefore the channel 28 height should be sufficiently large such that straining an underside 28c(shown in FIG. 4) of the channel 28 will not disrupt the gate oxide 14/semiconductor substrate 16 interface (i.e. the interface between a lower surface of the gate oxide 14 and an upper surface of the channel 28). At the same time, the channel 28 height should be sufficiently small such that a mechanical stress can be produced in the channel 28 by straining the underside 28cof the channel 28.

[0028] In one embodiment, the first removal is accomplished by a timed anisotropic etch, etching into the semiconductor substrate 16 for a predetermined amount of time in order to form the open regions 24aand 24b, wherein the depth 24dof the open regions 24aand 24b, and therefore the channel 28 height, is in the approximate range of 30 to 200 Angstroms.

[0029] After the first removal, a second set of sidewall spacers 26 is formed adjacent to the gate stack 13 and extends into the open regions 24aand 24bto protect a first sidewall 28aand a second sidewall 28bof the channel 28. A variety of techniques, including known techniques, can be used to form the second set of sidewall spacers 26. For example, a thin layer of a second sidewall spacer 26 material, such as a nitride, can be conformally deposited on an upper surface of the transistor 10 and a vertical anisotropic etch can be used to form the second sidewall spacers 26 to a desired distance from the sidewalls 28aand 28bof the channel 28. In one embodiment, the second sidewall spacers 26 and the first sidewall spacers 22 are comprised of different materials. For example, the first sidewall spacers 22 can be comprised of an oxide material and the second sidewall spacers 26 can be comprised of a nitride material.

[0030] FIG. 2 shows a second removal of a portion of the semiconductor substrate 16. In one embodiment, the second removal is accomplished by a vertical anisotropic etch into the semiconductor substrate 16, aligned with the second sidewall spacers 26. The second removal extends the open regions 24aand 24bfurther into the semiconductor substrate 16 forming open regions 30aand 30b. The open regions 30aand 30bare formed to a distance 30dbelow the second sidewall spacers 26 sufficient to allow a removal of a portion of the semiconductor substrate 16 which underlies the channel 28 (shown in FIG. 3). In one embodiment the second removal is accomplished by a timed anisotropic etch, etching into the semiconductor substrate 16 for a predetermined amount of time in order to form the open regions 30aand 30bto the distance 30dbelow the second sidewall spacers 26 in the approximate range of 50 nanometers to 1 micron.

[0031] FIG. 3 shows a third removal of a portion of the semiconductor substrate 16. In one embodiment, the third removal is accomplished by an isotropic etch into the semiconductor substrate 16. Since the third removal is isotropic, the third removal extends the open regions 30aand 30bfurther into the semiconductor substrate 16 forming an open region 32 which underlies a portion of the channel 28, exposing the underside 28cof the channel 28 while leaving the remainder of the channel 28 intact. During the third removal, the second sidewall spacers 26 continue to protect the sidewalls 28aand 28bof the channel 28.

[0032] In one embodiment, the third removal is accomplished by a timed isotropic etch, etching into the semiconductor substrate 16 for a predetermined amount of time until the open region 32 underlies a portion of the channel 28, exposing a portion of the channel underside 28c. In another embodiment, the third removal continues until the entire channel underside 28cis exposed. It is appreciated that the invention may be practiced wherein the channel underside 28cand an upper surface 16cof the semiconductor substrate 16 have profiles that are not flat, such as the triangularly-shaped profiles shown in FIG. 3.

[0033] In another embodiment however, the channel underside 28ccan be made to have a flat profile, for example by growing an etch stop layer (not shown) on the semiconductor substrate 16 at a desired channel 28 height prior to the third removal. In this embodiment, the third removal can be accomplished by an isotropic etch which forms the open region 32 underling the channel 28 and stops on the etch stop. The etch stop can then be removed, producing a flat profile on the underside 28cof the channel 28.

[0034] FIG. 4 shows a formation of a strain layer 34 on a portion of the exposed channel underside 28cin the open region 32. The strain layer 34 can be, but is not limited to, silicon germanium or silicon carbide. The invention may also be practiced with strain layers 34 comprised of another semiconductor material provided that the material has a similar dopant as the dopant in the channel 28 (for example, the strain layer 34 and the channel 28 may have different dopants that both produce p-type silicon), but a different crystal lattice structure or a different thermal expansion coefficient than the channel 28, either of which will produce a mechanical stress in the channel 28, thereby increasing the mobility of the carriers in the channel 28. In addition, the strain layer 34 can be comprised of an insulator having a different crystal lattice structure or a different thermal expansion coefficient than the channel 28.

[0035] It is appreciated that the strain layer 34 can be epitaxially grown on the channel underside 28c, i.e. the stain material 34 can be selectively grown only on crystallographic sites of the channel underside 28cto prevent the strain layer 34 from attaching to the second sidewall spacers 26, thereby facilitating a removal of the second sidewall spacers 26 (shown in FIG. 5).

[0036] Although the strain layer 34 is described as being grown only on the channel underside 28c, the invention may be practiced wherein the strain layer 34 grows on both the channel underside 28cand the upper surface 16cof the semiconductor substrate 16. The area remaining in the open region 32 after the formation of the strain layer 34 forms open regions 42aand 42b.

[0037] FIG. 5 shows a removal of the second sidewall spacers 26 and a formation of a source tip 36aand a drain tip 36b. The second sidewall spacers 26 are removed to expose the previously protected first sidewall 28aand second sidewall 28bof the channel 28. In one embodiment, the nitride second sidewall spacers 26 are removed by dipping the second sidewall spacers 26 in a solution of hot phosphoric acid. The phosphoric acid removes the nitride second sidewall spacers 26 while leaving the remainder of the transistor 10, such as the first sidewall spacers 22, the channel 28, and the strain layer 34, intact.

[0038] Once the sidewalls 28aand 28bare exposed, the source tip 36ais formed on the first sidewall 28ain the open region 42aand the drain tip 36bis formed on the second sidewall 28bin the open region 42b. In one embodiment, the tips 36aand 36bare epitaxially grown on the sidewalls, 28aand 28bto a lateral distance from the sidewalls 28aand 28bin the approximate range of 2 to 500 Angstroms.

[0039] The invention may also be practiced wherein the source tip 36agrows on both the first sidewall 28aand the strain layer 34 and the drain tip 36bgrows on both the second sidewall 28band the strain layer 34.

[0040] In one embodiment, the tips 36aand 36bare comprised of the same or a similar dopant, having an opposite polarity of the dopant contained in the channel 28. For example the tips 36aand 36bcan contain a dopant forming n-type silicon while the channel 28 can contain a dopant forming p-type silicon. The area remaining in the open regions 42aand 42bafter the formation of the tips 36aand 36bforms open regions 44aand 44b.

[0041] FIG. 6 shows a source 40aformed in the open region 44aand a drain 40bformed in the open region 44b. Various materials and methods of forming the source 40aand drain 40bmay be used to practice the invention. In one embodiment, the source 40aand the drain 40bare formed by growing polysilicon material in the open regions 44aand 44bwhile simultaneously growing dopants, for example phosphorus, arsenic or boron, onto crystallographic sites in the silicon. The temperature required to grow the dopants onto crystallographic sites in the silicon is in the approximate range of 600° C. to 700° C. In another embodiment the source 40aand drain 40bare formed by epitaxially growing a single crystal of silicon while simultaneously growing dopants onto crystallographic sites in the silicon. Although the source 40aand drain 40bare described as being comprised of silicon other appropriate semiconducting or conducting materials may be used to form the source 40aand drain 40b.

[0042] Growing dopants in silicon is in contrast to implanting dopants in silicon, which requires a high temperature activation anneal in the approximate range of 900° C. to 1100° C. to place the dopants on crystallographic sites in the silicon. This high temperature activation anneal tends to relax the strain layer to a lower strain level, usually resulting in dislocations and/or diffusion in the channel crystal, decreasing the desired mechanical stress in the channel. The growth temperature, being significantly lower than the activation anneal, makes it much easier to keep the strain intact and therefore produce the desired mechanical stress in the channel 28, thereby increasing the mobility of the carriers in the channel.

[0043] In one embodiment, the source 40aand drain 40bare comprised of silicon having a uniform dopant concentration, for example boron for p-type silicon or arsenic or phosphorous for n-type silicon. In other embodiments, the source 40aand drain 40bhave a graded increasing dopant distribution, a graded decreasing dopant distribution, or sections with higher dopant levels than other sections. Varying the dopant distribution within the source 40aand drain 40bvaries the conductivity of the source 40aand drain 40band can be used to help decrease a current leakage and capacitance in the transistor 10.

[0044] After the source 40aand the drain 40bare formed, salicide (self-aligned silicide) contacts 38 are formed above the source 40a, the drain 40band the gate conductor 12, providing good contact surfaces with low resistivity. A variety of techniques, including known techniques, can be used to form the salicide contacts 38. For example, the salicide contacts 38 can be formed by conformally depositing a metal such as nickel, tungsten or cobalt and providing heat so that the metal combines with any exposed silicon, i.e. the silicon from the source 40a, the drain 40b, and the gate conductor 12, to form the salicide. The unreacted metal, such as the metal adjacent to the isolation regions 18 and the first sidewall spacers 22 can be removed by selectively etching any unreacted metal.

[0045] FIG. 7 shows an embodiment of the invention. In this embodiment, the gate stack 13 is formed above the semiconductor substrate 16. Although a T-shaped gate stack 13 is shown, other configurations may be used to practice the invention. Adjacent to the gate stack 13 is the channel 28. The channel 28 is strained by a strain layer 34, producing a mechanical stress in the channel 28 thereby increasing the mobility of the carriers in the channel 28.

[0046] FIG. 8 shows another embodiment of the invention. The partially fabricated MOS transistor 11 shown in FIG. 8, can be accomplished in the same manner as is described above for the fabrication of the transistor 10 shown in FIG. 1. A difference between the transistor 10 and the transistor 11 is that in the transistor 11, the semiconductor substrate 16 is formed adjacent to an insulation layer 50. The insulation layer 50, like the isolation regions 18, is generally comprised of a dielectric material, such as an oxide. An advantage of forming the semiconductor substrate 16 adjacent to the insulation layer 50, is that the insulation layer 50 helps isolate the transistor 11 from other transistors or devices placed in proximity to the transistor 11.

[0047] FIG. 9 shows the second removal of a portion of the semiconductor substrate 16. All of the methods described above for the second removal in the transistor 10 shown in FIG. 2 can be used in the transistor 11 shown in FIG. 9. However, in the transistor 11, the insulation layer 50 can function as an etch stop, such that the second removal can be accomplished by an anisotropic etch, etching into the semiconductor substrate 16 until it reaches the insulation layer 50.

[0048] FIG. 10 shows the third removal of a portion of the semiconductor substrate 16. All of the methods described above for the third removal in the transistor 10 shown in FIG. 3 can be used in the transistor 11 shown in FIG. 10. However, in the transistor 11, the insulation layer 50 again functions as an etch stop, allowing for a removal of the remainder of the semiconductor substrate 16, excepting the channel 28.

[0049] FIG. 11 shows the formation of the strain layer 34 on a portion of the exposed channel underside 28cin the open region 32 of transistor 11. The strain layer 34 can be formed in the same manner as is described above for the transistor 10 in FIG. 4. However, in many cases the strain layer 34 is comprised of a material that does not grow on the insulation layer 50.

[0050] The transistor 11 shown in FIGS. 12 and 13 can be fabricated in the same manner as is described above for the transistor 10 in FIGS. 5 and 6.

[0051] Embodiments of the invention include a transistor having a strain layer formed on an underside of a channel in order to produce a mechanical stress in the channel, increasing the mobility of the carriers in the channel. Embodiments of the invention allow for the production of a greater amount of mechanical stress in the channel, and therefore a greater increase in the mobility of the carriers in the channel than other methods of straining the channel such as depositing a strain layer on the upper surface of the channel.

[0052] Although specific embodiments, including specific parameters, methods, and materials have been described, various modifications to the disclosed embodiments will be apparent to one of ordinary skill in the art upon reading this disclosure. Therefore, it is to be understood that such embodiments are merely illustrative of and not restrictive on the broad invention and that this invention is not limited to the specific embodiments shown and described.