Selective etching using sonication

[0002] This invention relates generally to selectively etching one material with respect to other materials.

[0003] In a variety of different circumstances, it may be desirable to etch polysilicon selectively. That is, it may be desirable to preferentially etch polysilicon while reducing the etching of other materials.

[0004] One example of a situation where such selectivity may be desirable is in connection with providing dual metal gate technology. Dual polysilicon gates are used in conventional complementary metal oxide semiconductor (CMOS) devices to engineer a desired threshold voltage that may be different between the NMOS and PMOS devices. Unfortunately, as the device's scale becomes smaller, this approach is not effective. When the polysilicon doping level is not sufficiently high, the polysilicon gate depletion effectively increases the gate dielectric thickness by several Angstroms. This negatively impacts the ability to scale gate dielectric thicknesses. Boron penetration and gate resistance may also be issues for such polysilicon gate technology.

[0005] One approach to this problem is to replace the polysilicon gate with a metal gate. More particularly, one metal gate may be utilized for the NMOS devices and a different metal gate may be utilized for the PMOS devices.

[0006] Thus, it may be desirable to form dual metal gate technology from conventional processing steps that use polysilicon. After the polysilicon has been defined to form the gate electrodes for a transistor, the polysilicon may be selectively removed. A different metal may be applied to form each of the NMOS and PMOS transistors.

[0007] Thus, there is a need for a better way to selectively etch materials.

[0008] FIG. 1 is an enlarged cross-sectional view of one embodiment of the present invention at an earlier stage of manufacture;

[0009] FIG. 2 is an enlarged cross-sectional view of the embodiment shown in FIG. 1 after further processing in one embodiment;

[0010] FIG. 3 is an enlarged cross-sectional view of the embodiment shown in FIG. 1 after further processing in one embodiment;

[0011] FIG. 4 is an enlarged cross-sectional view of the embodiment shown in FIG. 1 after further processing in one embodiment; and

[0012] FIG. 5 is an enlarged cross-sectional view of the embodiment shown in FIG. 1 after further processing in one embodiment.

[0013] Sonication during the polysilicon etching process may increase etch selectivity. In one embodiment, the sonication may be ultrasonic, namely the application of sonic energy in the frequency range between approximately 10 kilohertz (kHz) and 100 kHz. In a second embodiment, the sonication may be megasonic, namely the application of sonic energy in the frequency range between approximately 500-1000 kHz.

[0014] In one embodiment, the etch selectivity between p+ and n+ doped polysilicon regions may be significantly increased via sonication. For example, the etch rate of the n+ doped region may be as much as an order of magnitude greater, or more, depending on the actual power of sonication employed, than that of p+ doped or intrinsic polysilicon.

[0015] In one embodiment of a process according to the present invention, wet etching may be performed using sonics to improve etch selectivity. Such a wet etch may be performed by immersing a wafer in an etchant. For example, the wafer (or wafers) may be immersed in a tank, such as a chemical bath, that is equipped for sonication. The equipment for sonication may vary, but in one embodiment using ultrasonics or megasonics transducers located external to the tank may be used to provide sonic waves at the desired frequency.

[0016] While the chemistries used for etching may vary, in one embodiment a hydroxide source, such as tetramethylammonium hydroxide (TMAH) or NH₄OH, as two examples, may be used to etch a polysilicon layer. Other chemistries may also be used to etch polysilicon layers.

[0017] Referring to FIG. 1, conventional NMOS and PMOS transistor structures 12aand 12b, respectively, may be formed on a semiconductor structure 10. Each transistor structure 12 may include a polysilicon gate 14 over a gate dielectric 13. A nitride spacer 16 may be formed on the sidewalls of the gate 14 and an interlayer dielectric 18 may be situated over the structure 10 outside of the spacer 16 in one embodiment.

[0018] To implement a dual metal complementary metal oxide semiconductor device, it is desirable to remove one polysilicon gate 14 and to selectively replace it with another metal. In one embodiment of the present invention the NMOS and PMOS transistor structures 12 may receive different gate metal material.

[0019] Referring to FIG. 2, the polysilicon gate 14 for the NMOS structure 12ahas been removed by a selective etch which is highly selective relative to the p+ doped gate 14 of PMOS structure 12b, as well as to interlayer dielectrics, nitride spacers, metal gates, underlying gate oxides, silicon, and other high dielectric constant materials, to mention a number of examples.

[0020] The selective removal of the polysilicon gate 14 may involve using a 25 percent solution of tetramethylammonium hydroxide (TMAH) in one embodiment. In another embodiment tetraethylammonium hydroxide or another tetra(alkyl)ammonium hydroxide may be used. This etch, when coupled with sonic energy, is particularly selective of n-type doped polysilicon gate material. Thus, this embodiment is particularly applicable to removing the polysilicon gate 14 on the NMOS transistor structure 12a.

[0021] By making the polysilicon gates 14 taller than ultimately desired for both the NMOS and PMOS transistor structures 12, the slight etching, indicated at 20b, of the PMOS transistor 12bgate material 14 reduces the gate 14 height down to its desired height. As a result, the NMOS structure 12amay have its gate 14 removed and, even though the height of the gate 14 of the PMOS structure 12bis also reduced, it is reduced down to the ultimately desired height.

[0022] After the polysilicon gate 14 has been removed in the NMOS transistor structure 12a, a new gate material 15, such as a metal gate material, may be formed or deposited into the void 20a, in accordance with one embodiment of the present invention shown in FIG. 3. The entire structure, after a new metal is deposited, may be polished to reduce its height to correspond to the reduced height of the gate 14 of the PMOS transistor structure 12b.

[0023] In some embodiments, the PMOS transistor structure 12bmay be exposed and the p-type gate 14 may be etched away, as shown in FIG. 4, using any etchant that does not attack the replacement n-type gate material. In such an embodiment, no masking of the n-type gate material may be needed. The etchants described earlier may be used in some embodiments. The p-type polysilicon may then be replaced with a metal 22, as shown in FIG. 5.

[0024] Without sonication, the etch rate for n+ polysilicon is approximately twice the etch rate for p+ polysilicon. In other words, the etch rate between these polysilicon regions is not particularly selective.

[0025] With sonication, the etch rate for n+ polysilicon may be as great as 15 to 20 times the etch rate for p+ polysilicon in some embodiments. In other words, the etch rate between these polysilicon regions is very selective as a result of using sonication. In certain embodiments, sonication may increase the etch rate for n+ polysilicon tenfold. Thus use of sonication during etching causes a significantly increased etch rate for n+ polysilicon, but has no effect on the etch rate of p+ polysilicon.

[0026] This etching process may be completed, in some embodiments, with the PMOS gate exposed because of the high selectivity of n+ polysilicon. Referring to FIG. 3, as a result of the sonication, the n+ polysilicon gate 14 is completely removed, leaving the gate oxide layer 13 exposed. However, because of the reduced amount of time needed to etch the n+ polysilicon gate 14, as a result of sonication-induced selective etching, the p+ polysilicon gate 14 remains substantially intact.

[0027] Because of the high selectivity to NMOS polysilicon, the NMOS gate 14 of the complementary metal oxide semiconductor wafer may be removed without the need to protect the PMOS transistor structure 12b. This results in a savings in cost and time since additional masking would otherwise be required. This additional masking results in additional process steps which result in additional processing time and expense.

[0028] While an embodiment has been provided in which a dual metal gate technology is implemented, the present invention is not so limited. In fact, the techniques described herein may be utilized in a variety of semiconductor processing applications.

[0029] While the present invention has been described with respect to a limited number of embodiments, those skilled in the art will appreciate numerous modifications and variations therefrom. It is intended that the appended claims cover all such modifications and variations as fall within the true spirit and scope of this present invention.