Method for forming tri-gate FinFET with mesa isolation

[00002] The present invention relates generally to transistors and, more particularly, to fin field effect transistors (FinFETs).

[00003] The escalating demands for high density and performance associated with ultra large scale integration semiconductor devices require design features, such as gate lengths, below 100 nanometers (nm), high reliability and increased manufacturing throughput. The reduction of design features below 100 nm challenges the limitations of conventional methodology.

[00004] For example, when the gate length of conventional planar metal oxide semiconductor field effect transistors (MOSFETs) is scaled below 100 nm, problems associated with short channel effects, such as excessive leakage between the source and drain, become increasingly difficult to overcome. In addition, mobility degradation and a number of process issues also make it difficult to scale conventional MOSFETs to include increasingly smaller device features. New device structures are, therefore, being explored to improve FET performance and allow further device scaling.

[00005] Double-gate MOSFETs represent structures that have been considered as candidates for succeeding existing planar MOSFETs. In double-gate MOSFETs, two gates may be used to control short channel effects. A FinFET is a recent double-gate structure that exhibits good short channel behavior. A FinFET includes a channel formed in a vertical fin. The FinFET structure may be fabricated using layout and process techniques similar to those used for conventional planar MOSFETs.

[00006] Implementations consistent with the present invention provide a tri-gate FinFET formed using fully depleted silicon-on-insulator (FDSOI) mesa isolation. A tri-gate FinFET will have better short-channel control than double-gate and single-gate devices, and will have higher drive current than a double-gate FinFET for the same area.

[00007] Additional advantages and other features of the invention will be set forth in part in the description which follows and, in part, will become apparent to those having ordinary skill in the art upon examination of the following, or may be learned from the practice of the invention. The advantages and features of the invention may be realized and obtained as particularly pointed out in the appended claims.

[00008] According to the present invention, the foregoing and other advantages are achieved in part by a method of forming a fin field effect transistor that includes forming a mesa on a silicon-on-insulator wafer, forming a dummy gate with a first material in a first pattern over the mesa, and forming a first dielectric layer around the dummy gate. The method further includes removing the first material to create a trench shaped in the first pattern, forming a mask over a portion of the trench and the mesa, and etching a portion of the mesa that is exposed within the trench to form a fin. The method also includes forming a gate dielectric layer over the fin, forming a gate within the trench over the gate dielectric layer, and removing the first dielectric layer.

[00009] According to another aspect of the invention, a method of forming a tri-gate fin field effect transistor is provided. The method includes forming an oxide layer over a silicon-on-insulator wafer comprising a silicon layer, etching the silicon and oxide layers using a rectangular mask to form a mesa, and etching a portion of the mesa using a second mask to form a fin. The method further includes forming a gate dielectric layer over the fin and forming a tri-gate over the fin and the gate dielectric layer.

[00010] Other advantages and features of the present invention will become readily apparent to those skilled in this art from the following detailed description. The embodiments shown and described provide illustration of the best mode contemplated for carrying out the invention. The invention is capable of modifications in various obvious respects, all without departing from the invention. Accordingly, the drawings are to be regarded as illustrative in nature, and not as restrictive.

[00035] The following detailed description of the invention refers to the accompanying drawings. The same reference numbers in different drawings may identify the same or similar elements. Also, the following detailed description does not limit the invention. Instead, the scope of the invention is defined by the appended claims.

[00036] Consistent with the present invention, an exemplary process for forming a tri-gate FinFET is provided. The exemplary process forms a silicon mesa on a FDSOI wafer. A dummy gate that may include, for example, Si₃N₄, may be formed over the mesa and a layer of tetraethylorthosilicate (TEOS), or any other dielectric material, may then be formed over the mesa and the dummy gate. The dummy gate may then be removed, leaving a trench in the shape of the eventual gate and an active mask may be used to etch a fin channel in exposed portions of the mesa within the trench. A gate may then be formed in the trench over the fin to produce a tri-gate FinFET.

[00037] FIG. 1 illustrates a cross-section of a FDSOI wafer 100 formed in accordance with an exemplary embodiment of the present invention. FDSOI wafer 100, consistent with the present invention, may include a layer of semiconducting material 105 formed on a layer of oxide 110. The layer of semiconducting material 105 may include, for example, silicon, germanium or silicon-germanium, and may range, for example, from about 10 nm to about 50 nm in thickness. Layer 105 may be used to form a fin for a FinFET device as described in more detail below. Oxide layer 110 may be formed on a substrate comprising silicon, germanium or silicon-germanium (not shown). Oxide layer 110 may include, for example, silicon dioxide (SiO₂), though other oxide materials may be used. Another oxide layer 115 may be deposited or thermally grown on layer 105. Oxide layer 115 may include, for example, silicon dioxide (SiO₂), though other oxide materials may be used. Oxide layer 115 may range, for example, from about 10 to about 15 nm in thickness.

[00038] As shown in FIGS. 2A and 2B, a mesa 205 may be formed from layers 105 and 115. Mesa 205 may be formed, for example, using a rectangular active mask and conventional etching processes. For example, a conventional photoresist material may be patterned and etched to define a rectangular mask having dimensions ranging from about 200 nm to 1000 m in length and about 80 nm to about 250 nm in width. The areas not covered by the mask may then be etched, with the etching terminating on oxide layer 110. The photoresist material may then be removed. A layer 305 of material, such as, for example, Si₃N₄, may then be formed over mesa 205, as shown in FIG. 3. Layer 305 may be formed over mesa 205 using a conventional process, such as, for example, a conventional chemical-vapor deposition (CVD) process, and, then planarized using, for example, a conventional chemical-mechanical polishing (CMP) process. As further shown in FIGS. 4A and 4B, a dummy gate 405 may be formed from layer 305. Dummy gate 405 may be formed, for example, using conventional photolithographic and etching processes. The thickness t of the portion of dummy gate 405 located over mesa 205 may range, for example, from about 80 nm to about 120 nm. The portion of dummy gate 405 located over mesa 205, as shown in FIG. 4A, may be located over a channel region of a fin subsequently formed from mesa 205 (as further described below).

[00039] A layer 505 of tetraethylorthosilicate (TEOS), or any other dielectric material, may then be formed over mesa 205 and dummy gate 405. Layer 505 may then be polished back to expose the upper surface of dummy gate 405 using, for example, a CMP process, as illustrated in FIG. 5. As shown in FIG. 6, dummy gate 405 may be removed leaving a gate trench 605. Dummy gate 405 may be removed, for example, using a conventional etching process. As further shown in FIGS. 7A and 7B, an active mask 705 may be formed over mesa 205. Mask 705 may be formed using a conventional photoresist material and its length may extend about 100 nm beyond mesa 205 on each end and its width may range from about 15 nm to about 30 nm after photoresist trimming. Mask 705 may be used to remove portions of mesa 205 exposed in gate trench 605 to create fin 710. For example, the portions of mesa 205 not covered by mask 705 may be etched to form fin 710 that may comprise a channel region for the final tri-gate FinFET. Fin 710 may have a width, w, ranging, for example, from about 15 nm to about 30 nm. Mask 705 may then be removed.

[00040] As shown in FIG. 8A, a layer 805 of sacrificial oxide may be formed over fin 710. Sacrificial oxide layer 805 may be formed on fin 710 using any conventional process. In some exemplary embodiments, for example, sacrificial oxide layer 805 may be thermally grown on fin 710. As further shown in FIG. 8B, sacrificial oxide layer 805 may be removed using a conventional process, such as, for example, a conventional etching process, to remove defects from sidewalls of fin 710.

[00041] A gate dielectric 905 may then be formed on surfaces of fin 710 and exposed sidewalls of mesa 205 within gate trench 605, as shown in FIGS. 9A and 9B. Gate dielectric 905 may be thermally grown or deposited using conventional processes. Gate dielectric 905 may include SiO, SiO₂, SiN, SiON, HF₀₂, ZrO₂, Al₂O₃, HfSiO(x), HfSiO(x)N(1-x) an oxide/nitride stack, or other high-K dielectric materials. A gate 1005 may then be formed in gate trench 605, as shown in FIGS. 10A, 10B and 10C. Gate 1005 may include, for example, polysilicon or a metal, that may be deposited in gate trench 605 and then polished back to the upper surface of TEOS layer 505 using, for example, a conventional CMP process.

[00042] TEOS layer 505 and exposed portions of gate dielectric 905 may then be removed, stopping on mesa 205, to expose oxide layer 110, as shown in FIGS. 11A, 11B and 11C. TEOS layer 505 and exposed portions of gate dielectric 905 may be removed, for example, using conventional etching processes. Source 1105 and drain 1110 regions may then be formed from exposed portions of mesa 205 to complete the tri-gate FinFET formation process. As shown in FIG. 11A, the resulting gate 1005 is disposed on three sides of fin 710 (i.e., over the top surface of fin 710 and adjacent both side surfaces of fin 710), thus, producing a tri-gate FinFET. The tri-gate FinFET, consistent with the invention, will have better short-channel control than double-gate and single-gate devices and will have higher drive current than a double-gate FinFET for the same area.

[00043] FIGS. 12-17 illustrate an exemplary process for forming a strained FinFET fin using an epitaxial regrowth process consistent with another embodiment of the invention. As shown in FIG. 12, the exemplary process may begin with the formation of a seed layer 1210 and an oxide layer 1205 on a buried oxide (BOX) layer 1215. Seed layer 1210 may include, for example, germanium (Ge), though other semiconducting materials may be used, and may be formed using, for example, conventional deposition processes. Seed layer 1210 may range, for example, from about 200 Å to about 1000 Å in thickness. Oxide layer 1205 may include, for example, SiO or SiO₂, though other oxide materials may be used, and may be formed, for example, from a conventional CVD process. Oxide layer 1205 may range, for example, from about 800 Å to about 1200 Å in thickness.

[00044] As shown in FIG. 13, a trench 1305 may be formed in oxide layer 1205 using, for example, conventional photolithographic and etching processes. Trench 1305 may range, for example, from about 500 Å to about 5000 Å in width. As further shown in FIG. 14, a strained semiconducting layer 1405 may then be grown in trench 1305 using, for example, selective epitaxy. Growth of layer 1405 may start from Si_xGe_1-x (at “A”) and progress to silicon (at “C”). After growth of layer 1405, excess growth may be polished off using, for example, a conventional CMP process, to form fin 1505, as shown in FIG. 15.

[00045] As further shown in FIG. 16, oxide layer 1205 may be removed using, for example, a conventional etching process to form a strained fin 1505. A spacer 1605 may then be formed over strained fin 1505. Spacer 1605 may include a dielectric material, such as, for example, SiO, SiO₂, SiN, SiON, HFO₂, ZrO₂, Al₂O₃, HfSiO(x), or HfSiO(x)N(1-x) and may range, for example, from about 10 Å to about 40 Å in thickness. Seed layer 1210 may then be etched away, as shown in FIG. 17. Tensile strain in fin 1505, that is due to confinement from seed layer 1210, improves carrier mobility of the FinFET.

[00046] FIGS. 18-23 illustrate an exemplary process for forming multiple channel “nano-wires” in a fin of a FinFET consistent with a further embodiment of the invention. The formation of multiple channel nano-wires, having extremely small vertical and horizontal dimensions, can be very difficult. The exemplary process described with respect to FIGS. 18-23 enables the formation of such structures.

[00047] As shown in FIG. 18, the exemplary process may begin with the formation of a seed layer 1810 and an oxide layer 1805 on a buried oxide (BOX) layer 1815. Seed layer 1810 may include, for example, silicon, though other semiconducting materials may be used, and may be formed using, for example, conventional deposition processes. Seed layer 1810 may range, for example, from about 200 Å to about 500 Å in thickness. Oxide layer 1805 may include, for example, SiO or SiO₂, though other oxide materials may be used, and may be formed, for example, from a conventional CVD process. Oxide layer 1805 may range, for example, from about 800 Å to about 1200 Å in thickness.

[00048] As shown in FIG. 19, a trench 1905 may be formed in oxide layer 1805 using, for example, conventional photolithographic and etching processes. Trench 1905 may range, for example, from about 500 Å to about 2500 Å in width. Alternating layers of different semiconducting materials may then be grown in trench 1905 using, for example, selective epitaxy, as shown in FIG. 20. Two different semiconducting materials 2005 and 2010 may be grown, for example, in trench 1905. Semiconducting material 2005 may include silicon, for example, and semiconducting material 2010 may include germanium, for example. One skilled in the art will recognize that other semiconducting materials may be used for the different semiconducting materials 2005 and 2010. As further shown in FIGS. 21A and 21B, the excess material of the topmost layer of semiconducting material 2010 may be polished off using, for example, a conventional CMP process, thereby forming a planar top surface.

[00049] Oxide layer 1805 may be removed using, for example, a conventional etching process to form a fin 2205, as shown in FIGS. 22A and 22B. As further shown in FIGS. 23A, 23B and 23C, material 2005 may be removed using, for example, a conventional wet etching process, to create nano-wires 2305. Etching of material 2005 in fin 2205 may also result in undercutting 2310 of the FinFET source 2315 and drain 2320 regions, as shown in FIGS. 23B and 23C. Each of nano-wires 2305 represents an extremely small channel that can independently contact the eventually formed FinFET gate (not shown).

[00050] In the previous descriptions, numerous specific details are set forth, such as specific materials, structures, chemicals, processes, etc., in order to provide a thorough understanding of the present invention. However, the present invention can be practiced without resorting to the details specifically set forth herein. In other instances, well known processing structures have not been described in detail, in order not to unnecessarily obscure the thrust of the present invention. In practicing the present invention, conventional photolithographic and etching techniques may be employed, and hence, the details of such techniques have not been set forth herein in detail.

[00051] Only the preferred embodiments of the invention and a few examples of its versatility are shown and described in the present disclosure. It is to be understood that the invention is capable of use in various other combinations and environments and is capable of modifications within the scope of the inventive concept as expressed herein.