108A and 108B in the X direction.

Each of the dummy gate structures 108A and 108B includes a dummy gate dielectric layer 110 and a dummy gate electrode layer 112 formed over the dummy gate dielectric layer 110, as shown in FIGS. 2B-1 and 2B-2, in accordance with some embodiments. In some embodiments, the dummy gate dielectric layer 110 is made of one or more dielectric materials, such as silicon oxide (SiO), silicon nitride (SiN), silicon oxynitride (SiON), HfO₂, HfZrO, HfSiO, HfMO, HfAlO, and/or a combination thereof. In some embodiments, the dielectric material is formed using ALD, CVD, thermal oxidation, physical vapor deposition (PVD), another suitable technique, and/or a combination thereof.

In some embodiments, the dummy gate electrode layer 112 is made of semiconductor material such as polysilicon, poly-silicon germanium. In some embodiments, the dummy gate electrode layer 112 is made of a conductive material such as metallic nitrides, metallic silicides, metals, and/or a combination thereof. In some embodiments, the material for the dummy gate electrode layer 112 is formed using CVD, another suitable technique, and/or a combination thereof.

In some embodiments, the formation of the dummy gate structures 108A and 108B includes globally and conformally depositing a dielectric material for the dummy gate dielectric layer 110 over the semiconductor structure, depositing a material for the dummy gate electrode layer 112 over the dielectric material, planarizing the material for the dummy gate electrode layer 112, and patterning the dielectric material and the material for the dummy gate electrode layer 112 into the dummy gate structures 108A and 108B.

The patterning process includes forming a patterned hard mask layer (not shown) over the material for the dummy gate electrode layer 112 to overlap the channel regions of the fin structures 104A and 104B, in accordance with some embodiments. The material for the dummy gate electrode layer 112 and the dielectric material, uncovered by the patterned hard mask layer, is etched away until the source/drain regions of the fin structures 104A and 104B are exposed, in accordance with some embodiments.

Gate spacer layers 114 are formed along the sidewalls of the dummy gate structures 108A and 108B, as shown in FIG. 2B, in accordance with some embodiments. The gate spacer layers 114 are used to offset the subsequently formed source/drain features and separate the source/drain features from the gate structure, in accordance with some embodiments. In some embodiments, the gate spacer layers 114 are made of dielectric material, such as silicon oxide (SiO₂), silicon nitride (SiN), silicon carbide (SiC), silicon oxynitride (SiON), silicon carbon nitride (SiCN), silicon oxycarbonitride (SiOCN), and/or oxygen-doped silicon carbonitride (Si(O)CN).

In some embodiments, the formation of the gate spacer layers 114 includes globally and conformally depositing a dielectric material for the gate spacer layers 114 over the semiconductor structure, followed by an anisotropic etching process. In some embodiments, the etching process is performed without an additional photolithography process. The portions of the dielectric material that remain on the sidewalls of the dummy gate structures 108A and 108B serve as the gate spacer layers 114, in accordance with some embodiments.

Source/drain features 116 are formed over the source/drain regions of the fin structures 104A and 104B, as shown in FIGS. 2B-3 and 2B-4, in accordance with some embodiments. The formation of the source/drain features 116 includes recessing the source/drain regions of the fin structures 104A and 104B using the dummy gate structures 108A and 108B and the gate spacer layers 114 as masks to form source/drain recesses on opposite sides of the dummy gate structures 108A and 108B, in accordance with some embodiments. The source/drain recesses extend into the isolation structure 106, in accordance with some embodiments.

Afterward, the source/drain features 116 are grown on the exposed surfaces of the fin structures 104A and 104B in the source/drain recesses using an epitaxial growth process, in accordance with some embodiments. The epitaxial growth process may be molecular beam epitaxy (MBE), metal organic chemical vapor deposition (MOCVD), or vapor phase epitaxy (VPE), or another suitable technique. In some embodiments, the source/drain features 116 are made of any suitable material for an n-type semiconductor device and a p-type semiconductor device, such as Ge, Si, GaAs, AlGaAs, SiGe, GaAsP, SiP, SiC, SiCP, or a combination thereof. In some embodiments, the source/drain features 116 are in-situ doped during the epitaxial growth process.

For example, the n-type source/drain features for n-type semiconductor devices are doped with the n-type dopant during the epitaxial growth process. For example, the n-type dopant may be phosphorous (P) or arsenic (As). For example, the n-type source/drain features may be the epitaxially grown S1 doped with phosphorous to form silicon:phosphor (Si:P) source/drain features and/or arsenic to form silicon:arsenic (Si:As) source/drain feature. For example, the p-type source/drain features for p-type semiconductor devices are doped with the p-type dopant during the epitaxial growth process. For example, the p-type dopant may be boron (B) or BF₂. For example, the p-type source/drain features may be the epitaxially grown SiGe doped with boron (B) to form silicon germanium:boron (SiGe:B) source/drain feature.

A contact etching stop layer 118 is formed over the semiconductor structure 100 to covers the source/drain features 116, as shown in FIGS. 2B-3 and 2B-4, in accordance with some embodiments. In some embodiments, the contact etching stop layer 118 extends along and cover the top surface of the isolation structure 106 and the sidewalls of the gate spacer layers 114. In some embodiments, the contact etching stop layer 118 is made of dielectric material, such as silicon oxide (SiO₂), silicon nitride (SiN), silicon oxynitride (SiON), silicon carbide (SiC), oxygen-doped silicon carbide (SiC:O), oxygen-doped silicon carbonitride (Si(O)CN), or a combination thereof. In some embodiments, a dielectric material for the contact etching stop layer 118 is globally and conformally deposited using CVD (such as LPCVD, PECVD, HDP-CVD, or HARP), ALD, another suitable method, or a combination thereof.

An interlayer dielectric layer 120 is formed over the contact etching stop layer 118, as shown in FIGS. 2B-3 and 2B-4, in accordance with some embodiments. In some embodiments, the interlayer dielectric layer 120 is made of dielectric material, such as un-doped silicate glass (USG), or doped silicon oxide such as borophosphosilicate glass (BPSG), fluoride-doped silicate glass (FSG), phosphosilicate glass (PSG), borosilicate glass (BSG), and/or another suitable dielectric material. In some embodiments, the interlayer dielectric layer 120 and the contact etching stop layer 118 are made of different materials and have a great difference in etching selectivity.

In some embodiments, the dielectric material for the interlayer dielectric layer 120 is deposited using such as CVD (such as HDP-CVD, PECVD, HARP or FCVD), another suitable technique, and/or a combination thereof. The dielectric materials for the contact etching stop layer 118 and the interlayer dielectric layer 120 above the upper surface of the dummy gate structures 108A and 108B are removed using such as CMP, in accordance with some embodiments.

FIGS. 2C through 2C-6 illustrate a semiconductor structure 100 after the formation of final gate stacks 122A and final gate stacks 122B, in accordance with some embodiments. FIGS. 2C-1, 2C-2, 2C-3, 2C-4, 2C-5 and 2C-6 are cross-sectional views of the semiconductor structure 100 taken along line Y1-Y1, line Y1′-Y1′, line Y2-Y2, line Y2′-Y2′, line X1-X1 and line X1′-X1′ of FIG. 2C, respectively.

The dummy gate structures 108A and 108B are removed using one or more etching processes to form gate trenches, in accordance with some embodiments. The gate trenches expose the channel regions of the fin structures 104A and 104B, in accordance with some embodiments. In some embodiments, the gate trenches also expose the inner sidewalls of the gate spacer layers 114 facing the channel region, in accordance with some embodiments. In some embodiments, the etching process includes one or more etching processes. For example, when the dummy gate electrode layer 112 is made of polysilicon, a wet etchant such as a tetramethylammonium hydroxide (TMAH) solution may be used to selectively remove the dummy gate electrode layer 112. For example, the dummy gate dielectric layer 110 may be thereafter removed using a plasma dry etching, a dry chemical etching, and/or a wet etching.

Final gate stacks 122A and 122B are formed in the gate trenches, as shown in FIGS. 2C, 2C-1, 2C-2, 2C-5 and 2C-6, in accordance with some embodiments. The final gate stacks 122A are formed in the logic region 50A and extend across and surround the channel regions of the fin structures 104A, in accordance with some embodiments. The final gate stacks 122B are formed in the memory cell array region 50B and extend across and surround the channel regions of the fin structures 104B, in accordance with some embodiments. In some embodiments, the final gate stacks 122A and 122B extend in the Y direction. The final gate stacks 122A and 122B have longitudinal axes parallel to the Y direction, in accordance with some embodiments. That is, the dimensions (lengths) of the final gate stacks 122A and 122B in the Y direction are greater than the dimensions (widths) of the final gate stacks 122A and 122B in the X direction.

In some embodiments, each of the final gate stacks 122A and 122B includes a gate dielectric layer 124 and a gate electrode layer 126 formed over the gate dielectric layer 124, as shown in FIGS. 2C-1, 2C-2, 2C-5 and 2C-6, in accordance with some embodiments. The gate dielectric layer 124 is formed to partially fill the gate trenches, in accordance with some embodiments. In some embodiments, the gate dielectric layer 124 may include an interfacial layer and a high-k dielectric layer formed over the interfacial layer. The interfacial layer may be made of a chemically formed silicon oxide by one or more cleaning processes such as including ozone (O₃), ammonia hydroxide-hydrogen peroxide-water mixture, and/or hydrochloric acid-hydrogen peroxide-water mixture.

In some embodiments, the high-k dielectric layer is made of dielectric material with high dielectric constant (k value), for example, greater than 3.9. In some embodiments, the high-k dielectric layer includes hafnium oxide (HfO₂), TiO₂, HfZrO, Ta₂O₃, HfSiO₄, ZrO₂, ZrSiO₂, LaO, AlO, ZrO, TiO, Ta₂O₅, Y₂O₃, SrTiO₃(STO), BaTiO₃(BTO), BaZrO, HfZrO, HfLaO, HfSiO, LaSiO, AlSiO, HfTaO, HfTiO, (Ba,Sr)TiO₃(BST), Al₂O₃, Si₃N₄, oxynitrides (SiON), a combination thereof, or another suitable material. The high-k dielectric layer may be deposited using ALD, PVD, CVD, and/or another suitable technique.

The metal gate electrode layer 126 is formed to fill remainders of the gate trenches, in accordance with some embodiments. In some embodiments, the metal gate electrode layer 126 is made of more than one conductive material, such as a metal, metal alloy, conductive metal oxide and/or metal nitride, another suitable conductive material, and/or a combination thereof. For example, the metal gate electrode layer 126 may be made of Ti, Ag, Al, TiAlN, TaC, TaCN, TaSiN, Mn, Zr, TiN, TaN, Ru, Mo, WN, Cu, W, Re, Ir, Co, Ni, another suitable conductive material, or multilayers thereof.

The metal gate electrode layer 126 may be a multi-layer structure with various combinations of a diffusion barrier layer, work function layers with a selected work function to enhance the device performance (e.g., threshold voltage) for n-channel FETs or p-channel FETs, a capping layer to prevent oxidation of work function layers, a glue layer to adhere work function layers to a next layer, and a metal fill layer to reduce the total resistance of gate stacks, and/or another suitable layer. The metal gate electrode layer 126 may be formed using ALD, PVD, CVD, e-beam evaporation, or another suitable technique. The metal gate electrode layer 126 may be formed separately for n-channel FinFETs and p-channel FinFETs, which may use different work function materials.

A planarization process such as CMP may be performed on the semiconductor structure to remove the materials of the gate dielectric layer 124 and the metal gate electrode layer 126 formed above the top surface of the interlayer dielectric layer 120, in accordance with some embodiments. After the planarization process, the top surfaces of the metal gate electrode layer 126, the gate spacer layers 114, the contact etching stop layer 118 and the interlayer dielectric layer 120 are substantially coplanar, in accordance with some embodiments.

In the logic region 50A, portions of the final gate stacks 122A surrounding the fin structures 104A combine with the neighboring source/drain features 116 to form FinFETs (e.g., p-channel FinFETs or n-channel FinFETs), in accordance with some embodiments. The final gate stacks 122A engage the channel so that current can flow between the source/drain features 116 during operation. In the memory cell array region 50B, portions of the final gate stacks 122B surrounding the fin structures 104B combine with the neighboring source/drain features 116 to form FinFETs (e.g., p-channel FinFETs or n-channel FinFETs), in accordance with some embodiments. The final gate stacks 122B engage the channel so that current can flow between the source/drain features 116 during operation.

FIGS. 2D through 2D-6 illustrate a semiconductor structure 100 after the formation of gate cutting trenches 128A and gate cutting trenches 128B, in accordance with some embodiments. FIGS. 2D-1, 2D-2, 2D-3, 2D-4, 2D-5 and 2D-6 are cross-sectional views of the semiconductor structure taken along line Y1-Y1, line Y1′-Y1′, line Y2-Y2, line Y2′-Y2′, line X1-X1 and line X1′-X1′ of FIG. 2D, respectively.

A cutting process is performed on the final gate stacks 122A and 122B, as shown in FIGS. 2D, 2D-1 and 2D-2, in accordance with some embodiments. The cutting process includes patterning the semiconductor structure 100 to form gate cutting trenches 128A and 128B, as shown in FIGS. 2D, 2D-1, 2D-2, 2D-3, 2D-4, 2D-5 and 2D-6, in accordance with some embodiments. The final gate stacks 122A are cut into several segments by the gate cutting trenches 128A, in accordance with some embodiments. The final gate stacks 122B are cut into several segments by the gate cutting trenches 128B, in accordance with some embodiments. The gate cutting trenches may be also referred to as cut metal gate (CMG) pattern.

The gate cutting trenches 128A are formed in the spacing S1 (FIG. 2A) in the logic region 50A, and the gate cutting trenches 128B are formed in the spacing S2 (FIG. 2B) in the memory cell array region 50B, in accordance with some embodiments. In some embodiments, the gate cutting trenches 128A and 128B extend in the X direction. The gate cutting trenches 128A and 128B have longitudinal axes parallel to the X direction, in accordance with some embodiments. That is, the dimensions (lengths) of the gate cutting trenches 128A and 128B in the X direction are greater than the dimensions (widths) of the gate cutting trenches 128A and 128B in the Y direction.

The cutting process includes forming a patterned mask layer over the semiconductor structure 100 using a photolithography process. The patterned mask layer has trench patterns corresponding to the gate cutting trenches 128A and 128B, in accordance with some embodiments. One or more etching process are then performed using the patterned mask layer to remove portions of the final gate stacks 122A and 122B, the gate spacer layers 114, the contact etching stop layer 118, the interlayer dielectric layer 120, the isolation structure 106 and the substrate 102 exposed from the trench patterns, in accordance with some embodiments. The one or more etching processes may include dry etching such as reactive ion etch (RIE), neutral beam etch (NBE), inductive coupled plasma (ICP) etch, capacitively coupled plasma (CCP) etch, another suitable method, or a combination thereof.

During the one or more etching processes, the source/drain features 116 are kept at a distance from the gate cutting trenches 128A and 128B to avoid damage to the source/drain features 116, as shown in FIGS. 2D-3 and 2D-4, in accordance with some embodiments. Because the spacing S1 (FIGS. 2A and 2A-1) where the gate cutting trenches 128A are formed is greater than the spacing S2 (FIGS. 2A and 2A-2) where the gate cutting trenches 128B, the trench patterns for the gate cutting trenches 128A are defined to be wider (in the Y direction) than the gate cutting trench patterns for trenches 128B without causing the damage to the source/drain features 116 in the logic region 50A, in accordance with some embodiments.

In some embodiments, the gate cutting trenches 128A have a dimension D1 in the Y direction measured at the top of the metal gate electrode layer 126. In some embodiments, the dimension D1 is in a range from about 15 nm to about 90 nm. In some embodiments, the gate cutting trenches 128B have a dimension D3 in the Y direction measured at the top of the metal gate electrode layer 126. In some embodiments, the dimension D3 is in a range from about 7.5 nm to about 25 nm. In some embodiments, the dimension D1 is greater than the dimension D3. In some embodiments, the ratio (D3/D1) of the dimension D3 to the dimension D1 is in a range from about 0.3 to about 0.8, e.g., about 0.4 to about 0.6.

In some embodiments, the gate cutting trenches 128A have a depth D2 in the Z direction. In some embodiments, the gate cutting trenches 128B have a depth D4 in the Z direction. Because the dimension D1 of the gate cutting trenches 128A is greater than the dimension D3 of the gate cutting trenches 128B, more vertical etching occurs in the logic region 50A than in the memory cell array region 50B. As a result, in some embodiments, the depth D2 is greater than the depth D4. In some embodiments, the ratio (D4/D2) of the depth D4 to the depth D2 is in a range from about 0.6 to about 0.95. When the one or more etching process is complete, the bottom surfaces of the gate cutting trenches 128A stop inside the substrate 102, while the bottom surfaces of the gate cutting trenches 128B stop inside the isolation structure 106, in accordance with some embodiments.

FIGS. 2E through 2E-6 illustrate a semiconductor structure 100 after the formation of a first dielectric material 130, in accordance with some embodiments. FIGS. 2E-1, 2E-2, 2E-3, 2E-4, 2E-5 and 2E-6 are cross-sectional views of the semiconductor structure taken along line Y1-Y1, line Y1′-Y1′, line Y2-Y2, line Y2′-Y2′, line X1-X1 and line X1′-X1′ of FIG. 2E, respectively.

The first dielectric material 130 is formed over the semiconductor structure 100 to fill the gate cutting trenches 128A and 128B, as shown in FIGS. 2E, 2E-1, 2E-2, 2E-3, 2E-4, 2E-5 and 2E-6, in accordance with some embodiments. The gate cutting trenches 128B are overfilled with the first dielectric material 130, while the gate cutting trenches 128A are partially filled with the first dielectric material 130, in accordance with some embodiments. The portions of the first dielectric material 130 in the gate cutting trenches 128B serve as gate cutting structures 134B, in accordance with some embodiments.

In some embodiments, the first dielectric material 130 is silicon nitride (SiN), silicon oxynitride (SiON), silicon carbon nitride (SiCN), silicon oxide (SiO₂), silicon oxycarbonitride (SiOCN), oxygen-doped silicon carbide (SiC:O), oxygen-doped silicon carbonitride (Si(O)CN), or a combination thereof. In some embodiments, the first dielectric material 130 is deposited using ALD, CVD such as LPCVD, PECVD, HDP-CVD, or HARP, another suitable technique, or a combination thereof. In some embodiments, the first dielectric material 130 is deposited with a thickness T1 along the sidewalls of the fin structure 104A (measured in the Y direction), and over the top surface of the metal gate electrode layer 126, in accordance with some embodiments. In some embodiments, the thickness T1 is in a range from 3.8 nm to about 22.5 nm.

During the deposition process, in the trench 128B, a first portion of the dielectric material 130 formed on a sidewall of the trench 128B and a second portion of the dielectric material 130 formed on another sidewall of the trench 128B gradually approach each other, in accordance with some embodiments. The deposition process is performed until the first portion and the second portion of the dielectric material 130 entirely merged, in accordance with some embodiments. In the trench 128A, when the deposition process is complete, a third portion of the dielectric material 130 formed on a sidewall of the trench 128A has not yet merged with a fourth portion of the dielectric material 130 formed on another sidewall of the trench 128A, because the dimension D1 (FIG. 2D-1) of the gate cutting trenches 128A is greater than the dimension D3 (FIG. 2D-2) of the gate cutting trenches 128B, in accordance with some embodiments.

FIGS. 2F through 2F-6 illustrate a semiconductor structure 100 after the formation of a second dielectric material 132, in accordance with some embodiments. FIGS. 2F-1, 2F-2, 2F-3, 2F-4, 2F-5 and 2F-6 are cross-sectional views of the semiconductor structure taken along line Y1-Y1, line Y1′-Y1′, line Y2-Y2, line Y2′-Y2′, line X1-X1 and line X1′-X1′ of FIG. 2F, respectively.

The second dielectric material 132 is formed over the semiconductor structure 100 to overfill the remainders of the gate cutting trenches 128A, as shown in FIGS. 2F, 2F-1, 2F-3 and 2F-5, in accordance with some embodiments. The portions of the first dielectric material 130 and the second dielectric material 132 that are in the gate cutting trenches 128A serve as gate cutting structures 134A, in accordance with some embodiments. In some embodiments, the second dielectric material 132 is silicon oxide (SiO₂), silicon oxycarbonitride (SiOCN), oxygen-doped silicon carbide (SiC:O), oxygen-doped silicon carbonitride (Si(O)CN), silicon carbon nitride (SiCN), silicon oxynitride (SiON), silicon nitride (SiN), or a combination thereof.

In some embodiments, the second dielectric material 132 is deposited using ALD, CVD such as LPCVD, PECVD, HDP-CVD, HARP, or FCVD, another suitable technique, or a combination thereof. In some embodiments, the second dielectric material 132 and the first dielectric material 130 are made of different materials and have a great difference in etching selectivity. In some embodiments, the dielectric constant of the second dielectric material 132 is less than the dielectric constant of the first dielectric material 130. For example, the first dielectric material 130 is SiN and the second dielectric material 132 is SiO, SiON, SiCN or SiOCN. In some embodiments, the second dielectric material 132 has a thickness T2 in the Y direction, in accordance with some embodiments. In some embodiments, the thickness T2 is in a range from 7.5 nm to about 45 nm.

A planarization process is then performed on the semiconductor structure 100 to remove the second dielectric material 132 and the first dielectric material 130 formed above the top surface of the metal gate electrode layer 126, in accordance with some embodiments. The planarization may be CMP, etching back process, or a combination thereof. After the planarization process, the top surfaces of the second dielectric material 132, the first dielectric material 130, the metal gate electrode layer 126, the gate spacer layers 114, the contact etching stop layer 118 and the interlayer dielectric layer 120 are substantially coplanar, in accordance with some embodiments.

The gate cutting structures 134A are formed in the spacing S1 (FIG. 2A) in the logic region 50A, in accordance with some embodiments. The gate cutting structures 134B are formed in the spacing S2 (FIG. 2B) in the memory cell array region 50B, in accordance with some embodiments. In some embodiments, the gate cutting structures 134A and 134B extend in the X direction. The gate cutting structures 134A and 134B have longitudinal axes parallel to the X direction, in accordance with some embodiments. That is, the dimensions (lengths) of the gate cutting structures 134A and 134B in the X direction are greater than the dimensions (widths) of the gate cutting structures 134A and 134B in the Y direction.

The segments of the final gate stacks 122A are physically and electrically insulated from one another by the gate cutting structures 134A, and the segments of the final gate stacks 122B are physically and electrically insulated from one another by the gate cutting structures 134B, in accordance with some embodiments.

In some embodiments, the gate cutting structures 134A are bi-layer structures in which the second dielectric material 132 is nested within the first dielectric material 130. In some embodiments, the gate cutting structures 134B are single-layer structures including the first dielectric material 130 and are free of the second dielectric material 132. In some embodiments, the overall dielectric constant of the gate cutting structures 134A is less than the dielectric constant of the gate cutting structures 134B.

In some embodiments, the gate cutting structures 134A have a dimension D5 in the Y direction measured at the top of the metal gate electrode layer 126. In some embodiments, the dimension D5 is in a range from about 15 nm to about 90 nm. In some embodiments, the gate cutting structures 134B have a dimension D7 in the Y direction measured at the top of the metal gate electrode layer 126. In some embodiments, the dimension D7 is in a range from about 7.5 nm to about 45 nm. In some embodiments, the dimension D5 is greater than the dimension D7. In some embodiments, the ratio (D7/D5) of the dimension D7 to the dimension D5 is in a range from about 0.3 to about 0.8, e.g., from about 0.4 to about 0.6.

In some embodiments, the gate cutting structures 134A have a thickness D6 in the Z direction. In some embodiments, the gate cutting structures 134B have a thickness D8 in the Z direction. In some embodiments, the thickness D6 is greater than the thickness D8. In some embodiments, the ratio (D8/D6) of the thickness D8 to the thickness D6 is in a range from about 0.6 to about 0.95. In some embodiments, the bottom surface of the gate cutting structures 134A is located at a lower position than the bottom surface of the isolation structure 106. In some embodiments, the bottom surface of the gate cutting structures 134B is located at a position between the top surface and the bottom surface of the isolation structure 106.

In some embodiments, the first distance, which is measured from the top of the fin structure 104A to the bottom of the gate cutting structures 134A in the Z direction, is greater than about 140 nm. In some embodiments, the second distance, which is measured from the top of the fin structure 104B to the bottom of the gate cutting structures 134B in the Z direction, is greater than 120 nm. The first distance is greater than the second distance, in accordance with some embodiments.

FIGS. 2G through 2G-6 illustrate a semiconductor structure 100 after the formation of contact plugs 136A and contact plugs 136B, in accordance with some embodiments. FIGS. 2G-1, 2G-2, 2G-3, 2G-4, 2G-5 and 2G-6 are cross-sectional views of the semiconductor structure taken along line Y1-Y1, line Y1′-Y1′, line Y2-Y2, line Y2′-Y2′, line X1-X1 and line X1′-X1′ of FIG. 2G, respectively.

Contact plugs 136A and 136B are formed in and/or through the interlayer dielectric layer 120 and the contact etching stop layer 118 and land on and the source/drain features 116, as shown in FIGS. 2G, 2G-3, 2G-4, 2G-5 and 2G-6, in accordance with some embodiments. The contact plugs 136A are formed in the logic region 50A and are electrically connected to the source/drain features 116, in accordance with some embodiments. The contact plugs 136B are formed in the memory cell array region 50B and are electrically connected to the source/drain features 116, in accordance with some embodiments.

In some embodiments, some (labeled with 136A 1) of the contact plugs 136A are formed over and in direct contact with the gate cutting structures 134A. In some embodiments, the contact plug 136A 1 continuously extends over the source/drain features 116 and the gate cutting structure 134A, as shown in FIGS. 2G, 2G-3 and 2G-5. The contact plug 136A1 is also formed in and/or through the gate cutting structure 134A, in accordance with some embodiments. In some embodiments, some (labeled with 136B1) of the contact plugs 136B are formed over and in direct contact with the gate cutting structures 134B. In some embodiments, the contact plug 136B1 continuously extends over the source/drain features 116 and the gate cutting structure 134B, as shown in FIGS. 2G, 2G-4 and 2G-6. The contact plug 136B1 is also formed in and/or through the gate cutting structure 134B, in accordance with some embodiments.

In some embodiments, the formation of the contact plugs 136A and 136B includes patterning the interlayer dielectric layer 120, the contact etching stop layer 118 and the gate cutting structures 134A and 134B to form contact openings (where the contact plugs 136A and 136B are to be formed) using photolithography and etching processes until the source/drain features 116 are exposed. The etching processes may include dry etching such as RIE, NBE, ICP etch, CCP etch, another suitable method, or a combination thereof.

In the etching process for forming the contact openings, the second dielectric material 132 has a better etching resistance than the first dielectric material 130, and thus the loss of the gate cutting structures 134A and 134B caused by the lateral etching may be reduced, in accordance with some embodiments. If the materials of the gate cut structures lose too much, it may cause an increase in parasitic capacitance of the semiconductor devices.

A silicide layer (such as WSi, NiSi, TiSi and/or CoSi) may be formed on the exposed source/drain features 116, and then one or more conductive materials for the contact plugs 136A and 136B are deposited to fill the contact openings, in accordance with some embodiments. In some embodiments, one or more conductive materials are deposited using CVD, PVD, c-beam evaporation, ALD, electroplating (ECP), electroless deposition (ELD), another suitable method, or a combination thereof to overfill the contact openings. Afterward, the one or more conductive materials over the interlayer dielectric layer 120 are planarized using, for example, CMP. After the planarization process, the top surfaces of the contact plugs 136A and 136B, the second dielectric material 132, the first dielectric material 130, the metal gate electrode layer 126, the gate spacer layers 114, the contact etching stop layer 118 and the interlayer dielectric layer 120 are substantially coplanar.

The contact plugs 136A and 136B may have a multilayer structure including, for example, liner layers, glue layers, barrier layers, seed layers, metal bulk layers, another suitable layer, or a combination thereof. For example, a barrier layer (not shown) may optionally be deposited along the sidewalls and the bottom surfaces of the contact openings. The barrier layer is used to prevent the metal from the subsequently formed metal material from diffusing into the dielectric material (e.g., the interlayer dielectric layer 120 and the contact etching stop layer 118). The barrier layer may be made of tantalum (Ta), tantalum nitride (TaN), titanium (Ti), titanium nitride (TiN), cobalt tungsten (CoW), another suitable material, or a combination thereof. If the subsequently formed metal material does not easily diffuse into the dielectric material, the barrier layer may be omitted.

A glue layer (not shown) may optionally be deposited along the sidewalls and the bottom surfaces of the contact openings, and on the barrier layer (if formed). The glue layer is used to improve adhesion between the subsequently formed metal material and the dielectric material (e.g., the interlayer dielectric layer 120 and the contact etching stop layer 118). The glue layer may be made of tantalum nitride (TaN), titanium (Ti), titanium nitride (TiN), another suitable material, or a combination thereof.

A metal bulk layer is then deposited on the glue layer (if formed) to fill the remainder of the contact openings. In some embodiments, the metal bulk layer is formed using a selective deposition technique such as cyclic CVD process or ELD process, and it is not necessary to form a glue layer in the contact openings before depositing the metal bulk material. In some embodiments, the metal bulk layer is made of one or more conductive materials with low resistance and good gap-fill ability, for example, cobalt (Co), nickel (Ni), tungsten (W), titanium (Ti), tantalum (Ta), copper (Cu), rhodium (Rh), iridium (Ir), platinum (Pt), aluminum (Al), ruthenium (Ru), molybdenum (Mo), another suitable metal material, or a combination thereof.

In some embodiments, because the gate cutting structure 134A includes the second dielectric material 132 having lower etching resistance, the interface 134I1 between the contact plug 136A1 and the gate cutting structure 134A is located at a lower position than the interface 134I2 between the contact plug 136B1 and the gate cutting structure 134B.

In the manufacturing processes of the semiconductor structure including the logic devices and the memory devices (e.g., SRAM), the logic devices may focus more on device performance, e.g., speed, while the memory devices may focus more on device density. As the scale of the semiconductor structure continues to shrink, and the spacing between the contact plug and the metal gate electrode becomes smaller and smaller, one of the design challenges of forming the logic devices is to reduce the parasitic capacitance (e.g., out-fringing capacitance (Cof), contact-to-gate capacitance (Cco), etc.) to enhance the device performance, e.g., speed.

In accordance with the embodiments, the gate cutting structures 134A in the logic region 50A have a relatively great width (i.e., the dimension D5), and thus the lateral overlap area between final gate stacks 122A and the contact plugs 136A may be reduced. Therefore, the parasitic capacitance (e.g., Cof and/or Cco) of the logic devices may be reduced, thereby enhancing the performance of the logic devices.

If the ratio (D7/D5) is too small (e.g., the dimension D5 of the gate cutting structure 134A is too great), the risk of the damage to the source/drain feature 116 in the logic region 50A may be increased. If the ratio (D7/D5) is too great (e.g., the dimension D5 of the gate cutting structure 134A is too small, and/or the dimension D7 of the gate cutting structure 134B is too great), the parasitic capacitance of the logic devices may not be reduced sufficiently and/or the risk of the damage to the source/drain feature 116 in the memory cell array region 50B may be increased.

In addition, the gate cutting structure 134A includes the low-k dielectric material 132, which may further reduce the parasitic capacitance, which may further enhance the performance of the logic devices. Furthermore, the gate cutting structures 134B in the memory cell array region 50B have a relatively small width (i.e., dimension D6), and thus the density of the memory devices may be enhanced.

It should be understood that the semiconductor structure 100 may undergo further CMOS processes to form various features over the semiconductor structure, such as a multilayer interconnect structure (e.g., contact plugs to final gate stacks, conductive vias, metal lines, inter metal dielectric layers, passivation layers, etc.). In some embodiments, the logic device formed in the logic region 50A may be electrically connected to the memory device formed in the memory cell array region 50B through the conductive features of the multilayer interconnect structure.

FIGS. 3-1 and 3-2 illustrate a modification of the semiconductor structure 100 of FIGS. 2G-3 and 2G-4, in accordance with some embodiments of the disclosure. The semiconductor structure 200 of FIGS. 3-1 and 3-2 is similar to the semiconductor structure 100 of FIGS. 2G-3 and 2G-4, except that the interface between the contact plug 136A1 and the gate cutting structure 134A.

In some embodiments, because the second dielectric material 132 has a better etching resistance than the first dielectric material 130, the interface 134I3 between the contact plug 136A1 and the first dielectric material 130 of the gate cutting structure 134A is located at a higher position than the interface 134I4 between the contact plug 136A1 and the second dielectric material 132 of the gate cutting structure 134A. In some embodiments, a portion of the contact plug 136A1 is embedded within the first dielectric material 130 of the gate cutting structure 134A. In some embodiments, the interface 134I3 may be is located at substantially the same position with the interface 134I2 between the contact plug 136B1 and the gate cutting structure 134B.

FIGS. 4A, 4B and 4C are plan views illustrating the formation of a semiconductor structure 300 at various intermediate stages, in accordance with some embodiments of the disclosure. The embodiments of FIGS. 4A-4C are similar to the embodiments of FIGS. 2A through 2G, except the formation of a fin cutting structure.

FIGS. 4A through 4A-2 illustrate a semiconductor structure 300 after the formation of a fin cutting trench 140, in accordance with some embodiments. FIGS. 4A-1 and 4A-2 are cross-sectional views of the semiconductor structure 300 taken along line Y1-Y1 and line X1-X1 of FIG. 4A, respectively.

Continuing from FIGS. 2F through 2F-6, a cutting process is performed on the final gate stack 122A and the fin structures 104A, as shown in FIGS. 4A, 4A-1 and 4A-2, in accordance with some embodiments. The cutting process includes patterning the semiconductor structure 300 to form fin cutting trench 140, in accordance with some embodiments. The fin structures 104A are cut into several segments by the fin cutting trench 140, in accordance with some embodiments. In some embodiments, the fin cutting trench 140 further partially cut through the gate cutting structures 134A. The fin cutting trench may be also referred to as cut metal on oxide definition edge (CMODE) pattern.

In some embodiments, the fin cutting trench 140 extends in the Y direction. That is, the fin cutting trench 140 has longitudinal axes parallel to the Y direction, in accordance with some embodiments. That is, the dimension (length) of the fin cutting trench 140 in the Y direction are greater than the dimension (width) of the fin cutting trench 140 in the X direction.

The cutting process includes forming a patterned mask layer over the semiconductor structure 300 using a photolithography process. The patterned mask layer has a trench pattern corresponding to the fin cutting trench 140, in accordance with some embodiments. An etching process is then performed using the patterned mask layer to remove portions of the gate cutting structures 134A, the final gate stack 122A, the fin structures 104A and the substrate 102 that are exposed from the trench pattern, in accordance with some embodiments. The etching processes may include dry etching such as RIE, NBE, ICP etch, CCP, another suitable method, or a combination thereof. In some embodiments, the bottom surface of the fin cutting trench