This application claims the benefit of Korean Patent Application No. 10-2005-0009258, filed Feb. 1, 2005, the disclosure of which is hereby incorporated herein by reference in its entirety.
1. Field of the Invention
The present invention relates to a method of fabricating a semiconductor device, and more particularly, to a method of fabricating a Metal Oxide Semiconductor (MOS) transistor having a fully silicided metal gate electrode.
2. Description of the Related Art
In order to continue making advances in electronic products employing semiconductor devices, improvements in integration density, operating speed, and power consumption are required. Discrete devices such as MOS transistors are widely employed as switching devices for semiconductor devices. To meet the requirement of high integration, gates, source and drain junctions, and interconnections of the transistor should be reduced in size as much as possible. In addition, connections between the transistors should also be reduced in size.
However, transistor size reduction has several associated difficulties. For example, electrical resistance of the gate electrode increases as its size is reduced. In this case, an electrical signal applied to the gate electrode is delayed by a Resistance-Capacitance (RC) delay time. In addition, a short channel effect occurs due to the reduction of the channel length.
In the conventional art of employing polysilicon for the gate electrode, reduction of the size of the gate electrode further causes problems such as polysilicon depletion and boron penetration. Here, polysilicon depletion occurs in a depletion region adjacent to a gate insulating layer, i.e., a lower region within the polysilicon gate electrode. The polysilicon depletion region acts as an additional capacitance connected in series to the capacitance of the gate insulating layer. Consequently, the polysilicon depletion region causes the electrical equivalent thickness of the gate insulating layer to increase, which means a decrease in an effective gate voltage. In the conventional art, employing a thick gate insulating layer has a negligible effect since the thickness of the polysilicon depletion region is very small compared to the effective thickness of the thick gate insulating layer. However, as thinner gate insulating layers are used, the decrease in the effective gate voltage due to the polysilicon depletion region causes serious problems.
There are several advantages when a metal is used for the gate of the transistor instead of polysilicon. For example, the metal has a very high conductivity and may prevent gate depletion and boron penetration. However, the metal gate causes degradation of the gate insulating layer from metal ions and has a fixed work function, which makes it difficult to adjust a threshold voltage Vth. For example, a semiconductor device such as a complementary MOS (CMOS) transistor has an NMOS transistor region and a PMOS transistor region within a single chip. Threshold voltages of the NMOS and PMOS transistors should be adjusted to be different from each other. Consequently, a metal gate employed for the NMOS transistor region should be different from that employed for the PMOS transistor region, which makes the process very complicated.
To implement a high-performance MOS transistor suitable for a highly integrated semiconductor device, research into self-aligned silicide (i.e., salicide) technology has been done. Salicide technology is process technology for forming a metal silicide layer on the gate electrode and the source and drain regions to reduce their electrical resistance. In this case, a metal gate may be formed when the gate electrode is fully transformed into a metal silicide. In addition, when the gate electrode is transformed into the metal silicide in an N-doped or a P-doped state, a work function required from the NMOS or the PMOS can be obtained.
Referring to
Referring to
Consequently, the gate electrode 19 becomes silicided downward from the top so that a metal gate electrode 27 is formed. While the metal gate electrode 27 is formed, the source and drain regions 23 are also silicided downward from the top so that source and drain silicide layers 29 are formed. In this case, when the source and drain silicide layers 29 are deeper than a junction depth of the source and drain regions 23, leakage current occurs. That is, the source and drain silicide layers 29 must be formed to be shallower than the junction depth of the source and drain regions 23. Consequently, the metal gate electrode 27 is formed only on an upper region of the gate electrode 19.
A method of forming a metal gate electrode for improving the above-described problems is disclosed in U.S. Pat. No. 6,599,831 B 1 entitled “Metal Gate Electrode Using Silicidation and Method of Formation Thereof” to Maszara, et al.
According to Maszara, et al., a gate electrode and a capping layer are sequentially stacked on a predetermined region of a semiconductor substrate. A gate dielectric layer is interposed between the gate electrode and the semiconductor substrate. The gate electrode is formed of doped polysilicon. Spacers are then formed to cover sidewalls of the gate dielectric layer, the gate electrode, and the capping layer. Source and drain regions are formed in active regions of the semiconductor substrate using the capping layer and the spacers as ion implantation masks. The capping layer is selectively etched to expose the gate electrode. Subsequently, a metal layer covering the gate electrode and the source and drain regions is formed, and a silicidation process is carried out.
However, to prevent the spacers from being damaged while the capping layer is etched, the capping layer should be formed of a material having a high etch selectivity with respect to the spacers, but even so, it is not easy to remove the capping layer. For example, when the capping layer is an oxide layer, a trench isolation layer to be simultaneously exposed may be damaged. Alternatively, when the capping layer is a nitride layer, a trench liner to be simultaneously exposed may be damaged.
In addition, when the capping layer is not completely removed, the gate electrode cannot be fully transformed into silicide.
Consequently, a technique of completely transforming the gate electrode into silicide while preventing formation of a deep silicide layer in the source and drain regions is required.
Embodiments of the invention provide a method of fabricating a MOS transistor capable of preventing a deep silicide layer from being formed in source and drain regions while a gate electrode is fully transformed to a silicide.
One embodiment of the invention is directed to a method of fabricating a MOS transistor having a fully silicided metal gate electrode. The method includes forming an isolation layer defining an active region in a semiconductor substrate. An insulated gate pattern crossing over the active region is formed. Spacers are formed on sidewalls of the gate pattern. A selective epitaxial growth process is carried out on the gate pattern and the active regions on both sides of the gate pattern to form source and drain protrusion regions and a gate sacrificial pattern. A silicidation process is applied to the semiconductor substrate having the source and drain protrusion regions and the gate sacrificial pattern to form elevated source and drain silicide layers and a silicide sacrificial pattern. An interlayer-insulating layer is formed on the entire surface of the semiconductor substrate having the elevated source and drain silicide layers and the silicide sacrificial pattern. The interlayer-insulating layer is planarized to form a reduced gate pattern. A silicidation process is applied to the semiconductor substrate having the reduced gate pattern to form a fully silicided metal gate electrode.
The foregoing and other objects, features and advantages of the invention will be apparent from the detailed description of embodiments of the invention, as illustrated in the accompanying drawing. The drawings, however, are not necessarily to scale; rather emphasis has been placed on illustrating the principles of the invention.
FIGS. 3 to 10 are cross-sectional views illustrating a method of fabricating a MOS transistor having a fully silicided metal gate electrode in accordance with embodiments of the present invention.
The present invention will now be described more fully hereinafter with reference to the accompanying drawings, in which embodiments of the invention are shown. This invention may, however, be embodied in different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. In the drawings, the thickness of layers and regions are exaggerated for clarity. In addition, when a layer is described to be formed on another layer or on a substrate, the layer may be formed on the other layer or on the substrate, or one or more layers may be interposed between the layer and the other layer or the substrate. Like numbers refer to like elements throughout the specification.
FIGS. 3 to 10 are cross-sectional views illustrating a method of fabricating a MOS transistor having a fully silicided metal gate electrode in accordance with some embodiments of the present invention.
Referring to
The gate conductive layer is patterned to form a gate pattern 57 crossing over the active region. In this case, the process of forming the gate pattern 57 may include forming a hard mask pattern and a photoresist pattern sequentially stacked on the semiconductor substrate 51 having the gate conductive layer, and selectively etching the gate conductive layer, using the hard mask pattern and the photoresist pattern as etch masks. Subsequently, low concentration impurity ions are implanted into the active region, using the gate pattern 57 and the isolation layer 53 as ion implantation masks, to form lightly doped drain (LDD) regions 59. The low concentration impurity ions may be N-type impurity ions or P-type impurity ions.
Referring to
A cleaning process for removing surface-contaminated particles may be carried out on the semiconductor substrate 51 before the formation of the spacer insulating layer. The cleaning process may include a first cleaning step using a wet cleaning solution containing HF and a second cleaning step using a mixed solution of NH4OH, H2O2, and H2O. The exposed portion of the gate dielectric layer 55 may be etched and removed while the cleaning process is carried out. That is, the gate dielectric layer 55 may be present only under the gate pattern 57.
The spacer insulating layer may be formed of silicon oxide (SiO), silicon nitride (SiN), silicon oxynitride (SiON), or another similar material. The spacer insulating layer is anisotropically etched to form spacers 63 on the sidewalls of the gate pattern 57. For simplicity of description, it is hereinafter assumed that in this embodiment the spacer 63 is formed of a silicon oxide layer 61 and a silicon nitride layer 62 which are sequentially stacked. Consequently, the top surface of the gate pattern 57 is exposed, and the active regions on both sides of the gate pattern 57 are exposed.
Referring to
In other embodiments, the process of forming the gate recess region 57A and the source and drain recess regions 59A may be skipped.
Referring to
The gate sacrificial pattern 67 and the source and drain protrusion regions 69 may be formed of silicon (Si), silicon germanium compound (SiGe), silicon carbon compound (SiC), carbon (C) doped SiGe, phosphorus (P) doped SiGe, boron (B) doped SiGe, or another similar material.
Furthermore, the etching process and the SEG process may be repeated at least twice to form the gate sacrificial pattern 67 and the source and drain protrusion regions 69 to a desired thicknesses.
Referring to
The surface of the semiconductor substrate 51 having the source and drain protrusion regions 69 is cleaned to remove a native oxide layer and contaminated particles remaining on the source and drain protrusion regions 69 and the gate sacrificial pattern 67. The cleaning process may include a first cleaning step using a wet cleaning solution containing HF and then a second cleaning step using a mixed solution of NH4OH, H2O2, and H2O.
A source and drain metal layer 72 and a capping layer 74 are sequentially formed on the entire surface of the cleaned semiconductor substrate 51. The source and drain metal layer 72 may be chosen from nickel (Ni), cobalt (Co), tungsten (W), tantalum (Ta), titanium (Ti), hafnium (Hf), nickel tantalum (NiTa), nickel platinum (NiPt), sequentially stacked nickel and cobalt (Ni/Co), and sequentially stacked PVD-Co/CVD-Co, or formed of at least two stacked layers thereof. The PVD-Co is cobalt (Co) formed by a physical vapor deposition (PVD) method, and the CVD-Co is cobalt (Co) formed by a chemical vapor deposition (CVD) method. The source and drain metal layer 72 may be formed by a PVD method, a CVD method, or an atomic layer deposition (ALD) method. In addition, the capping layer 74 may be formed of a titanium nitride (TiN) layer. In this case, the titanium nitride layer (TiN) acts to prevent the source and drain metal layer 72 from being oxidized. However, in other embodiments, the formation of the capping layer 74 may be skipped.
Referring to
The elevated source and drain silicide layers 69A may penetrate into partial regions of the source and drain regions 71. In this case, when the elevated source and drain silicide layers 69A are deeper than the junction depth of the source and drain regions 71, leakage current occurs. That is, it is preferable to form the elevated source and drain silicide layers 69A shallower than the junction depth of the source and drain regions 71. In addition, the silicide sacrificial pattern 67A may penetrate into a partial region of the gate pattern 57.
Subsequently, the unreacted portions of the source and drain metal layer 72 on the spacer 63 and the isolation layer 53 are removed. The unreacted source and drain metal layer 72 can be removed using a mixed solution of sulfuric acid (H2SO4) and hydrogen peroxide (H2O2). While the unreacted source and drain metal layer 72 is removed, the capping layer 74 may also be stripped.
An interlayer-insulating layer 77 is formed on the semiconductor substrate 51 having the elevated source and drain silicide layers 69A and the silicide sacrificial pattern 67A.
Referring to
A gate metal layer 81 and a gate capping layer 83 covering the reduced gate pattern 57B are sequentially formed. The gate metal layer 81 may be chosen from Ni, Co, W, Ta, Ti, Hf, NiTa, NiPt, sequentially stacked nickel and cobalt (Ni/Co), and sequentially stacked PVD-Co/CVD-Co, or formed of at least two stacked layers thereof. The PVD-Co is Co formed by a PVD method, and the CVD-Co is Co formed by a CVD method, as mentioned above. The gate metal layer 81 may be formed by a PVD method, a CVD method, or an atomic layer deposition (ALD) method. In addition, the gate capping layer 83 may be formed of TiN. In this case, the titanium nitride (TiN) layer prevents the gate metal layer 81 from being oxidized. However, in other embodiments, the formation of the gate capping layer 83 may be skipped.
Referring to
Subsequently, the unreacted portion of the gate metal layer 81 on the spacer 63 and the interlayer-insulating layer 77 is removed. The unreacted gate metal layer 81 can be removed using a mixed solution of H2SO4 and H2O2. While the unreacted gate metal layer 81 is removed, the gate capping layer 83 may also be stripped.
The source and drain metal layer 72 and the gate metal layer 81 may be formed of the same metal material or different metal materials from each other. When the source and drain metal layer 72 and the gate metal layer 81 are formed of different metal materials from each other, the elevated source and drain silicide layers 69A and the fully silicided metal gate electrode 89 may be formed of silicide layers of different metal materials from each other.
According to some embodiment of the present invention described above, after an elevated source and drain silicide layers and a silicide sacrificial pattern are formed using an SEG process and a silicidation process, a planarization process such as a CMP process is carried out to remove the silicide sacrificial pattern. A reduced gate pattern is exposed due to the removal of the silicide sacrificial pattern. The reduced gate pattern is transformed to a fully-silicided metal gate electrode using the silicidation process. Accordingly, the formation of deep silicide layers in source and drain regions can be prevented when forming a fully silicided metal gate electrode. That is, the elevated source and drain silicide layers can be formed in a region shallower than the junction depth of the source and drain. Consequently, a MOS transistor having a fully silicided metal gate electrode can be fabricated, which may have a higher integration density and better performance compared to conventional MOS transistors.
Embodiments of the present invention have been disclosed herein and, although specific terms are employed, they are used and are to be interpreted in a generic and descriptive sense only and not for purpose of limitation. Accordingly, it will be understood by those of ordinary skill in the art that various changes in form and details may be made without departing from the spirit and scope of the present invention as set forth in the following claims.
Number | Date | Country | Kind |
---|---|---|---|
2005-09258 | Feb 2005 | KR | national |