KR20010008583A - Method of forming metal-gate electrode in semiconductor device - Google Patents

Abstract

PURPOSE: A metal gate electrode formation method is provided to be capable of prohibiting a salicidation reaction while preventing diffusion of F and H ions into a silicon nitride film, by forming a silicon nitride film of about 20 - 30 angstrom on the surface of doped polysilicon at room temperature. CONSTITUTION: A metal gate electrode formation method includes sequentially forming a gate insulating film(102) and a doped polysilicon layer(104) on a semiconductor substrate(100). RF N2 plasma process is performed in the same chamber so that they can react with doped silicon, thus forming a thin silicon nitride film(106), on the doped polysilicon layer, for preventing compound generated by means of diffusion of ions within a tungsten film(108) to be formed later and a salicidation reaction. Tungsten of a high melting point is deposited on the thin silicon nitride film. Next, the stacked tungsten layer, the thin silicon nitride film and the polysilicon layer are patterned to form a gate electrode(G).

Description

Method of forming metal gate electrode in semiconductor device

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of manufacturing a semiconductor device, and more particularly, to a method of forming a metal gate electrode of a semiconductor device in which a tungsten layer is formed as a high melting point metal on a doped polysilicon to be applied to high integration and high speed of a semiconductor device. .

In general, the gate electrode of the semiconductor device is conductive by using doped polysilicon. When the design rule decreases according to the high density of the semiconductor device, the sheet resistance increases due to the high specific resistance of the polysilicon. Then, when the sheet resistance of the gate electrode is increased, the signal transmission time is delayed in the integrated circuit, which causes a problem in improving the operation speed of the device.

To solve this problem, a gate electrode is formed by adding high melting point metals such as tungsten (W), titanium (Ti), and cobalt (Co) that have low specific resistance and stable at high temperature as a gate electrode material. .

1 is a vertical cross-sectional view for explaining a method of forming a tungsten gate electrode according to the prior art, with reference to this will be described the gate electrode manufacturing process of tungsten.

First, a gate insulating film 12 is formed on a silicon substrate 10 as a semiconductor substrate, and a doped polysilicon layer 14 is formed thereon. The tungsten layer 16 is formed by depositing tungsten on the doped polysilicon layer 14.

Then, a photolithography process using a gate mask is performed to pattern the tungsten layer 16, the doped polysilicon layer 14, and the gate insulating layer 12 that are sequentially stacked to form a gate electrode G.

In the deposition process of tungsten 16, tungsten is deposited by reducing WF ₆ to H ₂ , so that F and H are in the tungsten layer at concentrations of 10 ¹⁹ to 10 ²⁰ atoms / cm 3 and 10 ²¹ to 10 ²³ atoms / cm 3, respectively. It is contained. The fluorine (F) and hydrogen (H) atoms contained in the tungsten film 16 diffuse to the gate insulating film in the subsequent thermal process to increase its thickness or to fix fixed charges at the doped polysilicon layer and the gate insulating film interface. charge center phenomena occur to reduce GOI (Gate Oxide Integrity) characteristics.

In addition, a salicidation reaction occurs at the interface between the tungsten film 16 and the doped polysilicon film 14 by a diffusion reaction. As shown in ⓒ of FIG. 1, silicides such as W ₂ Si, WSi ₂ , and WSi ( silicide) material is produced. As a result, the specific resistance of the gate electrode is changed to change the electrical characteristics of the transistor.

In order to improve the electrical characteristics of the gate electrode having such a tungsten metal, there is a need for a conductive diffusion barrier that can suppress ion diffusion and salicide reactions of F and H. At this time, the conditions of the conductive diffusion barrier are first, thermal stability and less reactivity between tungsten and doped polysilicon, and second, due to the small work function difference between tungsten and doped polysilicon causes a flat band change Third, the electrical conductivity must be large.

Materials satisfying these conditions include a titanium nitride film (TiN), a tungsten nitride film (WN), and a silicon nitride film (SiN). Titanium nitride has a limitation in using titanium silicide due to the reactivity between titanium and doped polysilicon, and in the case of tungsten nitride, the etching process depends on the difference in etching selectivity between tungsten and doped polysilicon during the gate etching process. Had difficulty securing margins.

To this end, a silicon nitride film (SiN) is used as a conductive diffusion barrier on the doped polysilicon layer during the manufacturing process of the tungsten metal gate electrode. This silicon nitride film process typically used a method of depositing silicon (Si), annealing in an N ₂ O atmosphere, or NH ₃ plasma nitridation or implanting N ⁺ ions.

However, in the case of N ₂ O atmosphere annealing and NH ₃ plasma nitridation, a complex gas is used at high temperature, which makes the manufacturing process complicated and a problem due to a thermal budget in which the dopant concentration distribution in the doped polysilicon is changed. This is accompanied by. In addition, in the case of N ⁺ ion implantation, a silicon nitride film including many defects is often formed, and the diffusion preventing function and the etching characteristics are deteriorated.

In addition, in a tungsten-structured gate electrode having a design rule of 0.15 탆 or less, when the thickness of the silicon nitride film is thicker than about 30 mW, a flat band due to the dielectric properties of the film itself, a work function, etc. Change occurs, resulting in deterioration of the transistor's electrical and GOI characteristics.

An object of the present invention is to solve the above problems of the prior art RF N ₂ plasma at room temperature when forming a silicon nitride film for preventing diffusion and salicide reaction between the tungsten and the doped polysilicon layer of the gate electrode of tungsten metal By forming a silicon nitride film having excellent crystal state on the doped polysilicon surface of 30 Å or less by carrying out the process, it provides a method of forming a metal gate electrode of a semiconductor device that can simplify the manufacturing process and improve the electrical characteristics of GOI and transistor. have.

1 is a vertical cross-sectional view for explaining a method of forming a tungsten gate electrode according to the prior art,

2 is a vertical sectional view showing a structure of a tungsten gate electrode according to the present invention;

3A to 3E are process flowcharts illustrating a manufacturing process of the tungsten gate electrode illustrated in FIG. 2.

Explanation of symbols on the main parts of the drawings

100: silicon substrate 102: gate insulating film

104: doped polysilicon layer 106: silicon nitride thin film

108: tungsten layer

G: gate electrode

In order to achieve the above object, the present invention forms a gate electrode in which a doped polysilicon layer and a tungsten layer are stacked on a semiconductor substrate on which a gate insulating film is formed, and sequentially forms a gate insulating film and a doped polysilicon layer on a semiconductor substrate. Forming and performing an RF N ₂ plasma process on the doped polysilicon layer in the same chamber to react with the doped polysilicon to prevent a compound produced by ion diffusion and salicide reaction in the tungsten film to be formed subsequently. Forming a silicon nitride thin film, depositing a high melting point tungsten on the silicon nitride thin film to form a tungsten layer, and patterning the stacked tungsten layer, silicon nitride thin film and polysilicon layer to form a gate electrode of the semiconductor device. It characterized by comprising the step of forming.

According to the present invention, a silicon nitride thin film of about 20 to 30 microseconds is formed on the surface of the doped polysilicon at room temperature by an RF N ₂ plasma nitridation process as a subsequent treatment of the doped polysilicon deposition process in the tungsten metal gate manufacturing process. The electrical properties of the GOI and the transistor are improved by the diffusion preventing function of the silicon nitride film and the suppression of the salicide reaction to the F and H ions.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

2 is a vertical cross-sectional view showing a structure of a tungsten gate electrode according to the present invention, the present invention is a semiconductor insulating film 102 formed on the silicon substrate 100 as a semiconductor substrate, and a doped polysilicon layer sequentially stacked thereon A gate electrode composed of a 104 and a tungsten layer 108, an RF N ₂ plasma for suppressing the salicide reaction while preventing diffusion of conductive ions between the doped polysilicon layer 104 and the tungsten layer 108. The silicon nitride thin film 106 formed by the nitriding process is about 20-30 micrometers in thickness.

3A to 3E are process flowcharts illustrating a manufacturing process of the tungsten gate electrode shown in FIG. 2. Referring to this, a method of forming a tungsten gate electrode according to the present invention is as follows.

First, as shown in FIG. 3A, the gate insulating film 102 is formed on the silicon substrate 100 to have a thickness of 50 μs to 100 μs, and a doped polysilicon layer 104 is formed on the silicon substrate 100 to a thickness of 500 μs to 1000 μs. At this time, the doped polysilicon layer 104 is a CVD (chemical) at a temperature of 500 ℃ to 700 ℃ using a mixture ratio of 1.1: 1.5 to 1.5: 1.8 using SiH ₄ as a reactor and PH ₃ as a dopant vapor deposition) method.

As shown in FIG. 3B, an Ar plasma process is performed in an in-situ process to activate the surface of the doped polysilicon layer 104 to promote reactivity between N + ions and silicon. At this time, the process pressure and the flow rate for ensuring plasma stability are preferably set to 3 to 8 mTorr and 10 to 30 sccm, respectively.

As shown in FIG. 3C, an RF N ₂ plasma process may be performed on the doped polysilicon layer 104 in the same chamber to react with the doped polysilicon to cause ion diffusion and salicide reaction in the tungsten film to be formed. The silicon nitride thin film 106 which prevents the compound produced by this is formed in about 20-30 micrometers in thickness. At this time, in the nitriding treatment step, the process pressure and the flow rate for securing the stability of the plasma are preferably set to 2 to 5 mTorr and 10 to 50 sccm, respectively.

On the other hand, in general, a conventional silicon nitride film (SiNx) is produced by the following chemical reaction.

N ₂ → N + N

SiH ₄ → Si + 2H ₂

Si + N → SiNx

This chemical reaction proceeds at about 850-1000 ° C. where the free reaction energy value becomes negative.

However, when using an RF N ₂ plasma as in the present invention, the following chemical reaction required for SiN x formation is performed at room temperature by reaction between activated N ⁺ ions in the plasma and doped polysilicon surface atoms.

N → N ^{+ +} e ^-

N ⁺ + Si → SiNx

Here, N ⁺ is nitrogen ion activated in the plasma, and e ⁻ is electrons emitted during the plasma discharge.

In addition, when the plasma is formed by the RF discharge of the present invention, the formation of the same charge accumulation layer is suppressed as compared with the DC discharge plasma, and the repulsion phenomenon of the reaction ions (the electrostatic repulsion does not move the ions of the same charge to the reaction surface). Since the surface mobility between Si and N ⁺ is increased, a silicon nitride film having excellent crystallinity is formed at a high nitriding rate.

As shown in FIG. 3D, a high melting point tungsten is deposited on the silicon nitride thin film 106 to a thickness of 1000 GPa to 1500 GPa to form a tungsten layer 108. At this time, the deposition process of the tungsten layer 108 uses WF ₆ and H ₂ as the reactor body, the mixing ratio is ₂ to 3.5: 1 to 1.8, and the CVD method is used at a temperature of 350 ° C to 450 ° C.

3E, a photolithography process using a gate mask is performed to pattern the sequentially stacked tungsten layer 108, the silicon nitride thin film 106, and the polysilicon layer 104 to gate the semiconductor device. An electrode G is formed and the gate insulating film is etched according to the pattern to complete the gate electrode of the present invention.

As described above, the method of forming the gate electrode according to the present invention has the following advantages by forming a silicon nitride thin film on the doped polysilicon surface by in-situ plasma nitridation before tungsten deposition during the production of the tungsten gate electrode. Can be.

First, when the tungsten layer is deposited on the silicon nitride layer, the tungsten layer (W) / silicon nitride layer (SiNx) exhibits a uniform energy band while improving adhesion strength through chemical affinity between the W of the tungsten layer and the N atoms of the silicon nitride layer. As the / doped polysilicon layer interface is secured, the change of the flat band is reduced, thereby improving transistor characteristics.

Second, the present invention can improve GOI characteristics by preventing diffusion and suppression of salicide reactions of F and H ions in the tungsten layer of the silicon nitride film and P ions in the doped polysilicon.

Third, since the silicon nitride film of the present invention suppresses interfacial grooving due to consumption of doped polysilicon at the interface of tungsten / doped polysilicon, a uniform interface can be secured.

Fourth, the silicon nitride film of the present invention has an effect of doubling the resistance uniformity of the doped polysilicon by suppressing the dopant redistribution in the doped polysilicon during the subsequent thermal process.

Claims (5)

In forming a gate electrode in which a doped polysilicon layer and a tungsten layer are stacked on a semiconductor substrate on which a gate insulating film is formed, Sequentially forming a gate insulating layer and a doped polysilicon layer on the semiconductor substrate; A silicon nitride thin film is formed on the doped polysilicon layer by performing an RF N ₂ plasma process in the same chamber to react with the doped polysilicon to prevent a compound produced by ion diffusion and salicylation reaction in the tungsten film to be formed. Forming; Depositing tungsten with a high melting point on the silicon nitride thin film; And And forming a gate electrode of the semiconductor device by patterning the stacked tungsten layer, the silicon nitride thin film and the polysilicon layer. The method of forming a metal gate electrode of a semiconductor device according to claim 1, wherein the silicon nitride thin film has a thickness of 20 to 30 GPa. The semiconductor device of claim 1, wherein a plasma having a waveform of 13.56 MHz is used in the RF N ₂ plasma process, and a process pressure is 2 to 5 mTorr and a flow rate is 10 to 50 sccm in order to ensure plasma stability. Method of forming a metal gate electrode. The metal of the semiconductor device of claim 1, wherein the doped polysilicon layer is activated by an Ar plasma in an in-situ process to promote reactivity between N + ions and silicon prior to forming the silicon nitride thin film. Gate electrode formation method. The method of claim 4, wherein the pressure is 2 to 5 mTorr and the flow rate is 10 to 50 sccm during the Ar plasma process.