TW200828425A - Method for fabricating semiconductor device with gate stack structure - Google Patents

Abstract

A method for fabricating a semiconductor device includes forming a first conductive layer over a substrate, forming an intermediate structure over the first conductive layer, the intermediate structure formed in a stack structure comprising at least a first metal layer and a nitrogen containing metal silicide layer, and forming a second conductive layer over the intermediate structure.

Description

200828425 IX. Inventive Note: [Reference References for Related Applications] The present invention claims Korean Patent Application No. 10-2006-01 34326 and No. 10-, filed on December 27, 2006 and April 27, 2007. Priority of 2007-0041 28 No. 8 and all of these Korean patent applications are incorporated by reference. BACKGROUND OF THE INVENTION 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 semiconductor device having a gate stack structure. The present invention relates to a semiconductor device and a method of fabricating the same, and more particularly to a gate stack and a method of fabricating the same. [Prior Art] A tungsten polysilicon gate electrode formed by stacking polycrystalline germanium and tungsten has a very low electrical resistance, which is approximately polycrystalline germanium/tungsten telluride formed by stacking polycrystalline germanium and tungsten germanium (Poly-Si/ WSh) The resistance of the gate electrode is 1 / 5 to 1 / 1 0. Thus, the tungsten polysilicon gate electrode is required for the fabrication of the secondary _ 6 〇 n m memory device. Figures 1A through 1C depict a typical tungsten polysilicon gate stack structure. As shown in Fig. 1A, the tungsten polysilicon gate stack structure is formed by continuously stacking a polysilicon layer 11, a tungsten nitride (WN) layer 12, and a tungsten (W) layer 13. The WN layer 12 acts as a diffusion barrier. During the subsequent annealing process or gate re-oxidation process, the nitrogen in the WN layer 12 is decomposed between the tungsten layer 13 and the polysilicon layer 11 into a non-uniform insulating layer such as SiNx -5 - 200828425 and SiOxNy. The non-uniform insulating layer has a thickness ranging from about 2 nm to 3 nm. Thus, a component error of an image signal delay may occur at an operating frequency of several hundred megahertz (MHz) and an operating voltage of 1.5 V or less. Recently, a thin tungsten chevron (WSix) or titanium (Ti) layer as a diffusion barrier layer has been formed between the polycrystalline sand layer 11 and the WN layer 12 to prevent the tungsten layer 13 and the polysilicon layer. 11 forms Si-N bonds. As shown in FIG. 1B, if a tungsten germanide (WSix) layer 14 is formed between the polysilicon layer 1 1 and the WN layer 12, it is used during the formation of the WN layer 12. A nitrogen plasma forms a W-Si-N bond over the WSix layer 14. It is well known that W-Si-N is a good diffusion barrier layer having metallic properties. As shown in FIG. 1C, if titanium (germanium layer 15) is formed between the polysilicon layer 1 1 and the WN layer 12, the nitrogen plasma will be used in the reactive sputtering process during formation of the WN layer 12. Ti of the titanium layer 15 is converted into titanium nitride (TiN). The TiN layer acts as a diffusion barrier layer. As a result, although the WN layer 12 is decomposed during the subsequent thermal process, the TiN prevents diffusion of nitrogen toward the polysilicon 11 Come out, therefore, can effectively reduce the formation of Si-N. However, if the tungsten polysilicon gate is applied to the double polysilicon gate [that is, N of the N-type MOS field effect transistor (NM0SFET), a polycrystalline germanium gate and a P + -type polysilicon gate of a P-type MOS field effect transistor (PM0SFET), if the WSH/WN diffusion barrier structure is used in the tungsten polysilicon gate, it can be greatly increased Contact resistance between the tungsten layer and the P + -type polysilicon layer. Conversely, if the Ti/WN diffusion barrier structure is used in the tungsten polysilicon gate, the contact resistance between the tungsten layer and the P + -type polysilicon layer It is lower regardless of the polysilicon doping type. -6- 200828425 In this PM OS FET In the case of P + -type polysilicon, polycrystalline germanium depletion effect may occur in the reverse state of the actual operation mode. The polycrystalline germanium depletion effect may be dependent on the amount of boron retained in the P + -type polysilicon. The WSix/WN diffusion barrier structure may produce a larger polysilicon enthalpy effect than the Ti/WN diffusion barrier structure. Therefore, the WSix/WN diffusion barrier structure may degrade the transistor characteristics. As a result, the Ti/WN diffusion resistance The barrier structure can provide low contact resistance between the tungsten layer and the polysilicon layer and prevent P-type polysilicon from being depleted. Therefore, it is recommended to use the Ti/WN diffusion barrier structure. However, if a Ti/WN diffusion barrier structure is used, it may be The sheet resistance (Rs) of tungsten formed directly above the Ti/WN diffusion barrier structure is increased by about 1.5 to 2 times. Therefore, the increase of the sheet resistance (Rs) may affect the development of the tungsten polysilicon gate in the future. Embodiments of the present invention relate to a gate stack including a semiconductor device having an intermediate structure, wherein the intermediate structure has a low sheet resistance and a contact resistance and Effectively preventing outward diffusion of impurities, and a method for fabricating the gate stack. According to one aspect of the present invention, a method of fabricating a semiconductor device is provided, the method comprising forming a first conductive layer formed on a substrate; Forming an intermediate structure over the first conductive layer, the intermediate structure forming the stacked structure includes at least a first metal layer and a nitrogen of the nitrogen-containing metal telluride layer; and forming a second conductive layer over the intermediate structure. 200828425 Another aspect provides a method of fabricating a semiconductor device, the method comprising: forming a first conductive layer formed on a substrate, forming an intermediate structure over the first conductive layer, the intermediate structure forming the stacked structure comprising a first metal layer, a second metal layer, a metal telluride layer, and a third metal layer; and a second conductive layer is formed over the intermediate structure. [Embodiment] Fig. 2A is a graph showing the contact resistance between tungsten and polysilicon in a structure for each type of diffusion barrier. It can be observed that when a tungsten germanide (WSi〇/tungsten nitride (WN) or titanium (Ti)/WN structure is used in place of the tungsten nitride (WN) structure, the N-type impurity is doped greatly. The contact resistance indicated by RC between polycrystalline germanium (N+ POLY-Si) and tungsten (W). However, if the tungsten polysilicon gate is applied to a double polysilicon gate [ie, N-type metal oxide semiconductor field effect transistor] N + -type polysilicon gate of (NM0SFET) and P + -type polysilicon gate of P-type MOS field effect transistor (PM0SFET), if the WSix/WN structure is used in the tungsten polysilicon gate, The contact resistance between the tungsten and the P + -type polysilicon (P + P 〇 LY - Si) is greatly increased. Conversely, if the Ti / WN structure is used in the tungsten polysilicon gate, the tungsten and P + - The contact resistance between the polycrystalline turns exhibits a low level regardless of the polysilicon doping type. In the case of the P + -type polysilicon of the PMOS transistor, the polycrystalline germanium depletion effect can be generated in the inverted state of the actual operation mode. The polycrystalline germanium depletion effect is dependent on the amount of boron retained in the P + -type polycrystalline germanium. Figure 2B is a graph depicting the depth profile of the boron concentration for each type of gate stack. As described in the WSH/WN structure, the boron concentration is in the gate-8-200828425 pole insulating layer (eg oxide The surface of the layer and the polycrystalline crucible is as low as about 5 χ 1019 atoms/cm 3 . When the Ti/WN structure is used, the boron concentration measured at the same position is greater than about 8×10 19 atoms/cm 3 . As a result, in the WSix/WN structure. The WSix/WN structure reduces the characteristics of the transistor compared to the space in the Ti/WN structure. Therefore, it is preferable to use the Ti/WN structure, and the Ti/WN structure is provided in the W Low contact resistance with the polysilicon and prevention of P-type polysilicon vacancies. However, the application of the Ti/WN structure is limited. The sheet resistance (Rs) of W formed over the Ti/WN structure φ is increased by about 1.5 to 2. This limit will be described in more detail in Figure 2C. Figure 2C is a graph depicting the sheet resistance of a structure for each type of diffusion barrier. The sheet resistance of W is labeled Rs Usually, it can be in polycrystalline germanium layer, tantalum nitride (SiCh) layer, tantalum nitride (ShN4) layer and WS An amorphous nitrogen-containing tungsten (WNX) layer is formed over the ix layer, and thus, W having a low specific resistance (that is, in a range of about 15 μΩ - 〇ηη to 20 μΩ < ιη) can be formed thereon. The titanium nitride (TiN) and the large nitride (TaN) of the pure titanium (Ti), tungsten (W) and giant (Ta) and metal nitride φ materials form a W having a relatively small grain size. W having a specific resistance of about 30 μ Ω-cm is formed thereon. The increase in sheet resistance caused by the application of the Ti/WN structure may limit the future development of the tungsten polysilicon gate. According to various embodiments of the invention to be described below, the intermediate structure of the gate stack of different forms is formed with a plurality of thin layers comprising TI, W, bismuth (Si) or nitrogen (N) or each layer containing a large amount of nitrogen Thin layer. The intermediate structures act as diffusion barriers which reduce the contact resistance and the sheet resistance, as well as prevent penetration and outward diffusion of impurities. -9- 200828425 In the following examples, the term "layer/structure containing nitrogen or nitrogen-containing layer/structure (nitrogen c ο ntaining 1 aye r7 structure)" Structure and metal layer/structure containing nitrogen in a certain amount/weight ratio. Also, X in WSixNy represents the ratio of bismuth to tungsten, which ranges from about 0.5 to 3.0, and y represents the ratio of nitrogen to tungsten ruthenium, which ranges from about 0.01 to 10.00. Fig. 3A depicts a gate stack structure in accordance with a first embodiment of the present invention. The gate stack structure includes a first conductive layer 21, an intermediate structure 22, and a second conductive layer 23 formed in sequence. The first conductive layer 21 includes a polysilicon layer highly doped with a P-type impurity (e.g., boron) or an N-type impurity (e.g., phosphorus). The first conductive layer 21 may also include a polysilicon layer (Si〃xGex, wherein the X system is in a range between about 0.01 and 1.0) or a germanide layer. For example, the telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), tantalum (Ta), (Hf), cone (Zr), and platinum. One of the groups consisting of (Pt). The second conductive layer 23 includes a tungsten layer. The tungsten layer is about 100A to 2000A thick and is formed by performing a physical vapor deposition (PVD) method, a chemical vapor deposition (CVD) method, or an atomic layer deposition (ALD) method. The PVD method includes a sputtering deposition method using a tungsten sputtering target.

The intermediate structure 22 includes a titanium layer 22 A, a tungsten-containing tungsten (WN layer 22B, and a nitrogen-containing tungsten germanide (WShNy) layer 22C. In detail, the thickness of the titanium layer 22A is from about 10 A to about 8 Preferably, the seed layer 2 2 A has a thickness of from about 10 A to about 50 A. The titanium layer 22A is changed to TiN by subsequent WNX deposition to form a nitrogen-containing layer. The tungsten layer 22β, and some of its lower portion react with the first conductive layer 21, that is, the polysilicon layer thus forms a TiSix layer, and thus has a thickness as defined above. If the titanium layer 22A -10-

The thickness of 200828425 is large, and the thickness of the TiSh layer is also raised due to its volume expansion. Further, if the thickness of the titanium layer 22A is large, 22A can absorb the dopant of the polysilicon layer 21, for example, phosphorus or boron, which causes deterioration of device performance due to multiple depletion in the germanium layer 21. As described above, the nitrogen in the nitrogen-containing tungsten layer 22B is in the range of about 0.3 to 1.5 for tungsten. The nitrogen-containing tungsten layer is regarded as a tungsten layer of a certain content/weight ratio of tungsten nitride. Although it will be described in the following example, it is known that the nitrogen-containing tungsten layer 22B supplies nitrogen to the nitrogen-containing tungsten layer 2 2C. The nitrogen-containing tungsten layer 22B has a thickness of about 20A to 200A to the nitrogen of the nitrogen-containing tungsten telluride layer 2 2C. After the subsequent removal, the nitrogen-containing tungsten layer 22B becomes a pure tungsten layer or a tungsten layer containing a trace of nitrogen. The ratio of germanium to tungsten in the nitrogen-containing tungsten telluride layer 22C is in the range of 3.0, and the nitrogen of the nitrogen-containing tungsten germanide layer 22C is in the range of about 10% to about 60%. Here, the nitrogen content of the nitrogen-containing tungsten telluride layer is appropriately adjusted in the above manner. If the nitrogen content is too low, the nitrogen-containing tungsten carbide layer 22C cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten crucible 22C may be high, and thus the contact resistance becomes high. The nitrogen-containing tungsten silicide layer 22C represents a tungsten nitride layer (i.e., a tungsten-rhenium nitride layer) or a telluride layer containing a certain content/weight ratio. The thickness of the nitrogen-containing tungsten carbide layer 22C is formed to be in the range of about 200 Å. The recording of the nitrogen-containing tungsten layer 22B is carried out by performing a PVD method, a CVD method or an ALD method. The nitrogen-containing tungsten layer 22C is formed by performing a pVD method. The PVD method increases the thickness of the titanium layer by sputtering deposition or reactive sputtering, which is in a multi-proportion layer or contains three implementations of a telluride. Due to the fire treatment 〇: about 0 · 5 content is the reaction wall of the surface of 22C and the chemical layer, which leads to the tungsten enthalpy of the material deuterated nitrogen about 20A: layer 22A i 矽 t 积 进 -11-

200828425 ί. For example, by performing a sprinkling method with a sprinkling mine, 22 Α. The nitrogen-containing tungsten layer 22B is formed by performing a reaction formula with a tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing formation is carried out by performing a reactive sputtering deposition method with a plating target in a nitrogen atmosphere; 220 In particular, since the tungsten germanide layer 22C is not easily formed over the nitrogen-containing tungsten layer 22B, the PVD method is used (for example: Inversely forming) to form the nitrogen-containing tungsten telluride layer 22C. If the nitrogen-containing tungsten germanide layer 22C is formed by ^, the nitrogen-containing tungsten germanide layer 22C is uniformly grown in the nitrogen-containing tungsten layer method, so that tungsten oxide is present above the nitrogen-containing tungsten layer 22B ( The W?x) layer, the nitrogen-containing tungsten germanide layer 22C formed by the CVD method is attached to the agglomerate. However, the reactive sputtering deposition method was applied to the tungsten in the nitrogen atmosphere to allow the formation of the nitrogen-containing tungsten carbide to be uniform regardless of the underlying type. Figure 3B depicts an image obtained after the tungsten germanide layer on the nitrogen-containing tungsten layer by the PVD method. The PVD method is used to reactively deposit the layer of the moon layer over the nitrogen-containing tungsten layer. Reference letters WSiN and WN denote the containing | and the nitrogen-containing tungsten layer, respectively. According to a first embodiment of the present invention, the gate is stacked with a conductive layer 21, and a Ti/WNx/WSixNy intermediate structure 22 2 layer 23. The first conductive layer 21 includes a polysilicon and the second tungsten to form a tungsten polysilicon gate stack structure. In particular, the Ti/WNx/WSixNy intermediate structure includes the formation of the smear layer, the sputter deposition method, the tungsten sulphide sputter sulphide layer, the growth of the nitrogen-containing splatter, and the CVD method. Therefore, by virtue of the force, the conductor sputtering target layer 22C is uniformly formed. The nitrogen-containing deposition method is as follows: the nitrogen-containing tungsten stone, the tungsten germanide layer includes the first: the second conductive conductive Layer 23 comprises a metal layer, a -12-.200828425 second metal layer and a nitrogen-containing metal telluride layer stack structure. More specifically, the first metal layer, the second metal layer, and the nitrogen-containing metal telluride layer respectively comprise a pure metal layer, a nitrogen-containing metal layer, and a nitrogen-containing metal telluride layer. For example, the first metal layer, the second metal layer, and the nitrogen-containing metal telluride layer are respectively the titanium layer 22A' of the nitrogen-containing tungsten (WNX) layer 22B and the nitrogen-containing tungsten germanide (WShNy) layer 22C. It is also possible to form the intermediate structure including the above multiple layers in other different structures. For example, the first metal layer includes a giant (T a) layer in addition to the titanium layer, and the second metal layer includes a nitrogen-containing titanium tungsten layer in addition to the nitrogen-containing tungsten layer. The nitrogen-containing metal telluride layer includes, in addition to the nitrogen-containing tungsten telluride layer, a nitrogen-containing titanium telluride layer or a nitrogen-containing giant telluride layer. The macrolayer is formed by performing a PVD method including sputtering, a CVD method, or an ALD method. The nitrogen-containing titanium tungsten layer was formed by performing a reactive sputtering deposition method with a titanium tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing titanium sand layer and the nitrogen-containing giant telluride layer are formed by performing reactive sputtering deposition using individual titanium telluride and giant telluride sputtering targets in a nitrogen atmosphere. The macrolayer is formed to a thickness of about 1 to 80 Å. The Ta layer 22A preferably has a thickness of from about 10A to about 50 people. The Ta layer is formed by changing some of its upper portions to TaN by subsequent WNX deposition, and some of its lower portions react with the first conductive layer 21, that is, the polysilicon layer thus forms a TaSix layer, It has a thickness as defined above. If the thickness of the Ta layer is large, the thickness of the TaSh layer also increases and bulges due to its volume expansion. In addition, if the thickness of the T a layer is large, the τ a layer can absorb the dopant of the polycrystalline germanium layer 2 1 , such as shoulder or boron, so multiple depletion occurs in the polycrystalline sand layer 2 1 , resulting in component performance. Deterioration. Each of the nitrogen-containing tungsten layer, the nitrogen-containing sulphate layer, and the nitrogen-containing strontium-13-.200828425 layer has a thickness of about 20A to 200A and each layer has about 10% and 60%. The nitrogen content in the range between %. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing titanium or telluride layer cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing titanium or group telluride layer may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. Meanwhile, in the nitrogen-containing titanium tungsten layer, the ratio of titanium to tungsten is in the range of about 0.5 to 3.0. In the nitrogen-containing titanium sand layer, the ratio of sand to titanium is in the range of about 〇 5 φ to 3 · 0. In the nitrogen-containing group telluride layer, the ratio of ruthenium to osmium is in the range of about 0.5 to 3.0. Fig. 3C depicts a gate stack structure in accordance with a second embodiment of the present invention. In particular, the gate stack structure is an exemplary gate stack structure modified from the gate stack structure in accordance with the first embodiment of the present invention. In other words, the gate stack structure includes a gas-containing titanium layer in place of the titanium layer 22A described in FIG. 3A, which is identified as TiNx, where x is less than about 1. The gate stack structure according to the first embodiment includes a first conductive layer _ 0.001, an intermediate structure 202, and a second conductive layer 203. The first conductive layer 201 includes a polysilicon layer which is doped with a P-type impurity (for example, boron (B)) or an N-type impurity (for example, phosphorus (p)). In addition to the polysilicon layer, the first conductive layer 2 may also include a polycrystalline germanium (SihGex) layer 'where X is in the range of about 〇1 to 1 ’' or include a telluride layer. The telluride layer comprises a layer selected from the group consisting of nickel (nickel), chromium (co, cobalt (c), titanium (Ti), tungsten (w), tantalum (Ta), (Hf), chromium (Zr), and platinum (Pt). One of the groups formed.

& A conductive layer 2 0 3 includes a tungsten layer. The p v d method, the c V D method, and the ALD method are carried out to form a tungsten layer of about 100 A to 2, which is thick. The pvD-14-, 200828425 method includes a sputter deposition method using a tungsten sputtering target. The intermediate structure 202 includes a nitrogen-containing titanium (TiNx) layer 202A, a nitrogen-containing tungsten (WNX) layer 202B, and a nitrogen-containing tungsten germanide (WSixNy) layer 202C. More specifically, the nitrogen of the nitrogen-containing titanium layer 202A has a certain ratio to titanium, for example, a range of about 0.2 to 0.8. Here, the nitrogen-containing metal layer, i.e., the nitrogen-containing titanium layer 202A, has a ratio of nitrogen to titanium as described above to prevent the SiN from being generated in the TiNx layer. During the subsequent annealing treatment, since excessive Ti in the TiNx layer damages the Si-N bond formed between the polysilicon and TiNx, and thus SiN is removed, the generation of SiN can be prevented. This may be because TiN and SiN have a strong bond. Unlike the titanium layer 2 2 A described in Fig. 3A, the nitrogen-containing titanium layer 2 0 2 A is formed to have a thickness of about 10 to 150 Å. The nitrogen-containing titanium layer 202A represents a titanium nitride layer or a titanium layer containing a certain content/weight ratio of nitrogen. The nitrogen-containing tungsten layer 202B has a certain ratio of nitrogen to tungsten, for example, in the range of about 0.3 to 1.5. The nitrogen-containing tungsten layer 202B represents a tungsten nitride layer or a tungsten layer containing a certain content/weight ratio of nitrogen. Although described later, the nitrogen-containing tungsten layer 202B supplies nitrogen to the nitrogen-containing tungsten carbide layer 202C. The nitrogen-containing tungsten layer 202B is formed to have a thickness of about 20A to 200A. Due to the supply of nitrogen, the nitrogen-containing tungsten layer 202B becomes a pure tungsten layer or a tungsten-containing tungsten layer after annealing treatment. The ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 202C is in the range of between about 0.5 and 3.0, and the nitrogen content of the nitrogen-containing tungsten germanide layer 202C is in the range of from about 10% to about 60%. Inside. Here, the nitrogen content is appropriately adjusted as described above. If the nitrogen content is too low, since the nitrogen-containing tungsten silicide layer 202C cannot be successfully used as a diffusion barrier, a junction reaction occurs. On the other hand, if the nitrogen content is too high, it is included in the nitrogen-containing tungsten telluride layer 202C -15-200828425

The SiN content can be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The nitrogen-containing tungsten decide layer 202C represents a tungsten sand nitride layer or a tungsten ruthenide layer containing a certain content/weight ratio of nitrogen. The nitrogen-containing tungsten layer 202B is formed by performing a PVD method, a CVD method, or an ALD method. The nitrogen-containing titanium layer 202A and the nitrogen-containing tungsten germanide layer 202C are formed by a PVD method. The PVD method is carried out by a sputtering deposition method or a reactive sputtering deposition method. For example, the nitrogen-containing titanium layer 202A is formed by sputtering deposition using a titanium sputtering target in a nitrogen atmosphere. The nitrogen-containing tungsten layer 202B is formed by performing a reactive sputtering deposition method with a tungsten sputtering target φ in a nitrogen atmosphere. The nitrogen-containing tungsten telluride layer 202C is formed by performing a reactive sputtering deposition method using a tungsten telluride sputtering target in a nitrogen atmosphere. In particular, since the nitrogen-containing tungsten germanide layer 202C is not easily grown over the nitrogen-containing tungsten layer 202B, the PVD method (e.g., reactive sputtering deposition method) is used to form the nitrogen-containing tungsten germanide layer 202C. When the nitrogen-containing tungsten germanide layer 202C is formed by the CVD method, the nitrogen-containing tungsten germanide layer 202C cannot be uniformly grown over the nitrogen-containing tungsten layer 202B, thereby causing agglomeration. φ because the tungsten oxide (W layer) is present above the nitrogen-containing tungsten layer 202B, which weakens the adhesion of the nitrogen-containing tungsten carbide layer 202C formed by the CVD method, thereby causing the agglomeration. The reactive sputtering deposition method is performed in the nitrogen atmosphere with the tungsten telluride sputtering target to allow uniform formation of the nitrogen-containing tungsten carbide layer 202C regardless of the underlying type. When used similar to the first embodiment When the nitrogen-containing titanium layer 202A in the second embodiment of the titanium layer 22A is used, a low contact resistance can be obtained. The reason for obtaining the low contact resistance is because nitrogen is supplied to the nitrogen-containing titanium layer 202A to form a nitrogen-containing tungsten The layer 202B, whereby the upper portion of the nitrogen-containing titanium layer 202A is made strong, and at the same time - 16-200828425 prevents the agglomeration of the Ti-Si bond. The gate stack structure according to the second embodiment of the present invention includes the first conductive The layer 201, the TiNx/WNJWSixNy intermediate structure 202 and the second conductive layer 203. The first conductive layer 201 comprises a polysilicon and the second conductive layer 203 comprises tungsten, thereby forming a tungsten polysilicon gate stack structure. TiNx/WNx/WSixNy intermediate structure 202 to include Forming a stack of a gold layer, a second metal layer, and a nitrogen-containing metal telluride layer. The first and second metal layers are a metal φ layer containing a certain content/weight ratio of nitrogen and the nitrogen-containing metal deuteration The layer includes a certain content/weight ratio of nitrogen. For example, the first metal layer is the nitrogen-containing titanium layer 202 A. The second metal layer is the nitrogen-containing tungsten layer 202B. The metal telluride layer is the nitrogen-containing layer. The tungsten germanide layer 202C. The multilayer intermediate structure as described above may also be formed in other different structures. For example, the first nitrogen-containing metal layer includes a nitrogen-containing molybdenum layer (TaNx) layer in addition to the nitrogen-containing titanium layer, and The second nitrogen-containing metal layer includes a nitrogen-containing titanium tungsten (TiWNx) layer in addition to the nitrogen-containing tungsten layer. The nitrogen-containing metal halide layer includes nitrogen-containing titanium germanium in addition to the nitrogen-containing tungsten germanide layer. a (TlSixNy) layer or a nitrogen-containing group telluride (TaSixNy) layer. The nitrogen-containing germanium layer is formed by performing a PVD method including sputtering, a CVD method or an ALD method, by sputtering a target with titanium tungsten in a nitrogen atmosphere. A reactive sputtering deposition method is performed to form the nitrogen-containing titanium tungsten layer by using a nitrogen ring The nitrogen-containing titanium telluride layer and the nitrogen-containing molybdenum telluride layer are formed by reactive sputtering deposition using individual titanium telluride and molybdenum telluride sputtering targets. The thickness of the nitrogen-containing molybdenum layer is about 10A. Up to 80 A. Each of the nitrogen-containing titanium tungsten layer, the nitrogen-containing titanium telluride layer, and the nitrogen-containing giant telluride layer has a thickness of about 20A to 200A and each layer has about 10% and -17-200828425 The nitrogen content is in the range of 60%. Here, the nitrogen content is adjusted as above. If the nitrogen content is too low, the junction reaction occurs because the nitrogen-containing titanium or telluride succeeds as a diffusion barrier. On the other hand, if the amount is too high, the SiN contained in the nitrogen-containing titanium or giant telluride layer is high, and thus the contact resistance becomes high, resulting in deterioration of element performance. In the tungsten layer, the ratio of titanium to tungsten is in the range of about 0.5 to 3.0. In the nitroxide telluride layer, the ratio of niobium to titanium is within about 0.5 to 3.0. In the nitrogen-containing macrochemical layer, the ratio of cerium to molybdenum is in the range of about φ 3.0. Similar to the TiNx/WNx/WSixNy intermediate structure, including the intermediate structure including the Ni-Ti-containing layer, it has a low contact resistance and at the same time prevents a polysilicon from being depleted. Although the intermediate structure according to the embodiment is formed in three layers, the intermediate structure may further include a nitrogen-containing tungsten (WNX) layer over the layer. The additional provided contains a complex φ nitrogen having a thickness substantially the same as that of the first provided nitrogen-containing tungsten layer and nitrogen according to the TiNx/WNx/WSixNy intermediate structure of the second embodiment. As a result, the TiNx/WNx/WSixNy intermediate structure can have low and contact resistance and reduce the height of the gate stack structure. And the TiNx/WNx/WSixNy intermediate structure can reduce polysilicon caused by out-diffusion of impurities (for example, boron) doped in the first conductive layer. FIG. 3D depicts a gate electrode according to a third embodiment of the present invention. Structure. The gate stack structure includes a first conductive layer 21, an intermediate junction, and a second conductive layer 213. The first conductive layer 211 includes polycrystals of highly doped impurities (for example, boron (B)) or N-type impurities (for example, phosphorus (P)). The first conductive layer 211 may be in addition to the polysilicon. Including the inclusion of the stratified layer, the nitrogen content may be the content of the nitrogen-containing titanium in the range of 0.5 to the nitrogen-molybdenum layer resistance of the second solid tungsten-containing nitrogen-nitriding layer. Several layers of chip resistors, and the 201 is depleted. The stacked structure 2 12 has a P-type 3 sand layer. Crystalline -18- 200828425 (Si^Gex) layer, wherein the lanthanide is in the range of about o.oi to κο, or may also include a telluride layer. The telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), Ming (Co), titanium (Ti) 'tungsten (W), giant (amplitude &), and (11 〇, zirconium (2 〇 and platinum) (?1) One of the group consisting of. The second conductive layer 213 includes a tungsten layer. One of a PVD method, a CVD method, and an ALD method is performed to form a tungsten layer having a thickness of about ιοο to 2000 A. The PVD method includes A sputtering deposition method using a tungsten sputtering target is used. The intermediate structure 212 includes a titanium telluride (TiSix) layer 21 2A, a nitrogen-containing titanium (TiN layer 212B, a nitrogen-containing tungsten (WN layer 212C, and a nitrogen-containing tungsten germanium). WShNO layer 212D. According to the intermediate structures 22 and 202 described in the first and second embodiments, in addition to the titanium germanide layer, the nitrogen-containing titanium layer and the nitrogen-containing tungsten layer, they may be separately formed. a telluride layer, a nitrogen-containing tantalum layer, and a nitrogen-containing titanium tungsten layer. Further, in addition to the nitrogen-containing tungsten germanide layer, a nitrogen-containing titanium telluride layer or a nitrogen-containing giant telluride layer may be formed. The gate stack structure of the example is a structure caused by performing an annealing treatment on the gate stack structure according to the first and second embodiments of the present invention. The heat treatment accompanying the various processes (e.g., spacer formation and inner insulating layer formation) performed after forming the gate stack structure. Referring to Figures 3A and 3D to compare the intermediate structure 2 1 2 with The intermediate structure 22. When the titanium layer 22A reacts with the polysilicon from the first conductive layer 21, a titanium germanide layer 212A having a thickness of about 1 A to 30 A is formed. The ratio of germanium to titanium in the titanium germanide layer 212A is Between about 0.5 and 3.0. When nitrogen is supplied from the nitrogen-containing tungsten layer 22B to the titanium layer 22A, the -19-200828425 nitrogen-containing titanium layer 212B is formed. The thickness of the nitrogen-containing titanium layer 212B is about 10A. To 100A and having a ratio of nitrogen to titanium in the range of about 0.7 to 1.3. The ratio of nitrogen to titanium in the nitrogen-containing titanium layer 2 1 2B is about from the ratio of nitrogen to titanium in the titanium layer 22A. 0 is increased to about 0.7 to 1.3.

After the annealing, the nitrogen-containing tungsten layer 212C has a nitrogen content reduced to about 10% or less due to the denudation. The component symbol WN "D) represents the etched nitrogen-containing tungsten layer. The nitrogen-containing tungsten layer 21 2C is about 20 A to 200 Å thick. The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 212C is about 0.01. φ is in the range of 0.15. The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 2 1 2C is from about 0.3 compared to the ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 22C described in FIG. 3A. The range of distance between 1.5 and 1.5 is reduced to a range of about 0.01 to 0.15. The nitrogen-containing tungsten telluride layer 2 1 2D has substantially the same thickness and composition as the nitrogen-containing tungsten telluride layer 22C. The nitrogen-containing tungsten telluride layer 2 12D has a ratio of germanium to tungsten in the range of about 0.5 to 3.0 and a nitrogen content in the range of between about 10% and 60%. The thickness of the nitrogen-containing tungsten germanide layer 2 1 2D It is in the range between about 20 A and 200 A. φ Referring to Figures 3D and 3C to compare the intermediate structure 212 with the intermediate structure 202. During the annealing process, nitrogen is supplied from the nitrogen-containing tungsten layer 202B to the nitrogen-containing layer. Titanium layer 202A. As a result, the nitrogen-containing titanium layer 202A is transformed into a nitrogen-containing titanium layer 212B having minimal reaction with the titanium germanide layer 212A. The thickness of the titanium germanide layer 212A The ratio is in the range of about 1A to 30A, and the thickness of the nitrogen-containing titanium layer 212B is in the range of about 10A to 100A. The ratio of nitrogen to titanium in the nitrogen-containing titanium layer 212B is between about 0.7 and 1.3. The ratio of nitrogen to titanium in the nitrogen-containing titanium layer 21 2B is increased from about 0.2 to 0.8 -20 - . 200828425 to A range of between about 7 and 1 · 3. After the annealing, the nitrogen-containing tungsten layer 212C has a nitrogen content of about 10% or less due to erosion. The nitrogen-containing tungsten layer 212C is about 20A to 200A thick. The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 212C is in the range of about 0.01 to 0.15. Compared with the nitrogen to tungsten in the nitrogen-containing tungsten layer 202C described in FIG. 3C In proportion, the ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 212C is reduced from a range of between about 0.3 and 1.5 to a range of from about 〇1 to about 0.15. The nitrogen-containing tungsten telluride layer 2 1 2D has a rough Same as the thickness and composition of the nitrogen-containing tungsten-rhenium compound layer 202C. In detail, the nitrogen-containing tungsten germanide layer 21 2D has a ratio of germanium to tungsten of about 0.5 to 3.0 and about 10%. a nitrogen content in the range of 60%. The thickness of the nitrogen-containing tungsten carbide layer 2 1 2D is in a range between about 20 A and 200 A. The gate stack structure according to the third embodiment includes a first intermediate structure and a second The intermediate structure includes a first metal telluride layer and a first nitrogen-containing metal layer, and the second intermediate structure includes a second nitrogen-containing metal layer and a second nitrogen-containing metal telluride layer. For example, the first intermediate structure is formed by stacking the titanium germanide 0 layer 21 2A and the nitrogen-containing titanium layer 212B. The second intermediate structure is formed by stacking the nitrogen-containing tungsten layer 212C and the nitrogen-containing tungsten germanide layer 212D. Figure 3E depicts an image of the gate stack structure after the annealing process. The component symbols as described in the first to third embodiments represent the same elements. Therefore, the detailed description thereof will be omitted. Fig. 4A depicts a gate stack structure in accordance with a fourth embodiment of the present invention. The gate stack structure includes a first conductive layer 31, an intermediate structure 32, and a second conductive layer 33. The first conductive layer 31 includes a polycrystalline sand layer highly doped with a p-type impurity -21 - 200828425 (for example, boron) or an N-type impurity (for example, phosphorus). The first conductive layer 31 may also include a polysilicon layer (Si i.xGex, where X is in the range between about 〇1 and 1.0) or a germanide layer. For example, the telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), tantalum (Ta), (Hf), germanium (Zr), and platinum. One of the groups consisting of (Pt). The second conductive layer 33 includes a tungsten layer. The tungsten layer is about 100A to 2000A thick and is formed by performing a PVD method, a (: method, or a person: 10 method). The method includes a sputtering deposition method using a tungsten sputtering target. The titanium layer 32A and the nitrogen-containing tungsten germanide (WSuNy) layer 32B are included. In detail, the thickness of the titanium layer 32A is in the range of about 1 Torr to about 80 A. Preferably, the titanium layer 32A has about A thickness of 10A to about 50 A. The titanium layer 32A is changed to TiN by some subsequent upper portion by WSixNy deposition to form a nitrogen-containing tungsten germanide layer 32B, and some of its lower portion and the first conductive layer 31 The reaction, that is, the polysilicon layer thus forms a TiSh layer, and thus has a thickness as defined above. If the thickness of the titanium layer 3 2 A is large, the thickness of the TiSix layer also increases due to its volume expansion. In addition, if the thickness of the titanium layer 3 2 A is large, the titanium layer 3 2 A can absorb dopants, for example, phosphorus or boron of the polycrystalline germanium layer 31 and thus multiple depletion in the polycrystalline germanium layer 31. Deteriorating the performance of the device. The nitrogen-containing tungsten carbide layer 32B has a ratio of germanium to tungsten in the range of 0.5 to 3.0 and There is a nitrogen content of about 10% to 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may not succeed as a diffusion barrier due to the nitrogen-containing tungsten carbide layer 32B. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten carbide layer 32B may be high, and thus the contact resistance becomes high, resulting in deterioration of element properties. The telluride layer 32B represents a tungsten-22-200828425 germanium nitride layer or a tungsten germanide layer containing nitrogen in a certain amount/weight ratio. The nitrogen-containing tungsten germanide layer 32B is formed to have a thickness of about 20A to 200A. The titanium layer 32A is formed by a PVD method, a CVD method, or an ALD method. The nitrogen-containing tungsten germanide layer 32B is formed by a PVD method. The PVD method is performed by a sputtering deposition method or a reactive sputtering deposition method, for example, by The titanium sputtering target is subjected to a sputtering deposition method to form the titanium layer 3 2 A. The nitrogen-containing tungsten germanide layer 32B is formed by performing a reactive sputtering deposition method with a tungsten telluride sputtering target in a nitrogen atmosphere. In particular, since the nitrogen-containing tungsten germanide layer 32B φ can be uniformly formed and The pattern is independent, so the PVD method (for example, reactive sputtering deposition method) is used to form the nitrogen-containing tungsten germanide layer 32B. The gate stack structure according to the fourth embodiment of the present invention includes the first conductive layer 31. The Ti/WSlxNy intermediate structure 32 and the second conductive layer 33. The first conductive layer 31 includes polysilicon and the second conductive layer 33 includes tungsten, thereby forming a tungsten polysilicon gate stack structure. In particular, the Ti/ The WSixNy intermediate structure 32 includes a metal layer and a nitrogen-containing metal telluride layer. The metal layer includes a pure metal layer and the metal telluride layer φ includes a nitrogen-containing tungsten germanide layer. For example, the metal layer is the titanium layer 32A and the metal halide layer is the nitrogen-containing tungsten germanide layer 3 2 B. The multilayer intermediate structure according to the fourth embodiment can also be formed in other structures. The metal layer includes a tantalum layer in addition to the titanium layer, and the nitrogen-containing niobium metal telluride layer includes a nitrogen-containing titanium telluride (TiShNy) layer or a nitrogen-containing molybdenum telluride in addition to the nitrogen-containing tungsten telluride layer. (TaShNy) layer. The set of layers is formed by a PVD method including a sputtering deposition method, a CVD method, or an ALD method. The nitrogen-containing titanium telluride layer is formed by performing a reactive sputtering deposition method using a titanium telluride sputtering target in a nitrogen atmosphere. The nitrogen-containing group telluride layer was formed by performing a reactive sputtering deposition method using a molybdenum telluride sputtering target -23-200828425 in a nitrogen atmosphere. The molybdenum layer is about 10 Å to 80 Å thick. Preferably, the macrolayer has a thickness of from about i〇A to about 50A. The germanium layer changes some of its upper portion to TaN by subsequent WSHNy deposition to form a metal telluride layer, and some of its lower portion reacts with the first conductive layer 31, that is, the polysilicon layer thus forms TaSu The layer has a thickness as defined above. If the thickness of the giant layer is large, the thickness of the TaSu layer also increases due to its volume expansion. In addition, if the thickness of the giant layer is large, the group of layers can absorb the dopant of the polycrystalline germanium layer 31, such as phosphorus or boron, so multiple depletion occurs in the polycrystalline layer 31, resulting in components. Deterioration in performance. Each of the nitrogen-containing titanium telluride layer and the nitrogen-containing molybdenum telluride layer has a thickness of from about 20 A to about 200 A and each layer has a nitrogen content of from about 1% to about 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may not occur as a diffusion barrier due to the nitrogen-containing titanium or telluride layer. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing titanium or molybdenum telluride layer may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The ratio of niobium to titanium in the nitrogen-containing titanium niobium φ layer is in the range of about 0.5 to 3.0. The nitrogen-containing giant telluride layer has a rhodium to macro ratio of from about 0.5 to about 3.0. Fig. 4B shows a gate stack structure in accordance with a fifth embodiment of the present invention. The gate stack structure is modified from the gate stack structure according to the second embodiment. In other words, a titanium-containing titanium (TiNx) layer is used in place of titanium, where X is less than about 1. The gate stack structure includes a first conductive layer 301, an intermediate structure 302, and a second conductive layer 303. The first conductive layer 301 includes a polysilicon layer highly doped with a p-type impurity (e.g., boron) or an N-type impurity (e.g., phosphorus). The -24-200828425 - conductive layer 301 may also include a polycrystalline germanium layer (Si!-xGex, where x is in the range of between about 0.01 and 1.0) or a germanide layer. For example, the telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), molybdenum (Ta), hafnium (Hf), zirconium (2, and platinum). One of the group consisting of: The second conductive layer 303 comprises a tungsten layer. The PVD method, the CVD method or the ALD method is performed to form a tungsten layer of about 100 A to 2000 A thick. The PVD method includes use. Sputter deposition method of a tungsten sputtering target. The intermediate structure 302 includes a nitrogen-containing titanium (TiN〇 layer 302A and a nitrogen-containing tungsten germanium Φ compound (WSixNy) layer 302B. The nitrogen-containing titanium layer 302A has a range of about 0.2 to 0.8. The ratio of nitrogen to titanium and the thickness of about 10 A to 150 A. Here, the nitrogen-containing metal layer, that is, the nitrogen-containing titanium layer 302A has a ratio of nitrogen to titanium as described above to prevent SiN self. This TiNx layer 302A is produced. During the subsequent annealing process, the generation of SiN is prevented by the excessive Ti in the TiNx layer 302A being destroyed by the Sl_N bond formed between the polysilicon and TiNx, and thus the SiN is removed. This is possible because the TiN bonding is much stronger than the SiN bonding. The nitrogen-containing titanium layer 302A represents a titanium nitride layer or a nitrogen-containing titanium layer. In this example, in this example The nitrogen-containing titanium layer has a metal characteristic. The nitrogen-containing tungsten telluride layer 302B has a rhodium to tungsten ratio of from 0.5 to 3.0 and a nitrogen content of from about 10% to about 60%. Here, the nitrogen content is in the above manner If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing tungsten carbide layer 302B cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, it is included in the nitrogen-containing tungsten. The SiN content in the telluride layer 302B may be high, and thus the contact resistance becomes high, resulting in deterioration of device performance. The nitrogen-containing tungsten germanide layer 3 0 2 B represents a tungsten sand nitride layer or contains a certain content / a tungsten carbide layer having a weight ratio of nitrogen. -25- 200828425 The nitrogen-containing titanium layer 302A and the nitrogen-containing tungsten telluride layer 302B are formed by a PVD method. The PVD method is deposited by sputtering deposition or reactive sputtering. The method is carried out. For example, the nitrogen-containing titanium layer 302A is formed by performing a reactive plating deposition method in a nitrogen atmosphere, and the sputtering method is performed by sputtering with a tungsten telluride in a nitrogen atmosphere. Plating deposition method to form the nitrogen-containing tungsten carbide layer 302B. Because the PVD method allows The nitrogen-containing tungsten carbide layer 302B is uniformly formed regardless of the underlying type, so the PVD ^ (for example, the above-described reaction φ-type sputtering deposition method) is used to form the nitrogen-containing tungsten germanide layer 302B. The gate stack structure includes the first conductive layer 301, the TiNx/WSixNy intermediate structure 302, and the second conductive layer 303. The first conductive layer 301 and the second conductive layer 303 respectively include a polysilicon layer and a tungsten layer. Therefore, a tungsten polysilicon gate stack structure is provided. In particular, the TiNx/WSixNy intermediate structure comprises a metal layer and a nitrogen-containing metal telluride layer. The metal layer includes a metal layer containing nitrogen in a certain content/weight ratio, and the metal germanide layer includes a gold bismuth telluride layer containing nitrogen in a certain content/weight ratio. For example, the metal layer includes the nitrogen-containing titanium layer 302A, and the metal halide layer includes the nitrogen-containing tungsten germanide layer 30B. The multilayer intermediate structure according to the fifth embodiment can be formed in other different structures. The nitrogen-containing metal layer includes a nitrogen-containing group (TaNx) layer in addition to the nitrogen-containing titanium layer. The nitrogen-containing metal telluride layer includes a nitrogen-containing titanium telluride (TiSixNy) layer or a nitrogen-containing giant telluride (TaShNy) layer in addition to the nitrogen-containing tungsten germanide (WSixNy) layer. The nitrogen-containing tantalum layer is formed by a PVD method including a sputtering deposition method, a CVD method, or an ALD method. The nitrogen-containing titanium telluride layer was formed by performing a reactive sputtering deposition method using a titanium telluride sputtering target in a nitrogen atmosphere. The nitrogen-containing group telluride layer was formed by performing a reactive sputtering deposition method using a molybdenum telluride sputtering target in a nitrogen atmosphere of -26-, 200828425. The nitrogen-containing macrolayer has a thickness ranging from about 10A to 80A. Each of the nitrogen-containing titanium telluride layer and the nitrogen-containing telluride layer has a thickness of about 20A to 200A, and each layer has a nitrogen content of about 1% to 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may not occur as a diffusion barrier due to the nitrogen-containing titanium or group telluride layer. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing titanium or molybdenum telluride layer may be high, and φ thus causes the contact resistance to become high, resulting in deterioration of element performance. The ratio of niobium to titanium in the nitrogen-containing titanium telluride layer is in the range of between about 0.5 and 3.0. The nitrogen-containing macrochemical layer has a ratio of cerium to molybdenum ranging from about 0.5 to 3.0. Fig. 4C depicts a gate stack structure in accordance with a sixth embodiment of the present invention. The gate stack structure includes a first conductive layer 31, an intermediate structure 312, and a second conductive layer 313. The first conductive layer 311 includes a polysilicon layer highly doped with a P-type impurity (e.g., boron (6)) or an N-type impurity (e.g., phosphorus (P)). The first conductive layer 31 1 may include a polycrystalline germanium layer φ (Sil〃Ge〇, wherein the X system is in a range between about 〇1 and 1.0, or may include a vaporized layer) in addition to the polysilicon layer. The telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (C 〇, cobalt (Co), titanium (Ti), tungsten (,), giant (1 &), (11 〇, zirconium (2: 〇 and One of the group consisting of: The second conductive layer 313 includes a tungsten layer. The tungsten layer is formed by one of a PVD method, a CVD method, and an ALD method to form a tungsten layer of about ιοοΑ to 2000 Å. The PVD method includes a sputtering deposition method using a tungsten sputtering target. The intermediate structure 312 includes a titanium telluride (TiSix) layer 31 2A, a titanium-containing titanium (TiNx) layer 3 12B, and a nitrogen-containing tungsten germanide (wSixN germanium layer 312C). The intermediate structure may be formed in other different structures according to the materials described in the fourth and fifth embodiments according to -27-200828425. The gate stack structure according to the sixth embodiment is in accordance with the fourth and fifth aspects of the present invention. The gate stack structure of the embodiment is subjected to an annealing process, and the annealing is performed after forming the gate stack structure. Heat treatment accompanying various processes (for example, spacer formation and formation of an inner insulating layer). The formation of the nitrogen-containing tungsten telluride layer 3 2 B over the titanium layer 3 2 A is referred to in the reference 0 (Fig. 4A) After the annealing, a trace amount of nitrogen in the nitrogen-containing tungsten carbide layer 32B is decomposed in a boundary region between the tantalum layer 32 A and the nitrogen-containing tungsten carbide layer 32B. Therefore, as shown in FIG. 4C The titanium layer 32A is partially transformed into the nitrogen-containing titanium layer 312B, and the lower portion of the titanium layer 32A is reacted with the polysilicon from the first conductive layer 31 to form the titanium germanide layer 312A. The thickness of the telluride layer 312A is in the range of between about 1A and 30A, and wherein the ratio of germanium to titanium is in the range of between about 0.5 and 3.0.

The nitrogen-containing titanium layer 3 12B is about 10A to 100A thick and has a nitrogen to titanium ratio in the range between about 0.7 and 1.3 W. The nitrogen-containing tungsten germanide layer 312C has substantially the same thickness and composition as the nitrogen-containing tungsten carbide layer 32B. In detail, the nitrogen-containing tungsten telluride layer 3 12C has a rhodium to tungsten ratio of about 0.5 to 3.0 and a nitrogen content ranging between about 10% and 60%. The thickness of the nitrogen-containing tungsten telluride layer 3 1 2C is in the range of between about 20A and 200A. Referring to Figures 4C and 4B, the intermediate structure 3 1 2 and the intermediate structure 302 are compared. During the annealing process, nitrogen is supplied from the nitrogen-containing tungsten germanide layer 302B to the nitrogen-containing titanium layer 302A, thereby converting the nitrogen-containing titanium layer 302A to a minimum with the titanium germanide layer 312A. The nitrogen-containing titanium layer 312B is reacted. The thickness of the titanium telluride layer 3 12A is in the range of about 1 A to 30 A, and the thickness of the nitrogen-containing titanium layer 312B is in the range of about 10 A to 100 A. The ratio of nitrogen to titanium in the nitrogen-containing titanium layer 312B is in the range of about 0.7 to 1.3. The nitrogen to titanium ratio in the nitrogen-containing titanium layer 312B is increased from about 0.2 to 0.8 to about 0.7 and 1.3 as compared to the nitrogen to titanium ratio in the nitrogen-containing titanium layer 302B (see Figure 4C). The scope. φ The nitrogen-containing tungsten telluride layer 3 1 2C has substantially the same thickness and composition as the nitrogen-containing tungsten germanide layer 302 C. In detail, the nitrogen-containing tungsten telluride layer 3 12C has a rhodium to tungsten ratio of about 0.5 to 3.0 and a nitrogen content ranging between about 10% and 60%. The thickness of the nitrogen-containing tungsten carbide layer 3 1 2C is in the range of between about 20A and 200A. The gate stack structure according to the sixth embodiment includes a first intermediate structure and a second intermediate structure. The first intermediate structure includes a metal telluride layer and a nitrogen-containing metal layer ' and the second intermediate structure includes a nitrogen-containing metal telluride layer. For example, φ is formed by stacking the titanium germanide layer 3 1 2A and the nitrogen-containing titanium layer 3 1 2B. The second intermediate structure includes the nitrogen-containing tungsten telluride layer 3 12C. Fig. 5A depicts a gate stack structure in accordance with a seventh embodiment of the present invention. The gate stack structure includes a first conductive layer 41, an intermediate structure 42, and a second conductive layer 43. The first conductive layer 41 includes a polysilicon layer highly doped with a P-type impurity (e.g., boron) or an N-type impurity (e.g., phosphorus). The first conductive layer 41 may also include a polycrystalline germanium layer (Sll.xGex, where X is in the range between about 〇1 and 1.0) or a germanide layer. For example, the telluride layer includes: 29-.200828425 free nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), giant (Ta), give (Hf), pin One of a group consisting of (Zr) and platinum (Pt). The second conductive layer 43 includes a tungsten layer. The tungsten layer is about 100 A to 2000 A thick and is formed by performing a PVD method, a CVD method, or an ALD method. The PVD method includes a sputtering deposition method using a tungsten sputtering target. The intermediate structure 42 includes a titanium layer 42A, a nitrogen-containing tungsten germanide (WSixNy) layer 42B, and a nitrogen-containing tungsten (WNX) layer 42C. In detail, the thickness of the titanium layer 42A is in the range of about 10 A to about 80 A. Preferably, the titanium layer 4 2 A has a thickness of from about 1 〇Α to about 50 Å. The titanium layer 42A changes some of its upper portion to TiN by subsequent WNX deposition to form a nitrogen-containing tungsten layer 42C, and some of its lower portion reacts with the first conductive layer 41, that is, the polysilicon layer Since the TiSh layer is formed, it has a thickness as defined above. If the thickness of the titanium layer 42A is large, the thickness of the TiSh layer also increases due to its volume expansion. In addition, if the thickness of the titanium layer 42A is large, the titanium layer 42A can absorb the dopant of the polysilicon layer 41, such as phosphorus or boron, so that multiple depletion occurs in the polysilicon layer 41, resulting in deterioration of device performance. . The nitrogen-containing tungsten ruthenium compound layer 42B has a rhodium to tungsten ratio of about 0.5 to 3.0 and a nitrogen content of about 10% to 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing tungsten carbide layer 42B cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten carbide layer 42B may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The nitrogen-containing tungsten germanide layer 42B represents a tungsten germanium nitride layer or a tungsten germanide layer containing a certain content/weight ratio of nitrogen. The nitrogen-containing tungsten carbide layer 42B is formed to have a thickness of about 20A to 200A. -30至 200828425 The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 42C is in the range of between about 0.3 and 1.5. The nitrogen-containing tungsten layer 42C represents a tungsten nitride layer or a tungsten layer containing nitrogen in a certain content/weight ratio. The thickness of the nitrogen-containing tungsten layer 42C is in the range of about 20A to 200A. Although will be described later, it is known that the nitrogen-containing tungsten layer 42C supplies nitrogen to the nitrogen-containing tungsten carbide layer 42B. Therefore, after the annealing, the nitrogen-containing tungsten layer 42C becomes a pure tungsten layer having no nitrogen or a tungsten layer containing a trace of nitrogen. The titanium layer 42A φ and the nitrogen-containing tungsten layer 42C are formed by performing a PVD method, a CVD method, or an ALD method. The nitrogen-containing tungsten germanide layer 42B is formed by performing a PVD method. The PV D method is carried out by a sputtering deposition method or a reactive sputtering deposition method. For example, the titanium layer 42A is formed by performing a sputtering deposition method using a titanium sputtering target. The nitrogen-containing tungsten layer 4 2 C was formed by performing a reactive sputtering deposition method with a tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing tungsten sand layer 42B is formed by performing a reactive sputtering deposition method with a tungsten telluride sputtering target in a nitrogen atmosphere. In particular, since the above-described I-reactive sputtering deposition method is performed with the tungsten telluride sputtering target in the nitrogen atmosphere to allow uniform formation of the nitrogen-containing tungsten carbide layer 42B regardless of the underlying type, the PVD is used. The nitrogen-containing tungsten telluride layer 42B is formed by a method such as reactive sputtering deposition. A gate stack structure according to a seventh embodiment of the present invention includes the first conductive layer 41, the Ti/WSixNy/WNx intermediate structure 42 and the second conductive layer 43. The first conductive layer 41 includes polysilicon and the second conductive layer 43 includes tungsten, thereby forming a tungsten polysilicon gate stack structure. In particular, the Ti/WShNy/WNx intermediate structure includes a first metal layer, a nitrogen-containing metal telluride layer, and a second metal layer. The first metal layer comprises a layer of pure gold -31 - 200828425. The second metal layer includes a nitrogen-containing metal layer. The metal telluride layer includes a nitrogen-containing metal telluride layer. For example, the first metal layer is the titanium layer 42A. The second metal layer is the nitrogen-containing tungsten layer 42C. The metal telluride layer is the nitrogen-containing tungsten germanide layer 42B. The multilayer intermediate structure according to the seventh embodiment can also be formed in other structures. The first metal layer includes a giant layer in addition to the titanium layer. The second metal layer includes a nitrogen-containing titanium tungsten (TiWNO layer in addition to the nitrogen-containing tungsten layer. The metal sand layer includes a nitrogen-containing titanium-titanium (TiSixNy) layer in addition to the nitrogen-containing crane sand layer. Or a nitrogen-containing giant telluride (TaShNy) layer. The molybdenum layer is formed by a PVD method including a sputtering deposition method, a CVD method, or an ALD method. Reactive sputtering is performed by a titanium-tungsten sputtering target in a nitrogen atmosphere. Forming the nitrogen-containing titanium tungsten layer. The nitrogen-containing titanium telluride layer is formed by reactive sputtering deposition using a titanium telluride sputtering target in a nitrogen atmosphere by sputtering with a giant telluride in a nitrogen atmosphere. The target is subjected to reactive sputtering deposition to form the nitrogen-containing telluride layer. The molybdenum layer is about 10 to 80 people thick. Preferably, the molybdenum layer has a thickness of from about 10 to about 50 A. Some of the upper portions thereof are changed to TaN by subsequent wSixNy deposition to form a metal telluride layer, and some of the lower portions thereof react with the first conductive layer 41, that is, the polysilicon layer thus forms a TaSix layer, thus having The thickness as defined above. If the thickness of the layer is large, the TaSix The thickness also increases due to its volume expansion. Further, if the thickness of the molybdenum layer is large, the macro layer can absorb dopants, for example, phosphorus or boron of the polysilicon layer 41 and thus the polysilicon layer 4 Multiple depletion occurs in 1 , resulting in deterioration of device performance. Each layer of the nitrogen-containing titanium tungsten layer and the nitrogen-containing giant telluride layer is formed to a thickness of about 20A to 200A and each layer has about 10% to 60% nitrogen. Here, the nitrogen content is appropriately adjusted in the above manner by -32-200828425. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing telluride layer cannot be successfully used as a diffusion barrier. If the nitrogen content is too high, the SiN content contained in the nitrogen-containing telluride layer may be high, and thus the contact resistance becomes high, resulting in deterioration of element properties. The nitrogen-containing titanium tungsten layer has about 0.5 and 3 The ratio of titanium to tungsten in the range of 0. The ratio of germanium to titanium in the nitrogen-containing titanium telluride layer is in the range of about 0.5 to 3.0. The nitrogen-containing giant telluride layer has about 0.5 to 3.0. The ratio of titanium to titanium. Figure 5B depicts the invention in accordance with the invention. A gate stack φ structure of the eighth embodiment. The gate stack structure includes a first conductive layer 410, an intermediate structure 420, and a second conductive layer 403. The first conductive layer 401 includes a highly doped P- a polysilicon layer of a type of impurity (for example, boron) or an N-type impurity (for example, phosphorus). The first conductive layer 401 may also include a polysilicon layer (Si^xGex, wherein the X system is between about 0.01 and 1.0). Within the range) or a telluride layer. For example, the telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (C〇, cobalt (Co), titanium (Ti), tungsten (W), molybdenum (Ta), One of a group consisting of Hf), zirconium (Zr) and platinum (Pt). The second conductive layer 403 includes a tungsten layer. The tungsten layer is formed to have a thickness of about 1 Å to φ 2000 A and is formed by a PVD method, a CVD method, or an ALD method. The PVD method includes a sputtering deposition method using a tungsten concentrated plating target.

The intermediate structure 402 includes a titanium-containing titanium (TiNx) layer 402A, a nitrogen-containing tungsten germanide (WSixNO layer 402B, and a nitrogen-containing tungsten (WNX) layer 402C. More specifically, the nitrogen-containing titanium layer 4 0 2 A nitrogen There is a certain proportion of titanium (for example: in about 〇.  2 to 0. Within the scope of 8). Here, the nitrogen-containing metal layer, that is, the nitrogen-containing titanium layer 402A, has a ratio of nitrogen to titanium as described above to prevent SiN from being generated in the nitrogen-containing titanium layer 402A. Since excessive Ti in the nitrogen-containing titanium layer 402A during the subsequent annealing treatment destroys the Si-N-33-200828425 bond formed between the polycrystalline germanium and TiNx and thus removes SiN, the generation of SiN can be prevented. This is because the ΤιΝ link is more robust than the SiN link. The nitrogen-containing titanium layer 402A is formed to have a thickness of about 10A to 150A. The nitrogen-containing titanium layer 402A also includes a titanium nitride layer. The ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 402B is about 0. 5 and 3. The nitrogen content of the nitrogen-containing tungsten carbide layer 402B is in the range of about 10% to 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing tungsten carbide layer φ 402B cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten carbide layer 402B may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The nitrogen-containing tungsten germanide layer 402B also includes a tungsten germanium nitride layer or a tungsten germanide layer containing a certain content/weight ratio of nitrogen. The nitrogen-containing tungsten layer 402C has a certain ratio of nitrogen to tungsten (for example, at about 0. 3 to 1. Within the scope of 5). The nitrogen-containing tungsten layer 402C represents a tungsten nitride layer or a tungsten layer containing a certain content/weight ratio of nitrogen. Although described later, it is known that the nitrogen-containing tungsten layer 4〇2C supplies nitrogen to the nitrogen-containing tungsten silicide layer 4Q2B. The nitrogen-containing tungsten layer 402C is formed to have a thickness of about 20A to 200A. Due to the supply of nitrogen, the nitrogen-containing tungsten layer 402C becomes a pure tungsten layer or a tungsten-containing tungsten layer after the annealing. The nitrogen-containing tungsten layer 402C is formed by performing a PVD method, a CVD method, or an ALD method. The nitrogen-containing titanium layer 4〇2A and the nitrogen-containing tungsten germanide layer 402B are formed by a PVD method. The PVD method is carried out by sputtering deposition or a reactive sputtering deposition method. For example, the nitrogen-containing titanium layer 402 A is formed by performing a sputtering deposition method using a titanium sputtering target in a nitrogen atmosphere. The nitrogen-containing tungsten layer 402C is formed by performing a reactive sputtering deposition method with a tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing tungsten telluride layer 402B is formed by performing a reactive sputtering deposition method with a tungsten telluride sputtering target in a nitrogen atmosphere. In particular, since the nitrogen-containing tungsten carbide layer 402B can be uniformly formed regardless of the underlying type, the nitrogen-containing tungsten germanide layer 402B is formed using the PVD method (e.g., reactive sputtering deposition method). A gate stack structure according to an eighth embodiment of the present invention includes the first conductive layer 401, the TiNx/WSixNy/WNx intermediate structure 402, and the second conductive layer 403. The first conductive layer 401 includes polysilicon and the second conductive layer 403 includes tungsten, thereby forming a tungsten polysilicon gate stack structure. Specifically, the TiNx/WSixNy/WNx intermediate structure 402 is formed in a stacked structure including a first metal layer, a nitrogen-containing metal telluride layer, and a second metal layer. The first and second metal layers are nitrogen-containing metal layers, and the metal telluride layer is a nitrogen-containing metal halide layer. For example, the first metal layer is the nitrogen-containing titanium layer 402 A. The second metal layer is a nitrogen-containing tungsten layer 402C. The metal telluride layer is a nitrogen-containing tungsten telluride layer 4 0 2 B. φ The above-described multilayer intermediate structure can be formed by other different structures. For example, the first nitrogen-containing metal layer includes a nitrogen-containing macrolayer in addition to the nitrogen-containing titanium layer. The second nitrogen-containing metal layer includes a nitrogen-containing titanium tungsten layer in addition to the nitrogen-containing tungsten layer. The nitrogen-containing metal telluride layer includes a nitrogen-containing titanium telluride layer or a nitrogen-containing giant telluride layer in addition to the nitrogen-containing tungsten germanide layer. The nitrogen-containing macrolayer is formed by performing a PVD method including sputtering, a CVD method, or an ALD method. The nitrogen-containing titanium tungsten layer is formed by performing a reactive sputtering deposition method in a gas atmosphere with a titanium tungsten sputtering target. The nitrogen-containing titanium telluride layer is formed by reactive sputtering deposition method with individual titanium telluride and giant telluride sputtering target in a nitrogen atmosphere. And the -35- 200828425 nitrogen-containing giant telluride layer. The nitrogen-containing macrolayer is formed to have a thickness of about 1 ο A to 8 Ο A. Each of the nitrogen-containing titanium tungsten layer, the nitrogen-containing titanium telluride layer and the nitrogen-containing telluride layer has a thickness of 20A to 200A, and each layer has a nitrogen range of between about 10% and 60%. content. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing titanium or molybdenum telluride layer cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen contains a heavy cylinder, the SiN content contained in the gas-containing or group-containing sand layer may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. In the φ nitrogen-containing titanium tungsten layer, the ratio of titanium to tungsten is about 0.5 to 3. Within the range of 0. In the nitrogen-containing titanium telluride layer, the ratio of niobium to titanium is about 0. 5 to 3. Within the range of 0. In the nitrogen-containing bismuth compound layer, the ratio of political confrontation is about 0. 5 to 3. In the range of 0. Fig. 5C depicts a gate stack structure in accordance with a ninth embodiment of the present invention. The gate stack structure includes a first conductive layer 41 1 , an intermediate structure 4 1 2 , and a second conductive layer 413 . The first conductive layer 411 includes a polycrystalline, chopped layer highly doped with a P-type impurity (e.g., boron (B)) or a type 1 impurity (e.g., pity (?)). The first conductive layer 4 1 1 may include, in addition to the polysilicon layer, a polycrystalline germanium (Si nG ex) layer, wherein the X system is in the range of about 〇·〇ΐ and ι·〇, or includes a germanide layer. . The telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), molybdenum (Ta), (Hf), zirconium (Zr), and platinum (Pt). One of the groups formed. The second conductive layer 413 includes a tungsten layer. One of the PVD method, the CVD method, and the ALD method is performed to form a tungsten layer of about 100 A to 200 Å thick. The PVD method includes a sputtering deposition method using a tungsten sputtering target. The intermediate structure 4 1 2 includes a chic compound (T i S i X) layer 4 1 2 A, a nitrogen-containing -36-200828425 (TiNx) layer 412B, a nitrogen-containing tungsten germanide (wShNy) layer 412C, and a nitrogen-containing tungsten (WNX) layer 41 2D. The intermediate structure 4丨2 may be formed in a different structure according to the selection materials described in the seventh and eighth embodiments of the present invention. The gate stack structure according to the ninth embodiment is a structure resulting from annealing treatment of the gate stack structures according to the seventh and eighth embodiments of the present invention. The annealing includes subsequent heat treatment during various processes (e.g., spacer formation and formation of the inner insulating layer) performed after forming the gate stack structures. * Refer to Figures 5C and 5A to compare the intermediate structure 412 with the intermediate structure 42. When the titanium layer 42A reacts with the polysilicon from the first conductive layer 41, a titanium germanide layer 41 2A having a thickness of about 1 A to 30 A is formed. The ratio of germanium to titanium in the titanium germanide layer 21 2A is about 0. 5 and 3. Within the range of 0. When nitrogen is supplied from the nitrogen-containing tungsten layer 42B to the titanium layer 42A, the nitrogen-containing titanium layer 412B is caused. The nitrogen-containing titanium layer 412B has a thickness in the range of about 10A to 100A and has a thickness of about 0. 7 to 1. The ratio of nitrogen to titanium in the range of 3. The ratio of nitrogen to titanium in the nitrogen-containing titanium layer 4 1 2B is increased from about 0 to about 0, as compared to the ratio of nitrogen to titanium in the titanium layer 42A. 7 to 1. 3. The nitrogen-containing tungsten carbide layer 4 1 2C has substantially the same thickness and composition as the nitrogen-containing tungsten carbide layer 42C. In detail, the nitrogen-containing tungsten telluride layer 4 12C has about 0. 5 to 3. The ratio of 0 to tungsten in the range of 0 and the nitrogen content in the range between about 10% and 60%. The thickness of the nitrogen-containing tungsten telluride layer 4 1 2C is in the range of between about 2 Å and 200 Å. After the annealing, the nitrogen-containing tungsten layer 412D has a nitrogen content which is reduced to about 10% or less due to the etching. The component symbol WNX(D) indicates the etched -37·200828425 nitrogen-containing tungsten layer. The nitrogen-containing tungsten layer 4 1 2 D is about 20 A to 2 〇 0 people thick. The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 4 1 2D is about 0. 0 1 and 0. Within the range of 1 to 5 rooms. The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 4 1 2D is from about 0. Compared with the ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 42C described in FIG. 5A. 3 and 1. The range of 5 is reduced to about 0. 01 to 0. The scope of 15. Forming the nitrogen-containing tungsten telluride layer 42; 6 over the titanium layer 42A (see FIG. 5A), after the annealing, in a boundary region between the titanium layer 42A and the nitrogen-containing tungsten germanide 42B A trace amount of nitrogen in the nitrogen-containing tungsten telluride layer 42B is decomposed. As a result, as described in FIG. 5C, the titanium layer 42A is partially transformed into the nitrogen-containing layer 412B, and a portion below the seed layer 42A is reacted with the polysilicon from the first conductive layer 41 to form the titanium. Telluride layer 412A. Referring to Figures 5C and 5B, the intermediate structure 4 1 2 and the intermediate structure 04 are compared. The nitrogen-containing titanium layer 402A is converted into a nitrogen-containing titanium layer 412B having a minimum reaction with the titanium germanide layer 41 2A. The thickness of the titanium telluride layer 41 2A is in the range of about 1A to 30A, and the thickness of the nitrogen-containing titanium layer 412B is in the range of about 10A to 100A. The ratio of nitrogen to titanium in the nitrogen-containing titanium layer 41 2B is about 0. 7 and 1. Within the range of 3 rooms. The nitrogen-containing tungsten germanide layer 412C has a thickness and composition substantially the same as the nitrogen-containing tungsten germanide layer 42B. More specifically, the ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 4 1 2C is about 0. 5 to 3. Within the range of 0. The nitrogen-containing tungsten carbide layer 412C has a nitrogen content in the range of about 10% to 60% and a thickness of about 20A to 200A. After the annealing, the nitrogen-containing tungsten layer 412D has a nitrogen content reduced to about 10% or less due to the apocution effect. The nitrogen-containing tungsten layer 412D is about 20 to 200 people -38 to 200828425 thick. The ratio of nitrogen to tungsten in the nitrogen-containing tungsten layer 4 1 2D is in the range between. The gate stack structure according to the ninth embodiment includes a second intermediate structure. The first intermediate structure includes a first gold first nitrogen-containing metal layer, and the second intermediate structure includes a layer and a nitrogen-containing metal telluride layer. For example, the first intermediate junction of the nitrogen-containing tungsten germanide layer 412C and the nitrogen-containing tungsten layer 412C φ structure is formed by stacking 412A and the nitrogen-containing titanium layer 412B. Fig. 6A depicts a tenth embodiment in accordance with the present invention. The gate stack structure includes a first conductive layer 51 and a second conductive layer 53. The polycrystalline sand layer 51 of the first conductive layer 51 comprising a highly doped (e.g., boron) or N-type impurity (e.g., phosphorus) may also include a polycrystalline germanium layer (Sn. xGex, which is within a range of 1 · 0) or a telluride layer. For example, the one selected from the group consisting of nickel (Ni), chromium (C 〇, cobalt (Co), titanium (Ti), tungsten φ lead (Hf) 'chromium (Zr), and platinum (Pt) The second conductive layer 53 includes a tungsten layer. The tungsten layer is about thick and comprises a sputtering deposition method using a tungsten sputtering target by performing a PVD method, a CVD method, or an ALD method. The intermediate structure 52 includes titanium (Ti a layer 52A, a first 52B, a nitrogen-containing tungsten germanide (WSuNO layer 52C, and a second layer 52D. In detail, the thickness of the titanium layer 52A is in the range of %. Preferably, the titanium layer 52A has about 10A. The titanium layer 52A is about 0 by the deposition of ffi by the subsequent WNX. 01 and 0. 15 an intermediate structure and a telluride layer and a second nitrogen-containing metal. The gate stack structure 3 and the impurity P-type impurity layer in the second are formed by stacking. The first lead X is at about 0. 0 1 ; the telluride layer includes (W), giant (Ta), and 〇 100A to 2000A: formed. The PVD nitrogen-containing tungsten (WNx) layer has two nitrogen-containing tungsten (WN0 3 10A to about 80A to about 50A thick ί, some of its upper portion -39-200828425 is changed to TiN to form the first nitrogen-containing tungsten layer 52B, and Some of the lower portions thereof react with the first conductive layer 51, that is, the polysilicon layer thus forms a Ti Six layer, and thus have a thickness as defined above. If the thickness of the titanium layer 52A is large, the thickness of the TiSlx layer Also, the ridge is increased because of its volume expansion. Further, if the thickness of the titanium layer 52A is large, the titanium layer 52A can absorb the dopant of the polysilicon layer 51, for example, phosphorus or boron, and thus in the polysilicon layer 51. Multiple depletion occurs, resulting in deterioration of device performance. The ratio of nitrogen to tungsten in each of the first and second nitrogen-containing tungsten layers 5 2B and 5 2D is about 0. 3 and 1. Within 5 φ range. Each of the first and second nitrogen-containing tungsten layers is regarded as a tungsten nitride layer or a tungsten layer containing a certain content/weight ratio of nitrogen. Although will be described later, it is known that the first and second nitrogen-containing tungsten layers 52B and 52D supply nitrogen to the nitrogen-containing tungsten carbide layer 52C. Each of the first and second nitrogen-containing tungsten layers 52B and 52D has a thickness of from about 20 to 200 Å. Since nitrogen is supplied to the nitrogen-containing tungsten carbide layer 52C, each of the first and second nitrogen-containing tungsten layers 5 2B and 5 2D becomes a pure tungsten layer or a tungsten-containing tungsten layer after the subsequent annealing treatment. The ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 5 2C is about φ 〇. 5 and 3. The nitrogen content of the nitrogen-containing tungsten carbide layer 52C is in the range of from about 10% to about 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing tungsten carbide layer 52C cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten carbide layer 52C may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The nitrogen-containing tungsten telluride layer 5 2C represents a tungsten germanium nitride layer or a tungsten germanide layer containing a certain content/weight ratio of nitrogen. The nitrogen-containing tungsten telluride layer 52C is formed to have a thickness in the range of from about 20 Å to about 200 Å. -40- . 200828425 The titanium layer 52A and the first and second nitrogen-containing tungsten layers 52B and 52D are formed by a PVD method, a CVD or an ALD method. The nitrogen-containing tungsten telluride layer 52C is formed by a PVD method. . The PVD method is carried out by sputtering deposition or a reactive sputtering deposition method. For example, the seed layer 52A is formed by performing a sputtering deposition method on a titanium sputtering target. The first and second nitrogen-containing tungsten layers 52B and 52D are formed by performing a reactive sputtering deposition method in a gas-heat environment with a sprinkler. The nitrogen-containing tungsten telluride layer 52C is formed by performing a reactive sputtering deposition method using a tungsten telluride sputtering target in a nitrogen atmosphere. In particular, since the nitrogen-containing tungsten carbide layer 502C can be uniformly formed regardless of the underlying type, the PVD method (for example, reactive sputtering deposition method) can be used to form the nitrogen-containing tungsten germanide layer 5 02C. The gate stack structure according to the tenth embodiment includes the first conductive layer 51, the Ti/WNx/WSixNy/WNx intermediate structure 52, and the second conductive layer 53. The first conductive layer 51 and the second conductive layer 53 respectively comprise a polysilicon layer and a tungsten layer, thereby forming a tungsten polysilicon gate stack structure. In particular, the Ti/WNx/WShNy/WNx intermediate structure 52 includes a first metal layer, a second metal layer, a nitrogen-containing metal telluride layer, and a third metal layer. The first metal layer comprises a layer of pure metal, whereas the second and third layers of metal comprise a layer of nitrogen-containing metal. The nitrogen-containing metal telluride layer includes a metal telluride layer containing nitrogen in a certain amount/weight ratio. For example, the first metal layer is the titanium layer 52A, and the second and third metal layers are the first and second nitrogen-containing tungsten layers 5 2B and 52D, respectively. The metal telluride layer is the nitrogen-containing tungsten germanide layer 52C. The above multilayer intermediate structure may also be formed by other different structures. For example: the first metal layer comprises a giant layer in addition to the titanium layer. The second and third metal layers comprise, for example, a substantially -41- layer of a nitrogen-containing titanium tungsten layer in addition to the nitrogen-containing tungsten layer. 200828425 The same material. The nitrogen-containing metal telluride layer includes a titanium-containing nitride layer or a nitrogen-containing giant telluride layer in addition to the nitrogen-containing tungsten germanide layer. The set of layers is formed by performing a PVD method including sputtering, a CVD method, or an ALD method. The nitrogen-containing titanium tungsten layer was formed by performing a reactive-state plating deposition method by sputtering IE with strontium-tungsten in a nitrogen atmosphere. The nitrogen-containing titanium telluride layer and the nitrogen-containing giant telluride layer are formed by a single-time titanium oxide and molybdenum telluride sputtering target deposition method in a nitrogen atmosphere. The macrolayer is formed to a thickness of about 10A to 80a. Preferably, the layer of tantalum has a thickness of from about 1 〇A to about 50 Å. The macro layer is changed to TaN by some of its upper portion by subsequent WNX deposition to form a second metal layer, and some of its lower portion reacts with the first conductive layer 51, that is, the polycrystalline layer Thus, a layer of T a S ix is formed, and thus has a thickness as defined above. If the thickness of the giant layer is large, the thickness of the TaSix layer also increases due to its volume expansion. Further, if the thickness of the macro layer is large, the molybdenum layer can absorb the dopant of the polysilicon layer 51, for example, phosphorus or boron, so that multiple depletion occurs in the polysilicon layer 51, resulting in deterioration of device performance. Each of the nitrogen-containing titanium tungsten layer, the nitrogen-containing titanium telluride layer, and the nitrogen-containing molybdenum telluride layer has a thickness of about 20 to 200 A, and each layer has a range of about 1% to 60%. Nitrogen content. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may not occur as a diffusion barrier due to the nitrogen-containing titanium or telluride layer. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing titanium or telluride layer may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. In the nitrogen-containing titanium tungsten layer, the ratio of titanium to tungsten is about 0. 5 to 3. Within the range of 0. In the nitrogen-containing titanium telluride layer, the ratio of niobium to titanium is about 0. 5 to 3. Within the range of 0. In the nitrogen-containing group telluride layer, the ratio of lanthanum to lanthanum is about 0. 5 to 3. 0 of the range -42-, 200828425. Fig. 6B depicts a gate stack structure in accordance with an eleventh embodiment of the present invention. The gate stack structure includes a first conductive layer 501, an intermediate structure 502, and a second conductive layer 503. The first conductive layer 501 includes a polysilicon layer highly doped with a P-type impurity (e.g., boron) or an N-type impurity (e.g., phosphorus). The first conductive layer 510 may also include a polysilicon layer (S h .  x G e x, where X is about 0 · 0 1 and 1. Within the range of 0) or the telluride layer. For example, the telluride layer includes a material selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), giant (Ta), φ (Hf), and pin (Zr). And one of the groups consisting of platinum (Pt). The second conductive layer 503 includes a tungsten layer. The tungsten layer is about 100 to 2000 Å thick and is formed by performing a PVD method, a CVD method, or an ALD method. The PVD method includes a splash deposition method using a tungsten sputtering target. The intermediate structure 502 includes a titanium-containing titanium (TiN) layer 502A, a first nitrogen-containing tungsten (WNx) layer 502B, a nitrogen-containing tungsten germanide (WShNy) layer 502C, and a second nitrogen-containing tungsten (WN layer 502D). More specifically, the nitrogen-containing titanium layer 205A has a certain ratio of nitrogen to titanium (for example, at about 0. 2 to 〇.  Within the range of 8) and formed with a thickness of φ of about 1〇A to 150A. Here, the nitrogen-containing metal layer, i.e., the nitrogen-containing layer 50 2 A, has a ratio of nitrogen to cerium as described above to prevent s i N from being generated in the nitrogen-containing titanium layer 502A. Since excessive Ti in the nitrogen-containing titanium layer 205A during the subsequent annealing treatment breaks the Si-N bond formed between the polysilicon and TiNx and thus removes SiN, the generation of SiN can be prevented. This is because the ΤιΝ link is more robust than the SiN link. The nitrogen-containing titanium layer 5?2a represents a titanium nitride layer or a titanium layer containing a certain content/weight ratio of nitrogen. Each of the first and second nitrogen-containing tungsten layers 502B and 502D has a certain ratio of nitrogen to tungsten (for example, at about 〇·3 to 1. Within the scope of 5). Each of the first -43-200828425 and the second nitrogen-containing tungsten layers 5 0 2B and 5 02D also includes a tungsten nitride layer. Although described later, it is known that the first and second nitrogen-containing tungsten layers 205B and 502D supply nitrogen to the nitrogen-containing titanium layer 502A and the nitrogen-containing tungsten sulphide layer 502C. Each of the first and second nitrogen-containing tungsten layers 502B and 502D is formed to have a thickness of about 20 A to 200 Å. The first and second nitrogen-containing tungsten layers 502B and 502D become a pure tungsten layer or a tungsten layer containing a trace of nitrogen after the annealing due to the supply of nitrogen. The ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 502C is about φ 0. 5 and 3. The nitrogen content of the nitrogen-containing tungsten carbide layer 502C is in the range of from about 10% to about 60%. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing tungsten carbide layer 502C cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten carbide layer 502C may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The nitrogen-containing tungsten germanide layer 502C also includes a tungsten germanium nitride layer. The nitrogen-containing tungsten carbide layer 502C has a thickness of about 20A to 200A. φ The first and second nitrogen-containing tungsten layers 502B and 502D are formed by performing a PVD method, a CVD method, or an ALD method. The nitrogen-containing titanium layer 502A and the nitrogen-containing tungsten carbide layer 502C are formed by a PVD method. The PVD method is carried out by a sputtering deposition method or a reactive sputtering deposition method. For example, the nitrogen-containing titanium layer 502A is formed by sputtering deposition using a titanium sputtering target in a nitrogen atmosphere. Each of the first and second nitrogen-containing tungsten layers 502B and 502D is formed by performing a reactive sputtering deposition method using a tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing tungsten telluride layer 502C is formed by performing a reactive sputtering deposition method with a tungsten telluride sputtering target in a nitrogen atmosphere. In particular, since -44-200828425 is uniformly formed into the nitrogen-containing tungsten germanide layer 502C regardless of the underlying type, the PVD method (for example, reactive sputtering deposition method) is used to form the nitrogen-containing tungsten. Telluride layer 502C. A gate stack structure according to an eleventh embodiment of the present invention includes the first conductive layer 501, the TiNx/WNx/WSixNy/WNx intermediate structure 502, and the second conductive layer 503. The first conductive layer 501 includes a polysilicon and the second conductive layer 503 includes tungsten to form a tungsten polysilicon gate stack structure. .  Specifically, the TiNx/WNJWSixNy/WNx intermediate structure 502 is formed in a stacked structure including a first gold layer, a second metal layer, a nitrogen-containing metal φ telluride layer, and a third metal layer. The first, second and third metal layers are nitrogen-containing metal layers, and the nitrogen-containing metal telluride layer contains nitrogen in a certain content/weight ratio. For example, the first metal layer is the nitrogen-containing titanium layer 052 A, and the second and third metal layers are the first and second nitrogen-containing tungsten layers 052B and 50 2D, respectively. The metal telluride layer is the nitrogen-containing tungsten germanide layer 502C. The above multilayer intermediate structure may also be formed by other different structures. For example: φ In addition to the nitrogen-containing titanium layer, the first metal layer further includes a nitrogen-containing giant (TaNO layer. In addition to the nitrogen-containing tungsten layer, the second and third metal layers further include, for example, nitrogen-containing titanium tungsten. The (TiWNx) layer is substantially the same material. In addition to the nitrogen-containing tungsten germanide layer, the nitrogen-containing metal telluride layer further includes a nitrogen-containing titanium telluride (TiSixNy) layer or a nitrogen-containing germanide (TaShNy) layer. The nitrogen-containing giant layer is formed by performing a PVD method including sputtering, a CVD method or an ALD method, and the nitrogen-containing titanium tungsten layer is formed by performing a reactive sputtering deposition method with a titanium-tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing zirconia layer and the nitrogen-containing-45- are formed by a reactive sputtering and plating method in a nitrogen atmosphere with individual titanium telluride and molybdenum telluride sputtering targets. 200828425 Telluride layer. The nitrogen-containing tantalum layer is formed to a thickness of about 10 to 80 people. Each of the nitrogen-containing titanium tungsten layer, the nitrogen-containing titanium telluride layer, and the nitrogen-containing giant telluride layer is formed to a thickness of about 20 A to 200 A, and each layer has a range of between about 1% and 60%. Nitrogen content. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing titanium or telluride layer cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing titanium or telluride layer may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. In the nitrogen-containing titanium tungsten layer, the ratio of titanium to tungsten is between about 0.5 and 3. Within the range of 0. In the nitrogen-containing titanium telluride layer, the ratio of niobium to titanium is about 〇.  5 to 3. Within the range of 0. In the nitrogen-containing giant telluride layer, the ratio of lanthanum to giant is in about 〇.  5 to 3. Within the scope of 0. Fig. 6C depicts a gate stack structure in accordance with a twelfth embodiment of the present invention. The gate stack structure includes a first conductive layer 51, an intermediate structure 51 and a second conductive layer 513. The first conductive layer 511 includes a polysilicon layer highly doped with a P-type impurity (e.g., boron (B)) or an N-type impurity (e.g., phosphorus (P)). The first conductive layer 51 1 may include a polycrystalline silicon germanium (SinGex) layer in addition to the polycrystalline germanium layer, wherein the X system is about 〇·〇1 and 1. Within the range of 0, or including the telluride layer. The telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), station (Co), titanium (Ti), tungsten (W), molybdenum (Ta), (Hf), lanthanum (Z 〇 and platinum (Pt) One of the group consisting of. The second conductive layer 513 includes a tungsten layer. One of a PVD method, a CVD method, and an ALD method is performed to form a tungsten layer of about 1 〇〇A to 2000 Å thick. The pVD method A sputtering deposition method using a tungsten sputtering target is included. The intermediate structure 512 includes a titanium germanide (TiSi layer 512A, a nitrogen-containing gold-46-200828425 (TiNx) layer 512B, a first nitrogen-containing tungsten (WNx) layer 512C, a nitrogen-containing tungsten sand compound (WSixNy) layer 512D and a gas-containing tungsten layer 512E. The intermediate structure 5 1 2 can be formed in different structures according to the selection materials described in the tenth and eleventh embodiments of the present invention. The gate stack structure of the twelfth embodiment is a structure resulting from annealing treatment of the gate stack structures according to the tenth and eleventh embodiments of the present invention. The annealing includes after forming the gate stack structures. Heat treatment accompanying φ during various processes (eg, spacer formation and inner insulating layer formation) implemented. Refer to Sections 6 C and 6 A is for comparing the intermediate structure 51 to the intermediate structure 52. When the titanium layer 52A reacts with the polysilicon from the first conductive layer 51, a titanium germanide layer 512A having a thickness of about 1 to 30 A is formed. The ratio of bismuth to titanium in the titanium telluride layer 5 1 2 A is about 〇.  5 and 3. Within the range of 0. When nitrogen is supplied from the first nitrogen-containing tungsten layer 52B to the titanium layer 52A, the nitrogen-containing titanium layer 512B is formed. The nitrogen-containing titanium layer 512B has a thickness ranging from about 10A to 100A φ and has about 0. 7 to 1. The ratio of nitrogen to titanium in the range of 3. After the annealing, each of the first and second nitrogen-containing tungsten layers 512C and 51 2E has a nitrogen content reduced to about 10% or less due to the erosion. The component symbol WNx(D) represents the nitrogen-containing tungsten layer of the uranium. Each of the first and second nitrogen-containing tungsten layers 512C and 512E is about 2 Å to 200 Å thick. The ratio of nitrogen to tungsten in each of the first and second nitrogen-containing tungsten layers 512C and 51 2E is about 0. 0 1 and 0. Within the range of 1 to 5. The nitrogen-containing tungsten carbide layer 5 1 2 D has substantially the same thickness and composition as the gas-containing tungsten carbide layer 52C. In detail, the nitrogen-containing tungsten telluride layer - 47 - 200828425 5 1 2D has about 0. 5 to 3. The ratio of 0 to tungsten in the range of 0 and the nitrogen content of about 10% to 60%. The thickness of the nitrogen-containing tungsten germanide layer 512D is in the range of between about 20A and 200A. Referring to Figures 6C and 6B, the intermediate structure 51 is compared to the intermediate structure 502. During the annealing treatment, nitrogen is supplied from the nitrogen-containing tungsten layer 502B to the nitrogen-containing titanium layer 502A. As a result, the nitrogen-containing titanium layer 502A is transformed into a nitrogen-containing titanium layer 5 1 2 B which has a minimum reaction with the titanium germanide layer 5 1 2 A. The titanium telluride layer 512A has a thickness in the range of about 1A to 30A, and the φ-containing titanium nitride layer 512B has a thickness in the range of about 10A to 100A. The ratio of nitrogen to titanium in the titanium-containing titanium layer 512B is about 0. 7 and 1. Within the range of 3 rooms. After the annealing, when the first and second nitrogen-containing tungsten layers 502B and 50 2D are eroded, each of the first and second nitrogen-containing tungsten layers 512C and 512E has a reduction of about 10% or less. Nitrogen content. Each of the first and second nitrogen-containing tungsten layers 5 12C and 5 12E is about 20A to 200A thick. The ratio of nitrogen to tungsten in each of the first and second nitrogen-containing tungsten layers 512C and 512E is about φ about 0. 01 and 0. Within the range of 15 rooms. The nitrogen-containing tungsten carbide layer 5 1 2D has substantially the same thickness and composition as the nitrogen-containing tungsten carbide layer 502C. In detail, the nitrogen-containing tungsten telluride layer 5 1 2D has about 0. 5 to 3. The ratio of 0 to tungsten in the range of 0 and the nitrogen content of about 10% to 60%. The thickness of the nitrogen-containing tungsten carbide layer 512D is in the range of between about 20 A and 20 A. The gate stack structure according to the twelfth embodiment includes a first intermediate structure and a second intermediate structure. The first intermediate structure includes a metal telluride layer and a first nitrogen-containing metal layer, and the second intermediate structure includes a second nitrogen-containing metal -48-200828425 layer, a nitrogen-containing metal telluride layer, and a third nitrogen-containing metal layer. For example, the first intermediate structure is formed by stacking the titanium germanide layer 5 1 2A and the nitrogen-containing titanium layer 5 1 2B. The second intermediate structure is formed by stacking the nitrogen-containing layer 512C, the nitrogen-containing tungsten carbide layer 512D, and the nitrogen-containing tungsten layer 5 1 2E. .  Each of the intermediate structures according to the first to twelfth embodiments of the present invention includes a nitrogen-containing metal telluride layer (for example, a nitrogen-containing tungsten germanide layer) and also includes a plurality of thin layers (including titanium, tantalum, tungsten, and nitrogen). ). The nitrogen-containing tungsten telluride φ layer is formed by performing a reactive sputtering deposition method using a tungsten ruthenium sputter target in a nitrogen atmosphere. When the nitrogen-containing tungsten telluride layer is deposited, the reactive sputtering deposition method converts the titanium layer into the titanium nitride layer. In the case where the nitrogen-containing tungsten layer is formed over the titanium layer, the titanium layer is converted into the titanium nitride layer. Since the nitrogen-containing tungsten germanide layer acts as an amorphous diffusion barrier, when the tungsten layer is formed, the tungsten layer has a small specific resistance and a large grain size in the range of about 15 μ Ω-cm. Therefore, since the tungsten layer having a low specific resistance can be formed, the tungsten layer lowers the sheet resistance. Since the 0-titanium layer or the nitrogen-containing titanium layer is transformed into the titanium nitride layer when the nitrogen-containing tungsten layer or the nitrogen-containing tungsten germanide layer is formed, the first to twelfth embodiments according to the present invention The gate stack structure has low contact resistance and can reduce polysilicon vacancies. Furthermore, since the nitrogen-containing tungsten germanide layer is included in each of the intermediate structures, the gate stack structure has a low sheet resistance. Since the titanium layer or the titanium nitride layer is transformed into the titanium nitride layer, each of the plurality of layers included in the intermediate structures contains nitrogen. As a result, the contact resistance and the sheet resistance are low, and the height of each gate stack structure can be reduced. In addition, it is allowed to reduce the polysilicon 矽-49-200828425 depletion effect caused by the outward diffusion of impurities (e.g., boron) doped in the first conductive layer. Fig. 7A depicts a gate stack structure in accordance with a thirteenth embodiment of the present invention. The gate stack structure includes a first conductive layer 61, an intermediate structure 62, and a second conductive layer 63. The first conductive layer 61 includes a polysilicon layer highly doped with a P-type impurity (e.g., boron) or an N-type impurity (e.g., phosphorus). The first conductive layer 61 may also include a polycrystalline germanium layer (si^xGex, wherein the X system is about 〇. 〇1 and 1. Within the range of 0) or the telluride layer. For example, the telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), tantalum (Ta), bell φ (Hf), zirconium (Zr). And one of the groups consisting of platinum (pt). The second conductive layer 63 includes a tungsten layer. The tungsten layer is about 1 〇 〇 A to 2,000 Å thick and is formed by performing a PVD method, a CVD method, or an ALD method. The PVD method includes a sputtering deposition method using a tungsten sputtering target. The intermediate structure 62 includes a titanium (ITO) layer 62A, a first nitrogen-containing tungsten (WNxm 62B' tungsten germanide (WSix) layer 62C (where the X system is about 1. Within the range of 5 and 10) and the second nitrogen-containing tungsten (WNX) layer 62D. More specifically, the titanium layer 62A is formed to have a thickness ranging from about 10A to 80 people. Preferably, the titanium layer 62A has a thickness of φ of from about 1 〇A to about 5 Å. The titanium layer 62A is changed to T1N by some of its upper portion by subsequent WNX deposition to form a nitrogen-containing tungsten layer 62B, and some of its lower portion reacts with the first conductive layer 61, that is, the polycrystalline layer Thus, the TiSix layer is formed, and thus has a thickness as defined above. If the thickness of the layer 6 2 A is large, the thickness of the TiSh layer also increases due to its volume expansion. In addition, if the thickness of the titanium layer 62A is large, the titanium layer 62A can absorb dopants, for example, phosphorus or boron of the polysilicon layer 61 and thus multiple depletion in the polysilicon layer 61, resulting in deterioration of device performance. . The nitrogen of each of the first and second nitrogen-containing tungsten layers 62B and 62D has a certain ratio to tungsten •50-200828425 (for example, at about 0. 3 to 1. Within the scope of 5). Each of the first and second nitrogen-containing tungsten layers 62B and 62D also includes a tungsten nitride layer. Although described later, it is known that the first and second nitrogen-containing tungsten layers 62B and 62D have metallic properties. The first and second nitrogen-containing tungsten layers 62B and 62D supply nitrogen to the tungsten germanide layer 62C. Each of the first and second nitrogen-containing tungsten layers 62B and 62D is formed to have a thickness of about 2 Å to 200 Å. The first and second nitrogen-containing tungsten layers 62B and 62D become a pure tungsten layer or a tungsten-containing tungsten layer after the annealing due to the supply of nitrogen. The ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 62C is about 0. 5 and 3. Within the range of 0. The nitrogen-containing tungsten carbide layer 62C is formed to a thickness of about 20 Å to 100 Å. The titanium layer 62A, the first and second nitrogen-containing tungsten layers 62B and 62D, and the tungsten layer 63 are formed by performing a PVD method, a CVD method, or an ALD method. The nitrogen-containing tungsten telluride layer 62C is formed by performing a PVD method. The PVD method is carried out by a sputtering deposition method or a reactive sputtering deposition method. For example, the titanium layer 62A is formed by performing a sputtering deposition method using a titanium sputtering target. Each of the first and second nitrogen-containing tungsten layers 62B and 62D is formed by performing a reactive sputtering deposition method with a tungsten sputtering target in a nitrogen atmosphere. The nitrogen-containing tungsten germanide layer 62C is formed by performing a reactive sputtering deposition method using a tungsten telluride sputtering target. The tungsten layer 63 is formed by sputtering deposition using a tungsten sputtering target. The gate stack structure according to the thirteenth embodiment of the present invention includes the first conductive layer 61 and the Ti/WNWWSix/WNx intermediate structure 62. And the second conductive layer 63. The first conductive layer 61 includes polysilicon and the second conductive layer 63 includes tungsten, thereby forming a tungsten polysilicon gate stack structure. -51- 200828425 Specifically, the Ti/WNx/WSix/WNx intermediate structure 62 is formed in a stacked structure including a first metal layer, a second metal layer, a metal telluride layer, and a third metal layer. The first metal layer comprises a layer of pure metal. The second and third metal layers comprise a nitrogen-containing metal layer, and the metal germanide layer comprises a pure tungsten germanide layer. For example, the first metal layer is the titanium layer 62A, and the second and second metal layers are the first and second nitrogen-containing tungsten layers 62B and 62D, respectively. The metal telluride layer is the nitrogen-containing tungsten germanide layer 62C. The above multilayer intermediate structure may also be formed by other different structures. For example: φ In addition to the titanium layer, the first metal layer further includes a molybdenum layer. In addition to the tungsten germanide layer, the metal telluride layer further comprises a titanium telluride (TiSh) layer, wherein the X system is at 1. Within the range of 5 and 10, or the layer of giant salium (TaSu), where X is at 1. Within the range of 5 and 10. In addition to the nitrogen-containing tungsten layer, the second and third metal layers further comprise a layer of nitrogen-containing titanium tungsten (TiWNx). The tantalum layer is formed by performing a PVD method including sputtering, a CVD method, or an ALD method. The nitrogen-containing titanium tungsten layer was formed by performing a reactive sputtering deposition method with a titanium tungsten sputtering target in a nitrogen atmosphere. The titanium germanide layer and the giant sand layer are formed by performing a reactive sputtering φ plating deposition method on individual titanium telluride and giant telluride sputtering targets. The giant layer is formed to have a thickness of about 10A to 80A. Preferably, the macrolayer has a thickness of from about i〇A to about 50A. The macro layer is changed to TaN by some of its upper portion by subsequent WNX deposition to form a second metal layer, and some of its lower portion reacts with the first conductive layer S, that is, the polycrystalline layer is formed. The TaSix layer has a thickness as defined above. If the thickness of the giant layer is large, the thickness of the TaS layer is also increased due to the expansion of the volume. Further, if the thickness of the layer is large, the molybdenum layer can absorb the dopant. For example, phosphorus or boron of the polycrystalline germanium layer 6 1 and thus multiple depletions occur in the polycrystalline germanium layer 61, and the degradation of the device performance is caused by the -52-200828425. The nitrogen-containing titanium tungsten layer is about 20 A to 200 A thick. Each layer of the layer and the giant telluride layer is formed to a thickness of about 20 A to 200 A. The nitrogen-containing titanium tungsten layer has a nitrogen content ranging between about 10% and 60%. In the nitrogen-containing titanium tungsten layer, the titanium pair The ratio of tungsten is about 0. 5 to 3. Within the range of 0. In the titanium telluride layer, the ratio of bismuth to titanium is about 〇.  5 to 3. Within the range of 0. In the giant telluride layer, the ratio of 矽 to giant is about 0. 5 to 3. Within the range of 0. The tungsten germanide layer 62C is formed over the first nitrogen-containing tungsten φ layer 62B by performing a PVD method (for example, a sputtering deposition method). The sputtering deposition method is carried out with the tungsten germanide sputtering target to allow uniform formation of the tungsten germanide layer 62C regardless of the underlying type. Figure 7B depicts an image of the structure disposed after formation of a tungsten germanide layer over a nitrogen-containing tungsten layer by performing individual chemical vapor deposition (CVD) and physical vapor deposition (PVD) methods. Although the tungsten germanide layer CVD-WSh is not formed properly over the tungsten nitride layer WN by the CVD method, the tungsten germanium oxide can be uniformly formed over the tungsten nitride layer WN by the PVD method. Layer PVD-WSh. Therefore, since the tungsten layer having a low specific resistance can be formed over the tungsten germanide layer, the sheet resistance of the tungsten layer can be reduced. According to the gate stack structure of the thirteenth embodiment of the invention, when the nitrogen-containing tungsten layer 62B is formed over the titanium layer, the titanium layer is converted into a titanium nitride layer. According to the thirteenth embodiment of the present invention, since the titanium layer of the intermediate structure is transformed into the titanium nitride layer during formation of the nitrogen-containing layer, the gate stack structure can obtain low contact resistance and reduce the polysilicon enthalpy effect. Moreover, since the intermediate structure includes the tungsten germanide layer, the gate-53-200828425 pole stack structure can also obtain a low sheet resistance. Fig. 7C depicts the structure of the fourteenth embodiment in accordance with the present invention. The gate stack structure includes a first conductive layer 6 〇 1, a middle ring, and a second conductive layer 603. The first conductive layer 601 includes a poly-doped (eg, boron) or N-type impurity (eg, phosphorus) polycrystalline germanium-' conductive layer 601 may also include a polycrystalline 5 sigma layer (Sii_x〇ex, where 0 . 01 and 1. Within the range of 0) or the telluride layer. For example, the selection is selected from nickel (Ni), chromium ((:]:), cobalt ((: 〇), titanium (1^), tungsten (^¥) to (H f}, pin (Z r) And one of the groups consisting of platinum (P t ). The second conductive layer 603 comprises a crane layer. . The tungsten layer is about 2000 Å thick and is subjected to a PVD method, a CVD method or an ALD method. The PVD method includes a sputtering deposition method using a tungsten sputtering target. The intermediate structure 602 includes a nitrogen-containing titanium (TiN layer 602A, a tungsten (WNU) layer 602B, a tungsten germanide (WSix) layer 602C, and a second containing layer 602D. More specifically, the nitrogen of the nitrogen-containing titanium layer 602A For a ratio (for example: at about 0. 2 to 0. Within the range of 8) and formed with a thickness of φ 丨 5 〇. Here, the nitrogen-containing metal layer, i.e., the nitrogen-containing food, has a ratio of nitrogen to titanium as described above to prevent SiN from being produced in the 602A. Since excessive Ti in the nitrogen-containing nitrogen during the subsequent annealing treatment destroys the formation between the polycrystalline germanium and TiNx and thus removes SiN, the generation of SiN can be prevented. This is made possible by the stronger TiN ^ link. The nitrogen-containing titanium layer 602A also includes a layer. Each of the first and second nitrogen-containing tungsten layers 602B and 602D has a certain proportion (for example, at about 〇.  3 to 1. In the range of 5, the gate stack 3 structure 602 is mixed with the P-type layer. The Xth system is formed in a composite layer, tantalum (Ta), and 100A to ί. The first nitrogen-containing tungsten-nitrogen (WNX) titanium has a certain 10A to long layer 60 2 A, a titanium-containing titanium layer titanium layer 602A Si-N bond and a nitrogen ratio of the SiN 丨 titanium nitride-layer) Each of the first-54·200828425 and the second nitrogen-containing tungsten layers 602B and 602D also includes a tungsten nitride layer. The first and second nitrogen-containing tungsten layers 602B and 602D supply nitrogen to the tungsten carbide layer 602C. Each of the first and second nitrogen-containing tungsten layers 602B and 602D is formed to a thickness of about 20 A to 200 Å. The first and second nitrogen-containing tungsten layers 602 B and 602D become a pure tungsten layer or a tungsten layer containing a trace of nitrogen after the annealing due to the supply of nitrogen. The ratio of germanium to tungsten in the tungsten germanide layer 602C is about 〇. 5 and 3. Between the range. The tungsten germanide layer 602C has a thickness of about 20A to 200A. The first and second nitrogen-containing tungsten layers 602B and 602D are formed by performing a PVD method, a CVD method, or an ALD method. The nitrogen-containing titanium layer 602A and the tungsten germanide layer 602C are formed by performing a PVD method. The P, VD method is carried out by a sputtering deposition method or a reactive sputtering deposition method. For example, the nitrogen-containing titanium layer 602A is formed by sputtering deposition using a titanium sputtering target in a nitrogen atmosphere. The first and second nitrogen-containing tungsten layers 602B and φ 602D are formed by performing a reactive sputtering deposition method with a tungsten sputtering target in a nitrogen atmosphere. The tungsten germanide layer 602C is formed by performing a reactive sputtering deposition method using a tungsten telluride sputtering target. The tungsten layer 603 is formed by sputtering deposition using a tungsten sputtering target. A gate stack structure according to a fourteenth embodiment of the present invention includes the first conductive layer 601, the TiNx/WNx/WSix/WNx intermediate structure 602, and the second conductive layer 603. The first conductive layer 601 includes polysilicon and the second conductive layer 603 includes tungsten to form a tungsten polysilicon gate stack structure. Specifically, the TiNx/WNx/WSix/WNx-55-200828425 intermediate structure 602 is formed in a stacked structure including a first metal layer, a second metal layer, a metal telluride layer, and a third metal layer. The first, second and third metal layers are nitrogen-containing metal layers, and the metal sand layer is a pure metal sand layer. For example, the first metal layer is the nitrogen-containing titanium layer 602A, and the second and third metal layers are the first and second nitrogen-containing tungsten layers 602 B and 602D, respectively. The metal telluride layer is the tungsten silicide layer 602C. The above multilayer intermediate structure may also be formed by other different structures. For example: in addition to the nitrogen-containing titanium layer, the first metal layer further includes a nitrogen-containing molybdenum (TaN〇 layer. In addition to the tungsten germanide layer, the metal telluride layer further includes titanium germanium oxide (TiSh), The X system is about 1. Within the range of 5 and 10, or Tasix, where X is about 1. Within the range of 5 and 10. In addition to the nitrogen-containing tungsten layer, the second and third metal layers further include a nitrogen-containing titanium tungsten (TiWNx) / layer. The nitrogen-containing macrolayer was formed by performing a reactive sputtering method with a ruthenium sputtering target in a nitrogen atmosphere. The nitrogen-containing titanium tungsten layer was formed by performing a reactive sputtering deposition method using a titanium tungsten sputtering target in a nitrogen atmosphere. The titanium germanide layer and the germanide layer are formed by performing a reactive sputtering deposition method on individual titanium germanide and giant telluride sputtering targets. The nitrogen-containing tantalum layer is formed to have a thickness of about 10A to 150A φ. Each of the nitrogen-containing titanium tungsten layer, the titanium germanide layer, and the vaporized layer is formed to a thickness of about 20 to 200 people. The nitrogen content of the nitrogen-containing titanium tungsten layer is in the range of between about 10% and 60%. In the nitrogen-containing titanium tungsten layer, the ratio of titanium to tungsten is about 0.5 to 3. Within the range of 0. In the titanium germanide layer, the ratio of germanium to titanium is about 0. 5 to 3. Within the range of 0. In the eucalyptus layer, the ratio of bismuth to bismuth is about 0. Within the range of 5 to 3 · 0. The tungsten silicide layer 602C is formed over the first nitrogen-containing tungsten layer 602B by a PVD method (e.g., sputtering deposition method) in the intermediate structure 602 described above. The sputter deposition method is performed with the tungsten telluride sputter target to allow uniform formation of the tungsten sand compound - 56 - 200828425 layer 602C regardless of the underlying type. Fig. 7D depicts a gate stack structure in accordance with a fifteenth embodiment of the present invention. The gate stack structure includes a first conductive layer 6丨丨, an intermediate structure 6丄2, and a first conductive layer 613. The first conductive layer 611 includes a highly doped p-type impurity (for example, boron (B)) or an N-type impurity (. For example: a polycrystalline germanium layer of phosphorus (P). In addition to the polysilicon layer, the first conductive layer 611 may also include a polysilicon (Si^Gex) layer, wherein the X system is about 〇. Within the range of 〇1 and 1〇, or including a telluride layer. The telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt φ (Co), titanium (Ti), tungsten (w), giant (Ta), (Hf), zirconium (Zr), and platinum ( One of the groups consisting of Pt). The second conductive layer 613 includes a tungsten layer. One of the PVD method, the CVD method, and the ALD method is performed to form a tungsten layer of about ΐοοΑ to 2000A thick. The PVD method includes a sputtering deposition method using a tungsten sputtering target. The intermediate structure 6 1 2 includes a titanium telluride (TiSix) layer 6 1 2 A, a nitrogen-containing titanium (TiN〇 layer 612B, a first nitrogen-containing tungsten (WNX) layer 612C, a nitrogen-containing tungsten telluride (WSixNy) The layer 612D and the second nitrogen-containing tungsten layer 612E. The intermediate structure 612 can be formed in different structures according to the selection materials described in the φ thirteenth and fourteenth embodiments of the present invention. According to the fifteenth embodiment of the present invention The gate stack structure is a structure caused by annealing the gate stack structures according to the thirteenth and fourteenth embodiments of the present invention. The annealing is performed after forming the gate stack structures. Heat treatment accompanying various processes (eg, spacer formation and formation of inner insulating layer). Referring to Figures 7D and 7A to compare the intermediate structure 612 with the intermediate structure 62. When the titanium layer 62A is derived from The polysilicon of the first conductive layer 61 is reversed -57-200828425 to form a titanium telluride layer 6 1 2 A having a thickness of about 1 A to 30 A. The ratio of tantalum to titanium in the titanium chemostat layer 612A is About 0. 5 and 3. Within the range of 0. When nitrogen is supplied from the titanium layer 62A to the titanium layer 62A, the titanium-containing titanium layer 612B is caused. The nitrogen-containing titanium layer 6128 has a thickness ranging from about 10 to 1 且 and has a thickness of from about 0.6 to about 1. 2 range of nitrogen to titanium ratio. After the annealing, each of the first and second nitrogen-containing tungsten layers 61 2C and 61 2E has a nitrogen content reduced to about 10% or less by the etching. The φ element symbol WNx(D) represents the eroded nitrogen-containing tungsten layer. Each of the first and second nitrogen-containing tungsten layers 6 1 2 C and 6 1 2 E is about 2 〇 A to 2 Ο Ο A thick. The ratio of nitrogen to tungsten in each of the first and second nitrogen-containing tungsten layers 61 2C and 61 2E is in the range of about 0 · 0 1 and 0 · 15 . When the nitrogen from the first and second nitrogen-containing tungsten layers 602 B and 602D is decomposed, the tungsten germanide layer 602C is transformed into the nitrogen-containing tungsten germanide layer 6 1 2D. The ratio of germanium to tungsten in the nitrogen-containing tungsten germanide layer 6 1 2D is about 0. Within the range of 5 to 3 · 0. The nitrogen-containing tungsten telluride layer 6 1 2 D has a nitrogen content of about 10% φ to 60% and a thickness of about 20A to 200 people. Referring to Figures 7D and 7C, the intermediate structure 612 is compared to the intermediate structure 602. During the annealing treatment, nitrogen is supplied from the nitrogen-containing tungsten layer 602B to the nitrogen-containing titanium layer 602A. As a result, the nitrogen-containing titanium layer 602A was converted into a nitrogen-containing titanium layer 6 1 2B having a minimum reaction with the titanium germanide layer 6 1 2A. The thickness of the titanium telluride layer 6 1 2 A is in the range of about 1 A to 3 Å A, and the thickness of the nitrogen-containing titanium layer 612B is in the range of about 10 A to 100 Å. The ratio of nitrogen to titanium in the titanium-containing titanium layer 6 1 2 B is about 〇·7 and 1.  0 -58- 200828425 Within the range of 3, after the annealing, when the first and second nitrogen-containing tungsten layers 602B and 602D are invaded, each of the first and second nitrogen-containing tungsten layers 61 2C and 612E Has a nitrogen content that drops to about 10% or less. Each of the first and second nitrogen-containing tungsten layers 612C and 612E is about 20A to 200A thick. The ratio of nitrogen to tungsten in each of the first and second nitrogen-containing tungsten layers 61 2C and 6 12E is about 0. 0 1 and 0.  Within the range of 1 to 5. When the nitrogen from the first and second nitrogen-containing tungsten layers 602 B and 602D is eroded, the tungsten germanide layer 602C is transformed into the nitrogen-containing tungsten carbide layer φ 612D. The nitrogen-containing tungsten germanide layer 612D has about 0. 5 to 3. The ratio of zero to tungsten and about 10% to 60% of nitrogen. Here, the nitrogen content is appropriately adjusted in the above manner. If the nitrogen content is too low, the junction reaction may occur because the nitrogen-containing tungsten carbide layer 612D cannot be successfully used as a diffusion barrier. On the other hand, if the nitrogen content is too high, the SiN content contained in the nitrogen-containing tungsten carbide layer 612D may be high, and thus the contact resistance becomes high, resulting in deterioration of element performance. The thickness of the nitrogen-containing tungsten carbide layer 612D is in the range of between about 20 and 200 people. φ The gate stack structure according to the fifteenth embodiment includes a first intermediate structure and a second intermediate structure. The first intermediate structure includes a metal telluride layer and a first nitrogen-containing metal layer, and the second intermediate structure includes a second nitrogen-containing metal layer, a nitrogen-containing metal telluride layer, and a third nitrogen-containing metal layer. For example, the first intermediate structure is formed by stacking the titanium germanide layer 6 1 2A and the nitrogen-containing titanium layer 6 1 2B. The second intermediate structure is formed by stacking the nitrogen-containing tungsten layer 612C, the nitrogen-containing tungsten germanide layer 612D, and the nitrogen-containing tungsten layer 612E. The intermediate structure according to the first to fifteenth embodiments of the present invention can be implemented to control the gate electrode of the flash memory device in addition to the dynamic random access memory (DRAM) component. And the gate electrode of many logic components. Fig. 8 is a view showing a gate stack structure of a flash memory device in accordance with a sixteenth embodiment of the present invention. A tunneling oxide layer 702 corresponding to the gate insulating layer is formed over the substrate 707. A first polycrystalline sand electrode 703 for the floating gate FG is formed over the tunnel oxide layer 702. A dielectric layer 704 is formed over the first polysilicon electrode 703, and a second polysilicon electrode 507 for controlling the gate CG is formed over the dielectric layer 704. An intermediate structure 706 selected from the group consisting of intermediate structures of the various types described in the first to fifteenth embodiments of the present invention is formed over the second polysilicon electrode 705. The intermediate structure 706 includes a Ti/WNx/WSixNy intermediate structure in accordance with a first embodiment of the present invention. Therefore, the intermediate structure 706 is formed by continuously stacking the titanium layer 706A, the nitrogen-containing tungsten layer 706B, and the nitrogen-containing tungsten sulphide layer 706C. A tungsten electrode 707 and a hard mask 708 are formed over the intermediate structure 706. The symbol W and Η/M represent the tungsten electrode 707 and the hard cover 708, respectively. The gate stack structure of the flash memory device having the intermediate structure 706 as shown in Fig. 8 has a low sheet resistance and a contact resistance. Embodiments of the present invention can be applied to interconnections in various metals in addition to the gate electrode, for example, a bit line including an intermediate structure, a metal line, and a capacitor electrode. Furthermore, this embodiment of the invention can be applied to a gate stack structure of a semiconductor device that constitutes a dual polysilicon gate, wherein the double poly gate gate is composed of a first gate stack structure (including doping formed under the intermediate structure) a polycrystalline germanium electrode having an N-type impurity and a tungsten electrode formed over the intermediate structure) and a second gate stack structure of -60-200828425 (including a polycrystalline germanium electrode doped with a p-type impurity and formed over the intermediate structure) The tungsten electrode is composed of. Fig. 9 is a graph showing the sheet resistance (Rs) of the tungsten layer of each type of intermediate structure formed in accordance with the first to fifteenth embodiments of the present invention. The tungsten layer has a thickness of about 40 nm. It can be observed that the WSix/WNx intermediate structure is additionally applied by the CVD method and the PVD method over the Ti/WNX intermediate structure (ie,

Ti/WNx/CVD-WSix/WNx structure and 'Ti/WNx/PVD-WSh/WNx structure) • And force application □ WSixNy layer (ie, Ti/WNx/WSixNy structure), reduce the crane electrode Chip resistance. However, since the WSix layer cannot be appropriately grown over the WNx layer by the CVD method, it is necessary to form the WSH layer over the layer by the PVD method (for example, sputtering deposition method). The formation of the WSixNy layer was carried out by reactive sputtering deposition using a tungsten ruthenium sputter target and nitrogen. The sheet resistance of the Ti/WNx/CVD-WSix/WNx intermediate structure, the Ti/WNx/PVD-WSix/WNx intermediate structure, and the Ti/WNx/WSixNy intermediate junction φ tungsten electrode will be compared. The sheet resistance of the tungsten electrode is only low in the case of applying the Ti/WNx/PVD-WSix/WNx intermediate structure, and the Ti/WNx/WSixNy intermediate structure is the same as in the case of applying the WSix/WNx intermediate structure. In the case where the WSix layer is applied by the CVD method, the "3:^ layer cannot be formed uniformly over the WNx layer. As a result, agglomeration is generated over the WNx layer, thereby increasing the sheet resistance. Conversely 'If the sputtering deposition method using the WSix sputtering target or the reactive sputtering deposition method is used, the WS i X diffusion layer can be uniformly formed, thereby reducing the sheet resistance of the crane electrode. -61- 200828425 The 10A to 10C diagram describes the gate patterning process using the gate stack structure shown in Fig. 3A. The same component symbols identified in Fig. 3A denote the same components herein. Referring to Fig. 10A, a gate is formed over the substrate 800. a pole insulating layer 801, wherein an ion implantation process is performed in the substrate 80 1 to form an isolation layer, a well region, and a via. A patterned first conductive layer 2 1 is formed over the gate insulating layer 80 1 . An intermediate structure 22 is formed over the first conductive layer 21. A patterned second conductive layer 23 is formed over the intermediate φ structure 22. The patterned first conductive layer 21 includes a highly doped P-type impurity (for example, boron). Or polymorphism of N-type impurities (eg phosphorus) The patterned first conductive layer 21 may also include a polysilicon layer (Si uGe, wherein X is in the range of about 0.01 and 1.) or a germanide layer. For example, the germanide layer includes a layer selected from Nickel (Ni), chromium (C〇, cobalt (Co), titanium (Ti), tungsten (W), tantalum (Ta), (Hf), chromium (Z〇 and platinum (Pt) The intermediate structure 22 includes a patterned titanium layer (Ti) 22 A, a patterned nitrogen-containing tungsten φ (WN 〇 layer 22B, and a patterned nitrogen-containing tungsten lanthanide (WSixNy) layer 22C. The patterned second conductive layer 23 includes a tungsten layer. The tungsten layer is formed by performing a PVD method, a CVD method, or an ALD method. The method includes a sputtering deposition method using a tungsten sputtering target. A hard layer is formed over the patterned second conductive layer 23. The cover 802. The formation of the hard cover 802 may be omitted. The hard cover 802 includes silicon nitride (ShN4). A gate patterning process is performed to form the gate stack structure. In particular, although not shown, However, the first patterning process is performed using an etch barrier gate mask (not shown) formed by the photoresist layer to etch the -62-200828425 hard mask layer, the second conductor An electrical layer, a titanium layer including the intermediate structure 22, a plurality of layers of a nitrogen-containing tungsten layer and a nitrogen-containing tungsten germanide layer, and a portion of the first conductive layer. As a result, the gate insulating layer 810 and the substrate 80 The structure including the hard mask 802, the patterned second conductive layer 23, the intermediate structure 22, and the patterned first conductive layer 21 is formed over 0. Referring to FIG. 10B, the gate mask is removed, and then, A pre-spacer process is performed to prevent non-uniform etching and oxidation of the patterned second conductive layer 23 (i.e., the tungsten layer) and the intermediate structure 22. For example, a Si3N4 layer 803.φ is formed as a front spacer layer. Referring to FIG. 10C, a second gate patterning process is performed to etch the Si 3 Ν 4 layer 803 and a portion of the patterned first conductive layer 21. During the second gate patterning process, a portion of the Si3N4 layer 803 is etched using a dry etch to form spacers 803A on the sidewalls of the gate stack. The spacers 8〇3A are used as an etch barrier to etch the patterned first conductive layer 21. The symbol 2 1 A represents an electrode (for example, a polysilicon electrode). φ The first and second gate patterning processes using the front spacer layer as described above can be applied to the gate stack structure according to the second to fifteenth embodiments of the present invention. Figure 11 depicts another gate patterning process using the gate stack structure shown in Figure 3A. The same component symbols used in the first to fourth OC diagrams denote the same components. A gate insulating layer 801 is formed over the substrate 800, wherein an ion implantation process is performed in the substrate 800 to form an isolation layer, a well region, and a via. A patterned first conductive layer 2 1 b is formed over the gate insulating layer 80 1 . An intermediate structure 2 2 is formed over the patterned first conductive layer 2 1 B from the -63-200828425. A patterned second conductive layer 23 is formed over the intermediate structure 22. The patterned first conductive layer 21B includes a polysilicon layer highly doped with a p-type impurity (e.g., boron) or an N-type impurity (e.g., phosphorus). The patterned first conductive layer 21B may also include a polycrystalline sand stagger layer (.Sii^xGex, wherein the X system is in a range between about 0. 01 and 1.0) or a vaporized layer. For example, the telluride layer comprises a layer selected from the group consisting of nickel (Ni), chromium (Cr), cobalt (Co), titanium (Ti), tungsten (W), tantalum (Ta), (Hf), chromium (Zr), and platinum. One of the groups consisting of (Pt). φ The intermediate structure 22 includes a patterned titanium layer (Ti) 22A, a patterned nitrogen-containing tungsten (WNX) layer 22B, and a patterned nitrogen-containing tungsten germanide (WShNO layer 22C. The patterned second conductive layer 23 includes a tungsten layer. The tungsten layer is formed by performing a PVD method, a CVD method, or an ALD method. The PVD method includes a sputtering deposition method using a tungsten sputtering target. A hard mask 802 is formed over the patterned second conductive layer 23. The hard mask 802 is formed. The hard mask 802 includes tantalum nitride (Si3N4). A gate patterning process is performed to form the gate stack structure. Specifically, although not shown, it is used. An etched barrier gate mask (not shown) formed by the photoresist layer simultaneously etches the hard mask layer, the second conductive layer, the titanium layer including the intermediate structure 22, the nitrogen-containing tungsten layer, and the nitrogen-containing tungsten germanide layer a plurality of layers and portions of the first conductive layer. As a result, the hard mask 802, the patterned second conductive layer 23, the intermediate structure 22, and the patterning layer are formed over the gate insulating layer 801 and the substrate 800. The structure of a conductive layer 21B. The uranium engraving is selected immediately without using the front spacer layer. a pole patterning process' to replace the two-step gate patterning process using the front spacer layer. A gate patterning process that does not use the front spacer layer can be applied to the invention according to the invention -64-200828425 The gate stack structure of the second to fifteenth embodiments. According to an embodiment of the present invention, a plurality of thin layers (including titanium, tungsten, tantalum, and nitrogen or each layer included) disposed between the tungsten electrode and the polysilicon electrode The intermediate structure of nitrogen) allows for a sheet resistance as low as that of the poly-Si/WNx/W and poly-Si/WNx/WSh/W intermediate structures. Therefore, the height of the gate stack structure can be reduced, thereby Easily obtain process integration. Due to the reduction of boron penetration or boron out-diffusion, the polysilicon vacancy effect can be reduced, and therefore, the operating current of the PMOS transistor can be increased. Further, very low contact can be obtained between the tungsten electrode and the polysilicon electrode. Resistor, thus facilitating the manufacture of high speed components. As for the method of forming a tungsten polycrystalline gate for manufacturing high speed/high density/low power memory components, it can be implemented by a plurality of thin films (including titanium Intermediate structure of tungsten, tantalum and nitrogen, or each film comprising nitrogen) to achieve low contact resistance and low polysilicon vacancy effect. Although the invention has been described with reference to the specific embodiments, those skilled in the art will It is apparent that various changes and modifications can be made without departing from the spirit and scope of the invention as defined in the following claims. [FIG. 1A-1C] FIG. 1A to 1C depict a gate stack structure of a typical tungsten polysilicon gate. Figure 2A is a graph depicting the contact resistance between tungsten and polysilicon in the intermediate structure of each type. Figure 2B is a graph depicting the depth profile of the boron concentration for each type of gate stack structure. Figure 2C is a graph depicting the sheet resistance of the intermediate structure of each type. -65- 200828425 Figure 3A depicts a gate stack structure in accordance with a first embodiment of the present invention. Fig. 3B is an image obtained by forming a tungsten germanium nitride layer over a portion above the tungsten nitride layer by physical vapor deposition (PVD). Fig. 3C depicts a gate stack structure in accordance with a second embodiment of the present invention. Fig. 3D depicts a gate stack structure in accordance with a third embodiment of the present invention. φ Figure 3E depicts an image of the gate stack structure after the annealing process. Fig. 4A depicts a gate stack structure in accordance with a fourth embodiment of the present invention. Fig. 4B depicts a gate stack structure in accordance with a fifth embodiment of the present invention. Fig. 4C depicts a gate stack structure in accordance with a sixth embodiment of the present invention. Fig. 5A depicts a gate stack junction in accordance with a seventh embodiment of the present invention. Fig. 5B depicts a gate stack structure in accordance with an eighth embodiment of the present invention. Fig. 5C depicts a gate stack structure in accordance with a ninth embodiment of the present invention. Fig. 6A depicts a gate stack structure in accordance with a tenth embodiment of the present invention. Fig. 6B depicts a gate stack structure in accordance with an eleventh embodiment of the present invention. -66- 200828425 Figure 6C depicts a gate stack structure in accordance with a twelfth embodiment of the present invention. Fig. 7A depicts a gate stack structure in accordance with a thirteenth embodiment of the present invention. Figure 7B depicts an image of the structure disposed after formation of a tungsten germanide layer over a nitrogen-containing tungsten layer by performing individual chemical vapor deposition (Cvd) and physical vapor deposition (PVD) methods. Figure 7C depicts a gate stack _ structure in accordance with a fourteenth embodiment of the present invention. Fig. 7D depicts a gate stack structure in accordance with a fifteenth embodiment of the present invention. Fig. 8 depicts a gate stack structure in accordance with a sixteenth embodiment of the present invention. Fig. 9 is a graph showing the sheet resistance of a tungsten electrode of an intermediate structure of each type according to an embodiment of the present invention. 10A through 10C are cross-sectional views showing a gate stack structure according to an embodiment of the present invention to obtain a gate stack structure as shown in Fig. 3A. Fig. 1 is a cross-sectional view showing the gate patterning method using the gate stack structure shown in Fig. 3A. [Main component symbol description] 11 Polysilicon layer 12 Tungsten nitride (WN) layer 13 Tungsten (W) layer 14 Tungsten germanide (WSix) layer 21 First conductive layer -67· 200828425

21 A 21 B 22 22A 22B 22C 23 31 32 32A 3 2B 33 41 42 42A 42B 42C 43 51 52 52 A 52B 52C 52D 53 Electrode patterned first conductive layer · Intermediate structure Titanium layer Nitrogen-containing tungsten (WNx) layer containing nitrogen Tungsten telluride (WShNy) layer second conductive layer first conductive layer intermediate structure titanium layer nitrogen-containing tungsten germanide (WSixNy) layer second conductive layer first conductive layer intermediate structure titanium layer nitrogen-containing tungsten germanide (WSixNy) layer Nitrogen tungsten (WNX) layer second conductive layer first conductive layer intermediate structure titanium (T i) layer first nitrogen-containing tungsten (WNX) layer nitrogen-containing tungsten germanide (WSixNy) layer second nitrogen-containing tungsten (WN layer Two conductive layers -68- 200828425

61 first conductive layer 62 intermediate structure 62A titanium (Τι) layer 62B first nitrogen-containing tungsten (WNX) layer 62 C tungsten germanide (WSi germanium layer 62D second nitrogen-containing tungsten (WNX) layer 63 second conductive layer 201 A conductive layer 202 intermediate structure 202A nitrogen-containing titanium (TiO 2 ) layer 20 2 B nitrogen-containing tungsten (WNX) layer 202C nitrogen-containing tungsten germanide (WSixNy) layer 203 second conductive layer 211 first conductive layer 212 intermediate structure 212A titanium germanium Layer 212B Nitrogen-containing titanium (1:1 仏) layer 212C Nitrogen-containing tungsten (11) layer 21 2D Nitrogen-containing tungsten lanthanide (WSuNy) layer 213 Second conductive layer 301 First conductive layer 3 02 Intermediate structure 3 0 2 A Nitrogen-containing titanium (ITO) layer 302B nitrogen-containing tungsten germanide (WShNy) layer 303 second conductive layer-69- 200828425 conductive layer structure (TiSu) layer titanium (1^仏) layer tungsten germanide (WShNy) layer conductive layer Conductive layer structure

Titanium (;111) layer tungsten germanide (WSiXNy) layer

Tungsten (WNX) I conductive layer conductive layer structure

311 First 312 Intermediate 312A Titanium 矽 3 12B Nitrogen 312C Nitrogen 313 Second 401 First 402 Intermediate 402A Nitrogen 40 2B Nitrogen 402C Nitrogen 403 Second 411 First 412 Intermediate 412A Titanium 矽 412B Nitrogen 4 12C Nitrogen 412D Nitride (1^13 Titanium Titanium (TiNx) Titanium Telluride (WShNy) Titanium (1) Layer 413 Second Conductive Layer 501 First 502 Intermediate 502A Nitrogen 502B First 502C Nitrogen 502D Two conductive layer structure titanium (TiNx) i nitrogen-containing tungsten (, 1) layer tungsten germanide (WSixNy) layer containing nitrogen tungsten (WNX) layer -70-

200828425 503 511 512 5 12A 5 1 2 B 5 12C 512D 5 12E

601 602 602A 602B 602C 602D 603

612 612A 612B 6 1 2 C 612D 612E 613 Second Conductive Layer First Conductive Layer Intermediate Structure Titanium Telluride (TiSix) Layer Nitrogen Titanium (TiN〇 Layer First Nitrogen Containing Tungsten (WNX) Layer Nitrogen Tungsten Telluride (WShNy) Layer second nitrogen-containing tungsten layer second conductive layer first conductive layer intermediate structure nitrogen-containing titanium (ΊΠΝχ) layer 'first nitrogen-containing tungsten (WN〇 layer tungsten germanide (W Six) layer second nitrogen-containing tungsten (WNX) :^ Second conductive layer First conductive layer Intermediate structure Titanium telluride (TiSix) ^ Nitrogen-containing titanium (TiNx) layer First nitrogen-containing tungsten (WN〇 layer Nitrogen-containing tungsten telluride (WShNy) layer Second nitrogen-containing tungsten layer Second conductive layer substrate-71-701 200828425 702 Tunneling oxide layer 703 First polysilicon electrode 704 Dielectric layer 705 Second polysilicon electrode 706 Intermediate structure 706A Titanium layer 706B Nitrogen-containing tungsten layer 70 6C Nitrogen-containing tungsten Object layer 707 Crane electrode 708 Hard cover 800 Substrate 801 Gate insulating layer 802 Hard cover 803 S i 3 N 4 Layer 803 A Spacer CG Control gate FG Floating gate H/M Hard cover Rc Contact resistance Rs Sheet resistance W Crane Electrode-72-

Claims (1)

200828425 X. Patent application scope: 1. A method for manufacturing a semiconductor device, the method comprising: forming a first conductive layer over a substrate; forming an intermediate structure above the first conductive layer, wherein the intermediate structure forms a stacked structure At least a first metal layer and a nitrogen-containing metal telluride layer; and a second conductive layer formed over the intermediate structure. 2. The method of claim 1, wherein the intermediate structure package φ is formed to sequentially stack the first metal layer, the second metal layer and the nitrogen-containing metal sand layer above the first conductive layer. 3. The method of claim 1, wherein the intermediate structure comprises forming a first metal layer, a nitrogen-containing metal telluride layer, and a second metal layer sequentially over the first conductive layer. 4. The method of claim 1, wherein the intermediate structure comprises sequentially stacking the first metal layer, the second metal layer, the nitrogen-containing metal halide layer, and the third metal layer. 5. The method of claim 1, wherein the first metal layer comprises one of a pure metal layer and a nitrogen-containing metal layer. 6. The method of claim 5, wherein the pure metal layer comprises one of a titanium layer and a tantalum layer and the nitrogen-containing metal layer comprises one of a nitrogen-containing titanium layer and a nitrogen-containing molybdenum layer. The method of clause 5 wherein the pure metal layer forms a thickness of from about 10 A to about 50 A. 8. The method of claim 5, wherein the atomic ratio of nitrogen to the metal of the nitrogen-containing metal layer -73-200828425 ranges from about 0.2 to about 0.8. 9. The method of claim 2, wherein the second metal layer comprises one of a nitrogen-containing tungsten layer and a nitrogen-containing titanium tungsten layer. 10. The method of claim 3, wherein the second metal layer comprises one of a nitrogen-containing tungsten layer and a nitrogen-containing titanium tungsten layer. 1 1. The method of claim 4, wherein each of the second metal layer and the third metal layer comprises one of a nitrogen-containing tungsten layer and a nitrogen-containing titanium tungsten layer. 12. The method of claim 1, wherein the nitrogen-containing metal telluride # layer is formed by a reactive sputtering deposition method using a metal telluride sputtering target in a nitrogen atmosphere. The method of claim 1, wherein the nitrogen-containing metal telluride layer comprises one of a nitrogen-containing tungsten telluride layer, a nitrogen-containing titanium telluride layer, and a nitrogen-containing giant telluride layer. 14. The method of claim 13 wherein the nitrogen-containing metal telluride layer has a nitrogen content of from about 10% to about 60% and the atomic ratio of germanium to metal ranges from about 0.5 to about 3.0. The method of claim 1, wherein the first conductive layer comprises one selected from the group consisting of a polysilicon layer, a polysilicon layer, and a germanide layer, and the second conductive layer comprises Tungsten layer. a method of manufacturing a semiconductor device, the method comprising: forming a first conductive layer over a substrate; forming an intermediate structure over the first conductive layer, wherein the intermediate structure is in a stacked structure, including a first metal layer, a second metal layer, a metal telluride layer, and a 'third metal layer; and -74-200828425 form a second conductive layer over the intermediate structure. 17. The method of claim 16, wherein the metal is formed by performing a reactive sputtering deposition method. 18. The method of claim 16, wherein the metal ruthenium comprises one selected from the group consisting of a tungsten ruthenide layer, a titanium ruthenide layer, and a molybdenum telluride. The method of claim 16, wherein each of the first and third metal layers comprises a nitrogen-containing metal layer. The method of claim 19, wherein each of the second metal third metal layer comprises a nitrogen-containing tungsten layer and a nitrogen-containing titanium tungsten layer _ 2 1 as claimed in claim 20 5。 The method of the nitrogen-containing tungsten is from about 10% to about 60% of the nitrogen content, and the atomic ratio of nitrogen to tungsten is from 0.3 to about 1.5. 2 2. The method of claim 19, wherein the first gold has a nitrogen to metal atomic ratio ranging from about 0.2 to about 0.8. The method of claim 19, wherein the first gold 0 comprises one of a nitrogen-containing titanium layer and a nitrogen-containing molybdenum layer. 24. The method of claim 16, wherein the first gold comprises one of a titanium layer and a molybdenum layer. 25. The method of claim 16, wherein the first conductor comprises one selected from the group consisting of a polycrystalline chopped layer, a polycrystalline clam layer, and a clathrate layer, and the second conductive layer comprises tungsten Floor. The second layer, the genus layer, and the 〇 layer have a group surrounding the cladding layer of the subordinate layer.