Method for forming a metallization layer of a semiconductor device

[0002] 1. Field of the Invention

[0003] The present invention relates to a method for forming a metallization layer of a semiconductor device.

[0004] 2. Background of the Related Art

[0005] In a metal oxide semiconductor (MOS) device, metallization layers can be used to form or to interconnect a gate electrode, source and drain impurity diffused regions, a contact region, and each device with one another. Characteristics of an electrode metallization layer depend on the device feature dimensions, and the power source voltage. The resistance of a gate electrode increases as the number of features on a given chip increase. Capacitance also increases as metallization layer pitch is reduced. This, in turn, decreases signal transmission speed, thereby reducing an operation speed of the device.

[0006] Since resistance of a contact region can increase as much as K², where K is the number of individual devices on a chip, and since current density can increase by a factor of K, reliability of a metallization layer deteriorates as K increases. In addition, the reliability of a metallization layer can deteriorate due to electromigration (EM) effects. For these reasons, a material of low resistance is typically used to form a gate electrode, which can serve as a word line of a memory. The resistivity of doped polysilicon is typically greater than 200 μΩ·cm. If polysilicon is used as a material of a gate electrode, when the device dimensions are less than 1 μm, operation speed is reduced due to a delay in signal transmission, thereby deteriorating reliability of the semiconductor device.

[0007] To reduce resistance, a tungsten silicide WSi_xfilm may be used to form a gate electrode. The tungsten silicide WSi_xfilm has excellent step coverage, and its resistivity is about 100 μΩ·cm. Also, a WSi_xfilm may be easily deposited on polysilicon. In other words, polycide (polysilicon+refractory silicide) may be used as a gate electrode. However, it is known that a WSi_xfilm having a resistivity of about 100 μΩ·cm fails to sufficiently reduce resistance if the dimensions of a gate electrode are less than about 0.5 μm. To solve the resistivity problem, the use of other metallization layer materials such as W (resistivity of 10 μΩ·cm or less), TiSi₂(resistivity of 20 μΩ·cm or less), COSi₂(resistivity of 20 μΩ·cm or less), and TiN (resistivity of 30 μΩ·cm or less) have been studied.

[0008] Further, as the packing density of a semiconductor device increases, an aspect ratio of a contact hole for electrical connection of devices increases. The increased aspect ratio has also caused problems with step coverage.

[0009] A background art method for forming a metallization layer of a semiconductor device will be described with reference to FIGS. 1A to 1C, which are sectional views illustrating process steps of forming a metallization layer of a semiconductor device. As shown in FIG. 1A, an oxide film 2 is first deposited on a semiconductor substrate 1, and a photoresist PR is deposited on the oxide film 2. A region where a contact hole will be formed is then defined, and the photoresist PR in the region is selectively patterned by exposure and developing processes. Subsequently, the oxide film 2 is selectively removed, by an etching process using the patterned photoresist PR as a mask, so that a contact hole 3 is formed to expose the semiconductor substrate 1. At this time, the oxide film 2 forms an interlayer dielectric layer (ILD).

[0010] As shown in FIG. 1B, the photoresist PR is removed. A barrier metal layer 4 of Titanium(Ti)/Titanium Nitride (TiN) is then formed on the semiconductor substrate 1 in the contact hole 3 and on an entire surface of the oxide film 2. The barrier metal layer 4 is to prevent characteristics of a metallization layer, which will be formed later, from being varied.

[0011] For contact holes having a high aspect ratio, an overhang can occur at predetermined portions on sides A of the contact hole 3, or at a corner portion B of the bottom of the contact hole 3. In other words, the barrier metal layer 4 may not be fully deposited at the comer portion B of the bottom of the contact hole 3. Such an overhang frequently occurs when forming contact holes of a memory device with a capacity of 64M or more. Thus, the oxide film 2 is often exposed at the comer portion B of the bottom of the contact hole 3.

[0012] To prevent such an overhang from occurring, a collimator may be used to fully deposit the barrier metal layer 4 on the bottom corners of the contact hole 3. Alternatively, a deposition time may be extended to ensure the barrier metal layer 4 is deposited on the corner portion B of the bottom of the contact hole 3.

[0013] As shown in FIG. 1C, a tungsten layer 5 is then formed on an entire upper surface of the barrier metal layer 4, by a chemical vapor deposition (CVD) method. In the CVD method, a chemical material, which includes atoms of a material to be deposited, is provided to a reaction chamber. The chemical material (in a gas state) is reacted with another gas in the reaction chamber to generate a desired material. The desired material is then deposited on a surface of the substrate, and unnecessary materials (including gas) are exhausted from the reaction chamber.

[0014] The process steps of forming the tungsten layer 5 will be described in detail. First, a seed layer (not shown) of amorphous silicon is formed on the barrier metal layer 4 by flowing SiH₄gas on the entire surface of the substrate. The reaction path is SiH₄(gas)→amorphous Si+2H₂(gas). During this step, SiH₄gas flows for about five seconds at 25 Standard Cubic Centimeter per Minute (SCCM).

[0015] Subsequently, both SiH₄gas and WF₆gas flow on the barrier metal layer 4, and a first portion of the tungsten layer 5 is deposited. The reaction path is 3SiH₄(gas)+2WF₆(gas)→2W+3SiF₄+6H₂(gas). The tungsten layer 5 has property of 2 W. The tungsten (2 W) has a relatively fast deposition ratio and excellent adhesion.

[0016] A second portion of the tungsten layer 5 is then deposited on the first portion of tungsten at a slower deposition rate. In this step, H₂gas and WF₆gas flow on the first portion of the tungsten layer 5, and the second portion of the tungsten layer 5 is deposited by the reaction path 3H₂(gas)+WF₆(gas)→W+6HF (gas). The second portion of the tungsten layer is deposited at a slower deposition rate to achieve better step coverage.

[0017] The background art method for forming a metallization layer of a semiconductor device has several problems. Because the barrier metal layer 4 has an overhang, when the tungsten layer 5 is deposited on the barrier metal layer 4, a void 6 may be formed in the tungsten layer 5.

[0018] Also, WF₆gas may come into contact with TiN of the barrier metal layer in the contact hole during the course of the chemical reaction that occurs when depositing the tungsten layer. The WF₆gas may react with Ti, thereby generating TiF₃gas. The TiF₃gas may cause sparking to occur, which may cause an electrical short between adjacent metallization layers if an interval between the metallization layers is small. In addition, because the substrate may remain exposed at bottom comers of the contact hole 3, WF₆gas may directly diffuse into the semiconductor substrate. As a result, leakage current may occur, and characteristics of the device are deteriorated.

[0019] A collimator may be used to reduce the overhang effect, or a deposition time of the barrier metal layer may be increased to ensure the barrier metal layer is formed across an entire bottom surface of the contact hole 3. However, if the collimator is used, the formation process time increases. Also, if the deposition time of the barrier metal layer is increased to ensure the barrier metal layer covers the bottom comers of the contact hole, the overhang problem may become worse.

[0020] If the barrier metal layer is not formed at the comer portions of the bottom of the contact hole, WF₆gas may be distributed into the interface between Ti (in the barrier metal layer) and the oxide film in the contact hole during the chemical reaction for deposition of the tungsten layer. Fluorine (F) of the WF₆gas causes poor adhesion of the interface between the oxide film and the Ti, thereby deteriorating reliability in metallization of the semiconductor device.

[0021] It is an object of the present invention to substantially obviate at least the problems or disadvantages of the background art.

[0022] Another object of the present invention is to prevent an electrical short from forming between metallization lines.

[0023] A further object of the present invention is to improve the reliability of the device.

[0024] To achieve these and other advantages and in accordance with the purpose of the present invention, a method for forming a metallization layer comprising the steps of: forming an insulating film having a contact hole on a substrate; forming a barrier layer on the insulating film, including an inner surface of the contact hole; forming a semiconductor layer on a surface of the barrier layer; forming a first conductive layer having an overhang on the semiconductor layer; removing at least a portion of the overhang of the first conductive layer; and forming a second conductive layer in the contact hole.

[0025] Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objects and advantages of the invention may be realized and attained as particularly pointed out in the appended claims.

[0026] Preferred embodiments of the invention will be described in conjunction with the following drawing figures, wherein like reference numbers refer to like elements, and wherein:

[0027] FIGS. 1A to 1C are sectional views illustrating process steps of a background art method of forming a metallization layer of a semiconductor device; and

[0028] FIGS. 2A to 2F are sectional views illustrating process steps of a method of forming a metallization layer of a semiconductor device according to the present invention.

[0029] As shown in FIG. 2A, an ILD layer 12 is deposited on a semiconductor substrate 11 and a photoresist PR is deposited on the ILD layer 12. A region where a contact hole will be formed is then defined and the photoresist PR in the region is selectively patterned by exposure and developing processes. Subsequently, the ILD layer 12 is selectively removed by an etching process, using the patterned photoresist PR as a mask, so that a contact hole 13 is formed to expose the semiconductor substrate 11. If the contact hole 13 is part of a memory device having a capacity of 64M or more, an aspect ratio of the contact hole 13 may be more than 4, and its critical dimensions may be less than 0.4 μm.

[0030] As shown in FIG. 2B, the photoresist PR is then removed. A barrier metal layer 14 of Ti/TiN is formed on an entire surface of the device, including an upper surface of the semiconductor substrate 11, and inside the contact hole 13. An overhang occurs in the barrier metal layer 14 at upper portions A of the contact hole 13. The overhang A can cause problems in forming the barrier metal layer 14 at the corner portion B of the bottom of the contact hole 13. In general, such an overhang occurs when forming any film in the contact hole. In particular, however, an overhang tends to occur more frequently when fabricating a memory device having a capacity of 64M or more. Sides of the ILD layer 12, and the substrate 11 surface at a lower portion of the ILD layer 12 are exposed at the corner portion B of the bottom of the contact hole 13.

[0031] As shown in FIG. 2C, an amorphous silicon layer 15 is then formed on the barrier metal layer 14, including inside the contact hole 13. In the background art process described above, an amorphous silicon seed layer is formed with a SiH₄gas flow of a duration of about five seconds at flow rate of 25 SCCM. However, in a preferred embodiment of the present invention, a thicker amorphous silicon layer 15 is formed on the barrier metal layer 14 by either increasing the flow rate of the SiH₄gas, or increasing the duration of the flow time of the SiH₄gas, or both. In this case, the reaction path is SiH₄(gas)→amorphous Si+2H₂(gas). The amorphous silicon layer 15 is formed on the sides of the ILD layer 12 at the bottom of the contact hole 13 where the barrier metal layer 14 will not form, and at the comer portion of the bottom of the contact hole 13. The amorphous silicon layer 15 has a thickness of approximately 200 Ř300 Å.

[0032] As shown in FIG. 2D, SiH₄gas and WF₆gas flow on the amorphous silicon layer 15, and a first tungsten layer 16 is deposited on the amorphous silicon layer 15. The reaction path is 3SiH₄(gas)+2WF₆2W+3SiF₄(gas)+6H₂(gas). Note that the amorphous silicon layer 15 and the first tungsten layer 16 can be formed by first introducing SiH₄gas into the reaction chamber for a period of time necessary to form the amorphous silicon layer 15, then also introducing the WF₆gas to form the first tungsten layer 16.

[0033] Since the tungsten layer 16 has a relatively fast deposition ratio (2 W) and excellent adhesion, an overhang occurs, as shown in FIG. 2D. The first tungsten layer 16 is deposited on the amorphous silicon layer 15 until the contact hole 13 is completely plugged.

[0034] As shown in FIG. 2E, a portion of the first tungsten layer 16 in the contact hole 13, in which overhang occurs, is then selectively removed. This is accomplished by introducing a flow of WF₆gas. The reaction path is WF₆(gas)+W→WF₅(gas). At this time, the first tungsten layer 16 over the amorphous silicon layer 15 at the top of the contact hole 13 is removed. However, when the upper portion of the contact hole 13 is completely plugged by the first tungsten layer 16, because of the overhang that occurs when the first tungsten layer 16 is deposited, unnecessary etching or removal of the first tungsten layer 16 formed in the inside of the contact hole 13 can be avoided.

[0035] As shown in FIG. 2F, a second tungsten layer 17 is then formed on the amorphous silicon layer 15 and the first tungsten layer 16. The second tungsten layer 17 is formed by the reaction path 3H₂(gas)+WF₆(gas)→W+6HF (gas) in such a manner that both WF₆gas and H₂gas flow in the reaction chamber. The process steps shown in FIGS. 2C to 2F can be carried out consecutively in the same reaction chamber.

[0036] After forming the second tungsten layer 17, portions of the second tungsten layer 17 formed at locations other than the contact hole 13 may be selectively removed to form a tungsten plug. Subsequently process step for forming a metallization layer such as aluminum or its alloy can be performed. In other words, the tungsten layers 16 and 17 may be selectively patterned by photolithography etching for use as a metallization line. Alternatively, after forming a tungsten plug of the tungsten layers 16 and 17, a metallization material layer such as an aluminum or its alloy may be formed on the tungsten plug.

[0037] The method for forming a metallization layer of a semiconductor device according to the present invention has several advantages. Since the amorphous silicon layer is formed on the entire inner surface of the contact hole, even if the barrier metal layer is not formed on the corner portion of the bottom in the contact hole, WF₆gas cannot be diffused into TiN of the barrier metal layer, or into the semiconductor substrate. This, in turn, prevents an electrical short from forming between adjacent metallization layers, even where the interval between the metallization layers is small. Further, since nothing diffuses into the substrate, leakage current does not occur, and the semiconductor device has a high reliability. If an overhang occurs during the course of process steps of forming the tungsten layer, the portion where the overhang occurs is selectively removed. Therefore, it is possible to form a metallization without overhang and without a void.

[0038] The foregoing embodiments are merely exemplary and are not to be construed as limiting the present invention. The present teaching can be readily applied to other types of apparatuses. The description of the present invention is intended to be illustrative, and not to limit the scope of the claims. Many alternatives, modifications, and variations will be apparent to those skilled in the art. In the claims, means-plus-function clauses are intended to cover the structures described herein as performing the recited function and not only structural equivalents but also equivalent structures.