CROSS REFERENCE TO RELATED APPLICATIONSThis application is a division of co-pending application Ser. No. 10/170,813 filed on Jun. 13, 2002, which claims foreign priority to Japanese Application No. 2001-178539 filed on Jun. 13, 2001. The entire content of each of these applications is hereby incorporated by reference.
BACKGROUND OF THE INVENTIONThe present invention relates to a semiconductor device having a thin film capacitor and a method for fabricating the same.
Heretofore, in a general purpose DRAM, Ta.sub.2O.sub.5 having a high dielectric constant has been considered as a capacitor dielectric film in a memory cell. In this case, it is an ordinary practice that a lower electrode of a capacitor is formed of a polysilicon layer which can be formed to have a concavo-convex surface (for example, so called HSG (hemi-spherical grain) structure) in order to increase a capacitance per a unitary area. In order to form this polysilicon layer, a high temperature process on the order of 700 to 900 degrees Celsius is required.
On the other hand, in a logic mixed DRAM in which a logic section and a memory section are formed on the same chip, gate electrodes and source/drain diffused regions in the logic section are required to be provided with a cobalt (Co) silicide for a speedup of transistors.
The cobalt silicide layer can realize a low resistance, however, if the temperature is elevated, aggregation occurs in the cobalt silicide layer so that the resistance value of the gate electrodes and the diffused layers increases. Therefore, in a process after formation of the cobalt silicide layer, the process temperature cannot be elevated. For example, in the generation of the gate length of 0.15 micron, about 600 degrees Celsius is an upper limit.
Accordingly, if Ta.sub.2O.sub.5 is used to form a capacitor dielectric film in the memory section of the logic mixed DRAM and if a polysilicon layer is used to form a lower capacitor electrode, a high temperature process is required to form the polysilicon layer, with the result that the transistors in the logic section become deteriorated through the high temperature process. Therefore, in the generation of the gate length of 0.15 micron and in succeeding generations, it is not possible to use the polysilicon for the lower capacitor electrode. Under this situation, there is a demand for constitute the electrode of the capacitor with a metal or a metal nitride, for example, TiN (titanium nitride), W (tungsten) or Ru (ruthenium), which can be formed at a low temperature of not greater than 500 degrees Celsius where no aggregation occurs in the cobalt silicide.
Now, explanation will be made on a conventional method for forming a thin film capacitor, using a metal or a metal nitride for the lower capacitor electrode and also using Ta.sub.2O.sub.5 for the capacitor dielectric film.
A lower electrode of TiN, W or Ru is formed by a CVD (chemical vapor deposition) or a PVD (physical vapor deposition), and then, a Ta.sub.2O.sub.5 capacitor dielectric film is formed by a thermal CVD process. Thereafter, in order to reduce a leakage current in the Ta.sub.2O.sub.5 capacitor, a post anneal is carried out with a RTO (rapid thermal oxidation) or a UV—O.sub.3 oxidation at a temperature of not less than 500 degrees Celsius. Furthermore, an upper electrode of TiN or another is formed by the CVD process or the PVD process, and then, a patterning is carried out to have a desired shape. Thus, the thin film capacitor of a MIM structure having the capacitor dielectric film formed of Ta.sub.2O.sub.5 is obtained.
FIG. 14A is a diagrammatic section view of a capacitor formed of a capacitor dielectric film formed of Ta.sub.2O.sub.5 and upper and lower capacitor electrodes of TiN.FIG. 14B is a graph illustrating a relation between an electrode-to-electrode voltage (Vp) and a leakage current in the structure shown inFIG. 14A.FIG. 14B shows the leakage current at temperatures of 25 degrees Celsius, 85 degrees Celsius and 125 degrees Celsius. It would be seen from this figure that the leakage current remarkably increases when the temperature becomes not less than 85 degrees Celsius which is a device operation guarantee temperature.
Furthermore, in a LSI chip, it is considered to form a high dielectric constant thin film capacitor above interconnections in the LSI, so as to use it as a decoupling capacitor. The decoupling capacitor is provided to compensate for a voltage drop which is caused for a parasite inductance existing between a power supply and interconnections of the LSI.
Referring toFIG. 15, a conventional decoupling capacitor is shown. In the prior art, as shown inFIG. 15, a number of laminatedceramic capacitors93 are located around aLSI chip92 mounted on a printedcircuit substrate91, so that those capacitors function as the decoupling capacitor. However, a resonance frequency of the laminated ceramic capacitor is on the order of about 80 MHz, and therefore, when the LSI is speeded up to several hundred MHz to several GHz, a satisfactory electric charge compensation cannot be carried out, so that it does not function as the decoupling capacitor.
FIG. 16 illustrates a thin film capacitor used as the decoupling capacitor. A high dielectric constant thin film capacitor is formed over an uppermost interconnection layer in a semiconductor device, so as to constitute an on-chip decoupling capacitor.
As shown inFIG. 16, over a wiring conductor or interconnection (ground line)94 and a wiring conductor or interconnection (power supply line)95, a lower electrode film, a capacitor dielectric film and an upper electrode film are deposited in the named order by a CVD process, and then, patterned into a desired shape, so as to form a thin film capacitor composed of alower electrode96, a capacitordielectric film97 and anupper electrode98. This thin film capacitor constitutes a decouplingcapacitor99. Here, thelower electrode96 and theupper electrode98 are formed of TiN, and the capacitordielectric film97 is formed of Ta.sub.2O.sub.5. In addition, in order to reduce a leakage current, after the capacitordielectric film97 of Ta.sub.2O.sub.5 is formed, a UV—O.sub.3 anneal is carried out at a temperature of 500 degrees Celsius.
In the case of forming a high dielectric constant thin film capacitor over the uppermost interconnection layer in the semiconductor device to use it as the decoupling capacitor, the demand of a low inductance and a large capacitance is satisfied with a one-chip feature and use of a high dielectric constant capacitor.
As mentioned above, in the conventional thin film capacitor having the lower electrode formed of TiN, W or Ru and the capacitor dielectric film of Ta.sub.2O.sub.5, it is necessary to carry out the post-anneal in an oxidizing atmosphere since the leakage current is large in a condition just after the formation of Ta.sub.2O.sub.5. However, because of this post anneal, the lower electrode layer is oxidized so that a low dielectric constant film is formed. As a result, even if the thickness of the Ta.sub.2O.sub.5 film is reduced, a large capacitance cannot be obtained.
In addition, because of the oxidation occurring in the post anneal, a concavo-convex or a peeling occurs in the lower electrode layer, with the result that the yield of production lowers.
Furthermore, as shown inFIG. 14B, although the leakage current is no problem at a room temperature of 25 degrees Celsius, if the temperature is elevated to 85 degrees Celsius and further to 125 degrees Celsius, the leakage current increases, so that a sufficient capacitance characteristics cannot be ensured at a device operation guarantee temperature.
On the other hand, in the case that a high dielectric constant thin film capacitor is formed over the uppermost interconnection layer in the semiconductor device and is used as the decoupling capacitor, the demand of a low inductance and a large capacitance is satisfied with a one-chip feature and use of a high dielectric constant capacitor, but the present method for forming the thin film capacitor has a problem.
It has been proposed to use Ta.sub.2O.sub.5, SrTiO.sub.3 and (Ba, Sr)TiO.sub.3 formed by the PVD process or the CVD process, for the capacitor dielectric film of the above mentioned thin film capacitor. However, in order to obtain a capacitor dielectric film of an excellent quality having a large capacitance in the conventional method utilizing the PVD or CVD process, a high temperature of not less than 400 degrees Celsius is required. In addition, in order to reduce the leakage current, it is also necessary to carry out the post anneal in the oxidizing atmosphere at the temperature of not less than 450 degrees Celsius.
Furthermore, recently, a copper wiring conductor is used for interconnection. Therefore, in the case that a capacitor is formed over the uppermost interconnection layer, if the temperature is elevated to not less than 450 degrees Celsius, the interconnection layer is oxidized, with the result that the characteristics is deteriorated and the yield of production lowers.
In other word, the conventional method for forming the thin film capacitor can realize a large-capacitance, low-inductance, one-chip decoupling capacitor which meet with the speedup of the LSI.
BRIEF SUMMARY OF THE INVENTIONAccordingly, it is an object of the present invention to provide a thin film capacitor and a method for forming the same, which have overcome the above mentioned problems of the prior art.
Another object of the present invention is to provide a capacitor which can realize a large capacitance and a small leakage current as a capacitor for a DRAM cell in a memory section of a semiconductor device having the memory section and a logic section formed on the same chip, without deterioration of a transistor characteristics attributable to deterioration in a silicide layer formed in gate electrodes and on source/drain diffused layer regions in the semiconductor device.
Still another object of the present invention is to provide a semiconductor device having the above mentioned capacitor formed after an interconnection layer is formed (namely, above the interconnection layer) and a method for fabricating the same.
The above and other objects of the present invention are achieved in accordance with the present invention by a semiconductor device having a capacitor of a MIM (metal-insulator-metal) structure having a capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.1, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1).
According to another aspect of the present invention, there is provided a semiconductor device including a transistor having a gate electrode and source/drain diffused layers having a refractory metal silicide, wherein a capacitor of a MIM (metal-insulator-metal) structure having a capacitor dielectric film, which is formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), and which is provided on an insulator film formed on the source/drain diffused layer.
According to still another aspect of the present invention, there is provided a semiconductor device wherein a capacitor of a MIM (metal-insulator-metal) structure having a capacitor dielectric film, which is formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1-x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), and which is provided on an insulator film formed on an interconnection.
According to a further aspect of the present invention, there is provided a method for fabricating a semiconductor device, comprising the steps of forming a lower electrode of a capacitor, forming on the lower electrode, by means of an atomic layer deposition, a capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), and, after formation of the capacitor dielectric film, carrying out a heat treatment at a temperature not lower than a film deposition temperature in the atomic layer deposition.
According to a still further aspect of the present invention, there is provided a method for fabricating a semiconductor device, comprising the steps of forming a lower electrode of a capacitor, forming on the lower electrode, by means of an atomic layer deposition, a capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), and, after formation of the capacitor dielectric film, carrying out a heat treatment at a temperature which is not lower than a film deposition temperature in the atomic layer deposition but which is not higher than a temperature where no aggregation of a refractory metal silicide occurs in a gate electrode and source/drain diffused layer regions in the case that the refractory metal silicide is provided in the gate electrode and the source/drain diffused layer regions.
According to a further aspect of the present invention, there is provided a method for fabricating a semiconductor device having a capacitor of a MIM (metal-insulator-metal) structure formed on an insulator film formed on an interconnection, comprising the steps of forming a lower electrode of the capacitor, forming on the lower electrode, by means of an atomic layer deposition, a capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), and then, forming an upper electrode on the capacitor dielectric film.
The above and other objects, features and advantages of the present invention will be apparent from the following description of preferred embodiments of the invention with reference to the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGSFIG. 1 is a diagrammatic view for illustrating the structure of one embodiment of the thin film capacitor in accordance with the present invention;
FIG. 2 is a flow chart illustrating a process for forming a ZrO.sub.2 thin film by use of an ALD apparatus;
FIG. 3 is a graph illustrating a relation between an electrode-to-electrode voltage (Vp) of a thin film capacitor and a leakage current;
FIG. 4 is a graph illustrating a relation between teq (film thickness converted into SiO.sub.2) and an actual film thickness;
FIG. 5 is a graph illustrating a relation between teq (film thickness converted into SiO.sub.2) and an actual film thickness;
FIG. 6 is a graph illustrating a relation between a leakage current and teq (film thickness converted into SiO.sub.2);
FIGS. 7 and 8 are diagrammatic sectional views for illustrating a conventional method for forming a stacked MIM capacitor;
FIGS. 9A to 9N are diagrammatic sectional views for illustrating a method for forming a stacked MIM capacitor in accordance with the present invention;
FIG. 10 is a flow chart illustrating a process for forming a TiN thin film by use of an ALD apparatus;
FIG. 11 is a diagrammatic sectional view of a planar capacitor;
FIG. 12 is a diagrammatic sectional view of a box type capacitor;
FIG. 13 is a diagrammatic sectional view illustrating a fourth embodiment of the present invention;
FIG. 14A is a diagrammatic sectional view of a conventional capacitor formed of a capacitor dielectric film formed of Ta.sub.2O.sub.5 and upper and lower capacitor electrodes of TiN;
FIG. 14B is a graph illustrating a relation between an electrode-to-electrode voltage (Vp) and a leakage current in the conventional capacitor shown inFIG. 14A;
FIG. 15 illustrates a conventional decoupling capacitor; and
FIG. 16 illustrates a conventional thin film capacitor used as the decoupling capacitor.
DETAILED DESCRIPTION OF THE INVENTIONNow, embodiments of the present invention will be described with reference to the drawings.
Referring toFIG. 1, there is shown a diagrammatic view for illustrating the structure of one embodiment of the thin film capacitor in accordance with the present invention.
The first embodiment is a thin film capacitor of a MIM (metal-insulator-metal) structure, provided in a semiconductor device. This thin film capacitor includes alower electrode1, acapacitor dielectric film2 and anupper electrode3 stacked in the named order. Each of theupper electrode3 and thelower electrode1 is formed of at least one material selected from the group consisting of a metal and a metal nitride, represented by TiN, Ti, W, WN, Pt, Ir, Ru. Thecapacitor dielectric film2 is formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1-x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), and formed by means of an atomic layer deposition (abbreviated to “ALD”).
Here, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1) is an oxide of a solid solution of Zr and Hf. (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1) is an oxide of a solid solution of Zr and Ti. (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1) is an oxide of a solid solution of Hf and Ti. (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1) is an oxide of a solid solution of Zr, Ti and Hf.
Here, a method for forming the thin film capacitor in the case that the capacitor dielectric film is formed of ZrO.sub.2, will be described. First, a lower electrode thin film having a film thickness of 5 to 50 nm is formed of at least one material selected from the group consisting of TiN, Ti, W, WN, Pt, Ir, Ru, by means of a PVD process, a CVD process or a ALD process, and then, is patterned into a desired shape, so that alower electrode1 is formed.
Then, a capacitor dielectric film is formed of ZrO.sub.2 by means of the ALD process.FIG. 2 is a flow chart illustrating a process for forming a ZrO.sub.2 thin film by use of an ALD apparatus;
The ZrO.sub.2 thin film is deposited by using ZrCl.sub.4 as a starting material for Zr and H.sub.2O as an oxygen material, under a deposition temperature of 200 to 400 degrees Celsius.
First, ZrCl.sub.4 is supplied as a starting material into a chamber of the ALD apparatus, so that only a one-atom layer is deposited on a surface of the lower electrode thin film by causing a reaction. Next, the supply of ZrCl.sub.4 is stopped, and an inert gas represented by Ar or N.sub.2 is introduced into the chamber as a purge gas so that an excessive unreacted ZrCl.sub.4 is removed.
Thereafter, H.sub.2O is supplied to replace a Cl group which terminates Zr grown on the surface of the lower electrode thin film, with an OH group of H.sub.20. In this process, HCl is produced as a reaction subproduct. Then, the supply of H.sub.2O is stopped, and the inert gas represented by Ar or N.sub.2 is introduced into the chamber as the purge gas so that an unreacted H.sub.2O and the reaction subproduct HCl are removed.
Then, ZrCl.sub.4 is supplied again so that only another one-atom layer is deposited. The supply of ZrCl.sub.4 is stopped, and the purge gas is produced so that an unreacted ZrCl.sub.4 and the reaction subproduct HCl are removed.
In this manner, a cycle of the ZrCl.sub.4 supply, the purge, the H.sub.2O supply and the purge sequentially carried in the named order, is repeated necessary times, until thecapacitor dielectric film2 of ZrO.sub.2 having the film thickness of 5 to 15 nm is obtained.
After the ZrO.sub.2 thin film is formed, an upper electrode thin film having a film thickness of 5 to 50 nm is formed of at least one material selected from the group consisting of TiN, Ti, W, WN, Pt, Ir, Ru, by means of a PVD process, a CVD process or a ALD process, and then, is patterned into a desired shape, so that anupper electrode3 is formed. Thus, the thin film capacitor is obtained.
The thin film capacitor formed as mentioned above has a small leakage current and a large capacitance, because the capacitor dielectric film is formed of ZrO.sub.2 having a high electrical insulation property and a large dielectric constant, and because a leakage current increase is small even if the film thickness of ZrO.sub.2 is reduced.
In addition, if the ZrO.sub.2 thin film is formed by the ALD process, it is possible to carry out a film formation at a low temperature and to omit a post anneal which is carried out in an oxidizing atmosphere. Therefore, it is possible to avoid the drop of the capacitance, the increase of the leakage current and the drop in the yield of production, caused by oxidation of the lower electrode.
Referring toFIG. 3, there is shown a graph illustrating a relation between an electrode-to-electrode voltage (Vp) of a thin film capacitor and a leakage current in the case that the capacitor dielectric film is formed of ZrO.sub.2 and the upper and lower electrodes are formed of TiN. From comparison ofFIG. 3 withFIG. 14B, it would be seen that the leakage current evidently becomes small in comparison with the thin film capacitor having the capacitor dielectric film formed of Ta.sub.2O.sub.5, shown inFIG. 14A.
In the above mentioned embodiment, the capacitor dielectric film is formed of ZrO.sub.2 deposited by the ALD process. However, a similar advantage can be obtained in the case that the capacitor dielectric film is formed of at least one material selected from the group consisting of HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), and (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1).
In the case that the capacitor dielectric film is formed of HfO.sub.2, HfCl.sub.4 is used as a starting material for Hf, and H.sub.2O is used as an oxygen material.
In the case that the capacitor dielectric film is formed of (Zr.sub.x, Hf.sub.1−x)O.sub.2, ZrCl.sub.4 is used as a starting material for Zr, HfCl.sub.4 is used as a starting material for Hf, and H.sub.2O is used as an oxygen material.
In the case that the capacitor dielectric film is formed of (Zr.sub.y, Ti.sub.1−y)O.sub.2, ZrCl.sub.4 is used as a starting material for Zr, TiCl.sub.4 is used as a starting material for Ti, and H.sub.2O is used as an oxygen material.
In the case that the capacitor dielectric film is formed of (Hf.sub.z, Ti.sub.1−z)O.sub.2, HfCl.sub.4 is used as a starting material for Hf, TiCl.sub.4 is used as a starting material for Ti, and H.sub.2O is used as an oxygen material.
In the case that the capacitor dielectric film is formed of (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2, ZrCl.sub.4 is used as a starting material for Zr, TiCl.sub.4 is used as a starting material for Ti, HfCl.sub.4 is used as a starting material for Hf, and H.sub.2O is used as an oxygen material.
Now, a second embodiment of the present invention will be described.
According to this second embodiment, in a thin film capacitor of a MIM structure having a capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), deposited by means of the ALD process, an anneal is carried out after formation of the capacitor dielectric film.
The thin film capacitor of the MIM structure in accordance with the first embodiment having the capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), has a small leakage current and a large capacitance, in comparison with the thin film capacitor having the capacitor dielectric film formed of Ta.sub.2O.sub.5, and therefore, can be satisfactorily used as a memory capacitor in a DRAM cell. However, when the film thickness of the capacitor dielectric film is decreased, the capacitance increase is small (namely, the decrease of teq is small), and on the other hand, the leakage current increases.
Referring toFIG. 4, there is shown a graph illustrating a relation between teq (film thickness converted into SiO.sub.2) and an actual film thickness when the capacitor dielectric film formed of ZrO.sub.2 and the upper and lower electrodes are formed of TiN. A capacitance drop caused by a low dielectric constant layer contribution would be seen fromFIG. 4.
The reason for this is considered as follows: Just after the deposition of the ZrO.sub.2 thin film, a ZrO.sub.2 boundary layer having poor crystallinity exists, which act as a low dielectric constant layer. Therefore, even if the film thickness is decreased, the low dielectric constant layer contribution is large, with the result that the capacitance of the capacitor as a whole cannot be correspondingly increased. On the other hand, the leakage current increases with the decrease of the film thickness.
In this second embodiment, on the other hand, in the thin film capacitor formed in accordance with the first embodiment, namely, in the thin film capacitor of the MIM structure having the capacitor dielectric film formed of the ZrO.sub.2 thin film formed by the ALD process, the anneal is carried out at a temperature of 300 to 700 degrees Celsius after the deposition of the ZrO.sub.2 thin film. As a result, a further large capacitance and a small leakage current are realized in the thin film capacitor.
As mentioned above, the thin film capacitor having the ZrO.sub.2 thin film deposited on a metal electrode by the ALD process can be used as a memory capacitor in the DRAM cell. However, in order to meet with the demand for a further microminiaturization and a further high integration, it is preferred to obtain a further large capacitance and a small leakage current.
The inventors of the present application uncovered that if an ZrO.sub.2 thin film is only deposited on a metal electrode by the ALD process, an amorphous ZrO.sub.2 layer region exists at a boundary between the lower electrode and the ZrO.sub.2 layer, and it is not possible to obtain a characteristics intrinsic to the MIM structure thin film capacitor having the capacitor dielectric film of ZrO.sub.2. Namely, since the amorphous layer functions as a low dielectric constant layer in the MIM structure thin film capacitor, the obtained capacitance drops.
Furthermore, the inventors of the present application uncovered and confirmed that the amorphous ZrO.sub.2 layer is crystallized by carrying out the anneal after the deposition of the ZrO.sub.2 layer. With this crystallization, the ZrO.sub.2 layer is homogenized, so that the above mentioned low dielectric constant boundary layer is removed. As a result, it was confirmed that a further large capacitance and a small leakage current are obtained in comparison with the case that no anneal was carried out.
For example, the ZrO.sub.2 capacitor dielectric film is deposited on the lower electrode of TiN by the ALD process at a temperature of 200 to 400 degrees Celsius, and thereafter, for example, the upper electrode of TiN is formed, and then, is patterned into a desired shape, so that the MIM thin film capacitor is obtained. Thereafter, the MIM thin film capacitor thus obtained is annealed at a temperature which is not lower than a ZrO.sub.2 layer deposition temperature and which is in a range of 300 to 700 degrees Celsius.
Incidentally, the above mentioned amorphous layer is formed when the capacitor dielectric film is deposited, and the anneal may be carried out at any time after the deposition of the capacitor dielectric film. For example, a similar advantage can be obtained even if the anneal is carried out immediately after the deposition of the capacitor dielectric film or even if the anneal is carried out after the formation of the upper electrode.
In addition, the atmosphere in which the anneal is carried out is not important, but it is preferable to use a non-oxidizing atmosphere which does not cause a characteristics deterioration attributable to oxidation of the low electrode material, namely, N.sub.2, Ar, He, or a forming gas (H.sub.2+N.sub.2).
Referring toFIG. 5, there is shown a graph illustrating a relation between teq (film thickness converted into SiO.sub.2) and an actual film thickness in the case that the capacitor dielectric film is formed of ZrO.sub.2 and the upper and lower electrodes are formed of TiN. It could be seen fromFIG. 5 that, if the anneal is carried out, the low dielectric constant layer contribution becomes null. InFIG. 5, a black solid circle indicates the thin film capacitor formed with no anneal, which is the same as the thin film capacitor shown inFIG. 4. A white circle indicates the thin film capacitor formed by carrying out the anneal under the atmosphere of a mixed gas of hydrogen and nitrogen, and a triangle indicates the thin film capacitor formed by carrying out the anneal under the atmosphere of only nitrogen.
Referring toFIG. 6, there is shown a graph illustrating a relation between a leakage current and teq (film thickness converted into SiO.sub.2) in the case that the capacitor dielectric film is formed of ZrO.sub.2 and the upper and lower electrodes are formed of TiN. InFIG. 6, a white circle indicates the thin film capacitor formed with no anneal. A square indicates the thin film capacitor formed by carrying out the anneal under a mixed gas of hydrogen and nitrogen, and a triangle indicates the thin film capacitor formed by carrying out the anneal under only a nitrogen. It could be seen fromFIG. 6 that the case of carrying out the anneal under the mixed gas of hydrogen and nitrogen and the case of carrying out the anneal under only the nitrogen exhibit substantially the same leakage current value. Therefore, it would be seen that only the heat treatment is effective, and the atmosphere for the anneal does not give any influence.
As seen from the above, in the thin film capacitor of the second embodiment, the low dielectric constant layer (ZrO.sub.2 boundary layer having poor crystallinity) existing just after the deposition of the ZrO.sub.2 thin film, is improved in crystallinity by the anneal, so that it no longer functions as the low dielectric constant layer. As a result, the capacitance is increased (namely, teq is decreased). In addition, since crystallinity is improved in the whole of the ZrO.sub.2 thin film, the leakage current is decreased.
In the above mentioned second embodiment, the capacitor dielectric film is formed of ZrO.sub.2. However, a similar advantage can be obtained in the case that the capacitor dielectric film is formed of at least one material selected from the group consisting of HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), and (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1).
In addition, when the thin film capacitor of the second embodiment is formed in a semiconductor device having a transistor in which a refractory metal silicide is provided in a gate electrode and source/drain diffused layer regions, the anneal condition is required at a temperature which is not lower than the ZrO.sub.2 deposition temperature in the ALD process but which is not higher than a temperature where no aggregation of the refractory metal silicide occurs in the gate electrode and the source/drain diffused layer regions. The aggregation of the silicide is remarkable in a region having a small area, namely, in the gate electrode rather than the source/drain diffused layer regions. For example, in the device on the generation of the gate length 0.15 micron, the aggregation temperature is about 600 degrees Celsius. In this case, the anneal temperature is not lower than the ZrO.sub.2 deposition temperature in the ALD process but not higher than 600 degrees Celsius.
Now, a third embodiment of the present invention will be described.
According to this third embodiment, in a stacked MIM (metal-insulator-metal) capacitor in a DRAM or a logic mixed DRAM having a logic section and a memory section formed on the same chip, a lower electrode, a capacitor dielectric film and an upper electrode are sequentially formed in the ALD process by use of an ALD apparatus.
First, a conventional method for forming the stacked MIM capacitor in a DRAM or a logic mixed DRAM will be described. As shown inFIG. 7, a transistor is formed, and after acapacitor contact11 is formed, aninterlayer insulator film12 is deposited. Then, an opening is formed in theinterlayer insulator film12 by use of lithography, and a lower electrode (metal)13 is deposited. Thereafter, a resist14 is filled into the opening to protect the opening, and only an upper portion of the interlayer insulator film is removed by an etch-back process or a CMP (chemical mechanical polishing) process, so that capacitors are separated from each other. Then, as shown inFIG. 8, acapacitor dielectric film15 and anupper electrode16 are deposited, and a patterning is carried out to form a common electrode conductor (capacitor plate).
In this conventional method for forming the stacked MIM capacitor, when the lower electrode is selectively removed, it is necessary to charge the resist14 into the opening so as to prevent a portion which will become the lower electrode of the capacitor, from the etching. For the purpose of removing this resist, it is possible to remove this resist by using acid, in a MIS (metal-insulator-silicon) capacitor having the lower electrode formed of polysilicon. However, in the MIM capacitor having the lower electrode formed of a metal such as TiN, it is impossible to remove this resist by using acid (SPM (sulfuric acid—peroxide mixture). Therefore, the resist is removed by a plasma removal processing plus an organic removal processing. However, this removal method is difficult to completely remove depositions which were generated in the etching and a removal residue of the resist.
Furthermore, in conventional method for forming the stacked MIM capacitor, it is difficult to avoid a plasma damage to the surface of the lower electrode when the lower electrode is selectively removed and when the resist is removed. In addition, since impurity such as carbon in atmosphere of a clean room deposits on the surface of the lower electrode, it is also difficult to maintain a boundary between the lower electrode and the capacitor dielectric film in a good condition, with the result that the characteristics of the capacitor dielectric film is deteriorated.
In the third embodiment of the present invention, on the other hand, in a thin film capacitor of a MIM structure in which each of an upper electrode and a lower electrode is formed of at least one material selected from the group consisting of a metal and a metal nitride, represented by TiN, Ti, W, WN, Pt, Ir, Ru, and a capacitor dielectric film is formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), the lower electrode, the capacitor dielectric film and the upper electrode are sequentially formed in the same machine in the ALD process by using an ALD apparatus. With this arrangement, a boundary between the lower electrode and the capacitor dielectric film can be maintained in a good condition.
Now, a method in accordance with the third embodiment of the present invention for forming the stacked MIM capacitor in the DRAM or the logic mixed DRAM, will be described with reference to diagrammatic sectional views ofFIGS. 9A to 9N. Here, explanation will be made on a cylinder type capacitor, which is one kind of the stacked capacitor, and which includes a capacitor dielectric film formed of ZrO.sub.2 and upper and lower electrodes formed of TiN.
First, as shown inFIG. 9A, adevice isolation region22 is formed on a principal surface of a semiconductor substrate (silicon substrate)21 to confirm a device formation region. In the device formation region, agate electrode23 is formed on a not-shown gate insulator film, and a source/drain diffusedregion24 is formed in a surface region of the substrate at each side of the gate electrode. This gate electrode23 forms a word line in the DRAM. Then, aside wall25 is formed on each side of thegate electrode23, and Co (cobalt) or Ni (nickel) is deposited on thegate electrodes23 and the source/drain diffusedregions24 to form a cobalt silicide or a nickel silicide. Thus, MOS transistors are formed.
Then, as shown inFIG. 9B, aninterlayer insulator film26 is formed to cover the transistors, and an upper surface of theinterlayer insulator film26 is planarized.
As shown inFIG. 9C, a hole for a cell contact (capacitor contact)27 and a hole for a cell contact (bit contact)28 are formed in the interlayer insulator film to reach the source/drain diffusedregions24 by a lithography, and W (tungsten) is charged into the holes thus formed, to form a W plug.
As shown inFIG. 9D, a bit line conductor layer is formed on theinterlayer insulator film26 and thecell contacts27 and28, and is patterned to form a bit line electrically connected to the W plug of thecell contact28.
As shown inFIG. 9E, aninterlayer insulator film30 is deposited to cover thebit line29, and then, and an upper surface of theinterlayer insulator film30 is planarized.
As shown inFIG. 9F, a hole for acapacitor contact31 is formed in theinterlayer insulator film30 to reach thecell contact27 by a lithography, and W (tungsten) is charged into the hole thus formed, to form a W plug.
As shown inFIG. 9G, aninterlayer insulator film32 is formed to cover theinterlayer insulator film30 and thecapacitor contact31.
As shown inFIG. 9H, acylinder33 is formed in theinterlayer insulator film32 to reach thecapacitor contact31 by a lithography.
As shown inFIG. 9I, by using the ALD apparatus, alower electrode metal34 of TiN, acapacitor dielectric film35 of ZrO.sub.2 and anupper electrode metal36aof TiN are continuously formed in the named order within the same machine without being exposed to air.
Here, referring toFIG. 10, there is shown a flow chart illustrating a process for forming a TiN film which constitutes thelower electrode metal34 and theupper electrode metal36a, by use of the ALD apparatus.
For formation of the TiN film, TiCl.sub.4 and NH.sub.3 are used as a starting material gas. A film deposition temperature is on the order of 300 to 500 degrees Celsius.
First, TiCl.sub.4 is supplied to a chamber of the ALD apparatus. Thus, only a one-atom layer is deposited, by reaction, on a surface of theinterlayer insulator film32 including the surface of the cylinder formed in theinterlayer insulator film32. Next, the supply of TiCl.sub.4 is stopped, and a purge gas is introduced into the chamber so that an excessive unreacted TiCl.sub.4 is removed.
Thereafter, NH.sub.3 is supplied to replace a Cl group which terminates Ti grown on the surface of the interlayer insulator film, with a NH.sub.2 group. In this process, HCl is produced as a reaction subproduct. Then, the supply of NH.sub.3 is stopped, and the inert gas represented by Ar or N.sub.2 is introduced into the chamber as the purge gas so that an unreacted NH.sub.3 and the reaction subproduct HCl are removed.
Then, TiCl.sub.4 is supplied again so that only another one-atom layer is deposited. The supply of TiCl.sub.4 is stopped, and the purge gas in introduced so that an unreacted TiCl.sub.4 and the reaction subproduct HCl are removed. The supply of the purge gas is stopped, and NH.sub.3 is supplied.
In this manner, a cycle of the TiCl.sub.4 supply, the purge, the NH.sub.3 supply and the purge sequentially carried in the named order, is repeated necessary times, until the lowerelectrode metal film34 of TiN having the film thickness of 5 to 50 nm is obtained.
Thereafter, similarly to the procedure of the formation of the ZrO.sub.2 thin film in the first embodiment as shown inFIG. 2, thecapacitor dielectric film35 of ZrO.sub.2 having the film thickness of 5 to 15 nm is formed on the lowerelectrode metal film34 by alternately supplying ZrCl.sub.4 and H.sub.20.
Furthermore, the upperelectrode metal film36aof TiN having the film thickness of 5 to 50 nm is formed on thecapacitor dielectric film35 by alternately supplying TiCl.sub.4 and NH.sub.3, similarly to the procedure of the formation of the lowerelectrode metal film34 as shown inFIG. 10.
In the example shown inFIG. 9I, an upper electrode metal36bformed of W (tungsten) is deposited on the upperelectrode metal film36aso that the upper electrode is formed of a double layer consisting of a TiN layer and a W layer. In this case, the W layer is not necessarily required to be formed by use of the ALD process, but can be formed by use of a conventional CVD process or a sputtering. The ALD process expends a time since it is necessary to alternately supply the different gases. Use of the CVD process or the sputtering is effective for a mass production of the device.
This can be applied to thelower electrode34. Namely, it is not necessary to form the whole of thelower electrode34 by use of the ALD process. Thelower electrode34 can be formed first by the sputtering, for example, and then, only a portion of thelower electrode34 corresponding to the boundary layer is formed by the ALD process, and thereafter, thecapacitor dielectric film35 and theupper electrode36aare succeedingly formed by the ALD process.
In other words, thelower electrode34 and theupper electrode36aare sufficient if only a boundary portion to thecapacitor dielectric film35 is formed by the ALD process. Therefore, it is sufficient if the film thickness of each of thelower electrode34 and theupper electrode36ato be formed by the ALD process has at least one-atom layer thickness. Accordingly, only a boundary portion to thecapacitor dielectric film35, of each of thelower electrode34 and theupper electrode36a, is formed by the ALD process to have at least one-atom layer thickness, and the other portion of each of thelower electrode34 and theupper electrode36acan be formed by use of the CVD process or the sputtering so that the whole film thickness of each of thelower electrode34 and theupper electrode36abecomes 5 to 50 nm.
Thereafter, as shown inFIG. 9J, by a patterning using the CMP, the etch-back or the lithography, the stacked structure thus formed is divided into a plurality of individualcylinder type capacitors37 each formed of thelower electrode34, thecapacitor dielectric film35 and theupper electrode36.
Succeedingly, as shown inFIG. 9K, an insulatingfilm38 is deposited on the whole surface to isolate thelower electrode34 and theupper electrode36 from each other.
As shown inFIG. 9L, anopening39 is formed through the insulatingfilm38 at only a position above theupper electrode36 and to reach theupper electrode36. At this time, theopening39 never extends to thelower electrode34.
As shown inFIG. 9M, acommon interconnection layer40 is formed to cover the upper surface and to charge into theopening39 so that theupper electrodes36 are connected in common to thecommon interconnection layer40.
Furthermore, as shown inFIG. 9N, aninterlayer insulator film41 is formed to cover the upper surface and thecommon interconnection layer40, and afirst metal layer42 is formed on theinterlayer insulator film41 so as to constitute a first level interconnection.
In this third embodiment, since the lower electrode, the capacitor dielectric film and the upper electrode are continuously formed in the same machine (the same chamber) by use of the ALD apparatus which can control the composition of the film with an atom-layer level, it is possible to perfectly prevent the chemical and physical damages to the surface of the lower electrode, which would otherwise occur when the lower electrode is patterned and when the resist is removed. In addition, it is also possible to minimize deposition of carbons included in the air within the clean room, to the surface of the lower electrode and the surface of the capacitor dielectric film. Because of these reasons, it is possible to maintain a boundary between the lower electrode and the capacitor dielectric film and a boundary between the capacitor dielectric film and the upper electrode in a good condition. Furthermore, since the boundary between the lower electrode and the capacitor dielectric film is maintained in the good condition, it is possible to minimize the capacitance drop and the increase of the capacitor dielectric film leakage current.
In addition, after the MIM structure capacitor is formed, if the anneal is carried out similarly to the capacitor of the second embodiment, it is possible to realize a capacitor having a further large capacitance and a small leakage current. In this case, the anneal is carried out at a temperature which is not lower than the ZrO.sub.2 forming temperature in the ALD process but which is not higher than a temperature where no aggregation of a refractory metal silicide occurs in a gate electrode and source/drain diffused layer regions in the case that the refractory metal silicide is provided in the gate electrode and the source/drain diffused layer regions.
The present invention is in now way limited to the cylinder type capacitor, but can be applied to a planar capacitor and a box type capacitor.
Referring toFIG. 11, there is shown a diagrammatic sectional view of a planar capacitor to which the present invention can be applied. Adevice isolation region52 is formed on a principal surface of a semiconductor substrate (silicon substrate)51 to confirm a device formation region. In the device formation region, agate electrode53 is formed on a not-shown gate insulator film, and a source/drain diffusedregion54 having a cobalt silicide is formed in a surface region of the substrate at each side of the gate electrode. Aninterlayer insulator film55 is formed to cover thegate electrode53 and the principal surface of the substrate.
Abit contact56 is formed in theinterlayer insulator film55 to reach one of each pair of source/drain diffusedregions54. Abit line57 is formed on theinterlayer insulator film55 to be electrically connected to thebit contact56. Furthermore, aninterlayer insulator film58 is deposited to cover thebit line57 and theinterlayer insulator film55.
On theinterlayer insulator film58, alower electrode59 formed of a metal of a metal nitride, acapacitor dielectric film60 formed of ZrO.sub.2, and anupper electrode61 formed of a metal of a metal nitride, are continuously formed in the named order by the ALD apparatus. Thelower electrode59 is electrically connected to acapacitor contact62 formed through theinterlayer insulator films58 and55 to reach the other of each pair of source/drain diffusedregions54. Aside wall66 is formed on a side surface of each planar capacitor thus formed which is composed of thelower electrode59, thecapacitor dielectric film60 and theupper electrode61.
Acommon interconnection layer63 is formed on theupper electrode61 of each planar capacitor to interconnect theupper electrode61 of the planar capacitors. Aninterlayer insulator film64 is formed to cover thecommon interconnection layer63, and afirst metal layer65 is formed on theinterlayer insulator film64 to constitute a first level metal interconnection.
Referring toFIG. 12, there is shown a diagrammatic sectional view of a box type capacitor to which the present invention can be applied. Adevice isolation region72 is formed on a principal surface of a semiconductor substrate (silicon substrate)71 to confirm a device formation region. In the device formation region, agate electrode73 is formed on a not-shown gate insulator film, and a source/drain diffusedregion74 having a cobalt silicide is formed in a surface region of the substrate at each side of the gate electrode. Aninterlayer insulator film75 is formed to cover thegate electrode73 and the principal surface of the substrate.
Abit contact76 is formed in theinterlayer insulator film75 to reach one of each pair of source/drain diffusedregions74. Abit line77 is formed on theinterlayer insulator film75 to be electrically connected to thebit contact76. Furthermore, aninterlayer insulator film78 is deposited to cover thebit line77 and theinterlayer insulator film75. Acapacitor contact82 is formed through theinterlayer insulator films78 and75 to reach the other of each pair of source/drain diffusedregions74.
On thecapacitor contact82, a lower electrode79aof W (tungsten) is formed. To cover the lower electrode79aand theinterlayer insulator film78, a lower electrode79bformed of a metal of a metal nitride, acapacitor dielectric film80 formed of ZrO.sub.2, and anupper electrode81 formed of a metal of a metal nitride, are continuously formed in the named order by the ALD apparatus. Thus, the box type capacitor is composed of the lower electrodes79aand79b, thecapacitor dielectric film80 and theupper electrode81.
On theupper electrode81, aninsulator film86 is formed to isolate theupper electrode81 and thelower electrode79 from each other. On the aninsulator film86, acommon interconnection layer83 is formed to interconnect the upper electrode81bof the planar capacitors. Aninterlayer insulator film84 is formed to cover thecommon interconnection layer83, and afirst metal layer85 is formed on theinterlayer insulator film84 to constitute a first level metal interconnection.
Now, a fourth embodiment of the present invention will be described.
In this fourth embodiment, in a semiconductor device having a MIS thin film capacitor formed on an insulator film formed to over an uppermost interconnection layer, a capacitor dielectric film of the MIS thin film capacitor is formed by the ALD process which can carry out a film deposition at a low temperature and which makes the post-anneal in an oxidizing atmosphere unnecessary, and the MIS thin film capacitor is located to function as a decoupling capacitor for power supply lines.
Referring toFIG. 13, there is shown a diagrammatic sectional view of a part of a semiconductor device, for illustrating the fourth embodiment of the present invention.
Aninterlayer insulator film103 is formed to cover an uppermost interconnection (ground line)101, another uppermost interconnection (power supply line)102 and an underlying interlayer insulator film. On theinterlayer insulator film103, alower electrode105, acapacitor dielectric film106 and anupper electrode107 are formed in the named order to constitute adecoupling capacitor104. Thelower electrode105 is connected through acontact108 to the uppermost interconnection (ground line)101, and theupper electrode107 is connected through acontact109 to the uppermost interconnection (power supply line)102.
Now, a method for forming the decoupling capacitor shown inFIG. 13 will be described.
In a logic device formed in a known fabricating process, on theinterlayer insulator film103, a lower electrode film is formed of at least one material selected from the group consisting of TiN, Ti, TaN, Ta, W, WN, Pt, Ir, Ru, by means of a sputtering process or a ALD process, and then, is patterned into a desired shape, so that thelower electrode105 is formed at a position above the uppermost interconnection (ground line)101 and is connected to thecontact108 reaching the uppermost interconnection (ground line)101.
Thereafter, a capacitor dielectric film is deposited to cover thelower electrode105 by use of the ALD process at a film deposition temperature of 200 to 400 degrees Celsius. This capacitor dielectric film is constituted of a single-layer film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), or alternatively, a multi-layer film formed of at least two materials selected from the group mentioned above. Then, the capacitor dielectric film is patterned into a desired shape to form thecapacitor dielectric film106.
Then, to cover thecapacitor dielectric film106, an upper electrode film is formed of at least one material selected from the group consisting of TiN, Ti, TaN, Ta, W, WN, Pt, Ir, Ru, by means of a sputtering process or a ALD process, and then, is patterned into a desired shape, so that theupper electrode107 is formed to partially cover thecapacitor dielectric film106 and is connected to thecontact109 reaching the uppermost interconnection (power supply line)102. Thus, the thin film capacitor functioning as the decoupling capacitor is formed in the semiconductor device.
In the embodiment shown inFIG. 13, the lower electrode is connected to the uppermost interconnection (ground line) and the upper electrode is connected to the uppermost interconnection (power supply line). However, the present invention is in no way limited to this fashion. It is a matter of course to persons skilled in the art that a similar advantage can be obtained in the case that the lower electrode is connected to the uppermost interconnection (power supply line) and the upper electrode is connected to the uppermost interconnection (ground line).
In addition, in the embodiment shown inFIG. 13, the thin film capacitor is formed over the uppermost interconnection of the semiconductor device to function as the decoupling capacitor. However, the thin film capacitor can be formed within the inside of the semiconductor device or on a lower surface of the semiconductor device.
As mentioned above, in the fourth embodiment, since a capacitor dielectric film having a high dielectric constant is formed by the ALD process which can carry out a film deposition at a low temperature and which makes the post-anneal in an oxidizing atmosphere unnecessary, the thin film capacitor can be formed in the semiconductor device with no characteristics deterioration attributable to oxidation of the interconnection layer and with no drop in yield of production.
Incorporation of the thin film capacitor in accordance with the present invention into the semiconductor device as the decoupling capacitor can overcome the problem mentioned hereinbefore of the conventional on-chip decoupling capacitor and can simultaneously realize a low inductance and a large capacitance which are advantages of the on-chip decoupling capacitor.
As mentioned above, in the MIM thin film capacitor in accordance with the present invention, since a capacitor dielectric film is formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), it is possible to reduce a leakage current and to increase a capacitance value.
Furthermore, in the MIM thin film capacitor in accordance with the present invention, since an anneal is carried out after a capacitor dielectric film is formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1-x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), it is possible to further reduce a leakage current and to further increase a capacitance value.
In the stacked MIM thin film capacitor in accordance with the present invention provided in a DRAM or a logic mixed DRAM, since a lower electrode, a capacitor dielectric film and an upper electrode are continuously formed in the ALD process, a silicide layer formed in a gate electrode and a source/drain diffused region is never deteriorated, and a sufficient capacitance value (a maximum capacitance value per a unitary area and a minimum leakage current value per a unitary area) can be ensured as a capacitor of each DRAM cell provided in a memory section.
Moreover, according to the present invention, since a thin film capacitor having a capacitor dielectric film formed of at least one material selected from the group consisting of ZrO.sub.2, HfO.sub.2, (Zr.sub.x, Hf.sub.1−x)O.sub.2 (0<x<1), (Zr.sub.y, Ti.sub.1−y)O.sub.2 (0<y<1), (Hf.sub.z, Ti.sub.1−z)O.sub.2 (0<z<1), (Zr.sub.k, Ti.sub.i, Hf.sub.m)O.sub.2 (0<k, l, m<1, k+l+m=1), is formed over an uppermost interconnection layer in a semiconductor device, it is possible to realize a low-inductance, large-capacitance, one-chip decoupling capacitor, which meets with a speedup of an LSI.
The invention has thus been shown and described with reference to the specific embodiments. However, it should be noted that the present invention is in no way limited to the details of the illustrated structures but changes and modifications may be made within the scope of the appended claims.