Antireflection coating and semiconductor device manufacturing method

Abstract
An antireflection coating has two-layer structure including lower and upper silicon nitride films (p-SiN films) formed by plasma CVD. For the lower p-SiN film, the real part of its complex index of refraction is set in the range not less than 1.9 nor more than 2.5, the imaginary part is set in the range of not less than 0.9 nor more than 1.7, and the film thickness is set in the range of not less than 20 nm nor more than 60 nm. For the upper p-SiN film, the real part of its complex index of refraction is set in the range not less than 1.7 nor more than 2.4, the imaginary part is set in the range of not less than 0.15 nor more than 0.75, and the film thickness is set in the range of not less than 10 nm nor more than 40 nm.
Description




BACKGROUND OF THE INVENTION




1. Field of the Invention




The present invention relates to an antireflection coating used in a semiconductor device manufacturing process and to a semiconductor device manufacturing method including a process of forming the same.




2. Description of the Background Art




In general, the interconnection in semiconductor devices is formed by pattern-etching a film of interconnection conductor such as metal, silicide, etc. The method in which resist is applied directly on the interconnection conductor film and the interconnection conductor film is directly etched using the resist as a mask is no longer in common use; the following method is widely used instead. That is to say, a silicon oxide film (also called an oxide film hereinafter) or a silicon nitride film (also called a nitride film hereinafter) is formed on the interconnection conductor film and the oxide film or the like is first patterned by using the resist. Subsequently the interconnection conductor film is pattern-etched by using the patterned oxide film etc. as a mask.




In this process, an antireflection coating (ARC) is provided to prevent the exposure light for exposing the resist from reflecting at a film lying under the resist.

FIG. 45

is a schematic sectional view showing a semiconductor device (or a semiconductor substrate)


101


P, which is used to explain a conventional antireflection coating


10


P. The semiconductor device


101


P has a tungsten silicide (WSi) film


4


P for the gate electrode, an oxide film


3


P, a nitride film


2


P, an antireflection coating


10


P and a resist


1


P stacked in this order. The nitride film


2


P is made of stoichiometric silicon nitride (Si


3


N


4


).




The antireflection coating


10


P is composed of a single layer of silicon nitride formed by plasma CVD (Chemical Vapor Deposition: hereinafter the silicon nitride film is referred to also as a p-SiN film). The antireflection coating


10


P is also called a p-SiN film


10


P hereinafter using the same reference character. In order to allow the p-SiN film


10


P to function as the antireflection coating, it is preferred that its complex index of refraction (n−i×k) approximately satisfy the conditions 1.5≦n≦2.5, 0.2≦k≦1.5. The characters n, k and i are, respectively, the real part and the imaginary part of the complex index of refraction and the imaginary number (imaginary unit).




It is assumed here that, for the p-SiN film


10


P, n=2.13, k=0.83, and its film thickness is 47.5 nm (475 angstroms).





FIG. 46

is a contour diagram showing the results of calculation (simulation) of the reflectance R of the exposure light in the semiconductor device


101


P using the p-SiN film


10


P. The horizontal and vertical axes in

FIG. 46

respectively show the film thickness of the oxide film


3


P and that of the nitride film


2


P in the semiconductor device


101


P. The reflectance R was calculated considering the multiple reflection of the light, so that the calculated results almost perfectly agree with experimental results. While the calculation of the reflectance R assumes that the exposure light is at a wavelength of 248 nm, similar calculation results are obtained at other wavelengths.




It can be seen from

FIG. 46

that the reflectance R varies between about 2 and 8% with the film thickness variations of the oxide film


3


P and the nitride film


2


P. The variation of the reflectance R directly affects the patterning dimensions of the resist. This is described below.





FIG. 47

is a schematic sectional view showing another semiconductor device


102


P used to explain the antireflection coating


10


P. The semiconductor device


102


P has a polycrystalline silicon film


5


P for the gate electrode, the oxide film


3


P, the p-SiN film


10


P, the resist


1


P and a top antireflection coating (TARC)


6


P stacked in this order.





FIG. 48

is a graph showing the results of calculation of the resist patterning dimension in the semiconductor device


102


P using the p-SiN film


10


P. The horizontal and vertical axes in

FIG. 48

respectively show the thickness of the oxide film


3


P of the semiconductor device


102


P and the patterning dimension (line width) of the resist


1


P;

FIG. 48

shows the results of calculation at varied exposure light energies. The conditions used for exposure of the resist


1


P are: NA (Numerical Aperture)=0.65; annular ⅔; exposure mask L/S (Line and Space) pattern 200 nm. It can be seen from

FIG. 48

that the dimension of the resist


1


P varies by 17 nm as the film thickness of the oxide film


3


P varies.




While the two semiconductor devices


101


P and


102


P differ in detailed structure, they are common in that a transparent insulating film is provided under the antireflection coating


10


P. Therefore the description above applies to both of the semiconductor devices


101


P and


102


P.




Achieving further size reduction of the semiconductor devices requires finer dimensional accuracy. For example, a semiconductor device with a line width of 130 nm is required to satisfy a total dimensional accuracy of about ±10% (13 nm), so that the above-described variation of the resist patterning dimension, caused by the film thickness variations of the oxide film


3


P and nitride film


4


P, must be restricted to about ±5 nm. That is to say, the existing process is insufficient.




SUMMARY OF THE INVENTION




A first aspect of the present invention is directed to an antireflection coating provided between an underlying layer and a resist in a semiconductor device manufacturing process. According to the first aspect, the antireflection coating comprises: a lower plasma silicon nitride film provided on the underlying layer, the lower plasma silicon nitride film being formed by a plasma chemical vapor deposition and containing more silicon than stoichiometric silicon nitride (Si


3


N


4


); and an upper plasma silicon nitride film provided on the lower plasma silicon nitride film, the upper plasma silicon nitride film being formed by the plasma chemical vapor deposition and containing more silicon than the stoichiometric silicon nitride (Si


3


N


4


); wherein the lower plasma silicon nitride film has a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.9 nor more than 2.5 and in the range of not less than 0.9 nor more than 1.7, and the lower plasma silicon nitride film has a thickness in the range of not less than 20 nm nor more than 60 nm, and wherein the upper plasma silicon nitride film has a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.7 nor more than 2.4 and in the range of not less than 0.15 nor more than 0.75, and the upper plasma silicon nitride film has a thickness in the range of not less than 10 nm nor more than 40 nm.




Preferably, according to a second aspect, in the antireflection coating, the complex index of refraction of the lower plasma silicon nitride film is (2.1−i×1.2) and the thickness of the lower plasma silicon nitride film is 50 nm, and the complex index of refraction of the upper plasma silicon nitride film is (2.0−i×0.3) and the thickness of the upper plasma silicon nitride film is 30 nm.




Preferably, according to a third aspect, the antireflection coating further comprises a silicon oxide film formed on the upper plasma silicon nitride film.




Preferably, according to a fourth aspect, in the antireflection coating, the silicon oxide film has a thickness of about 4 to 10 nm.




According to a fifth aspect, a semiconductor device manufacturing method comprises the steps of: (a) forming an underlying layer; (b) forming an antireflection coating on the underlying layer; and (c) forming a resist on the antireflection coating and patterning the resist through exposure and developing; the step (b) comprising the steps of (b-1) forming a lower plasma silicon nitride film on the underlying layer by a plasma chemical vapor deposition, the lower plasma silicon nitride film containing more silicon than stoichiometric silicon nitride (Si


3


N


4


), and (b-2) forming an upper plasma silicon nitride film on the lower plasma silicon nitride film by the plasma chemical vapor deposition, the upper plasma silicon nitride film containing more silicon than the stoichiometric silicon nitride (Si


3


N


4


), wherein the lower plasma silicon nitride film has a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.9 nor more than 2.5 and in the range of not less than 0.9 nor more than 1.7, and the lower plasma silicon nitride film has a thickness in the range of not less than 20 nm nor more than 60 nm, and wherein the upper plasma silicon nitride film has a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.7 nor more than 2.4 and in the range of not less than 0.15 nor more than 0.75, and the upper plasma silicon nitride film has a thickness in the range of not less than 10 nm nor more than 40 nm.




Preferably, according to a sixth aspect, in the semiconductor device manufacturing method, the complex index of refraction of the lower plasma silicon nitride film is (2.1−i×1.2) and the thickness of the lower plasma silicon nitride film is 50 nm, and the refractive index of the upper plasma silicon nitride film is (2.0−i×0.3) and the thickness of the upper plasma silicon nitride film is 30 nm.




Preferably, according to a seventh aspect, in the semiconductor device manufacturing method, the step (b-1) and the step (b-2) are continuously performed in the same chamber.




Preferably, according to an eighth aspect, in the semiconductor device manufacturing method, the step (b) further comprises a step (b-3) of exposing an exposed surface of the upper plasma silicon nitride film to a plasma of a gas which contains oxygen to form a thermal oxide film.




Preferably, according to a ninth aspect, the semiconductor device manufacturing method further comprises the steps of; (d) forming a silicon oxide film covering the upper and lower plasma silicon nitride films, and (e) applying at least one of an etching process and a polishing process to the silicon oxide film.




According to the first aspect, the reflectance of the exposure light can be reduced during the resist exposure so that the resist, and hence the underlying layer, can be accurately patterned. Moreover, the light intensity distribution of the standing wave of the exposure light produced in the resist can be improved to make the patterned resist less apt to break. Furthermore, since the lower and upper plasma silicon nitride films are formed by using plasma chemical vapor deposition, the two plasma silicon nitride films can be continuously formed in the same chamber. That is to say, the formation of the antireflection coating offers good productivity.




According to the second aspect, the reflectance of the exposure light can be reduced during the resist exposure, regardless of the film thickness of the underlying layer.




According to the third aspect, it is possible to prevent the lower and upper plasma silicon nitride films from lessening the effect of acid generated in a chemically amplified resist.




According to the fourth aspect, in the semiconductor device manufacturing method, the silicon oxide film is formed so thin that the silicon oxide film and the upper and lower plasma silicon nitride films can be removed without the need to perform separated process steps (i.e. they can be removed simultaneously).




According to the fifth aspect, the reflectance of the exposure light can be reduced in the step (c), so that the resist, and hence the underlying layer, can be accurately patterned. Also, the light intensity distribution of the standing wave of the exposure light produced in the resist can be improved to prevent the patterned resist from breaking.




According to the sixth aspect, the reflectance of the exposure light can be reduced during the resist exposure, independently of the film thickness of the underlying layer.




According to the seventh aspect, the formation of the antireflection coating offers good productivity.




According to the eighth aspect, the thermal oxide film prevents the lower and upper plasma silicon nitride films from lessening the acid effect of the chemically amplified resist. Also, in the semiconductor device manufacturing process, the thermal oxide film is formed so thin (about 4 nm) in the step (b-3) that the thermal oxide film and the upper and lower plasma silicon nitride films can be removed without the need for separated process steps (i.e. they can be removed simultaneously). Moreover, the step (b-3) uses a plasma of an oxygen-containing gas, so that the lower and upper plasma silicon nitride films and the thermal oxide film can be continuously formed in the chamber used to form the lower and upper plasma silicon nitride films, thus providing good productivity in the antireflection coating formation.




According to the ninth aspect, in the step (e), the upper and lower plasma silicon nitride films can be used as a stopper in the process of etching or/and polishing the silicon oxide film. Accordingly, the stoichiometric silicon nitride (Si


3


N


4


), which has conventionally been used as the stopper, can be formed thinner or removed. In general, the plasma silicon nitride film offers higher deposition rate than the stoichiometric silicon nitride, so that the stopper can be formed with good productivity. Also, in general, the plasma silicon nitride film produces smaller film stress than the stoichiometric silicon nitride, so that the film stress in the semiconductor device can be reduced.




The present invention has been made to solve the problems explained earlier, and a first object of the present invention is to provide an antireflection coating and a semiconductor device manufacturing method which enable the resist to be accurately patterned and also enable the resist to be formed in such shape as will not easily break.




A second object of the present invention is to provide an antireflection coating and a semiconductor device manufacturing method which can achieve the first object regardless of the film thickness of the underlying layer beneath the antireflection coating.




A third object of the present invention is to provide an antireflection coating and a semiconductor device manufacturing method which are suitable for use with a chemically amplified resist.




These and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.











BRIEF DESCRIPTION OF THE DRAWINGS





FIG. 1

is a schematic diagram used to explain the reflection of light.





FIG. 2

is a schematic sectional view used to explain a semiconductor device to which an antireflection coating according to a first preferred embodiment is applied.





FIG. 3

is a diagram showing the correlation between the imaginary part k and the real part n in the complex indices of refraction of the upper and lower p-SiN films of the first preferred embodiment.





FIG. 4

is a schematic sectional view showing a semiconductor device used to explain the setting of the imaginary part k in the complex index of refraction of the lower p-SiN film of the first preferred embodiment.





FIG. 5

is a graph showing the results of calculation about the reflectance at the interface between the resist and the lower p-SiN film of the first preferred embodiment.





FIG. 6

is a graph showing the results of calculation about the reflectance at the interface between the resist and the upper p-SiN film of the first preferred embodiment.





FIG. 7

is a schematic sectional view used to explain a semiconductor device to which the antireflection coating of the first preferred embodiment is applied.





FIG. 8

is a contour diagram used to explain the reflectance in the semiconductor device shown in FIG.


7


.





FIG. 9

is a graph used to explain the reflectance in the semiconductor device shown in FIG.


7


.





FIG. 10

is a schematic sectional view used to explain a semiconductor device to which the antireflection coating of the first preferred embodiment is applied.





FIG. 11

is a graph showing the results of calculation of the patterning dimension of the resist in the semiconductor device shown in FIG.


10


.





FIG. 12

is a schematic sectional view showing a semiconductor device having the resist patterned by using the antireflection coating of the first preferred embodiment.





FIG. 13

is a diagram showing the light intensity distribution in the resist produced during pattern exposure in the presence of the antireflection coating of the first preferred embodiment.





FIG. 14

is a schematic sectional view used to explain a semiconductor device to which the antireflection coating of the first preferred embodiment is applied.





FIG. 15

is a contour diagram used to explain the reflectance in the semiconductor device shown in FIG.


14


.





FIG. 16

is a graph used to explain the reflectance in the semiconductor device shown in FIG.


14


.





FIG. 17

is a schematic sectional view used to explain a semiconductor device to which the antireflection coating of the first preferred embodiment is applied.





FIGS. 18

to


20


are contour diagrams used to explain the reflectance in the semiconductor device shown in FIG.


17


.





FIGS. 21 and 22

are schematic diagrams used to explain the reaction in which a lone-pair electrons in the silicon nitride film attract acid.





FIG. 23

is a schematic sectional view used to explain a semiconductor device to which an antireflection coating according to a second preferred embodiment is applied.





FIGS. 24

to


29


are schematic sectional views used to explain a semiconductor device manufacturing method according to a third preferred embodiment.





FIGS. 30

to


35


are schematic sectional views used to explain a semiconductor device manufacturing method according to a fourth preferred embodiment.





FIGS. 36

to


38


are schematic sectional views used to explain a semiconductor device manufacturing method according to a fifth preferred embodiment.





FIG. 39

is a schematic plan view used to explain the semiconductor device of the fifth preferred embodiment.





FIG. 40

is a schematic sectional view to explain a semiconductor device according to a sixth preferred embodiment.





FIGS. 41

to


44


are schematic sectional views used to explain the semiconductor device manufacturing method according to the sixth preferred embodiment.





FIG. 45

is a schematic sectional view of a semiconductor device used to explain a conventional antireflection coating.





FIG. 46

is a contour diagram showing the reflectance of the exposure light in the semiconductor device shown in FIG.


45


.





FIG. 47

is a schematic sectional view showing another semiconductor device used to explain the conventional antireflection coating.





FIG. 48

is a graph showing the patterning dimension of the resist in the semiconductor device shown in FIG.


47


.





FIG. 49

is a diagram showing the light intensity distribution produced in the resist during pattern exposure in the presence of the conventional antireflection coating.





FIG. 50

is a sectional view used to explain a conventional semiconductor device manufacturing method.





FIG. 51

is a schematic plan used to explain a conventional semiconductor device.





FIG. 52

is a sectional view used to explain a conventional semiconductor device manufacturing method.











DESCRIPTION OF THE PREFERRED EMBODIMENTS




<First Preferred Embodiment>




First,

FIG. 1

is a schematic diagram used to explain the reflection of light. As shown in

FIG. 1

, the first and second layers


201


and


202


are in contact with each other, where the first layer


201


has a complex index of refraction (n1−i×k1) and the second layer


202


has a complex index of refraction (n2−i×k2). The characters n1 and n2 are the real parts of the complex indices of refraction, which determine the light velocity in the substance. The characters k1 and k2 are the imaginary parts of the complex indices of refraction, which correspond to the absorption coefficient of the light. A larger value of the imaginary part indicates a higher light blocking property and a smaller value of it indicates a higher transmittance. Hereinafter the complex index of refraction may simply be called an index of refraction.




Light traveling in the second layer


202


to the first layer


201


is reflected at the interface between the layers


201


and


202


, where the reflectance R is given as:






R
=




(


n





1

-

n





2


)

2

+


(


k





1

-

k





2


)

2





(


n





1

+

n





2


)

2

+


(


k





1

+

k





2


)

2













The reflectance R becomes smaller as the two real parts n1 and n2 become closer to each other and also as the two imaginary parts k1 and k2 become closer to each other.





FIG. 2

is a schematic sectional view used to explain a semiconductor device (or a semiconductor substrate)


101


using an antireflection coating


10


according to a first preferred embodiment. In the semiconductor device


101


, the antireflection coating


10


is composed of an inorganic material and disposed between, and in contact with, a resist


1


and a silicon oxide film (which may be also called an oxide film hereinafter)


3


formed as an underlying layer (which is optically transparent to the exposure light wavelength)


14


. More specifically, the antireflection coating


10


has a double-layer structure including silicon nitride films


11


and


12


formed by plasma chemical vapor deposition (CVD: which may be also called p-SiN films hereinafter). Hereinafter, the p-SiN film


11


in contact with the underlying layer


14


may be referred to also as a lower p-SiN film


11


and the p-SiN film


12


disposed on the lower p-SiN film


11


, on the side of the resist


1


, may be referred to also as an upper p-SiN film


12


.




The p-SiN films


11


and


12


formed by plasma CVD have a chemical structure expressed as Si


x


N


y


(x/y≧3/4), where Si excessively exists over the number of chemical covalent bonds of the stoichiometric silicon nitride (Si


3


N


4


). For the formation of the p-SiN films


11


and


12


, an excess of Si can be included by setting the ratio of flow rate of the SiH


4


gas, with respect to that of the N


2


O gas, in the CVD material gas, to be higher than the ratio for the stoichiometric silicon nitride. The p-SiN films


11


and


12


have etching property intermediate between the stoichiometric silicon nitride (Si


3


N


4


) and the silicon oxide film. For example, while it is very difficult to etch the stoichiometric silicon nitride (Si


3


N


4


) under silicon-oxide etching conditions, the p-SiN films


11


and


12


can be etched (though at slower rates than the silicon oxide film) by using the silicon-oxide etching conditions.




In particular, for the lower p-SiN film


11


, the real part n in the complex index of refraction is set in the range of not less than 1.9 nor more than 2.5 (preferably around 2.1), the imaginary part k is set in the range of not less than 0.9 nor more than 1.7 (preferably around 1.2), and the film thickness is set in the range of not less than 20 nm nor more than 100 nm (preferably around 50 nm). The lower p-SiN film


11


thus blocks the exposure light traveling toward the transparent underlying layer


14


to prevent the reflectance R from being affected by the film thickness variation of the underlying layer


14


.




For example, the lower p-SiN film


11


having the complex index of refraction (n−i×k) shown above can be formed by CVD by setting the film formation conditions as follows: flow rates of SiH


4


gas, N


2


gas and N


2


O gas respectively at 220 sccm (=220×10


−3


liters/min), 5000 sccm, 150 sccm; RF power at 130 W; formation pressure at 4.25 Torr (1 Torr=133.322 Pa); formation temperature at 400° C.




For the upper p-SiN film


12


, the real part n in the complex index of refraction is set in the range of not less than 1.7 nor more than 2.4 (preferably around 2.0), the imaginary part k is set in the range of not less than 0.15 nor more than 0.75 (preferably around 0.3), and the film thickness is set in the range of not less than 10 nm nor more than 60 nm (preferably around 30 nm). The upper p-SiN film


12


thus reduces the reflectance R through the interference effect.




For example, the upper p-SiN film


12


having the complex index of refraction (n−i×k) shown above can be formed by CVD by setting the film formation conditions as follows: flow rates of SiH


4


gas, N


2


gas and N


2


O gas respectively at 95 sccm, 5000 sccm, 150 sccm; RF power at 160 W; formation pressure at 5 Torr; formation temperature at 400° C.




In the CVD, the film quality uniformity and the deposition rate can be controlled by adjusting the pressure and the RF power.




For example, the semiconductor device


101


of

FIG. 2

can be obtained by: forming the oxide film


3


by CVD on the tungsten silicide film


4


formed by sputtering, sequentially forming the lower and upper p-SiN films


11


and


12


under the conditions shown above, and applying the resist


1


.





FIG. 3

is a diagram showing the correlation (experimental results) between the imaginary part k and the real part n of the p-SiN films


11


and


12


. In

FIG. 3

, the horizontal axis shows the imaginary part k and the vertical axis shows the real part n. The white squares in

FIG. 3

(□) show experimental values. As shown by the white squares and the characteristic curve C


0


connecting them, the real part n increases as the imaginary part k increases in the p-SiN films


11


and


12


(i.e. they have a positive correlation). Thus the correlation shown in

FIG. 3

is obtained first and then the complex indices of refraction of the lower and upper p-SiN films


11


and


12


are set on the basis of their correlation.




Specifically, for example, considering the variations in characteristics of the p-SiN films which are caused depending on the film forming apparatus, the characteristic curve C


0


is shifted by ±0.2 in the vertical axis direction and the region AR


3


between the shifted characteristic curves C


1


and C


2


is defined as the correlation region of the real part n and the imaginary part k. Then the regions AR


1


and AR


2


are selected for the complex indices of refraction desired for the p-SiN films


11


and


12


. The real part n and the imaginary part k of the complex index of refraction of the lower p-SiN film


11


are defined in the region where the two regions AR


1


and AR


3


overlap. Similarly, the real part n and the imaginary part k of the complex index of refraction of the upper p-SiN film


12


are defined in the region where the two regions AR


2


and AR


3


overlap. When the correlation between the real part n and the imaginary part k in the complex indices of refraction of the p-SiN films


11


and


12


has thus been obtained in advance, it is not necessary to repeatedly perform the experiments, and the complex indices of refraction (n−i×k) can be realistically and practically determined by taking the correlation into account.




The settings of the complex indices of refraction and the film thicknesses of the lower and upper p-SiN films


11


and


12


are now described in greater detail. The calculations (simulations) about the exposure light reflection etc. take into account the multiple reflection of the light in the antireflection coating


10


etc., so that the calculated results and experiments almost perfectly agree. While the calculations were carried out at an exposure light wavelength of 248 nm, similar results are obtained at other wavelengths.




First, the real parts n of the complex indices of refraction of the p-SiN films


11


and


12


are described. In general, the real part n of the complex index of refraction of the resist


1


is about 1.8 to 1.9 and the real parts n of the p-SiN films


11


and


12


are about 2.0 to 2.5; the resist


1


and the p-SiN films


11


and


12


thus have close real parts n. Values around 2.2 are adopted as the real parts n of the p-SiN films


11


and


12


considering that a currently common p-SiN film has a real part n of 2.23 and that the value around 2.2 is nearly intermediate between the real part n of the resist


1


and those of the p-SiN films


11


and


12


.




Next, the imaginary parts k of the complex indices of refraction (n−i×k) of the p-SiN films


11


and


12


are described. To allow the p-SiN films


11


and


12


to function as the antireflection coating, it will be necessary that the values be set as 0.2≦k≦1.5, as in the conventional antireflection coating


10


P (see FIG.


45


). However, since the imaginary part k of the resist


1


, about 0.01 to 0.02, considerably differs from the imaginary parts k of the p-SiN films


11


and


12


, the imaginary parts k of the p-SiN films


11


and


12


are defined as shown below.




The imaginary part k of the lower p-SiN film


11


is described first.

FIG. 4

is a schematic sectional view showing a semiconductor device (or a semiconductor substrate)


101


B used to explain the setting of the imaginary part k of the lower p-SiN film


11


. The semiconductor device


101


B is used to examine the effect of the thickness of the oxide film


3


on the reflectance R; the semiconductor device


101


B has a stacked-layer structure of the tungsten silicide film


4


oxide film


3


lower p-SiN film


11


resist


1


.





FIG. 5

is a graph showing the results of calculation about the reflectance R at the interface between the lower p-SiN film


11


and the resist


1


in the semiconductor device


101


B. In this graph, the horizontal axis shows the thickness of the lower p-SiN film


11


and the vertical axis shows the variation width (maximum-minimum) of the reflectance R with the thickness of the oxide film


3


varying in the range from 100 to 300 nm. The graph shows the results calculated using the imaginary part k of the lower p-SiN film


11


as a parameter, where k=0.7, 0.86, 1, 1.15, 1.3. It is assumed here that the real part n of the complex index of refraction of the lower p-SiN film


11


has the value 2.23 shown above.




It can be seen from

FIG. 5

that the variation width of the reflectance R becomes smaller as the p-SiN film


11


becomes thicker and as the imaginary part k of the lower p-SiN film


11


becomes larger. For example, when the lower p-SiN film


11


is too thick, the aspect ratio in etching increases and then the productivity is lowered. Considering this, the thickness of the lower p-SiN film


11


will preferably be around 50 nm where the change (or the changing rate) of the variation width of the reflectance R is relatively small, and the thickness will practically be in the range of not less than 20 nm nor more than 60 nm.




As can be seen from the correlation shown in

FIG. 3

described earlier, setting a large imaginary part k in the complex index of refraction of the lower p-SiN film


11


causes its real part n to vary away from the real part n of the resist


1


(=about 1.8 to 1.9). Therefore it will be preferred that the lower p-SiN film


11


have an imaginary part k around 1.15. Thus, a more realistic or practical example of the complex index of refraction of the lower p-SiN film


11


is (2.1−i×1.2).




Next, the imaginary part k of the upper p-SiN film


12


is described.

FIG. 6

is a graph showing the results of calculation about the reflectance R and the thickness of the upper p-SiN film


12


in the semiconductor device


101


(see FIG.


2


). In this graph, the horizontal axis shows the thickness of the upper p-SiN film


12


and the vertical axis shows the reflectance R. This graph shows the results calculated by using the imaginary part k of the upper p-SiN film


12


as a parameter, where k=0.1, 0.2, 0.3, 0.5, 0.7, 0.86 and 1.0. It is assumed here that the real part n of the upper p-SiN film


12


has the value 2.23 shown above and the complex index of refraction and the thickness of the lower p-SiN film


11


are respectively (2.23−i×1.15) and 50 nm as shown above. The thickness of the oxide film


3


is 150 nm.




It can be seen from

FIG. 6

that when the imaginary part k of the upper p-SiN film


12


is small, the reflectance R takes very small values (minimum values) around the film thickness of 20 nm (it is almost zero). For example, when the imaginary part k is in the range from about 0.20 to 0.70 and the film thickness is in the range from about 10 nm to 40 nm, the reflectance R can be a practically usable value of 2 to 3%, or lower. Also, for example, when the imaginary part k is in the range from about 0.20 to 0.40 and the film thickness is in the range from about 15 nm to 30 nm, the reflectance R can be 1% or less.




As shown in the correlation diagram of

FIG. 3

, the real part n rapidly varies when the imaginary part k is too small. Accordingly, considering stable reproducibility of the film characteristics (particularly the complex index of refraction in this case), a more realistic or practical example of the complex index of refraction of the upper p-SiN film


12


is (2.0−i×0.3). The deposition rate of the p-SiN film is generally so high (about 5 to 10 nm/sec) that a film about 10 nm thick is formed in several seconds. Since the plasma is unstable in such a short film formation time, it is difficult to obtain stable film thickness and stable film quality. Therefore, values around 30 nm, e.g. in the range of not less than 20 nm nor more than 35 nm, are more realistic or practical film thicknesses for the upper p-SiN film


12


.




Next, the results of calculation of the reflectance etc. carried out using a semiconductor device (or a semiconductor substrate) having the antireflection coating


10


is explained, where a semiconductor device


101


C having the structure shown in

FIG. 7

is used. The semiconductor device


101


C has a stacked-layer structure of the silicon substrate


15


/oxide film


16


/polycrystalline silicon film


5


/oxide film


3


as transparent underlying layer


14


/lower p-SiN film


11


/upper p-SiN film


12


/resist


1


. The antireflection coating


10


of the semiconductor device


101


C is formed on the transparent underlying layer


14


; in regard to the reflection of the exposure light, the structure of the semiconductor device


101


C is the same as that of the semiconductor device


101


shown in FIG.


2


.




In the calculation of the reflectance in the semiconductor device


101


C, the indices of refraction of the silicon substrate


15


, the oxide film


16


and the polycrystalline silicon film


5


were assumed to be (2.42−i×3.18), 1.51, and (1.63−i×3.58), respectively. The indices of refraction of the oxide film


3


forming the underlying layer


14


, the lower p-SiN film


11


, the upper p-SiN film


12


, and the resist


1


were assumed to be 1.51, (2.2−i×1.17), (2.0−i×0.32), and (1.8−i×0.01), respectively. The thickness of the oxide film


16


and that of the polycrystalline silicon film


5


were assumed to be 17 nm and 20 nm, respectively. In the calculation of the reflectance R, the film thickness of the polycrystalline silicon film


5


is equivalent to infinity, since it serves as a perfectly light blocking film at thicknesses of 20 nm or larger.





FIG. 8

is a contour diagram showing the results of calculation about the relation between the thicknesses of the lower and upper p-SiN films


11


and


12


and the reflectance R in the semiconductor device


101


C. The calculation shown in

FIG. 8

assumes the thickness of the oxide film


3


to be 90±10 nm.

FIG. 8

indicates that the reflectance R can be 2 to 3%, or lower, in the range where the thickness of the lower p-SiN film


11


is about 7 nm or larger and that of the upper p-SiN film


12


is from about 17 to 50 nm, for example. Also, for example, in the range in which the film thickness of the lower p-SiN film


11


is 20±3 nm and that of the upper p-SiN film


12


is 30±3 nm (see the rectangular area surrounded by the broken line in FIG.


8


), the reflectance R can be 1% or lower.





FIG. 9

is a graph showing the results of calculation about the relation between the thickness of the oxide film


3


as the underlying layer


14


and the reflectance R in the semiconductor device


101


C. It is assumed in the calculation shown in

FIG. 9

that the thicknesses of the lower and upper p-SiN films


11


and


12


are respectively 20±3 nm and 30±3 nm (see the rectangular area surrounded by the broken line in FIG.


8


).

FIG. 9

indicates that the reflectance can be 3% or lower in the range where the thickness of the oxide film


3


is from about 70 to 120 nm, for example, and the reflectance R can be 1% or lower in the range of 90 ±10 nm shown above.




The antireflection coating


10


and a semiconductor device manufacturing method using the same enables accurate formation of the resist


1


. This is now described referring to

FIGS. 10 and 11

.

FIG. 10

is a schematic sectional view used to explain a semiconductor device


101


D to which the antireflection coating


10


is applied, where the semiconductor device


101


D has a stacked-layer structure of the polycrystalline silicon


5


/oxide film


3


as the underlying layer


14


/lower p-SiN film


11


/upper p-SiN film


12


/resist


1


/top antireflection coating (so-called TARC)


6


.





FIG. 11

is a graph showing the results of calculation about the patterning dimension of the resist


1


in the semiconductor device


101


D.

FIG. 11

corresponds to

FIG. 48

explained earlier, where the horizontal and vertical axes show the film thickness of the oxide film


3


in the semiconductor device


101


D and the patterning dimension (line width) of the resist


1


;

FIG. 11

shows results calculated at varying exposure light energies. The exposure conditions used for the resist


1


are: NA (Numerical Aperture)=0.65, annular ⅔, and exposure mask L/S (Line and Space) pattern 200 nm. According to

FIG. 11

, the variation of the line width of the resist


1


can be suppressed to 6 nm with the film thickness variation of the oxide film


3


. In the case of the conventional single-layer antireflection coating


10


P, the line width variation of the resist


1


P was 17 nm (see

FIG. 48

)




As compared with the conventional single-layer antireflection coating


10


P, the antireflection coating


10


and a semiconductor device manufacturing method using the same satisfactorily suppress breakage or collapse of the patterned resist


1


. This effect is now explained referring to

FIGS. 12

,


13


and


49


.

FIG. 12

is a schematic sectional view showing a semiconductor device


101


E having the resist


1


patterned by using the antireflection coating


10


.

FIG. 13

is a diagram showing the light intensity distribution (calculated results) in the resist


1


produced during the pattern exposure in the presence of the antireflection coating


10


(i.e. in the semiconductor device


101


. For comparison,

FIG. 49

shows the light intensity distribution (calculated results) in the resist


1


P in the presence of the conventional antireflection coating


10


P instead of the antireflection coating


10


. In

FIGS. 13 and 49

, the horizontal and vertical axes respectively show the width and thickness directions of the resists


1


and


1


P, and the origin (the zero point) on the horizontal axis corresponds to the center in the width direction of the patterned resists


1


and


1


P.




As shown in

FIG. 12

, the semiconductor device


101


E has a stacked-layer structure of the polycrystalline silicon


5


/oxide film


3


/lower p-SiN film


11


/upper p-SiN film


12


/resist


1


. It is assumed in the calculation of

FIG. 13

that the complex indices of refraction of the lower and upper p-SiN films


11


and


12


are respectively (2.1−i×1.2) and (2.0−i×0.35), their film thicknesses are respectively 50 nm and 30 nm, and the oxide film


3


is 100 nm thick. The calculation shown in

FIG. 49

assumed the conventional antireflection coating


10


P to be a p-SiN film having a complex index of refraction of (2.1−i×0.8) and a film thickness of 48 nm.




It is also assumed that the resist


1


is a positive resist and is pattern-exposed by using a mask pattern for isolated line of 120 nm in width. The exposure conditions used for the resist


1


are: NA=0.65, annular ⅔, and exposure mask L/S pattern 200 nm.




During the exposure of the resist


1


, the exposure light is reflected at the interface between the resist


1


and the antireflection coating


10


and at the interface between the resist


1


and the air (i.e. at the exposed surface of the resist


1


) and standing wave is produced in the resist


1


. Due to the standing wave, the light intensity, in other words the exposure intensity of the resist


1


, is distributed as shown in

FIGS. 13 and 49

in the resist


1


. Since the exposed areas of the positive resist are dissolved when developed, the resist


1


then exhibits a sectional shape corresponding to the distribution shown in

FIG. 13

or


49


.




That is to say, in

FIG. 49

, the conventional antireflection coating


10


P allows the intensive exposure light to penetrate into the areas of the resist


1


that are to remain (i.e. these areas are intensively exposed) and the resultant section of the resist


1


in the thickness direction shows periodic undulations with large amplitude (in other words, it has considerably narrowed parts). Therefore the resist is susceptible to breakage at the narrowed parts.




In contrast, as shown in

FIG. 13

, the antireflection coating


10


can reduce the amplitude of the undulations explained above, so that the narrowed parts of the resist


1


can be wider than those obtained with the conventional antireflection coating


10


P. The antireflection coating


10


thus suppresses the breakage or collapse of the patterned resist


1


.




Also, since the lower and upper plasma silicon nitride films


11


and


12


are formed by plasma CVD, the two p-SiN films


11


and


12


can be continuously formed in the same chamber. That is to say, according to the antireflection coating


10


and a semiconductor device manufacturing method using the same, the antireflection coating offers superior productivity in formation.




The antireflection coating


10


can be applied also to an underlying layer


14


other than the oxide film


3


to provide the same effects as those obtained when the underlying layer


14


is the oxide film


3


.





FIG. 14

is a schematic sectional view showing a semiconductor device


101


F to which the antireflection coating


10


is applied. The semiconductor device


101


F has a stacked-layer structure in which the oxide film


3


in the semiconductor device


101


C of

FIG. 7

has been replaced by a silicon nitride film (hereinafter also called a nitride film)


2


, where the nitride film


2


corresponds to the underlying layer


14


. The nitride film


2


is made of stoichiometric silicon nitride (Si


3


N


4


) having a refractive index of 2.27. The nitride film


2


is formed by, for example, thermal CVD, prior to the formation of the antireflection coating


10


.





FIG. 15

is a contour diagram showing the reflectance R which, like that shown in

FIG. 8

, was obtained by calculation. The calculation of

FIG. 15

assumed the thickness of the nitride film


2


to be 115±10 nm.

FIG. 15

indicates that the reflectance R can be 2 to 3%, or lower, in the range in which the thickness of the lower p-SiN film


11


is about 5 nm or larger and the thickness of the upper p-SiN film


12


is from about 20 nm to 50 nm, for example. Also for example, in the range in which the thickness of the lower p-SiN film


11


is 20±3 nm and that of the upper p-SiN film


12


is 30±3 nm (the rectangular range surrounded by the broken line in FIG.


15


), the reflectance R can be 1% or lower.





FIG. 16

is a graph showing the relation between the thickness of the nitride film


2


and the reflectance R, which was calculated in the same way as that shown in FIG.


9


. The calculation of

FIG. 16

assumed that the film thicknesses of the lower and upper p-SiN films


11


and


12


were respectively 20±3 nm and 30±3 nm (see the rectangular region surrounded by the broken line in FIG.


15


). According to

FIG. 16

, the reflectance R can be 3% or lower in the range in which the thickness of the nitride film


2


is from about 95 to 130 nm, and the reflectance R can be 1% or lower in the range of 115±10 nm shown above.





FIG. 17

is a schematic sectional view showing a semiconductor device


101


G to which the antireflection coating


10


is applied. The semiconductor device


101


G has a stacked-layer structure of the tungsten silicide film


4


/nitride film


2


/oxide film


3


/lower p-SiN film


11


/upper p-SiN film


12


/resist


1


, where the oxide film


3


and the nitride film


2


correspond to the transparent underlying layer


14


. The semiconductor device


101


G has a structure in which the conventional single-layer antireflection coating


10


P in the semiconductor device


101


P shown in

FIG. 45

has been replaced by the antireflection coating


10


.





FIG. 18

is a contour diagram showing the reflectance R obtained by calculation like that of FIG.


8


. The calculation of

FIG. 18

assumed that the thickness of the oxide film


3


and that of the nitride film


2


were respectively 80±4 nm and 70±3.5 nm (which correspond to the rectangular region surrounded by the broken line in

FIG. 19

described later), and that the index of refraction of the tungsten silicide film


4


was (2.42−i×3.18).

FIG. 18

shows that the reflectance R can be 2 to 3%, or lower, in the range in which the thickness of the lower p-SiN film


11


is about 4 nm or larger and that of the upper p-SiN film


12


is from about 20 nm to 55 nm, for example. Also, for example, in the range in which the thickness of the lower p-SiN film


11


is 20±3 nm and that of the upper p-SiN film


12


is 30±3 nm (the rectangular range surrounded by the broken line in FIG.


18


), the reflectance R can be 1% or lower.





FIG. 19

is a contour diagram showing the results of calculation about the relation between the reflectance R and the thicknesses of the oxide film


3


and the nitride film


2


as the underlying layer


14


. The calculation of

FIG. 19

assumed that the film thicknesses of the lower and upper p-SiN films


11


and


12


were respectively 20±3 nm and 30±3 mm (see the rectangular region surrounded by the broken line in FIG.


18


).




As can be seen by comparing FIG.


19


and

FIG. 46

about the conventional semiconductor device


101


P, the antireflection coating


10


enables lower reflectance than the conventional single-layer antireflection coating


10


P. Furthermore, the reflectance can be 2 to 3% or lower, or even 1% or lower, at appropriate, various combinations of the thickness of the oxide film


3


and that of the nitride film


2


.




Also,

FIG. 20

is a contour diagram like FIG.


19


. The calculation of

FIG. 20

assumed that the complex indices of refraction of the lower and upper p-SiN films


11


and


12


were respectively (2.15−i×1.2) and (2.00−i×0.3), and the film thicknesses were 50 nm and 30 nm.

FIG. 20

shows that the reflectance R can be 1% or lower (even as very low as 0.05 to 0.15%), regardless of the film thicknesses of the oxide film


3


and the nitride film


2


. Further, in

FIG. 20

, the difference between the maximum and minimum values of the reflectance R is much smaller than that with the conventional single-layer antireflection coating


10


P, which shows that the film thickness variations of the oxide film


3


and the nitride film


2


exert less influence upon the variation of the reflectance R.




As can be seen from

FIGS. 8

,


9


,


15


,


16


,


18


and


19


, by controlling the film thickness of the underlying layer


14


, the lower p-SiN film


11


can be formed thinner (e.g. 20 nm) while suppressing the reflectance R. In this case the antireflection coating


10


can be removed easily.




<Second Preferred Embodiment>




Now, in fine patterning processes with line widths of 200 nm or smaller, KrF (248 nm), ArF (193 nm), etc. are used as the exposure light; chemically amplified resists are used with such exposure lights.




The chemically amplified resists generate acid (H


+


) on exposure, which alters the solubility in developing. By the way, the chemical covalent bond of the nitride film contains lone-pair electrons (see FIG.


21


). Accordingly, when the chemically amplified resist exists on the nitride film, the lone-pair electrons of the nitride film attract the acid (H


+


) generated during the resist exposure (see FIG.


22


). That is to say, the nitride film is basic and therefore lessens the acid, which may hinder accurate resist patterning. Accordingly, a second preferred embodiment explains an antireflection coating suitable for use with the chemically amplified resists.





FIG. 23

is a schematic sectional view used to explain a semiconductor device


102


to which an antireflection coating


10


B of the second preferred embodiment is applied. The semiconductor device


102


has a stacked-layer structure of tungsten silicide film


4


/oxide film


3


/nitride film


2


/lower p-SiN film


11


/upper p-SiN film


12


/surface oxide film


13


/resist


1


. In this structure, the three layers, the lower p-SiN film


11


/upper p-SiN film


12


/surface oxide film


13


, correspond to the antireflection coating


10


B, and the oxide film


3


and the nitride film


2


correspond to the transparent underlying layer


14


. The underlying layer


14


may be formed only with the oxide film


3


or the nitride film


2


. It is assumed that the resist


1


is chemically amplified.




The surface oxide film


13


is formed as shown below. For example, after the formation of the upper p-SiN film


12


, a gas which contains oxygen, e.g. N


2


O gas, is introduced into the chamber and plasma is generated, and the exposed surface of the upper p-SiN film


12


is thermally oxidized with oxygen produced by the plasma. More specifically, the surface oxide film


13


is formed under the following conditions: N


2


O gas flow rate 1000 sccm; RF power 160 W; pressure 4.25 Torr; temperature 400° C. In this process, the surface oxide film


13


is formed to a thickness of about 10 nm or smaller (preferably about 4 to 5 nm, or smaller) so that it will not prevent the antireflection function of the lower and upper p-SiN films


11


and


12


.




With the antireflection coating


10


B, the surface oxide film


13


interposed between the chemically amplified resist


1


and the upper p-SiN film


12


prevents the acid (H


+


) in the resist


1


generated during the exposure from being attracted to the lone-pair electrons in the upper p-SiN film


12


. The acid function in the resist


1


is thus not hindered and the resist


1


can be precisely patterned into desired shape.




Furthermore, since the thickness of the surface oxide film


13


is 10 nm or smaller, the surface oxide film


13


and the p-SiN films


12


and


11


can be etched continuously (simultaneously) in the process for etching the antireflection coating


10


B. In other words, when the surface oxide film


13


is thicker than 10 nm, the surface oxide film


13


and the p-SiN films


12


and


11


must be separately etched in two steps, which would complicate the etching process.




The surface oxide film


13


can be formed by, instead of the thermal oxidation, depositing an oxide film itself by CVD. However, the thermal oxidation can form the thin surface oxide film


13


, 10 nm or thinner, more accurately than CVD. This is because, when thermal oxidation is applied to the surface, the oxidation reaction stops at a depth of about 4 to 5 nm, since the nitride film is resistant to oxidation enough to be used as the antioxidation film. On the other hand, it is technically difficult to stably obtain a film of 10 nm or thinner by the CVD having poor controllability.




In this respect, the thermal oxidation method, capable of forming the surface oxide film


14


as thin as 4 to 5 nm thick, facilitates the continuous etching to the surface oxide film


13


and the p-SiN films


12


and


11


. Also, the upper p-SiN film


12


and the surface oxide film


13


(and even the lower p-SiN film


11


) can be continuously formed in the same chamber by generating oxygen plasma in the chamber used to form the p-SiN film


12


. In this way, forming the surface oxide film


13


by thermal oxidation enhances the productivity in the formation of the antireflection coating


10


B.




The surface oxide film


13


, having a thickness of about 4 to 5 nm, can satisfactorily prevent the acid in the resist


1


from diminishing. In other words, the surface oxide film


13


can be still thinner as long as it can prevent the acid from being lessened. Other films which are not basic may be used in place of the surface oxide film


13


. The antireflection coating


10


can be used when the resist


1


is not chemically amplified.




<Third Preferred Embodiment>




In recent devices, the gate oxide film in MOSFETs is as thin as about 2 nm, so that the gate oxide film may be etched during the process of pattern-etching the gate electrode. It is thus difficult to leave the gate oxide film on the active region to protect Si in the active region. For this reason, as shown in the sectional view of

FIG. 50

, a silicon oxide film


3


P is formed on the silicide film


4


P for the gate electrode. The oxide film


3


P is pattern-etched by using the resist


1


P and then the resist


1


P is removed. Next the silicide film


4


P is etched by using the patterned oxide film


3


P as a mask. In this process, the etching conditions are selected so that the oxide film


3


P used as a mask will not be etched away, which suppresses etching damage to the gate oxide film


16


P.




A third preferred embodiment shows a method for patterning the gate electrode (or a semiconductor device manufacturing method) by using the above-described antireflection coatings


10


and


10


B referring to the sectional views of

FIGS. 24

to


29


.




First, a gate oxide film


16


; a film forming the gate electrode later, e.g. the tungsten silicide film


4


; the oxide film


3


corresponding to the underlying layer


14


; the lower and upper p-SiN films


11


and


12


forming the antireflection coating


10


; and the resist


1


are sequentially formed on the silicon substrate


15


. The resist


1


is then patterned through exposure and developing (see FIG.


24


). Next, by using the patterned resist


1


as a mask, the antireflection coating


10


and the oxide film


3


are pattern-etched (see FIG.


25


). Unlike the stoichiometric nitride film (Si


3


N


4


), the p-SiN films


11


and


12


formed by plasma CVD can be etched under etching conditions selected for the oxide film


3


(though the rate is slower), so that the upper and lower p-SiN films


12


and


11


and the oxide film


3


can be continuously (or simultaneously) etched. The rate of continuous etching can be controlled by adjusting the etching conditions on the basis of the etching conditions for the oxide film


3


.




Next, the resist


1


is removed (see

FIG. 26

) and the silicide film


4


is etched by using the patterned antireflection coating


10


and oxide film


3


as a mask (see FIG.


27


). The residue of the antireflection coating


10


is removed by thermal phosphoric acid (see FIG.


28


). The thermal phosphoric acid treatment may be unwanted when the antireflection coating


10


has been etched away during etching to the silicide film


4


. Then common process is performed to form side walls (see

FIG. 29

) and source/drain, etc. (not shown) to complete the MOSFET.




The antireflection coating


10


enables the resist


1


, and hence the gate electrode, to be precisely patterned. The antireflection coating


10


B provides the same effects (this applies also to the fourth to sixth preferred embodiments described later).




<Fourth Preferred Embodiment>




In recent devices, gate electrodes are disposed very close to each other. Therefore a (contact) hole may be formed between, and partially over, adjacent gate electrodes. Such structure is called SAC (Self Align Contact).




In the SAC structure, the hole is formed by etching an interlayer film made of SiO


2


, so that a nitride film, having large selectivity to SiO


2,


is formed on the gate electrodes and used as the etching stopper to protect the gate electrodes from etching. Recently, an oxide film may be formed between the nitride film and the gate electrode to alleviate stress caused because the nitride film and the gate electrode are made of different materials.




As stated above, the recent SAC structure has a transparent oxide film and nitride film formed on the gate electrodes. A fourth preferred embodiment shows a method for forming the SAC structure (or a semiconductor device manufacturing method) by using the above-described antireflection coatings


10


and


10


B referring to the sectional views of

FIGS. 30

to


35


.




First, a semiconductor device having the SAC structure is described referring to FIG.


30


. In the semiconductor device, a gate oxide film


16


is formed on the silicon substrate


15


, and gate electrodes


4


composed of, for example, tungsten silicide, are formed on the gate oxide film


16


. The oxide film


3


, nitride film


2


and p-SiN film


11


are formed in this order on each gate electrode


4


. In this structure, the oxide film


3


and the nitride film


2


correspond to the underlying layer. The gate electrode


4


, the oxide film


3


, the nitride film


2


and the p-SiN film


11


are covered by a nitride film


17


, and the nitride film


17


is covered by an interlayer film


18


made of SiO


2


. The interlayer film


18


has a hole


20


; the hole


20


is formed to expose the nitride film


17


in the space between the adjacent gate electrodes


4


and the oxide film


3


formed on the gate electrodes


4


etc. The hole


20


extends in this space and also partially above the gate electrodes


4


(more specifically above the shoulders of the p-SiN film


11


). The nitride film


2


serves as an etching stopper during formation of the hole


20


.




Next, a method for manufacturing this semiconductor device is described referring to

FIGS. 31

to


35


. First, the gate oxide film


16


; the tungsten silicide film as a film forming the gate electrode


4


later; the oxide film


3


; the nitride film


2


; the lower and upper p-SiN films


11


and


12


forming the antireflection coating


10


; and the resist


1


are sequentially formed on the silicon substrate


15


. The resist


1


is then patterned through exposure and developing (see FIG.


31


). Subsequently, by using the patterned resist


1


as a mask, the antireflection coating


10


and the underlying layer


14


are pattern-etched (see FIG.


32


).




Then the resist


1


is removed (see

FIG. 33

) and the silicide film


4


is etched by using the patterned antireflection coating


10


and underlying layer


14


as a mask (see FIG.


34


). In this process, part of the antireflection coating


10


, e.g. part of the lower p-SiN film


11


, may remain. Then the nitride film


17


and the interlayer film


18


are sequentially formed to cover the exposed surface. Resist


19


is then formed on the interlayer film


18


and an opening is formed in the resist


19


; in the plan view, the opening is formed in an area including parts (shoulders) of the adjacent gate electrodes


4


(i.e. in the area where the contact


20


is formed: see FIG.


35


). The interlayer film


18


is etched by using the opened resist


19


as a mask to form the hole


20


(see FIG.


30


).




The antireflection coating


10


enables precise patterning of the resist


1


and hence of the gate electrode.




In this process, the use of the antireflection coating


10


, or the p-SiN films


11


and


12


, reduces the film stress in the semiconductor device. This is because of the following reason. First, since the upper and lower p-SiN films


11


and


12


can be used also as the etching stopper, the nitride film


2


serving as the etching stopper can be formed thinner than in the absence of the antireflection coating


10


; or the nitride film


2


may be omitted. This reduces the film stress of the nitride film


2


. Now, the p-SiN film formed by plasma CVD generally produces smaller film stress than the stoichiometric Si


3


N


4


film, so that the stress reducing effect by thinning the nitride film


2


can be obtained even when the p-SiN films


11


and


12


are used. Also, the p-SiN films


11


and


12


generally offer higher deposition rate than the nitride film


2


of stoichiometric silicon nitride (Si


3


N


4


), leading to superior productivity in the stopper formation.




<Fifth Preferred Embodiment>




A fifth preferred embodiment explains a dual damascene process (or a semiconductor device manufacturing method) using the antireflection coatings


10


and


10


B referring to the sectional views of

FIGS. 36

to


38


. The description shows a dual damascene process in which a trench is formed after formation of a hole (called “via first”).




First, the nitride film


2


, the oxide film


3


as an interlayer film, the lower p-SiN film


11


and the upper p-SiN film


12


are sequentially formed to cover the interconnection


22


on the substrate. The nitride film


2


and the oxide film


3


correspond to the underlying layer


14


. Then the antireflection coating


10


and the oxide film


3


are patterned by using resist to form a hole


23


above the interconnection


22


(see FIG.


36


). Then the nitride film


2


is pattern-etched to extend the hole


23


so that it reaches the interconnection


22


. Resist (not shown) is then applied to cover the antireflection coating


10


and the entire surface is etched back. The resist (also called a resist plug)


24


is thus buried in the hole


23


(see FIG.


37


). It is buried to prevent the bottom of the hole


23


from being exposed for a long time to the etching plasma during the formation of a trench described later.




Next, the resist


1


is formed to cover the antireflection coating


10


and the hole


23


and is patterned through exposure and developing into a shape corresponding to the trench


25


(see

FIG. 38

) connected to the hole


23


. The antireflection coating


10


and the oxide film


3


are etched by using the resist


1


as a mask, so as to form the trench


25


(see FIG.


38


).




While negative resist and positive resist are both applicable as the resist


1


used to form the trench


25


, the negative resist offers more accurate patterning of the resist


1


. This is because of the following reason. That is to say, during formation of the trench, the resist


1


is removed above the hole


23


. Therefore, if positive resist is used (it is removed in the exposed areas), the exposure light must be applied above the hole


23


. However, the exposure light may be scattered by the hole


23


, or may not sufficiently reach into the hole


23


. In contrast, since the negative resist is exposed to the light in the areas to remain, i.e. since the exposure light is not applied onto the hole


23


, the resist


1


can be patterned more accurately. It is therefore preferable to use the negative resist as the resist


1


.





FIG. 39

is a schematic plan view showing a semiconductor device having the hole


23


and the trench


25


formed by using the antireflection coating


10


. For comparison,

FIG. 51

shows a plan view of a similar semiconductor device manufactured by using the conventional single-layer antireflection coating


10


P. Since the lower p-SiN film


11


of the antireflection coating


10


is more light-blocking than the conventional single-layer antireflection coating


10


P (since it has a larger imaginary part k), the exposure light is less likely to be transmitted through the antireflection coating


10


and light scattering is therefore less likely to occur at the interconnection metal. Therefore the antireflection coating


10


allows the resist


1


, and hence the trench


25


, to be accurately patterned. More specifically, this enhances the straightness (linearity) of the trench


25


around the hole


23


. It is supposed that, when the conventional antireflection coating


10


P is used, the trench


25


is widened around the hole


23


as shown in

FIG. 51

because the exposure light is weakened by scattering around the hole


23


.




<Sixth Preferred Embodiment>




Recent devices widely include element isolation called shallow trench isolation (STI). A sixth preferred embodiment shows a method for forming element isolation having the STI structure (or a semiconductor device manufacturing method) by using the antireflection coatings


10


and


10


B.




First, a semiconductor device having an STI element isolation is described referring to the schematic section of FIG.


40


. This semiconductor device has a silicon substrate


15


in which a trench


26


is formed; the trench


26


is filled with an insulating film (e.g. SiO


2


film)


27


forming the element isolation. The insulating film


27


protrudes above the surface of the silicon substrate


15


.




Next, a method for manufacturing this semiconductor device is described referring to the sectional views of

FIGS. 41

to


44


. First, the oxide film


3


, nitride film


2


, lower p-SiN film


11


, upper p-SiN film


12


and resist


1


are formed in this order on the silicon substrate


15


. The oxide film


3


and the nitride film


2


correspond to the underlying layer


14


. The nitride film


2


serves as a stopper during CMP process described later. Next, the resist


1


is patterned to form an opening in the part where the trench


26


is to be formed (see FIG.


41


). For comparison with

FIG. 41

,

FIG. 52

shows a conventional manufacturing method which does not use the antireflection coating


10


.




The nitride film


2


and the oxide film


3


are etched by using the resist


1


as a mask. The silicon substrate


15


is then etched by using the resist


1


, the nitride film


2


and the oxide film


3


as a mask to form the trench


26


(see FIG.


42


). Alternatively, the silicon substrate


15


may be etched after removal of the resist


1


by using the nitride film


2


and the oxide film


3


as a mask. The resist


1


is then removed and the insulating film


27


is formed to fill the trench


26


and to cover the antireflection coating


10


(see FIG.


43


), and then the insulating film


27


is polished by CMP for planarization (see FIG.


44


). Subsequently the residue of the nitride film


2


and the oxide film


3


(and the antireflection coating


10


if any) is removed to complete the semiconductor device shown in FIG.


40


.




The antireflection coating


10


enables the resist


1


, and hence the trench


27


of the STI structure, to be accurately patterned.




Moreover, when polishing the insulating film


27


, the upper and lower p-SiN films


12


and


11


can be used as the stopper, as well as the nitride film


2


. Therefore the nitride film


2


, which is used also in conventional manufacturing methods, can be formed thinner or removed. The p-SiN films


11


and


12


generally offer higher deposition rate than the nitride film


2


made of stoichiometric silicon nitride (Si


3


N


4


), so that the productivity in the stopper formation can be enhanced. Also, thinning the nitride film


2


reduces the film stress. In general, since the p-SiN film produces smaller film stress than the stoichiometric Si


3


N


4


film, the above-described stress reducing effect produced by thinning the nitride film


2


can be obtained even when the p-SiN films


11


and


12


are used.




It is also possible to use the upper and lower plasma silicon nitride films


12


and


11


as a stopper during both of the CMP and etching processes in the series of manufacturing process steps.




While the invention has been described in detail, the foregoing description is in all aspects illustrative and not restrictive. It is understood that numerous other modifications and variations can be devised without departing from the scope of the invention.



Claims
  • 1. An antireflection coating provided between an underlying layer and a resist in a semiconductor device manufacturing process, said antireflection coating comprising:a lower plasma silicon nitride film provided on said underlying layer, said lower plasma silicon nitride film being formed by a plasma chemical vapor deposition and containing more silicon than stoichiometric silicon nitride (Si3N4); and an upper plasma silicon nitride film provided on said lower plasma silicon nitride film, said upper plasma silicon nitride film being formed by said plasma chemical vapor deposition and containing more silicon than said stoichiometric silicon nitride (Si3N4); said lower plasma silicon nitride film having a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.9 nor more than 2.5 and in the range of not less than 0.9 nor more than 1.7, said lower plasma silicon nitride film having a thickness in the range of not less than 20 nm nor more than 60 nm, said upper plasma silicon nitride film having a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.7 nor more than 2.4 and in the range of not less than 0.15 nor more than 0.75, said upper plasma silicon nitride film having a thickness in the range of not less than 10 nm nor more than 40 nm.
  • 2. The antireflection coating according to claim 1,wherein said complex index of refraction of said lower plasma silicon nitride film is (2.1−i×1.2) and the thickness of said lower plasma silicon nitride film is 50 nm, and said complex index of refraction of said upper plasma silicon nitride film is (2.0−i×0.3) and the thickness of said upper plasma silicon nitride film is 30 nm.
  • 3. The antireflection coating according to claim 1, further comprising a silicon oxide film formed on said upper plasma silicon nitride film.
  • 4. The antireflection coating according to claim 3, wherein said silicon oxide film has a thickness of about 4 to 10 nm.
  • 5. A semiconductor device manufacturing method comprising the steps of:(a) forming an underlying layer; (b) forming an antireflection coating on said underlying layer; and (c) forming a resist on said antireflection coating and patterning said resist through exposure and developing; said step (b) comprising the steps of, (b-1) forming a lower plasma silicon nitride film on said underlying layer by a plasma chemical vapor deposition, said lower plasma silicon nitride film containing more silicon than stoichiometric silicon nitride (Si3N4), and (b-2) forming an upper plasma silicon nitride film on said lower plasma silicon nitride film by said plasma chemical vapor deposition, said upper plasma silicon nitride film containing more said silicon than said stoichiometric silicon nitride (Si3N4), said lower plasma silicon nitride film having a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.9 nor more than 2.5 and in the range of not less than 0.9 nor more than 1.7, said lower plasma silicon nitride film having a thickness in the range of not less than 20 nm nor more than 60 nm, said upper plasma silicon nitride film having a complex index of refraction in which the real part and the imaginary part have values respectively in the range of not less than 1.7 nor more than 2.4 and in the range of not less than 0.15 nor more than 0.75, said upper plasma silicon nitride film having a thickness in the range of not less than 10 nm nor more than 40 nm.
  • 6. The semiconductor device manufacturing method according to claim 5,wherein said complex index of refraction of said lower plasma silicon nitride film is (2.1−i×1.2) and the thickness of said lower plasma silicon nitride film is 50 nm, and said complex index of refraction of said upper plasma silicon nitride film is (2.0−i×0.3) and the thickness of said upper plasma silicon nitride film is 30 nm.
  • 7. The semiconductor device manufacturing method according to claim 5, wherein said step (b-1) and said step (b-2) are continuously performed in the same chamber.
  • 8. The semiconductor device manufacturing method according to claim 5, wherein said step (b) further comprises a step (b-3) of exposing an exposed surface of said upper plasma silicon nitride film to a plasma of a gas which contains oxygen to form a thermal oxide film.
  • 9. The semiconductor device manufacturing method according to claim 5, further comprising the steps of;(d) forming a silicon oxide film covering said upper and lower plasma silicon nitride films, and (e) applying at least one of an etching process and a polishing process to said silicon oxide film.
Priority Claims (1)
Number Date Country Kind
2001-013262 Jan 2001 JP
US Referenced Citations (8)
Number Name Date Kind
5488246 Hayashide et al. Jan 1996 A
5721167 Subramanian et al. Feb 1998 A
5783365 Tsujita Jul 1998 A
5872054 Tsujita Feb 1999 A
6083852 Cheung et al. Jul 2000 A
6300671 Moore et al. Oct 2001 B1
6316372 DeBoer et al. Nov 2001 B1
6365320 Foote et al. Apr 2002 B1
Non-Patent Literature Citations (3)
Entry
Qizhi He, et al. “Inorganic Antireflective Coating Process for Deep-UV Lithography”, SPIE 98 vol. 3334, pp. 337-346.
Atsumi Yamaguchi, et al. “Improvement of CD control and resolution by optimizing inorganic ARC process”, SPIE 2000 vol. 4000, pp. 905-914.
Qun Ying Lin, et al. “Dual Layer Inorganic SiON Bottom ARC for 0.25 μ m DUV Hard Mask Applications”, SPIE 99 vol. 3678, pp. 186-197.