Background of the Invention
1. Field of the Invention
[0001] This invention relates to a method of producing a photomask. The invention can be
utilized for photomasks used for forming various pattern images, a method of producing
such photomasks, a method of exposing using such photomasks and a method of manufacturing
semiconductor devices using such photomasks. For example, it can be utilized for photomasks
used for the techniques of forming various patterns in semiconductor device manufacture
processes, a method of producing such photomasks and a method of exposing using such
photomasks. Further, it can be applied to an exposing apparatus and also to a method
of semiconductor device manufacture, for instance a method of manufacturing such semiconductor
devices as memory devices, logic operational devices, CCD devices, LCD devices, etc.
2. Description of the Related Art
[0002] The related art will now be described by taking the field of semiconductor devices
as an example. When manufacturing a semiconductor device, various patterns are formed.
In such semiconductor device manufacture, a pattern transfer process or commonly termed
lithographic process is used mainly for transferring a photomask pattern on a resist
material on a semiconductor wafer.
[0003] With a recent trend for finer semiconductor device structures, it is becoming more
and more difficult to obtain a fine pattern with a desired resolution. As an example,
the prior art photomask has a problem that the resist pattern size obtained by transfer
is not in accord with a fine mask pattern size and is smaller than a desired value.
Accordingly, in the prior art the mask pattern size is set to be greater than the
resist pattern size that is obtainable by the transfer. Further, to solve the above
problem and for the purpose of resolution improvement, investigations have been conducted
about making shorter the wavelength of light of exposing in exposure apparatus used
in the lithographic process, phase shifting masks for shifting the light phase, a
shape change illumination process, in which the shape of a light source is changed,
a pupil filter process, in which a filter is provided in an emission pupil in condenser
lens system, and a FLEX process, in which exposing is done a plurality of times at
different focus positions.
[0004] A general example of the prior art method will now be given. Fig. 1 is a graph illustrating
the concept underlying the prior art method example. In this method, the transfer
resist pattern size is determined by experiments or simulation with a plurality of
different defocus values, thus obtaining a mask pattern size versus defocus curve
1. From this curve 1, the range of mask pattern size in a design tolerance range 2
is obtained. From this mask pattern size range depth of focus 3 is obtained, the numerical
value of which indicates the performance of the lithography.
[0005] Further, there is a ED tree method which correlatively deals with the depth of focus
and the exposure dose latitude. An example of this method is shown in Fig. 2. Curves
41 to 46 as shown, represent the relation between the exposure dose and the defocus
for respective percentage changes in the transfer resist pattern size from the design
mask pattern size. Assuming a design tolerance condition of the transfer resist pattern
size that the percentage change therein from the design value is within ±10 % (see
curves 43 and 44 in Fig. 2) and that the exposure dose latitude that is necessary
is 20 % in range as shown by 5 in Fig. 2, the depth of focus is as shown by 3 in the
Figure.
[0006] In simulation evaluation in various prior art techniques, the parameters that are
evaluated are mostly two in number.
[0007] In
EP O 313 013 a lithographic process analysis and control system is described which provides a
modeled version of a lithographic process in the three dimensions of feature width,
focus and exposure. This system uses the model to quickly determine the range of focus
and exposure limits for obtaining the desired feature width.
[0008] In the prior art methods of evaluation as exemplified above, no considerations are
given to the fluctuations of the exposure dose and also the fluctuations of the mask
pattern size of the mask. Therefore, the depth of focus that is obtainable is greatly
aloof from the actual process condition, and it is greater than the actual depth of
focus. Besides, in this method it is impossible to obtain quantitative evaluation
of other parameters such as the exposure dose latitude and the mask pattern size latitude.
[0010] In the prior art method, the evaluation by simulation required an enormous amount
of experiments for making up for the great aloofness from the actual process condition.
The experiment requires time and cost, and in this case it is difficult to obtain
efficient and systematic evaluation. Particularly, in the distal device development
without established apparatus or material, there is a serious drawback that it is
very difficult to find out the relationship among various techniques.
[0011] Further, as patterns are finer, a serious problem is posed by adverse effects of
mask pattern size fluctuations on the transfer pattern, which has previously been
any problem. In the prior art, it has been impossible to carry out evaluation by taking
the mask pattern size fluctuations into considerations. Giving no considerations to
the mask pattern size fluctuations means that the evaluation has heretofore been made
under the assumption that the mask pattern size of the mask is in accord with the
design mask pattern size and is fixed at all times. Actually, however, there are fluctuations
in the mask production process, and it is impossible to perfectly eliminate the mask
pattern size fluctuations. Hence, it has been indispensable for proper condition setting
to let the mask pattern size fluctuations be reflected on the evaluation, but this
has heretofore been impossible.
Object and Summary of the Invention
[0012] An object of the invention is to provide a method of producing a photomask which
can solve the problems discussed above in the prior art, permit mutual correlation
of a large number of parameters, for instance three or more parameters, to be found
out, permit optimum condition to be obtained from such correlation, permit reduction
of aloofness from the actual process condition, permit quantitatively grasping various
performances, permit reduction of time and cost, permit taking the influence of mask
pattern size fluctuations into considerations and permit actual optimization.
[0013] To attain the above object of the invention, the present invention provides a method
as specified in claim 1.
[0014] In another method not forming part of the invention, when obtaining the exposure
dose latitude, combinations of plurality of defocus and mask pattern size values in
predetermined ranges of the defocus latitude and the mask pattern size latitude are
set. The combinations are suitably as finely set as possible in the predetermined
ranges of the defocus latitude and the mask pattern size latitude. On the basis of
and for each of these combinations, transfer patterns are obtained by varying the
exposure dose. The transfer patterns for the various exposure doses may be obtained
by a calculation operation using simulation or the like, by actual measurement in
experiments or by combining both these means. Suitably, the operation is carried out
in the mode of obtaining transfer resist patterns in the photolithography. The transfer
patterns which are obtainable by the above various means are checked as to whether
they meet the design tolerance condition. For example, a check is made as to whether
the size, area, shape, etc. of the patterns meets the tolerance condition at the time
of the design (i.e., at the time of semiconductor device design in case of a mask
for the semiconductor device formation). According to this check is obtained the exposure
dose latitude as the range of exposure dose meeting the tolerance condition in all
the predetermined ranges of the defocus latitude and the mask pattern size latitude.
At least one of mask parameters is set such as to maximize the exposure dose latitude.
The mask parameter may be the shape and size of the design pattern, the transmissivity
and phase of the transmitting area, the transmissivity and phase of the light shielding
area, etc.
[0015] A photomask with a transmitting area thereof having two portions for transmitting
light at mutually different phases is a commonly called phase shift mask. In this
case, it is possible to increase the resolution by setting the phase difference most
suitably to 180°. It is possible to provide other phase differences depending on the
design.
[0016] The phase difference may be provided by forming a phase shift section or a commonly
called shifter section by changing the thickness of a portion of a light transmissive
substrate, of glass for instance, by engraving that portion. Alternatively, it may
be provided by forming a phase shift section or commonly called shifter section with
a film of a phase shift material (a resist, SiO
2, etc.) such that the thickness of the film provides a phase difference of 180°, for
instance. Further, a portion for providing for a phase difference may be formed by
changing the optical path length by changing the refractive index with doping. Various
further means may also be used to this end.
[0017] A photomask with the light shielding area thereof transmits light in such an extent
as not to sensitize a sensitizer exposed in the photolithographic process and also
transmits light at a different phase from that of the transmitting area.
[0018] In a method of designing a photomask, the mask parameter is one member or a combination
of two or more members of the group consisting of the mask pattern size, the phase
of the transmitting area, the transmissivity of the light shielding area and the phase
of the light shielding area.
[0019] A method of designing a photomask having a transmitting area and a light shielding
area comprises the steps of setting a plurality of combinations of defocus and mask
pattern size values in predetermined ranges of the focus latitude and the mask pattern
size latitude, obtaining transfer patterns by varying the exposure dose, and checking
whether the transfer patterns meet a design tolerance condition, thereby obtaining
the exposure dose latitude as the range of exposure dose meeting the tolerance condition
in all the predetermined ranges of the focus latitude and the mask pattern size latitude,
a mask parameter being set for maximizing the obtained exposure dose latitude.
[0020] When obtaining the exposure dose latitude, combinations of pluralities of defocus
and mask pattern size values in predetermined ranges of the focus latitude and the
mask pattern size latitude are set. The combinations are suitably as finely set as
possible in the predetermined ranges of the focus latitude and the mask pattern size
latitude. On the basis of and for each of these combinations, transfer patterns are
obtained by varying the exposure dose. The transfer patterns for the various exposure
doses may be obtained by a calculation operation using simulation or the like, by
actual measurement in experiments or by combining both these means.
[0021] The transfer patterns obtained as above are typically transfer resist patterns in
the photolithography, and the operation is suitably carried out in the mode of obtaining
transfer resist patterns in the photolithography. The transfer patterns which are
obtainable by the above various means are checked as to whether they meet the design
tolerance condition. For example, a check is made as to whether the size, area, shape,
etc. of the patterns meets the tolerance condition at the time of the design (i.e.,
at the time of the semiconductor device design in case of a mask for the semiconductor
device formation). According to this check is obtained the exposure dose latitude
as the range of exposure dose meeting the tolerance condition in all the predetermined
ranges of the defocus latitude and the mask pattrern size latitude. At least one of
mask parameters is set such as to maximize the exposure dose latitude. The mask parameter
may be the shape and size of the design pattern, the transmissivity and phase of the
transmitting area, the transmissivity and phase of the light shielding area, etc.
[0022] In the above way, it is possible to obtain the condition for optimizing the mask
design, thus permitting production of the optimum mask.
[0023] When obtaining the mask design optimization condition, it is possible to simultaneously
set the optimizing condition for the exposing condition.
[0024] A method of exposing using a photomask having a transmitting area and a light shielding
area comprises the steps of setting combinations of pluralities of defocus and mask
pattern size values in predetermined ranges of the defocus latitude and the mask pattern
size latitude, obtaining a transfer pattern by varying the exposure dose, and checking
whether the transfer pattern meets a design tolerance condition, thereby obtaining
the exposure dose latitude as the range of exposure dose meeting the tolerance condition
in all the predetermined ranges of the defocus latitude and the size latitude, the
exposure being set for maximizing the obtained exposure dose latitude. Such exposure
parameter may be lens numerical aperture (NA), partial coherency, exposure wavelength,
light source shape, pupil filter structure, etc.
[0025] A further method of exposing with a photomask having a transmitting area and a light
shielding area comprises the steps of setting combinations of pluralities of defocus
and mask pattern size values in predetermined ranges of the defocus latitude and the
mask pattern size latitude, obtaining tansfer patterns by varying the exposure dose,
and checking whether the transfer patterns meet a design tolerance condition, thereby
obtaining the exposure dose latitude as the range of exposure dose meeting the tolerance
condition in all the predetermined ranges of the defocus latitude and the mask pattern
size latitude, a mask parameter being set for maximizing the obtained exposure dose
latitude.
[0026] When optimizing the photomask design condition for obtaining the exposure dose latitude
according to the predetermined defocus latitude and mask pattern size latitude, the
condition of exposing using the photomask is optimized simultaneously, thereby obtaining
the optimum exposing condition.
[0027] The optimum condition of exposing is advantageously determined like the setting of
the optimum condition of the mask formation.
[0028] A method of manufacturing a semiconductor device in photolithography uses a photomask
having a transmitting area and a blocking area, the photomask being obtained by setting
combinations of pluralities of defocus and mask pattern size values in predetermined
ranges of the defocus latitude and the mask pattern size latitude, obtaining transfer
patterns by varying the exposure dose, and checking whether the transfer patterns
meet a design tolerance condition, thereby obtaining the exposure dose latitude as
the range of exposure dose meeting the tolerance condition in all the predetermined
ranges of the defocus latitude and the mask pattern size latitude.
[0029] The semiconductor device may suitably be finely integrated LSIs, for instance logic
devices, CCD devices, LCD devices, memory devices, etc.
[0030] A photomask having a transmitting area and a blocking area is obtained, by a further
method not forming part of the invention, by setting combinations of pluralities of
exposure dose and mask pattern size values in predetermined ranges of the exposure
dose latitude and the mask pattern size latitude, obtaining transfer patterns by varying
the defocus in the neighborhood of the just focus, and checking whether the transfer
patterns meet a design tolerance condition, thereby obtaining the defocus latitude
as the range of defocus meeting the tolerance condition in all the predetermined ranges
of the predetermined ranges of the defocus latitude and the mask pattern size latitude,
a mask parameter being set for maximizing the obtained exposure dose latitude.
[0031] When obtaining the defocus latitude, combinations of pluralities of exposure dose
and mask pattern size values in predetermined ranges of the exposure dose latitude
and the mask pattern size latitude are set. The combinations are suitably as finely
set as possible in the predetermined ranges of the exposure dose latitude and the
mask pattern size latitude. On the basis of and for each of these combinations, transfer
patterns are obtained by varying the defocus in the neighborhood of the just focus.
(In the specification, by the term "just focus" is meant a focus position of the greatest
light intensity distribution and contrast.) The transfer patterns for the various
defocuses may be obtained by a calculation operation using simulation or the like,
by actual measurement in experiments or by combining both these means. Suitably, the
operation is carried out in the mode of obtaining transfer resist patterns in the
photolithography. The transfer patterns which are obtained by the above various means
are checked as to whether they meet the design tolerance condition. For example, a
check is made as to whether the size, area, shape, etc. of the patterns meets the
tolerance condition at the time of the design (i.e., at the time of semiconductor
device design in case of a mask for the semiconductor device formation). According
to this check is obtained the defocus latitude as the range of defocus meeting the
tolerance condition in all the predetermined ranges of the exposure dose latitude
and the mask pattern size latitude. At least one of mask parameters is set such as
to maximize the defocus latitude. The mask parameter may be the shape and size of
the design pattern, the transmissivity and phase of the transmitting area, and transmissivity
and phase of the light shielding area, etc.
[0032] A photomask with the transmitting area thereof has two portions for transmitting
light at mutually different phases.
[0033] As for the phase difference, it is most suitably set to 180° for resolution increase.
It is possible to provide other phase differences as well depending on the design.
[0034] The phase difference may be provided by forming a phase shift section or a commonly
called shifter section by changing the thickness of a portion of a light transmissive
substrate, of glass for instance, by engraving that portion. Alternatively, it may
be provided by forming a phase shift section or commonly called shifter section with
a film of a phase shift material (a resist, SiO
2, etc.) such that the thickness of the film provides a phase difference of 180°, for
instance. Further, a portion for providing for a phase difference may be formed by
changing the optical path length by changing the refractivity with doping. Various
further means may also be used to this end.
[0035] A photomask with the light shielding area thereof transmits light in such an extent
as not to sensitize a sensitizer in the photolithographic process and also transmits
light at a different phase from that of the transmitting area.
[0036] In a method of producing a photomask, the mask parameter is one member or a combination
of two or more members of the group consisting of the mask pattern size, the phase
of the transmitting area, the transmissivity of the light shielding area and the phase
of the light shielding area.
[0037] A method of producing a photomask having a transmitting area and a light shielding
area comprises the steps of setting combinations of pluralities of exposure dose and
mask pattern size values in predetermined ranges of the exposure dose latitude and
the mask pattern size latitude, obtaining transfer patterns by varying the defocus
in the neighborhood of the just focus, and checking whether the transfer patterns
meet a design tolerance condition, thereby obtaining the defocus latitude as the range
of defocus meeting the tolerance condition in all the predetermined ranges of the
exposure dose latitude and the mask pattern size latitude, a mask parameter set for
maximizing the obtained defocus latitude.
[0038] When obtaining the defocus latitude, combinations of pluralities of exposure dose
and mask pattern size values in predetermined ranges of the exposure dose latitude
and the mask pattern size latitude are set. The combinations are suitably as finely
set as possible in the predetermined ranges of the exposure dose latitude and the
mask pattern size latitude. On the basis of and for each of these combinations, transfer
patterns are obtained by varying the defocus in the neighborhood of the just focus.
The transfer patterns for the various defocuses may be obtained by a calculation operation
using simulation or the like, by actual measurement in experiments or by combining
both these means.
[0039] The transfer patterns that are obtained are typically transfer resist patterns in
photolithography, and suitably the operation is carried out in the mode of obtaining
transfer resist patterns in the photolithography. The transfer patterns which are
obtained by the above various means are checked as to whether they meet the design
tolerance condition. For example, a check is made as to whether the size, area, shape,
etc. of the pattern meets the tolerance condition at the time of the design (i.e.,
at the time of semiconductor device design in case of a mask for the semiconductor
device formation). According to this check is obtained the defocus latitude as the
range of defocus meeting the tolerance condition in all the predeteremined ranges
of the exposure dose latitude and the mask pattern size latitude. At least one of
the mask parameters is set such as to maximixe the defocus latitude. The mask parameters
may be the shape and size of the desgin pattern, the transmissivity and phase of the
transmitting area, the transmissivity and phase of the light shielding area, etc.
[0040] In the above way, the mask design optimizing condition is obtained, thus permitting
the manufacture of the optimum mask.
[0041] When obtaining the mask design optimization condition, the optimization condition
for the exposing condition can be obtained simultaneously.
[0042] A method of exposing in a photolithographic process, comprises the steps of setting
combinations of pluralities of exposure dose and mask pattern size values in all predetermined
ranges of the exposure dose latitude and the the mask pattern size latitude, obtaining
transfer patterns by varying the defocus, and checking whether the transfer patterns
meet a design tolerance condition, thereby obtaining the defocus latitude as the range
of defocus meeting the tolerance condition in all the predetermined ranges of the
exposure dose latitude and the mask pattern size latitude, an exposure parameter being
set for maximizing the defocus latitude. The exposure parameter may be lens numerical
aperture (NA), partial coherency, exposure wavelength, light source shape, pupil filter
shape, etc.
[0043] A method of exposing using a photomask having a transmitting area and a light shielding
area comprises the steps of setting combinations of pluralities of exposure dose and
mask pattern size values in predetermined ranges of the exposure dose latitude and
the mask pattern size latitude, obtaining transfer patterns by varying the defocus
in the neighborhood of the just focus, checking whether the transfer patterns meet
a design tolerance condition, and thereby obtaining the defocus latitude as the range
of defocus meeting the tolerance condition in all the predetermined ranges of the
exposure dose latitude and the mask pattern size latitude, a mask parameter being
set for maximizing the obtained defocus latitude.
[0044] The optimum condition of the photomask can be set by means described above.
[0045] When optimizing the photomask design condition for obtaining the defocus latitude
according to the predetermined exposure dose latitude and the mask pattern size latitude,
the condition of exposing using the photomask is optimized simultaneously to obtain
the optimum exposing condition.
[0046] The optimum condition of exposing can be advantageously determined in a similar way
to the setting of the optimum condition of the mask formation.
[0047] A method of manufacturing a semiconductor device in photolithography using a photomask
having a transmitting are and a light shielding area comprises the steps of setting
combinations of pluralities of exposure dose and mask pattern size values in predetermined
ranges of the exposure dose latitude and the mask pattern size latitude, obtaining
transfer patterns by varying the defocus in the neighborhood of the just focus, and
checking whether the transfer patterns meet a design tolerance condition, thereby
obtaining the defocus latitude as the range of defocus meeting the tolerance condition
in all the predetermined ranges of the exposure dose latitude and the mask pattern
size latitude, a mask parameter being set for maximizing the defocus latitude.
[0048] The semiconductor device may suitably be finely integrated LSIs, for instance logic
circuits, CCDs, LCDs, memory devices, etc.
[0049] The above object of the invention is further attained by a method of producing a
photomask having a transmitting area and a light shielding area, by setting combinations
of pluralities of exposure dose and defocus values in predetermined ranges of the
exposure dose latitude and the defocus latitude, obtaining transfer patterns by varying
the mask pattern size, and checking whether the transfer patterns meet a design tolerance
condition, thereby obtaining the mask pattern size as the range of mask pattern size
meeting the tolerance condition in all the predetermined ranges of the exposure dose
latitude and the defocus latitude, a mask parameter being set for maximizing the mask
pattern size latitude.
[0050] When obtaining the mask pattern size latitude, combinations of pluralities of exposure
dose and defocus values in predetermined ranges of the exposure dose latitude and
the defocus latitude are set. The combinations are suitably as finely set as possible
in the predetermined ranges of the exposure dose latitude and the defocus latitude.
On the basis of and for each of these combinations, transfer patterns are obtained
by varying the mask pattern size in the neighborhood of a predetermined mask pattern
size. The transfer patterns for the various mask pattern sizes may be obtained by
a calculation operation using simulation or the like, by actual measurement in experiments
or by combining both these means. Suitably, the operation is carried out in the mode
of obtaining transfer resist patterns in the photolithography. The transfer patterns
which are obtainable by the above various means are checked as to whether they meet
the design tolerance condition. For example, a check is made as to whether the size,
area, shape, etc. of the patterms meet the tolerance condition at the time of the
design (i.e., at the time of semiconductor device design in case of a mask for the
semiconductor device formation). According to the check is obtained the mask pattern
size latitude as the range of mask pattern size meeting the tolerance condition in
all the predetermined ranges of the exposure dose latitude and the defocus latitude.
At least one of mask parameters is set such as to maximize the mask pattern size latitude.
The mask parameter may be the shape and size of the design pattern, the transmissivity
and phase of the transmitting area, the transmissivity and phase of the light shielding
area, etc.
[0051] The above object of the invention is further attained by a method of producing a
photomask with the transmitting area thereof having two portions for transmitting
light at mutually different phases.
[0052] The phase difference is most suitably 180° for obtaining an increased resolution.
It is possible to provide other phase differences as well depending on the design.
[0053] The phase difference may be provided by forming a phase shift section or a commonly
called shifter section by changing the thickness of a portion of a light transmissive
substrate, of glass for instance, by engraving that portion. Alternatively, it may
be provided by forming a phase shift section or commonly called shifter section with
a film of a phase shift material (a resist, SiO
2, etc.) such that the thickness of the film provides a phase difference of 180°, for
iinstance. Futher, a portion providing for a phase difference may be formed by changing
the Optical path length by changing the refractivity with doping. Various further
means may also be used to this end.
[0054] The above object of the invention is further attained by a method of producing a
photomask with the light shielding area thereof transmitting light in such an extent
as not to sensitize a sensitizer exposed in the photographic process and also transmitting
light at a different phase from that of the transmitting area.
[0055] The above object of the invention is further attained by a method of producing a
photomask, in which the mask parameter is one member or a combination of two or more
members of the group consisting of the mask pattern size, the phase of the transmitting
area, the transmissivity of the light shielding area and the phase of the light shielding
area.
[0056] The above object of the invention is in general attained by a method of producing
a photomask having a transmitting area and a light shielding area, which comprises
the steps of setting combinations of pluralities of exposure dose and defocus values
in predetermined ranges of the exposure dose latitude and the defocus latitude, obtaining
transfer patterns by varying the mask pattern size latitude, and checking whether
the transfer patterns meet a design tolerance condition, thereby obtaining the mask
pattern size latitude as the range of mask pattern size meeting the tolerance condition
in all the predetermined ranges of the exposure dose latitude and the defocus latitude,
a mask parameter being set for maximizing the mask pattern size latitude.
[0057] According to the invention, when obtaining the mask pattern size latitude, combinations
of pluralities of exposure dose and defocus values in predetermined ranges of the
exposure dose latitude and the defocus latitude are set. The combinations are suitably
as finely set as possible in the predetermined ranges of the exposure dose latitude
and the defocus latitude. On the basis of and for each of these combinations, transfer
patterns are obtained by varying the mask pattern size. The transfer patterns for
various defocuses may be obtained by a calculating operation using simulation or the
like, by actual measurement in experiments or by combining both these means.
[0058] The transfer patterns are typically transfer resist patterns in photolithography,
and the operation is suitably carried out in the mode of obtaining transfer resist
patterns in the photolithography. The transfer patterns which can be obtained by the
above various means are checked as to whether they meet the design tolerance condition.
For example, a check is made as to whether the size, area, shape, etc. of the patterns
meets the tolerance condition at the time of the design (i.e., at the time of semiconductor
device design in case of a mask for the semiconductor device formation). According
to this check is obtained the mask pattern size latitude as the range of mask pattern
size meeting the tolerance condition in all the predetermined ranges of the exposure
dose latitude and the mask pattern size latitude. At least one of mask parameters
is set such as to maximize the mask pattern size latitude. The mask parameter may
be the shape and size of the design pattern, the transmissivity and shape of the transmitting
area, the transmissivity and phase of the light shielding area, etc.
[0059] In the above way, the mask design optimization condition is obtainable for producing
an optimum mask.
[0060] When obtaining the mask design optimizatin condition, the optimization condition
of the exposing condition can be obtained simultaneously.
[0061] A method of exposing in a photographic process comprises the steps of setting combinations
of pluralities of exposure dose and defocus values in predetermined ranges of the
exposure dose latitude and the defocus latitude, obtaining transfer patterns by varying
the mask pattern size, and checking whether the transfer patterns meet a design tolerance
condition, thereby obtaining the mask pattern size latitude as the range of mask pattern
size meeting the tolerance condition in all the predetermined ranges of the exposure
dose latitude and the defocus latitude, an exposure parameter being set for maximizing
the mask pattern size latitude.
[0062] A method of exposing using a photomask having a transmitting area and a light shielding
area comprises the steps of setting combinations of pluralities of exposure dose and
defocus values in predetermined ranges of the exposure dose latitude and the defocus
latitude, obtaining transfer patterns by varying the mask pattern size, and checking
whether the transfer patterns meet a design tolerance condition, thereby obtaining
the mask pattern size latitude as the range of mask pattern size meeting the tolerance
condition in all the predetermined ranges of the exposure dose latitude and the defocus
latitude, a mask parameter being set for maximizing the mask pattern size latitude.
[0063] The optimum condition of the photomask can be set by the means described above.
[0064] A method of exposing uses a photomask having a transmitting area and a light shielding
area, in which when optimizing the photomask design condition for obtaining the mask
pattern size latitude according to the predetermined defocus latitude and exposure
dose latitude, the condition of exposing using the photomask is optimized simultaneously
to obtain the optimum exposing condition.
[0065] The optimum condition of exposing when using the mask can be advantageously determined
in a manner similar to that of setting the optimum condition of the mask formation.
[0066] A method of manufacturing a semiconductor device in photolithography uses a photomask
having a transmitting area and a light shielding area and comprises the steps of setting
combinations of pluralities of exposure dose and defocus values in predetermined ranges
of the exposure dose latitude and the defocus latitude, obtaining transfer patterns
by varying the mask pattern size, and checking whether the transfer patterns meet
a design tolerance condition, thereby obtaining the mask pattern size latitude as
the range of mask pattern size meeting the tolerance condition in all the predetermined
ranges of the exposure dose latitude and the defocus latitude, a mask parameter being
set for maximizing the mask pattern size latitude.
[0067] The semiconductor device may be finely integrated LSIs, for instance logic circuits,
CCDs, LCDs, memory devices, etc.
[0068] Referring to Figure 3, when obtaining the defocus latitude, combinations of pluralities
of exposure dose and mask pattern size values in predetermined ranges of the exposure
dose latitude (target exposure dose latitude I) and the mask pattern size latitude
(mask line width latitude II as target transfer pattern size) are set. The combinations
are suitably as finely set as possible in the predetermined ranges of the exposure
dose latitude I and the mask pattern size latitude II. Further, on the basis of and
for each of these combinations, transfer patterns are obtained by varying the defocus
in the neighborhood of the just focus. The transfer patterns are obtained by simulation,
by experiments or by combining both these means. In consequence, it can be known that
the transfer pattern that is obtained at each point meets or does not meet a resolution
condition. The boundary R between the zone, in which the resolution condition is met,
and the zone otherwise, can be obtained by taking a plurality of (as many as possible)
points. In Fig. 3, the resolution condition is met in the zone under the boundary
R. On the basis of this, the depth of focus DOF III on the R surface is obtained,
and from this value the defocus latitude is obtained.
[0069] On this basis, a mask parameter is set for maximizing the defocus latitude.
[0070] In the method of photomask manufacture according to the invention, the optimum value
of the photomask is determined by combining the mask pattern size latitude with pluralities
of data in predetermined ranges of the exposure dose latitude and the defocus latitude
and obtaining the permissible range of the combinations. It is thus possible to obtain
the optimum condition for the mask formation.
[0071] In the method of exposing the photomask which is formed under the above optimum condition
is used. Thus, the photomask can be used for satisfactory pattern formation. Further,
it is possible to make exposing under the optimized proper exposing condition.
[0072] In the method of manufacturing a semiconductor device the photomask which is formed
under the above optimum condition is used. Thus, it is possible to obtain a semiconductor
device having satisfactory performance, and the method of suited for finer integration.
Brief Description of the Drawings
[0073]
Figs. 1 and 2 are views for describing a method of exposing in the prior art;
Fig. 3 is a view for describing the constitution of the invention; and
Fig. 4 is a contour line view corresponding to threshold providing transfer pattern
size, as obtained from exposure dose and resist sensitivity, for describing embodiments
of the invention.
Detailed Description of the Preferred Embodiments
[0074] Now, embodiments of the invention will be described in detail with reference to the
drawings without any sense of limiting the invention as well as examples which are
useful for understanding the invention but do not form part of the invention.
[0075] In the transfer pattern shape formation by exposing on a resist material formed on
a material to be exposed, for instance a semiconductor wafer, what is used for reduction
projection may be referred to as reticle, and what is used for unity magnification
projection may be referred to as mask. Also, what corresponds to a master plate may
be called mask, and what is obtained by duplicating such a master plate may be called
mask. In the specification, the reticles and masks having the above various meanings
are collectively referred to as mask.
[0076] A first exemplary embodiment will now be described in detail.
[0077] This is an example of an attenuated phase-shifting mask under exposing conditions
of exposure wavelength of 248 nm, NA of 0.45 and
σ of 0.3.
[0078] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the mask pattern size latitude was set to ± 0.05 µm (on 5 times reticle).
In case of transferring a 0.3 µm contact hole, five combinations of the half-light
shielding area amplitude transmissibity and the mask pattern size (on 5 times reticle)
were set as 25 % and 1.50 µm, 30 % and 1.60 µm, 35 % and 1.75 µm, 40 % and 1.85 µm,
and 45 % and 1.95 µm. With each of these different masks, the following was made.
[0079] In the first place, for each of these five masks the exposure dose was set such that
a transfer contact hole of 0.3 µm could be obtained under the conditions of defocus
of 0 µm and no mask pattern size deviation.
[0080] Then, the defocus was set to 0 µm, ± 0.25 µm, ± 0.50 µm, ± 0.75 µm and ± 1.00 µm.
The mask pattern size deviation on the 5 times reticle was set to -0.05 µm, 0.00 µm
and 0.05 µ m. The deviation of the exposure dose from the setting value was set to
-20 %, -15 %, -10 %, -5 %, 0 %, 5 %, 10 %, 15 % and 20 %. The simulation parameters
as noted above are desirably continuous values, but discrete values were provided
by taking the calculation time into considerations. For all these combinations of
parameter values, light intensity distributions were obtained by using a light intensity
simulator based on the scalar diffraction theory.
[0081] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0082] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check is made as to whether the size meets a predetermined contact
hole size tolerance condition. The resist pattern size tolerance condition was set
such that the deviation from a design mask pattern size of 0.3 µm was within 10 %.
[0083] From the result of the check, the range of exposure dose, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size values meet
a predetermined tolerance condition, was obtained and made to be the exposure dose
latitude.
[0084] Among the five combinations of the half-light shielding area transmissivity and mask
pattern size obtained in the above procedure, the greatest exposure dose latitude
could be obtained with the amplitude transmissivity of the half-light shielding area
transmissivity of 45 % and the mask pattern size of 1.95 µ m.
[0085] The mask obtained in this example sufficiently met the depth of focus of 2.0 µm and
the mask pattern size latitude of ± 0.05 µm (on 5 times reticle). Further, it had
sufficiently great exposure dose latitude to permit exposing with sufficient latitude,
and it was possible to obtain a sharp resist pattern shape.
[0086] While this example was applied to the attenuated phase-shifting mask, this is by
no means limitative, similar effects are obtainable with masks of conventional systems
and also with phase shift masks of other systems. Further, while this example concerned
with contact hole patterns, this is only exemplary, and similar effects are obtainable
with other patterns than the contact hole patterns.
[0087] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0088] Further, as for the defocus latitude, mask pattern size latitude and mask pattern
size latitude, the values in this example are by no means limitative, and it is possible
to adopt other conditions as well.
[0089] A second exemple will now be described.
[0090] This is an example of the optimization of the exposing condition of a half-tone system
phase shift mask. In this example, with exposure wavelength of 248 nm and NA of 0.45,
judgment is made as to which of exposing conditions of σ of 0.3 nd 0.5 is to be adopted.
[0091] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the mask pattern size latitude was set to ± 0.05 µm (on 5 times reticle).
Further, the amplitude transmissivity of the half-light shielding area of the half-tone
system phase shift mask was set to 40 %, and the mask pattern size (on 5 times reticle)
when transferring a contact hole of 0.3 µm was set to 1.85 µ m in case of σ of 0.3
and to 1.75 µm in case of σ of 0.5. Under these two different exposing conditions,
the following was made.
[0092] In the first place, under each of the two exposing conditions the exposure dose was
set such that a transfer contact hole of 0.3 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size deviation.
[0093] Then, the defocus was set to 0 µm, ± 0.25 µm, ± 0.50 µm, ± 0.75 µm and ±1.00µm. The
mask pattern size deviation on the 5 times reticle was set to -0.05 µm, 0.00 µm and
0.05 µ m. The deviation of the exposure dose from the setting value was set to -20
%, -15 %, -10 %, -5 %, 0 %, 5 %, 10 %, 15 % and 20 %. The simulation parameters as
noted above are desirably continuous values, but discrete values were provided by
taking the calculation time into considerations. For all these combinations of parameter
values, light intensity distributions were obtained by using a light intensity simulator
based on the scalar diffraction theory.
[0094] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, where obtained and
made to be transfer resist patterns.
[0095] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.3 µm was within
10 %.
[0096] From the result of this check, the range of exposure dose, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size valuee meet
a predetermined tolerance condition, was obtained and made to be the exposure dose
latitude.
[0097] Of the above two different values of σ set in the above procedure, the maximum exposure
dose latitude could be obtained with σ = 0.3. Thus, σ = 0.3 was adopted.
[0098] The mask obtained in this example sufficiently met the depth of focus of 2.0 µm and
the mask pattern size latitude of ± 0.05 m (on 5 times reticle). Further, it had sufficiently
great exposure dose latitude to permit exposing with sufficient latitude, and it was
possible to obtain a sharp resist pattern shape.
[0099] While this example was applied to the half-tone system phase shift mask, this is
by no means limitative, similar effects are obtainable with masks of conventional
system and also with phase shift masks of other systems. Further, while this example
concerned with contact hole patterns, this is only exemplary, and similar effects
are obtainable with other patterns than the contact hole patterns.
[0100] Further, while transfer resist patterns were obtained by a process of obtaining the
contour lines of the light intensity distribution obtained by the light intensity
simulation based on the scaler diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distribution
by experiments. In obtaining transfer resist patterns from light intensity distribution,
it is possible use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0101] Further, as for the defocus latitude, mask pattern size latitude and resist patern
size latitude, the values in this example are by no means limitative, and it is possible
to adopt other conditions as well.
[0102] A third exemple embodiment will now be described.
[0103] This is an example of an application to the optimization of the exposing condition
for a photomask having a transmitting area and a light shielding area. In this example
a judgment was made as to which one of four different exposing conditions is to be
adopted, the conditions being such that σ of 0.3 and 0.5 with exposure wavelength
of 248 nm and NA of 0.45 and σ of 0.3 and 0.5 with exposure wavelength of 365 nm and
MA of 0.57.
[0104] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the mask pattern size latitude was set to ± 0.05 µm (on 5 times reticle).
In case of transferring a 0.5 µm contact hole, with the mask pattern size set to 2.5
µm (on 5 times reticle) the following was made under each of the four exposing conditions.
[0105] In the first place, under each of the four exposing conditions the exposure dose
was set such that a transfer contact hole of 0.5 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size deviation.
[0106] Then, the defocus was set to 0 µm, ± 0.25 µm and ± 0.50 µm, ± 0.75 µm and ± 1.00
µm. The mask pattern size deviation on the 5 times reticle was set to -0.05 µm, 0.00
µm and 0.05 µm. The deviation of the exposure dose from the setting value was set
to -20 %, -15 %, -10 %, -5 % %, 0 %, 5 %, 10 %, 15 % and 20 %. The simulation parameters
as noted above are desirably continuous values, but discrete values were provided
by taking the calculation time into considerations. For all these combinations of
parameter values, light intensity distributions were obtained by using a light intensity
simulator based on the scalar diffraction theory.
[0107] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0108] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check is made as to whether the size meets a predetermined contact
hole size tolerance condition. The resist pattern size tolerance condition was set
such that the deviation from a design mask pattern size of 0.5 µm was within 10 %.
[0109] From the result of the check, the range of exposure dose, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size values meet
a predetermined tolerance condition, was obtained and made to be exposure dose latitude.
[0110] Among the four exposing conditions obtained in the above procedure, the maximum light
exposure latitude was obtained in the case of the exposure wavelength of 248 nm, NA
of 0.45 and σ of 0.3. This condition was thus adopted.
[0111] The mask obtained in this example sufficiently met the depth of focus of 2.0 µm and
the mask pattern size latitude of ± 0.5 µm (on 5 times reticle). Further, it has sufficiently
great exposure dose latitude to permit exposing with sufficient latitude, and it was
possible to obtain a sharp resist pattern shape.
[0112] While this example was applied to a photomask having a transmitting area and a light
shielding area, this is by no means limitative, and similar effects are obtainable
with phase shift masks, etc. Further, while this example concerned with contact hole
patterns, this is only exemplary, and similar effects are obtainable with other patterns
than the contact hole patterns.
[0113] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0114] Further, as for the defocus latitude, mask pattern size latitude and resist pattern
size latitude, the values in this embodiment are by no means limitative, and it is
possible to adopt other conditions as well.
[0115] A fourth example will now be described.
[0116] This is an example of a photomask having an exposing area and a transmitting area
under the exposing conditions of exposing wavelength of 365 nm, NA of 0.57 and σ of
0.3.
[0117] As lithographic process setting values, the defocus latitude was set to 1.0 µm (±
0.5 µm), and the mask pattern size latitude was set to ± 0. 5 µm (on 5 times reticle).
In case of transferring a 0.5 µm contact hole, four different mask pattern sizes of
2.40 µm, 2.45µm, 2.50µm and 2.55µm were set on 5 times reticle. With each of these
mask pattern sizes, the following was made.
[0118] In the first place, for each of these four masks, the exposure dose was set such
that a transfer contact hole of 0.5 µm could be obtained under the conditions of defocus
of 0 µm and no mask pattern size deviation.
[0119] Then, the defocus was set to 0
µm ±0.25 µm and ± 0.5
µ m. The mask pattern size deviation on the 5 times reticle was set to -0.05 µm, 0.00
µm and 0.05 µm. The deviation of the exposure dose from the setting value was set
to -20 %, -15 %, -10 %, -5 %, 0 %, 5 %, 10 %, 15 % and 20 %. The simulation parameters
as noted above are desirably continuous values, but discrete values were provided
by taking the calculation time into considerations. For all these combinations of
parameter values, light intensity distributions were obtianed by using a light intensity
simulator based on the scalar diffraction theory.
[0120] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer patterns.
[0121] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.5 µm was within
10 %.
[0122] From the result of the check, the range of exposure dose, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size values meet
a predetermined tolerance condition, was obtained and made to be the exposure dose
latitude.
[0123] Among the four patterns obtained by the above procedure, the maximum exposure latitude
coupled be obtained with the mask pattern size of 2.40 µm. Thus, this condition was
adopted.
[0124] The mask obtained in this example sufficiently met the depth of focus of 1.0 µm and
the mask pattern size latitude of ± 0.05 µm (on 5 times reticle). It had sufficiently
great exposure dose latitude to permit exposing with sufficient latitude, and it was
possible to obtain a sharp resist pattern shape.
[0125] While this example was applied to a photomask having a transmitting area and a light
shielding area, this is by no means limitative, and similar effects are obtainable
with phase shift masks or the like. Further, while this example concerned with contact
hole patterns, this is only exemplary, and similar effects are obtainable with other
patterns than the contact hole patterns.
[0126] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction therory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation precesses. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0127] Further, as for the defocus latitude, mask pattern size latitude and resist pattern
size latitude, the values in this embodiment are by no means limitative, and it is
possible to adopt other conditions as well.
[0128] A fifth example will now be described.
[0129] This is an example of a half-tone system phase shift mask under the exposing conditions
of exposure wavelength of 246 nm, NA of 0.45 and σ of 0.3.
[0130] As lithographic process setting values, the exposure dose latitude was set to ± 5
%, and the mask pattern size latitude was set to ± 0.05 µm (5 times reticle). For
transferring a contact hole of 0.3 µm, five different combinations of the amplitude
transmissivity of the half-light shielding area and mask pattern size (on 5 times
reticle) were set as 25 % and 1.50 µm, 30 % and 1.60 µm, 35 % and 1.75 µm, 40 % and
1.85 µm, and 45 % and 1.95 µm. With each of these five different masks, the following
was made.
[0131] In the first place, for each of these five masks the exposure dose was set such that
a transfer contact hole of 0.3 µm could be obtained under the conditions of defocus
of 0 µm and no mask pattern size deviation.
[0132] Then, the deviation of the exposure dose from the setting value was set to -5 %,
0 % and 5 %. The mask pattern size deviation on 5 times reticle was set to -0.05 µ
m, 0.00 µm, and 0.05 µm. Further, the defocus was set to 0 µm, ±0.25 µm, ±0.50µm,
±0.75µm, ±1.00µm, and ± 1.25 µm. The simulation parameters as noted above are desirably
continuous values, but discrete values were provided by taking the calculation time
into considerations. For all these combinations of parameter values, light intensity
distributions were obtained by using a light intensity simulator based on the scalar
diffraction theory.
[0133] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0134] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.3 µm was within
10 %.
[0135] From the result of the check, the range of defocus, in which the resist patterns
obtained with all the combinations of the exposure dose and mask pattern size values
meet a predetermined tolerance condition, was obtained and made to be the defocus
latitude.
[0136] Among the five combinations of the half-light shielding area transmissivity and mask
pattern size obtained by the above procedure, the maximum defocus latitude could be
obtained with the combination of the amplitude transmissivity of the half-light shielding
area of 45 % and the mask pattern size of 1.95 µm.
[0137] The mask obtained in this example sufficiently met the exposure dose latitude of
± 5 % and the mask pattern size latitude of ± 0.05 µm (on 5 times reticle). Further,
it had sufficiently great defocus latitude to permit exposing with sufficient latitude,
and it was possible to obtain a sharp resist pattern shape.
[0138] While this example was applied to the half-tone system phase shift mask, this is
by no means limitative, and similar effects are obtainable with conventional system
masks and other system phase shift masks. Further, while this example concerned with
contact hole patterns, this is only exemplary, and similar effects are obtainable
with other patterns than the contact hole patterns.
[0139] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0140] Further, as for the exposure dose latitude, defocus latitude and mask pattern size
latitude, the values in this example are by no means limitative, and it is possible
to adopt other conditions as well.
[0141] A sixth example will now be described.
[0142] This is an example of an optimization of the exposing condition of a half-tone system
shift mask. In this example, with exposure wavelength of 248 nm and NA of 0.45, a
judgment was made as to which one of the exposing conditions of σ of 0.3 and 0.5 is
to be adopted.
[0143] As lithographic process setting values, the exposure dose latitude was set to ± 5
%, and the mask pattern size latitude was set to ± 0.05 µm (5 times reticle). Further,
the amplitude transmissivity of the half-tone system phase shift mask was set to 40
%, and the mask pattern size (on 5 times reticle) when transferring a contact hole
of 0.3 µm was set to 1.85 µm in case of σ of 0.3 and 1.75 µm in case of σ of 0.5.
Under each of these two exposing conditions, the following was made.
[0144] In the first place, under each of the two exposing conditions the exposure dose was
set such that a transfer contact hole of 0.3 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size deviation.
[0145] Then, the deviation of the exposure dose from the setting value was set to -5 %,
0 % and 5 %. The mask pattern size deviation on 5 time reticle was set to -0.05 µ
m, 0.00 µm, and 0.05 µm. Further, the defocus was set to 0 µm, ± 0.25 µm ± 0.50 µm,
± 0.75 µm, ± 1.00 µm, and 1.25 µm. The simulation parameters as noted above are ±
desirably continuous values, but discrete values were provided by taking the calculation
time into considerations. For all these combinations of parameter values, light intensity
distributions were obtained by using a light intensity simulator based on the scalar
diffraciton theory.
[0146] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0147] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set set such that the deviation from a design mask pattern size of 0.3 was within
10 %.
[0148] From the result of the check, the range of defocus, in which resist patterns obtained
with all the combinations of the exposure dose and mask pattern size values meet a
predetermined tolerance condition, was obtained and made to be the defocus latitude.
[0149] Of the two values of σ obtined by the above procedure, the maximum defocus latitude
could be obtained with σ = 0.3. Thus, σ = 0.3 was adopted.
[0150] The mask obtained in this example sufficiently met the exposure dose latitude of
±5 % and the mask pattern size latitude of ± 0.05 µm (on 5 times reticle). Further,
it has sufficiently great defocus latitude to permit exposing with sufficient latitude,
and it was possible to obtain a sharp resist pattern shape.
[0151] While this example was applied to the half-tone system phase shift mask, this is
by no means limitative, and similar effects are obtainable with masks of conventional
systems and also with phase shift masks of other systems. Further, while this example
concerned with contact hole patterns, this is only exemplary, and similar effects
are obtainable with other patterns than the contact hole patterns.
[0152] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0153] Further, as for the exposure dose latitude, mask pattern size latitude and resist
pattern size latitude, the values in this example are by no means limitative, and
it is possible to adopt other conditions as well.
[0154] A seventh example will now be described.
[0155] This is an example of an optimization of the exposing conditions of a photomask having
a transmitting area and a light shielding area. In this example, a judgment was made
as to which one of four different exposing conditions, namely those in which σ is
0.3 and 0.5 with exposure wavelength of 248 nm and NA of 0.45 and those in which σ
is 0.3 and 0.5 with exposure wavelength of 365 nm and NA of 0.57, is to be adopted.
[0156] As lithographic process setting values, the exposure dose latitude was set to ± 5
%, and the mask pattern size latitude was set to ± 0.05 µm (on 5 times reticle). Further,
the mask pattern size when tansferring a contact hole of 0.5 µm was set to 2.5 µm
(on 5 times reticle)L. Under each of these four exposing conditions, the following
was made.
[0157] In the first place, under each of the four exposing conditions, the exposure dose
was set such that a transfer contact hole of 0.5 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size deviation.
[0158] Then, the deviation of the exposure dose from the setting value was set to -5 %,
0 % and 5 %, and the deviation of the mask pattern size on the 5 times reticle was
set to -0.05 µm, 0.00 µm, and 0.05 µm. The defocus was set to 0 µm, ± 0.25 µm, ± 0.50
µm, ± 0.75 µm, ± 1.00 µm, and ± 1.25 µm. The simulation parameters as noted above
are desirably continuous values, but discrete values were provided by taking the calculation
time into considerations. For all these combinations of parameter value, light intensity
distributions were obtained by using a light intensity simulator based on the scalar
diffraction theory.
[0159] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0160] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.5 µm was within
10 %.
[0161] From the result of the check, the range of defocus, in which resist pattens obtained
in all the combinations of the exposure dose and mask pattern size values Meet a predetermined
tolerance condition, was obtained and made to be the defocus latitude.
[0162] Among the four exposing conditions obtained in the above procedure, the maximum defocus
latitude could be obtained in the case with exposure wavelength of 248 nm, NA of 0.45
and σ of 0.3. This condition was adopted.
[0163] The mask obtained in this example sufficiently met the exposure dose latitude of
± 5 % and mask pattern size latitude of ± 0.05 µm (on 5 times reticle). Further, it
had sufficiently great defocus latitude to permit exposing with sufficient latitude,
and it was possible to obtain a sharp resist pattern shape.
[0164] While this example was applied to the photomask having a transmitting area and a
light shielding area, this is by no means limitative, and similar effects are obtainable
with phase shit masks and the like as well. Further, while this example concerned
with contact hole patterns, this is only exemplary, and similar effects are obtainable
with other patterns than the contact hole patterns.
[0165] Futher, while the transfer patterns were obtained by the process of obtaining the
contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0166] Further, as for the exposure dose latitude, mask pattern size latitude and resist
pattern size latitude, the values in this example are by no means limitative, and
it is possible to adopt other conditions as well.
[0167] An eighth example will now be described.
[0168] This is an example of a photomask having a transmitting area and a light shielding
area under exposing conditions of exposure wavelength of 365 nm, NA of 0.57 and σ
of 0.3.
[0169] As lithographic process setting values, the exposure dose latitude was set to ± 5
%, and the mask pattern size latitude was set to ± 0.05 µm (5 times reticle). In addition,
the mask pattern size on 5 times reticle when transferring 0.5 µm contact hole was
set to 2.40 µm, 2.45 µm, 2.50 µm and 2.55 µm. With these four different patterns,
the following was made.
[0170] In the first place, for each of the four masks the exposure dose was set such that
a transfer contact hole of 0.3 µm could be obtained under the conditions of defocus
of 0 µm and no poattern size deviation.
[0171] Then, the deviation of the exposure dose from the setting value was set to -5 %,
0 % and 5 %, and the deviation of the mask pattern size on the 5 times reticle was
set to -0.05 µm, 0.00 µm and 0.05 µm. The defocus was set to 0 µm, ± 0.25 µm, ± 0.50
µm, ± 0.75 µm, ± 1.00 µm, and ± 1.25 µm. The simulation parameters as noted above
are desirably continuous values, but discrete values were provided by taking the calculation
time into considerations. For all these combinations of parameter values, light intensity
distributions were obtained by using a light intensity simulator based on the scalar
diffraction theory.
[0172] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in exposing apparatus and
empirically obtained resist sensitivity. Further, contour lines 6 as shown in Fig.
4, corresponding to the light intensity threshold values, where obtained and made
to be transfer resist patterns.
[0173] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that deviation from a design mask pattern size of 0.5 µm was within 10
%.
[0174] From the result of the cheek the range of defocus, in which resist patterns obtained
with all the combinations of the exposure dose and mask pattern size values meet a
predetermined tolerance condition, was obtained and made to be the defocus latitude.
[0175] Among the four different pattern sizes obtained in the above procedure, the maximum
defocus latitude could be obtained with the mask pattern size of 2.40 µm. This condition
was adopted.
[0176] The mask obtained in this example sufficiently met the exposure dose latitude of
± 5 % and the mask pattern side latitude of ± 0.05 µm (5 times reticle). Further,
it had sufficiently great defocus latitude, and it was possible to obtain a sharp
resist pattern shape.
[0177] While this example was applied to the photomask having a transmitting area and a
light shielding area, this is by no means limitative, and similar effects are obtainable
with phase shift masks and the like as well. Further, while this example concerned
with contact hole patterns, this is only exemplary, and similar effects are obtainable
with other patterns than the contact hole pattens.
[0178] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0179] Further, as for the exposure dose latitude, mask pattern size latitude and resist
pattern latitude, the values in this example are by no means limitative, and it is
possible to adopt other conditions as well.
[0180] A first embodiment of the invention will now be described.
[0181] This embodiment is an example of application of the invention to a half-tone system
phase shift mask under exposing conditions of exposure wavelength of 248 nm, NA of
0.45 and 0 of 0.3.
[0182] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the exposure dose latitude of ± 5 %. When transferring a contact hole
of 0.3 µm, five combinations of the amplitude transmissivity of the half-light shielding
area and mask pattern size (on 5 times reticle) are set as 25 % and 1.50 µm, 30 %
and 1.60 µm, 35 % and 1.75 µm, 40 % and 1.85 µm, and 45 % and 1.95 µm. With each of
these five different masks, the following was made.
[0183] In the first place, for each of these five masks the exposure dose was set such that
a transfer contact hole of 0.3 µm could be obtained under the conditions of defocus
of 0 µm and no mask pattern size deviation.
[0184] Then, the defocus was set to 0 µm, ±0.25µm, ±0.50µm, 0.7 µm, ± 1.00 µm, and the deviation
of the exposure dose from the setting value was set to -5 %, 0 % and 5 %. The deviation
of the mask pattern size on the 5 times reticle was set to -0.20 µm, -0.15 µm, -0.10
µm, -0.05 µm, 0.00 µm, 0.05 µm, 0.10 µm, 0.15 µm, and 0.20 µm. The simulation parameters
as noted above are desirably continuous values, but discrete values were provided
by taking the calculation time into considerations. For all these combinations of
parameter values, light intensity distributions were obtained by using a light intensity
simulator based on the scalar diffraction theory.
[0185] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity values, were obtained and made to be
transfer resist patterns.
[0186] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.3 µm was within
10 %.
[0187] From the result of the check, the range of mask pattern size, in which resist patterns
obtained with all the combinations of the defocus and exposure dose values meet a
predetermined tolerance condition, was obtained and made to be the exposure dose latitude.
[0188] Among the five combinations of the half-light shielding area transmissivity and mask
pattern size obtained by the above procedure, the maximum mask pattern size latitude
could be obtained with the combination of the amplitude transmissivity of the half-light
shielding area transmissivity of 45 % and the mask pattern size of 1.95 µm.
[0189] The mask obtained in this embodiment sufficiently met the defocus latitude of 2.0
µm and the exposure latitude of ± 5 %. Further, it had sufficiently great mask pattern
size latitude to permit exposing with sufficient latitude, and it was possible to
obtain a sharp resist pattern shape.
[0190] While this embodiment was applied to the half-tone system phase shift mask, this
is by no means limitative, and similar effects are obtainable with masks of conventional
systems and also with phase shift masks of other system. Further, while this embodiment
concerned with contact hole patterns, this is only exemplary, and similar effects
are obtainable with other patterns than the contact hole patterns.
[0191] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0192] Further, as for the exposure dose latitude, defocus latitude and mask pattern size
latitude, the values in this embodiment are by no means limitative, and it is possible
to adopt other conditions as well.
[0193] A second embodiment of the invention will now be described.
[0194] This embodiment is an example of application of the invention to optimizing the exposing
conditions of a half-tone system phase shift mask. In this embodiment, a judgment
was made as to which one of two exposing conditions, namely with σ of 0.3 and 0.5,
with exposure wavelength of 248 nm and NA of 0.45, is to be adopted.
[0195] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the exposure dose latitude was set to ± 5 %. Further, the amplitude
transmissivity of the half-light shielding area of the half-tone system phase shift
mask was set to 40 %, and the mask pattern size (on 5 times reticle) when transferring
a contact hole of 0.3 µm, was set to 1.85 µm for σ of 0.3 and 1.75 µm for σ of 0.5.
Under each of these two exposing conditions, the following was made.
[0196] In the first place, for each mask the exposure dose was set such that a transfer
contact hole of 0.3 µm could be obtained under the conditions of defocus of µm and
no mask pattern size deviation.
[0197] Then, as simulation parameters, the defocus was set to 0 µm, ± 0. 25 µm, ±0. 50 µm,
±0.75 µm, and ±1.00 µm, and the deviation of the exposure dose from the setting value
was set to -5 %, 0 % and 5 %. Further, the deviation of the mask pattern size on the
5 times reticle was set to -0.20 µm, -0.15µm, -0.10µm, -0.05µm, 0.00µm, 0.05µm, 0.10µm,
0.15 µm, and 0.20 µm. The simulation parameters as noted above are desirably continuous
values, but discrete values were provided by taking the calculation time into considerations.
For all these combinations of parameter values, light intensity distributions were
obtained by using a light intensity simulator based on the scalar diffraction theory.
[0198] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0199] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made ass to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.3 was within
10 %.
[0200] From the result of the check, the range of mask pattern size, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size values meet
a predetermined tolerance condition, was obtained and made to be the mask pattern
size latitude.
[0201] Of the two values of σ obtianed in the above procedure, the maximum mask pattern
size latitude could be obtained with σ = 0.3. Thus, σ = 0.3 was adopted.
[0202] The mask obtained in this embodiment sufficiently met the defocus latitude of 2.0
µm and the exposure dose latitude of ± 5 %. Further, it had sufficient mask pattern
size latitude to permit exposing with sufficient latitude, and it was possible to
obtain a sharp resist pattern shape.
[0203] With the exposing conditions and masks as determined in the above, it was possible
to sufficiently meet the defosus latitude of 2.0 µm and the exposure dose latitude
of ± 10 % and obtain a sharp resist pattern shape. In consequence, it was possible
to manufacture semiconductor devices with satisfactory yield.
[0204] While this embodiment was applied to the half-tone system phase shift mask, this
is by no meams limitative, and similar effects are obtainable with masks of conventional
systems and also with phase shift masks of other systems. Further, while this embodiment
concerned with contact hole patterns, this is only exemplary, and similar effects
are obrtianable with other patterns than the contact hole patterns.
[0205] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distibutions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0206] Further, as for the exposure dose latitude, the defocus latitude and the mask pattern
size latitude, the values in this embodiment are by no means limitative, and it is
possible to adopt other conditions as well.
[0207] A third embodiment of the invention will now be described.
[0208] This embodiment is an example of application of the invention to optimizing the exposing
conditions of a usual photomask having a transmitting area and a light shielding area.
In this embodiment, a judgment was made as to which one of four different exposing
conditions, namely those with σ of 0.3 and 0.5 with exposure wavelength of 248 nm
and NA of 0.45 and those with of 0.3 and 0.5 with exposure wavelength of 365 nm and
NA of 0.57, is to be adopted.
[0209] As lithographic process setting values, the defocus latitude was set to 2.0 µm, and
the exposure dose latitude was set to ± 5 %. Further, the mask pattern size (on 5
times reticle) when transferring a contact hole of 0.5 µm was set to 2.5 µm. Under
these four exposing conditions, the following was made.
[0210] In the first place, for each mask the exposure dose was set such that a transfer
contact hole of 0.5 µm could be obtained under the conditions of defocus of 0 µm and
no mask pattern size deviation.
[0211] Then, as simulation parameters the defocus was set to 0 µm, ±0.25 µm, ± 0.50 µm,
± 0. 75µm, and ±1.00 µm, and the deviation of the exposure dose from the setting value
was set to -5 %, 0 % and 5 %. The deviation of the mask pattern size on the 5 times
reticle was set to -0.20 µm, -0.15µm, -0.10µm, -0.05µm, 0.00µm, 0.05µm, 0.10µm, 0.15
µm, and 0.20 µm. These simulation parameters are desirably continuous values, but
discrete values were provided by taking the calculation time into considerations.
For all these combinations of parameter values, light intensity distributions were
obtained by using a light intensity simulator based on the scalar diffraction theory.
[0212] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0213] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.5 µm was within
10 %.
[0214] From the result of the check, the range of mask pattern size, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size values meet
a predetermined tolerance condition, was obtained and made to be the mask pattern
size latitude.
[0215] Among the four exposing conditions obtained in the above procedure, the maximum mask
pattern size latitude could be obtained in the case with exposure wavelength of 248
nm, NA of 0.45 and σ of 0.3. This condition was thus adopted.
[0216] The mask obtained in this embodiment sufficiently met the defocus latitude of 2.0
µm and the exposure dose latitude of ± 5 %. Further, it had sufficiently great mask
pattern size latitude, and it was possible to obtain a sharp resist pattern shape.
[0217] With the exposing conditions and masks as determined in the above, it was possible
to sufficiently meet the defocus latitude of 2.0 µm and the exposure dose latitude
of ± 10 % and obtain a sharp resist pattern shape. In consequence, it was possible
to manufacture semiconductor devices with satisfactory yield.
[0218] While this embodiment was applied to the usual photomask having a transmitting area
and a light shielding area, this is by no means limitative, and similar effects are
obtainable with phase shift masks and the like as well. Further, while this embodiment
concerned with contact hole patterns, this is only exemplary, and similar effects
are obtainable with other patterns than the contact hole patterns.
[0219] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulator or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer expeiments.
[0220] Further, as for the exposure dose latitude, the defocus latitude and the mask pattern
size latitude, the values in this embodiment are by no means limitative, and it is
possible to adopt other conditions as well.
[0221] A fourth embodiment of the invention will now be described.
[0222] This embodiment is an example of application of the invention to a photomask having
a transmitting area and a light shielding area under the exposing conditions of exposure
wavelength of 365 nm, NA of 0.57 and σ of 0.3.
[0223] As lithographic process setting values, the defocus latitude was set to 1.0 µm (±
0.5 µm), and the exposure dose latitude was set to ±5 %. Further, the mask pattern
size on 5 times reticle for transferring a contact hole of 0.5 µm was set to 2.40
µm, 2.45 µm, 2.50 µm, and 2.55 µm. With each of these four patterns, the following
was made.
[0224] In the first place, for each of the four masks the exposure dose was set such that
a transfer contact hole of ± 0.5 µm could be obtained under the conditions of defocus
of 0
µm and no mask pattern size deviation.
[0225] Then, the defocus was set to 0
µm,± 0.25
µm, and ± 0.50 µm, and the deviation of the exposure dose from the setting value was
set to -5 %, 0 % and 5 %. Further, the deviation of the mask pattern size on the 5
times reticle was set to -0.20 µm, -0.15 µm, -0.10 µm, -0.05 µm, 0.00 µm, 0.05 µm,
0.10 µm, 0.15 µm, and 0.20 µm. The simulation parameters as noted above are desirably
continuous values, but discrete values were provided by taking the calculation time
into considerations. For all these combinations of parameter values, light intensity
distributions were obtained by using a light intensity simulator based on the scalar
diffraction theory.
[0226] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines 6 as shown in
Fig. 4, corresponding to the light intensity threshold values, were obtained and made
to be transfer resist patterns.
[0227] As the diameter of the contour linee, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.5 µm was within
10 %.
[0228] From the result of the check, the range of mask pattern size, in which resist patterns
obtained with all the combinations of the defocus and mask pattern size values meet
a predetermined condition, was obtained and made to be the mask pattern size latitude.
[0229] Of the four patterns obtained in the above procedure, the maximum mask pattern size
latitude could be obtained with the mask pattern size of 2.40 µm. Thus, this condition
was adopted.
[0230] The mask obtained in this embodiment sufficiently met the depth of focus of 1.0 µm
and the exposure dose latitude of ± 5 %. Further, it had sufficient exposure pattern
latitude to permit exposing with sufficient latitude, and it was possible to obtain
a sharp resist pattern shape.
[0231] While in this embodiment the invention was applied to the photomask having a transmitting
area and a light shielding area, this is by no means limitative, and similar effects
are obtainable with phase shift masks and the like as well. Further, while this embodiment
concerned with contact hole patterns, this is only exemplary, and similar effects
are obtainable with other patterns than the contact hole patterns.
[0232] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0233] Further, as for the exposure dose latitude, the defocus latitude and the mask pattern
size latitude, the values in this embodiment are by no means limitative, and it is
possible to adopt other conditions as well.
[0234] A further exemple will now be described.
[0235] This is an example of optimizing the exposing condition of a photomask which has
a usual light shielding area for substantially perfectly shielding light and a transmitting
area. In this example, in flex exposing with exposure wavelength of 365 nm, NA of
0.57, σ of 0.6 and number of light emission times of 2, a judgment was made as to
which one of two different exposing conditions with focal point pitch of 1.5 µm and
2.0 µm is to be adopted.
[0236] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the mask pattern size latitude was set to ±0.05 µm (on 5 times reticle).
Further, the mask pattern size when transferring a contact hole of 0.5 µm was set
to 0.5 µm. Under each of these two exposing conditions, the following was made.
[0237] In the first place, under each of the two exposing conditions the exposure dose was
set such that a transfer contact hole of 0.5 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size condition.
[0238] Then, the defocus was set to 0 µm, ± 0.25 µm, ± 0.50 µm, ± 0.75 µm, and ± 1.00 µm,
and the deviation of the mask pattern size on the 5 times reticle was set to -0.05
µm, 0.00 µm, and 0.05 µm. Further, the deviation of the exposure dose from the setting
value was set to -20 %, -15 %, -10 %, -5 %, 0 %, 5 %, 10 %, 15 %, and 20 %. The simulation
parameters as noted above are desirably continuous values, but discrete values were
provided by taking the calculation time into considerations. For all these combinations
of parameter values, light intensity distributions were obtained by using a light
intensity simulator based on the scalar diffraction theory.
[0239] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines corresponding
to the light intensity threshold values were obtained and made to be transfer resist
patterns.
[0240] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.5 µm was within
10 %.
[0241] With these values of the defocus, mask pattern size and resist pattern size, the
range of exposure dose in the predetermined tolerance condition range, i.e., the exposure
dose latitude, could be obtained.
[0242] Of the two exposing conditions obtained in the above procedure, the maximum exposure
dose latitude could be obtained with the focal point pitch of 1.5 µm.
[0243] The method of exposing in this example sufficiently met the depth of focus of 2.0
µm and the mask pattern size latitude of ±0.05µm (on 5 times reticle). Further, because
of sufficiently great exposure dose latitude, it was possible to obtain exposing with
sufficient latitude, and it was possible to obtain a sharp resist pattern shape.
[0244] While this example was applied to the photomask having a usual light shielding area
and a transmitting area, this is by no means limitative, and similar effects are obtainable
with shift masks and the like as well. Further, while this example concerned with
contact hole patterns, this is only exemplary, and similar effects are obtainable
with other patterns than the contact hole patterns.
[0245] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simiulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0246] Further, as for the defocus latitude and the mask pattern size and resist pattern
size tolerance conditions, the values in this example are by no means limitative,
and it is possible to adopt other conditions as well.
[0247] A further example of the invention will now be described.
[0248] This is an example of an application optimizing the exposing condition with respect
to a photomask having a usual light shielding area substantially perfectly shielding
light and a transmitting area. In this example, in flex exposing with exposure wavelength
of 365 nm, NA of 0.57, σ of 0.6 and number of exposing times of 2, a judgment was
made as to which one of two different exposing conditions, i.e., focal point pitches
of 1.5 µm and 2.0 µm, is to be adopted.
[0249] As lithographic process setting values, the exposure dose latitude was set to ± 5
%, and the mask pattern size latitude was set to ±0.05 µm (on 5 times reticle). Further,
the mask pattern size when transferring a contact hole of 0.5 µm, was set to 0.5 µm.
Under each of these two exposing conditions, the following was made.
[0250] In the first place, under each of the two exposing conditions, the exposure dose
was set such that a transfer contact hole of 0.5 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size deviation.
[0251] Then, the deviation of the exposure dose from the setting value was set to -5 %,
0 % and 5 %. The deviation of the mask pattern size on the 5 times reticle was set
to -0.05 µm, 0.00 µm, and 0.05 µm. The defocus was set to 0 µm, ± 0.25 µm, ± 0.50
µm, ± 0.75 µm, ± 1.00µm, and ±1.25 µm. The simulation parameters as noted above are
desirably continuous values, but discrete values were provided by taking the calculation
time into considerations. For all these combinations of parameter values, light intensity
distributions were obtained by using a light intensity simulator based on the scalar
diffraction therory.
[0252] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines corresponding
to the light intensity threshold values were obtained and made to be transfer resist
patterns.
[0253] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predermined contact
hole size tolerance condition. The resist pattern size tolerance condition was set
such that the deviation from a design mask pattern size of 0.5 µm was within 10 %.
[0254] With these exposure dose, mask pattern size and resist pattern size values, the range
of defocus in the predetermined tolerance condition ranges, that is, the defocus latitude,
could be obtained.
[0255] Of the two exposing conditions obtained in the above procedure, the maximum defocus
latitude could be obtained with the focal point pitch of 2.0 µm. This condition was
thus adopted.
[0256] The method of exposing in this example sufficiently met the exposure dose latitude
of ± 5 % and the mask pattern size latitude of ± 0.05 µm (on the 5 times reticle).
Further, because of sufficiently great defocus latitude, it was possible to obtain
exposing with sufficient latitude and also obtain a sharp resist pattern shape.
[0257] While this example was applied to the photomask having a usual light shielding area
and a transmitting area, this is by no means limitative, and similar effects are obtainable
with phase shift masks and the like as well. Further, while this example concerned
with contact hole patterns, this is only exemplary, and similar effects are obtainable
with other patterns than the contact hole patterns.
[0258] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by experiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0259] As for the exposure dose latitude, and the mask pattern size and resist pattern size
tolerance conditions, the values in this example are by no means limitative, and it
is possible to adopt other conditions as well.
[0260] A fifth embodiment of the invention will now be described.
[0261] This embodiment is an example of an application of the invention to optimizing the
exposing condition with respect to a photomask having a usual light shielding area
substantially perfectly shielding light and a transmitting area. In this embodiment,
in flex exposing with exposure wavelength of 365 nm, NA of 0.57, σ of 0.6 and number
of exposing times of 2, a judgment was made as to which one of two exposing conditions,
i.e., focal point pitches of 1.5 µm and 2.0 µm, is to be adopted.
[0262] As lithographic process setting values, the defocus latitude was set to 2.0 µm (±
1.00 µm), and the mask pattern size latitude was set to ± 0.05 µm (on 5 times reticle).
Further, the mask pattern size when transferring a contact hole of 0.5 µm was set
to 0.5 µm. Under each of these two exposing conditions, the following was made.
[0263] In the first place, under each of these exposing conditions, the exposure dose was
set such that a transfer contact hole of 0.5 µm could be obtained under the conditions
of defocus of 0 µm and no mask pattern size deviation.
[0264] Then, the deviation of the exposure dose from the setting value was set to -5 %,
0, % and 5 %, and the defocus was set to 0 µm, ± 0.25 µm, ± 0.50 µm, ± 0.75 µm, and
± 1.00 µm. The deviation of the mask pattern size on the 5 times reticle was set to
-0.2µm, -0.15µm, -0. 10µm, -0.05 µm, 0.00µm, 0.05µm, 0.10µm, 0.15µm, and 0.20µm. These
simulation parameters as noted above are desirably continuous values, but discrete
values were provided by taking the calculation time into considerations. For all these
combinations of parameter values, light intensity distributions were obtained by using
a light intensity simulator based on the scalar diffraction theory.
[0265] With these light intensity distributions, light intensity threshold values giving
transfer pattern sizes were obtained from exposure dose in the exposing apparatus
and empirically obtained resist sensitivity. Further, contour lines corresponding
to the light intensity threshold values are obtained and made to be transfer resist
patterns.
[0266] As the diameter of the contour lines, the contact hole size in the resist patterns
was obtained. Then, a check was made as to whether the size meets a predetermined
contact hole size tolerance condition. The resist pattern size tolerance condition
was set such that the deviation from a design mask pattern size of 0.5 µm was within
10 %.
[0267] With these defocus, exposure dose and resist pattern size values, the range of mask
pattern size in predetermined tolerance condition ranges, i.e., the mask pattern size
latitude, could be obtained.
[0268] Of the two exposing conditions obtained in the above procedure, the maximum mask
pattern size latitude could be obtained with the focal point pitch of 1.5 µm. This
condition was thus adopted.
[0269] The method of exposing in this embodiment sufficiently met the defocus latitude of
1.5 µm and the exposure dose latitude of ± 5 %. Further, it permitted mask production
with sufficient mask pattern size latitude, and it also permitted a sharp resist pattern
shape to be obtained. Thus, it was possible to manufacture semiconductor devices with
satisfactory yield.
[0270] While this embodiment was applied to the photomask having a usual light shielding
area and a transmitting area, this is by no means limitative, and similar effects
are obtainable with phase shift masks and the like as well. Further, while this embodiment
concerned with contact hole patterns, this is only exemplary, and similar effects
are obtainable with other patterns than the contact hole patterns.
[0271] Further, while the transfer resist patterns were obtained by a process of obtaining
the contour lines of the light intensity distributions obtained by the light intensity
simulation based on the scalar diffraction theory, this is by no means limitative,
and it is possible to use light intensity simulator based on the vector diffraction
theory or other theories. Further, it is possible to obtain light intensity distributions
by expeiments. In obtaining transfer resist patterns from light intensity distributions,
it is possible to use development simulators or other calculation processes. Further,
it is possible to obtain transfer resist patterns directly by transfer experiments.
[0272] Further, as for the exposure dose latitude, defocus latitude and resist pattern size
tolerance condition, the values in this embodiment are by no means limitative, and
it is possible to adopt other conditions as well.
[0273] As has been described in the foregoing, according to the invention it is possible
to provide a method of producing a photomask which permits finding out correlations
of many parameters, for instance three or more parameters, to one another, permit
obtaining the optimum condition from these correlations, permit the aloofness from
the actual condition to be reduced, permit quantitative grasping of various kinds
of performance, permit influence of mask pattern size fluctuations, etc. into considerations,
and permit actual optimization.
[0274] For example, by utilizing the invention it is possible specifically to obtain quantitative
evaluation of parameters in the exposing process and those in the mask manufacture
in consideration of the actual process. Particularly, by selecting the exposure dose
latitude, defocus and mask pattern size as parameters and accurately evaluating the
correlation of these parameters to one another in various exposing conditions, it
is possible to manufacture semiconductor devices with satisfactory yield or build
processes of manufacture with high efficiency and at low cost.