Patent Application
20070229790
This application claims priority to German Patent Application 10 2006 013 459.1, which was filed Mar. 23, 2006, and is incorporated herein by reference.
Arrangement for the transfer of structural elements of a photomask onto a substrate and method for the transfer of structural elements of a photomask onto a substrate.
The invention under consideration concerns an arrangement for the transfer of structural elements of a photomask onto a substrate. Furthermore, the invention concerns a method for the transfer of structural elements of a photomask onto a substrate.
With the progressive miniaturization of integrated circuits, components with increasingly small structure sizes on a substrate are needed. To this end, a predetermined pattern of a mask is transferred onto a substrate in a lithographic process. Nowadays, structures of a few tens of nm in width and length are transferred onto wafer surfaces. In competition with other semiconductor manufacturers, both the throughput and also the precision of the transfer are decisive in economic success. The throughput is ensured by a “step and scan” method. However, defects in the precision of the structure transfer, in particular, in the control of the length and width of the structures to be portrayed, reduce the yield of functional chips.
Two main sources are responsible for inaccuracies during the structure transfer. Both mask inaccuracies as well as nonuniformities over the image field, caused by the projection system, contribute to undesired variation in the structure dimensions on the wafer. Mask defects are, above all, with high Mask Error Enhancement Factor (MEEF) values (≧3.5), which are to be typically expected with small kl factors, of decisive influence on the structure size control. This applies, above all, to the critical chip structures, whose line widths are characterized by the “Critical Dimension” (CD). If we are not dealing with lines, but rather two-dimensional structures, such as contact holes, then both their width as well as their length, or their width and the aspect ratio, which is determined by the relationship of width to length, must be controlled.
In order to ensure an improvement in the structure size control, high demands are placed on the structure size precision on the mask with high MEEF values. In this way, the costs of mask production are driven up. Therefore, attempts are being made to find ways to improve the structure size control by other methods that do not involve a considerable rise in manufacturing costs for lithographic masks.
An approach to improved CD control provides for correcting the illumination dose during the scanning process. First, the CD variation over the image field is measured and a dose matrix is prepared, which contains an optimal dose for each point in the image field. The dose along the scanning direction can be modulated by varying the scanning rate or by varying the pulse dose. Furthermore, a modulation of the dose can be effected along the slit direction by introducing gray filters. In a mathematical sense, dose variation Δ dose in the form Δ dose=f1(X)×f2(Y) can be realized for the two-dimensional illumination field with the X and Y coordinate directions, where, for example, f1(X) describes the dose variation along the scanning direction, and f2(Y) the dose variation along the slit direction, realized by means of gray filters.
In general, the optimal dose, doseopt(X, Y), can be approximated only more or less poorly by a dose variation Δdose=f1(X)×f2(Y) in the product form. In practice, the dose variation along the scanning direction f1(X) caused by the high scanning rates, which can be up to 500 mm/s, is inaccurately adjustable. This makes it difficult to approximate the optimal dose distribution, if, as is common in practice, comparatively high CD variations along the scanning direction in the image field, which are based on mask defects, are to be corrected. With this method, the aspect ratios, such as hole width to hole length, cannot be controlled with two-dimensional structures, such as contact holes, even if a good approximation of the optimal dose is possible.
Thus, an adjustment of the local dose in the illumination field may adjust the width of a contact hole to its theoretical value, but it will also change the possibly previously corrected length of the contact hole. In general, therefore the control of both the length and the width of two-dimensional structures will be required. This, however, is impossible with an adjustment of the dose adapted locally in the illumination field.
This characteristic of not being able to control simultaneously both the length and the width of two-dimensional structures is common to the method with many other previously proposed possibilities for CD control.
Another method for CD control provides for adjusting the intensity distribution of the light striking the mask, in accordance with the previously measured line width distribution in the illumination field by local manipulation of the refractive index and the absorption coefficient of the glass carrier. By means of a laser beam, the local refractive index and the absorption variations in the glass carrier are thereby introduced. With illumination with actinic light, fractions of the light intensity are removed from the beam path of the projection system by absorption and light scattering. By variation of the spatial density of the introduced variations of the refractive index and absorption coefficient, the intensity effective on the mask plane can be subjected to fine-grain modulation. In particular, intensity or dose variations of the general form Δdose (X, Y), that is, not only as for the method described above in the product form Δdose=f1(X)×f2(Y), can be introduced. The CD correction accuracy is accordingly greater.
With the method, the entire system, mask-illumination system and projection objective, can be optimized. CD variations caused by the projection system can also be automatically corrected, which leads to a limited usability of the corrected masks. Thus, the mask adapted by this method cannot be used if it is used in another projection objective or when using another illumination in the same projection objective, if CD variations caused by the projection objective or the individually used illumination adjustment cannot be neglected. That leads to the masks having to be re-written specifically for the projection objective, wherein new costs arise. Like the above-described irradiated dose adjustment with an additional gray filter in the slit direction, it is also impossible to correct both the length and the width of two-dimensional structures simultaneously.
Another method consists in separating the mask and the correcting element physically from one another. A transparent optical element, which modulates the effective intensity on the plane of the mask structures, either by means of laser beams or by placement of light-absorbing structures, is thereby introduced before the mask. The transparency of the light-absorbing structures, adjusted to the previously measured CD variation on the wafer plane, thereby permits a homogenization of the structure sizes on the wafer plane. At the same time, by the physical separation of the mask and the correcting element, it becomes possible to use masks in individually different projection objectives. Only the correcting elements must then be replaced when using the same or a similar mask in individually different projection objectives or when using another illumination adjustment. The costs are reduced by the feasibility of using multiple masks.
Here, just as with the previously described methods, only the effective intensity or the dose can be modulated, whereby it is not possible to correct the homogeneity of the length and the width for two-dimensional structures.
Therefore, there is the demand to further improve arrangements and methods for the transfer of structural elements of a photomask onto a substrate.
An embodiment of an arrangement for the transfer of structural elements of a photomask onto a substrate comprises an illumination device which produces radiation, a photomask with a plurality of structural elements, whereby the radiation from the illumination device transfers the structural elements of the photomask onto a photoresist placed on a substrate. The arrangement moreover comprises an optical element, wherein the optical element produces a local variation of a degree of transmission of the radiation.
An embodiment of a method for the transfer of structural elements onto a substrate comprises the provision of a photomask with a plurality of structural elements, a substrate on which a photoresist is formed, an optical element, and an illumination device that produces radiation for the transfer of the structural elements of the photomask. The method, moreover, comprises the placement of the optical element between the photomask and the substrate or between the illumination device and the photomask, the transfer of the structural elements of the photomask onto the photoresist formed on the substrate, wherein the optical element produces a local variation of a degree of transmission of the radiation.
An embodiment of a method for the transfer of structural elements onto a substrate comprises the provision of a photomask with a plurality of structural elements placed thereon, a first substrate on which a photoresist is formed, and an illumination device that produces radiation for the transfer of the structural elements of the photomask. The method also comprises the transfer of the structural elements of the photomask onto the photoresist formed on the first substrate and the measurement of the image elements on the first substrate, obtained by the transfer of the structural elements of the photomask onto the photoresist formed on the first substrate. The method comprises, moreover, the determination of deviations of the obtained image elements on the first substrate in comparison with nominal structures, the production of an optical element that corrects the deviation of the obtained image elements on the first substrate in comparison with the nominal structures, a second substrate on which a photoresist is formed, the placement of the optical element between the photomask and the second substrate or between the illumination device and the photomask, and the transfer of the structural elements of the photomask onto the photoresist that is formed on the second substrate, wherein the optical element causes a local variation in the degree of transmission of the radiation.
Other advantageous embodiments of an arrangement for the transfer of structural elements of a photomask onto a substrate and the method for the transfer of structural elements onto a substrate are possible and are apparent to one skilled in the art from the following detailed description of the embodiment examples.
Embodiments of the arrangement for the transfer of structural elements of a photomask onto a substrate and embodiments of the method for the transfer of structural elements onto a substrate are explained in more detail, below with reference to the drawings. Shown are:
a, a top view of a two-dimensional mask structure of a photomask;
b, the diffraction pattern of the mask structure of the photomask shown in
c, the illumination pupil of an illumination device 4, in the form of a quadrupole illumination device;
d, the result of convolution of the frequency spectrum of the photomask, in accordance with
a, a top view of a resist contour in a photoresist, which one obtains with a structure transfer of a rectangular dark structure on the photomask without the use of stacked antireflection layers;
b, a top view of a resist contour in a photoresist which one obtains with a structure transfer of a rectangular dark structure on the photomask, using stacked antireflection layers;
a and 6b, an example of the transmission behavior of stacked antireflection layers;
a,
7
b, and 7c, an example of the transmission behavior of stacked antireflection layers; and
The first lens system 15 is placed between the illumination device 4 and the photomask 2. The optical element 6 is placed between the photomask 2 and the second lens system 20. The second lens system 20 is placed between the optical element 6 and the substrate 5.
The illumination device 4 comprises a light source, which produces ultraviolet (UV) or deep ultraviolet (DUV-Deep UltraViolet) light, or another type of radiation, which is suitable for a photolithographic process. The light source can comprise, for example, an ArF laser, which generates light with a wavelength of 193 nm. The illumination device 4 can be designed to produce an oblique illumination of the photomask 2. This can be produced, for example, by an arrangement of one or more light sources at a distance from the optical axis 50. The illumination device 4 for the production of an oblique illumination can, for example, comprise a dipole illumination device, a quadrupole illumination device, or an annular (ring-shaped) illumination device.
The photomask 2 comprises a mask pattern with structural elements 3, which are to be transferred onto the substrate 5. The photomask 2 typically comprises a thin quartz plate, on which dark structures 30, such as light-absorbing elements, for example, chromium, and light-attenuating elements, such as molybdenum-silicate, are applied.
The optical element 6 can be connected firmly with the photomask 2 in that it is placed, for example, on a pellicle frame of the photomask 2. Alternatively, the optical element 6, however, can also be affixed with the aid of an arrangement independent of the photomask 2, between the photomask 2 and the second lens system 20. The optical element 6 has a carrier 8, which can be made of an optically transparent material, such as quartz glass. At least one antireflection layer stack 9 is placed on a surface of the carrier 8, facing the photomask 2. The at least one antireflection layer stack 9, however, can also be placed on a surface of the carrier 8, turned away from the photomask 2.
The stacked antireflection layers 9 can comprise several layers. The stacked antireflection layers 9 can comprise a first layer 10, which is placed on the surface of the carrier 8, a second layer 11, placed on the first layer 10, and a third layer 12, placed on the second layer 11. The radiation falling on the stacked antireflection layers 9 is attenuated, as a function of the angle of incidence of the radiation, with respect to a surface 7 of the optical element 6.
The substrate 5 can comprise a wafer, which is coated with a photoresist (photosensitive coating) 21, so that after carrying out a photolithographic process, an image of the mask pattern on the photosensitive coating 21 is produced on the wafer.
When operating the arrangement 1, a radiation 1000, produced by the illumination device 4, passes the first lens system 15, the photomask 2, the optical element 6, and the second lens system 20, and projects an image of the mask pattern onto the photoresist 21, which is placed on the substrate 5. The photoresist 21 can then be developed or etched, so as to produce a resist contour of the photoresist 21. The resist contour of the photoresist 21 can be transferred to the substrate 5 by etching processes known in the art.
The mask pattern of the photomask 2 leads to a situation in which the radiation, incident on the photomask 2, is split up, behind the photomask 2, into diffraction orders. The diffraction orders are present in the far radiation field, behind the photomask 2, in an angular distribution, specific to the mask pattern and the illumination device 4.
The diffraction orders of the radiation, incident on the stacked antireflection layers 9 and diffracted at the photomask 2, are attenuated as a function of the shape of the stacked antireflection layers 9 and as a function of the angle of incidence of the radiation with respect to the surface 7 of the optical element 6.
The carrier 8 of the optical element 6 is made up of several sections 8a, 8b, 8c, 8d. Some of the stacked antireflection layers 9, with different layer thicknesses for the first layer 10, the second layer 11, and the third layer 12, are formed on individual sections 8a, 8b, 8c, 8d of the carrier. The individual sections 8a, 8b, 8c, 8d of the carrier 8, and the corresponding stacked antireflection layers 9 are associated with individual areas on the surface of the substrate 5, wherein the association is determined by the specific shape of the arrangement 1. Local pupil filters, effective for individual sections in the illumination field, are realized by the development of the carrier 8 with several sections 8a, 8b, 8c, 8d, which have some of the stacked antireflection layers 9; the filters permit the correction of both the length and the width of two-dimensional structures of the photoresist 21 in accordance with a previously measured nonhomogeneity of length and width distributions of the structures of the photoresist 21 on the substrate 5. The mode of operation of the pupil filter is explained in more detail in the description, with reference to
In order to ensure that the layer thickness variations of the individual stacked antireflection layers 9 can be adjusted locally on the carrier, a laser-aided, chemical vapor deposition method (CVD), for example, can be used for the formation of the individual layers. The local temperature distribution and thus the local deposition rate of the layer material is influenced by a locally variable intensity irradiation of the laser. Thus, it is possible to adjust the thickness of the layer material to be deposited locally, in a purposeful manner and accurate to a nanometer.
Another possibility is the placement of diaphragms (“stencils”) with variable openings before the carrier 8 to be coated or to move it under such stencils, while controlling the time. Thus, the material flow of the layer material to be deposited can be controlled locally on the carrier 8 and, in this way, an exact layer thickness control can be achieved. However, other methods can also be used to apply the layers on the carrier 8.
For the mode of operation of the optical element 6, described above, it is important that with the individual method, the required layer thickness variations can be adjusted to several nm to 10 nm. Lesser demands are thereby made of the spatial resolution of the local layer thickness variations. It is sufficient if the layer thickness control primarily attains a lateral resolution in the range of approximately 0.1 to 1 nm.
The standardized wave vector gives the direction of propagation of the diffraction order 1003 of the radiation 1000, directly before the optical element 6. The diffraction order 1003 of the radiation 1000 strikes the optical element 6 at an angle θ, with respect to the surface 7 of the optical element 6. If the x and y components of the normalized wave vector z,1 are designated as {right arrow over (k)}x=sin({circle around (x)}x) and {right arrow over (k)}y=sin({circle around (x)}y), then the result of the normalization is {right arrow over (k)}z=−√{square root over (1−(sin2({circle around (x)}x+sin2({circle around (x)}y)). A thickness modulation of the stacked antireflection layers 9 (not shown in
of the stacked antireflection layers 9 (not shown in
a shows a top view of the two-dimensional mask structure and a two-dimensional mask pattern of a photomask 2. The mask structure comprises two-dimensional, periodic dark structures 30 with structure periods placed vertically with respect to one another and two-dimensional, periodic structural elements 3, with structure periods placed vertically with respect to one another. Adjacent structural elements 3 have a first distance (pitch) px along a first direction X, and adjacent structural elements 3 have a second distance py along a second direction Y, wherein the first distance px is different from the second distance py. The first distance px=220 nm and the second distance py=180 nm in the example under consideration.
b shows the diffraction pattern 101 to 109 of the mask structure of the photomask 2, which are formed with light normally incident on the photomask plane in a representation in which the diffraction intensities are plotted versus the angles or the direction cosines sin θx and sin θy of the diffraction orders. This representation illustrates the frequency spectrum of the photomask 2, which is present in the exit pupil plane.
As a result of the lower pitch py along the Y direction, in comparison with the pitch px along the X direction, the diffraction orders 103 and 107 are at a shorter distance from the central diffraction order 101 than the diffraction orders 105 and 109.
The circle 110 symbolizes the maximum opening of an objective of the arrangement 1 (not shown in
c illustrates the frequency spectrum of an illumination device 4, which is designed as a quadrupole illumination device, in a pupil representation, wherein the frequency spectrum is plotted versus the angles and the direction cosines sin θx and sin θy. The areas 201 to 204 thereby represent the intensities of the illumination device 4 in the illumination pupil.
d shows the result of a convolution of the frequency spectrum of the photomask 2, in accordance with
The areas 303 and 304 of the pupil filling, lying further inside, are associated with the pitch px; the areas 301 and 302 of the pupil filling lying further outside are associated with the pitch py.
The circle 309 symbolizes the maximum opening of an objective of the arrangement 1. Areas lying outside this circle 309 do not contribute to the structure transfer.
In accordance with one embodiment, an optical element 6, with stacked antireflection layers 9, which produces an angle-dependent transmission modulation, is found behind the photomask 2. Depending on the angle of incidence of the diffraction orders, with respect to the surface 7 of the optical element 6, the intensity of the individual diffraction order is modulated, wherein the stacked antireflection layers 9 act as a rotationally symmetrical pupil filter. The individual transparency as a function of the angle of incidence can thereby be adjusted purposefully by the layer thicknesses of the individual layers 10, 11, 12 of the stacked antireflection layers 9. Since the layer thicknesses of the individual layers 10, 11, 12 of the stacked antireflection layers 9 can be varied locally, that is, as a function of the lateral position behind the photomask 2, a specifically adapted pupil filter is realized for each position in the image field.
The pupil filter has a multiplicative effect on the pupil filling. The example shown in
a shows a top view of a resist contour 500a in a photoresist placed on a substrate obtained with a transfer of structural elements placed on a photomask, without the use of stacked antireflection layers, in accordance with one embodiment. The resist contour 500a thereby represents the image of a rectangular dark structure 30 of the photomask and can be designated as an image element of the structure transfer located on the substrate. The rectangular dark structure is a dark structure of an arrangement of dark structures on the photomask, placed periodically along the X and Y directions, as shown, for example, in
The extent of the resist contour along the X direction is 100 nm and the extent of the resist contour along the Y direction is 64 nm.
The resist contour 500a, obtained by the structure transfer, is then compared with a nominal structure. The dimensions of the nominal structure in the X and the Y directions are the lengths of a resist contour that one would like to obtain with a structure transfer of the rectangular dark structure. For example, it may be desirable to extend the length of the resist contour in the Y direction. However, it may also be desirable to extend the length of the resist contour in the X direction.
In order to effect the desired change in the ratio of the length of the resist contour in the X direction to that in the Y direction, an optical element is produced, which enables the lengths of the resist contour to be corrected in accordance with the desired nominal structure. In the example under consideration, it is desired that the length of the resist contour be extended in the Y direction. To this end, the optical element is designed with at least one antireflection layer stack, wherein the layer thicknesses of the individual layers of the antireflection layers stacked are designed in such a way that the stacked layers act as a pupil filter with a transparency that decreases towards the outside.
Another substrate with a photoresist placed thereon is provided, and the optical element is placed between the photomask and the other substrate.
Then a transfer of the structural elements of the photomask onto the photoresist placed on the other substrate is performed. The result of this structure transfer is shown in
The aspect ratio, which is determined by the ratio of the width in the Y direction to the width in the X direction, is 0.64 for the case without a pupil filter, whereas it is increased to 0.74 when using the pupil filter.
The pupil filter therefore produces an extension of the structure width in the second direction Y, relative to the structure width in the first direction X. If the reverse is desired, i.e., an extension of the structure width in the first direction X relative to the structure width in the second direction Y, then a pupil filter is used in which the regions 301 and 302 of the pupil area, shown in
a shows the transmission behavior of an example of stacked antireflection layers, which attentuate the structure periods (large pupil coordinates of the corresponding diffraction orders). The transmission is shown as a function of radial pupil coordinate sin α in the exit pupil plane, which is typically four times larger, with an enlargement factor of M=4, than radial pupil coordinate sin β in the entry pupil plane of the second lens system 20; that is, in the plane directly behind the optical element. The first layer of the stacked antireflection layers is made of magnesium fluoride and has a layer thickness of 1877.6 nm. The second layer is made of tantalum pentoxide and has a layer thickness of 855.7 mm. The third layer is made of magnesium fluoride and has a layer thickness of 1660.7 nm.
b shows the transmission behavior of another example of stacked antireflection layers, which attenuates diffraction orders lying further inside in the pupil. The first layer of the antireflection layer stack is made of magnesium fluoride and has a layer thickness of 1346.8 nm. The second layer is made of tantalum pentoxide and has a layer thickness of 388.6 nm. The third layer is made of magnesium fluoride and has a layer thickness of 1711.5 nm.
By adjusting the layer thicknesses of the individual layers of the stacked antireflection layers, it is possible to realize almost any pupil filter. In this way, the aspect ratio for contact holes can be easily affected.
If a thin plate that can be covered with the layer system is not originally provided in the optical design, then care must be taken that the carrier of the optical element on which the stacked antireflection layers are to be applied be formed only very thinly. Otherwise, aberrations are induced which can no longer be simply corrected.
The design of the stacked antireflection layers, that is, the layer thicknesses of the individual layers and the layer sequence, can be designed in such a way that the spherical aberrations induced by the carrier of the optical element are corrected at the same time that the required angle-dependent transmission modulation is corrected. In order to effect both corrections, the transmission modulation and the compensation of the spherical phase errors, stacked antireflection layers which consist of more than three layers may also be required.
With reference to
a,
7
b, and 7c show the transmission behavior of an individual stacked antireflection layers 9, with respect to the angle of incidence. The indicated angle range of 0° to 13.5° corresponds to the maximum opening of an objective of the arrangement 1 with a numerical aperture NA of 0.93.
The stacked antireflection layers 9, which forms the basis of
The stacked antireflection layers 9, which forms the basis of
The stacked antireflection layers 9, which forms the basis of
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2006 013 459.1 | Mar 2006 | DE | national |
