WO2019001639A1

WO2019001639A1 - Lithographic exposure device

Info

Publication number: WO2019001639A1
Application number: PCT/DE2018/100590
Authority: WO
Inventors: Ralph Semmler; Dirk Döring; Christoph HUSEMANN; Dirk Jahn; Andreas Kaiser
Original assignee: Carl Zeiss Jena Gmbh
Priority date: 2017-06-29
Filing date: 2018-06-27
Publication date: 2019-01-03
Also published as: DE102017114504B4; DE102017114504A1

Abstract

The invention relates to a lithographic exposure device having an exposure unit (5) comprising an imaging matrix (2), a telecentric lens (1) and a workbench (4) having a rest surface (4.1) for the relative arrangement and the relative movement of a substrate having a photoresist layer (3) towards the lens (1) in a movement direction (R), wherein the imaging matrix (2) is imaged in an image plane (BB) which is tilted relative to a plane (E) defined by the rest surface (4.1), or an actual length of the depth of field (T) is at least 1.5 times greater than the calculated length (IR) resulting from the image-side aperture angle (CCB).

Description

Lithography exposure device

A lithographic exposure apparatus includes an exposure unit and a workbench, wherein the exposure unit can be moved relative to the workbench in a predetermined travel direction to stepwise illuminate a photoresist layer disposed on the workbench several times.

The yield in photolithography is proportional to the light energy introduced during the total exposure time per pixel into a photoresist layer of a substrate (e.g., wafer or portafilter). In maskless photolithography, individual pixels of an imager array are typically imaged via an exposure lens as an image pixel on the photoresist layer, respectively. In this case, the image generator matrix and the exposure objective, which together form an exposure unit, are moved in a travel direction equal to a longitudinal extent of the imaging matrix relative to the photoresist layer of the substrate which rests on a workbench, so that each image pixel can be exposed stepwise as often as the image pixel Imager matrix has pixels in the respective longitudinal extent. The more pixels the imager matrix has in the direction of travel, the more frequently an image pixel is exposed and the more light energy is introduced into an image pixel. Consequently, large quantities of light energy can be registered with the use of multiple image generator matrices. A parallel use, in which a simultaneous exposure of each image pixel by one pixel of the image generator matrices is constructively difficult. A use in succession, in which the exposure of the image pixels by the pixels of the individual imaging arrays successively takes place, is more time-consuming. Both variants make the exposure unit comparatively more complex and expensive.

In parallel or alternatively, the objective is designed so that the light output emitted per pixel of the image generator matrix is transmitted as completely as possible to one image pixel each. As emitted light power, of which one speaks correctly only with primary radiators, the light intensity reflected by a secondary radiator, such as a micromirror array (DMD), should also be understood here. As the exposure lens, a telecentric lens whose optical scheme is shown in FIG. 1 is usually used, which images an imaging matrix 2 positioned in the object-side focal plane BE of a first lens group L into the image-side focal plane BE ₂ ^{x of} a second lens group L ₂ a substrate with a photoresist layer is arranged. The aperture diaphragm AP stands in the pupil plane PE of the objective coinciding with the image-side focal plane BEi ^{x of} the first lens group and the object-side focal plane BE _{2 of} the second lens group L ₂ and thus determines the object-side and image-side opening angle oto, OCB via the size of its opening from which the object-side and the image-side numerical aperture of the objective 1 result.

In contrast to the photolithographic exposure with masks, in which the object structure to be imaged have an arbitrary size and thus can be kept very small, so that it is imaged onto the photoresist layer with a standardized magnification of 0.25 ×, the object structure is exposed to an imaging matrix the size of the pixels is limited downwards. The image structures to be imaged on the photoresist layer are generally not substantially larger than one pixel, so that the magnification, determined by the focal length ratio f ₂ / fi, is less than 1.5.

The light conductance, determined by the given pixel size and the given radiation angle ocp of the pixels of the image generator matrix, remains constant over the beam path through the objective 1, which is why an image-side aperture angle OCB equal to the object-side aperture angle oco results at a magnification of, for example, 1: 1. Accordingly, on the one hand, the image-side numerical aperture is larger than is necessary for a required resolution, and on the other hand, this results in only a small depth of field. Frequently, the surface flatness of a photoresist layer to be exposed is greater in tolerance than the depth of field range is large, so that at least some of the pixels P are not sharply imaged on the photoresist layer. That is, the light power transmitted by the lens, which lies within the depth of field range, is completely introduced only into the image pixels. The result is that high demands on the flatness of the photoresist layer must be made so that it is within the only small depth of field over the entire extent of the substrate, which can hardly be fulfilled practically. Alternatively, to enable a focus tracking, so that the photoresist layer is always in the depth of field exposure, is technically and control technology very expensive and only feasible if the distance change to the lens has a shallow rise, so that the change in distance over the longitudinal extent of the imaging matrix in the direction of travel only small is.

In order to increase the depth of field, it is common practice to adjust the opening of the aperture stop so that a smaller image-side numerical aperture results, which still leads to a sufficient resolution and a greater depth of field at a given wavelength of the light to be imaged.

However, this inevitably also reduces the object-side aperture, so that the light output emitted by one pixel is no longer completely coupled into the objective. In order to finally enter a same light energy into a picture pixel with a lower light output, then only the respective exposure duration and / or the number of exposures per picture pixel can be increased. The former leads to an increased process time, the latter leads to a higher equipment complexity.

The object of the invention is to provide a lithographic exposure apparatus with which the input of the light output emitted from one of the pixels of an imager matrix into an image pixel on a photoresist layer which is outside a depth of field range calculated from the image-side aperture is increased.

This object is achieved for a lithography exposure apparatus according to claim 1. Advantageous embodiments are specified in the subclaims 2 to 5. The invention will be explained in more detail with reference to embodiments and drawings.

Show:

Fig. 1 is an optical scheme of a telecentric lens, as it is identical or in modified form in an inventive

Lithography exposure device is used and is used to explain the task,

2 shows a first embodiment of a lithography exposure apparatus with an objective according to FIG. 1 and an imager matrix tilted for this purpose, FIG.

Fig. 3 shows a second embodiment with a support surface of a

Worktable tilted exposure unit, comprising a lens according to Fig.1 and

4 shows a third embodiment in which the lens, modified to a

Lens according to Fig. 1, as a detuned achromatic is executed.

A first embodiment of a lithography apparatus according to the invention, as shown in FIG. 2, comprises a workbench 4 and an exposure unit 5 with a telecentric objective 1 and an imager matrix 2, similar to the prior art.

The telecentric lens 1 includes a first lens group Li having a first focal length fi (a distinction between object-side and image-side focal length should be omitted simplicity), a first object-side focal plane BE and a first image-side focal plane ΒΕ-ι ', and a second lens group L. ₂ , with a second focal length f2, a second object-side focal plane BE ₂ and a second image-side focal plane BE ₂ '.

The first image-side focal plane BEi 'and the second object-side BE ₂ coincide and form a common pupil plane PE, in which an aperture diaphragm AP is arranged. The size of the aperture of the aperture Ap is selected so that there is an object-side opening angle oco, which is adapted to the radiation angle ap of the pixel P. From the object-side opening angle oco and the imaging scale f ₂ / fi of the objective 1, an image-side opening angle CCB is obtained, from which, given the wavelength of the light emitted by the pixels P, a calculated length IR for a depth of field T results. The image-side opening angle CCB is larger than is necessary for the resolution of the image pixels BP resulting from the imaging.

The worktable 4 has a support surface 4.1 which defines a plane E and serves for the relative positioning of a substrate and thus a photoresist layer 3 to be exposed to the objective 1 for the purpose of carrying out the exposure process. During the exposure, the work table 4 and the exposure unit 5 are relatively moved relative to each other so that each image pixel BP of a pattern to be exposed on the photoresist layer 3 is repeatedly exposed. The image pixels BP are each successively acted upon by the light output of the pixels P arranged next to one another in the direction of travel R.

The fact of the multiple exposure by pixels P arranged side by side in the direction of travel R enables the object according to the invention to be achieved in a first manner in which the imager matrix 2 is imaged in an image plane BB which is tilted to a plane E defined by the support surface 4.1. The image pixels BP are thereby, while they are quasi scanned by the pixels P of a row of pixels P of the imager matrix 2 arranged in the travel direction R, at least temporarily within the depth of field T lying around the image plane BB, even if they have a greater tolerance from the plane E deviate as the computational length IR of the depth of field T is large. In this way, it is possible to enter more light output in image pixels BP, which are otherwise outside the depth of field T, for individual exposures.

In order to image the image generator matrix 2 into an image plane BB, which, contrary to the prior art, is tilted to a plane E defined by the contact surface 4.1, as in FIG 2, the image generator matrix 2 is arranged on the optical axis A, intersecting the first object-side focal plane BE and enclosing a first tilt angle β with the first tilt angle β and the travel direction R lying in one plane (drawing plane). Consequently, the pixels P are imaged in an image plane BB tilted to the image-side focal plane BE ₂ '. The defined by the support surface 4.1 level E is aligned in this case parallel to the second image-side focal plane BE ₂ '.

Alternatively, in a second exemplary embodiment, as shown in FIG. 3, the imaging matrix 2 is arranged in the object-side focal plane BEi and the optical axis A of the objective 1 encloses an angle γ of less than 90 ° with the plane E defined by the support surface 4.1. Consequently, the pixels P are imaged in an image plane BB coinciding with the image-side focal plane BE ₂ -, which is tilted with respect to the plane E.

The two aforementioned embodiments are practically simple to carry out. They require no modification of previously calculated lenses, but can be used quasi with an adapted to the beam angle ocp of the pixel P of an imager matrix 2 object-side opening angle oco and thus adapted numerical aperture despite small depth of field T to fulfill the object of the invention.

The solution of the invention is carried out in a second manner in which the telecentric lens 1 is modified, so that the beam guidance of different beam portions within the lens 1 is changed either wavelength-dependent or angle-dependent in the geometric beam guidance so that the lens 1 an actual length Desτ of the depth of field T results as a superposition of the depths of the different beam portions, which is greater than the resulting from the image-side opening angle OCB computational length IR.

This second way of solving the problem will be explained with reference to the following embodiments. In contrast to poorly corrected lenses, targeted measures are taken here, which selectively extend the computational depth of field either in the axial direction or generate several axially offset foci and thus depth of field without worsening the deteriorating resolution (in the transverse direction) below the required resolution ,

A third embodiment, shown in Fig. 4, relates to an embodiment with the wavelength specific in the axial direction different focal positions are created.

Usually, the objective 1 is corrected as an achromat for two wavelengths λι, λ _{2 of} the transmitting light, that is, these two wavelengths λ-ι, λ ₂ in a same image-side focal plane, here the second image-side focal plane BE ₂ 'are mapped. In order to obtain an actual length Ιτ of the depth of field T which is greater than the length IR calculated from the image-side opening angle OCB of the depth of field Ti, T ₂ for one of the two wavelengths λι, λ _2, the objective 1 is calculated such that for the two wavelengths λ-ι, λ ₂ different second image-side focal planes ΒΕ ₂ 'λι, ΒΕ ₂ ' λ ₂ result, which are axially offset so far to each other that two partial depth of field Ti, T ₂ result in wavelength, either partially overlap axially or adjoin one another. The actual length Ιτ of the depth of field T of the lens 1, which is formed by both partial field depth ranges Ti, T ₂ , thereby increases up to a doubling. Strictly speaking, the use of broadband light means that the edge regions of the wavelength distribution may also be shifted more than the monochromatic depth of field, because at each intervening wavelength another wavelength can be found so that their monochromatic depth of field ranges Ti, T ₂ overlap. This results in a closed overlap area. The possible shift is limited by the increasing loss of resolution and contrast and by the limited spectral sensitivity of the photoresist. So useful polychromatic depth of field T can be increased up to a factor of 4 monochromatic depth of field Ti, T ₂ .

A fourth exemplary embodiment, not shown in the drawings, relates to an embodiment with which different focal positions are created in the axial direction by geometric-optical or wave-optical influencing of the beam bundle components. In addition to the introduction of a defined amount of spherical aberrations of max. three Lambda by deliberate detuning of the lens, the introduction of a special filter in the pupil plane is possible. The effect of introducing spherical aberration is based on the fact that different aperture angles of the bundle in the image space have different cutting widths and thus thus cause longitudinal errors.

Apart from continuous phase shifts, discrete phase steps are also conceivable. In addition to spherical aberrations as an example of continuous phase shifts, there are also possibilities, in particular by means of discrete steps, that ray bundle components no longer interfere with one another and allow them to form separate, larger depth-of-field areas corresponding to their respective smaller numerical apertures. In the image plane, these larger depths of field overlap to a common larger depth of field.

LIST OF REFERENCE NUMBERS

1 lens

2 imager matrix

3 photoresist layer

4 workbench

4.1 bearing surface

5 exposure unit

Li first lens group

BE object-side focal plane of the first lens group L

At the image-side focal plane of the first lens group Li

fi first focal length

L ₂ second lens group

BE ₂ object-side focal plane of the second lens group L ₂

BE ₂ 'image-side focal plane of the second lens group L ₂

f ₂ second focal length ocp radiation angle of a pixel P

oco Object-side opening angle

OCB image-sided opening angle

ß first tilt angle between the image generator matrix 2 and the object-side focal plane BEi γ second tilt angle between the optical axis A and E level

T depth of field

Ti, T ₂ , .... T _n partial depth of field

IR arithmetic length of the depth of field T

IT actual length of the depth of field T

A optical axis

BB image plane

P pixels (the imager matrix 2) BP image pixels

E level

AP aperture stop

PE pupil level

R travel direction

Claims

claims

1 . A lithographic exposure apparatus comprising an exposure unit (5) comprising an imager array (2) having a plurality of pixels (P) and a telecentric objective (1) having an optical axis (A), and a workbench (4) having a support surface (4.1) Relative arrangement and relative movement of a substrate with a photoresist layer (3) to the lens (1) in a direction of travel (R), wherein the telecentric lens (1) has a first lens group (Li) with a first image-side focal plane (ΒΕ-ι ') and a second lens group (L ₂ ) having a second object-side focal plane (BE ₂ ) and the first image-side focal plane (ΒΕ-ι ') and the second object-side focal plane (BE ₂ ) coincide in a pupil plane (PE), in which an aperture diaphragm (AP) is arranged, with an aperture, which results in an object-side opening angle (oto) of the lens (1), which is adapted to a radiation angle (ccp) of the pixels (P), and which results from d an image-side aperture angle (ote) resulting from an object-side aperture angle (oto) and a magnification of the objective (1) is greater than that required for a required resolution of image pixels (BP) exposed by the image of the pixels (P), and from image-sided opening angle (ote) results in a calculated length (IR) for a depth of field (T), characterized in that

in that the imager matrix (2) is imaged into an image plane (BB) which is tilted to a plane (E) defined by the support surface (4.1), or an actual length (Ιτ) of the depth of field (T) at least 1.5x is greater than the resulting from the image-side opening angle (ote) computational length (l _R ).

2. Lithographiebelichtungsvorrichtung according to claim 1, characterized in that the image generator matrix (2) on the optical axis (A) a first object-side focal plane (BEi) intersecting and with this a first tilt angle (ß) including, wherein the first tilt angle (ß ) and the direction of travel (R) lie in one plane, so that the pixels (P) are tilted into the image plane tilted to a second image-side focal plane (ΒΕ ₂ ') of the objective (1) (BB) and the plane defined by the support surface (4.1) level (E) is aligned parallel to the second image-side focal plane (ΒΕ ₂ ').

3. lithography exposure apparatus according to claim 1, characterized in that the image generator matrix (2) in the first object-side focal plane (BE-i) is arranged and the optical axis (A) of the lens (1) has a second tilt angle (γ) of less than 90 ° the plane (E) defined by the bearing surface (4.1) enclosing the second tilt angle (ß) and the direction of travel (R) in a plane, so that the pixels (P) in the plane with the second image side focal plane (ΒΕ ₂ ') coincident image plane (BB) which is tilted with respect to the plane (E).

4. lithography exposure device according to claim 1, characterized in that the lens (1) causes a different beam guidance for beam components of different wavelengths, so that the actual length (IT) of the depth of field (T) as a superposition of the depths of the beam bundles shares is greater than that from the image-side opening angle (ae) resulting computational length (IR).

5. lithography exposure apparatus according to claim 1, characterized in that the lens (1) causes a different beam guidance for beam portions of different aperture angle ranges, so that the actual length (IT) of the depth of field (T) as a superposition of the depths of the beam bundles shares is greater than that from the image-side opening angle (ae) resulting computational length (IR).