JPH01120820A - Dummy wafer for aligner - Google Patents

Abstract

PURPOSE:To correct a system error satisfactorily and maintain highly accurate alignment by a method wherein reversible photosensitive material which has a sensitivity to light applied to the surface of a 1st object and facilitates writing and erasing is employed. CONSTITUTION:A magneto-optical recording material layer 44 made of reversible material such as DyFe, TbDyFe, TbFe or TbFeCo which facilitates writing and erasing is formed on the surface of a metal film 43 by sputtering. As the magneto-optical recording material layer 44 is normally easy to be oxidized and deteriorated when it is brought into contact with the air directly, a protective film 44a is provided on the magneto-optical recording material layer 44. As the material of the protective film, nitride such as SiN or oxide such as Al2O3 is recommended because pin holes are hardly produced in the film made of such materials. Then a pattern region composed of an exposed region and a nonexposed region is formed. As lights transmitted through and reflected by those regions have different polarization angles, by utilizing this phenomenon, those regions are observed by a polarization microscope and a pattern with clear contrast can be detected highly accurately.

Description

Detailed Description of the Invention (Field of Industrial Application) The present invention relates to a dummy wafer for an exposure device used in the manufacture of semiconductor devices such as ICs and LSIs, and particularly to a dummy wafer for a reticle, a mask (hereinafter referred to as a "reticle"), etc. The first object and the second object are exposed and transferred when a pattern such as an electronic circuit formed on the object surface is transferred directly or through an optical means such as a projection lens onto a second object surface such as a sea urchin surface. The present invention relates to a dummy wafer for use in an exposure apparatus for performing positioning with an object, that is, alignment with a high density.

(Prior Art) An exposure apparatus for manufacturing semiconductor devices such as ICs and LSIs is required to have two basic performances: resolution performance and overlay performance. The former is a semiconductor substrate (hereinafter referred to as a "wafer")
The ability to form fine patterns on a photoresist surface applied to a surface, and the latter is the ability to accurately form a pattern on a photomask compared to the pattern formed on a surface during the moe process. It is the ability to align and transfer.

Exposure apparatuses are broadly classified into, for example, contact, proximity, mirror 1:1 projection, stepper, X-ray aligner, etc., depending on their exposure method, and the optimal overlay method for each has been devised and implemented.

Generally, for semiconductor device manufacturing, an exposure method that has a good balance between resolution performance and superimposition performance is preferred, and for this reason, reduction projection type exposure apparatuses, so-called steppers, are currently frequently used.

The resolution performance required for future exposure equipment is 0.5
The exposure method that can achieve this performance is, for example, a stepper using an excimer laser as a light source,
Proximity type aligner that uses a line as the exposure source,
There are three EB direct writing methods. Of these, the former two methods are preferable from the viewpoint of productivity.

On the other hand, the nail alignment accuracy generally ranges from 173 to the minimum line width for printing.
A value of 115 is required, and achieving this accuracy is generally as difficult or more difficult than achieving resolution performance.

In general, the following error factors exist in relative positioning, that is, alignment, between the pattern on the reticle surface and the pattern on the eight surfaces.

(1-1) The pattern (or mark) on the eight faces of the sea urchin varies depending on the type of device and process, its cross-sectional shape, physical properties, etc.
A wide variety of changes in optical properties.

(1-2) In order to reliably perform alignment with a predetermined accuracy in response to a variety of processes, the alignment detection system (optical system, signal processing system) must have a degree of freedom.

(1-3) In order to give the alignment optical system a degree of freedom, it is necessary to configure it independently of the projection lens, and as a result, the alignment between the reticle and the wafer becomes indirect.

In general, it is important to minimize these error factors and maintain a good balance.

Next, specific examples of the above-mentioned error factors will be shown.

(2-1) By setting the alignment light to the same wavelength as the exposure wavelength, a TTL ON Ax js alignment system can be constructed. This is because the J2 shadow lens has good aberration correction for this wavelength, so for example, an alignment optical system can be placed above the reticle, and the projected image of the wafer pattern and the pattern on the reticle surface can be aligned. −
Both can be aligned while simultaneously observing within the field of view, and exposure can be performed at the position where alignment is completed. Therefore, there are no systematic error factors in this method.

However, in this method, there is no choice in the alignment wavelength, and in processes such as absorption resist, there is a disadvantage in processing when the signal light from the wafer is extremely attenuated.

(2-2) In an off-axis type stepper, the wafer alignment optical system can be freely designed without being subject to any restrictions of the projection lens, and this degree of freedom can enhance process adaptability. However, with this method, it is not possible to simultaneously observe the reticle and the sea urchin; the reticle is aligned to a predetermined standard using a reticle alignment microscope, and the wafer is aligned using a wafer alignment microscope (hereinafter referred to as the "wafer microscope"). Alignment is performed to the reference inside the microscope. For this reason, error factors exist between the reticle and the wafer. Furthermore, after wafer alignment, in order to overlap the wafer pattern with the projected image of the reticle, a predetermined distance (this is called the "reference length" or r B
a5e 1ine J) wafer must be moved. Therefore, this method results in an increased system error factor.

In an apparatus having an alignment method that includes such system errors, it is necessary to strive to maintain these error factors stably, and at the same time, periodically check and correct the amount.

For example, the reference length (hereinafter referred to as "baseline") is the distance between the optical axis of the projection lens and the optical axis of the alignment microscope.

) is usually in the order of tens of I11. In general, even if strict temperature control is carried out to suppress the thermal expansion of the material that connects these layers, changes over time occur in units of 0.1 μm to 0.01 μm. In addition to the aforementioned baseline, factors that cause such changes over time include the projection magnification of the lens, the alignment position of the reticle, and the arrangement coordinate system of the wafer stage.

Magnification and 5HOT as conventional system error correction methods
Resist is applied to a reference wafer with guaranteed alignment, aligned and exposed according to the normal procedure, taken out of the device, developed, and the amount of error is read using vernier-like marks, and the value calculated from there. The most common method was to input the value manually into the device as an offset value.

However, this method has the problem that the operating rate of the apparatus decreases, the operator's effort increases, and furthermore, a series of operations requires a lot of time, so that the aging process progresses further as time passes after exposure.

(Problem to be solved by the invention) In contrast, the latent image (
pattern created by exposed and unexposed areas)
A method has been proposed in which alignment is performed by forming a latent image, detecting the imaging state of the latent image with a microscope attached to the exposure apparatus itself, reading the reference length, and measuring the magnification of the projection lens.

However, in the past, because the resist was configured to be optimal for exposure, fine pattern formation, and subsequent processes, this method did not provide sufficient contrast when detecting latent images, resulting in high-precision detection. There were problems such as not being able to.

The main object of the present invention is to solve these problems by using a dummy wafer having a photosensitive material sensitive to the light irradiating the reticle.

That is, when the present invention superposes a reticle as a first object and a wafer as a second object, various error factors in the superposition, such as changes in magnification of the projection optical system over time, changes in the reference length over time, etc. The object of the present invention is to provide a dummy wafer for an exposure apparatus that can satisfactorily correct system errors such as changes over time in the arrangement coordinates of an X-Y stage, and can always perform highly accurate overlay.

(Means for solving the problem of inversion) A dummy wafer for an exposure apparatus according to the present invention has a first object and a second object facing each other without alignment, and the pattern on the first object surface is transferred to the second object surface. When performing exposure transfer to the second
By forming an image of an alignment pattern provided on the first object surface instead of the object and detecting the imaging state of the alignment pattern image, the pattern on the first object surface is transferred to the second object surface. The material is used to determine the system error during exposure transfer to the first object surface, and has a reversible photosensitive material that is sensitive to the light that irradiates the first object surface and is capable of writing and erasing. It is.

That is, the present invention first exposes and transfers a pattern on a reticle surface as a first object onto a dummy wafer surface as a third object, and then sets the imaged state of the transferred image at that time as a second object. It is characterized in that the system error of the entire apparatus is automatically or semi-automatically corrected within the apparatus itself based on the detection result at this time.

(Example) Fig. 1 is a schematic diagram of an embodiment of the exposure apparatus in which a dummy wafer for an exposure apparatus according to the present invention is used, and Fig. 2 shows a part of the wafer stage and wafer transport system shown in Fig. 1. FIG.

In this embodiment, a so-called off-axis alignment type exposure apparatus is taken as an example.

In FIG. 1, a light beam from an illumination system 4 illuminates a reticle 2 as a first object held by a platen 6. In FIG. Then, a pattern such as an electronic circuit formed on the surface of the reticle 2 is projected and transferred onto the surface of the wafer 3 as a second object using the projection lens 1.

The reticle 2 and the wafer 3 can be exchanged by means of transport means (not shown). A reticle reference mark 5 is arranged around the lower part of the reticle 2 with a slight gap from the reticle 2 for correctly arranging the reticle 2 with respect to the coordinate system of the apparatus. A reticle microscope 7 for aligning the reticle 2 with respect to the reticle reference mark 5 is provided at a position facing the reticle 2 .

Note that the reticle microscope 7 and the reticle reference mark 5 are provided, for example, at two locations symmetrically about the center of the reticle.

For reticle alignment, a reticle microscope 7 reads the relative positional error between the reticle set mark and the reticle reference mark 5 provided on the 2nd surface of the reticle, and a reticle stage 8 movable in the XYθ directions adjusts the relative positional error between the reticle 2 and the platen 6. This is done by driving in a direction where the value approaches zero. Then, when the relative position error at this time falls within a predetermined tolerance range, the process ends. Note that 7a is an image pickup tube, a CCD, or the like.

Wafer alignment microscope near projection lens 1! 0 is placed. The wafer 3 is held by vacuum suction on a wafer holder 11, and the wafer holder 11 is held by a θZ stage 12 that can move in the rotation direction and the vertical direction, and the θZ stage 12 can move in the XY force directions. It is configured as follows.

Note that the θ direction here indicates the rotation direction around the Z axis.

At the end of the XY stage 13, an optical mirror 14Y for detecting position coordinates in the Y direction and a laser interferometer (hereinafter referred to as "interferometer") 15Y for making a light beam incident on the optical mirror 14Y are arranged. has been done. Similarly, an optical mirror 14X (not shown) and an interferometer 15X (not shown) are arranged for detecting position coordinates in the X direction. The position of the XY stage 13 and the XY position coordinates of the wafer 3 are read using these two interferometers 15X and 15Y.

The above data processing such as reticle alignment, wafer alignment, and stage position information is performed routinely. Also, for sequential operations etc., control device 2
This is done at 0.

Creation of jobs, commands to the device, settings of parameters, etc. are performed manually on the console using the keyboard 22 on the display 21.

In FIG. 2, 30 is the optical axis of the #2 shadow lens l, and for convenience, it will be explained below by aligning it with the origin 0 of the XY stage 13.

The Y-axis and the X-axis are the mirror surfaces 31 of the optical mirrors 14X and L4Y.
It is represented by the directions of X and 31Y.

In this embodiment, the optical axis 32 (point P) of the wafer microscope 10 is placed on the Y axis (X=0.Y=-1) for convenience.

Optical axis 30 of projection lens 1 and optical axis 32 of wafer microscope m10
The distance gt is the so-called reference length (Base 1ine). The stroke of the XY stage 13 is set to the maximum diameter of the wafer in the X direction and to (D+1) or a value close to it in the Y direction, so that the wafer microscope 10 can observe the entire surface of the wafer 3 and The entire area is exposed.

Projection lens! The projectable area is the area indicated by a circle 35 on the surface of the wafer 3, but since the reticle 2 is generally rectangular, the effective area is a rectangular area 36. This area 3
Reference numeral 6 denotes an area where the pattern on the second surface of the reticle is projected and transferred onto the third surface of the wafer in one exposure.

The XY stage 13 is drawn so that the center of the wafer 3 mounted thereon coincides with the optical axis 30 of the projection lens 1, but the XY stage 13 is positioned at ±D/D/D in the X direction with respect to this position.
2. It is movable in the Y direction within a range of +D/2° - (D/2 +1).

In this embodiment, in order to collect the exposed wafer 2 and place the unexposed wafer on the wafer chuck, the XY stage 13 is moved to point Q, X = -D/2, as shown by a two-dot chain line. ,
It will be moved to the position Y = -(o/2 + 1)c7).

On the other hand, for a normal wafer continuous processing routine, wafers are transferred from a wafer cassette 23 containing unexposed wafers to a pre-alignment stage 25 by a conveying means such as a supply belt 24.
After transferring the wafer to the top and positioning it on the pre-alignment stage 25 based on the outer shape of the wafer, the supply hand 26
The wafer is placed on the wafer holding table 11 at the wafer transfer position Q.

On the other hand, the wafer that is on the XY stage 13 and has already been exposed is placed on a collection belt 28 by a collection hand 27 and stored in a wafer cassette 29 on the collection side.

A dummy wafer 40 of the present invention used for calibration during this time is stored on a standby stage 41. The dummy wafer 40 is configured to have a reversible material such as a magneto-optical recording material or a photochromic material on its substrate.

When a predetermined time input into the console by the operator or when a predetermined number of wafers has been processed, the normal wafer processing routine is temporarily stopped and the exposed wafer is held. After being removed from the stand 11, the dummy wafer 40 is transferred to the wafer holding stand 11 by the supply hunt 26.

In this case, the dummy wafer 40 is substantially concave with the manufacturing wafer.
Preferably, it is a circular thin plate having the same shape and dimensions.

Further, it is desirable that the parallelism and flatness of the front and back surfaces of the substrate be better than that of a wafer.

FIG. 3 shows a microscopic cross-sectional view of a dummy wafer 40 using the magneto-optical recording material of this example. Board 42
The thickness is approximately 1 mm, and the material is preferably a material such as quartz that has a small coefficient of thermal expansion and is easy to work with.

In addition, L E (Low Expansion)
It may be a material such as glass that has a coefficient of thermal expansion substantially concave to that of the Si wafer, a metal material, or the Si wafer itself. A metal film 43 such as AI or Cr is formed on the surface of the substrate 42 by vapor deposition or sputtering as necessary to provide an appropriate reflectance.

As a reversible material that can be written and erased on the surface of the metal 11Q43, for example, DyFe, TbD, etc.
A magneto-optical recording material layer 44 of yFe, TbFe, TbFeCo, etc. is formed by sputtering.

The film thickness of the magneto-optical recording material layer 44 at this time is preferably as thin as possible in order to reduce the energy required for reversing the magnetization direction, but it is important to note that a certain degree of reflectivity is required during observation and that the film thickness The number of participants is preferably about 300 to 800 from the viewpoint of uniformity.

Further, in this embodiment, the magneto-optical recording material 44 is magnetized only in the direction perpendicular to the substrate 42.

The magneto-optical recording material in this example is such that when a certain level of light energy is irradiated in a zero magnetic atmosphere, the polarity is reversed only in the exposed area, and when a reverse magnetic field of above the above level is applied, the polarity is reversed in the same direction. Any material may be used as long as it has a reversible property of forming an aligned state with uniform polarity.

For example, Gd and Tb are used as magneto-optical recording materials.

An alloy of a heavy rare earth metal such as Dy and a transition metal such as Fe or Co is preferable, but in order to lower the exposure energy, it is sufficient if the Curie temperature is 2QQ'C or less, but DyFeQTbDyFe etc., which have a low Curie temperature of about 70°C, are preferable. Particularly preferred.

Since the magneto-optical recording material 44 is normally easily oxidized and deteriorates when directly exposed to air, a protective film 44a is provided on the magneto-optical recording material 44. The material for the protective film is a nitride such as SiN or Al2O.

Oxides such as oxides are preferred because pinholes are less likely to occur in the film.

These materials generally have poor transmittance for ultraviolet to deep ultraviolet light used as exposure light, so the absorption in this layer is reduced and the heat capacity of this layer is reduced to dissipate the heat absorbed here. In order to facilitate conduction to the magneto-optical recording material layer 44, the film thickness is preferably as thin as possible, and is preferably about 300 to 800, for example.

□ At this time, a pattern area consisting of an exposed area and an unexposed area is formed, and since the light transmitted or reflected through these areas has different polarization angles, this can be used to examine these areas with a polarizing microscope. observation, and detect patterns with clear contrast with high precision.

At the time when the dummy wafer 40 is sent to the XY stage 13 side for calibration, the magneto-optical recording material layer 44 has its magnetic lines of force uniformly aligned, for example, with N poles at all positions. They are aligned toward the surface. (This state is schematically indicated by an arrow in FIG. 3.) Calibration is performed on the magneto-optical recording material layer 44.
The pattern on two surfaces of the reticle is projected and exposed. Then, only the exposed portion of the magneto-optical recording material layer 44 undergoes a reversal phenomenon in the direction of magnetic force, resulting in the formation of a pattern on the surface of the reticle 2. In order to cause this reaction, it is necessary to surround the magneto-optical recording material layer 44 with a zero magnetic field having a predetermined intensity and direction.

There are the following methods for creating such a magnetic field near the image plane of the projection lens 1 during exposure.

(3-1) A method of using the substrate 42 of the dummy wafer 40 itself as a magnet.

(3-2) A method in which a magnet is placed at the tip of the projection lens l or at the annular zone around the exposure light beam.

(3-3) Method of placing a magnet in the wafer holding table 11.

FIG. 4 shows a cross-sectional view of the structure of the wafer holding table 11 of an example h'1 obtained by the method (3-3).

The wafer holding table 11 consists of an upper plate 45 and a lower plate 46, which are connected at the periphery with screws 47. A chamber (space) 48 is formed between the upper plate 45 and the lower plate 46,
A magnet plate 49 is glued therein to the underside of the top plate. The lines of magnetic force generated from this form a magnetic field for the dummy wafer 40 placed on the upper surface of the upper plate 45.

Therefore, in this case, the upper plate 45 is preferably made of a non-magnetic material such as ceramic. The wafer 3 or the dummy wafer 40 is supported on the upper surface of the upper plate 45, and a well-known vacuum suction is used for flattening at this time.

A dummy wafer 40 is supported on a wafer holding table 11 having a magnetic field, and is moved through the projection lens 1 by moving the XY stage 13.
The magneto-optical recording material layer 4 of the dummy wafer 40 is exposed by projecting the pattern on the two surfaces of the reticle.
4, a pattern (hereinafter referred to as a "magnetization image") of regions in which the directions of magnetic force are opposite to each other is formed.

FIGS. 5 and 6 are schematic diagrams of an embodiment in which a dummy wafer according to the present invention is formed using a photochromic material.

In FIG. 5, reference numeral 42 denotes a substrate with a thickness of about 1 mm, and a photochromic layer 78 is formed on the substrate 42. As shown in FIG.

This layer is made by dissolving in a base material such as PMMA, which is coated with, for example, sutokudoran-based, fulgide-based, dihydrodrene-based, thioindigo-based, aziridine-based, bipyridine-based, polycyclic aromatic-based, tetrabenzopentacene-based, etc. It is applied to a thickness of about 1 μm using a method such as a steel coating.

When these photochromic materials are irradiated with light, the structural formula of the substance changes and the transmittance changes. Furthermore, it has reversibility, allowing it to return to its original state by irradiating it with light of a different wavelength or by heating it. For example, spiropyrans change from colorless to purple when irradiated with ultraviolet rays, and return to their original colorless color when irradiated with light around 570 r+m or heated.

In the present example, it is preferable to provide an antireflection film and/or an absorption layer for the first wavelength between the photochromic layer 78 and the substrate 42.

Alternatively, a metal mirror may be provided between the substrate 42 and an antireflection film or an absorption layer to provide an appropriate reflectance.

In FIG. 6, the substrate 42 itself is a dummy wafer, and is made of, for example, so-called photochromic glass into which silver halide or the like is dissolved. Quartz or the like is preferable as the photochromic glass substrate from the viewpoint of transmittance of ultraviolet light normally used as exposure light and the above-mentioned thermal expansion. This photochromic glass becomes colored when exposed to ultraviolet light, and returns to its original colorless state after a certain period of time.

In addition, the photochromic material in each of the above examples is a substance that strongly reacts to light of the first wavelength and causes an optical change, but does not react at all to light of the second wavelength. Any type of configuration may be used as long as it is composed of more than 100%.

Next, an example will be described in which a magnetized image formed on the dummy wafer 40 is observed with the wafer microscope 10 and its position is detected, taking as an example a case where a magneto-optical recording material is used as the reversible material.

The wafer microscope 10 is originally used for observing patterns on the wafer 3 surface and detecting the position of the alignment pattern on the wafer 3. On the other hand, since magnetization image detection is an additional function, the wafer microscope 10 must have both of these detection capabilities, and the detected values of both must match with appropriate accuracy. is necessary.

In other words, when the target mark for on-plane alignment and the target mark for the magnetization image are located at exactly the same position, the difference between the detection values for each is a sufficiently small value that is consistent with the overlay accuracy of the entire device. Is required.

FIG. 7 shows a structural diagram mainly consisting of the optical system of the wafer microscope 10. The wafer microscope 10 will be explained from the light source side according to the light path. Although the main optical system is configured in the housing 50, the light source section 51 is arranged at a different position and guided to the main body side light introduction section by an optical fiber 52.

The reason why the lamp house (light source section) 51 is separated from the main body is to prevent heat generated in the lamp house and thermal expansion of the structure material from being transmitted to the main body.

The light emitted from the optical fiber 52 passes through an illumination lens 53, a first polarization filter 54, and an illumination filter 55, and enters a polarization splitting prism 57 along the optical axis a.

The principal ray passes through the half surface 58, is polarized at right angles by the mirror surface 59, heads downward along the optical axis, is focused by the objective lens 61, and illuminates the surface of the wafer 3 or dummy wafer 40.

The light reflected on the surface of the sea urchin 8 goes upward along the optical axis and is reflected on the mirror surface 59, further reflected on the half (polarized, optical mirror) surface 58, and further reflected on the optical axis C on the mirror surface 59. The light then advances along the optical axis d, passes through a relay lens 61 and an erector lens 62, and forms an image on a reference mark 65.

Further, the light transmitted through the reference mark 65 is passed through a second polarizing plate 67,
After passing through the imaging lens 68 and the mirror 69, the image is again focused on the light-receiving surface of the imaging tube or image sensor 70. Therefore, the image sensor 70 can simultaneously observe an image of the surface of the wafer 3 and an image of the pattern on the reference mark 65, which is generated by blocking the reflected light from the wafer 3 with the pattern on the reference mark 65.

The electrical signal charged and converted by the image pickup tube 70 is processed by the image processing circuit 72 via the CCU 71, and the relative position information of the pattern on the wafer 3 and the reference pattern on the reference mark 65 is outputted and sent to the control device 20.

Here, the second polarizing filter (although it may be the first) 67 can be rotated to a predetermined angle by a pulse motor 75 and a belt 76.

When observing a normal wafer, by making the first polarizing plate 54 and the second polarizing plate 67 optically in the same phase direction, the light receiving side can receive the maximum amount of light. It can also be used as a light amount adjustment means according to rate changes.

When observing the magnetization image of the dummy wafer 40, the difference between the forward and reverse directions of magnetic force is only the polarization angle of the reflected light, so by rotating the second polarizing plate 67 to an angle that matches the phase angle, the magnetization image can be observed. It can be observed in an image pickup tube as a contrast pattern of light and dark.

Note that at the position of the second polarizing plate, the light carrying the information of the wafer 3 and the light carrying the information of the reference mark 65 pass through in common, so even if the polarizing plate is tilted due to rotation, ,
For this purpose, the amount of axis deviation is the same, and no relative positional deviation occurs between the images of both the wafer pattern and the reference mark pattern.
Therefore, no alignment error occurs.

As described above, the positions of patterns on wafers and dummy wafers are detected using a wafer microscope.

In this way, this example uses a reversible material that reacts with exposure light, can optically detect patterns formed by exposure (exposed and non-exposed areas), and can be written and erased (in this example, By using a dummy wafer with a magneto-optical recording material (although a photochromic material or the like may also be used), system errors that change over time can be corrected, and the overlay accuracy between the reticle and the wafer can be improved. I'm trying.

Furthermore, by using reversible materials, these correction operations can be made routine within the device and automated.

Next, the system error correction procedure in this embodiment will be explained using the flowchart of FIG. 8, taking as an example the case where a magneto-optical recording material is used.

First, dummy wafers 40 coated with a magneto-optical recording medium and aligned with polarities aligned in the same direction are supplied to the band 2.
6, the wafer is transferred from the standby stage 41 to the wafer holding table 11. Then, the movement of the XY stage 13 and exposure are repeated in accordance with the preset pattern layout. At this time, the position coordinates of the XY stage are read and stored simultaneously with the exposure. To simplify the explanation below, during exposure,
It is assumed that the XY stage 13 is accurately positioned at the designed grid position at the time of detection.

(In other words, the position error of the xX stage 13 is zero) In reality, the error is allowed within the guaranteed linearity range during detection, and in that case, the following ΔX and ΔY are calculated by adding and subtracting the detection error amount and the stage position error. It is clear that

The reticle 2 serving as the original image is a reticle for a real device, and position detection marks 87L and 87R are arranged at positions corresponding to the scribe lines on both sides of the real device pattern 86, as shown in FIG. 9, for example. There is. These position detection marks 87L and 87R are projected onto the dummy wafer 40 by exposure to form a magnetized image.

FIG. 10 is an explanatory diagram of the pattern on the surface of the dummy wafer 40 at this time. In FIG. 10, for example, in the E-shot, a mark 94 of the E-shot marks 92 for position detection indicates information YR in the Y-axis direction, and a mark 95 indicates information XR in the X-axis direction.

Similarly, the E-/-1 mark 91 also shows information XL and YL in the X-axis direction and Y-axis direction. The same applies to other shots.

When step ant exposure is repeated until the final shot is reached, the XY stage 13 is moved to a position where each position detection mark can be observed by the wafer microscope 10 in order from the A shot on the dummy wafer 40, and the wafer microscope 10
The amount of relative positional deviation with respect to the reference mark within 0 is detected and stored.

Next, a means for calculating the amount of change in magnification of the projection lens 1, the amount of change in the baseline, etc. from the amount of positional deviation detected here will be described.

Let the stage coordinates at the time of exposure of an arbitrary shot (here, the E shot is taken up) be (xE, YE). In order to observe the E-shot mark 91, use the xY stage 1.
3 to the coordinates (X, +xβ/2°YE+l+yβ/2).

Here, β is the projection magnification, and x and y are the mark intervals on the two surfaces of the reticle shown in FIG. 9, respectively. The amount of deviation from the reference mark at that time is ΔXL.

Let it be ΔYL. Similarly, in order to observe the E shot mark 92, move the XY stage 13 to the coordinates (X, -xβ/2.Y
, +I-yβ/2), and the amount of deviation at that time is set to ΔxR9ΔYR.

If Δβ is a projection magnification error, then Δβ=(ΔXL−ΔXR)/(X/β). Further, if the inclination angle of the reticle projected image with respect to the stage is ΔRθ, then ΔRθ=(ΔYL−ΔYR)/(X/β). If the X component of the amount of change in the baseline is Δx3cop,..., and the Y component is ΔY5cope, then X, Cope = (ΔX
L +ΔXR) /2Y 5cope = (ΔYL+Δ
YR)/2. However, this amount includes not only a simple change in the position of the lens and the wafer microscope 10, but also errors in the alignment of the reticle 2 with respect to the projection lens 1, the habit of the stage, etc.
It is calculated as an amount that includes all system errors in matching the pattern actually aligned with the wafer microscope 10 and the reticle 2.

In the above example, only one shot was simply described, but in order to reduce the detection error, abnormal data is deleted and averaged from the data of all detected shots, and the projection magnification error Δβ and the reticle projection image are Stage tilt angle ΔR
Calculate θ, compare each amount with the tolerance, and if it exceeds, reset the projection magnification or reposition the reticle, and offset the ff1xY stage so that it does not overlap with the already formed mark again. It is better to move and expose all shots.

Then, the projection magnification of the projection lens, the reticle projected image, and the stage feed are checked again by the above-mentioned means, and if they are within the allowable range, the amount of deviation of the mark position and the XY stage 13 are checked.
A first correction grid is set by performing least squares approximation, abnormal data scattering, etc. on the relationship with the coordinates of .

As described above, the dummy wafer 40 is unloaded from the XY stage 13, and the polarity is returned to the aligned state depending on the position of the standby stage 41.

At the same time as the dummy wafer 40 is carried out, the wafers to be sent to the actual process are processed. The polarizing microscope is switched to a normal wafer microscope 10 so that the alignment marks prepared in the previous process can be detected. Then, the wafer carried out from the carrier is pre-aligned with reference to the orientation flat, and the supply hunt 24 moves the wafer to the wafer holding table 1.
Transfer to 1. Then, the shot alignment mark positions of a specified number of shots are detected, the wafer is rotated, and the specified number of shots are measured to detect the amount of positional deviation. Using statistical processing means, the amount of positional deviation of the wafer shot is A correction grid is created as a second correction grid.

Then, correction is performed using the first correction grating obtained by the previous magnetization image detection to create a third correction grating.

Then, move the wafer to the projection lens 1 position,
Step-and-exposure is repeated while performing correction using the third correction grating in order from the first shot. When all shots are completed, the wafer is collected by the collection hunt 27, and the next wafer is continued using the above-mentioned means.

On the other hand, the dummy wafer 40 returned to the standby stage 41 has its magnetized image erased until its next turn.

Next, a method of erasing the magnetized image on the surface of the dummy wafer 40 will be explained. Although various methods can be applied to erase the magnetized image, for example, the following five methods can be applied satisfactorily.

(4-1) Applying a strong magnetic field in the opposite direction to that when writing the magnetized image.

(4-2) Raise the temperature by m degrees to the Curie temperature, peel it in the opposite direction, and apply a magnetic field.

(4-3) A weak magnetic field in the opposite direction is applied, and a part of the exposure light is divided and re-irradiated outside the optical glaze of the projection lens.

(4-4) Reverse the direction of the magnetic field and re-irradiate the exposure light through the projection lens.

(4-5) A strong current is passed through the electromagnet provided on the chuck or the lens barrel in the opposite direction to that during exposure.

FIG. 11 is a schematic diagram of an embodiment showing the method (4-1) described above. On the right side of the figure, a strong magnetic force (for example, 1 to 5 KDe
, Oersted) is held by a magnet holder 82. When the dummy wafer 40 carrying the magnetized image passes under the magnet 81 while being collected by the collecting belt 28, the magnetized image is erased.

FIG. 12 is a schematic diagram of an embodiment showing the method (4-2) described above. In the same figure, the dummy wafers 40 collected by the collection hunt 27 are temporarily stopped on a heat insulating table 104 in a heat insulating chamber 89. The dummy wafer 40 is heated by the heater 88 in the insulating chamber 89 and raised to the Curie temperature (approximately 60 degrees). When the Curie temperature is reached, the magnetized image of the dummy wafer 40 is erased by the peeling magnets generated from the magnet plates 49 disposed above and below the holding chamber 89.

FIG. 13 is a schematic diagram of eight examples of the method (4-3) described above. In the figure, a half mirror 84 is provided in the optical path of the illumination system 4 to guide a part of the light beam from, for example, an excimer laser to a different optical path from the shadow lens 1 . Then, the dummy wafer 4 collected by the collection belt 28
0 temporarily stops on the magnet holder 82 where a weak magnetic field is formed by the perforated disc magnet 83 and the magnet plate 49. Therefore, at the same time as the wafer 3 is exposed, the light beam taken out by the half mirror 84 passes through the optical path 7 as shown in the figure and irradiates the entire surface of the dummy wafer 40. As a result, the magnetized image on the dummy wafer 40 is erased.

FIG. 14 is a schematic diagram of an embodiment showing the method (4-4) described above. A magnetic field in the opposite direction is formed by passing a current in the coil 79 in the opposite direction to that used when forming the magnetized image. Here, the location on the dummy wafer 40 where the magnetized image has been formed is reirradiated with a light beam from, for example, an excimer laser via the projection lens 1. Then, the magnetized image on the dummy wafer 40 is erased.

FIG. 15 is a schematic diagram of an embodiment showing the method (4-5) described above. By flowing more electric current v&It in the direction opposite to that during formation of the magnetized image of the coil 79, a strong magnetic field in the opposite direction to that during exposure is formed. The magnetized image is erased by moving the dummy wafer 40 again under the projection lens 1 in this strong magnetic field.

FIG. 16 is a schematic diagram of an embodiment in which the wafer holding table shown in FIG. 4 is provided with a magnet used as a magnetic force line generating means on the projection lens 1 side. Although a permanent magnet may be fixed to the projection lens barrel 97 with screws 47 or the like in the same manner as the magnetic force line chuck, an example using an electromagnet is shown here.

An electromagnet is constructed by winding a coil 79 around a coil core 98 made of iron or the like, and the coil core 98 is fixed to the lower part of the projection lens barrel 97 with screws 47. Therefore, the coil 79
A magnetic field is formed in the vicinity of the dummy wafer 40 by applying an electric current v&80 to the dummy wafer 40 .

By using an electromagnet instead of a permanent magnet as the magnet and changing the direction and magnitude of the current 80 flowing through the electromagnet to change the direction and magnitude of the magnetic field, it is possible to both write and erase a magnetized image.

Next, an example of a method for erasing a latent image formed on the surface of a dummy wafer when a photochromic material is used as the dummy wafer in this embodiment will be described.

FIG. 17 is a schematic diagram of an embodiment in which Subirondolan-based photochromic material is used. In this embodiment, the latent image is erased using light with a wavelength of around λ=570 nm.

In the figure, the dummy wafer 40 collected by the collection belt 28 is stopped at a predetermined position. The shutter 100 is normally closed, but opens only when erasing the latent image to emit a bright line around the ultra-high pressure mercury lamp (wavelength λ = 578nIIl).

I'm leaving it to you. ) from the direct and elliptical mirror 10
2, the entire surface of the tummy wafer 40 is irradiated via the condenser lens 110. At this time, shutter 100
The time it is open (the time required to erase the latent image) is approximately 5~
It takes about 10 seconds.

Note that the shutter 100 is not necessary if the exposed wafer 3 is completely subjected to molecular illumination by the ultra-high pressure mercury lamp 101 (for example, when the collection path is independent or there is a dedicated space for erasing the latent image).

Further, in this embodiment, for example, a copper vapor laser (wavelength λ=578 nm) or the like may be used instead of the ultra-high pressure mercury lamp 101.

FIG. 18 is a schematic diagram of an embodiment in which the latent image is erased by heating the dummy wafer 40. In the figure, a storage room 89 is provided in the middle of the recovery route. The dummy wafer 40 once enters the insulating chamber 89 through the door 103 and stops on the insulating table 104.

The latent image is then erased by heating with a heater 88. At this time, when a spiropyran type material is used as the photochromic material, the heating temperature is about 6000°C.

FIGS. 19 and 20 are schematic diagrams of an embodiment of the present invention, each showing the process from supplying to collecting the dummy wafer 40.

In FIG. 19, dummy wafers 40 are stored in the supply side cassette 23 along with a predetermined number of wafers. However, the positional information of the dummy wafer 40 in the supply side wafer cassette 23 is input into the console by the operator in advance.

After the exposure processing of the wafer, when the next exposure processing target becomes the dummy wafer 40, the main body of the exposure apparatus waits for calibration according to a command from the control device 20. The dummy wafer 40 follows the same path as the wafer (pre-alignment stage 25→supply hand 26) and is placed on the wafer holding table 11. After completing a series of calibrations, the wafer follows the same path as the wafer (from the collection hand 27 to the collection belt 28) and is stored in the collection side wafer cassette 29. In this way, by storing the dummy wafers 40 together with a predetermined number of wafers in the supply side wafer cassette 23 and flowing the dummy wafers along the same route as the wafers, it is possible to transfer the dummy wafers 40 in units of wafer cassettes without adding a new transport device for the dummy wafers 40. The exposure equipment is being calibrated.

This embodiment has the advantage that a dummy wafer containing a photochromic material can be easily applied to an existing off-axis type exposure apparatus using a software method without adding any special hardware means. ing.

Further, FIG. 20 shows a case where the dummy wafer 40 cannot be reproduced on the exposure device or a large number of dummy wafers 40 are required.
FIG. 7 is a schematic diagram of an embodiment in which a dummy wafer-dedicated cassette and its transport device are provided separately from the supply and recovery side wafer cassettes 23 and 29.

While a series of exposure processing sequences are being repeated, when a predetermined time manually entered by the operator on the console in advance (or when a predetermined number of wafers has been processed), a normal wafer processing routine is started according to a command from the control device 20. Once stopped, the exposed wafer is placed on the wafer holding table 1.
It will be removed from 1. Thereafter, the dummy wafer 40 is transferred from the supply side dummy wafer cassette 202 to the standby stage 41 by a conveying means such as a belt 200, and the supply hand 26
The wafer is placed on the wafer holding table 11 by the following steps. After completing a series of calibrations, the dummy wafer 40 is recovered into the recovery side dummy wafer cassette 203 following the path of wafer holding table 11 → supply hand 26 → standby stage 41 → belt 200. Of course, at this time, it is also possible to add a standby stage 11, for example, a dummy wafer reproducing device, and after reproducing the calibrated dummy wafers, the dummy wafers can be recovered into the recovery side dummy wafer cassette 203.

Note that in FIGS. 19 and 20, the dummy wafer clearly has a recognition means different from that of the Si wafer, and the exposure apparatus in this embodiment has, for example, a pre-alignment stage 25.
The wafer has means for detecting the characteristics and identifying whether the Si wafer is a dummy wafer. For example, Si wafers are easily identified by giving them physical, optical, electrical, and other characteristics.

In the above embodiments, for example, a dummy wafer containing a magneto-optical recording material is inserted into a wafer cassette containing regular wafers, guided through a normal wafer transport path to a wafer holding table, and subjected to surface correction. It may also be configured to perform an action.

The present invention can be applied to conventional exposure equipment by using some manual support, by changing some units of the magnetic wafer chuck and polarizing microscope, and by erasing the magnetized image outside the equipment. can.

Although this embodiment shows an example in which the present invention is applied to a stepper having an optical projection lens, it can naturally be applied to an exposure apparatus having a precise stage position measuring means such as an X-ray stepper.

In addition to magneto-optical recording materials and photochromic materials, recording media with reversible writing and erasing properties include
Various recording materials can be used, such as amorphous silicon and chalcogen.

(Effects of the Invention) According to the present invention, a dummy wafer having a reversible material capable of writing and erasing, such as a magneto-optical recording material or a photochromic material, is used, and a magnetized image forms a latent image on the surface of the dummy wafer. By detecting the imaging state of the magnetized image and latent image, system errors such as reticle rotation, projection lens magnification, reference length, stage arrangement quirks, etc., as well as system errors caused by changes over time can be eliminated. The apparatus itself has the advantage of being able to automatically or semi-automatically perform good correction and superimpose the reticle and wafer with high precision.

[Brief explanation of the drawing]

1 is a schematic diagram of one embodiment of the present invention applied to an exposure apparatus, FIG. 2 is an explanatory diagram of a portion including the wafer stage and wafer transport system of FIG. 1, and FIG. 4th. Fifth. FIG. 6 is a schematic diagram of an embodiment of a dummy wafer according to the present invention,
FIG. 7 is a schematic diagram of the optical system of the wafer microscope, FIG. 8 is a flowchart showing the order of system error correction according to the present invention, FIG. 9 is an explanatory diagram of the pattern on the reticle surface, and FIG.
The figure is an explanatory diagram of a pattern on a dummy wafer surface according to the present invention, No. 11. 12th. 13th゜14th. FIG. 15 is a schematic diagram of an embodiment showing a method of erasing a magnetized image formed on a dummy wafer according to the present invention, and FIG. 16 is a schematic diagram showing writing and erasing of a magnetized image on a dummy wafer surface according to the present invention. ,
17th. FIG. 18 is a schematic diagram of an embodiment showing a method of erasing a latent image formed on a dummy wafer according to the present invention, and FIG.
．． FIG. 20 is a schematic diagram of an embodiment showing the steps from supplying to collecting dummy wafers according to the present invention. In the figure, 1 is an I2 shadow lens, 2 is a reticle, and 3 is a wafer. 4 is the illumination system, 5 is the reticle reference mark, 6 is the platen,
7 is a reticle microscope, 8 is a reticle stage, 10 is a wafer microscope, 11 is a wafer holding stand, 12 is a θZ stage, 13 is an XY stage, 40 is a dummy wafer, 23 is a wafer cassette, 24 is a supply belt, 25 is a pre-alignment stage , 26 is the supply hunt, 27 is the collection hand,
28 is a collection belt, and 29 is a wafer cassette. Patent applicant Canon Co., Ltd. Figure 1 Figure 3 Seki No. 9 Figure Isamu 8 + uq 49 Tsuta 14 Illustration] Figure 5 Figure 16 Kasumi Figure 17

Claims (2)

[Claims] (1) When aligning a first object and a second object so that they face each other, and exposing and transferring the pattern on the first object surface onto the second object surface, the first object is used instead of the second object. By forming an image of an alignment pattern provided on the object plane and detecting the imaging state of the alignment pattern image, the first
It is used to find the system error when exposing and transferring the pattern on the object surface onto the second object surface, and is sensitive to the light that irradiates the first object surface, and can be written and erased. 1. A dummy wafer for exposure equipment, characterized by comprising a reversible photosensitive material. (2) A dummy wafer for an exposure apparatus according to claim 1, wherein the photosensitive material is a magneto-optical recording material or a photochromic material.