JPH10256350A - Semiconductor manufacturing method and apparatus therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately detect an alignment mark position, without being influences by ruggedness of resist or thin film on a semiconductor wafer surface or resist ununiformity. SOLUTION: An alignment mark is detected on a work 1 when the work 1 is positioned such that the autocorrelation operation is made in terms of the alignment mark width about a photographed alignment mark image signal to obtain the position coordinates of the mark from the signal waveform in the operation result. An alignment mark on a semiconductor wafer 1 is e.g. imaged by CCD cameras 9, 10 to obtain image signals thereof about which an autocorrelation processor then computes the autocorrelation in terms of the alignment mark width and a mark position processor computes the position coordinates of the alignment mark from the signal waveform obtd. in the autocorrelation operation result.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an alignment mark formed on a semiconductor wafer to be processed when the semiconductor wafer is manufactured by applying a resist, exposing, developing, cleaning and the like. The present invention relates to a semiconductor manufacturing method and apparatus for performing a positioning operation and automatic focusing at that time, a positioning operation when spinning a semiconductor wafer, and a semiconductor wafer transfer operation.

[0002]

2. Description of the Related Art In a semiconductor manufacturing process, an alignment mark (hereinafter, referred to as an alignment mark) formed on a semiconductor wafer or a mask in an exposure process is detected, and based on the alignment mark position, a semiconductor wafer or a mask is detected. Aligning the mask is performed.

As a method of detecting an alignment mark in such alignment, for example, there is the following technique.

[0004] First, "Nikkei Micro Device" 1990
The technology described in the November issue, P112 to P113, scans a stage on which a semiconductor wafer is placed while irradiating a laser beam to a diffraction grating alignment mark drawn on the semiconductor wafer, and scans the alignment mark from the alignment mark. The diffracted light is received by a detector, and the signal photoelectrically converted by the detector is A / D (analog / digital) converted and stored as a digital signal in a waveform memory. The digital signal stored in the waveform memory is processed by a waveform processor to detect the position of the alignment mark.

Second, the technique described in Japanese Patent Application Laid-Open No. 63-274802 discloses that a semiconductor wafer set on a stage of a projection exposure apparatus emits light of at least three wavelengths or light containing these lights. The reflected light from the alignment mark drawn on the semiconductor wafer is dispersed, the optical path length of the reflected three wavelengths of reflected light is changed, and the reflected lights of these wavelengths are combined on the same surface. This is a method of detecting the position of the alignment mark by forming an image.

However, in the first technique, since a multilayer thin film and a resist are formed on the surface of the semiconductor wafer, there are irregularities and resist unevenness.

For this reason, in the method of using single-wavelength light for illuminating a semiconductor wafer, incident light causes multiple interference in a thin film or a resist on the surface of the semiconductor wafer, and the influence of unevenness of the thin film or the resist and unevenness of the resist. As a result, the accuracy of alignment mark position detection is significantly reduced.

In the second technique, there is no problem of being affected by the unevenness of the thin film and the resist, and the unevenness of the resist. However, the illumination system and the optical system are complicated and large-scale, and it is difficult to adjust and control them.

On the other hand, as another method for detecting an alignment mark, for example, there is the following technique. That is, thirdly, according to the technique described in Japanese Patent Application Laid-Open No. 2-122517, an auxiliary mark is formed on a semiconductor wafer at a predetermined interval from an alignment mark, and light or light is applied to the auxiliary mark. Pre-alignment is performed to scan the particle beam to detect the global position of the semiconductor wafer, and light,
Alternatively, it is a method of detecting a position shift of a semiconductor wafer by scanning a particle beam.

Fourth, the technique described in Japanese Patent Publication No. 63-35094 has a predetermined relative positional relationship with respect to an alignment mark (micro mark) on a semiconductor wafer, and is more effective than an alignment mark. A pilot mark is formed in a range over a large area. First, an electron beam is scanned on the pilot mark to detect its position, and then an electron beam is irradiated on the alignment mark in accordance with the position detection result and the relative positional relationship. In this method, the position of the alignment mark is detected by scanning.

A fifth technique is that a reduction projection exposure apparatus (stepper exposure) for exposing a pattern with light uses a CCD.
(Solid-state imaging device) Television image (T
V image), a pre-alignment mark is detected, the CCD camera is moved to the alignment mark position in accordance with the detected information, and fine alignment is performed based on the TV image.

The sixth technique will be described with reference to a wafer alignment apparatus using light shown in FIG.

A .theta. Stage 3 is disposed inside a vacuum chamber 2 for performing processing on the semiconductor wafer 1. The .theta.
A holder 4 for holding the semiconductor wafer 1 is provided on the stage 3. In addition, an alignment window 5 is formed in the upper part of the vacuum chamber 2.

An XYZ stage and a linear scale (hereinafter abbreviated as XYZ stage) 6 are provided outside the vacuum chamber 2.
Two camera detection systems 7 and 8 are attached.

Each of these camera detection systems 7 and 8 includes a CCD
Cameras 9 and 10 are attached, and a lamp house as a light source is connected via optical fibers 11 and 12.

With such a configuration, the XYZ stage 6 is driven in step # 1 according to the alignment flowchart shown in FIG. 40, and the two camera detection systems 7, 8 are set to the design value of each alignment mark position on the semiconductor wafer 1. To move up.

Next, in step # 2, as shown in FIG. 41, the images of the alignment marks 13 and 14 are captured at low magnification by the two CCD cameras 9 and 10, and in the next step # 3 each alignment data is obtained from each image data. The position coordinates of the marks 13 and 14 are measured.

Next, in step # 4, based on the measurement results of the respective alignment marks 13 and 14, the XYZ stage 6 is driven again to move the two camera detection systems 7 and 8 above the mark positions. At 5, as shown in FIG. 41, the image of the mark 15 is captured at high magnification by the two CCD cameras 9 and 10, and at the next step # 6, the position coordinates of the mark 15 are measured from the image data.

Then, in the next step # 7, the amount of displacement of the semiconductor wafer 1 is obtained from the position coordinates of the mark 15, and the semiconductor wafer 1 is aligned.

However, in the third and fourth techniques,
In each case, a mark for pre-alignment is required, and furthermore, since the pre-alignment mark is scanned and then the fine alignment mark is scanned to detect the position, the alignment process takes a long time.

When detecting a pre-alignment mark, the electron beam must be scanned over a wide range, and the scattered electrons at that time may affect other patterns on the semiconductor wafer 1.

In the fifth technique, a pre-alignment mark is required as in the third and fourth techniques, and the pre-alignment mark detection mechanism and the fine alignment mark detection mechanism are separate mechanisms. And the control becomes complicated.

Further, since the fine alignment is performed after the pre-alignment, it takes time for the alignment process.

In the sixth technique, two camera detection systems 7,
8 requires a mechanism for switching between a low magnification and a high magnification, and further detects the position of each of the alignment marks 13 and 14 at a low magnification and then switches to a high magnification to mark 15
Since the accurate position coordinates are obtained, the alignment process also takes time.

On the other hand, when performing alignment of a semiconductor wafer, focus is automatically adjusted on the alignment mark when the alignment mark is detected.

As this automatic focusing method, there is a method in which a focus position detecting system is provided separately from the mark detecting system, and there are a pneumatic method and an optical method.

The pneumatic type uses an air micrometer 16 as shown in FIG. 42, for example.
The semiconductor wafer 1 is mounted on the stage 17.
A shaft 18 is connected to a lower portion of the stage 17, and a pulse motor 21 is connected to the shaft 18 via a lever 19, a feed screw 20, and the like.

A piezoelectric element 22 is provided along the axis 18, and a micro displacement gauge 23 is provided below the stage 17.

Above the stage 17, a projection lens 24
, And an air micrometer 16 is provided below the projection lens 24.

With such a configuration, the projection lens 24
When air is blown out from the surroundings and hits the semiconductor wafer 1,
Since the variation in the distance between the projection lens 24 and the semiconductor wafer 1 is captured as the variation in the air pressure in the supply-side air chamber, this pressure difference is used for controlling the position of the projection lens 24.

In the optical system, as shown in FIG. 43, light (laser light) emitted from a light source 25 passes through an optical lens 26 and a light transmitting slit 27, and a slit image is projected onto the semiconductor wafer 1 by a projection lens 28. . The reflected light from the semiconductor wafer 1 is re-imaged on the light receiving slit 31 via the optical lens 29 and the vibrating mirror 30 as a pair of optical systems, and this is detected by the light receiving element 32 to perform the focus position detection.

FIG. 44 (a) shows another focusing method.
A half mirror 35 is disposed between the camera lens 33 and the image sensor 34 as shown in FIG. 7, and an image is detected by a focusing sensor 36 via the half mirror 35 to determine whether the image is in focus. Then, the motor 38 is driven via the amplifier 37 based on the determination result, and the focusing mechanism 39 is driven.
Is operated to move the camera lens 33 back and forth to control the camera lens 33 to be at the in-focus position. That is, since the output of the focus sensor 36 changes according to the position of the camera lens 33 as shown in FIG. 44 (b), the output of the focus sensor 36 is controlled, for example, from zero to the focus position. I do.

As another focusing method, there is a method using image processing.

In this method, several striations projected on an object are observed, distance is obtained from the width and sharpness of the striations, and the information is fed back to the position control system.

As a method of obtaining the sharpness from the pattern image of the object, for example, as shown in FIG.
The video signal output by the observation of 0 is differentiated by the differentiating circuit 41, and its absolute value is obtained by the next absolute value circuit 42 to obtain the edge signal of the pattern in the image. Then, a value of the edge signal exceeding the threshold value set in the threshold circuit 43 is sent to the integrating circuit 45 through the gate circuit 44, and the amount is integrated in the entire screen or in a window provided in the screen. Then, the overall strength of the edge signal is obtained, and the optimum lens position is obtained by the hill-climbing method so as to maximize the strength.

That is, this method is a method of observing a pattern image at the lens position while sending the lens position in minute steps, and performing the processing of the integrating circuit 45 from the differentiating circuit 41 to obtain the maximum value of the edge signal.

However, in the pneumatic automatic focusing shown in FIG. 42, air is blown onto the semiconductor wafer 1 so that if dust or the like is mixed in the air, the dust will adhere to the circuit pattern and cause defective products. Becomes Also, since the gap between the nozzle and the semiconductor wafer 1 cannot be made large,
This is disadvantageous for transporting the semiconductor wafer 1.

The optical system shown in FIGS. 43 and 44 is susceptible to the irregularities on the surface of the semiconductor wafer 1 and the reflectivity of the light under the wafer, which causes a reduction in the accuracy of automatic focusing.

Further, in these pneumatic and optical automatic focusing methods, a focus position detection system must be provided separately from the mark detection system, so that the mechanism and control system are complicated.

The automatic focusing using image processing is mechanically simple because the mark detection system and the focus position detection system are the same, but the time required for focusing is shorter than that of the pneumatic and optical systems. Take it. In addition, in the method of setting the threshold value of the image processing shown in FIG. 45, when the pattern density in the image changes, the threshold value changes and the focusing cannot be performed.

On the other hand, in the semiconductor manufacturing process, there is a process in which a semiconductor wafer is rotated at a high speed to carry out resist coating, development, cleaning and the like.

FIG. 46 is a configuration diagram of a semiconductor manufacturing apparatus provided with such a spin processing device. FIG. 46A is a diagram viewed from above, and FIG. 46B is a cross-sectional configuration diagram.

In the cup 46, resist application, development,
Processing such as cleaning is performed. A spin chuck 47 is arranged in the cup 46, and a rotation shaft of a spin motor 48 is connected to the spin chuck 47. A vacuum suction hole 49 is formed in the spin chuck 47. The spin chuck 47 and the spin motor 48 move up and down by a vertical mechanism (not shown), and the spin chuck 47 is disposed in the cup 46.

A wafer chuck 50 is disposed above the spin chuck 47. The wafer chuck 50 holds, for example, a dummy semiconductor wafer (hereinafter, referred to as a dummy wafer) 51 such as the semiconductor wafer 1 and moves in the XY directions by a linear drive mechanism (not shown) to position the dummy wafer 51 and the like on the spin chuck 47. It has become something.

In such an apparatus, the rotation centers of the spin chuck 47 and the wafer chuck 50 are aligned using the dummy wafer 51.

That is, the wafer chuck 50 holds the dummy wafer 51 having a center hole formed in the center, and is moved in the XY directions by the operation of the linear drive mechanism. Then, the center hole of the dummy wafer 51 and the spin chuck 47 are formed.
By visually matching the vacuum suction holes 49 of
The rotation centers of the spin chuck 47 and the wafer chuck 50 are aligned.

However, in such positioning of the rotation center, the rotation axis of the spin chuck 47 and the vacuum suction hole 49 are misaligned, and the dummy wafer 51 is visually observed.
It is difficult to perform positioning with high accuracy due to an error when the center hole is aligned with the vacuum suction hole 49 of the spin chuck 47.

On the other hand, in semiconductor manufacturing, there is a process of spin cleaning a semiconductor wafer by a spin processing apparatus.

FIG. 47 is a configuration diagram of such a spin processing apparatus, wherein FIG. 47A is a diagram viewed from above, and FIG. 47B is a cross-sectional configuration diagram.

The spin chuck 52 is provided with a wafer chuck section 53. This spin chuck 52
It is connected to a drive source (not shown) and rotates at high speed.

Each of the fixing pins 5
4, the semiconductor wafer 1 is mounted.

A cam plate 55 is provided on the spin chuck 52.
, And each lever 57 is provided on a shaft 56 of the cam plate 55. Each end of the lever 57 is provided with a respective wafer clamp pin 58.

The levers 57 are provided rotatably about respective shafts 59, and rotate the respective wafer clamp pins 58 in an arc to press the side surfaces of the semiconductor wafer 1.

With such a configuration, the semiconductor wafer 1
Are mounted on the wafer chuck portion 53 via the fixing pins 54, the levers 57 rotate around the shafts 59, respectively. Due to the rotation of these levers 57, each wafer clamp pin 58 makes an arc movement in the direction of arrow A as shown in FIG. 47 (a), and presses the semiconductor wafer 1 from its side.

Thereafter, the spin chuck 52 rotates at a high speed to clean the semiconductor wafer 1.

However, in the case of high-speed rotation, since the side surface of the semiconductor wafer 1 is pressed by the respective wafer clamp pins 58, dust is generated every time the side surface of the semiconductor wafer 1 comes into contact with the respective wafer clamp pins 58, and the wafer is rotated. It accumulates on the clamp pins 58 and adheres again to the semiconductor wafer 1.

On the other hand, in a semiconductor manufacturing apparatus, a semiconductor wafer or the like is carried as a carrier between processes.

FIG. 48 is a configuration diagram of such a carrier transport device, and FIG. 49 is a configuration diagram viewed from above.

The two carrier transporting devices 60 and 61 have a function of sequentially feeding a plurality of carriers 62 one by one in the direction of the arrow (a), for example, and are connected in series.

The carrier transport devices 60 and 61 include:
Each of the carrier transport drive devices 62 is provided along the longitudinal direction of the carrier transport device gantry 64.

The carrier transport driving device 62 has a function of moving the carrier vertical drive device 63 in the direction of the arrow (b) along the longitudinal direction of the carrier transport device base 64.

With such a configuration, the carrier up-down driving device 63 places one carrier 62 on its upper part. The carrier transport driving device 62 moves the carrier vertical driving device 63 with the carrier 62 placed thereon, and sequentially transports the carriers 62 one by one, for example, in the order of arrows A to D.

When the transport distance of the carrier 62 is longer than the transport distance of one carrier transport device and the transport distance is extended, as described above, two or more carrier transport devices 60 and 61 are used. Concatenate.

In this case, two carrier transfer devices 60,
49, a carrier transfer device 65 is arranged between them.

The carrier transfer device 65 is provided with a carrier 62
, A vertical driving device 67 that drives the hand 66 and moves up and down, and a driving device 68 that moves the vertical driving device 67 along the connecting direction of the carrier transport devices 60 and 61. I have.

Accordingly, the carrier transfer device 65 receives and holds the carrier 62 conveyed by the carrier conveying device 61, for example, with the hand 66, and transfers the hand 66 and the vertical driving device 67 and the like to the carrier conveying device 60 side. It moves and transfers the carrier 62 held by the hand 66 to the carrier transport device 60.

In some cases, two carrier transfer devices 60 and 61 are connected in a right angle direction, for example, as shown in FIG. Also in this case, the carrier transfer device 65 is disposed between the two carrier transfer devices 60 and 61 to change the transfer direction of the carrier 62.

However, when connecting a plurality of carrier transfer devices 60 and 61, a carrier transfer device 65 must be arranged between them. Since the carrier transfer device 65 includes the hand 66, the vertical drive device 67, and the drive device 68 as described above, it is large, requires a wide installation space, and is expensive.

On the other hand, in the manufacturing process of the semiconductor element,
An element is formed through a number of processes such as pattern formation by film formation and etching.

The film formation method includes a sputtering method and a CVD method, and the etching method includes a wet etching method and a dry etching method. Among them, the dry etching method includes the RIE method and the CDE method.

Each of these methods has its own characteristic, and in the formation stage, a necessary method is used in combination for each process.

For example, in the metal wiring forming step, the steps are as follows.

(1) a metal film forming step, (2) a resist pattern forming step, (3) a tri-etching step using a reactive gas, (4) a resist stripping step, and (5) a cleaning step.

Of these steps, steps (1), (3), and (4) are performed using a vacuum apparatus, and the other steps are performed at atmospheric pressure. ing.

As described above, a unique device is required for each process, and the method of combining the processes differs depending on the process.

For this reason, the devices are installed alone, and no connection is made between the devices. There is a problem that impurities such as dust and moisture and metal are mixed. The retention method is important.

For example, in a wiring pattern processing step in forming a metal wiring, the semiconductor substrate is held in the air after each processing, and a problem occurs due to the holding in the air. That is, post-processing of the semiconductor substrate after the process is an important factor, and a particular problem in this process is corrosion of the Al-based metal wiring.

The problem of corrosion of the Al-based metal wiring is that an acid is generated due to chlorine contamination of the semiconductor substrate and a reaction between moisture in the air due to exposure to the air after the dry etching step.
This acid corrodes the Al-based metal.

In response, attempts have been made to reduce the chlorine contamination and prevent the occurrence of corrosion by performing a heat treatment in a vacuum after the dry etching step.

For this reason, it is necessary to newly install a vacuum heat treatment apparatus after (1) the step of forming a metal film, which complicates the process and increases the time required for the processing, and the installation space of the apparatus is increased. Becomes larger.

Further, when forming a metal film in forming a multi-layer metal wiring, and particularly when forming a buried wiring in a connection hole, a problem arises because the steps are not connected. For example, at the time of connection between the lower active layer and the wiring layer, there is a problem that the resistance is increased and the adhesion strength is reduced due to the influence of the natural oxide film on the active layer surface.

Further, a natural oxide film or foreign matter between layers in a multilayer metal wiring causes an increase in resistance and a reduction in wiring reliability.

As described above, a problem occurs due to the fact that the respective steps are not connected. In particular, in the formation of a multilayer metal wiring, the cleanliness of the surface is important.

In view of the above, there is a semiconductor manufacturing apparatus for performing a plurality of processes using one vacuum apparatus, that is, a multi-chamber system.

FIG. 51 is a configuration diagram of such a multi-chamber system.

The transfer chamber 70 includes a plurality of process chambers 71 to 74 and a loading chamber 7.
5. The unloading chamber 76 is connected, and the process chambers 71 to 74 and the loading chamber 7 are connected.
5. The unloading chamber 76 is connected.

The plurality of process chambers 71 to 74 perform processes such as formation of a metal film, tri-etching using a reactive gas, and resist stripping on a semiconductor substrate as an object to be processed.

The loading chamber 75 and the unloading chamber 76 allow the semiconductor substrate to enter and leave the transfer chamber 70.

In the transfer chamber 70, a robot arm 77 is provided. This robot arm 7
Reference numeral 7 has a function of moving the semiconductor substrate between the process chambers 71 to 74, the loading chamber 75, and the unloading chamber 76.

However, if such a multi-chamber system is used, the process can be made continuous, but it becomes large in size as compared with an apparatus used in an individual process, and the structure of the robot arm 77 for moving the semiconductor substrate is also increased. In some cases, a failure may occur due to the complexity of the process, and the continuous process may be interrupted.

[0091]

As described above, the accuracy of the alignment mark position detection is low due to the influence of the unevenness of the thin film and the resist on the semiconductor wafer surface and the unevenness of the resist.

Further, fine alignment is performed after pre-alignment, or alignment is performed by switching from low magnification to high magnification, so that the alignment process takes time.

In the automatic focusing for the alignment mark, the air pressure type causes a defective product due to adhesion of dust in the air, and the optical type reduces the automatic focusing accuracy due to the influence of the reflectivity of the wafer under the wafer in the optical type. In addition, the image processing method requires more time than the pneumatic method and the optical method, and if a threshold value is provided, it is not possible to cope with the focusing when the pattern density in the image changes.

Further, in the spin processing apparatus, the rotational center of the spin chuck 47 and the wafer chuck 50 can be positioned with high precision by axial displacement between the rotary shaft of the spin chuck 47 and the vacuum suction hole 49 or by visual alignment. Is difficult to do.

When cleaning is performed by the spin processing apparatus, the side surface of the semiconductor wafer 1 and each of the wafer clamp pins 58 are removed.
Each time a contact occurs, dust is generated, and the dust adheres to the semiconductor wafer 1 again.

Further, in the case of transporting a carrier, a carrier transfer device 65 must be provided, which requires a large installation space and is expensive.

In the multi-chamber system, the size of the multi-chamber system is larger than those used in the individual steps, and a failure occurs due to the complicated structure of the robot arm 77, so that the continuous process may be interrupted.

Accordingly, it is an object of the present invention (claims 1 to 26) to provide a semiconductor manufacturing method and apparatus which can manufacture a highly reliable semiconductor element.

The present invention (claims 1 to 3) provides a semiconductor manufacturing method and apparatus capable of detecting an alignment mark position with high accuracy without being affected by a thin film on a semiconductor wafer surface, unevenness of a resist, and uneven resist. The purpose is to do.

Another object of the present invention (claims 4 to 10) is to provide a semiconductor manufacturing method and apparatus which can shorten the alignment processing time.

It is still another object of the present invention to provide a semiconductor manufacturing method and apparatus capable of accurately performing focusing on an alignment mark.

Another object of the present invention (claims 17 to 19) is to provide a semiconductor manufacturing apparatus capable of positioning a rotation center of a spin chuck and a wafer chuck with high accuracy.

Another object of the present invention is to provide a semiconductor manufacturing apparatus capable of preventing contamination of a semiconductor wafer without accumulating dust generated when the semiconductor wafer is held in a clamp pin. And

Another object of the present invention (claims 22 and 23) is to provide an inexpensive semiconductor manufacturing apparatus which does not take up space even if the transport distance or the transport direction changes.

The present invention (claims 24, 25, 26)
SUMMARY OF THE INVENTION It is an object of the present invention to provide a semiconductor manufacturing apparatus capable of moving a workpiece by a simple mechanism and performing a continuous process of semiconductor manufacturing and capable of miniaturization.

[0106]

According to the first aspect of the present invention, there is provided a semiconductor manufacturing method having a function of detecting an alignment mark formed on an object when positioning the object at a predetermined position. An auto-correlation operation using an alignment mark width as a parameter is performed on an image signal obtained when an image is captured, and the position coordinates of the alignment mark are obtained from a signal waveform resulting from this operation. Is the way.

According to the second aspect, in the semiconductor manufacturing apparatus having the function of detecting the alignment mark formed on the object when positioning the object at a predetermined position, the imaging means for imaging the alignment mark An autocorrelation calculation processing means for performing an autocorrelation calculation using the alignment mark width as a parameter with respect to the image signal output from the imaging means; and And a mark position coordinate processing means for obtaining position coordinates.

According to a third aspect of the present invention, there is provided a semiconductor manufacturing apparatus having a function of detecting an alignment mark formed on an object to be processed and positioning the object at a predetermined position based on the position of the alignment mark. Imaging means for imaging a mark; autocorrelation arithmetic processing means for performing an autocorrelation arithmetic operation on the image signal output from the imaging means using the alignment mark width as a parameter; and an arithmetic processing result of the autocorrelation arithmetic processing means. A mark position coordinate processing means for obtaining the position coordinates of the alignment mark from the signal waveform, and the processing object is moved based on the position coordinates obtained by the mark position coordinate calculation processing means to adjust the positional deviation of the processing object. And a movement adjusting means for achieving the above object.

According to a fourth aspect of the present invention, in a semiconductor manufacturing method having a function of detecting an alignment mark formed on a workpiece when positioning the workpiece at a predetermined position, the workpiece is disposed at a predetermined position. A semiconductor that achieves the above object by predicting the position of an alignment mark by an autoregressive model obtained based on a past positional deviation amount when the alignment is performed, and detecting the position of the alignment mark according to the predicted position of the alignment mark It is a manufacturing method.

According to a fifth aspect of the present invention, there is provided a semiconductor manufacturing method in which a parameter of an auto-regression model is determined in advance based on a past positional deviation amount of an object to be processed, and a design position of an alignment mark is corrected using the auto-regression model of the parameter. is there.

According to a sixth aspect of the present invention, there is provided a semiconductor manufacturing method having a learning function of storing past displacement amounts of an object to be processed and changing parameters of an autoregressive model to optimal values based on the displacement amounts. .

According to the present invention, the average value and the deviation of the positional deviation amounts are obtained by storing the past positional deviation amounts of the object to be processed, and when the average value and the deviation exceed a predetermined threshold value. A semiconductor manufacturing method for changing parameters of an autoregressive model.

According to the eighth aspect of the present invention, in a semiconductor manufacturing apparatus for detecting an alignment mark formed on a workpiece when positioning the workpiece at a predetermined position, a positional deviation of the workpiece relative to the predetermined position in the past is provided. Position correction means for predicting the position of the alignment mark by an autoregressive model obtained based on the amount, and mark detection means for detecting the alignment mark in accordance with the position of the alignment mark predicted by the position correction means. This is a semiconductor manufacturing apparatus that aims to achieve the above object.

According to the ninth aspect, in a semiconductor manufacturing apparatus which scans an object with an electron beam to draw on the object, the semiconductor manufacturing apparatus obtains the position based on a past positional deviation amount with respect to a predetermined position of the object. Position correcting means for predicting the position of the alignment mark by the obtained autoregressive model, and mark detection for scanning the object with the electron beam in accordance with the position of the alignment mark predicted by the position correcting means to obtain the actual position of the alignment mark Means, and a drawing correction means for obtaining a positional shift amount of the object from the actual position of the alignment mark detected by the mark detecting means, and correcting an irradiation position of the electron beam based on the positional shift amount. Accordingly, the present invention is directed to a semiconductor manufacturing apparatus for achieving the above object.

According to the tenth aspect, in the semiconductor manufacturing apparatus for performing a predetermined process on the object to be processed housed in the vacuum chamber, the predetermined value is obtained based on the past positional deviation amount of the object to be processed from the predetermined position. Position correction means for predicting the position of the alignment mark using the autoregressive model,
Imaging means for imaging the alignment mark; a moving mechanism for relatively moving the position of the imaging means and the object according to the alignment mark position predicted by the position correcting means; and a moving mechanism for moving the imaging means after the movement by the moving mechanism. The above object is attained by providing positioning means for obtaining an actual position of an alignment mark from image data obtained by imaging, obtaining an amount of displacement of the object from this position, and positioning the object at a predetermined position. This is a semiconductor manufacturing apparatus to be tried.

According to the eleventh aspect, in the semiconductor manufacturing method for imaging an alignment mark formed on an object to detect the position of the alignment mark when positioning the object at a predetermined position, The alignment mark is set at a plurality of locations and the alignment mark is imaged at each location. Each mark image data is processed at each of the plurality of intervals to obtain a variance value, and an image is captured from an approximate curve obtained from these variance values. This is a semiconductor manufacturing method that seeks to achieve the above object by obtaining the focal position of the device.

According to the twelfth aspect, each variance value is obtained by processing each mark image data at a plurality of intervals, and the position corresponding to the maximum value in the approximate curve obtained by these variance values is defined as the true focal position. Semiconductor manufacturing method.

According to the thirteenth aspect, when an image of an alignment mark is taken at a plurality of intervals with the object to be processed, the imaging position of the alignment mark is set at a plurality of locations along a predetermined direction on the object to be processed. Semiconductor manufacturing method.

According to a fourteenth aspect of the present invention, in a semiconductor manufacturing apparatus for detecting an alignment mark position by imaging an alignment mark formed on a target object when positioning the target object at a predetermined position. Imaging means for setting an interval with the body at a plurality of locations and imaging an alignment mark formed on the object to be processed at each of the plurality of intervals;
Focusing means for processing each mark image data at each interval obtained by imaging by the imaging apparatus to obtain each variance value, and obtaining a focal position of the imaging apparatus from an approximate curve obtained by these variance values. This is a semiconductor manufacturing apparatus that aims to achieve the above object.

According to the fifteenth aspect, the focusing means obtains each variance value by processing each mark image data at a plurality of intervals, and determines a position corresponding to the maximum value in the approximate curve obtained by these variance values. This is a semiconductor manufacturing apparatus having a true focal position.

According to the sixteenth aspect, when the alignment mark is imaged by setting the interval between the object to be processed and the imaging means at a plurality of positions, the imaging position of the alignment mark is set along a predetermined direction on the object to be processed. This is a semiconductor manufacturing apparatus set at a plurality of locations.

According to the seventeenth aspect, the workpiece held by the workpiece chuck is positioned and mounted on the spin chuck, and the spin chuck body is rotated at a high speed to perform predetermined processing on the workpiece. In a semiconductor manufacturing apparatus, a sensor having a spot-shaped detection region with respect to the rotation center of a spin chuck, and a dummy processing body mounted on the spin chuck and having a positioning portion for a processing object formed at a portion corresponding to the rotation center of the spin chuck And a positioning means for positioning the workpiece chuck at a position where a positioning unit is detected by a sensor when the dummy workpiece is mounted on a spin chuck. .

According to the eighteenth aspect, in the semiconductor manufacturing apparatus, the positioning portion is a projection formed at the center of the positioning processing body.

According to the nineteenth aspect, the positioning unit is a semiconductor manufacturing apparatus including a hole formed at the center of the positioning processing body and a signal source provided at the center position of the spin chuck body.

According to the twentieth aspect, there is provided a semiconductor manufacturing apparatus for mounting an object to be processed on a spin chuck, pressing the object to be processed from a side surface by a clamp pin, and rotating the spin chuck at a high speed to perform processing on the object. ,
A semiconductor manufacturing apparatus which achieves the above object by providing a contact position switching means for switching a contact position of a clamp pin with respect to an object to be processed.

According to the twenty-first aspect, the contact position switching means is a semiconductor manufacturing apparatus for causing a clamp pin to move in an arc.

According to a twenty-second aspect, in a semiconductor manufacturing apparatus provided with a transport device for sequentially transporting a plurality of workpieces in a predetermined direction by moving a transport table on which workpieces are placed by a transport driving device, The transport length of the transport drive unit in the device is formed longer than the transport unit body,
In addition, the transport driving device in the transport device is arranged at a position where the transport table can be transferred to the transport drive device of another transport device to be connected, and the plurality of transport devices are connected in series to achieve the above object. Semiconductor manufacturing equipment.

According to the twenty-third aspect, in a semiconductor manufacturing apparatus provided with a transport device for sequentially transporting a plurality of workpieces in a predetermined direction by moving a transport table on which workpieces are placed by a transport driving device, The transfer length of the transfer drive device in the apparatus is formed longer than the transfer device main body, and a transfer direction changing mechanism for changing at least the transfer direction of the object to be processed is arranged between each transfer device, and each transfer device is set to a predetermined length. The semiconductor manufacturing apparatus is to be arranged at an angle to achieve the above object.

According to the twenty-fourth aspect, in a semiconductor manufacturing apparatus having a plurality of processing chambers for performing processing on an object to be processed, the semiconductor processing apparatus is rotatably provided between the processing chambers and holds the object by this rotating operation. An arm mechanism for transferring between the processing chambers, and a plurality of opening / closing mechanisms for opening and closing the processing chambers for each processing chamber in response to a transfer operation of the object to be processed between the processing chambers by the arm mechanism. This is a semiconductor manufacturing apparatus provided to achieve the above object.

According to the twenty-fifth aspect, the arm mechanism is a semiconductor manufacturing apparatus provided in an arm standby chamber formed between the processing chambers.

According to the twenty-sixth aspect, a plurality of processing chambers for performing processing on objects to be processed arranged in a circle, an arm standby chamber formed in a central portion of these processing chambers, and an arm standby chamber An arm mechanism rotatably provided for holding the object to be processed by the rotating operation and transporting the object between the processing chambers, and each of the processing chambers in response to the transfer operation of the object between the processing chambers by the arm mechanism. And a plurality of opening / closing mechanisms for opening / closing the processing chamber for each case.

According to the first aspect, an image of an alignment mark formed on an object to be processed such as a semiconductor wafer is taken, and an autocorrelation operation using the alignment mark width as a parameter is performed on an image signal obtained by this image pickup. Then, the position coordinates of the alignment mark are obtained from the signal waveform resulting from the arithmetic processing. Thus, the alignment mark position can be detected with high accuracy without being affected by the thin film on the semiconductor wafer surface, the unevenness of the resist, and the unevenness of the resist.

According to the second aspect, the alignment mark formed on the object to be processed is imaged by the imaging means, and the autocorrelation operation using the alignment mark width as a parameter is performed by the autocorrelation operation processing means on the image signal. Do
The position coordinates of the alignment mark are obtained by the mark position coordinate calculation means from the signal waveform resulting from the calculation processing.

According to the third aspect, the alignment mark formed on the object to be processed is imaged by the imaging means, and the autocorrelation operation using the alignment mark width as a parameter is performed on the image signal by the autocorrelation operation processing means. Do
The position coordinates of the alignment mark are determined by the mark position coordinate calculation processing means from the signal waveform resulting from the calculation processing, and the processing object is moved by the movement adjusting means based on the position coordinates to adjust the positional deviation of the processing object. .

According to the fourth aspect, when an object to be processed such as a semiconductor wafer is arranged at a predetermined position, the position of the alignment mark is predicted by an auto-regression model obtained based on the past positional deviation amount of this arrangement. The position of the alignment mark is detected according to the predicted position of the alignment mark. Thereby, the alignment processing time can be shortened.

In this case, according to the fifth aspect, the parameters of the auto-regression model are obtained in advance based on the past positional deviation amount of the object to be processed, and the design position of the alignment mark is corrected using the auto-regression model of the parameters. .

According to the present invention, the past positional deviation amounts of the object to be processed are stored, and the parameters of the autoregressive model are changed to optimum values based on these positional deviation amounts.

According to the seventh aspect, the average value and the deviation of the past positional deviation amounts of the object to be processed are obtained, and when the average value and the deviation exceed a predetermined threshold value, the auto-regression model is calculated. Change parameters.

According to the eighth aspect, the position of the alignment mark is predicted by the position correction means using an autoregressive model obtained based on the past positional deviation amount of the object to be processed from the predetermined position. The alignment mark is detected by the mark detection means according to the position of the mark.

According to the ninth aspect, when the object is scanned with the electron beam to perform drawing on the object, the position correction means uses the past positional deviation amount with respect to the predetermined position of the object. The position of the alignment mark is predicted by the obtained autoregressive model, and the alignment mark is detected by the mark detecting means according to the predicted position of the alignment mark. Then, the amount of displacement of the object to be processed is determined from the actual position of the alignment mark by the drawing correction means, and the irradiation position of the electron beam is corrected based on the amount of displacement. Thereby, the alignment processing time can be shortened, and the drawing accuracy by scanning the electron beam can be increased.

According to the tenth aspect, when a predetermined process is performed on the object stored in the vacuum chamber,
The position of the alignment mark is predicted by an autoregressive model obtained based on the past positional deviation amount of the object to be processed by the position correction means, and the position of the imaging means and the object to be processed are determined according to the predicted alignment mark position. The position is relatively moved by the moving mechanism, and after this movement, the actual position of the alignment mark is obtained by the positioning means from the image data obtained by the imaging by the imaging means, and the positional deviation amount of the object to be processed is determined from this position. Then, the object is positioned at a predetermined position. Thus, the object to be processed can be stored at a predetermined position in the vacuum chamber with high accuracy, and a predetermined process can be performed.

According to the eleventh aspect, when focusing on the alignment mark formed on the object to be processed, the distance between the image pickup device and the object to be processed is set at a plurality of locations, and the alignment mark is imaged at each location. Then, each mark image data is processed at each of the plurality of intervals to obtain a variance value, and a focal position of the imaging apparatus is obtained from an approximate curve obtained from the variance values. Thereby, focusing on the alignment mark can be performed with high accuracy.

In this case, according to the twelfth aspect, each variance value is obtained by processing each mark image data at a plurality of intervals, and the part corresponding to the maximum value in the approximate curve obtained by these variance values is determined. Focus position.

According to the thirteenth aspect, when the alignment mark is imaged at a plurality of locations as described above, the imaging position of the alignment mark is set at a plurality of locations along a predetermined direction on the object to be processed. Image.

According to the fourteenth aspect, when focusing on the alignment mark formed on the object to be processed, the distance between the image pickup device and the object to be processed is set at a plurality of locations, and the alignment mark is imaged at each location. Then, each mark image data is processed by the focusing means at each of the plurality of intervals to obtain a variance value, and the focal position of the imaging apparatus is obtained from an approximate curve obtained from the variance values.

In this case, according to the fifteenth aspect, the focusing means processes each mark image data at a plurality of intervals to obtain each variance, and corresponds to the maximum value in the approximate curve obtained from these variances. The point where this occurs is the true focal position.

According to the sixteenth aspect, the imaging position of the alignment mark is set at a plurality of positions along the predetermined direction on the object to be processed.

According to the seventeenth aspect, when the spin chuck is rotated at a high speed to perform a predetermined process on the workpiece, the spin chuck is provided with a positioning portion for the workpiece at a portion corresponding to the rotation center of the spin chuck. The formed dummy processing body is mounted, and the processing object chuck is positioned by the positioning means such that the positioning portion of the dummy processing body is detected by the sensor. If the positioning portion of the dummy processing body is detected by the sensor in this way, the rotation center of the spin chuck and the wafer chuck can be positioned with high accuracy.

In this case, according to the eighteenth aspect, the positioning portion is a projection formed at the center of the positioning processing body.

Further, according to the nineteenth aspect, the positioning section comprises a hole formed at the center of the positioning processing body and a signal source provided at the center position of the spin chuck main body. Light and the like emitted from the sensor are detected by a sensor, and the rotation center of the spin chuck and the wafer chuck is aligned.

According to the twentieth aspect, when the object to be processed is pressed from the side surface by the clamp pin and the spin chuck is rotated at a high speed to perform the processing on the object to be processed, the contact position of the clamp pin to the object to be processed is determined. Switching is performed by position switching means. This prevents contamination of the semiconductor wafer without accumulating dust generated when the semiconductor wafer is held in the clamp pins.

In this case, according to the twenty-first aspect, the contact position switching means moves the clamp pin in an arc.

According to the twenty-second aspect, when the objects to be processed are sequentially fed one by one and transported, the transport length of the transport driving device is formed to be longer than the transport device body. When a plurality of transport devices are connected, the transport drive device is arranged at a position where the transport table can be transferred to a transport drive device of another transport device to be connected. Thereby, even if the transport distance changes, the transport can be performed without taking up space.

According to the twenty-third aspect, when the objects to be processed are sequentially fed one by one and transported, the transport length of the transport driving device in the transport device is formed longer than the transport device main body.
Then, a transfer direction changing mechanism for changing at least the transfer direction of the object to be processed is arranged between the transfer devices. Thereby, even if the transport direction changes, transport can be performed without taking up space.

According to the twenty-fourth aspect, the arm mechanism is rotatably provided between the plurality of processing chambers for performing processing on the processing object, and the processing apparatus holds the processing object by the rotation of the arm mechanism to interpose the processing chamber. The opening / closing mechanism provided in each processing chamber is operated in response to the transfer operation of the object to be processed by the arm mechanism, and the object to be processed is taken in and out of each processing chamber. Thus, the object to be processed can be moved by a simple mechanism to perform continuous processing of semiconductor manufacturing.

In this case, according to the twenty-fifth aspect, the arm mechanism is provided in an arm standby chamber formed between the processing chambers.

According to the twenty-sixth aspect, a plurality of processing chambers for performing processing on an object to be processed are arranged around the arm standby chamber provided with the arm mechanism, and the arm mechanism is rotated and operated. In response to the operation of the arm mechanism, each opening / closing mechanism for each processing chamber is opened and closed, and the object to be processed is transported between the processing chambers.

[0158]

BEST MODE FOR CARRYING OUT THE INVENTION

(1) Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. The same parts as those in FIG. 39 are denoted by the same reference numerals, and detailed description thereof will be omitted.

In the semiconductor manufacturing method according to the first aspect of the present invention, when the alignment mark formed on the semiconductor wafer 1 is detected when the semiconductor wafer 1 is positioned at a predetermined position, the alignment mark is imaged. An autocorrelation operation is performed on the image signal at this time using the width of the alignment mark as a parameter, and the position coordinates of the alignment mark are obtained from the signal waveform resulting from this operation.

FIG. 1 is a block diagram of a semiconductor manufacturing apparatus according to the second and third aspects of the present invention to which such a semiconductor manufacturing method is applied.

A lamp house 100 is connected to the camera detection systems 7 and 8 via optical fibers 11 and 12, and the light intensity of the lamp house 100 is adjusted by a light amount adjusting unit 101. I have.

Each of the camera detection systems 7 and 8 includes a driver 1
The optical magnification switching unit 103 is connected to the optical magnification switching unit 103 via an optical magnification switching unit 102. The optical magnification switching unit 103 includes a driver 102
To switch each of the camera detection systems 7 and 8 to either the low magnification mode or the high magnification mode.

The stage control unit 104 controls the drive of the θ stage 3 and the XYZ stage 6 via a driver 105. Among them, the position of the XYZ stage 6 from the linear scale 6a is controlled with respect to the XYZ stage 6. Has a function of taking in the measured value through the detector 106 and performing feedback control.

On the other hand, each of the CCDs 9 and 10 is controlled by the camera controller 107 to take a picture.
0 is output from the camera controller 1
07 and stored in the image input unit (image memory) 108.

The image processing unit 109 performs an autocorrelation operation on the image data stored in the image input unit 108 using the alignment mark width formed on the semiconductor wafer 1 as a parameter. It has a function of obtaining the position coordinates of the alignment mark from the waveform.

FIG. 2 is a specific configuration diagram of the image processing unit 109.

The image processing section 109 has the functions of a correlation parameter storage section 110, an autocorrelation calculation processing section 111, and a mark position coordinate calculation processing section 112.

The correlation parameter storage unit 110 stores a mark width w as a correlation parameter.

The autocorrelation calculation processing section 111 reads out the mark width w, which is the correlation parameter stored in the correlation parameter storage section 110, and performs autocorrelation calculation processing on the image data stored in the image input section 108. It has a function to perform.

The autocorrelation function is such that the input signal is Bo (i) and the output signal after the autocorrelation operation processing is B
(i)

(Equation 1) Becomes

Here, N is the number of signal data from m-th to n-th.

The mark position coordinate calculation processing section 112 has a function of obtaining the position coordinates of the alignment mark from the signal waveform of the calculation processing result of the autocorrelation calculation processing section 11.

The CPU 113 controls the operation of the light amount adjustment unit 101, the optical magnification switching unit 103, the stage control unit 104, and the image processing unit 109, and uses the position coordinates of the alignment marks obtained by the image processing unit 109 to adjust the position coordinates of the stage control unit 104. Has the function of passing to

When receiving the position coordinates of the alignment mark, the stage control unit 104 drives the driver 105 based on the position coordinates to operate the θ stage 3,
It has a function of adjusting θ of the semiconductor wafer 1.

Next, the wafer alignment operation in the apparatus configured as described above will be described.

The semiconductor wafer 1 is aligned with an orientation flat by an orientation flat alignment device (not shown), and is also transferred into the vacuum chamber 1 by a robot (not shown).

The light radiated from the lamp house 100 is
Camera detection system 7, 8 through each optical fiber 11, 12
And the semiconductor wafer 1 from these camera detection systems 7 and 8
Irradiated on the surface.

At this time, the light amount adjusting section 101 adjusts the light emission intensity of the lamp house 100 and adjusts the light amount of the base of the semiconductor wafer 1.

The optical magnification switching unit 103 drives the driver 102 to switch each of the camera detection systems 7 and 8 to either the low magnification mode or the high magnification mode.

After auto-focusing by the camera detection systems 7 and 8, each CCD is passed through these camera detection systems 7 and 8.
The cameras 9 and 10 capture two alignment mark images drawn on the surface of the semiconductor wafer 1 and output respective image signals. These image signals are digitally converted by the image input unit 108 and stored as image data.

Each alignment mark image picked up by each of the CCD cameras 9 and 10 has, for example, a cross shape as shown in FIG. 4, and a signal waveform of one line corresponding to this mark image has a mark edge portion as shown in FIG. In addition, due to the influence of unevenness and unevenness of the thin film and the resist, the noise component, that is, the swell at both ends is included.

However, the autocorrelation calculation processing section 111
The mark width w, which is the correlation parameter stored in the correlation parameter storage unit 110, is read, and the image input unit 108
Equations (1) to (3) for the image data stored in
The autocorrelation function of is calculated.

That is, the autocorrelation operation processing unit 111
The image data stored in the image input unit 108 is used as an original signal, and the autocorrelation functions of the equations (1) to (3) are operated on the original signal.

For example, if the original signal has the signal waveform shown in FIG. 6, the signal waveform B1 (i) shown in FIG. 7 is obtained by calculating the autocorrelation function of the equation (1). The signal waveform B2 shown in FIG.
(i) is obtained.

Thereafter, the autocorrelation calculation processing section 111 calculates
By calculating (3) and calculating the multiplied value of each signal waveform B1 (i) and B2 (i), the signal waveform B resulting from the processing of the autocorrelation function shown in FIG. 9 is obtained.

In the signal waveform B, a mark edge portion is remarkably appeared after noise is removed.

However, the mark position coordinate processing unit 11
2 detects the position coordinates of the alignment mark by detecting the mark edge portion of the signal waveform B resulting from the calculation processing by the autocorrelation calculation processing unit 11.

Then, the coordinates of the alignment mark position are passed to the stage control unit 104 by the CPU 113.

When receiving the position coordinates of the alignment mark, the stage control unit 104 drives the driver 105 based on the position coordinates to operate the θ stage 3 and
The θ adjustment of the semiconductor wafer 1 is performed.

As a result, alignment with respect to semiconductor wafer 1 is performed.

As described above, in the first embodiment, the alignment marks formed on the semiconductor wafer 1 are imaged by the CCD cameras 9 and 10, and the image signals are aligned by the autocorrelation operation processing unit 111. An autocorrelation operation is performed using the mark width as a parameter, and a mark position coordinate operation processing unit 1
Since the position coordinates of the alignment mark are obtained by the method 12, the position of the alignment mark can be detected with high accuracy without being affected by the thin film on the surface of the semiconductor wafer 1, the unevenness of the resist, and the unevenness of the resist.

In addition, by driving the θ stage 3 of the semiconductor wafer 1 based on the detected position coordinates of the alignment mark, the positional deviation with respect to the semiconductor wafer 1 can be adjusted to achieve highly accurate alignment.

Further, since the image processing method is adopted, the configuration is simple, and the adjustment and control thereof are easy.

Although the first embodiment of the present invention has been described with reference to the case where the present invention is applied to alignment of a semiconductor wafer in semiconductor manufacturing, the present invention is not limited to this. The present invention may be applied to an apparatus for identifying and inspecting a component by using processing.

(2) Next, a second embodiment of the present invention will be described with reference to the drawings.

According to a fourth aspect of the present invention, there is provided a method of manufacturing a semiconductor device, comprising: positioning the semiconductor wafer at a predetermined position; detecting the alignment mark formed on the semiconductor wafer; The position of the alignment mark is predicted by an auto-regression model obtained based on the past positional deviation amount when it is arranged at a predetermined position, and the position of the alignment mark is detected according to the predicted position of the alignment mark.

In this case, the parameters of the auto-regression model are obtained in advance based on the past positional deviation amount of the semiconductor wafer 1, and the design position of the alignment mark is corrected using the auto-regression model of the parameters.

Further, the past displacement amounts of the semiconductor wafer 1 are stored, and the parameters of the autoregressive model are changed to optimal values based on these displacement amounts.

Further, the past positional deviation amounts of the semiconductor wafer 1 are stored, and the average value and deviation of these positional deviation amounts are obtained. When these average values and deviation exceed predetermined threshold values, the auto regression model Change parameters.

FIG. 10 is a block diagram of a pattern drawing apparatus according to the eighth and ninth aspects of the present invention to which the semiconductor manufacturing method is applied.

The sample chamber 120 is provided on an anti-vibration table 121. The XY stage 1 is provided in the sample chamber 120.
The XY stage 122 is provided with a holder 123 for holding the semiconductor wafer 1 as a sample.

An electron gun 124 for forming an electron beam system is provided above the sample chamber 120. An electromagnetic illumination device 125, a first shaping aperture 126, and a second shaping aperture 127 are arranged on the traveling path of the electron beam emitted from the electron gun 124.
8, an electromagnetic reduction lens 129 and an electromagnetic objective lens 130 are arranged.

Further, a blanking electrode 131, a shaping deflector 132, a main deflector 133, and a sub deflector 134 are arranged on the traveling path of the electron beam.

On the other hand, main controller 135 operates pattern data generator 137 in accordance with the drawing data stored in drawing data section 136, and sends the drawing pattern data generated in pattern data generator 137 to beam controller 138. It is something to send.

The beam controller 138 has a function of controlling the operation of the blanking electrode 131, the shaping deflector 132, the main deflector 133, and the sub deflector 134 according to the drawing pattern data.

Accordingly, the electron beam emitted from the electron gun 124 becomes a uniform illumination beam by the electromagnetic illuminator 125, is shaped into a square by the first shaping aperture 126, and then is formed into a rhombus and a rectangle by the electromagnetic projection lens 128. It is projected on the second shaping aperture 127.

At this time, the irradiation position of the electron beam on the second shaping aperture 127 is controlled by the shaping deflector 132 so that the irradiation shape and area of the electron beam on the second shaping aperture 127 are in accordance with the CAD data.

The electron beam pattern that has passed through the second shaping aperture 127 is reduced and projected by the electromagnetic reduction lens 129 and the electromagnetic objective lens 130. At this time, the electron beam position with respect to the drawing pattern position is determined by the main deflector 133 and the It is controlled by the sub deflector 134.

In this case, the main deflector 133 controls the position in the frame, which is the drawing irradiation area, with reference to the position of the XY stage 122 on which the semiconductor wafer 1 is mounted, and the sub-polarizer 134 controls the position in the frame. Is controlled for a drawing range obtained by finely dividing the drawing.

Note that the blanking electrode 131 controls the presence or absence of an electron beam.

The electron detector 139 is provided above the sample chamber 120 and has a function of detecting reflected electrons generated when the semiconductor wafer 1 is irradiated with an electron beam.

[0212] The mark detection device 140 includes the electronic detector 13
9 has a function of receiving a detection signal output from the reference numeral 9 and processing the detection signal to detect a position of an alignment mark drawn on the semiconductor wafer 1.

The Z sensor 141 is provided above the sample chamber 120 and above the semiconductor wafer 1, and has a function of measuring the height of the semiconductor wafer 1. The height measurement value of the semiconductor wafer 1 is sent to the beam control device 138.

The length measuring device 142 has a function of measuring the position of the semiconductor wafer on the XY plane from the position of the XY stage 122, and this measured value is transmitted to the beam controller 138 and the XY stage controller 143. Has been sent.

The XY stage control device 143 operates in accordance with the position data transmitted from the main control device 135.
It has a function of driving and controlling the stage 122.

The position correcting device 144 has a function of predicting the position of an alignment mark on the semiconductor wafer 1 by using an autoregressive model obtained based on the amount of positional deviation of the semiconductor wafer 1 from a predetermined position in the past. Each function of the function coefficient storage unit 145 and the position correction unit 146 is provided.

The correction function coefficient storage section 145, in which coefficients are set by the main controller 135, has a function of storing the parameters (order, coefficient) of the autoregressive model.

The position correction unit 146 receives a position data command from the main control unit 135 and receives the correction function coefficient storage unit 14.
5 has a function of receiving the parameters stored in 5, calculating an autoregressive model using these parameters to correct the position data, and sending the corrected position data to the XY stage 143.

By sending the corrected position data to the XY stage controller 143, the XY stage controller 143 controls the drive of the XY stage 122 so as to correct the electron beam irradiation position based on the corrected position data. I do.

The main controller 135 includes the terminal device 14
7 is connected.

Next, an alignment method (registration) of the semiconductor wafer 1 will be described with reference to a flowchart of mark position detection shown in FIG.

First, an autoregressive model is determined.

Before drawing on the semiconductor wafer 1, in step # 10, the amount of misalignment of the semiconductor wafer 1, which is a sample, when transferred from an autoloader (not shown) to the XY stage 122 in advance is shown in FIG. Data is collected by detecting the alignment mark 1a drawn on the semiconductor wafer 1.

That is, the electron detector 139 detects reflected electrons generated when the semiconductor wafer 1 is irradiated with an electron beam, and sends a detection signal corresponding to the amount of reflected electrons to the mark detection device 140.

The mark detection device 140 receives the detection signal output from the electronic detector 139, and
The position of the upper alignment mark 1a is detected, and the position data is sent to the main controller 135.

Next, main controller 135 proceeds to step # 11.
In, the parameters (order, coefficient) of the autoregressive model are determined based on the collected position data, and the parameters (order, coefficient) are set in the correction function coefficient storage unit 145 of the position correction device 144 in the following step # 12. .

Here, a method of determining the parameters (order, coefficient) of the autoregressive model will be described.

.., N is the displacement data of the semiconductor wafer 1, the auto-regression model is as follows: Xn + 1 = a1 × n + a2 × n-2 +... + Am × xn-m + 1 + wn + 1 (4) m is the order, ai is the coefficient of the model, and wn + 1 is white noise.

The order m in the above equation (4) may be selected so that the value of the FPE represented by the following equation is minimized.

[0230]

(Equation 2) According to the above procedure, the order and the coefficient of the autoregressive model are obtained for the displacement data (Δx, Δy, Δθ) of the semiconductor wafer 1, and the autoregressive model is detected when the alignment (registration) mark on the semiconductor wafer 1 is detected. Using the model, the position correction unit 146 corrects the design position of the alignment mark 1a.

Next, a procedure for detecting a mark in semiconductor wafer alignment (registration) will be described.

The main controller 135 sends the design position (x, y) of the alignment mark 1a to the position corrector 144 in step # 13.

The position corrector 14 of the position corrector 144
In step # 14, the auto-regression model of the above equation (4) is calculated to calculate the current positional deviation amount (Δx, Δy, Δθ) from the past positional deviation data, and the following equation is calculated to calculate the alignment. The mark position (u, v) of the mark 1a is predicted and sent to the XY stage control device 143.

[0234]

(Equation 3) In step # 15, the XY stage control device 143 drives the XY stage 122 in accordance with the predicted mark position (u, v) of the alignment mark 1a, and moves the position of the alignment mark 1a.

Next, the electron beam emitted from the electron gun 124 scans according to the mark position (u, v). At this time, the electron detector 139 detects the reflected electrons from the semiconductor wafer 1 and outputs the detection signal. Is output.

The mark position detecting device 140 performs step #
At 16, the position of the alignment mark 1a is measured in response to the detection signal output from the electronic detector 139.

Next, main controller 135 proceeds to step # 1.
At 7, the position shift amount of the semiconductor wafer 1 is obtained based on the measured mark position, and the position shift amount is sent to the beam control device 138.

This beam control device 138 performs step #
At 18, the shaping deflector 132 according to the displacement amount,
The operation of the main deflector 133 and the sub deflector 134 is controlled, the position of the electron beam emitted from the electron gun 124 is corrected, and the pattern drawing on the semiconductor wafer 1 is performed.

Note that main controller 135 may store the past displacement amounts of semiconductor wafer 1 and change the parameters of the autoregressive model to optimal values based on these displacement amounts.

Further, main controller 135 proceeds to step # 19.
, The past displacement amounts of the semiconductor wafer 1 are stored, and the average value and deviation 3σ of these displacement amounts are obtained. In the next step # 20, the average value and deviation 3σ are obtained.
Exceeds the predetermined threshold value, the process proceeds to step # 21 to execute steps # 10 to # 12 again, recalculate the parameters of the autoregressive model, obtain the optimal parameters, and obtain the correction function coefficient storage unit 145. May be stored.

As described above, in the second embodiment, the position of the alignment mark 1a is predicted by an auto-regression model obtained based on the past displacement amount when the semiconductor wafer 1 is placed at a predetermined position. According to the predicted position of the alignment mark 1a, the alignment mark 1
Since the position a is detected, the alignment processing time can be shortened, and drawing accuracy by scanning the electron beam can be increased by shortening the alignment processing time.

That is, the position of the alignment mark 1a of the semiconductor wafer 1 placed on the XY stage 122 is predicted by a statistical processing method using an autoregressive model, and an electron beam is scanned at this position to obtain an accurate mark. Since the position is detected, pre-alignment is unnecessary, and the alignment processing time can be shortened.

The parameters (order,
Coefficient) is always set to the optimum value by the learning function, so that reliability can be improved.

Further, since a mechanism for detecting the pre-alignment mark is not required, the structure and control are simplified.

(3) Next, a third embodiment of the present invention will be described with reference to the drawings. The same parts as those in FIG. 1 are denoted by the same reference numerals, and detailed description thereof will be omitted.

FIG. 13 is a view showing the configuration of a wafer alignment apparatus according to the tenth aspect of the present invention to which the semiconductor manufacturing method is applied.

The position correcting device 150 has a function of predicting the position of an alignment mark on the semiconductor wafer 1 by using an autoregressive model obtained based on the amount of positional deviation of the semiconductor wafer 1 from a predetermined position in the past. Each function of the function coefficient storage unit 151 and the position correction unit 152 is provided.

The correction function coefficient storage unit 150
The coefficient is set by 3 and the parameter (order, coefficient) of the autoregressive model is stored.

The position correction unit 152 receives a position data command from the CPU 113, receives a parameter stored in the correction function coefficient storage unit 151, and calculates an autoregressive model using the parameter to correct the position data. And
It has a function of sending the corrected position data to the stage control unit 104.

By transmitting the corrected position data to the stage control unit 104, the stage control unit 104
The drive control of the XYZ stage 6 is performed so that the irradiation position of the light from each of the optical lens systems 7 and 8 is corrected based on the corrected position data.

Next, an alignment method of the semiconductor wafer 1 will be described with reference to a flowchart of mark position detection shown in FIG.

First, an autoregressive model is determined.

Before processing the semiconductor wafer 1, in step # 30, the amount of displacement of the semiconductor wafer 1 when the semiconductor wafer 1 has been transferred from the autoloader (not shown) to the XY stage 122 in the vacuum chamber 2 in advance is determined. 1
The data is collected by detecting the alignment mark 1a depicted in FIG.

That is, each of the CCD cameras 9 and 10 images each alignment mark 1a formed on the semiconductor wafer 1 through each of the optical lens systems 7 and 8, and outputs an image signal.

[0255] The image processing section 109 is provided for each CCD camera 9,
Image processing is performed on the image data obtained by each of the ten imagings,
The position of each alignment mark 1a is detected.

Next, in step # 31, the CPU 113 determines the parameters (order, coefficient) of the autoregressive model based on the collected position data, and in step # 32, determines the parameters (order, coefficient) in the position correcting device. 144 is set in the correction function coefficient storage unit 145.

According to the above procedure, the order and coefficient of the auto-regression model are obtained for the displacement data (Δx, Δy, Δθ) of the semiconductor wafer 1, and the auto-regression model is detected when the alignment mark on the semiconductor wafer 1 is detected. , The alignment mark 1a in the position correction unit 150
Correct the design position of.

Next, a procedure for detecting a mark in semiconductor wafer alignment will be described.

In step # 33, the CPU 113 sends the design position (x, y) of the alignment mark 1a to the position correction device 150.

The position corrector 15 of the position corrector 150
In step # 34, an autoregressive model is calculated to calculate the current position shift amount (Δ
x, Δy, Δθ), and calculates the following equation to predict the mark position (u, v) of the alignment mark 1a, and sends it to the stage control unit 104.

This stage control section 104 executes step #
In 35, the XYZ stage 7 is driven according to the predicted mark position (u, v) of the alignment mark 1a,
The position of the alignment mark 1a is moved.

Next, the optical lens systems 7 and 8 are switched to the high magnification mode in step # 36. Each C
The CD cameras 9 and 10 capture images of the alignment marks 1a on the semiconductor wafer 1 through the optical lens systems 7 and 8 in the high magnification mode and output image signals.

[0263] The image processing unit 109 includes:
Image processing is performed on each image signal from 10 and the mark positions of two alignment marks 1a are measured.

Next, in step # 37, the CPU 113 calculates the amount of displacement of the semiconductor wafer 1 based on the measured mark position, and sends the amount of displacement to the stage control unit 104.

This stage control section 104 executes step #
At 38, the XYZ stage 6 and the θ stage 3 are driven in accordance with the amount of displacement to perform alignment with the semiconductor wafer 1.

Note that the CPU 113 may store the past positional deviation amounts of the semiconductor wafer 1 and change the parameters of the autoregressive model to optimal values based on these positional deviation amounts.

In step # 39, the CPU 113 stores the past positional deviation amounts of the semiconductor wafer 1 and obtains the average value and the deviation 3σ of these positional deviation amounts. In the next step # 40, these average values and deviations are calculated. When the deviation 3σ exceeds the predetermined threshold value, the process proceeds to step # 41 and steps # 30 to # 32 are executed again, the parameters of the autoregressive model are recalculated, the optimum parameters are obtained, and the correction function coefficient is stored. The information may be stored in the unit 151.

As described above, according to the third embodiment, the alignment processing time can be shortened and the semiconductor wafer 1 can be moved to a predetermined position in the vacuum chamber 2 as in the second embodiment. Can be stored with high accuracy and a predetermined process can be performed.

Note that the second and third embodiments are
The present invention is not limited to application to an electron beam lithography apparatus or a wafer alignment apparatus, and may be used for other semiconductor manufacturing apparatuses, for example, alignment of a reduction projection exposure apparatus or an inspection apparatus, or length measurement SEM.
The present invention can also be applied to alignment of devices other than semiconductors that require alignment.

(4) Next, a fourth embodiment of the present invention will be described.

The semiconductor manufacturing method according to claims 11 to 13 of the present invention relates to a method for detecting an alignment mark position by imaging an alignment mark 1a formed on a semiconductor wafer 1 with an imaging device, for example, a CCD camera.
The distance between the semiconductor wafer 1 and the CCD camera is set at a plurality of locations, the alignment mark 1a is imaged at each location, and each mark image data is processed at each of the plurality of intervals to obtain a variance value. The focus position of the imaging device is obtained from the obtained approximate curve.

In this case, each variance value is obtained by processing each mark image data at a plurality of intervals, and the position corresponding to the maximum value in the approximate curve obtained by these variance values is defined as the true focal position.

In the case where the distance between the semiconductor wafer 1 and the CCD camera is set at a plurality of positions to image the alignment mark 1a, the imaging position of the alignment mark 1a is set at a plurality of positions along the predetermined direction on the semiconductor wafer 1. Set.

FIG. 15 is a configuration diagram of an alignment apparatus to which a semiconductor manufacturing apparatus according to claims 14 to 16 of the present invention is applied. The same parts as those in FIG. 1 are denoted by the same reference numerals, and detailed description thereof will be omitted.

The two camera detection systems 7 and 8 are arranged above the semiconductor wafer 1 so that the two alignment marks 1a drawn on the semiconductor wafer 1 as shown in FIG. Are located.

Also, these camera detection systems 7 and 8 are XYZ
By controlling the operation of the stage 6 and the stage control unit 160, the space between the stage 6 and the semiconductor wafer 1 is set at a plurality of positions, for example, as shown in FIG.
As shown in the figure, three intervals are set at the measurement start point Z1, the measurement points Z2 and Z3, respectively.

On the other hand, the image processing unit 161
08, each image data for each CCD camera 9, 10 and each measurement point Z1, Z2, Z3 is read, and these image data are processed to obtain a variance value v. It has a function of determining a position corresponding to the maximum value of the approximated curve as a true focal position of each of the CCD cameras 9 and 10.

Next, the wafer alignment operation in the apparatus configured as described above will be described.

The semiconductor wafer 1 is aligned in an orientation flat by an orientation flat alignment device (not shown), and is also transferred into the vacuum chamber 2 by a robot (not shown).

The light radiated from the lamp house 100 is
Camera detection system 7, 8 through each optical fiber 11, 12
And the semiconductor wafer 1 from these camera detection systems 7 and 8
Irradiated on the surface.

At this time, the light quantity adjusting section 101 adjusts the light emission intensity of the lamp house 100 and adjusts the light quantity of the base of the semiconductor wafer 1.

The optical magnification switching unit 103 drives the driver 102 to switch each of the camera detection systems 7 and 8 to the low magnification mode or the high magnification mode.

The two camera detection systems 7 and 8 are arranged corresponding to the two alignment marks 1a drawn on the semiconductor wafer 1 as shown in FIG.

Each CC through these camera detection systems 7 and 8
The D cameras 9 and 10 are connected to the 2
The image of the alignment mark 1a at the location is observed.

Next, automatic focusing using this alignment mark 1a is performed.

That is, the stage control section 160
Driving the Z stage 6, the camera detection systems 7 and 8 are
Is set to the measurement start point Z1 shown in (1).

Each of the CCD cameras 9 and 10 captures two images of the alignment mark 1a drawn on the surface of the semiconductor wafer 1 through the camera detection systems 7 and 8, respectively, and outputs respective image signals. These image signals are digitally converted by the image input unit 108 and stored as image data.

[0288] The image processing section 161 is provided for each of the CCD cameras 9 and 10 stored in the image input section 108 and for each measurement point Z.
1 Each other image data is read, a first-order differentiation process is performed on the image data, and a variance of the signal data is calculated and obtained.

FIG. 18 shows a signal profile of the k-th line with respect to the image data and a signal profile obtained by first-order differentiation.

The image processing section 161 performs first-order differentiation processing on the image data and the variance of the signal data for the entire screen of the image data or for some lines.
An averaging process is performed by integrating it.

Subsequently, the stage control section 160
The stage 6 is driven to sequentially set the camera detection systems 7 and 8 to measurement points Z2 and Z3 shown in FIG.

At these measurement points Z2 and Z3, each CC
The D cameras 9 and 10 capture images of two alignment marks 1a drawn on the surface of the semiconductor wafer 1 through the camera detection systems 7 and 8, respectively, and output respective image signals. These image signals are digitally converted by the image input unit 108 and stored as image data.

[0293] The image processing section 161 is provided for each of the CCD cameras 9 and 10 stored in the image input section 108 and for each measurement point Z.
2 and Z3, each image data is read out, a first-order differentiation process is performed on these image data in the same manner as described above, and the variance of the signal data is calculated and obtained.

Next, the image processing section 161 determines whether each measurement point Z 1,
A quadratic curve approximation equation v as shown in FIG. 19 is obtained from the relationship with the Z-axis direction using the variance values v for each of Z2 and Z3.

The quadratic curve approximation equation v is represented by the following equation: v = aZ ² + bZ + c (10)

The image processing section 161 detects the maximum value of the variance in the quadratic curve approximation equation v, and determines the position corresponding to this maximum value as the true focal position of each of the CCD cameras 9 and 10.

FIGS. 20 and 21 show experimental data on the dispersion value of the focusing. FIG. 20 shows the underlayer Poly of the semiconductor wafer 1, and FIG.
Underlayer Al.

As can be seen from these experimental data, the waveforms of the dispersion values of both the bases Poly and Al of the semiconductor wafer 1 are quadratic curves, and show the maximum value at the focal position.

The true focal positions of the CCD cameras 9 and 10 are sent to the stage controller 160 through the CPU 113. The stage control unit 160 drives the XYZ stage 6 in the Z-axis direction to control each of the camera detection systems 7 and 8 for each C
It moves to the true focal position of the CD cameras 9 and 10.

Thereafter, the CCD cameras 9 and 10 take images of the two alignment marks 1a drawn on the surface of the semiconductor wafer 1 through the camera detection systems 7 and 8, respectively.
Each image signal is output. These image signals are digitally converted by the image input unit 108 and stored as image data.

[0301] The image processing section 161 converts each image data stored in the image input section 108 into each alignment mark 1a.
Is calculated, and the amount of displacement of the semiconductor wafer 1 in the rotation direction is calculated and obtained. This displacement amount is passed to the stage control unit 160 by the CPU 113.

When the stage controller 160 receives the amount of displacement of the semiconductor wafer 1 in the rotation direction, the stage controller 160 drives the driver 105 based on the amount of displacement to operate the θ stage 3 and adjust the θ of the semiconductor wafer 1. I do.

As described above, in the fourth embodiment, the CCD cameras 9 and 10 are set at the measurement points Z1, Z2 and Z3 with respect to the semiconductor wafer 1, and the measurement points Z1 and Z2 are set.
, Z2, and Z3, the alignment marks 1a are imaged, the mark image data is processed to obtain variance values, and each CCD corresponds to the maximum variance value in the quadratic curve approximation equation obtained from these variance values. Since it was determined as the true focal position of cameras 9 and 10,
Focusing on the alignment mark 1a can be performed with high accuracy.

That is, since the image processing method is used, the configuration is simple, and the focusing can be performed without being affected by the unevenness, pattern density, and type of the base of the semiconductor wafer 1.

Further, since the focus position is determined by the three measurement points Z1, Z2, Z3, focusing can be performed at high speed and with high accuracy.

The fourth embodiment may be modified as follows.

For example, three measurement points Z1, Z2, Z3
It is needless to say that the accuracy of the true focal position Zm is improved by increasing the number of measurement points instead of obtaining the true focal position.

In order to increase the focusing accuracy, each of the camera detection systems 7 and 8 is centered on the determined focal position Zm.
Is scanned in a range of ± ΔZ, and a variance value in this range is obtained to detect the maximum value, whereby a more accurate true focus position can be determined.

In the fourth embodiment, each CCD
The camera 9 and 10 are set at each of the measurement points Z1, Z2 and Z3 at one position on the semiconductor wafer 1. However, the measurement start point Z1 is set at a plurality of positions in the X-axis and Y-axis directions. You may.

FIGS. 22 and 23 show such a measurement starting point Z1.
Shows the focus position (focus position) when the position is changed in the X-axis and Y-axis directions. Among them, FIG.
FIGS. 23A and 23B show the focus positions of the CCD camera 9 in the X-axis and Y-axis directions, and FIGS. 23A and 23B show the focus positions of the CCD camera 10 in the X-axis and Y-axis directions.

In the fourth embodiment, the case where the automatic focusing method is applied to the wafer alignment apparatus has been described. However, the inspection apparatus using the image processing of the semiconductor and the parts using the image processing other than the semiconductor are used. The present invention can also be applied to a device for identifying and inspecting a device.

(5) Next, a fifth embodiment of the present invention will be described. The same parts as those in FIG. 46 are denoted by the same reference numerals.

FIG. 24 is a block diagram of a semiconductor manufacturing apparatus provided with a spin processing device according to claims 17 to 19 of the present invention.

In the cup 46, processes such as resist application, development, and cleaning on the semiconductor wafer 1 are performed.

[0315] In the cup 46, the spin chuck 4 is provided.
7 is arranged, and the rotating shaft of the spin motor 48 is connected to the 7. The spin chuck 47 has a vacuum suction hole 49 formed therein.

The spin chuck 47 and the spin motor 48 are moved up and down by a vertical mechanism (not shown), and the spin chuck 47 is disposed in the cup 46.

[0317] Above the spin chuck 47, a wafer chuck 50 is disposed. The wafer chuck 50 holds, for example, a dummy semiconductor wafer (hereinafter, referred to as a dummy wafer) 170 such as the semiconductor wafer 1 and is moved in the X and Y directions by a linear driving mechanism of the transfer device 171.
The dummy wafer 170 and the like are positioned at the rotation center of the spin chuck 47.

A position detecting jig 172 is mounted above the spin chuck 47.

The position detection jig 172 is vacuum-sucked to the spin chuck 47 when aligning the rotation centers of the spin chuck 47 and the wafer chuck 50, and coincides with the rotation center of the spin chuck 47. A protrusion 173 is formed at a position where the protrusion 173 is located.

On the other hand, the dummy wafer 170 is provided with a reflection type position sensor 175 via an L-shaped sensor position adjustment mechanism 174.

The position sensor 175 outputs a beam spot, detects the reflected beam spot, and detects the position of the object to be detected, and outputs the beam spot toward the rotation center of the dummy wafer 170. I have. The beam spot diameter is formed to be substantially the same as the projection 173 of the position detection jig 172.

The position sensor 175 has a maximum output voltage when the beam spot diameter matches the projection 173.

A voltmeter 176 is connected to the output terminal of the position sensor 175, and the measured voltage value is
77.

The center detecting section 177 inputs the measured voltage value of the voltmeter 176 to detect the maximum position, and as shown in FIG. 25, changes the position of the dummy wafer 170 at which the measured voltage value becomes maximum to (x1 , Y1), and then, if the position where the measured voltage value of the voltmeter 176 becomes the maximum by rotating the spin chuck 47 by 180 ° is (x2, y2), ((x1 + x2) / 2, (y1 + y2) / 2 3) has a function of detecting that the rotation centers of the wafer chuck 50 and the spin chuck 47 coincide with each other.

[0325] The controller 178 includes the spin motor 48.
And a function of controlling the transfer operation of the transfer device 171 and the like, and has a function of rotating the spin chuck 47 by 180 ° when the measured voltage value of the voltmeter 176 indicates the maximum.

Next, a description will be given of the positioning operation of the apparatus configured as described above.

The position detecting jig 172 on which the projection 173 is formed is attracted to the spin chuck 47.

A dummy wafer 170 provided with a position sensor 175 is attached to the wafer chuck 50 by fitting the outer peripheral portion of the wafer chuck 50 into a taper.

The transfer device 171 slides the wafer chuck 50 in the X and Y directions in the XY plane.

By sliding the wafer chuck 50 in the X and Y directions in this manner, when the beam spot of the position sensor 175 and the projection 173 on the position detecting jig 172 match as shown in FIG. The output voltage of the sensor 175 indicates the maximum.

The output voltage of the position sensor 175 is measured by the voltmeter 176, and the measured voltage value is sent to the center detecting section 177.

The center detecting section 177 inputs the measured voltage value of the voltmeter 176 to detect the maximum position, and sets the position of the dummy wafer 170 at which the measured voltage value becomes maximum to be (x1, y1). .

The position (x1, y1) of the dummy wafer 170 is read from the transfer unit 171.

[0334] Here, when the fact that the maximum of the measured voltage value has been detected is sent from the center detection unit 177 to the controller 178, the controller 178 rotates the spin motor 48 to rotate the spin chuck 47 by 180 °.

When the spin chuck 47 rotates,
The position sensor 175 irradiates the beam spot, receives the reflected beam spot from the position detecting jig 172, and outputs a voltage corresponding to the beam spot.

When the spin chuck 47 rotates, the center detector 177 inputs the measured voltage value of the voltmeter 176 to detect the maximum value, and detects the maximum value of the measured voltage value. Position (x
2, y2).

[0337] The center detecting section 177
70 positions (x1, y1) and spin chuck 4
Based on (x2, y2), it is detected that the rotation centers of the wafer chuck 50 and the spin chuck 47 coincide with each other at the position ((x1 + x2) / 2, (y1 + y2) / 2).

Therefore, the transfer device 171 moves the wafer chuck 50 in the X and Y directions so as to be at this position, and makes the rotation centers of the wafer chuck 50 and the spin chuck 47 coincide.

As described above, in the fifth embodiment, when processing the semiconductor wafer 1 by rotating the spin chuck 47 at a high speed, the position detecting jig 172 on which the projection 173 is formed is attracted to the spin chuck 47. Then, the dummy wafer 170 provided with the position sensor 175 is fitted to the wafer chuck 50, and the rotation of the spin chuck 47 and the wafer chuck 50 is performed based on the position at which the position sensor 175 detects the protrusion 173 and indicates the maximum output voltage. Since the center is aligned, the spin chuck 4
Regardless of the shape of 7, the position of the rotation center of the spin chuck 47 and the wafer chuck 50 can be aligned with high precision.

Since such high-precision positioning can be performed, the accuracy required for the spin chuck 47 does not need to be so high, and the number of manufacturing steps for the spin chuck 47 can be reduced, and the cost can be reduced.

Further, the configuration is simple in that one position sensor 175 is used, and the processing for the output voltage of this position sensor 175 does not require complicated calculation or control.

Note that the fourth embodiment of the present invention relates to
It may be modified as follows.

For example, the position sensor 175 is not limited to being provided on the dummy wafer 170, but may be provided on the spin channel 47 as shown in FIG.

Further, as shown in FIG.
9 formed on the wafer chuck 5
0 and the position sensor 175 is
It is provided on the spin channel 47 via 81. Then, the light emitting sensor 182 is disposed on the support arm 181 at a position facing the position sensor 175.

With such a configuration, when the rotation centers of the spin chuck 47 and the wafer chuck 50 coincide with each other, the light from the light emitting sensor 182 incident on the position sensor 175 has the maximum light amount.

Therefore, the wafer chuck 5 showing the maximum light quantity
Based on the zero position and the position indicating the maximum light amount when the spin chuck 47 is rotated, the position that is the center of rotation between the spin chuck 47 and the wafer chuck 50 can be detected.

(6) Next, a sixth embodiment of the present invention will be described. The same parts as those in FIG. 47 are denoted by the same reference numerals, and detailed description thereof will be omitted.

FIG. 29 is a block diagram of a substrate holding device in a semiconductor manufacturing apparatus according to claims 20 and 21 of the present invention. FIG. 1A is a diagram viewed from above, and FIG. 1B is a cross-sectional configuration diagram.

This semiconductor manufacturing apparatus comprises a spin chuck 5
2 and the semiconductor wafer 1
Is pressed from the side by three wafer clamp pins 58, and the spin chuck 52 is rotated at high speed to perform spin cleaning on the semiconductor wafer 1.

The substrate holding device in the semiconductor manufacturing apparatus is provided with contact position switching means 190 for switching the contact positions of the three clamp pins 58 with the semiconductor wafer 1.

This contact position switching means 190 is
5 and the lever 57 are driven by the drive source 191 to rotate the lever 57 about the shaft 59 so that the clamp pin 58
30 has a function of making an arc movement in the A direction or the B direction as shown in FIG.

Specifically, the contact position switching means 190 first holds the semiconductor wafer 1 by causing the clamp pins 58 to circularly move in the direction A, and then causes the clamp pins 58 to circularly move in the direction B during spin cleaning. 1 is held.

Next, the operation of the device configured as described above will be described.

When the semiconductor wafer 1 is mounted on the wafer chuck 53 via the fixing pins 54, each of the three levers 57 rotates about a shaft 59.

In other words, these levers 57 are
By the drive of 91, the circular motion is made in the direction A around the shaft 59 via the cam plate 55 as shown in FIG. 30.

Each of the three wafer clamp pins 58 presses the semiconductor wafer 1 from the side surface of the semiconductor wafer 1 at the contact point a of the wafer clamp pins 58 by the circular motion in the direction A.

Next, each lever 57 makes an arc motion in the direction B around the shaft 59 via the cam plate 55 by driving of the drive source 191 as shown in FIG. The three wafer clamp pins 58 are moved at the contact point b of the wafer clamp pins 58 by the circular motion in the B direction.
The semiconductor wafer 1 is pressed down from the side surface.

Thereafter, the spin chuck 52 rotates at a high speed, and the semiconductor wafer 1 is cleaned.

As described above, by performing the spin cleaning by switching the contact point of each wafer clamp pin 58 to the semiconductor wafer 1 from a to b, at the time of cleaning the semiconductor wafer 1,
The cleaning of each wafer clamp pin 58, particularly the contact point a, is performed simultaneously.

As described above, in the sixth embodiment, the semiconductor wafer 1 is moved from the side to each wafer clamp pin 5.
8, when the spin chuck 52 is rotated at a high speed to perform the spin cleaning on the semiconductor wafer 1, the contact points a of the wafer clamp pins 58 on the semiconductor wafer 1 are switched from a to b.
Can be prevented from accumulating in each of the wafer clamp pins 58, and the contamination of the semiconductor wafer 1 can be prevented.

It should be noted that each wafer clamp pin 58 is not limited to the circular motion, but may be any mechanism having two or more contact points with the semiconductor wafer 1.

In addition to the spin cleaning for the semiconductor wafer 1, resist coating, developing,
It may be applied to perform processing such as washing.

(7) Next, a seventh embodiment of the present invention will be described. The same parts as those in FIG. 47 are denoted by the same reference numerals, and detailed description thereof will be omitted.

FIG. 31 is a block diagram of a carrier transport device in a semiconductor manufacturing apparatus according to claim 22 of the present invention. FIG. 1A is a side view of a transfer unit of the carrier transfer apparatus, and FIG. 1B is a view of the transfer unit viewed from above.

[0365] The carrier 200 serving as the main body of the transport unit is provided with a carrier sta- sis 201 on the upper part thereof, and each carrier guide 202 for mounting the carrier 62 on the carrier sta- sis 201 is provided.

The gantry 200 is provided with a carrier transport driving device 203. The carrier transport driving device 203 is formed to be longer than the longitudinal direction of the gantry 200 and is provided on one wall side in the gantry 200. In FIG. 31, the longitudinal direction of the gantry 200 is defined as the X-axis direction.

[0367] The portion of the carrier transport driving device 203 protruding from the gantry 200 is a transfer position 204 of the carrier 62 as shown in Fig. 31 (b).

[0368] On this carrier transport driving device 203,
A carrier transport table 206 is provided via a carrier vertical drive device 205. The carrier up / down drive device 205 has a function of moving up / down the carrier transport table 206 in the Z-axis direction.

When sequentially feeding each carrier 62 from, for example, right to left in the drawing, the carrier transport control device 207 controls the carrier up / down drive device 205 to ascend and places the carrier 62 on the carrier transport table 206 to carry the carrier guide. It has a function of controlling the transport of the carrier up-down drive unit 205 to a predetermined position in this state, and lowering the carrier up-down drive unit 205 when transporting the carrier up and down to the predetermined position.

Next, the carrier transport operation of the apparatus configured as described above will be described.

In this carrier transfer device, for example, after the leftmost carrier 62 in the drawing is transferred to another location by an AGV (automated guide vehicle) or a carrier transfer device (not shown), the remaining carriers 62 are sequentially transferred. You.

That is, the carrier transport control device 207
Controls the ascending / descending drive device 205 to place the carrier 62 on the carrier transport table 206 and let the carrier 62 escape from the carrier guide 202.

In this state, the carrier transport control device 207
Controls the drive of the carrier transport driving device 203 in the X-axis direction, and controls the transport of the carrier vertical drive device 205 to a predetermined position.

Then, when the carrier 62 is transported to a predetermined position, the carrier transport control device 207 controls the carrier up / down driving device 205 to descend, and lowers the carrier 62 to the carrier guide 202.

Next, the transport unit of the carrier transport device is set to 2
An embodiment in which the link is connected will be described with reference to the configuration diagram of FIG. FIG. 3A is a side view, FIG. 3B is a view from above, and FIG. 3C is a view from the longitudinal direction of the transport unit. The same parts as those in FIG. 31 are denoted by the same reference numerals, and detailed description thereof will be omitted.

When two transport units are connected, the carrier transport driving device 203 in one transport unit is provided on one wall side in the gantry 200, and the carrier transport driving device 203 in the other transport unit is mounted on the gantry 200.
It is provided on the other wall side in 00.

[0377] When these transport units are connected in series, as shown in FIG.
3, a portion where the other transfer position 204 and the other carrier transport driving device 203 are arranged to face each other, that is, a transfer position 208 is formed.

Next, the carrier transport operation in which two transport units are connected in series will be described.

In one transport unit (on the right side in the drawing), the carrier transport control device 207 raises the carrier vertical drive device 205 to control the carrier transport table 20.
The carrier 62 is put on the carrier 6 and escapes from the carrier guide 202.

In this state, the carrier transport control device 207
Controls the drive of the carrier transport driving device 203 in the X-axis direction, and moves the carrier vertical driving device 205 to the transfer position 208.
Transport control.

Then, the carrier 62 is moved to the delivery position 208.
When the carrier is transported to the carrier 6, the carrier transport control device 207 controls the
Lower 2

Subsequently, in the other transport unit (on the left side in the drawing), the carrier transport control device 207 controls the carrier vertical drive device 205 at the transfer position 208 to raise the carrier 6 on the carrier transport table 206.
2 is put away from the carrier guide 202.

In this state, the carrier transport control device 207
Controls the drive of the carrier transport driving device 203 in the X-axis direction, and controls the transport of the carrier vertical drive device 205 to a predetermined position.

When the carrier 62 is transported to a predetermined position, the carrier transport control device 207 lowers the carrier vertical drive device 205 to lower the carrier 62.

Thereafter, the operation of transporting the carriers 62 is repeated, and the carriers 62 are sequentially fed one by one.

As described above, in the seventh embodiment, the transport length of the carrier transport driving device 203 is set to
Since the transfer unit is formed to be longer than 0 and the transfer units are connected in series, for example, two times, the transfer position 208 of the carrier 62 is formed. Therefore, a carrier transfer device separately installed for the transfer of the carrier becomes unnecessary, and the transfer distance of the carrier 62 changes. However, it can be transported inexpensively without taking up space.

In the above embodiment, a plurality of transport units may be connected instead of connecting two transport units in series.

(8) Next, an eighth embodiment of the present invention will be described.

FIG. 33 is a block diagram of a carrier transport device in a semiconductor manufacturing apparatus according to claim 23 of the present invention. The same parts as those in FIG. 31 are denoted by the same reference numerals, and detailed description thereof will be omitted.

[0390] This carrier transport device shows an arrangement in which two transport units (mounts 200, 200) are connected in a perpendicular direction.

A transport direction changing mechanism 210 as shown in FIG. 34 is arranged between these transport units. Note that, in the figure, A, which indicates the positional relationship between each transfer position P, Q at which the carrier 62 is transferred, and each transport unit,
B is described.

The transport direction changing mechanism 210 has a function of changing the transport direction of the carrier 62. The transport direction changing mechanism 210 includes a first carrier transport drive device 211, a second carrier transport drive device 212, and a carrier rotation drive device 213. It is configured.

The first carrier transport driving device 211 has a function of transporting the carrier 62 in the X-axis direction, for example, from the delivery position P to a line on the next delivery position Q.

The second carrier transport driving device 212 has a function of transporting from the carrier transfer position P line to the next transfer position Q in the Y-axis direction.

The carrier rotation driving device 213 has a function of rotating the carrier 62 to change the transport direction of the carrier 62.

Next, a description will be given of the carrier transport operation of the apparatus configured as described above.

In one transport unit (lower right in the drawing),
A transport operation similar to that of the transport unit (FIG. 31) is performed.

That is, the carrier transport control device 207
Controls the ascending / descending drive device 205 to place the carrier 62 on the carrier transport table 206 and let the carrier 62 escape from the carrier guide 202.

In this state, the carrier transport control device 207
Controls the drive of the carrier transport drive device 203 in the X-axis direction, and controls the transport of the carrier vertical drive device 205 to the transfer position P.

Then, when the carrier 62 is transported to the delivery position P, the carrier transport control device 207 lowers the carrier vertical drive device 205 to lower the carrier 62. Then, the carrier transport control device 207 controls the drive of the carrier transport drive device 203 and returns the carrier vertical drive device 205 and the carrier transport table 206 to their original positions.

Thereafter, the first carrier transport driving device 21
1 conveys the carrier 62 in the X-axis direction from the delivery position P to a line on the next delivery position Q.

Next, the carrier rotation driving device 213 rotates the carrier 62 by 90 ° in the CCW direction and
Change the transport direction of.

Further, the second carrier transport driving device 212
Is transported in the Y-axis direction from the carrier transfer position P line to the next transfer position Q, and is transferred to the other transfer unit (upper left in the drawing).

As described above, in the eighth embodiment, when the carriers 62 are transported one by one in order,
The transport length of the carrier transport driving device 203 is formed to be longer than the gantry 200, and the carrier 6
Since the transport direction changing mechanism 210 for changing the transport direction of No. 2 is arranged, a carrier transfer device separately installed for carrier delivery is not required, and a space is taken even if the transport direction of the carrier 62 changes to a right angle direction or the like. And can be transported at low cost.

Note that the eighth embodiment may be modified as follows.

For example, when the transport units are transported by being connected at right angles, the first carrier transport driving device 211 and the second
Since the carrier transport driving device 212 has a constant stroke, it may be configured to use an air cylinder in addition to the motor.

[0407] Further, the carrier rotation driving device 213 may be configured to use a rotor actuator in addition to the motor.

[0408] Further, the carrier up-down driving device 205 may be configured to use an air cylinder in addition to the motor.

(9) Next, a ninth embodiment of the present invention will be described.

FIG. 35 is a block diagram of a multi-chamber system in a semiconductor manufacturing apparatus according to claims 24 and 25 of the present invention.

The loading chamber 220 and a plurality of process chambers, for example, the process chambers 221 and 22
2 are arranged linearly.

The arm standby chambers 223 and 224 are disposed between the loading chamber 220 and the process chambers 221 and 222, respectively.

The arm standby chambers 223 and 224
The robot arm 225 for transferring the semiconductor wafer 1 between the loading chamber 220 and each of the process chambers 221 and 222 is provided therein.

Each of these robot arms 225 rotates about a rotation axis 226, and causes the tip 225a holding the semiconductor wafer 1 to move in an arc so as to be transferred between, for example, the loading chamber 220 and the process chamber 221. It has become.

Each arm standby chamber 223, 224
A valve 227 is provided between the load chamber 220 and each of the process chambers 221 and 222.

The valves 227 partition between the arm standby chambers 223 and 224, the loading chamber 220 and the process chambers 221 and 222, respectively.
And it opens and closes.

FIG. 36 shows such an arm standby chamber 22.
It is a block diagram seen from the side of 3,224. A rotation mechanism 228 is connected to a rotation shaft 226 of the robot arm 225.

The semiconductor wafer 1 is placed on the upper and lower pins 229 in each of the process chambers 221 and 222. The upper and lower pins 229 are vertically moved by a wafer vertical drive mechanism 230.

Each of the valves 227 is provided with a valve up / down drive mechanism 231. The valve up / down drive mechanism 231 moves each valve 227 up and down to open and close each chamber.

The controller 232 controls the robot arm 2 according to the transfer operation of the semiconductor wafer 1 between the chambers.
It has a function to drive and control the 25 rotation mechanisms 228, the wafer vertical drive mechanism 230, and the valve vertical drive mechanism 231.

Next, the wafer transfer operation of the multi-chamber system configured as described above will be described.

(A) For example, during processing, each wafer upper and lower pin 229 in each of the process chambers 221 and 222
Are in a lowered state, and hold the semiconductor wafer 1.

Each valve 227 is raised by a valve up / down drive mechanism 231 to close the valve between the loading chamber 220 and the process chamber 221 and between the process chambers 221 and 222.

Each robot arm 225 is in a standby position in the arm standby chamber 223 at a standby position.

(B) When the process is completed, each wafer upper and lower pin 229 in each of the process chambers 221 and 222 is
The semiconductor wafer 1 is driven by the drive of the wafer vertical drive mechanism 230.
Is raised to a predetermined position while holding.

Each valve 227 is lowered by the valve up / down drive mechanism 231 to open the valve between the loading chamber 220 and the process chamber 221 and between the process chambers 221 and 222.

Each robot arm 225 is waiting at a standby position in the arm standby chamber 223.

(C) Subsequently, with the upper and lower pins 229 raised and the valve 227 opened, the robot arm 225 moves in an arc, and holds the tip 225 a at the arm process position, for example, holding the semiconductor wafer 1 in the process chamber 221. Move to where it is.

For example, each of the process chambers 221 and 22
The robot arm 225 between the two moves in an arc by the driving of the rotation mechanism 228, and moves the tip 225a upward while the semiconductor wafer 1 in the process chamber 221, for example, is held.

(D) Here, when the robot arm 225 holds the semiconductor wafer 1, the upper and lower pins 229 are lowered by the driving of the wafer vertical drive mechanism 230.

(E) Next, while holding the semiconductor wafer 1, the robot arm 225 makes a circular motion by driving the rotation mechanism 228, and holds the next arm process position, for example, the semiconductor wafer 1 in the process chamber 222. Move up.

(F) Next, when the robot arm 225 reaches above the position where the semiconductor wafer 1 is held, the upper and lower pins 229 in the process chamber 222 are
It rises by driving. Thereby, the semiconductor wafer 1
Is transferred to upper and lower pins 229 in the process chamber 222.

(G) Next, the robot arm 225 makes a circular motion by driving the rotating mechanism 228 and stands by at a standby position in the arm standby chamber 224.

(H) Next, each process chamber 221, 22
2, the wafer upper and lower pins 229 respectively hold and lower the semiconductor wafer 1, and the valves 227 are raised by a valve up and down driving mechanism 231, between the loading chamber 220 and the process chamber 221, and in each process. The valve between the chambers 221 and 222 is closed.

Thereafter, each process chamber 221, 22
Process processing is performed in 2.

As described above, in the ninth embodiment, the arm standby chambers 223 and 22 are located between the loading chamber 220 and the process chambers 221 and 222.
4, the semiconductor wafer 1 is held and transferred between the chambers 220 to 222 by the rotation operation of the robot arm 225, and each of the chambers 220 to 222 is connected to the arm standby chamber in response to the transfer operation of the semiconductor wafer 1. Since the valves 227 provided between the chambers 223 and 224 are opened and closed, the semiconductor wafer 1 is transported between the chambers 220 to 222 by a simple mechanism, and continuous processing of semiconductor manufacturing can be performed.

That is, a semiconductor manufacturing process under vacuum, for example, a continuous process that does not involve opening to the atmosphere, such as film formation, etching, and annealing, can be performed.

The semiconductor wafer 1 is located between each chamber.
Since the transfer speed of the semiconductor device is increased and the transfer mechanism is simplified, the continuous processing in the semiconductor manufacturing process has a special effect.

Further, if the transfer configuration is provided by the robot arm 225 in the arm standby chambers 223 and 224, the size of the intermediate chamber between the process chambers can be reduced, and the size can be reduced as compared with the conventional multi-chamber system. Is possible.

(10) Next, a tenth embodiment of the present invention will be described.

FIG. 37 is a block diagram of a multi-chamber system in a semiconductor manufacturing apparatus according to claims 24 and 25 of the present invention.

The loading chamber 240 and a plurality of process chambers, for example, the process chambers 241, 24
Are arranged in a straight line.

The arm standby chambers 243, 244,... Are arranged between the loading chamber 240 and the process chambers 241, 242, respectively.

The arm standby chambers 243, 244
The robot arm 245 for transferring the semiconductor wafer 1 between the loading chamber 240 and each of the process chambers 241 and 242 is provided therein.

Each of the robot arms 245 rotates about a rotation axis 246, and moves in an arc while holding the semiconductor wafer 1, for example, in the loading chamber 24.
0 and the process chamber 241.

Each arm standby chamber 243, 244
A valve 247 is provided between the load chamber 240 and each of the process chambers 241, 242.

The valves 247 partition between the arm standby chambers 243 and 244, the loading chamber 240, and the process chambers 241 and 242, respectively.
And it opens and closes.

Also, in the atmosphere adjacent to the loading chamber 240 and the process chamber of the final process,
Each of the robot arms 248 and 249 is arranged.

[0449] These robot arms 248 and 249
The semiconductor wafer 1 is transferred into the loading chamber 240 by a circular motion similarly to each robot arm 245, or transferred from the final process chamber to the outside.

With such a configuration, the semiconductor wafer 1 is transferred into the loading chamber 240 by the circular motion of the robot arm 248, and then the robot arm 245 in each arm standby chamber 243 is moved according to the processing time. The semiconductor wafer 1 is transferred between the process chambers 241, 242,... By the circular motion.

As a result, the semiconductor wafer 1 can be continuously processed in a vacuum atmosphere.

(11) Next, an eleventh embodiment of the present invention will be described.

FIG. 38 is a configuration diagram of a multi-chamber system in a semiconductor manufacturing apparatus according to claim 26 of the present invention.

The loading chambers 250 and 3 are annularly
Two process chambers 251 to 253 are arranged.

At the center of these chambers 250 to 253, a cross-shaped arm standby chamber 254 is arranged.

In this arm standby chamber 254, a cross-shaped robot arm 255 is provided. The robot arm 255 rotates about a rotation shaft 255a, and a holding portion for the semiconductor wafer 1 is provided at each cross-shaped tip.

Also, the loading chamber 250, between the loading chamber 250 and the process chamber 251, and between the process chambers, 251 to
Between 253, valves 256 to 261 are provided, respectively.

The valves 256 to 261 partition between the arm standby chamber 254, the loading chamber 250, and the process chambers 251 to 253, and open and close.

With such a configuration, when the semiconductor wafer 1 is loaded from the outside into the loading chamber 250, the robot arm 255 holds the semiconductor wafer 1 in the loading chamber 250.

At this time, the robot arm 255 simultaneously holds the semiconductor wafer 1 in each of the process chambers 251 to 252.

Further, each of the valves 257 to 261 is opened.

Then, after holding the semiconductor wafer 1, the robot arm 255 rotates about the rotation axis 255a, and conveys, for example, the semiconductor wafer 1 in the loading chamber 250 to the arm process position of the process chamber 251 and at the same time, The semiconductor wafer 1 in the process chamber 251 is transferred to the next arm processing position of the process chamber 252, and the semiconductor wafer 1 in the process chamber 252 is transferred to the next arm processing position of the process chamber 253.

The semiconductor wafer 1 in the process chamber 253 is unloaded to the outside.

Thereafter, each of the valves 257 to 261 is closed, and a process is performed in each of the process chambers 251 to 253.

As described above, in the eleventh embodiment, the process chambers 251 to 253 are arranged circumferentially around the arm standby chamber 254, and the robot arm 255 is rotated and the valves 25
Since the semiconductor wafers 7 to 261 are opened and closed in response to the operation of the robot arm 255, the semiconductor wafer 1 is transferred between the chambers 220 to 222 by a simple mechanism, and the semiconductor wafer 1 is continuously processed, for example, under vacuum. A semiconductor manufacturing process, for example, a continuous process that does not involve opening to the atmosphere such as film formation, etching, and annealing can be performed.

When the processing times of the process chambers 251 to 253 are substantially equal, the semiconductor wafers 1 in the process chambers 251 to 253 can be simultaneously transferred, so that the productivity of semiconductor manufacturing can be further improved.

[0467]

As described in detail above, claims 1 to 5 of the present invention.
According to 26, a semiconductor manufacturing method and apparatus capable of manufacturing a highly reliable semiconductor element can be provided.

According to the first to third aspects of the present invention, there is provided a semiconductor manufacturing method and apparatus capable of detecting an alignment mark position with high accuracy without being affected by a thin film on the surface of a semiconductor wafer, unevenness of a resist, and uneven resist. Can be provided.

According to claims 4 to 10 of the present invention, it is possible to provide a semiconductor manufacturing method and apparatus capable of shortening the alignment processing time.

Further, according to the ninth aspect of the present invention, it is possible to provide a semiconductor manufacturing apparatus capable of shortening the alignment processing time and increasing the drawing accuracy by scanning an electron beam.

Further, according to the tenth aspect of the present invention, it is possible to provide a semiconductor manufacturing apparatus which can store an object to be processed at a predetermined position in a vacuum chamber with high accuracy and perform a predetermined process.

According to claims 11 to 16 of the present invention,
It is possible to provide a semiconductor manufacturing method and a semiconductor manufacturing method capable of accurately performing focusing on an alignment mark.

According to claims 17 to 19 of the present invention,
It is possible to provide a semiconductor manufacturing apparatus capable of positioning the rotation center of the spin chuck and the wafer chuck with high accuracy.

According to claims 20 and 21 of the present invention,
A semiconductor manufacturing apparatus capable of preventing contamination of a semiconductor wafer without accumulating dust generated when the semiconductor wafer is held in a clamp pin can be provided.

[0475] According to claims 22 and 23 of the present invention,
An inexpensive semiconductor manufacturing apparatus can be provided without taking up space even if the transport distance or the transport direction changes.

According to the twenty-fourth, twenty-fifth, and twenty-sixth aspects of the present invention, it is possible to provide a semiconductor manufacturing apparatus capable of moving an object to be processed by a simple mechanism so as to be capable of continuous processing of semiconductor manufacturing.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a first embodiment of a semiconductor manufacturing apparatus according to the present invention.

FIG. 2 is a specific configuration diagram of an image processing unit in the apparatus.

FIG. 3 is a diagram showing a relationship between input and output signals of an autocorrelation calculation processing unit.

FIG. 4 is a diagram showing scan lines for alignment marks.

FIG. 5 is a signal waveform diagram when an alignment mark is scanned.

FIG. 6 is a diagram showing an example of a waveform of an original signal when an alignment mark is scanned.

FIG. 7 is a waveform diagram of an autocorrelation calculation process on an original signal.

FIG. 8 is a waveform diagram of an autocorrelation calculation process on an original signal.

FIG. 9 is a waveform chart after an autocorrelation calculation process.

FIG. 10 is a configuration diagram of a pattern drawing apparatus according to a second embodiment to which the semiconductor manufacturing method according to the present invention is applied.

FIG. 11 is a flowchart of mark position detection of the apparatus.

FIG. 12 is a schematic view showing an alignment mark detection operation.

FIG. 13 is a configuration diagram of a wafer alignment apparatus according to a third embodiment to which the semiconductor manufacturing method according to the present invention is applied.

FIG. 14 is a flowchart of mark position detection of the apparatus.

FIG. 15 is a configuration diagram of a wafer alignment apparatus according to a fourth embodiment to which a semiconductor manufacturing method according to the present invention is applied.

FIG. 16 is a layout diagram of each camera detection system.

FIG. 17 is a diagram showing measurement points of each camera detection system.

FIG. 18 is a diagram showing a signal profile of a first derivative in a focusing operation.

FIG. 19 is a waveform diagram of a quadratic curve approximation equation based on a variance of a signal profile of a first derivative.

FIG. 20 is a diagram showing experimental data of a quadratic curve approximation equation.

FIG. 21 is a diagram showing experimental data of a quadratic curve approximation equation.

FIG. 22 is a diagram showing the variance of the signal profile of the first derivative.

FIG. 23 is a diagram showing the variance of the signal profile of the first derivative.

FIG. 24 is a configuration diagram showing a fifth embodiment of the semiconductor manufacturing apparatus according to the present invention.

FIG. 25 is a diagram showing an operation of aligning the rotation centers of a wafer chuck and a spin chuck.

FIG. 26 is a diagram showing detection of a protrusion by a position sensor.

FIG. 27 is a configuration diagram showing a modified example of the device.

FIG. 28 is a configuration diagram showing a modification of the same device.

FIG. 29 is a configuration diagram of a substrate holding device according to a sixth embodiment of the semiconductor manufacturing apparatus according to the present invention.

FIG. 30 is a view showing an arc movement of a wafer clamp pin in the same apparatus.

FIG. 31 is a configuration diagram of a carrier transport device of a semiconductor manufacturing apparatus according to a seventh embodiment of the present invention.

FIG. 32 is a configuration diagram of a carrier transport device connected in two rows in a semiconductor manufacturing apparatus according to the present invention.

FIG. 33 is a configuration diagram of a carrier transport device of an eighth embodiment of the semiconductor manufacturing apparatus according to the present invention.

FIG. 34 is a configuration diagram of a transport direction changing mechanism in the apparatus.

FIG. 35 is a configuration diagram of a multi-chamber system of a ninth embodiment of a semiconductor manufacturing apparatus according to the present invention.

FIG. 36 is a configuration diagram of the arm standby chamber in the same system as viewed from the side.

FIG. 37 is a configuration diagram of a multi-chamber system of a tenth embodiment of a semiconductor manufacturing apparatus according to the present invention.

FIG. 38 is a configuration diagram of a multi-chamber system according to an eleventh embodiment of the semiconductor manufacturing apparatus according to the present invention.

FIG. 39 is a configuration diagram of a conventional wafer alignment apparatus.

FIG. 40 is an alignment flowchart of the apparatus.

FIG. 41 is a schematic view showing an alignment operation of the apparatus.

FIG. 42 is a configuration diagram of an automatic focus position detection system using an air micrometer.

FIG. 43 is a configuration diagram of an optical automatic focus position detection system.

FIG. 44 is a configuration diagram of an automatic focus position detection system using a focus sensor.

FIG. 45 is a configuration diagram of an automatic focus position detection system by image processing.

FIG. 46 is a configuration diagram of a semiconductor manufacturing apparatus provided with a spin processing device.

FIG. 47 is a configuration diagram of a spin processing apparatus that performs cleaning.

FIG. 48 is a configuration diagram of a carrier transport device.

FIG. 49 is a configuration diagram when two carrier transport devices are connected in series.

FIG. 50 is a configuration diagram when two carrier transport devices are connected in a right angle direction.

FIG. 51 is a configuration diagram of a multi-chamber system.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Semiconductor wafer, 3 ... theta stage, 6 ... XYZ stage, 7,8 ... Camera detection system, 9,10 ... CCD, 47 ...
Spin chuck, 48 Spin motor, 50 Wafer chuck, 55 Cam plate, 57 Lever, 58 Wafer clamp pin, 62 Carrier, 104 Stage control unit, 107 Camera controller, 108 Image input unit (image Memory), 109 image processing unit, 110 correlation parameter storage unit, 111 autocorrelation calculation processing unit, 112
... Mark position coordinate calculation processing unit, 120.
XY stage, 124 electron gun, 126 first forming aperture, 127 second forming aperture, 128 electromagnetic projection lens, 130 electromagnetic objective lens, 132 molding deflector, 133 main deflector, 134 sub Deflectors, 135 ...
Main controller, 137 ... Pattern data generator, 138
... Beam control device, 139 electronic detector, 140 mark detection device, 143 XY stage control device, 144
Position correction device, 145 ... correction function coefficient storage unit, 146 ...
Position correction unit, 150: position correction device, 151: correction function coefficient storage unit, 152: position correction unit, 161: image processing unit, 170: dummy wafer, 172: position detection jig, 1
73: Projection, 174: Sensor position adjustment mechanism, 175: Position sensor, 176: Voltmeter, 177: Center detection unit, 17
8. Controller, 190 Contact position switching means, 191
... Drive source, 200 ... Stand, 201 ... Carrier stay
202: Carrier guide, 203: Carrier transport drive, 204: Delivery position, 205: Carrier vertical drive, 206: Carrier transport table, 207
... Carrier transport control device, 210 ... Transport direction changing mechanism,
211: first carrier transport driving device, 212: second carrier transport driving device, 213: carrier rotation driving device, 220: loading chamber, 221, 222 ...
Process chamber, 223, 224: arm standby chamber, 225: robot arm, 227: valve.

Continuing from the front page (72) Inventor Katsutoshi Higuchi 1 Koga Toshiba-cho, Saiwai-ku, Kawasaki City, Kanagawa Prefecture Inside the Toshiba Tamagawa Plant (72) Inventor Hiroshi Nishimura 33-33 Shinisogo-cho, Isogo-ku, Yokohama-shi, Kanagawa Prefecture (72) Inventor Takao Ino, 33, Shinisogocho, Isogo-ku, Yokohama-shi, Kanagawa Prefecture, Japan

Claims (26)

[Claims] 1. A semiconductor manufacturing method having a function of detecting an alignment mark formed on a target object when positioning the target object at a predetermined position. Performing a self-correlation calculation using the width of the alignment mark as a parameter, and obtaining position coordinates of the alignment mark from a signal waveform resulting from the calculation. 2. A semiconductor manufacturing apparatus having a function of detecting an alignment mark formed on a target object when positioning the target object at a predetermined position, comprising: an imaging unit configured to image the alignment mark; Autocorrelation calculation processing means for performing an autocorrelation calculation on the image signal output from the imaging means using the alignment mark width as a parameter; and a position of the alignment mark from a signal waveform of a calculation processing result of the autocorrelation calculation processing means. And a mark position coordinate processing means for obtaining coordinates. 3. A semiconductor manufacturing apparatus having a function of detecting an alignment mark formed on an object to be processed and positioning the object at a predetermined position based on the position of the alignment mark. An autocorrelation calculation processing means for performing an autocorrelation calculation using the alignment mark width as a parameter on an image signal output from the imaging means; a signal waveform of a calculation processing result of the autocorrelation calculation processing means A mark position coordinate calculating means for calculating the position coordinates of the alignment mark from the image processing apparatus; and moving the object based on the position coordinates obtained by the mark position coordinate calculating means to adjust the positional deviation of the object. A semiconductor manufacturing apparatus, comprising: 4. A semiconductor manufacturing method having a function of detecting an alignment mark formed on an object to be processed when positioning the object to be processed at a predetermined position. A method of manufacturing a semiconductor device, comprising: predicting a position of the alignment mark by using an autoregressive model obtained based on a past displacement amount, and detecting the position of the alignment mark according to the predicted position of the alignment mark. 5. The method according to claim 1, wherein a parameter of the auto-regression model is obtained in advance based on a past positional shift amount of the object, and a design position of the alignment mark is corrected using the auto-regression model of the parameter. Item 5. The semiconductor manufacturing method according to Item 4. 6. A learning function for storing past displacement amounts of the object to be processed and for changing parameters of the autoregressive model to optimal values based on the displacement amounts. The semiconductor manufacturing method according to the above. 7. The past displacement amount of the object to be processed is stored, an average value and a deviation of these displacement amounts are obtained, and when the average value and the deviation exceed a predetermined threshold value, the auto-regression is performed. 7. The method according to claim 6, wherein parameters of the model are changed. 8. A semiconductor manufacturing apparatus for detecting an alignment mark formed on a processing object when positioning the processing object at a predetermined position, the semiconductor manufacturing apparatus comprising: Position correction means for predicting the position of the alignment mark using the obtained autoregressive model, and mark detection means for detecting the alignment mark in accordance with the position of the alignment mark predicted by the position correction means. A semiconductor manufacturing apparatus characterized by the above-mentioned. 9. A semiconductor manufacturing apparatus which scans a target object with an electron beam to draw on the target object, wherein the self-produced object is obtained based on a past positional deviation amount of the target object with respect to a predetermined position. Position correction means for predicting the position of the alignment mark by a regression model; and a mark for obtaining the actual position of the alignment mark by scanning the object with the electron beam in accordance with the alignment mark position predicted by the position correction means. Detecting means; and a drawing correcting means for obtaining a positional shift amount of the object from an actual position of the alignment mark detected by the mark detecting means, and correcting an irradiation position of the electron beam based on the positional shift amount. And a semiconductor manufacturing apparatus comprising: 10. A semiconductor manufacturing apparatus for performing a predetermined process on a processing target housed in a vacuum chamber, wherein an autoregressive model obtained based on a past positional deviation amount of the processing target from a predetermined position. Position correcting means for estimating the position of the alignment mark, imaging means for imaging the alignment mark, and relative positions of the imaging means and the object to be processed according to the alignment mark position predicted by the position correcting means. A moving mechanism for moving the object, and after moving by the moving mechanism, an actual position of the alignment mark is obtained from image data obtained by imaging by the imaging means, and a positional deviation amount of the object to be processed is obtained from this position. And a positioning means for positioning the object to be processed at a predetermined position. . 11. A semiconductor manufacturing method for imaging an alignment mark formed on the object to detect the position of the alignment mark when positioning the object at a predetermined position. Is set at a plurality of locations, the alignment mark is imaged at each location, and each mark image data is processed at each of the plurality of intervals to obtain variance values, and the imaging device is obtained from an approximate curve obtained from these variance values. A method of manufacturing a semiconductor device, comprising: obtaining a focal position of a semiconductor device. 12. A method of processing each mark image data at a plurality of intervals to obtain each variance value, and a position corresponding to a maximum value in an approximate curve obtained by these variance values is set as a true focal position. The semiconductor manufacturing method according to claim 11, wherein 13. When imaging the alignment mark by setting a plurality of intervals with the object to be processed, the imaging position of the alignment mark is set to a plurality of positions along a predetermined direction on the object to be processed. The method according to claim 11, wherein
The semiconductor manufacturing method according to the above. 14. A semiconductor manufacturing apparatus for imaging an alignment mark formed on the object to detect the position of the alignment mark when positioning the object to be processed at a predetermined position. Imaging means for setting an interval at a plurality of locations and imaging the alignment mark formed on the object at each of the plurality of intervals; and processing each mark image data at each interval obtained by imaging by the imaging apparatus. And a focusing means for determining each variance value and calculating a focal position of the imaging device from an approximate curve obtained from the variance values. 15. The focusing means processes each mark image data at a plurality of intervals to obtain each variance, and determines a true focal position corresponding to a maximum value in an approximate curve obtained by these variances. The semiconductor manufacturing apparatus according to claim 14, wherein: 16. When imaging the alignment mark by setting an interval between the object to be processed and the imaging means at a plurality of positions, a plurality of imaging positions of the alignment mark are set along a predetermined direction on the object to be processed. 15. The semiconductor manufacturing apparatus according to claim 14, wherein the position is set at a location. 17. An object to be processed held by an object to be processed chuck is positioned and mounted on a spin chuck, and
A semiconductor manufacturing apparatus that performs a predetermined process on the object by rotating the spin chuck body at a high speed; a sensor having a spot-shaped detection region with respect to a rotation center of the spin chuck; A dummy processing body in which a positioning portion of the processing target is formed at a portion corresponding to a rotation center of the chuck; And a positioning means for positioning the object chuck. 18. The positioning unit according to claim 17, wherein the positioning unit is a projection formed at the center of the positioning processing body.
The semiconductor manufacturing apparatus according to the above. 19. The semiconductor according to claim 17, wherein the positioning section comprises a hole formed at the center of the positioning processing body and a signal source provided at a center position of the spin chuck body. Manufacturing equipment. 20. A semiconductor manufacturing apparatus which mounts an object to be processed on a spin chuck, presses the object to be processed from the side by a clamp pin, and rotates the spin chuck at a high speed to perform processing on the object. A semiconductor manufacturing apparatus comprising: contact position switching means for switching a contact position of a clamp pin with respect to the object to be processed. 21. The semiconductor manufacturing apparatus according to claim 20, wherein said contact position switching means moves said clamp pin in an arc. 22. A semiconductor manufacturing apparatus, comprising: a transport device for sequentially transporting a plurality of workpieces in a predetermined direction by moving a transport table on which workpieces are placed by a transport driving device; A position where the transfer length of the transfer drive device is formed to be longer than the transfer unit body, and the transfer table in the transfer device can transfer the transfer table to the transfer drive device of another connected transfer device. Wherein a plurality of the transport units are connected in series to each other. 23. A semiconductor manufacturing apparatus comprising: a transport device for sequentially transporting a plurality of workpieces in a predetermined direction by moving a transport table on which workpieces are mounted by a transport driving device; The transfer length of the transfer drive device is formed to be longer than the transfer device main body, and a transfer direction changing mechanism that changes at least the transfer direction of the object to be processed is disposed between the transfer devices. A semiconductor manufacturing apparatus, wherein the semiconductor manufacturing apparatus is arranged at a predetermined angle. 24. A semiconductor manufacturing apparatus provided with a plurality of processing chambers for performing processing on an object to be processed, wherein the semiconductor manufacturing apparatus is rotatably provided between the processing chambers, and holds the object by this rotating operation. An arm mechanism for transporting between the processing chambers, a plurality of opening / closing mechanisms for opening and closing the processing chambers for each of the processing chambers in response to a transfer operation of the object to be processed between the processing chambers by the arm mechanism, A semiconductor manufacturing apparatus comprising: 25. The semiconductor manufacturing apparatus according to claim 24, wherein the arm mechanism is provided in an arm standby chamber formed between the processing chambers. 26. A plurality of processing chambers for processing an object to be processed arranged in a circle, an arm standby chamber formed in a central portion of the processing chamber, and a rotatably provided in the arm standby chamber. An arm mechanism for holding the object to be processed by the rotation operation and transporting the object between the processing chambers; and transferring the object to be processed between the processing chambers by the arm mechanism. And a plurality of opening and closing mechanisms for opening and closing the processing chamber for each process.