JP4406419B2 - Pattern formation method - Google Patents

Description

  The present invention relates to a pattern forming method for forming a resist pattern on a substrate to be processed in photolithography for manufacturing a semiconductor device, and more particularly, a pattern for determining a development end point while monitoring a monitor area during resist development. It relates to a forming method.

  As a method of reducing dimensional variation between wafers, there is a method of controlling the development time from the monitoring result of a monitor area provided separately from the device pattern at the time of development (see, for example, Patent Document 1). In this method, since the monitor area arranged on a specific chip on the wafer is monitored, if the monitored chip has a specific condition for some reason, the size of the monitored chip is desired. Although the value is finished, there is a problem that the average dimension of the wafer is greatly shifted.

  In order to solve this problem, the present inventors monitor a latent image before development, extract an average chip, monitor the monitor pattern of the chip during development, and finish development in an optimal development time. A method (Patent Document 2) has already been proposed. However, with this method, sufficient accuracy cannot be obtained when the thickness of the underlying film varies. In addition, since monitoring is performed as a change amount of the latent image, when the monitor area is a pattern that is affected by the focus variation (for example, the device pattern itself), if both the exposure amount and the focus vary, a typical chip Could not be obtained with high accuracy.

On the other hand, there is (Patent Document 3) as a method that can be monitored during development even when there is a change in the base film thickness. In this method, the development time and the exposure amount are finally obtained from specific points such as extreme values of intensity change. However, in recent years, with the miniaturization of the pattern size, the singular point of the pattern intensity change changes not only by the exposure amount but also by the focus fluctuation, so that an optimum development time can be obtained from one piece of information called the singular point. I can't.
JP-A-10-300428 Japanese Patent Application No.11-273212 Patent No. 2818689

  As described above, in the conventional method in which the development time is controlled from the monitoring result of the monitor area provided separately from the device pattern, the average of the wafer is measured when the monitored chip has a specific condition for some reason. There is a problem that the typical dimensions are greatly shifted. In addition, the method of extracting an average chip by monitoring a latent image cannot obtain sufficient accuracy when the film thickness of the underlayer varies. Further, in the case of a pattern in which the monitor area is affected by the focus variation, there is a problem that a representative chip cannot be obtained with high accuracy if both the exposure amount and the focus are varied. In addition, with the miniaturization of the pattern size, the optimum development time cannot be obtained from one piece of information that is a singular point of the reflected light intensity change of the pattern.

  The present invention has been made in consideration of the above circumstances, and the object of the present invention is to obtain a representative chip on a wafer with high accuracy and to obtain a development end point of the chip with high accuracy. Another object of the present invention is to provide a pattern forming method capable of improving dimensional accuracy and yield.

(Constitution)
In order to solve the above problems, the present invention adopts the following configuration.

  That is, the present invention relates to a pattern forming method in which a device pattern is exposed to a resist on a substrate to be processed, and thereafter a device pattern is formed on the resist by development. Irradiating the monitor region with light of a plurality of wavelengths during development of the resist, and predicting a development stop time during which the device pattern is finished to a desired dimension based on wavelength dispersion of reflected light intensity reflected from the monitor region And a step of supplying a development stop solution onto the substrate during the predicted development stop time to stop the development.

  According to another aspect of the present invention, there is provided a pattern forming method of exposing a device pattern to a resist on a substrate to be processed, and then forming the device pattern on the resist by development. When developing the resist, the monitor region is irradiated with light having a plurality of wavelengths, the intensity change of the reflected light reflected from the monitor region is converted into a phase change, and the device pattern has a desired dimension based on the converted phase. And a step of supplying a development stop solution onto the substrate during the predicted development stop time to stop the development.

  Here, preferred embodiments of the present invention include the following.

  (1-1) The monitor area is the memory cell area in the device pattern.

  (1-2) The monitor area should be arranged at a position showing the average flatness in the chip based on the flatness information at the time of leveling in the exposure apparatus.

  (2-1) When extracting the chip, the deviation between the focus value in the monitor area and the set focus value is read based on the information on the flatness at the time of leveling in the exposure apparatus, and the deviation value is equal to or less than a predetermined value. Extracting chips.

  (2-2) When extracting chips, extract chips that have average conditions based on the results of measuring the reflected light intensity of the monitor area for multiple wavelengths for multiple chips. To do.

  (2-3) When extracting chips, the chips that are average conditions are extracted based on the result of measuring the latent image depth in the monitor area for a plurality of chips.

  (3-1) The process of predicting the development stop time involves irradiating the monitor area with light of a plurality of wavelengths, and predicting the development stop time from the time that is a singular point of the intensity change of the reflected light reflected from the monitor area. To do.

  (3-2) The process of predicting the development stop time is obtained by irradiating the monitor area with light of a plurality of wavelengths and using the exposure apparatus as a singular point of intensity change of the reflected light reflected from the monitor area. The development stop time is predicted from the focus value of the monitor area.

  (3-3) The process of predicting the development stop time is to irradiate the monitor area with light of a plurality of wavelengths, and predict the development stop time from the reflectance at a specific time of the intensity change of the reflected light reflected from the monitor area. To do.

  (3-4) The step of predicting the development stop time is obtained by irradiating the monitor area with light of a plurality of wavelengths, and the reflectance at a specific time of the intensity change of the reflected light reflected from the monitor area and the exposure apparatus. The development stop time is predicted from the focus value of the monitor area.

  (3-5) The process of predicting the development stop time involves irradiating the monitor area with light of a plurality of wavelengths, and the reflectivity and the intensity of the reflected light at a specific time of the intensity change of the reflected light reflected from the monitor area. Predict the development stop time from the time that is the singular point of change.

  (4) The specific time for the intensity change of the reflected light reflected from the monitor area and the time that becomes the singular point for the intensity change of the reflected light are set in different time zones during development.

  (5-1) The process of predicting the development stop time irradiates the monitor area with light of a plurality of wavelengths, converts the intensity change of the reflected light reflected from the monitor area into a phase change, and the phase becomes a desired value. When this happens, the development stop time is taken.

  (5-2) The step of predicting the development stop time irradiates the monitor region with light of a plurality of wavelengths, converts the intensity change of reflected light reflected from the monitor region into a phase change, Estimate development stop time.

  (5-3) The step of predicting the development stop time irradiates the monitor region with light of a plurality of wavelengths, converts the intensity change of the reflected light reflected from the monitor region into a phase change, To predict development stop time.

  (5-4) The step of predicting the development stop time irradiates the monitor area with light of a plurality of wavelengths, converts the intensity change of the reflected light reflected from the monitor area into a phase change, The development stop time is predicted from the time when the phase becomes zero.

(Function)
As described above, with the recent miniaturization of processing dimensions, precise dimensional control is required even between wafers. The present invention is characterized by the following three points in addition to controlling the development time from the result of measuring the intensity of reflected light from the monitor region in order to reduce the variation in dimensions between wafers.

  First, in order to avoid monitoring chips that have special conditions, chips that are processed under typical conditions in the wafer are extracted before the development process. At this time, if there is a high possibility that the focus is fluctuated by the chip, a chip whose focus is close to the set amount is extracted based on the focus information in the exposure apparatus. Thereafter, a chip having an average exposure amount condition is extracted from the extracted chips. In this case, for example, the reflected light intensity is measured over multiple wavelengths from the monitor region, and the chips processed under average conditions are extracted by measuring the latent image depth for a plurality of chips. Next, the intensity of reflected light from the monitor area of the extracted chip is measured, and the development time is controlled. Thereby, dimensional variations between wafers can be suppressed, and dimensional accuracy and yield can be improved.

  Secondly, in order to predict the development time, the monitor region is irradiated with light of a plurality of wavelengths, and the development stop time at which the device pattern is finished to a desired dimension is predicted based on the reflected light intensity change of the monitor region. That is, the development time is controlled from the result of measuring the reflected light intensity from the monitor region over multiple wavelengths. By performing measurement at multiple wavelengths in this way, it is possible to improve the accuracy of each measurement.

  Third, in order to predict the development time, the monitor region is irradiated with light of a plurality of wavelengths, the intensity change of the reflected light reflected from the monitor region is converted into a phase change, and the device is based on the converted phase. Predict the development stop time when the pattern is finished to the desired dimension. That is, by focusing on the phase rather than the intensity change of the reflected light, the development stop time can be expressed as a general amount independent of the wavelength. In this case, complicated processing is eliminated and the analysis can be simplified.

  According to the present invention, when developing time control is performed in order to reduce the dimensional variation between wafers, a chip processed under typical conditions in a wafer is subjected to focus information and a monitor area in the exposure apparatus. It is extracted by measuring the film thickness. Then, development is started, and by monitoring the intensity change in the monitor area of this chip, it is possible to obtain the optimum development time. Thereby, the dimensional variation between wafers can be greatly reduced, and the dimensional accuracy and yield are greatly improved.

  Also, when determining the optimum development time, know both the exposure amount and the focus amount during exposure by measuring the singularity time and the reflectance at a specific development time from the wavelength dispersion with respect to the development time. And an optimum development time can be obtained. In addition, by replacing the intensity with the phase, it is possible to monitor with physical quantities that do not depend on the film thickness of the base or the wavelength of the probe light, simplifying the analysis by eliminating the complicated process of obtaining dimensions from wavelength dispersion. Can do.

  The details of the present invention will be described below with reference to the illustrated embodiments.

(First embodiment)
(Constitution)
FIG. 1 is a schematic configuration diagram showing a developing device used in the first embodiment of the present invention. In the actual lithography process, there is a film structure such as an oxide film below the resist and the antireflection film. However, in this embodiment, the case where the lower layer is Si is shown for simple explanation. Therefore, since the wafer is in a very flat state, the chip close to the center of the wafer is exposed with the focus amount set by the exposure apparatus.

  In addition to the units necessary for performing normal development to rinsing, a monitor head 11 for monitoring is disposed in the developing unit 10. The monitor head 11 has a fiber 13 for introducing light from a halogen lamp as a light source 12 of probe light (a light source having a broad wavelength dispersion in this range is preferable in order to perform measurement in a wavelength range of 400 to 800 nm). The wafer 20 can be illuminated by a lens in the monitor head 11.

  The reflected light from the wafer 20 is divided into two parts, a CCD camera 31 and a spectroscope 32 by a half mirror in the monitor head 11. Of the divided light, the light focused on the CCD camera 31 is analyzed by the image analysis unit 33 and used for position detection. Further, the light introduced into the spectroscope 32 is analyzed by the intensity analysis unit 34 and used for obtaining the wavelength dispersion of the reflected light intensity.

  The monitor head 11 can scan the wafer 20 in the X and Y directions by a signal from the control unit 35 so that the monitor area arranged in each chip in the wafer 20 can be monitored. Based on the shot map, the monitor area of a specific chip is moved to a position where it can be detected. Then, an image is acquired by the CCD camera 31 and sent to the image analysis unit 33 to determine whether the region measured by the spectroscope 32 matches the monitor region (memory cell region). The monitor head 11 is moved to match.

  Here, as a reference image element for determining whether or not they match, a resist latent image of the currently processed layer can be selected. However, since the latent image has a low contrast, if this is used as a reference image, a pattern is obtained. There is a possibility that the accuracy of matching is lowered or pattern matching cannot be performed. In such a case, it is also possible to select the arrangement information of the already processed resist lower layer pattern as reference information for pattern position detection. In this case, from the information on where the monitor pattern of the upper resist pattern is arranged in the lower layer pattern and the pattern matching result using the arrangement information of the lower pattern as reference information, the monitor pattern of the resist pattern Detect position.

  Each chip on the wafer 20 has a mark region 22 arranged in the periphery, and a plurality of memory cell regions 23 (200 nm L / S patterns) are arranged in a device region 21 inside the chip region 22. The monitor area is the memory cell area 23 of the device area 21.

  In the present embodiment, the developing device is used to monitor the latent image before development and the latent image during development, and measure the wavelength dispersion of the reflected light intensity in the monitor region.

(Function)
A development sequence in the present embodiment is shown in FIG. First, when the wafer 20 is transferred to the developing unit 10 (step S1), the position of the notch is detected and the position of the wafer 20 in the rotational direction is adjusted (step S2). Thereafter, the monitor head 11 moves to the monitor area of the specific chip (i) based on the shot map at the time of exposure and the layout of the exposure mask (step S3). Here, as shown in FIG. 4, nine chips near the center are to be monitored. However, if a chip that is out of focus at the time of exposure is monitored, the accuracy of extracting a representative chip decreases. To do. Therefore, based on the information from the exposure apparatus, the chip whose focus condition is shifted is removed from the monitor target (step S4).

  A specific method will be described next. When exposing the chip during exposure, the exposure apparatus sets the focus value while performing leveling. In other words, the focus is on the average of the flatness of the wafer. Therefore, it is desirable that the monitor area is at its average location. Since there is a high possibility that the average location is near the center of the chip, the location of the memory cell region (4) in the device region 21 of FIG. Further, it is judged from the reading of the interferometer of the exposure apparatus whether the flatness of the monitor area (4) in the chips 1 to 9 is an average level, and the chip that is greatly deviated is monitored for the latent image. And not. In this embodiment, among the nine chips, chip 1 and chip 5 were not at an average level, and thus were removed from the chips to be monitored. Of course, this step may be omitted if the focus does not shift.

  Note that in a general exposure apparatus, the pattern shape is exposed to detect the shape of the surface in the direction orthogonal to the scan, and exposure is performed while changing the focus in accordance with the average position. There is a focus measurement point in the middle, and so-called leveling is performed in which the focus is adjusted for each point. This leveling information is stored in the exposure apparatus. Therefore, by referring to the leveling information stored in the exposure apparatus, it is possible to know the focus condition during exposure of each chip.

  Next, the pattern positioning (step S5) method will be described. The observation area after the head movement in S3 is as shown in FIG. The optical system of the monitor head 11 is adjusted so that a region 52 to be detected by a spectroscope enters the center of the field of view 51 of the CCD camera. Therefore, if monitoring is performed in the state of FIG. 3A, the intensity of reflected light from other than the monitor region 53 is also detected. Therefore, after the monitor head 11 is moved, a deviation amount (Δx, Δy) between the center of the monitor region 53 and the center of the region 52 detected by the spectroscope is obtained and monitored at a position as shown in FIG. The head 11 is moved. As a result, the wavelength dispersion of the reflected light intensity from only the monitor region 53 can be detected by the spectrometer 32.

  Here, an example in which the position adjustment is performed using the CCD camera 31 has been described. However, if the stage accuracy of the monitor head 11 is sufficient, once the position adjustment is performed, it is not necessary to adjust thereafter. If the illumination light is a substantially monochromatic light with a narrow band filter, the intensity is not measured by the spectroscope 32, but the position of the CCD camera 31 is detected by pattern matching and the CCD floor is used. The tone may be the reflected light intensity.

  Next, the film thickness of the monitor area is measured (step S6). The film thickness of the monitor region is obtained from the chromatic dispersion of the measured reflected light intensity. FIG. 5 shows an example of wavelength dispersion of the measured reflected light intensity. Since the reflected light intensity is determined by the film thickness and optical constant of the formed film, the film thickness can be measured from the wavelength dispersion as shown in FIG. Here, the film reduction amount of the monitor area to be measured is related to the exposure amount as shown in FIG. Therefore, if the amount of film reduction in the monitor region is measured, it can be determined how much exposure the chip (i) being measured is exposed. Here, the method for obtaining the film reduction amount in the monitor region from the chromatic dispersion is shown, but when the lower layer of the antireflection film is flat, the film loss amount in the monitor region is obtained from the reflected light intensity of a specific wavelength. Also good.

  The measurement result of the film thickness in the monitor area is shown in FIG. As described above, chip 1 and chip 5 are excluded from measurement because the focus conditions are deviated. Since the average value is 11.5 nm from the result of this measurement, the chip representing the wafer is set to chip 2 (step S8), and the monitor area of this chip is measured during development.

  When the chip to be measured is determined, development is started (step S9). After the developer is deposited, the monitor head 11 moves to the chip determined in S8 (here, chip 2), adjusts the position, and starts monitoring (step S10). In the monitoring during development, the wavelength dispersion of the reflected light intensity from the monitor region is measured for each time as in the case of measuring the film thickness of the latent image.

  A method for determining the end of development (step S11) will be described based on the waveform during actual development. At the same time, changes in reflected light intensity at a number of wavelengths are measured, of which a description will be given based on the waveform at a wavelength of 550 nm. FIG. 8 shows the relationship (experimental results) between the development time at a wavelength of 550 nm and the reflectance of the monitor area. From this reflectance change, an interference waveform is observed in the early stage of development, and a gradual intensity change is observed in the latter half.

  In order to explain this, FIG. 9 shows the result of theoretically determining this change. In this calculation (FIG. 9), in the first 10 seconds, the area of the exposed portion (extraction size 190 nm) is reduced, the film is completely removed in 10 seconds, and then the development in the lateral direction is performed at a rate of 2 nm in 10 seconds. The process is performed under the condition that the desired dimension (remaining dimension 200 nm) is reached in 60 seconds. That is, in FIG. 8, the film thickness of the exposed portion is reduced in the initial stage, and the second half represents the change in the horizontal dimension.

  Further, FIG. 10 shows the result of determining the relationship of the change in reflectance (remaining dimension 210 nm → 190 nm) with respect to the lateral dimension change in the latter half. From this result, it is understood that the reflectance is calculated for several wavelengths, the pattern dimension is obtained, and the time when the dimension reaches a predetermined value may be set as the development end point. However, in FIG. 10, since the amount of change in the reflectance with respect to the dimension is very small, it is likely that the process will take time to accurately obtain the dimension from the reflectance calculated for several wavelengths.

  In this embodiment, since the lower layer of the antireflection film is made of Si, the reflectance is not affected by the lower layer. However, since the lower layer is actually an oxide film, the reflection is caused by a change in the thickness of the oxide film. Since the rate changes greatly, there is a high possibility that processing will take more time. For this reason, the optimum development time is obtained from the time that is a singular point of intensity change and the reflectance at a specific development time. Here, the two quantities are measured because there are two physical quantities, exposure amount and focus, as factors determining the dimensions. Of course, when the focus value is known by the exposure apparatus, either the singular point of intensity change or the reflectance at a predetermined time may be measured. Usually, the singular point of intensity change has higher measurement accuracy. In such a case, the singular point of intensity change is measured.

  In the present embodiment, since the lower layer of the antireflection film is Si, a chip that does not meet the focus condition is removed from the monitor target based on information from the exposure apparatus (step S4). Therefore, since it is known that the focus is achieved, the optimum development time is obtained from the singular point of the intensity change.

  Next, a method for calculating the time to be a singular point will be described. In order to detect a singular point in FIG. 8, the differential value of the intensity change is obtained, and the time when the differential value becomes a certain value or less may be obtained. In this embodiment, since the lower layer of the antireflection film is known as Si, a wavelength suitable for obtaining a singular point can be obtained in advance. Here, for this reason, the singular point was obtained using a wavelength of 550 nm. However, since the lower layer is actually an oxide film, the wavelength suitable for obtaining the singular point may not be known in advance due to the film thickness variation.

  In that case, reflectivity measurement is performed for a plurality of wavelengths, and after extracting a wavelength having a large change in slope near a singular point, a time during which the differential value of the intensity change is less than a certain value may be obtained. . It is also possible to select a plurality of wavelengths having a large change in slope near the singular point, obtain a singular point from each wavelength, and improve the accuracy of obtaining the singular point. In this case, the number of wavelengths that can be analyzed is determined by the processing time.

  FIG. 11 shows the relationship between the exposure amount and the time that becomes a singular point when the focus fluctuates. In this embodiment, it is known that the focus is just focused in S4, and it can be seen from the intensity change in FIG. 8 that the time that becomes a singular point is 10 seconds. Therefore, the development time is 54 seconds from the relationship in FIG. Desired. Therefore, the development is completed in 54 seconds (step S12).

  In this embodiment, the chip 2 representing the wafer is extracted (S8), and the development stop time is obtained by monitoring the chip during development (S11). The latent image of the extracted chip is monitored and developed. It is also possible to determine the stop time. For example, in the chip 2, as a result of monitoring the latent image, the film thickness is obtained as 11.5 nm (FIG. 7), and the exposure amount can be obtained from this result and FIG. Furthermore, the development stop time can be obtained based on the relationship shown in FIG.

(effect)
As described above, in this embodiment, when developing time control is performed in order to reduce the dimensional variation between wafers, the chips processed under typical conditions in the wafer are subjected to focus information and monitoring in the exposure apparatus. Extraction is performed by measuring the film thickness of the region. Then, by monitoring the intensity change in the monitor area of this chip during development, it is possible to obtain the optimum development time. Thereby, the dimensional variation between wafers can be greatly reduced, and the dimensional accuracy and yield are greatly improved.

(Second Embodiment)
(Constitution)
The developing device used in this embodiment is the same as that of the first embodiment, as shown in FIG. In an actual lithography process, there is a film structure such as an oxide film below the resist and the antireflection film. In this embodiment, the lower layer is an oxide film (thickness target is 300 nm). Therefore, the flatness of the wafer is not as high as in the first embodiment, and even a chip near the center of the wafer may not be exposed with the focus amount set by the exposure apparatus.

  Also in the present embodiment, the latent image before development and monitoring during development are performed, and the wavelength dispersion of the reflected light intensity in the monitor region is measured.

(Function)
The development sequence in this embodiment is basically the same as that in the first embodiment, as shown in FIG. Steps S1 to S5 are exactly the same as those in the first embodiment, and a description thereof will be omitted. However, in the present embodiment, among S9 chips, chip2, chip5, and chip9 are not at an average level in S4, so these are excluded from the chips to be monitored.

  When measuring the film thickness of the monitor region in step S6, the film thickness of the monitor region is obtained from the chromatic dispersion of the measured reflected light intensity. FIG. 13 shows an example of wavelength dispersion of the measured reflected light intensity. Since the reflected light intensity is determined by the film thickness and the optical constant of the constituent film, the film thickness can be measured from the wavelength dispersion as shown in FIG. The film reduction amount of the monitor area measured here is related to the exposure amount as shown in FIG. Therefore, if the amount of film reduction in the monitor region is measured, it can be determined how much exposure the chip (i) being measured is exposed. Here, the method for obtaining the film reduction amount in the monitor region from the chromatic dispersion is shown, but when the lower layer of the antireflection film is flat, the film loss amount in the monitor region is obtained from the reflected light intensity of a specific wavelength. Also good.

  The measurement results of the film thickness in the monitor area are shown in FIG. As described above, chip2, chip5, and chip9 are excluded from measurement because the focus conditions are shifted. Since the average value is 11.4 nm from the result of this measurement, the chip representing the wafer is set to chip 1 (step S8), and the monitor area of this chip is measured during development.

  When the chip to be measured is determined, development is started (step S9). After the developer is deposited, the monitor head 11 moves to the chip determined in S8 (here, chip 1), adjusts the position, and starts monitoring (step S10). In the monitoring during development, the wavelength dispersion of the reflected light intensity from the monitor region is measured for each time as in the case of measuring the film thickness of the latent image.

  A method for determining the end of development (step S11) will be described based on the calculation result of the waveform during development. At the same time, the reflected light intensity changes at a number of wavelengths are measured, of which a description will be given based on the waveform at a wavelength of 470 nm. FIG. 15 shows the result of theoretically determining this change. Here, the central value of the thickness of the base oxide film is 300 nm, and the waveform when there is a variation of ± 5 nm is also plotted. In this calculation (FIG. 15), in the first 10 seconds, the area of the exposed portion (dimension 190 nm) is reduced, the film is completely removed in 10 seconds, and then development proceeds in the lateral direction at a rate of 2 nm in 10 seconds. , Under the condition that the desired dimension (200 nm) is reached in 60 seconds. From this result, it can be seen that the reflectance changes greatly when the thickness of the underlying oxide film varies.

  FIGS. 16A to 16C show how the wavelength dispersion of the reflectance changes when the film thickness and dimensions change. From these figures, it can be seen that when the reflectance is calculated for each wavelength and fitted to the base film thickness and the pattern dimensions, the base film thickness and dimensions are determined. Then, it can be seen that the point in time when the dimension reaches a predetermined value may be set as the development end point.

  More specifically, the intensity of each reflected light with respect to a plurality of wavelengths from the monitor region is detected, and it is determined by fitting in FIGS. 16A to 16C that, for example, the base film thickness is 300 nm. To do. In this case, referring to FIG. 16B, the result of measuring the reflected light intensity (reflectance) at each wavelength is fitted. For example, if the required pattern size is 200 nm, the reflectance at each wavelength is determined. May be determined as the end of development at a time point where A coincides with the curve of A. As can be seen from FIG. 16, since the change in reflectance with respect to the dimension is small, it is less accurate to determine the dimension from the change in reflectance at one wavelength, but it is sufficient to detect at a plurality of wavelengths. Accuracy is obtained.

  By the way, the fitting as described above may take time to process. In this case, the optimum development time is obtained from the time that is a singular point of intensity change and the reflectance at the specific development time. Here, the two quantities are measured because there are two physical quantities, exposure amount and focus, as factors determining the dimensions. Of course, when the focus value is known by the exposure apparatus, either the singular point of intensity change or the reflectance at a predetermined time may be measured. Usually, the singular point of intensity change has higher measurement accuracy. In such a case, the singular point of intensity change is measured.

  In this embodiment, since the lower layer of the antireflection film is an oxide film, the exposure is not always performed with the focus set by the exposure apparatus. Therefore, the optimum development time is obtained from the singular point of the intensity change and the reflectance at the specific development time (development time 30 seconds).

  Next, a method for calculating the time to be a singular point will be described. In order to detect a singular point from the waveform as shown in FIG. 15, the differential value of the intensity change is obtained, and the time when the differential value becomes a certain value or less may be obtained. In this embodiment, since the lower layer of the antireflection film is an oxide film, a wavelength suitable for obtaining a singular point cannot be determined in advance when the variation of the oxide film thickness is large. Therefore, reflectance measurement is performed for a plurality of wavelengths, and after extracting a wavelength having a large change in slope near a singular point, a time during which the differential value of the intensity change is less than a certain value may be obtained. Here, when the singular point of the intensity change was obtained in this way, it was 11 seconds. It is also possible to select a plurality of wavelengths having a large change in slope near the singular point, obtain a singular point from each wavelength, and improve the accuracy of obtaining the singular point. In this case, the number of wavelengths that can be analyzed is determined by the processing time.

  Next, a method for obtaining a dimension from the reflectance at a specific development time will be described. Here, from the previously known wavelength dispersion of the oxide film thickness and dimensions as shown in FIG. 16, the dimension at the development time of 30 seconds was 205 nm. Here, 30 seconds is selected as the specific development time. Since it is known in advance that the time that becomes the singular point of intensity change is before 20 seconds, the first 25 seconds is the time that becomes the singular point. This is because the processing speed is faster when the processing is divided, such as performing processing for accurately obtaining the size and performing processing for obtaining the dimension from the reflectance after 30 seconds. Therefore, the specific time may be set in consideration of the processing time of both of them.

  FIG. 17 shows the relationship between the exposure amount and the dimension at the development time of 30 seconds when the focus fluctuates, and FIG. 18 shows the relationship between the exposure amount and the time as a singular point when the focus fluctuates. Since the dimension at a development time of 30 seconds was measured as 205 nm and the time required for a singular point was measured as 11 seconds, the exposure amount was measured as 20.5 mJ and the focus was measured as 0.3 μm. The development time is calculated from this, but because the focus is basically set in the exposure device so that it is in a just state, it is considered that the monitored area has shifted, so the development time is calculated. Obtained from the exposure amount. Therefore, the development time is calculated as 54 seconds from the relationship shown in FIG. Therefore, the development is completed in 54 seconds (step S12).

  In this embodiment, the chip 1 representing the wafer is extracted (S8), and the development stop time is obtained by monitoring the chip during development (S11). The latent image of the extracted chip is monitored and developed. It is also possible to determine the stop time. For example, the chip 1 is required to obtain a film reduction of 11.4 nm as a latent image monitoring result (FIG. 14), and the exposure amount can be obtained from this result and FIG. Furthermore, the development stop time can be obtained based on the relationship shown in FIG.

(effect)
As described above, according to the present embodiment, when developing time control is performed in order to reduce dimensional variation between wafers, a chip processed under typical conditions in the wafer is focused on the exposure apparatus. Information is extracted by measuring the film thickness of the monitor area. Then, when the development is started and the intensity change in the monitor region of the chip is monitored, the time that becomes a singular point and the reflectance at a specific development time are measured from the wavelength dispersion with respect to the development time. As a result, it is possible to know both the exposure amount and the focus amount at the time of exposure, and it is possible to obtain an optimal development time corresponding to the exposure amount and the focus amount. As a result, dimensional variations between wafers can be greatly reduced, and dimensional accuracy and yield are greatly improved.

(Third embodiment)
(Constitution)
The developing device used in this embodiment is the same as that of the first embodiment, as shown in FIG. In the first embodiment, the development end point is determined from the reflectance, but in this embodiment, the reflected light intensity change is replaced with the phase, and the development end point is predicted from the phase.

  In this embodiment, the latent image before development and the latent image during development are monitored, and the wavelength dispersion of the reflected light intensity in the monitor region is measured. In monitoring during development, the change in reflected light intensity is converted into a phase, and the end point of development is predicted from the phase.

(Function)
The development sequence in this embodiment is basically the same as that in the first embodiment, as shown in FIG. Steps S1 to S9 are exactly the same as those in the first embodiment, and a description thereof will be omitted.

  After the developer is deposited by the start of development, the monitor head 11 moves to the chip determined in S8 (here, chip 2), adjusts the position, and starts monitoring (step S10). In the monitoring during development, the wavelength dispersion of the reflected light intensity from the monitor region is measured for each time as in the case of measuring the film thickness of the latent image. Then, the phase is obtained based on the reflected light intensity change.

  A method for determining the end of development (step S11) will be described based on the waveform during actual development. At the same time, changes in reflected light intensity at a number of wavelengths are measured, of which a description will be given based on the waveform at a wavelength of 550 nm. FIG. 19 shows the relationship (experimental result) between the development time at a wavelength of 550 nm and the reflectance of the monitor region. From this change in reflectance, it can be seen that an interference waveform is seen at the initial stage of development. A singular point can be seen at the end of the interference waveform, which is the point at which the film reduction in the exposed portion is almost completed. A method for converting the intensity of reflected light into a phase with this point set to 0 is shown below. Here, the reason for focusing on the phase is that the physical quantity does not depend on the structure of the base or the wavelength of the probe light.

  As shown in FIG. 20, the phase where the intensity change is maximum is −φ, the intensity at that time is Imax, the phase where the intensity change is minimum is −φ + π, the intensity is Imin, and the singularity is The phase is 0, and the intensity at that time is Io. If defined in this way, the relationship between the intensity I and the phase θ is expressed by the following equation.

I = a · sin (θ + φ) + b
a = (Imax−Imin) / 2
b = (Imax + Imin) / 2
Here, when the calculation is performed by substituting the intensity at the phase -φ, -φ + π, 0 as the boundary condition, the relationship between the phase and the reflected light intensity is obtained as shown in FIG. Therefore, it was possible to convert the reflected light intensity into the phase. This conversion improves accuracy by using a plurality of wavelengths. It is better to convert using as many wavelengths as the analysis time allows. Based on the relationship of FIG. 21, the relationship between development time and phase is shown in FIG. Since the time when the reflectance at a specific wavelength reaches a certain value is the time required for finishing to a desired size, the development is terminated when the specific phase is reached based on the relationship shown in FIG.

  Here, since the reflected light intensity of 0.88 when measured at a wavelength of 550 nm is an intensity finished to a desired dimension, the development is terminated at a phase of 16.4 degrees. Therefore, development is completed in 30 seconds. (Step S12).

  In this embodiment, the phase at which the development end point is measured is set to a specific value, but the present invention is not limited to this. In the second embodiment, the development end point obtained from the time as a singular point and the reflectance at a specific time is obtained. However, it is also possible to obtain the development end point by replacing the reflectance at the specific time with the phase. In addition, it is also possible to set two specific times and obtain the development end point from the phase at that time.

(effect)
As described above, according to the present embodiment, when developing time control is performed in order to reduce dimensional variation between wafers, a chip processed under typical conditions in the wafer is focused on the exposure apparatus. Information is extracted by measuring the film thickness of the monitor area. By starting development and converting the intensity change in the monitoring area of the chip into a phase and monitoring it, it becomes possible to obtain the optimum development time, and the dimensional variation between wafers can be greatly reduced. This greatly improves the dimensional accuracy and yield. Further, by replacing the intensity with the phase, it is possible to perform monitoring with a physical quantity that does not depend on the film thickness of the base or the wavelength of the probe light. Therefore, the complicated process of obtaining the dimension from the chromatic dispersion is eliminated, and the analysis can be simplified.

  In addition, this invention is not limited to each embodiment mentioned above. In the embodiment, a chip having an average exposure condition (focus, exposure amount) is extracted. However, when the exposure condition of each chip on the wafer is the same, the step of extracting the chip is not necessarily required. Further, the monitor area is not necessarily limited to the device pattern (memory cell area in the embodiment) in the device pattern area, and may be a monitor-dedicated pattern. Furthermore, the monitor area is not necessarily limited to the device pattern area, but may be formed in the mark area.

  In addition, various modifications can be made without departing from the scope of the present invention.

1 is a schematic configuration diagram showing a developing device used in each embodiment of the present invention. 6 is a flowchart for explaining a monitor chip determination and development sequence. The figure for demonstrating the position adjustment of a monitor head. The figure for demonstrating the scan of the chip | tip before image development. The figure which shows the example of the wavelength dispersion of the measured reflectance. The figure which shows the relationship between exposure amount and the film reduction amount of the latent image of a monitor area | region. The figure which shows the list of the result of having measured the latent image film thickness of the monitor area | region. The figure which shows the relationship (experimental result) between development time and the reflectance of a monitor area | region. The figure which shows the relationship (calculation result) between development time and the reflectance of a monitor area | region. The figure which shows the relationship between the line dimension of 200 nmL / S, and a reflectance. The figure which shows the relationship between the exposure amount when a focus changes, and the time which becomes a singular point. The figure which shows the relationship between the exposure amount and the development time when a pattern dimension is finished to a desired value. The figure which shows the example of the wavelength dispersion of the measured reflectance. The figure which shows the list of the result of having measured the latent image film thickness of the monitor area | region. The figure which shows the reflectance change of a 200 nmL / S pattern at the time of base film thickness fluctuation | variation. The figure which shows the change of a reflectance when a film thickness, a dimension, and a wavelength are changed. The figure which shows the relationship between exposure amount and the dimension in the development time of 30 second for 200 nmL / S. The figure which shows the relationship between the exposure amount when a focus changes, and the time which becomes a singular point. The figure which shows the relationship between development time and the reflected light intensity (wavelength 550nm) of a monitor area | region. The figure for demonstrating a mode that the intensity change of reflected light is converted into a phase. The figure which shows the relationship between a phase and the reflected light intensity normalized. The figure which shows the relationship between development time and a phase.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Development unit 11 ... Monitor head 12 ... Light source 13 ... Fiber 20 ... Wafer 21 ... Device area 22 ... Mark area 23 ... Memory cell area 31 ... CCD camera 32 ... Spectroscope 33 ... Image analysis part 34 ... Intensity analysis part 35 ... Control unit 36 ... Exposure apparatus 51 ... Field of view of CCD camera 52 ... Area detected by spectroscope 53 ... Monitor area

Claims (13)

In a pattern formation method of forming a device pattern on a resist by development after exposing the device pattern to a resist on a substrate to be processed,
Exposing the monitor area together with the device area during exposure of the device pattern;
Irradiating the monitor region with light of a plurality of wavelengths when developing the resist, and predicting a development stop time during which the device pattern is finished to a desired dimension based on wavelength dispersion of reflected light intensity reflected from the monitor region; ,
Supplying a development stop solution onto the substrate during the predicted development stop time, and stopping the development. In a pattern formation method of forming a device pattern on a resist by development after exposing the device pattern to a resist on a substrate to be processed,
Exposing the monitor area together with the device area during exposure of the device pattern;
When developing the resist, the monitor region is irradiated with light of a plurality of wavelengths, the intensity change of the reflected light reflected from the monitor region is converted into a phase change, and the device pattern has a desired dimension based on the converted phase. A process for predicting the development stop time to be finished;
Supplying a development stop solution onto the substrate during the predicted development stop time, and stopping the development.   3. The pattern forming method according to claim 1, wherein the monitor region is a memory cell region in a device pattern.   3. The pattern forming method according to claim 1, wherein the monitor region is arranged at a position showing an average flatness in the chip based on information on the flatness at the time of leveling in an exposure apparatus. .   The step of predicting the development stop time includes irradiating the monitor region with light of a plurality of wavelengths, and predicting the development stop time from the reflectance at a specific time of the intensity change of the reflected light reflected from the monitor region. The pattern forming method according to claim 1, wherein:   The step of predicting the development stop time includes irradiating the monitor region with light of a plurality of wavelengths, the reflectivity at a specific time of the intensity change of the reflected light reflected from the monitor region, and the monitor region obtained from the exposure apparatus The pattern forming method according to claim 1, wherein the development stop time is predicted from the focus value.   The step of predicting the development stop time includes irradiating the monitor area with light of a plurality of wavelengths, and reflecting the reflectivity at a specific time of the intensity change of the reflected light reflected from the monitor area and the specific change of the intensity of the reflected light. The pattern forming method according to claim 1, wherein a development stop time is predicted from a point time.   8. The specific time of the intensity change of reflected light reflected from the monitor region and the time that becomes a singular point of the intensity change of reflected light are set in different time zones during development. Pattern forming method.   The step of predicting the development stop time includes irradiating the monitor area with light of a plurality of wavelengths, converting the intensity change of the reflected light reflected from the monitor area into a phase change, and when the phase reaches a desired value. The pattern forming method according to claim 2, wherein a development stop time is predicted from   The step of predicting the development stop time irradiates the monitor region with light of a plurality of wavelengths, converts the intensity change of reflected light reflected from the monitor region into a phase change, and develops the development stop time from the phase of a specific time. The pattern forming method according to claim 2, wherein the pattern is predicted.   The step of predicting the development stop time irradiates the monitor region with light of a plurality of wavelengths, converts the intensity change of the reflected light reflected from the monitor region into a phase change, and stops the development from the phase of the plurality of times. The pattern forming method according to claim 2, wherein the time is predicted.   The step of predicting the development stop time irradiates the monitor region with light of a plurality of wavelengths, converts the intensity change of the reflected light reflected from the monitor region into a phase change, and changes the phase to phase 0 and a specific time The pattern forming method according to claim 2, wherein the development stop time is predicted from the determined time.   The pattern forming method according to claim 1, wherein the development stop time is predicted between an exposure process and a development process.

Images