DE102007000980A1 - Camera i.e. color camera, characteristic curve determining method for inspection machine, involves adjusting exposure time of camera to value, and allocating sparkles or associated exposure time to camera signal measured in camera - Google Patents

Abstract

The method involves transmitting illuminating sparkle from an illuminating device (12) with a predetermined sparkle frequency. An exposure time of a camera (10) i.e. color camera, is adjusted to a value, at which a predetermined set of sparkles is detected. The sparkles or the associated exposure time are allocated to a camera signal that is measured in the camera. The exposure time of the camera is adjusted to another value, at which another predetermined set of sparkles is detected. The latter set or the associated exposure time is assigned to another signal that is measured in the camera.

Description

The The present invention relates to a method for determining the characteristic a camera of an inspection machine, in particular for inspection a surface of a wafer according to the The preamble of claim 1.

In Semiconductor manufacturing becomes wafers during the manufacturing process processed sequentially in a variety of process steps. With As the density of integration increases, so do the demands on the Quality of structures to be formed on the wafer. To For that purpose, it is an advantage to look at the quality also individual process steps, for example lithography steps, during the manufacturing process and even before a downstream one Process step can reliably assess. Is already after execution of a process step and before the Completion of a manufacturing process that detects a wafer or on the wafer formed structures are faulty, so the wafer can be immediately sorted out without any further process steps must be executed. The one found faulty Wafer can be treated separately until a satisfactory Quality is achieved. In this way, the efficiency can be and yield in semiconductor manufacturing can be increased.

to Inspection of such surfaces are particularly suitable optical devices. There are known optical devices which by image recognition many different structures on the Can recognize the surface of a wafer. in this connection the wafer is illuminated and with a camera (matrix or line scan camera) sampled. Such wafer inspection devices can work in different types of lighting. For example a bright field illumination is possible in which the Surface of the wafer illuminated and that of the surface reflected light is captured by a camera. A dark field illumination is made possible by the fact that the surface of the Wafers illuminated and that of defects, particles and the like on the surface scattered light captured by a camera becomes. Camera signal fluctuations of the used for lighting Lighting source can be distracting affect the accuracy of the image evaluation.

As well disturbing, however, are influences that are made Differences of the camera used in the inspection machine result. Therefore, it has already been proposed the properties of each camera used to take into account that the characteristic of the camera with a constant light source through Variation of the exposure time or with a light meter is determined. The characteristic can then be used in the evaluation of the result of the taken into account with her recorded surface become.

Often, however, a type of wafer inspection machines is used in which the surface of the wafer is illuminated stroboscopically, ie with successive flashes. In this case, a region on the surface of the wafer is illuminated by a flash of light, an image is detected by the illuminated region and then the wafer and the illumination light beam are displaced relative to one another for a next image acquisition process. For this type of lighting was in the DE10359723A1 It has already been proposed to improve the accuracy by reducing the influence of camera signal fluctuations of the flashes of light used for illumination on the accuracy of image acquisition. A detection of the characteristic with a constant light source is eliminated in this type of lighting but.

task It is therefore the object of the present invention to provide a method for determining the characteristic of a camera of an inspection machine with flash lighting propose.

According to the invention this object by a method having the features according to claim 1 solved.

According to the The invention thus becomes a method for determining the characteristic of a Camera of an inspection machine, in particular for inspecting a Surface of a wafer proposed, wherein a lighting device is provided for illuminating the surface of the wafer. The camera is placed in the inspection machine so that it for detecting the reflected from the surface of the wafer Light can be used. The illumination device becomes stroboscopic, d. H. operated so that it sends out consecutive flashlights. In this case, a high flash frequency is selected, the preferred between 10 and 50 Hz, in particular at 20 Hz. The exposure time the camera and the flash frequency of the lighting device so matched to each other, that with the set flash frequency each exposure time a first number of flashes can be recorded. Then this first number of flashes can be measured in the camera Camera signal to be assigned. By changing the Exposure time can then be recorded a second number of flashes and this second number of flashes a second, in the Be assigned to the camera measured camera signal.

This process can be continued with further exposure times, so that by a variation of the exposure time of the camera received flashes for a plurality of exposure times can be assigned to the respective number of flashes of the associated measured value of the camera signal of the camera. This results in the characteristic curve of the camera to be measured from a number of different number of flashes, / N1 to NE) to which the respectively measured camera signals (S1 to SN) are assigned.

typically, the final number is much larger than the first Number of flashes. If the first number is less than 10, in particular 1, the final number can be in particular between 10 and 50. For example, the exposure time of the camera after the Setting the flash frequency to 20 Hz can be set so that she picks up the first number exactly one flash, which then a first, in the camera measured value of Kamerig nals is assigned. By a subsequent variation of the exposure time can now 2, 3, 4, ... flashes each exposure time with the Camera are recorded, which in turn each of the in assigned to the camera signal value.

When For example, the final number can be set to 50, so over this method is recorded in this case a characteristic of the camera with the discrete numbers of flashes 1, 2, 3, ... 50 respectively Camera signals are assigned. Because the brightness values of the flashes can be constant, so that the individual behavior of the camera be taken into account in future measurements.

While The flash unit must not be switched off when recording the characteristic curve be and must be throughout the recording of the characteristic flash at 20 Hz. Therefore, only after all measurements, the flash unit switched off and the black level (the dark current) for all exposure times determined.

The Determination of the characteristic curve is achieved by the measured values (Camera signal) are plotted over the number of flashes. The number of flashes on the X-axis and the camera signal on the Y-axis (Current or voltage) applied. The measured values become a Just fit. The fit procedures are conventional mathematical Method. The correction characteristic results pointwise by the measured camera signals (Y values) the X values of the correction characteristic curve are, above that, the difference between the fit Straight line and the measured camera signal (Y-value).

In a further advantageous embodiment of the invention can measure the camera signals for any number of Flashing be performed several times. For this purpose, the measurement for the first number of flashes consecutively several times be performed and from the obtained values of the camera signal for the first number an average is formed, the then the first number of flashes is assigned. Likewise will for the second number and the further number of flashes proceed to the final number. Thus the mean values for every number of flashes formed. Also, the mean can be be formed that several measurements from the first number to to the final number and the averages of the Camera signal values of the camera only after completion of all measurement series be determined for each number.

In Another embodiment is additionally the value of the camera signal for each of the exposure times measured with the illumination source switched off. Thus receives for each exposure time, a correction value, so called dark current, the camera. This allows the determined camera signal value Corrected for each exposure time and the characteristic more accurate be measured.

The Lighting device and the camera are preferably positioned that a first defined test object of the illumination device illuminated and reflected from the first defined test object Light is captured by the camera. In this way the characteristic becomes taken with a first number of bases. This process can be repeated with a second test object, the characteristic also included with a second number of bases can be from the first number of vertices is different.

To monitor to be able to ensure that the lighting device with a constant Brightness works before and after the detection of the characteristic curve the camera signals the lighting device at a large Number of flashes (<= Final number) are measured and compared.

Becomes a color camera with multiple color channels used, so can use this method in a simple way for everyone the color channels, in particular the channels R, G, B, the characteristic curve by determining the camera signal for every number of flashes for each of these channels be determined.

Further Advantages and advantageous embodiments of the invention are the subject the following figures and their parts description. Show it in detail:

1 a schematic arrangement in an inspection machine for inspecting a surface of a disc-shaped object;

2 schematically a characteristic of the camera;

3 schematically the characteristics of a color camera for the color channels R, G and B; and

4 schematically the camera dark current for different exposure times.

1 schematically shows in one possible embodiment, an inspection machine for inspecting the surface 16 a wafer 14 , The inspection machine further comprises the illumination device 12 , each having an illuminating light beam 20 as a flash can radiate to an area on the surface 16 of the wafer 14 to illuminate. The wafer 14 is on a recording device 18 applied, which is designed so that it is movable. This ensures that the wafer 14 relative to the illumination light beams 20 can be moved continuously or clocked.

The inspection machine further includes a camera 10 , with a picture of the illuminated area on the surface 16 of the wafer 14 can be detected. The camera 10 can be designed as a matrix or line camera that can capture monochrome and / or color images, preferably with R, G and B color components. The from the camera 10 Captured image data can be output to a data processing system, stored there and / or further processed.

The one of the lighting device 12 light beam emitted as a flash of light 20 passes over a half-transparent mirror 21 on the surface 16 where it reflects differently depending on the reflectivity of the surface area and into the camera 10 is shown. To determine the characteristic of the camera 10 becomes the lighting device 12 operated stroboscopically or is designed as a flashlamp. The flash frequency is preferably selected to be high, in particular between 10 and 50 Hz, for example at 20 Hz. The exposure time of the camera 10 is then adjusted, and with the lighting device 12 tuned so that during the exposure exactly one flash is recorded. This flash is then assigned a camera signal value that is in the camera 10 detected for this one flash. Subsequently, the exposure time of the camera 10 changed so that during the exposure exactly two flashes of lighting device 12 can be detected. Because in the camera 10 the intensity received by the flashes per exposure time can be summed correspondingly to a different camera signal for these two flashes. This process is repeated until the desired final number of flashes, for example 50, has been reached.

This procedure results in the characteristic curve 22 the camera, which is exemplary and schematic in 2 is shown. The x-axis shows the number of flashes emitted by the flash lamp and recorded by the camera for each exposure time. The number of flashes is proportional to that of the illumination device 12 intensity emitted per exposure time. On the y-axis is the signal value of the camera 10 applied in random units. With this characteristic is the for each camera 10 individual connection between the camera 10 detected intensity and the resulting camera signal known.

The accuracy of the characteristic curve can be further improved by averaging the individual camera signals (S1... SN) recorded for each number of flashes (N1... NE). For this purpose, for example, the measurement of the camera signal for each number of times, example, 100-fold, are made, whereby an average value of the camera signal is calculated from the individual detected camera signals. The exposure time of the camera 10 For this purpose, for example, is dimensioned such that just a flash of lighting device 12 is recorded. One measures this one flash 100 times and forms the average of the won camera signals for a flash. Then the exposure time of the camera 10 so measured that just two flashes of the lighting device 12 be recorded. Measure these two flashes 100 times and form the average of the obtained camera signals for two flashes. This will continue until the desired or specified number of flashes has been reached and thus the characteristic curve 22 is obtained.

If a color camera is used as the camera, then, as in 3 shown, the characteristic for each of the color channels are determined. For this purpose, the camera signal for each color channel R, G and B is determined with the described method, from which immediately the characteristics 24 . 26 and 28 result.

In order to further improve the method and thus the result of the determined characteristic curve, it is also possible before or after the determination of the characteristic curve 22 . 24 . 26 . 28 a camera dark current 30 be determined for each exposure time. This is then when determining the characteristic 22 . 24 . 26 . 28 considered. As in 5 For this purpose, a camera dark current is represented schematically for each of the exposure times B1 to BN used 30 determined with switched off lighting device. In order to keep the dispersion of the signal as low as possible, the camera dark current can 30 measured repeatedly for each exposure time and from this the mean value for this illumination time can be determined. For example, the exposure time B1 is set so that exactly one flash is detected and for the time B1, the camera dark current without light flash 100 times gemes will be. From this, the average value for the camera dark current for the exposure time B1 can then be calculated. Subsequently, the exposure time B2 is adjusted so that exactly two flashes can be detected and for the time B2, the camera dark current without light flash is measured 100 times. From this, the average value for the camera signal for the exposure time B2 can then be calculated. With this method, the mean camera dark currents for all exposure times B1 to BN are determined. When using a color camera, the camera dark current for all color channels R, G, B can thus also be determined as a function of the exposure time. Thus, each camera signal for each exposure time can be corrected individually by the camera dark current, resulting in an improved characteristic.

In the 1 illustrated arrangement of the camera 10 and lighting device 12 is just an example. It is also possible to choose a different arrangement, for example an oblique positioning of the illumination device 12 and the camera 10 relative to the surface 16 , It just needs to be ensured that the surface of the 16 reflected rays from the camera 10 can be detected. In the selection of the reflective surface, preference is given to using a test object whose reflectivity is known. Depending on the requirement, different test objects can also be used with which the characteristic curve can then be recorded with different interpolation points.

10

camera

12

lighting device

14

wafer

16

surface

18

recording device

20

Illuminating light beam

21

semipermeable mirror

22

curve

24

curve R

26

curve G

28

curve B

30

Camera dark current

32

Point

QUOTES INCLUDE IN THE DESCRIPTION

This list The documents listed by the applicant have been automated generated and is solely for better information recorded by the reader. The list is not part of the German Patent or utility model application. The DPMA takes over no liability for any errors or omissions.

Cited patent literature

• - DE 10359723 A1 [0005]

Claims (12)

Method for determining the characteristic curve ( 22 . 24 . 26 . 28 ) of a camera ( 10 ) of an inspection machine, in particular for inspecting a surface ( 16 ) of a wafer ( 14 ), wherein a lighting device ( 12 ) for illuminating the surface ( 16 ) of the wafer ( 14 ) and the camera ( 10 ) for detecting the surface ( 16 ) of the wafer ( 14 Reflected light are provided, characterized by the following steps: - emission of illumination flashes from the illumination device ( 12 ) with a predetermined flash frequency; - Setting the shutter speed of the camera ( 10 ) to a value at which each exposure time of the camera ( 10 ) a predetermined first number (N1) of flashes is detected; Assigning the first number (N1) of flashes or the associated exposure time to one in the camera ( 10 ) measured first camera signal (S1); - Setting the shutter speed of the camera ( 10 ) to a value at which a predetermined second number (N2) of flashes is detected; Assigning the second number (N2) of flashes or the associated exposure time to one in the camera ( 10 ) measured second signal value (S2). Method according to claim 1, characterized in that the exposure time of the camera ( 10 ) is set after detecting the second number of flashes to a value with which a further number of flashes can be detected and this further number can be assigned a signal value measured in the camera and this process is repeated until an end count (NE ) of flashes to which a signal end value (SE) is assigned. Method according to claim 2, characterized in that that the first number (N1) is equal to 1, that the second number (N2) is equal to 2 and the end number (NE) is much larger is and is in particular between 10 and 50. Method according to claim 3, characterized the final number (NE) is equal to 20. Method according to one of claims 1 to 4, characterized in that the measurement of the camera signal for each number (N1 ... NE) of flashes is performed several times and from the obtained camera signal values (S1 ... SE) for every number of flashes a camera signal average is formed which is assigned to the respective number of flashes. Method according to one of Claims 1 to 5, characterized in that, before or after the assignment of the camera signal values for all exposure times, the following step is carried out: - switching off the illumination source ( 12 ); - Measuring the camera dark current ( 30 ) in the camera ( 10 ) for each of the exposure times used (B1, B2, ... BN). A method according to claim 6, characterized in that with the illumination source switched off ( 12 ) for each exposure time (B1, B2, ... BN) measured camera dark current ( 30 ) the camera signal (S1 ... SN) assigned for each number of flashes is individually corrected. Method according to one of claims 2 to 7, characterized in that the illumination device ( 12 ) is positioned so that it illuminates a first defined test object and the camera ( 10 ) is positioned so as to detect the light reflected from the first defined test object and to record the characteristic with a first number of vertices. Method according to claim 8, characterized a second, differently reflecting from the first test object Test object illuminated and the characteristic with a second number of Support points is recorded. Method according to one of claims 1 to 9, characterized in that before the measurement of the first number and after the measurement of the final number, a measurement of the camera signals the flash light source is performed so that detects can be whether the flash source during the measurement works with constant brightness. Method according to claim 10, characterized that a measurement, whether the flash light source with constant brightness works at regular intervals during the actual measurement is done so a Drift of the flash units can be considered. Method according to one of claims 1 to 11, characterized in that the camera is a color camera with several Color channels is and for each of the color channels, in particular the channels R, G, B, determines the characteristic becomes.