JP5494303B2 - TFT array inspection apparatus and defect strength calculation method - Google Patents

Description

  The present invention relates to a TFT array inspection for inspecting an array of a TFT substrate such as a liquid crystal substrate, and more particularly to data processing with a detection intensity suitable for defect detection of a TFT array.

  In a liquid crystal array inspection apparatus, as a picked-up image obtained by picking up an image on a liquid crystal substrate, an optical pick-up image obtained by optical pick-up or a charged particle beam such as an electron beam or an ion beam is two-dimensionally displayed on the substrate. A scanned image obtained by scanning can be used.

  In the manufacturing process of the TFT array substrate used in the TFT display device, an inspection is performed to check whether the manufactured TFT array substrate is driven correctly (Patent Documents 1 and 2).

  For example, an inspection signal is applied to an array of substrates to be inspected to bring the array into a predetermined potential state, and the substrate is scanned by two-dimensional irradiation with a charged particle beam such as an electron beam or an ion beam. There is known an array inspection apparatus for inspecting an array of TFTs based on a scanning image obtained in (1). In the TFT array inspection, secondary electrons emitted by electron beam irradiation are detected by converting them into analog signals using a photomultiplier or the like, and an array defect is determined based on the signal intensity of the detection signals.

  In the TFT substrate, an array and a pixel are formed corresponding to each other, and the pixel can be driven by applying a drive signal to the array. In the TFT array inspection, generally, an electron beam is irradiated to a plurality of locations for one pixel, and secondary electrons emitted from each irradiation point are detected.

  Since the signal intensity of the detection signal of secondary electrons detected from one pixel varies, it is known to perform median processing to assign a plurality of signal intensity to one pixel (Patent Document 3).

JP 2004-271516 A JP 2004-309488 A JP-A-2005-321308 (paragraphs 0026 to 0029)

  In the array inspection of the TFT substrate, the signal intensity of the detection signal obtained for each pixel is used as the defect intensity used for defect detection of the pixel, and the defect intensity is compared with a predetermined threshold value. .

  The signal intensity of the detection signal varies depending on various factors such as the irradiation state of the electron beam applied to the pixel and the detection state of the secondary electron detector.

  In order to improve the detection accuracy and signal strength of the detection signal, multiple detection signals are detected from one pixel, the detected multiple detection signals are assigned to each pixel, and the signal strength of the assigned detection signals is used for defect detection. It is known that the defect strength is used.

  When assigning signal intensities of a plurality of detection signals to each pixel, if the signal intensity of each detection signal fluctuates, the assigned defect intensity also fluctuates due to the fluctuation of the signal intensity, so that the detection accuracy of defect detection decreases. There's a problem.

  In addition, the measurement of detecting secondary electrons by irradiating an electron beam is repeated several times on the same substrate, the signal intensity of a plurality of detection signals is obtained for each pixel, and the signal increased by these plurality of detection signals It is known to compare the intensity with a threshold value.

  When acquiring multiple detection signals by measuring multiple times for a single pixel, if the signal strength of the detection signal obtained by each measurement varies, the repeatability of the defect strength calculated for each pixel decreases, and defects are detected. Detection accuracy will be reduced.

  Further, when the pixel defect strength is obtained by the average value of the signal intensities of a plurality of detection signals, the obtained defect strength signal strength itself is higher than the minimum value obtained in each measurement by the average processing, or The value is lower than the maximum value. The direction of change in which the signal strength of the defect strength increases or decreases due to this averaging process is the direction in which the defective pixel is determined as a normal pixel in the defect detection performed by comparison with the threshold value. There is a risk of lowering the detection accuracy.

  Therefore, the present invention solves the above-described problem, and occurs when the signal intensity of each detection signal fluctuates when determining the defect intensity used for defect detection in each pixel based on the signal intensity of a plurality of detection signals. The purpose is to reduce the fluctuation of the defect intensity of each pixel and prevent the detection accuracy of the defect detection from being lowered.

  Further, according to the present invention, when obtaining the defect intensity used for defect detection in each pixel based on the signal intensity of the plurality of detection signals, when obtaining a plurality of detection signals by measuring a plurality of times for each pixel. It is an object of the present invention to reduce a decrease in defect intensity repetition accuracy and a decrease in detection accuracy of defect detection caused by fluctuations in the signal intensity of a detection signal obtained in each measurement.

  According to the present invention, for a pixel whose defect intensity is to be determined, a predetermined range is defined in the vicinity of the pixel, and a function for interpolating the intensity distribution in the vicinity of the pixel using a detection signal within the predetermined range is obtained. The defect intensity of the pixel is obtained from (minimum value or maximum value). The defect intensity is a signal intensity for detecting a defect in the pixel, and the defect is detected by comparing the defect intensity with a predetermined threshold value.

  According to the present invention, the intensity distribution in the vicinity of a pixel can be interpolated with a function, and the defect intensity of the pixel can be obtained from the extreme value of the obtained function. By obtaining the extreme value of the function that interpolates the intensity distribution, the fluctuation of the defect intensity of each pixel is suppressed with respect to the fluctuation of the signal intensity of each detection signal, and the deterioration of the detection accuracy of the defect detection is prevented.

  Further, the present invention obtains a function that interpolates the intensity distribution in the vicinity of the pixel, and obtains an extreme value of this function, thereby reducing the repetition accuracy of the defect intensity caused by repeating measurement for one pixel and the defect. Reduction in detection accuracy of detection is reduced.

  The present invention applies an inspection signal of a predetermined voltage to the TFT substrate, drives the array, detects secondary electrons obtained by irradiating the TFT substrate with an electron beam, and determines the signal intensity of the detection signal of the secondary electrons. Of TFT substrate array inspection for detecting array defect based on pixel defect intensity, and TFT array inspection apparatus for detecting array defect based on pixel defect intensity Embodiments.

  According to an aspect of the defect strength calculation method of the present invention, an inspection signal of a predetermined voltage is applied to the TFT substrate to drive the array, detect secondary electrons obtained by irradiating the TFT substrate with an electron beam, and detect the secondary electrons. In TFT substrate array inspection in which the TFT substrate array is inspected based on the signal strength of the detection signal, the TFT substrate array inspection is a defect using the defect strength calculated from the signal strength of the detection signal for the pixel corresponding to the array. Perform by detection.

  The defect intensity is calculated by irradiating a plurality of locations on a TFT substrate with an electron beam to detect secondary electrons, obtaining a plurality of detection signals per pixel, and detecting the detection signals from the plurality of detection signals. The step of setting an extraction range for extracting a detection signal used for calculating the intensity distribution of at least three extraction ranges of each pixel, an extraction range including the pixel, and at least two extraction ranges including a part of the pixel , The step of calculating the average intensity of the detection signals included in each extraction range, and the calculated average intensity and the position on the TFT substrate representing the extraction range for calculating the average intensity are approximated for at least three coordinate points. A step of obtaining a function to be interpolated and a step of obtaining an extreme value of the obtained function are set as defect strengths used for defect detection of the pixel.

  In addition, in the calculation of the defect intensity according to the present invention, the detection signal from the TFT substrate is not limited to one detection but can be applied to a plurality of detections that are repeated a plurality of times on the same substrate. A defect intensity average value is calculated by averaging a plurality of defect intensities obtained by a plurality of detections, and a defect is detected using the calculated defect intensity average value.

  The present invention calculates an average value from an extraction range of a pixel for which defect strength is obtained and an extraction range in the vicinity of the pixel, calculates an extreme value from a function approximating the average value, and determines the calculated extreme value as a defect of the pixel. By using the defect strength for detecting the signal, it is possible to reduce the influence of fluctuations in the signal strength of the detection signal and variations in the signal strength among the detection signals obtained by a plurality of detections.

  Further, since the extreme value is closer to the threshold value used for defect detection than the signal intensity at each detection point interpolated by the function for obtaining the extreme value, the detection accuracy of defect detection is increased.

  The aspect of the TFT array inspection apparatus of the present invention is to drive the array by applying an inspection signal of a predetermined voltage to the TFT substrate, detect secondary electrons obtained by irradiating the TFT substrate with an electron beam, and detect secondary electrons. In a TFT substrate array inspection apparatus that inspects an array of TFT substrates based on the signal intensity of the detection signal, a defect detection unit that detects defects using the defect intensity calculated from the signal intensity of the detection signal for pixels corresponding to the array And an average intensity calculation unit that calculates the average intensity of the detection signal within a predetermined extraction range from the signal intensity of the detection signal, and a defect intensity calculation unit that calculates the defect intensity from the signal intensity of the detection signal, On the substrate, a plurality of locations of one pixel are irradiated with an electron beam to detect secondary electrons and obtain a detection signal.

  The average intensity calculation unit sets an extraction range for extracting a detection signal used for calculating an intensity distribution of the detection signal from a plurality of detection signals per pixel, and for each pixel, an extraction range including the pixel, and the pixel In at least three extraction ranges of at least two extraction ranges including a part of the detection signal, the average intensity of the detection signal included in each extraction range is calculated.

  The defect strength calculator obtains a function that interpolates by approximating at least three coordinate points, with the average strength and the position on the TFT substrate representing the extraction range for calculating the average strength as coordinate points. The extreme value of the function is obtained, and the obtained extreme value is calculated as the defect intensity used for defect detection of the pixel.

  The average strength calculation unit and the defect strength calculation unit repeatedly calculate the defect strength for the same substrate a plurality of times. The defect strength average value calculation unit calculates the defect strength average value by averaging the defect strengths obtained by repeated calculations, and detects defects using the calculated defect strength average value.

  According to the present invention, in the aspect of the defect strength calculation method and the aspect of the TFT array inspection apparatus, various extraction ranges can be set for calculating the defect strength.

  In the calculation of the defect intensity of the present invention, an example of the extraction range is an extraction range including a detection point in a pixel, and any axis in the x-axis direction, y-axis direction, or diagonal direction on the TFT substrate with respect to this pixel. The function that interpolates the points of the average intensity obtained in each of the extraction ranges, which are at least three of the two extraction ranges adjacent in the direction, is obtained by calculating the average intensity and the extraction range obtained in these extraction ranges. This is a quadratic function passing through three coordinate points represented by representative positions.

  In the calculation of the defect intensity according to the present invention, another example of the extraction range includes an extraction range including detection points in the pixel, and 4 adjacent to the pixel in the x-axis direction and the y-axis direction on the TFT substrate. One extraction range, or an extraction range including a pixel, and eight extraction ranges adjacent to the pixel in the x-axis direction, the y-axis direction, and the diagonal direction on the TFT substrate. The function that interpolates the obtained average intensity points approximates a curved surface that passes through the five or nine coordinate points represented by the average intensity obtained within these extraction ranges and the position that represents the extraction range. Function to interpolate.

  Further, in the calculation of the defect intensity according to the present invention, various arrangements of detection signals in the extraction range can be determined. An example of the extraction range is a range in which four detection signals arranged in a rectangular shape are extracted, and the average intensity of the extraction range is obtained from the average value of the signal strengths of the four detection signals.

  According to the present invention, when determining the defect intensity used for defect detection in each pixel based on the signal intensity of a plurality of detection signals, the defect intensity of each pixel caused by the fluctuation of the signal intensity of each detection signal. It is possible to reduce fluctuations and prevent a decrease in detection accuracy of defect detection.

  In addition, according to the present invention, when the defect intensity used for defect detection in each pixel is obtained based on the signal intensity of the plurality of detection signals, a plurality of detection signals are acquired by measuring a plurality of times for each pixel. At this time, it is possible to reduce a decrease in defect strength repeatability and a decrease in detection accuracy of defect detection caused by fluctuations in the signal strength of detection signals obtained in each measurement.

It is a schematic block diagram for demonstrating the structural example of the TFT array test | inspection apparatus of this invention. It is a flowchart for demonstrating the outline of the calculation operation | movement of the defect intensity | strength of this invention. It is explanatory drawing for demonstrating the outline of the calculation operation | movement of the defect strength of this invention. It is explanatory drawing for demonstrating the outline of the calculation operation | movement of the defect strength of this invention. It is explanatory drawing for the structure of the signal processing part of the TFT array test | inspection apparatus of this invention. It is a flowchart for demonstrating calculation of the defect strength of this invention. It is a figure for demonstrating the defect intensity | strength extraction range of this invention. It is explanatory drawing for demonstrating the example of calculation of the defect strength of this invention. It is explanatory drawing for demonstrating the example of calculation of the defect strength of this invention. It is explanatory drawing for demonstrating the example of calculation of the defect strength of this invention. It is explanatory drawing for demonstrating the example of calculation of the defect strength of this invention. It is explanatory drawing for demonstrating another example of calculation of the defect strength of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Hereinafter, the configuration of the TFT array inspection apparatus of the present invention will be described with reference to FIG. 1, the outline of the defect intensity calculation operation of the present invention will be described with reference to FIGS. 2 to 4, and the present invention will be described with reference to FIG. The configuration of the signal processing unit of the TFT array inspection apparatus will be described, calculation examples of the defect strength of the present invention will be described with reference to FIGS. 6 to 10, and another calculation example of the defect strength of the present invention will be described with reference to FIG. A modification of the defect strength calculation example of the present invention will be described with reference to FIG.

[Configuration of TFT array inspection equipment]
FIG. 1 is a schematic diagram for explaining a configuration of a TFT array inspection apparatus according to the present invention.

  FIG. 1 is a schematic block diagram for explaining a configuration example of a TFT array inspection apparatus of the present invention. In the example shown in FIG. 1, a TFT substrate such as a liquid crystal substrate is irradiated with an electron beam, secondary electrons emitted from the TFT substrate are detected, and the defect intensity used for defect detection is determined from the signal intensity of the detection signal of the secondary electrons. The example of a structure to acquire is shown.

  In FIG. 1, a TFT array inspection apparatus 1 includes a stage 2 on which a TFT substrate 100 such as a liquid crystal substrate is placed and can be conveyed in the XY directions, and an electron gun 3 disposed above the stage 2 and away from the stage 2. And a detector 4 for detecting secondary electrons emitted from a pixel (not shown) of the panel 101 of the TFT substrate 100.

  The stage 2 is controlled to be driven by a stage drive control unit 6, and the electron gun 3 is controlled to be irradiated with an electron beam and scanned on the TFT substrate 100 by an electron beam scanning control unit 5. The detection signal of the secondary electrons detected by the detector 4 is processed by the signal processing unit 10, and the obtained defect intensity or defect intensity average value is sent to the defect detection unit 20 to be used for pixel defect detection.

  Since the pixels and the array are formed on the panel of the TFT substrate and each pixel is driven by applying a voltage to the array, the pixel defect detection corresponds to the array inspection for the pixel.

  The drive operation of each part of the electron beam scanning control unit 5, the stage drive control unit 6, the signal processing unit 10, and the defect detection unit 20 is controlled by the control unit 7. The control unit 7 has a function of performing control including the entire operation of the TFT array inspection apparatus 1, and can be configured by a CPU that performs these controls and a memory that stores a program that controls the CPU.

  The stage 2 mounts the TFT substrate 100 and is movable in the X-axis direction and the Y-axis direction by the stage drive control unit 6, and the electron beam irradiated from the electron gun 3 is an electron beam scanning control unit 5. Can be swung in the X-axis direction or the Y-axis direction. The stage drive control unit 6 and the electron beam scanning control unit 5 can scan the electron beam on the TFT substrate 100 and irradiate each pixel on the panel 101 of the TFT substrate 100 by single or cooperative operation.

[Outline of defect strength calculation by TFT array inspection system]
2 to 4 are a flowchart and a schematic diagram for explaining the outline of the defect strength calculation operation of the present invention.

  An outline of a procedure for detecting a defect of a pixel by the TFT array inspection apparatus of the present invention will be described with reference to a flowchart of FIG. 2 and explanatory diagrams of FIGS.

  First, the TFT substrate is irradiated with an electron beam and scanned, secondary electrons emitted from the pixel by this electron beam irradiation are detected, and the detection intensity of the detection signal is measured. The detection intensity obtained by detecting the secondary electrons emitted from the pixel has the property that it varies depending on the potential applied to the pixel. The TFT array inspection apparatus 1 uses this property to detect pixel defects and perform array inspection. Depending on the voltage pattern applied to the array, the combination of driven pixels can be changed, thereby changing the defect type to be detected.

  The measurement of the detection intensity of the detection signal by electron beam scanning on the TFT substrate is performed a plurality of times on the same TFT substrate, and a plurality of detection intensities of the detection signal at the same detection point can be obtained.

  FIG. 3A shows an example of a detection point 110 that irradiates an electron beam on the panel 101. In the example shown here, an example is shown in which four electron beams are irradiated to one pixel to obtain detection signals from four detection points 110. FIG. 3A shows an example in which measurement is performed n times. Therefore, in this case, n detection intensities are obtained for the same detection point 110 (S1).

  Next, using the detected intensity, the intensity distribution in the vicinity of the pixel for which the defect is to be detected is calculated. This intensity distribution indicates the distribution state of the detection intensity in the vicinity of the target pixel, and FIG. 4A schematically shows the intensity distribution 104 when the detection signal is obtained for the entire surface of the pixel. .

  In FIG. 4A, the pixel P22 having the x-axis position “2” and the y-axis position “2” has a defect, and the detection intensity level obtained from this pixel is lower than the other pixels. Yes. Here, although the detection intensity level of the defective part is shown in a low state, it is not necessarily low, and the detection intensity level of the defective part may be high depending on the defect type.

  In the present invention, this intensity distribution is calculated by a function that interpolates the discrete detection intensity obtained at the detection point. For example, in FIG. 4A, by using detection intensities at a plurality of detection points 110, a function of an intensity distribution curve 121 for interpolating detection intensities in the y-axis direction, and an intensity distribution curve for interpolating detection intensities in the x-axis direction. 122 functions are obtained. 4B shows an example of a function of an intensity distribution curve 121 that interpolates the detected intensity in the y-axis direction, and FIG. 4C shows an example of a function of the intensity distribution curve 122 that interpolates the detected intensity in the x-axis direction. Show.

  The function that interpolates the detection intensity is obtained using the detection intensity of the detection point in the extraction range including the detection point of the target pixel for which the intensity distribution is obtained, and in the adjacent extraction range including a part of the target pixel, Using the obtained detection intensity and the position of the extraction range as coordinate points, a function that approximates these coordinate points can be calculated.

  For example, in FIG. 3B, an extraction range A including four detection points of the target pixel P, an extraction range B including four detection points including a part of the target pixel P, and the other of the target pixel P For the extraction range C including four detection points including a part of the average value, the average intensity is calculated as the detection intensity in each extraction range, and the coordinates composed of the average intensity representing each of these extraction ranges and the position of the extraction range A function for approximating points and interpolating is calculated. The position representing the extraction range is a calculation point for calculating the average intensity, and can be, for example, the center or the center of gravity of a plurality of detection points included in the extraction range (S3).

  Next, the extreme value of the intensity distribution is obtained using the obtained interpolation function. The position representing the extreme value in the intensity distribution curve represents the candidate position of the defective part. 4B and 4C, the extreme values of the functions of the intensity distribution curves 121 and 122 appear at the position of the pixel P22, indicating that the pixel P22 may be a defect. Whether or not the pixel P22 is a defective pixel is determined by comparing the defect intensity, which is the signal intensity of the detection signal of the pixel P22, with a threshold value (S4).

  The obtained extreme value is used as the defect strength of the pixel. FIG. 3C shows the defect intensity calculated for each pixel. For example, in the first measurement, the defect intensity I11a is calculated for the pixel P11, and the defect intensity I21a is calculated for the pixel P21 (S5). When the number of times of measurement is one, defect detection is performed using the calculated defect strength (S8).

  When a plurality of measurements are performed on the same TFT substrate, n defect intensities of defect intensities I11a to I11n are calculated for the pixel P11 by the first to nth measurements, and a defect is detected for the pixel P21. N defect intensities of intensities I21a to I21n are calculated (S5).

  After calculating the defect strength with respect to the total number of measurements (S6), the average of the calculated plurality of defect strengths is obtained to calculate the average defect strength, and this value is taken as the defect strength. FIG. 3D shows the average defect intensity value Iijm (S7). Defect detection is performed using the calculated defect strength average value Iijm as the defect strength (S8).

[Signal processing section of TFT array inspection equipment]
Next, a configuration example of the signal processing unit provided in the TFT array inspection apparatus of the present invention will be described with reference to FIG.

  The signal processing unit 10 illustrated in FIG. 5 includes a storage unit 14 that receives and stores the signal strength of the detection signal, and a signal strength at a detection point that is within the extraction range from the signal strength of the detection signal stored in the storage unit 14. An average intensity calculation unit 11 that calculates the average intensity by calculating the average of the read signal intensity, and a storage unit 15 that stores the average intensity calculated by the average intensity calculation unit 11 together with a position representative of the position of the extraction range, The average intensity stored in the storage unit 15 and the position of the extraction range, which is a calculation point for calculating the average intensity, are read out, and the read value is used as a coordinate point, and a function for approximating and interpolating these coordinate points is calculated. A function calculation unit 12a, an extreme value calculation unit 12b that calculates an extreme value (minimum value or maximum value) of the calculated function, and an extreme value calculated by the extreme value calculation unit 12b as a defect strength, and the extreme value exists. Pixels to A defect intensity average value calculating section that reads out defect intensity stored in the same pixel from the storage section 16, calculates an average of the read defect intensity, and calculates a defect intensity average value. 13.

  The function calculation unit 12a and the extreme value calculation unit 12b constitute a defect strength calculation unit 12 that calculates the defect strength. When the number of measurements is only one and the value of the defect strength used for the average calculation is one, the defect strength average value calculation unit 13 uses the defect strength calculated by the defect strength calculation unit 12 as it is. Output as an average value.

[Example of defect strength calculation for TFT array inspection equipment]
In the calculation example shown here, two extraction ranges that are close to the extraction range set for the target pixel for which the defect intensity is calculated are set. The defect intensity is calculated by calculating the average intensity for each of the three extraction ranges including the extraction range including the detection point in the target pixel and two adjacent adjacent extraction ranges set for the extraction range. Using the calculated point representing the position of the position and the average intensity as coordinate points, a quadratic function passing through these three coordinate points is calculated, the extreme value of the calculated quadratic function is calculated, and the calculated extreme value is defined as the defect strength. .

  Hereinafter, calculation examples of the defect strength of the present invention will be described with reference to FIGS. Here, an example in which an extraction range adjacent in the y-axis direction with respect to the extraction range set for the target pixel is shown as an adjacent neighboring extraction range, and the extraction range of the target pixel in the x-axis direction is shown. An example in which adjacent extraction ranges are set will be described later with reference to FIG.

  FIG. 6 is a flowchart for explaining the calculation of the defect strength of the present invention. First, a TFT substrate is irradiated with an electron beam for scanning, secondary electrons emitted from the pixel by this electron beam irradiation are detected, the detection intensity of the detection signal is measured, and the measured detection intensity is measured on the TFT substrate. It is memorized with the position of (S10). The detection intensity of the detection signal by scanning the electron beam onto the TFT substrate is measured a plurality of times on the same TFT substrate, and the detection intensities of the plurality of detection signals are obtained for the same detection point (S11).

  For the target pixel, an extraction range including the detection point in the pixel and two adjacent extraction ranges adjacent to the extraction range are set. FIG. 7 is a diagram for explaining the extraction range. FIG. 7 shows only nine of the many pixels of the panel. FIG. 7 shows an example in which one pixel is irradiated with four electron beams and detection signals are acquired from four detection points. Here, an example of four detection points is shown, but the number of detection points can be changed by changing the irradiation interval of the electron beam and increasing or decreasing the number of irradiation points irradiated to one pixel. In FIG. 7A, one pixel 102 has four detection points 110.

  In FIG. 7B, an extraction range including four detection points adjacent in a rectangular shape with respect to the detection points on the panel is set, and the average intensity is calculated and extracted from the signal intensity of the detection points within the extraction range. The signal intensity in the range is obtained, and the defect intensity of each pixel is calculated using this average intensity.

  As the calculation point 111 representing the position of the average intensity calculated in the extraction range, the center point of each pixel and the center point of adjacent pixels in the pixels aligned in the y-axis direction are set. An explanation will be given taking pixel A in FIG. An extraction range 112A including four detection points in the pixel A and two extraction ranges 112B and 112C adjacent to the extraction range 112A in the y-axis direction are set for the pixel A. The extraction range 112B includes two detection points on the upper side among the four detection points in the extraction range 112A and two detection points on the lower side of the adjacent pixel above the pixel A. The extraction range 112C includes: Among the four detection points in the extraction range 112A, two detection points on the lower side and two detection points on the upper side of the adjacent pixel on the lower side of the pixel A are included.

  The extraction ranges 112A to 112C calculate the average intensity by averaging the detection intensities of the four detection points included in each extraction range. The calculated average intensity is treated as the signal intensity at the positions of the calculation points 111A to 111C set in each extraction range. The average intensity at each calculation point described above is calculated and stored for all detection points detected by the panel (S12, S13).

  Next, the defect intensity in each pixel is obtained using the average intensity calculated in the processes of S14 to S16.

  For the target pixel whose defect intensity is to be calculated, three average intensities are read out: the average intensity of the calculation point corresponding to the target pixel, and the average intensity of calculation points adjacent to the calculation point on both sides in the y-axis direction ( S14), a quadratic function passing through three points with the position and average intensity of these calculation points as coordinate points is obtained. 8 and 9, the average intensity points A to C of the calculation points 111A to 111C are plotted in a coordinate system in which the horizontal axis is the position in the y-axis direction and the vertical axis is the average intensity. The curves in FIGS. 8 and 9 indicate quadratic functions passing through these calculation points 111A to 111C.

  Note that when the target pixel is the outer peripheral pixel of the panel, the average intensity used for the calculation is two, so the average intensity of the pixels outside the outer peripheral pixel is obtained by extrapolation, and a quadratic function is obtained using this average intensity. May be calculated (S15).

  The extreme value is obtained for the obtained quadratic function, and the extreme value is stored as the defect intensity of the target pixel (S16).

  When calculating the defect strength for all pixels (S16) and performing multiple measurements on the same TFT substrate, calculate the defect strength for all measurements (S18), and then calculate the calculated multiple defect strengths. Then, the average of the defect strength is calculated to calculate the average value of the defect strength, and this value is set as the defect strength (S19).

  8 and 9 show examples of obtaining the minimum value. Here, an example will be described in which the signal intensity of the detection signal becomes low when the array has a defect.

  The defect detection of a pixel is performed by comparing the signal intensity of the detection signal of the pixel with a predetermined threshold value. Here, the signal intensity used for defect detection is proved as the defect intensity. 8 and 9, th1 and th2 (th1 <th2) are set as threshold values to be compared with the defect intensity in defect detection.

  FIG. 8A shows a case where the signal intensity of the pixel A is smaller than the threshold value th1. In this case, when the signal intensity of the pixel A is compared with the threshold value th1 as the defect intensity, it is detected as a defect. At this time, if the signal intensity of the pixel A fluctuates upward, the signal intensity of the pixel A becomes A ′, and may not be detected as a defect when compared with the threshold value th.

  In the present invention, the extreme value of the quadratic function obtained for the target pixel is used as the defect intensity used for defect detection of the pixel. In FIG. 8A, the quadratic function when the signal intensity of the pixel A fluctuates upward to become the signal intensity A ′ is indicated by a broken line, and the minimum value of the quadratic function is indicated by a point Q. Since the average intensity of the point Q obtained from the minimum value of the quadratic function is equal to or smaller than the average intensity of the pixel A, a defect can be detected by comparing the average intensity of the point Q with the threshold value th1. It becomes possible.

  FIG. 8B shows a case where defect detection is performed by using a threshold value th2 higher than the threshold value th1. Although it becomes possible to detect defects by increasing the threshold level in this way, an increase in the threshold level increases the possibility of erroneous detection of detecting a non-defective pixel as a defective pixel, and the accuracy of defect detection. There is a problem that decreases. For example, in the example of FIG. 9, the detection intensity of the pixel A is not detected at the threshold value th1, but is detected at the threshold value th2 at which the threshold level is increased.

  In the present invention, the defect detection accuracy can be improved without increasing the threshold level by using the extrema of the quadratic function obtained for the target pixel as the defect intensity.

  FIG. 10 shows an example in which the defect intensity is calculated for a defective pixel and a pixel adjacent to the defective pixel. FIG. 10 shows an example in which the second pixel in the x-axis direction and the y-axis direction is a defective pixel.

  FIGS. 10C and 10D show the case where the defect intensity of the defective pixel is calculated, FIG. 10C shows the extraction range, and FIG. 10D shows the average intensity and quadratic function of each extraction range. , And the minimum value. Here, since the defect strength obtained from the minimum value Q is smaller than the threshold value, the defect is detected.

  On the other hand, FIGS. 10A and 10B and FIGS. 10E and 10F show the case where the defect intensity of the pixel adjacent to the defective pixel is calculated, and FIGS. 10A and 10E show the extraction range. FIGS. 10B and 10F show the average intensity, the quadratic function, and the minimum value of each extraction range. Here, since the defect strength obtained from the minimum value Q is larger than the threshold value, no defect is detected.

[Another calculation example of defect strength of TFT array inspection equipment]
Another example of calculating the defect strength according to the present invention will be described with reference to FIG.

  In the calculation example described above, an extraction range adjacent in the y-axis direction is set with respect to the extraction range of the target pixel whose defect strength is to be calculated, and a quadratic function is obtained based on the average intensity obtained in these extraction ranges. This is an example in which the extrema of the quadratic function is the defect strength.

  On the other hand, the calculation example shown in FIG. 11 sets an extraction range adjacent in the x-axis direction to the extraction range of the target pixel for which the defect strength is obtained, and based on the average intensity obtained in these extraction ranges. This is an example in which a quadratic function is obtained and the extrema of the quadratic function is used as the defect strength.

  The calculation procedure can be similarly performed by using the x axis instead of the y axis in the flowchart shown in FIG.

  FIG. 11 shows an example in which the defect intensity is calculated for a defective pixel and a pixel adjacent to the defective pixel. FIG. 11 illustrates an example in which the second pixel in the x-axis direction and the y-axis direction is a defective pixel.

  11C and 11D show a case where the defect intensity of the defective pixel is calculated, FIG. 11C shows the extraction range, and FIG. 11D shows the average intensity and quadratic function of each extraction range. , And the minimum value. Here, since the defect strength obtained from the minimum value Q is smaller than the threshold value, the defect is detected.

  On the other hand, FIGS. 11A, 11B, 11E, and 11F show the case where the defect intensity of a pixel adjacent to the defective pixel is calculated, and FIGS. 11A and 11E show the extraction range. FIGS. 11B and 11F show the average intensity, quadratic function, and local minimum value of each extraction range. Here, since the defect strength obtained from the minimum value Q is larger than the threshold value, no defect is detected.

[Variation of another calculation operation of defect strength of TFT array inspection equipment]
A modified example of the defect strength calculation example of the present invention will be described with reference to FIG.

  The above calculation example shows an example in which the extraction range of the target pixel for which the defect strength is calculated is adjacent to the y-axis direction or the x-axis direction. The average strength may be calculated to obtain a quadratic function, and the extreme value of the quadratic function may be used as the defect strength.

  FIGS. 12A and 12B show an example in which extraction ranges adjacent in the oblique direction are used.

  In addition, each example described above shows an example in which the extraction range of the target pixel for which the defect strength is obtained is adjacent in the y-axis direction, the x-axis direction, or the diagonal direction, but the outer periphery of the extraction range of the target pixel. An average intensity of adjacent extraction ranges may be calculated, a function that interpolates by approximating these calculation points with a curved surface may be obtained, and an extreme value of this function may be used as the defect intensity.

  FIGS. 12C and 12D illustrate an example in which extraction ranges that are adjacent to each other in the x-axis direction and the y-axis direction with respect to the target pixel extraction range are used. In this case, there are a total of five extraction ranges.

  FIGS. 12E and 12F show an example in which extraction ranges adjacent to both sides of the target pixel extraction range in the x-axis direction, the y-axis direction, and the oblique axis directions are used. . In this case, there are a total of nine extraction ranges.

  In the calculation examples of FIG. 12C to FIG. 12F, the extraction ranges are arranged on a plane, so that the average intensity of each calculation point is approximated and corrected by a function representing the plane.

  The table below shows a comparison between the conventional calculation and the calculation according to the present invention with respect to an example in which the defect strength average value and the range (range) of the variation in defect strength are measured three times for the same substrate.

  In addition, the numerical value of the comparison result is an extraction for performing comparison by both calculations, and the magnitude of the numerical value is not meaningful.

  From the above comparison results, the defect strength repeatability is improved and the average value of the defect strength itself can be lowered, indicating an improvement in the defect detection accuracy.

  In the present invention, the TFT substrate can be a liquid crystal substrate or an organic EL, and can be applied to a film forming apparatus for forming various semiconductor substrates in addition to a film forming apparatus for forming a liquid crystal substrate or an organic EL.

DESCRIPTION OF SYMBOLS 1 Array inspection apparatus 2 Stage 3 Electron gun 4 Detector 5 Electron beam scanning control part 6 Stage drive control part 7 Control part 10 Signal processing part 11 Average intensity calculation part 12 Defect intensity calculation part 12a Function calculation part 12b Extreme value calculation part 13 Defect intensity average value calculation unit 14 Storage unit 15 Storage unit 16 Storage unit 20 Defect detection unit 100 Substrate 101 Panel 102 Pixel 104 Intensity distribution 110 Detection point 111 Calculation point 112A, 112B, 112C Extraction range 121 Intensity distribution curve 122 Intensity distribution curve Q Local minimum

Claims (10)

Driving the array by applying an inspection signal of a predetermined voltage to the TFT substrate, detecting secondary electrons obtained by irradiating the TFT substrate with an electron beam, and based on the signal intensity of the detection signal of the secondary electrons In TFT substrate array inspection to inspect the array of substrates,
The array inspection of the TFT substrate is performed by defect detection using a defect intensity calculated from the signal intensity of the detection signal for pixels corresponding to the array,
The calculation of the defect strength is
On the TFT substrate, a plurality of detection signals per pixel are obtained by irradiating an electron beam to a plurality of locations for one pixel to detect secondary electrons,
Set an extraction range for extracting a detection signal used for calculation of the intensity distribution of the detection signal from the plurality of detection signals,
For each pixel, in at least three extraction ranges of the extraction range including the pixel and at least two extraction ranges including a part of the pixel, the average intensity of the detection signal included in each extraction range is calculated,
Obtain a function that approximates at least three points with the calculated average intensity and a position representing the extraction range as coordinate points,
Find the extreme value of the function,
A defect intensity calculation method, wherein the obtained extreme value is used as a defect intensity used for defect detection of the pixel.   The defect strength calculation is repeated a plurality of times for the same substrate, the defect strength obtained in each calculation is averaged to calculate a defect strength average value, and the defect is detected using the calculated defect strength average value. The defect strength calculation method according to claim 1. The extraction range includes a total of three extraction ranges including the pixel and two extraction ranges adjacent to the pixel in the x-axis direction, the y-axis direction, or the oblique direction on the TFT substrate. Yes,
The defect strength calculation method according to claim 1, wherein the function is a quadratic function passing through three points representing each extraction range. The extraction range is
An extraction range including the pixel, and four extraction ranges adjacent to the pixel in the x-axis direction and the y-axis direction on the TFT substrate;
Or
An extraction range including the pixel, and eight extraction ranges adjacent to the pixel in the x-axis direction, the y-axis direction, and the diagonal direction on the TFT substrate;
The defect strength calculation method according to claim 1, wherein the function is a function that approximates a curved surface that passes through five points or nine points representing the extraction range. The extraction range is a range for extracting four detection signals arranged in a rectangular shape,
The defect intensity calculation method according to claim 1, wherein the average intensity is obtained from an average value of signal intensity of the four detection signals. Driving the array by applying an inspection signal of a predetermined voltage to the TFT substrate, detecting secondary electrons obtained by irradiating the TFT substrate with an electron beam, and based on the signal intensity of the detection signal of the secondary electrons In a TFT substrate array inspection apparatus for inspecting an array of substrates,
An average intensity calculator that calculates an average intensity of the detection signal within a predetermined extraction range from the signal intensity of the detection signal;
A defect intensity calculation unit that calculates the defect intensity from the signal intensity of the detection signal, and a defect detection unit that performs defect detection using the calculated defect intensity for pixels corresponding to the array,
On the TFT substrate, a detection signal is obtained by irradiating an electron beam to a plurality of locations per pixel to detect secondary electrons,
The average intensity calculator
Set the extraction range to extract the detection signal used to calculate the intensity distribution of the detection signal from multiple detection signals per pixel,
For each pixel, in at least three extraction ranges of the extraction range including the pixel and at least two extraction ranges including a part of the pixel, the average intensity of the detection signal included in each extraction range is calculated,
The defect strength calculator is
Determining a function that approximates at least three points with the calculated average intensity and a position representing the extraction range as coordinate points, obtaining an extreme value of the function, and using the obtained extreme value for defect detection of the pixel It is characterized by calculating as
TFT array inspection device. It has a defect strength average value calculation unit,
The average strength calculator and the defect strength calculator repeat the calculation of the defect strength multiple times for the same substrate,
The defect strength average value calculating unit calculates a defect strength average value by averaging the defect strength obtained by the calculation performed repeatedly, and detects a defect using the calculated defect strength average value. The TFT array inspection apparatus according to claim 6. The extraction range includes three extraction ranges including the extraction range including the pixel and two extraction ranges adjacent to the pixel in any of the x-axis direction, the y-axis direction, and the oblique direction on the TFT substrate. Range,
The TFT array inspection apparatus according to claim 6, wherein the function is a quadratic function passing through three points representing each extraction range. The extraction range is
An extraction range including the pixel, and four extraction ranges adjacent to the pixel in the x-axis direction and the y-axis direction on the TFT substrate;
Or
An extraction range including the pixel, and eight extraction ranges adjacent to the pixel in the x-axis direction, the y-axis direction, and the diagonal direction on the TFT substrate;
The TFT array inspection apparatus according to claim 6, wherein the function is a function that approximates a curved surface that passes through five points or nine points that represent the extraction range. The extraction range is a range for extracting four detection signals arranged in a rectangular shape,
10. The TFT array inspection apparatus according to claim 6, wherein the average intensity is obtained from an average value of the signal intensities of the four detection signals. 11.

Images