JP4287863B2 - Review SEM - Google Patents

Description

  The present invention relates to a method and an apparatus for manufacturing an element having a fine pattern on a substrate, and more particularly to an inspection apparatus and an inspection method for a fine pattern such as a circuit in a semiconductor device or a liquid crystal. The present invention relates to a pattern inspection apparatus, an inspection method, and a semiconductor manufacturing method using the same.

  A semiconductor wafer inspection will be described as an example.

  A semiconductor device is manufactured by repeating a process of transferring a pattern formed mainly on a photomask on a semiconductor wafer by lithography and etching. In the manufacturing process of semiconductor devices, the quality of lithography processing, etching processing, and other various processing processes, and the generation of foreign matter greatly affect the yield of semiconductor devices. Therefore, the manufacturing process is performed by detecting abnormalities and defects early or in advance. In order to provide feedback, a method for inspecting a pattern on a semiconductor wafer in the manufacturing process has been conventionally performed.

  As a method for inspecting defects existing in patterns on a semiconductor wafer, a defect inspection apparatus that irradiates a semiconductor wafer with white light and compares the same kind of circuit patterns in a plurality of LSIs using an optical image has been put into practical use. For example, in an inspection method using an optical image, as described in Japanese Patent Laid-Open No. 3-167456, an optically illuminated region on a substrate is imaged by a time delay integration sensor, and the image is input in advance. A method of detecting a defect by comparing with design information is disclosed.

  In addition, it detects the diffracted light or scattered light by irradiating laser light, and discriminates the diffracted light from the regular circuit pattern from the scattered light from irregularly shaped foreign matters or defective parts to detect only foreign matters or defects. A method is disclosed in JP-A-9-138198.

  Furthermore, as circuit patterns become finer, circuit pattern shapes become more complex, and materials become more diverse, it becomes difficult to detect defects using optical images, so circuit patterns can be created using electron beam images with higher resolution than optical images. Methods for comparative inspection have been proposed. As a pattern inspection system using electron beams, Journal of Vacuum Science & Technology B, Vol. 9, No. 6, pages 3005 to 3009, 1992 (Journal of Vacuum Science & Technology B, Vol. 9, No.6, pp. 3005-3009 (1991)), Journal of Vacuum Science and Technology B, Vol. 10, No. 6, 2804-2808, 1992 (Journal of Vacuum Science & Technology B, Vol. 10, No. 6, pp. 2804-2808 (1992)), JP-A-5-258703, US Pat. No. 5,502,306 and JP-A-10-234543, Irradiate a conductive substrate (X-ray mask, etc.) with an electron beam having an electron beam current 1000 times or more (10 nA or more) of SEM, and detect any secondary electrons, reflected electrons, or transmitted electrons generated. The faith Method of automatically detecting discloses a defect by comparing an image formed from the inspection. Hereinafter, this inspection method is referred to as an electron beam visual inspection method.

  In the electron beam appearance inspection, an image having a higher resolution than that of the optical appearance inspection or laser inspection can be obtained, so that it is possible to detect minute foreign matters and defects on a minute circuit pattern. In addition, due to the potential contrast in which the surface potential difference is reflected in the secondary electron emission efficiency due to the effect of charging by electron beam irradiation, circuit pattern conduction or non-conduction on the surface or lower layer, wiring or transistor short circuit, etc. It is possible to detect electrical defects. The potential contrast and a technique using the same are disclosed in “Electron, Ion Beam Handbook” (Nikkan Kogyo Shimbun), pages 662-623.

  In the optical appearance inspection and the inspection using a laser beam, a foreign matter or a shape defect of a circuit pattern is detected. The detected foreign matter or defect needs to be observed in detail to identify the cause of occurrence. For this reason, a method is adopted in which position information of foreign matters or defects detected by various inspection apparatuses is received from a communication network or medium and observed at high magnification and high resolution using an optical microscope image, a laser microscope image, or an electron beam image. . As patterns and detected defects and foreign objects become finer, electron beam images can be observed with higher resolution and surface detail than optical microscope or laser microscope images, so the observation method using electron beam images is widely adopted. ing. This observation apparatus using an electron beam image is hereinafter referred to as a review SEM.

  The review SEM and the observation method in the review SEM are disclosed in JP-A-9-184715 and JP-A-10-135288. As described above, for foreign matters or pattern defects detected by the optical appearance inspection and laser inspection apparatus, the position to be observed can be specified by exchanging position information of these inspection apparatuses, and observed by an electron beam image. By using a review SEM that can be made, it has become possible to observe a detailed shape.

  However, in the electron beam external appearance device, even if there is no abnormality in the surface shape, an electrical defect in which the transistors are electrically shorted or disconnected in the lower layer can be detected by potential contrast. The electric defect detected by this electron beam type visual inspection is to be observed in detail with a normal SEM, the optical observation apparatus or the review SEM described above, but the potential contrast cannot be obtained and the observation cannot be performed. There was a problem. In order to solve this problem, the above-described conventional technique has not proposed a review SEM that achieves both high-resolution observation of foreign matter and pattern shape defects and detailed observation of electrical defects that can be observed by potential contrast.

JP-A-3-167456 JP-A-5-258703 US Pat. No. 5,502,306 JP-A-10-234543 JP-A-9-184715 JP-A-10-135288 Journal of Vacuum Science & Technology B, Vol. 9, No. 6, pp. 3005-3009, 1992 (Journal of Vacuum Science & Technology B, Vol. 9, No. 6, pp. 3005- 3009 (1991)) Journal of Vacuum Science & Technology B, Vol. 10, No. 6, No. 2804-2808, 1992 (Journal of Vacuum Science & Technology B, Vol. 10, No. 6, pp. 2804- 2808 (1992)) "Electron, Ion Beam Handbook" (Nikkan Kogyo Shimbun), pages 662-623

  As mentioned in the above-mentioned prior art, in addition to optical appearance inspection and laser inspection method for various fine circuit patterns including semiconductor devices etc., electron beam appearance inspection is applied, resulting in foreign matter and patterns In addition to defects such as shape defects, it has become possible to detect various defects that cannot be detected or discriminated only by the surface shape, such as electrical defects such as open / short circuits of various transistors and the like, and poor hole pattern conduction.

  In order to improve the circuit pattern manufacturing process based on the results of detecting defects with the electron beam inspection system, the details of the detected defects are observed in detail, and the details of foreign matter, shape defects, electrical defects, etc. It is necessary to identify the cause of occurrence of defects based on the observation results and classification results.

  When observing surface shape abnormalities and foreign objects using a review SEM, an SEM image obtained by irradiating the observed region under high-resolution conditions with a narrowed electron beam and detecting the generated secondary electrons The method of observing is effective. In the SEM image, the surface unevenness information is included in the reflected electrons generated when the observation object is irradiated with the electron beam. Therefore, when observing the surface height of the observation object, the unevenness information A method of detecting and observing reflected electrons including information is effective. Therefore, in order to observe the shape and material information and the surface unevenness information at the same time, a plurality of detectors for distinguishing and detecting secondary electron signals and reflected electron signals having different properties are installed and detected respectively. The method is effective. On the other hand, the observation conditions such as the electron beam current and the scanning speed suitable for observing the electrical defect based on the potential contrast are different from the conditions suitable for observing the surface shape defect and the foreign matter.

  In the above prior art, the review SEM for observing the surface shape and the shape of the foreign matter with high resolution has been described, but the observation conditions suitable for observing the electrical defects cannot be easily set. was there. Therefore, the electron optical conditions have to be set manually according to the purpose of observation, and there is a problem that adjustment takes a lot of time and labor.

  Also, in the review SEM, position information such as defects detected by other various inspection apparatuses is exchanged by communication or a medium, etc., and moved to the position of the defect portion to form an image. In order to observe, it is necessary to position with high precision. In the conventional method and apparatus, correction such as rotation is performed by alignment using a pattern of two points on the wafer to be observed, but in the exchange of position information between various apparatuses, the magnification and position of the coordinate system There has been a problem that a slight shift of information occurs and the position shifts when the defect portion is viewed.

  In addition, in various inspection apparatuses, inspections are successively performed on wafers manufactured every moment in the semiconductor manufacturing process. For this reason, the number of detected foreign matters and defects becomes enormous. Although it is necessary to grasp the detailed contents of the detected defect and the like, when all the defects on all the wafers are observed with the conventional review SEM, there is a problem that it takes an enormous amount of time. For this reason, it is necessary to select a portion to be observed for efficiently grasping the feature from a large number of defects. Furthermore, it was necessary to automatically observe and classify selected defects one after another at high speed.

  The first object of the present invention is to inspect a fine pattern such as a circuit on a substrate surface of a semiconductor device or the like by irradiating with white light / laser light or an electron beam, and detect surface irregularities or shape defects, An object of the present invention is to provide a pattern inspection apparatus and a pattern inspection method for efficiently inspecting and observing various foreign matters and defects such as foreign matters and electrical defects in a short time.

  A second object of the present invention is to provide a pattern inspection apparatus and a pattern inspection method for observing and classifying various foreign matters and defects as described above at high speed and with high accuracy.

  Further, the third object of the present invention is to apply a process to a semiconductor device of various types and multi-steps and other fine circuit patterns at an early stage so that the process failure of the semiconductor device or the like is faster and more efficient than the conventional device and method. It is an object of the present invention to provide an inspection apparatus and method that can contribute to reducing the defect rate while improving the reliability of a semiconductor device or the like by reflecting the result in manufacturing conditions.

  In a substrate having a fine circuit pattern including a semiconductor device, by performing various inspections of a method of comparing an image obtained by irradiating white light, laser light, and an electron beam with an adjacent equivalent pattern, It is possible to automatically detect foreign objects on the circuit pattern, defects in shape, and electrical open / short defects. However, what is provided by the inspection apparatus is positional information on the occurrence of a defect and the size of the defect. In order to obtain the detailed content of the defect, it is necessary to acquire a high-resolution image again at the location where the defect has occurred, and classify the content of the defect related to the shape, unevenness, material, etc. from the image information.

  Conventionally, in both a length measuring SEM and an observation SEM, in order to observe a fine shape with high resolution, a method of acquiring and adding a plurality of images at the same location while reducing the beam current applied to the sample Was forming an image. In such an image forming method, a fine shape can be observed, but it has been difficult to observe an electrical defect due to a potential contrast. That is, when an open or short defect detected by the electron beam visual inspection apparatus is observed with a review SEM, there are many cases where the defect cannot be recognized.

  When observing an electrical defect using a potential contrast, the present inventors need to charge the observation region, and in order to make charging more, it is desirable that the irradiation beam current is large. I found. However, since it is necessary to obtain a potential contrast and to perform detailed surface observation in the same observation, it is necessary to select a beam current that can maintain the resolution. It was also found that various electrical defects can be observed with the same apparatus by appropriately changing the beam scanning speed according to the contents of the electrical defects.

  In the conventional apparatus, various parameters of the apparatus are fixed only for the purpose of observing the fine shape. In order to change the optical conditions and detection conditions of electron beam irradiation, beam axis adjustment, brightness adjustment, etc. There was a problem of requiring a huge amount of man-hours. However, as already described, it is necessary to observe electrical defects as well as fine shapes with the same device, electron beam irradiation conditions for observing the shapes, conditions for observing irregularities, electrical defects It is necessary to be able to easily set the conditions for observing each.

  The present inventors set the conditions suitable for fine shape observation as “shape observation mode”, the conditions suitable for observing the uneven state as “unevenness observation mode”, and the conditions suitable for observing electrical defects. It was found that different electron beam irradiation conditions, lens conditions, signal detection conditions, etc. can be easily set by creating a file as “potential contrast observation mode” and storing / registering it.

  As the electron beam irradiation condition, in the “shape observation mode”, a small beam current is advantageous for high-resolution observation. On the other hand, in the “potential contrast observation mode”, the amount of secondary electrons reaching the detector changes according to the electrical characteristics by charging the surface, and the energy of secondary electrons generated from the sample surface changes. Therefore, since the observation is made by utilizing the phenomenon that the brightness of the SEM image differs between the normal circuit pattern and the pattern in which the defect has occurred, the beam current to be irradiated is large in order to further charge the sample surface. Is more advantageous. Furthermore, by changing the scanning speed of the electron beam according to the content of the electrical defect, it is possible to control the charged state of the sample surface and to reveal the defective part from the SEM image. In the “concave / convex observation mode”, since it is difficult to be influenced by the beam current to be irradiated and the scanning speed, it is not necessary to control the beam irradiation conditions.

  Next, the inventors have found the following about conditions for detecting a signal that is secondarily generated from the sample surface when the sample surface is irradiated with the electron beam. In the “shape observation mode”, secondary electrons generated from the sample surface are detected, and in the “unevenness observation mode”, reflected electrons are detected. In the “potential contrast observation mode”, it is possible to obtain each piece of information by detecting only secondary electrons having a predetermined energy among the secondary electrons. Therefore, when the electron beam is irradiated, the secondary electrons and backscattered electrons generated from the sample surface are detected as described above, and the information on “shape”, “concave / convex”, and “potential contrast” is revealed individually. It was found that it can be detected.

  In addition, by observing the shape by detecting secondary electrons in the “shape observation mode” above, if the reflected electrons are detected separately, “irregularities” can be observed at the same time. "When detecting secondary electrons with high energy, it was found that" irregularities "can be observed at the same time if the reflected electrons are detected separately.

  Next, since electrical defects are not detected in the optical appearance inspection and laser irradiation inspection, the “shape observation mode” is set in the defect inspection after the optical appearance inspection and laser inspection. It was found that in the defect inspection after performing the electron beam inspection to be detected, by setting the “potential contrast observation mode”, appropriate observation conditions are automatically set when observing the defect. .

  In addition, an input screen for selecting the “shape observation mode” and “potential contrast observation mode” is provided on the operation screen, and the electron beam irradiation condition and the signal detection condition are set according to the selection by selecting from the screen. In this way, it was found that the target defect can be easily observed. Furthermore, with regard to the “unevenness observation mode”, if it is selected whether or not to observe, “unevenness” can be arbitrarily observed both in the “shape observation mode” and in the “potential contrast observation mode”. .

  By carrying out these inspection methods and using an inspection apparatus having these functions, it becomes possible to observe the fine shape and the unevenness information and the electrical defect arbitrarily according to the purpose. Hereinafter, means for realizing the inspection method of the present invention will be described.

  According to the present invention, first, electron beam irradiation and signal detection conditions suitable for potential contrast observation can be set. As conditions suitable for potential contrast observation, the beam current was made relatively large, the irradiation energy was made relatively low, the scanning speed was fast for open defects, and was relatively slow for short defects. In addition, in the detection system, an energy filter is provided so that the secondary electrons generated in the normal part and the defect part due to the influence of the surface potential are different in the amount of secondary electrons. A function of detecting only a signal having an energy larger than a predetermined energy applied to is provided. As a result, compared with the conventional method in which observation is performed under the same conditions with a small beam current and without energy filtering, an electrical defect can be recognized by the potential contrast.

  Second, various electron beam irradiation conditions and signal detection conditions suitable for recognizing electrical defects described in the first means are registered as “potential contrast observation mode”, and various types suitable for fine shape observation. The electron beam irradiation conditions and signal detection conditions are registered as a “shape observation mode”, and the signal detection conditions suitable for observing the surface roughness / height state are tabulated and filed as “unevenness observation mode”. By doing so, it can be easily set under any observation condition, and almost no adjustment is required.

  Third, using the electron beam irradiation conditions suitable for observing the potential contrast described in the first means, an energy filter is provided in the signal detection system, and filtering is performed by arranging a plurality of detectors. In addition, signals of various energies are detected separately. As a result, the shape information is observed with a detection signal centered on secondary electrons, the surface roughness information is detected with a detection signal centered on reflected electrons, and the potential contrast is observed with a detection signal centered on high-energy secondary electrons. Thus, “potential contrast mode”, “shape observation mode”, and “unevenness observation mode” can be set by changing only the detection conditions under the same electron beam irradiation conditions. That is, optical axis adjustment is not necessary.

  Fourth, when alignment of position information by two dies is performed, an image of an alignment circuit pattern in the first die and an alignment circuit pattern in the second die is acquired, and positional deviation and rotation correction are performed. By obtaining not only the information but also the magnification error of the coordinate system, positioning can be performed with high accuracy even when visualizing a defect at a location away from the aligned die.

  Fifth, after performing position and rotation alignment using a predetermined pattern, further observation of a predetermined foreign matter or defect and further correction of position information from the position of the foreign matter / defect in the field of view can further increase the position. The position can be corrected accurately.

  Sixth, when inspection is performed with the inspection device and the inspection result is output to the outside, the detection result information includes the code of the device that has been inspected together with the coordinate data of the defect, and inspection with the review SEM By recognizing the type of apparatus inspected from the result data, it is possible to automatically set the observation conditions such as the “potential contrast observation mode” and the “shape observation mode”. That is, when defect data inspected by an optical visual inspection device or laser inspection device that cannot detect an electrical defect is read, the “shape observation mode” is automatically selected, and the electrical When the defect data inspected by the electron beam type visual inspection equipment that can detect general defects is read, the “potential contrast observation mode” is set according to the inspection conditions etc. This makes it easy to set conditions for implementation.

  Seventh, on the screen for setting the observation conditions in the review SEM, an input unit for selecting the observation conditions such as “potential contrast mode”, “shape observation mode”, “unevenness observation mode” is arranged in the screen and the observation is performed. By setting an input unit for selecting a location to be performed on the screen, it is possible to easily realize the automatic setting of the review.

  Eighth, in the process of automatically performing observation, image observation, and defect content classification for a plurality of defect locations in the review SEM, a means for converting the observation order so as to move one after another in the order of near defects is provided. It is.

  Ninth, on the screen for setting review conditions, a main menu for selecting inspection conditions is provided so that review conditions can be set by simple menu selection, and detailed apparatus conditions are submenus in the main menu. By not displaying directly, the setting of observation conditions was simplified. Thereby, it is possible to input main parameters for observation only by selecting a menu without complicated input operations.

  With the means described above, inspecting a substrate having a fine circuit pattern such as a semiconductor device, and inspecting the detected defect in detail, not only the conventional surface shape observation but also an electrical defect. It becomes possible to observe. And, by observing the observation conditions suitable for surface shape observation and the conditions suitable for observing electrical defects as device conditions in advance, it can be selected in the setting condition file that determines the inspection conditions. Therefore, appropriate observation conditions can be easily set without complicated optical axis adjustment. Further, by performing magnification correction by two points and position information correction by actual defects and foreign matters in alignment or the like, it is possible to display a defect with high accuracy. Furthermore, by displaying a condition file screen having an input unit for determining an observation mode and an observation location on the operation screen of the observation apparatus, setting of observation conditions can be simplified.

  Using these observation methods and observation apparatuses, by carrying out observation after inspecting a substrate having a circuit pattern, for example, a semiconductor device in the manufacturing process with an inspection apparatus, it is difficult to observe with conventional techniques. Defects can be easily observed, and furthermore, observation condition setting and device adjustment, which took an enormous amount of time for observing electrical defects, can be performed with a simple operation called review condition file selection. Observation parameter setting, which has taken enormous time in file creation, can be efficiently performed in a short time. As a result, detailed information such as pattern shape defects and fine foreign matters, electrical continuity defects and circuit shorts caused by process processing in the manufacturing process of semiconductor devices, etc. can be grasped at an early stage. It becomes possible to reveal problems that are latent in As a result, it is possible to take measures against the cause of defects in the manufacturing process of various substrates including a semiconductor device at a higher speed and with higher accuracy than the conventional method and the conventional apparatus, and at the same time, it is possible to secure a high yield, that is, a non-defective product rate, and to generate defects It is possible to shorten TAT from detection to countermeasure.

  The conventional inspection method and review apparatus can only obtain an image obtained by observing an object with an uneven surface from above, and enormous adjustment and time are required to observe a defect that can be observed by potential contrast. It was virtually impossible to simultaneously perform shape observation and potential contrast observation. On the other hand, when observing various defects generated on a substrate such as a semiconductor device having a circuit pattern using the pattern inspection apparatus of the present invention, the electrical defects are observed under appropriate conditions in the potential contrast observation mode. I was able to do that. As a result, the contents of each of the shape defect and the electrical defect can be distinguished and classified. As a result, it is possible to grasp the contents of defects generated on the wafer to be inspected, and to obtain information necessary for specifying a process or an apparatus to be countermeasured.

  In the above review, by applying an automatic sequence and review condition setting function, it becomes possible to automatically review a plurality of wafers one after another unattended, and the above review can be executed efficiently.

  Furthermore, by providing a function to automatically select the observation mode, such as the shape observation mode after the optical appearance inspection and the potential contrast mode after the electron beam appearance inspection, the user can make complicated selections and operations. Appropriate observation conditions can be set without executing this function. In addition, by selecting a portion to be reviewed before the review by the review sampling function, it becomes possible to effectively obtain information on the entire wafer even when observing and classifying the same number of review points.

  Because of these effects, the contents of defects generated on the wafer can be grasped efficiently at high speed and with high accuracy. By applying this inspection to the substrate manufacturing process, the above conventional technology was inefficient, so the observation time It is possible to solve problems such as long defects or defects that cannot be observed, and to grasp the problem accurately and immediately. The defect rate of the substrate can be reduced and the productivity can be increased. In addition, by applying the above inspection, the occurrence of an abnormality can be detected quickly, and measures can be taken earlier than before, so that a large number of defects can be prevented in advance, and as a result, Since the occurrence of defects itself can be reduced, the reliability of semiconductor devices and the like can be improved, the development efficiency of new products and the like can be improved, and the manufacturing cost can be reduced.

  Hereinafter, a pattern inspection apparatus and a pattern inspection method of the present invention will be described in detail with reference to the drawings.

  In this embodiment, an inspection method for forming an image using an electron beam for a circuit pattern that has already been inspected by another inspection apparatus, and classifying the presence / absence of a defect on the circuit pattern and the content of the defect from the image, The inspection apparatus will be described. Here, as an example, a case where a circuit pattern of a semiconductor device formed on a wafer is inspected will be described.

  First, FIG. 1 shows an outline of a pattern inspection system using the pattern inspection apparatus according to the present invention. The pattern inspection system 1 receives information such as an inspection device 2 for finding an unknown foreign matter or defect, coordinates of the defect found by the inspection device 2, etc., and details the contents of the defect to observe an image or observation of the defect The review device 3 that outputs the classification result according to the above, similarly receives the information such as the coordinates of the defect, the analysis device 4 that performs cross-section processing and various analysis analyzes thereafter, the tester 5 that performs the wafer operation test, the image and the coordinates It is divided into a data collection and analysis system 6 that mediates and manages data. Examples of the inspection device 2 include devices such as a laser type foreign matter inspection device 7, an optical appearance inspection device 8, and an electron beam appearance inspection device 9. Examples of the review device 3 include an optical review device 10 and a review SEM 11. Examples of the analysis apparatus 4 include an FIB processing apparatus 12, a high resolution SEM 13, an AES analysis apparatus 14, a TEM 15, and the like. Examples of the tester 5 include a memory tester 16 and a logic tester 17. In the data collection and analysis system 6, a server 18 and a data storage unit 19 are included.

  For wafers in the manufacturing process, all or some of the wafers are inspected using any one of the inspection apparatuses 2. However, when inspections are regularly performed in a predetermined process or when a defect occurs, Depending on the case, the inspection may be performed in an arbitrary process. In such an inspection, when the inspection is performed using the laser type foreign matter inspection apparatus 7, foreign matters and scratches are mainly detected. When the inspection is performed by the optical appearance inspection device 8, a pattern shape defect, a defect under a thin film that transmits white light, and the like are detected in addition to the foreign matter. When inspected by the electron beam type visual inspection 9, in addition to foreign matters and pattern shape defects generated on the surface, electrical defects such as contact hole conduction failure, wiring disconnection, wiring short circuit and transistor short circuit are detected. The These detected defects are output to the server 18 of the data collection analysis system 6 and stored in the data storage unit 19. Since each device of the inspection apparatus 2 has a function of searching for a defect location, the output data is inspected by the number of defects, the coordinate information of each defect, the size of the defect, and each inspection device of the inspection apparatus 2. This is information such as the inspection conditions and the like.

  In order to take measures against defects occurring in the manufacturing process, it is necessary to grasp the detailed contents of the defects in order to identify the cause of the defects, and for this purpose, it is necessary to obtain information such as defect images or defect contents. Therefore, the process of observing the defective part detected by the review apparatus 3 is required. When reviewing a defective portion with the review device 3, the coordinate data of the defect obtained by the inspection device 2 is read via the server 18. The optical review device 10 can obtain images and classification results by an optical microscope. However. In order to grasp the contents of fine foreign matters and defects on a fine circuit pattern, observation is performed with a review SEM 11 with higher resolution. Details of the configuration of the review SEM 11 and the observation method will be described later.

  In the analysis device 4, by processing a cross section or a plane using the FIB 12, observing the surface or cross section with a high resolution SEM 13 or TEM 15, or analyzing the composition of the component at a desired location by the AES analysis device 14 or the like, Further, a fine structure can be analyzed. These analyzes are performed by cutting out the wafer, and the analyzed wafer is destroyed. Moreover, since the analysis location is cut out and analyzed, the structure and coordinate system of the sample stage are greatly different. Therefore, it is difficult to display the field of view as it is based on the coordinate information of the defect detected by the inspection apparatus 2. That is, although these analysis apparatuses 4 can perform detailed analysis, the analysis target wafer is destroyed and the analysis time is required for a long time. Therefore, a wafer that has been defective and is extracted becomes an object of analysis. Further, the tester 5 conducts an electrical test on each die of the completed wafer, examines various operation states, and selects them.

  The inspection system 1 collects data on the result of inspecting the wafer in such a manufacturing process by the inspection device 2, the result of examining the contents of foreign matter / defects by the review device 3, and the result of conducting the electrical test by the tester 5 Factors that suppress the yield by analyzing and analyzing the malfunction and yield status of the tester 5 and the inspection status of various processes, the contents of the generated foreign matter and defects, collected by the analysis system 6 Identify processes and devices that require reduction, predict yield, etc. Accordingly, it is important to classify the defect detected and detected by the inspection apparatus 2 by acquiring an image including detailed information of the defect with high accuracy, high speed, and high efficiency.

  The conventional review SEM 11 was intended to observe a fine shape with high resolution by reducing the electron beam current to a small size. For this reason, it was possible to observe with high resolution foreign matter and defects having irregularities on the surface detected by the laser type foreign matter inspection apparatus 7 or pattern shape defects detected by the optical appearance inspection device 8. However, electrical defects of the semiconductor transistor such as contact hole conduction failure, wiring disconnection, wiring short circuit, and transistor short circuit detected by the electron beam visual inspection 9 could not be observed. In the review apparatus 3, since it is used for the purpose of observing the detected defect part in detail and classifying and outputting the contents of the defect, whether or not the defect detected by another inspection apparatus is a false detection. It is important to recognize what the content of the defect was. Therefore, when considering the inspection results of the present embodiment, foreign matter (including identification of the type of foreign matter), pattern shape defect (including disconnection / short circuit, scratch, shape abnormality, etc.), contact hole non-conduction, lower layer transistor A review SEM and a review (inspection) method that make it possible to recognize the defect contents from the SEM image and to classify them by distinguishing them will be described.

  Hereinafter, the review SEM 20 and the review method in the present invention corresponding to the review SEM 11 of the inspection system 1 described in FIG. 1 will be described below.

  FIG. 2 shows an example of the pattern inspection apparatus according to the present invention, and shows an example of the configuration of the review SEM 20. The apparatus includes an electron optical system 21, a stage mechanism system 22, a wafer transfer system 23, a vacuum exhaust system 24, an optical microscope 25, a control system 26, and an operation unit 27. The electron optical system 21 includes an electron gun 28, a condenser lens 29, an objective lens 30, a first detector 31, a second detector 32, a third detector 33, an energy filter 34, a deflector 35, and a reflector 36. The wafer height detector 37 is configured. The stage mechanism system 22 includes an XY stage 38, a holder 39 for placing a wafer as a sample, a holder 39, and a retarding power source 40 for applying a negative voltage to the wafer 51. A position detector by laser length measurement is attached to the XY stage 38. The wafer transfer system 23 includes a cassette mounting portion 41 and a loader 42, and a holder 39 moves back and forth between the loader 42 and the XY stage 38 with the wafer 51 mounted thereon. The control system 26 includes a signal detection system control unit 43, a beam deflection correction control unit 44, an electron optical system control unit 45, a wafer height sensor detection system 46, and a mechanism / stage control unit 47. The operation unit 27 includes an operation screen and operation unit 48, an image processing unit 49, and an image / inspection data storage unit 50.

  Next, the operation of each unit in FIG. 2 will be described using the flow shown in FIG. This flow is a flow for automatically executing a review.

  First, in step 58, the wafer cassette on which the wafer 51 is placed on an arbitrary shelf is placed on the cassette mounting portion 41 in the wafer transfer system 23 of FIG.

  Next, in step 59, in order to designate the wafer 51 to be reviewed from the operation screen 48, the shelf number in the cassette on which the wafer 51 is set is designated. In the review, as already described, since inspection is performed by another inspection apparatus and observation by an electron beam image is executed based on inspection result information including positional information such as defects, the operation screen 48 is used. Select the inspection result file. In the selection, it is possible to read an inspection result file by communication via a network or the like, or to read an inspection result file from a medium such as FD or MO. In any case, by specifying the inspection result file name, various data of the inspection result can be read into the data input unit 56 and converted into the data format and coordinate system used in the review SEM 20 by the data conversion unit 57. The Further, the review condition file name is input from the operation screen 48. This review condition file is configured by combining various parameters for determining the content of the review. The contents and setting method of the review condition file will be described later. When reviewing a plurality of wafers automatically and continuously, an inspection result file name and a review condition file name are input for each wafer to be inspected in the same manner as described above. This condition setting can be arbitrarily executed up to 25 sheets × 2 cassettes, and continuous automatic review can be performed. The input of conditions necessary for executing these reviews is completed, and in step 60, an automatic review sequence is started.

  In step 61, when the automatic review is started, first, the set wafer 51 is transferred into the review apparatus. In the wafer transfer system 23, the holder 39 on which the wafer 51 is placed has the same size and shape as the wafer 51 regardless of whether the diameter of the wafer to be inspected is different or the wafer shape is different, such as an orientation flat type or notch type. It can be handled by exchanging them together. The wafer to be inspected is placed on the holder 39 by a wafer loader 42 including an arm, a preliminary vacuum chamber, etc. from the cassette, held and fixed, and transferred to the inspection chamber together with the holder.

  In step 62, when the wafer 51 is loaded, the electron beam irradiation condition is set to each part by the electron optical system controller 45 based on the input review condition. Then, an electron beam image of a predetermined portion of the wafer 51 is acquired, and the focus / astigmatism is adjusted from the image. At the same time, the height of the wafer 51 is obtained from the height detector 37, the correlation between the height information and the in-focus condition of the electron beam is obtained, and the wafer height is not executed every time when acquiring the electron beam image thereafter. It automatically adjusts to the in-focus condition based on the detection result. Thereby, high-speed continuous electron beam image acquisition became possible.

  When the electron beam irradiation conditions and the focus / astigmatism adjustment are completed, in step 63, alignment is performed using two points on the wafer. FIG. 4 is further used to show the alignment method in the review SEM 20 in the present embodiment. In order to perform the automatic alignment necessary for automatically executing the review, using a wafer having a pattern equivalent to the wafer to be observed in advance, the chip to be aligned, the pattern formed for alignment, or suitable for alignment The preset coordinates and dictionary information can be read in the optical microscope image and SEM image of the pattern and the review condition input in step 59. In the review, it is necessary to view the position of the defect with high accuracy. For this reason, alignment is automatically performed using the alignment conditions and the alignment image registered in advance by the above method before visualizing the defective portion.

  As described above, the optical microscope image 75 and the electron beam image 76 at the designated location are registered in the memory as the location where the alignment pattern exists in advance, and the image name and the two coordinates (X1) of the designated alignment target are registered. , Y1), (X2, Y2) are registered in the review condition file. In the alignment, if there is an alignment pattern on the first chip 73, the sample stage is moved to the coordinates registered on the inspection condition file, an image is first acquired with an optical microscope, and an already registered image is obtained by image processing. A part that matches the optical microscope image is automatically searched, and if detected, the coordinates of the detection point are calculated by calculation. An electron beam image 76 of the same location is acquired based on the detected coordinates, and a location that matches the registered electron beam image is automatically searched by image processing. If detected, the coordinates (X1, Y1) of the detection point are obtained. Calculated by calculation and stored as the first coordinate. Next, with respect to the second chip 74 at a parallel position on the circuit pattern matrix, the stage moves to a position where an alignment pattern is likely to exist. Therefore, as in the first point, the optical microscope image 75 ′ and the electron beam image 76 ′ are searched for a location that matches the registered image by image processing, and the coordinates (X2, Y2) of the detected location are calculated by calculation. And stored as the second coordinate. The expected position of the second point calculated from the alignment mark position (X1, Y1) on the first chip and the chip size data, and the coordinate position of (X2, Y2) obtained by actually acquiring the image The rotation amount θ of the circuit pattern array on the substrate to be inspected with respect to the stage movement direction is obtained from the coordinate shift between the two points, that is, the shift amount in the X direction and the Y direction. Determine the correction amount. Further, the position of the alignment pattern (X1, Y1; X2, Y2) is detected, the offset value from the coordinate value and the magnification of the coordinate system are calculated, and the inspection pattern and the inspection start point are defined as a circuit pattern on the wafer to be inspected. It is possible to detect according to the position. Conventionally, only the rotation amount θ of the pattern array is calculated in this alignment, and the magnification correction of the coordinate system has not been performed. As a result, in the field-of-view of the defect after the alignment is performed, the defect comes off the center of the field of view as the distance from the alignment execution chip increases. In this embodiment, the position offset and the magnification error in the X direction are calculated not only from the rotation amount θ but also from the shift amount of the pattern coordinates, and the magnification error in the Y direction is also calculated from the magnification error. The problem of falling off was solved, and it became possible to observe defects with high accuracy.

  When the alignment is completed, in step 64, the rotation and coordinate values are corrected based on the alignment result, and the position is moved to the position of the defect to be reviewed based on the various information of the already read inspection result file. A method for selecting a defect to be reviewed will be described later. After the movement of the XY stage 38 is completed, a highly accurate position is obtained from a laser length meter attached to the XY stage 38, and based on this stage position information, an electron beam irradiation is performed by the beam deflection control 44 so that the defect becomes the center of the visual field. The position is corrected and a beam is irradiated onto the wafer 51. Taking into account the coordinate error of other inspection devices, the defect is in the field of view, so the first image is acquired at a relatively low magnification. For example, a low-magnification image is acquired with a visual field size of about 10 to 15 μm □. In the present embodiment, in the automatic review sequence, the observation magnification is set to a fixed magnification registered in advance in the review condition file.

  In step 65, when an image of a portion corresponding to the defect portion coordinates is acquired, the image is transferred to the image processing unit 49 via the signal detection system 43. After that, it moves to an adjacent chip, acquires an image of the same pattern in the adjacent chip under the same conditions, and similarly transfers the image to the image processing unit 49. The image processing unit 49 performs alignment and image comparison of these two images, and recognizes the detailed position and defect size of the defective part. Based on this result, it moves again to the coordinates of the defective portion, and adjusts the electron beam irradiation position by the beam deflection correction 43 so that the defective portion becomes the center based on the defect position information in the image. Get an image. For example, in the case of image reacquisition, an image with a visual field size of about 1 to 3 μm □ is acquired. In the present embodiment, as described above, the observation magnification is set to a fixed magnification registered in advance in the review condition even in high magnification acquisition. In addition to the method of the present embodiment, for example, when recognizing a defective part in the image acquisition at the low magnification in step 64, the size of the defect can be grasped with high accuracy at the same time. Accordingly, it is possible to obtain the defect size and automatically change the magnification so that the defect has an area of, for example, 69% of the image.

  In step 66, the acquired high-magnification image is stored in the image / data storage unit 50 as necessary. Set whether to save in advance or not in the review condition file. Moreover, it is possible to simultaneously store a plurality of types of images from a plurality of detectors according to the setting as required. For example, it is possible to simultaneously store an image by secondary electrons detected by the second detector 32 and an image by reflected electrons detected by the first detector 31.

  In step 67, at the same time that the image is stored, the image processing unit 49 extracts defect features from the image information and automatically classifies the defect contents. The classified result is encoded into a numerical value of, for example, 0 to 255, and the code number is written in a location corresponding to the defect classification code in the inspection result file. As a result, the code indicating the classification result is automatically written for the reviewed defect in the inspection result file.

  As shown in step 71, a series of operations of moving to the defect location → acquisition of low magnification → defect portion recognition → acquisition of high magnification → acquisition of high magnification → storage of defect → defect classification → writing classification result to inspection defect file is reviewed Repeat several minutes. In step 68, when the above-described series of operations are completed for all the defects designated for review on one wafer, the inspection result file (file in which the classification result is written) of the wafer is automatically saved and designated. Then, the inspection result file is output, and in step 69, the wafer is unloaded.

  Thereafter, a series of operations from the wafer loading to the inspection result file output and the unloading are repeated for the designated number of wafers as shown in step 72, and when the review is completed for all the wafers designated for review, At 70, the automatic sequence ends.

  A method for setting the review conditions in the automatic review sequence will be described below. The review condition file includes an observation mode configured by electron beam irradiation conditions and signal detection conditions for determining the target image quality, review sampling including the number of review points, and ADR (Auto Defect Review) for automatically performing the review. ) / ADC (Auto Defect Classification) conditions. ADR / ADC conditions include alignment images, observation magnification (low magnification, high magnification), electron beam irradiation conditions in each mode (electron beam current, condenser lens conditions, objective lens conditions, detectors, energy filter conditions, etc. ), Automatic classification conditions (image processing conditions), image storage conditions, and the like.

  Among the various review conditions, the contents of the observation mode for determining the image quality to be acquired, that is, the electron beam irradiation conditions and the detection conditions will be described with reference to FIGS.

  As described above, when observing an electrical defect using a potential contrast, the present inventors need to charge an observation region, and in order to charge more, a beam current to be irradiated is large. I found that it is preferable. However, since it is necessary to obtain a potential contrast and to perform detailed surface observation in the same observation, it is necessary to select a beam current that can maintain the resolution. For this reason, it was found that various electrical defects can be observed with the same apparatus by appropriately changing the beam operating speed in accordance with the contents of the electrical defects.

  In the conventional apparatus, various parameters of the apparatus are fixed only for the purpose of observing the fine shape. In order to change the optical conditions and detection conditions of electron beam irradiation, beam axis adjustment, brightness adjustment, etc. There was a problem of requiring a huge number of man-hours. However, as described above, in order to classify and grasp the contents of defects detected by various inspection apparatuses, it is necessary to observe electrical defects as well as fine shapes with the same apparatus. Therefore, it was necessary to be able to easily set the electron beam irradiation conditions, the conditions for observing irregularities, and the conditions for observing electrical defects.

  The inventors set the conditions suitable for fine shape observation as “shape observation mode”, the conditions suitable for observing the uneven state as “unevenness observation mode”, and the conditions for observing electrical defects as “ It has been found that different electron beam irradiation conditions, lens conditions, signal detection conditions and the like can be easily set by creating a file as “potential contrast observation mode” and storing / registering it. Further, as the electron beam irradiation condition, in the “shape observation mode”, a small beam current is advantageous for high-resolution observation. In the “potential contrast observation mode”, the amount of secondary electrons that reach the detector changes according to the electrical characteristics by charging the surface, and the energy of secondary electrons generated from the sample surface also changes. Since the observation is made by utilizing the phenomenon that the brightness of the SEM image differs between the normal circuit pattern and the circuit pattern in which the defect has occurred, it is advantageous that the beam current irradiated to increase the charge of the sample surface is larger. is there. Furthermore, depending on the content of the defect, by changing the scanning speed of the electron beam, it is possible to control the surface charging state and reveal the defective portion from the SEM image. On the other hand, since the “irregularity observation mode” is not easily affected by the beam current to be irradiated and the scanning speed, it is not necessary to control the beam irradiation conditions.

  Next, the present inventors have found that the conditions for detecting a signal generated secondarily from the sample surface when the electron beam is irradiated onto the sample surface are different. In “shape observation mode”, secondary electrons generated from the sample surface are detected, and in “unevenness observation mode”, reflected electrons are observed. In the “potential contrast observation mode”, it is possible to obtain each piece of information by detecting only secondary electrons having an energy higher than a predetermined energy among the secondary electrons. Therefore, when the electron beam is irradiated, the secondary electrons and reflected electrons generated from the sample surface are individually detected as described above, so that information on “shape”, “concave / convex”, and “potential contrast” can be individually obtained. It has been found that it can be detected and revealed.

  In addition, by observing the shape by detecting secondary electrons in the “shape observation mode” above, if the reflected electrons are detected separately, “irregularities” can be observed at the same time. It was found that “irregularities” can be detected at the same time by detecting reflected electrons separately when detecting secondary electrons with high energy.

  FIG. 5 divides the image quality of electron beam images into three types according to these results: shape observation mode (secondary electron image), potential contrast observation mode (potential contrast image), and unevenness observation mode (reflection electron image). In each case, conditions suitable for observation are shown. First, the secondary electron image is suitable for observation of materials and shapes, and is suitable for observation of ordinary foreign matters and shape defects. When observing this secondary electron image, it is desirable that the current of the irradiated primary beam is in the range of several pA to several tens of pA. On the other hand, when observing a potential contrast image, the difference in the amount and energy of secondary electrons that reach the detector will result in a contrast if the surface has a potential difference due to charging. It needs to be about 10 pA to 200 pA. In the case of a reflected electron image suitable for concavo-convex observation, it is not necessary to specify a beam current.

  Further, the scanning speed of the electron beam is such that, in normal shape observation, images of TV scans are displayed by adding tens of images, for example, 64 images. Images can be obtained. Therefore, a secondary electron image for shape observation is acquired by adding several tens of images by TV scanning. On the other hand, in the potential contrast image, the inventors have found that an appropriate scanning speed varies depending on the type of defect of the object to be observed. For example, in the case of an open defect in a circuit in which a transistor or a contact hole is not conductive, it can be observed as an image obtained by adding several tens of images by the same TV scan as in the case of shape observation.

  However, in the case of a defect in which the transistors and the transistor and the wiring are short-circuited, the difference between the normal part and the defective part is more likely to be caused by irradiating the electron beam at a low scanning speed. For example, by adding 8 images with a slow scan 8 times slower than the TV scan, an S / N image equivalent to the image with the TV scan added 64 images can be obtained, and the contrast of the short defect portion can be reduced. Obtainable. Therefore, in the potential contrast mode, when the content of a defect is expected, one of the above scanning speeds is set for observation. Further, the scanning speed is set according to the type of defect to be observed. Alternatively, when a defect cannot be recognized through observation with a TV scan, the image is acquired again at a different scan speed. Further, in the mode for observing the shape, the electron beam irradiation energy at the time of observation is set from 800 V to 1 KV. On the other hand, in the potential contrast observation mode, since the low acceleration beam is more likely to give a potential contrast, the irradiation energy is set to 500V to 800V. In order to set irradiation energy, it can control by changing the retarding power supply 40 applied to a sample stand.

  Next, in order to individually detect secondary electrons, high-energy secondary electrons and reflected electrons for potential contrast observation corresponding to the above various observation modes, separate signal detection systems are used according to the characteristics of each signal. It was made to detect with. First, in the normal secondary electron image, the energy of the secondary electrons is about several volts, but in the case where a potential contrast is formed, the energy of the charged potential on the surface is added. Further, the reflected electrons have energy substantially equal to the irradiation energy. As described with reference to FIG. 2, the review SEM 20 has three detectors. The first detector 31 is for detecting reflected electrons with high energy. An energy filter 34 is provided between the second detector 32 and the third detector 33, and is applied by applying a retarding potential of several tens to several tens of volts to the energy filter 34. Electrons having an energy higher than the detected potential are detected by the third detector 33, and electrons having lower energy are detected by the second detector 32. Therefore, a normal secondary electron image and a potential contrast image are separated at the same time. An image can be formed by detecting from the detector. Further, reflected electrons including surface irregularity information can be simultaneously detected by separate detectors to form an image.

  Therefore, in the review SEM according to the present invention, the above-mentioned electron beam irradiation conditions and detection system conditions (selection of detectors, etc.) can be automatically executed by selecting the observation mode as the review conditions. That is, the electron beam irradiation conditions and signal detection conditions suitable for shape observation, the electron beam irradiation conditions and signal detection conditions suitable for potential contrast observation, and the signal detection conditions for backscattered electron detection in the above two modes are set in advance. Find it after adjustment and register it. When a desired image type is selected in the review condition, the selected condition is automatically set.

  The detailed contents of the observation mode and the image quality acquired thereby have been described so far. Next, other review condition setting methods and operation screens will be described with reference to FIG.

  FIG. 6 shows an operation screen display when setting the review conditions. When performing the review, as indicated by reference numeral 77, first, the cassette containing the wafer to be observed and the shelf number in the cassette are selected. When selecting, it is only necessary to click with a mouse or the like the cassette location containing the wafer to be inspected. Next, a defect file when the wafer has already been inspected is selected from the selection dialog 78. At this time, in order to set after confirming the content of the selected defect file, for example, the number of defects, the defect distribution, etc., the review SEM 20 displays the distribution and summary on the screen 79 by clicking once with a mouse or the like. Added to click once. This confirmation and setting method is the same even if, for example, a method of displaying an outline by right-clicking and setting by left-clicking is used. Then, a review recipe that determines the content of the review is selected. As in the case of a defective file, the selection method is selected from files registered in advance in the dialog 80. Also in this case, the review recipe outline is displayed on the screen when clicked once with a mouse or the like so that it can be set after confirming the contents of the review recipe, and is set when clicked twice. In the review recipe, the defect observation mode 81 described so far is set. A sampling method for selecting the number of defects to be reviewed and a defect to be observed is set. In addition, as detailed review conditions, alignment images for executing automatic alignment, observation magnification (when observing at a fixed magnification, low magnification and high magnification observation magnification, when making it variable according to defect size rather than fixed) , Low magnification and defect ratio in the screen can be input), automatic defect classification conditions (for example, parameters such as the contents of defects that have been learned and matching rate), image storage conditions (number of images to be stored and types of images) ) And the conditions of electron beam irradiation and detection system can be confirmed and changed by including in this review detailed condition.

  In this way, it is possible to easily set the wafer to be reviewed from the screen, the defect coordinate data to be reviewed, and the conditions for executing the review, and also select the contents of the recipe. Since it is a system, it can be easily created without complicated input and setting, and can be managed as a file.

  Further, when setting a review recipe, it is necessary to designate a defect observation mode. For example, it is known in advance that a potential contrast defect cannot be detected in an optical appearance inspection apparatus or a laser scattering inspection apparatus. Yes. Therefore, in the review using the inspection result of the optical inspection apparatus or the laser inspection apparatus, the shape defect observation mode can be automatically set. Similarly, in the electron beam appearance inspection, there is a high possibility that defects detected by the potential contrast are included. Therefore, in the review using the inspection result of the electron beam appearance inspection apparatus, the potential contrast observation is performed. Make sure the mode is set. Since the defect file normally contains the code of the inspection apparatus, this makes it possible to automatically determine the observation mode when the defect file of the inspection result is selected.

  Further, in the review SEM 20, in order to improve the speed of review, when observing a plurality of defects, the order of observation is selected in the closest order. FIG. 7 is a diagram showing an example of the review order. Conventionally, for example, as shown in FIG. 7A, defects are observed in the order in which defects are detected or in order for each chip. However, in order to automatically review more than 500 defects per hour, for example, one after another, it is necessary to minimize the moving distance and moving time of the stage, as shown in FIG. Thus, it moves to the nearest defect. This makes it possible to review faster than the conventional method.

  Conventionally, the presence or absence of a defect is not detected until an electrical test is performed on the wafer or chip for which the manufacturing process has been completed by the tester 5, and in the analysis for identifying the cause of the defect, the analysis apparatus 4 is enormous. The meeting was held over time, but by applying the electron beam visual inspection device 9 and the review SEM 20 in the inspection system 1, it becomes possible to detect electrical defects early in the manufacturing process, and The review SEM 20 makes it possible to grasp the contents of defects at high speed and with high accuracy. Therefore, conventionally, it took several months from the occurrence of a defect until it was detected and countermeasures were taken, but by applying the inspection apparatus and inspection method of the present invention, one to several days after the occurrence of the defect. As a result, the efficiency of semiconductor development and manufacturing has been greatly improved.

  Furthermore, as a result of classifying the defects by the review SEM 20, it is possible to check the inspection conditions under which the inspection is performed in the electron beam type visual inspection apparatus 9. In other words, since the presence or absence of erroneous detection in the detection result and the size of the detected defect can be recognized with high accuracy, the defect content and sensitivity detected by the electron beam visual inspection apparatus 9 can be monitored, and the detection sensitivity When a problem occurs in the inspection apparatus 9, it can be fed back to the inspection apparatus 9.

  Using the inspection apparatus and inspection method described so far, based on the position information of the defect detected by another inspection apparatus, an electron beam image of the defective part is acquired, the contents are classified, and the image is stored. In the apparatus apparatus, it is possible to efficiently perform reviews and operations for determining review conditions. In addition, a potential contrast image that has been difficult to observe can be easily set by setting a mode. As a result, it has been difficult to observe so far, and as a result, information on electrical defects, shape defects, irregularities, and the like can be distinguished, and the contents of defects generated on the wafer can be accurately grasped. In addition, since it is possible to automate the flow until the results are output as distinguished as described above, high-speed automatic observation can be performed, and reviews can be performed several hundred times faster than conventional reviews. It became possible to carry out.

  Furthermore, by selecting wafers to be reviewed, selecting defect files containing defect location information inspected by other inspection equipment, and selecting and setting review conditions, it is possible not only for automatic reviews but also for setting review conditions. Simplification and high speed have been achieved, so when setting review conditions for a large number of semiconductor products in a large number of process steps, it is possible to immediately set and execute the conditions without delaying the review. Become. Therefore, the time required for the operator can be saved, the waiting time of the product is greatly shortened, and the TAT for detecting the occurrence of defects can be shortened.

  In this way, when the inspection apparatus detects variations in various manufacturing conditions and occurrences of abnormalities by performing in-line inspection methods and apparatuses for reviewing wafers inspected by various inspection apparatuses in the manufacturing process of semiconductor devices and the like, Since not only the number of defects but also what kind of content and shape defects can be grasped in detail accurately and immediately, a large number of defects can be prevented in advance. In addition, by applying the inspection apparatus and method according to the present invention, it becomes possible to determine the inspection conditions of the wafer to be inspected efficiently and accurately in a short time, and as a result, more accurate inspection can be applied, resulting in high occurrence of defects. Sensitivity can be detected. In addition, the time required to perform reviews and grasp the contents of defects can be greatly reduced, so product waiting time and operator occupation time can be shortened, and defects can be detected earlier than conventional devices and methods. Productivity can be increased.

  As described above, with regard to the typical apparatus configuration and inspection method of the present invention, the specific inspection flow, the operation of each part, the flow for determining inspection conditions, and the operation screen and operation method for inspection and inspection condition setting Although the embodiments have been described, the present invention can be applied to an inspection method and an inspection apparatus that combine a plurality of features described above without departing from the scope of the present invention.

1 is a schematic configuration diagram showing an example of a pattern inspection system according to the present invention. The figure explaining one Example of the pattern inspection apparatus by this invention. The figure which shows an example of the test | inspection flow by this invention. The figure which shows an example of the alignment method. The figure which shows the conditions for every observation mode. The figure which shows the inspection condition setting content and the operation method. The figure which shows an example of a review order.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Inspection system, 2 ... Inspection apparatus, 3 ... Review apparatus, 4 ... Analysis apparatus, 5 ... Tester, 6 ... Data collection system, 7 ... Laser type foreign material inspection apparatus, 8 ... Optical appearance inspection apparatus, 9 ... Electron beam Type visual inspection device, 10 ... optical review device, 11 ... review SEM, 12 ... FIB processing device, 13 ... high resolution SEM, 14 ... AES analyzer, 15 ... TEM, 16 ... memory tester, 17 ... logic tester, 18 DESCRIPTION OF SYMBOLS ... Data collection server, 19 ... Data storage unit, 20 ... Review SEM, 21 ... Electro-optical system, 22 ... Stage mechanism system, 23 ... Wafer transfer system, 24 ... Vacuum exhaust system, 25 ... Optical microscope, 26 ... Control system , 27 ... operation unit, 28 ... electron gun, 29 ... condenser lens, 30 ... objective lens, 31 ... first detector, 32 ... second detector, 33 ... third detector, 34 ... Lugo filter, 35 ... deflector, 36 ... reflector, 37 ... wafer height detection system, 38 ... XY stage, 39 ... holder, 40 ... retarding power supply, 41 ... cassette loading section, 42 ... wafer loader, 43 ... signal Detection system control unit, 44 ... Beam deflection correction control unit, 45 ... Electronic optical system control unit, 46 ... Wafer height sensor detection system, 47 ... Mechanism / stage control unit, 48 ... Operation screen / operation unit, 49 ... Image processing 50 ... Data storage unit 51 ... Wafer 52 ... Primary electron beam 53 ... Reflected electron 54 ... Secondary electron 55 ... High energy component secondary electron 56 ... Data input unit 57 ... Data conversion unit 73: alignment first point designation chip, 74: alignment second point designation chip, 75, 75 '... optical microscope image for alignment, 76, 76' ... electron beam image for alignment.

Claims (13)

A scanning electron microscope for acquiring a secondary electron or reflected electron signal generated by irradiating the sample with an electron beam;
An image processing unit that performs image processing on the signal obtained from the scanning electron microscope;
A data storage unit for storing images,
The image processing unit acquires defect position information obtained by inspecting a sample by an inspection apparatus including an electron beam inspection apparatus and inspection conditions of the inspection apparatus, and displays the inspection conditions on an operation screen. In the observation mode selected from the plurality of observation modes including the potential contrast observation mode based on the displayed inspection condition, an image of a position corresponding to the position information of the defect is higher in resolution than the inspection apparatus. A review SEM having a function of acquiring, storing the image in the data storage unit, and automatically classifying the content of the defect into a shape defect and an electrical defect based on the image. A scanning electron microscope for acquiring a secondary electron or reflected electron signal generated by irradiating the sample with an electron beam;
An image processing unit that performs image processing on the signal obtained from the scanning electron microscope;
A data storage unit that stores a plurality of images based on the signal obtained from the scanning electron microscope, and automatic defect classification conditions for automatically classifying each of the plurality of images into shape defects and electrical defects And comprising
The image processing unit acquires position information of a plurality of defects obtained by inspection of a sample by an inspection apparatus including an electron beam inspection apparatus and inspection conditions of the inspection apparatus , and displays the inspection conditions on an operation screen. In the observation mode selected from the plurality of observation modes including the potential contrast observation mode based on the displayed inspection conditions, images of the respective positions corresponding to the position information of the plurality of defects are inspected. Each of the images is stored in the data storage unit, and each of the images is automatically converted into an image including a shape defect and an image including an electrical defect based on the automatic defect classification condition. Review SEM characterized by classifying into: A stage for holding the sample;
An electron optical system for irradiating the sample with an electron beam;
A detection system including a plurality of detectors for detecting a signal generated secondarily by the electron beam applied to the sample;
An image processing unit that performs image processing on the signal obtained from the detection system;
An electron optical system controller for controlling the electron optical system;
Data for storing a review condition file in which an image obtained by image processing of the image processing unit and an electron beam irradiation condition for obtaining an image for inspecting a shape defect and an electrical defect are registered in advance. A storage unit,
The electron optical system control unit sets an electron beam irradiation condition of the electron optical system based on the review condition file,
The image processing unit acquires position information of a plurality of defects obtained by inspection of a sample by an inspection apparatus including an electron beam inspection apparatus and inspection conditions of the inspection apparatus , and displays the inspection conditions on an operation screen. In the observation mode selected from the plurality of observation modes including the potential contrast observation mode in the review condition file based on the displayed inspection condition, the inspection is performed according to the electron beam irradiation condition in the review condition file. Acquiring images of each position corresponding to the position information of the plurality of defects at a higher resolution than the apparatus;
The review SEM, wherein the image processing unit automatically classifies the image at each position into an image including a shape defect and an image including an electrical defect. In the review SEM of claim 1,
The data storage unit stores a review condition file for inspecting shape defects and electrical defects,
The review condition file includes a first electron beam irradiation condition in the scanning electron microscope for acquiring an image including a shape defect and a second condition in the scanning electron microscope for acquiring an image including an electric defect. The review SEM characterized by including electron beam irradiation conditions. In a review SEM according to claim 4,
The review SEM characterized in that an image acquired with a higher resolution than the inspection apparatus is an image obtained under the first or second electron beam irradiation conditions. In the review SEM of claim 2,
The data storage unit stores a review condition file for inspecting shape defects and electrical defects,
The review condition file includes a first electron beam irradiation condition in the scanning electron microscope for acquiring an image including the shape defect, and a second in the scanning microscope for acquiring an image including the electrical defect. The review SEM characterized in that the electron beam irradiation condition and the defect automatic classification condition are included. In the review SEM of claim 6,
The review SEM characterized in that an image acquired with a higher resolution than the inspection apparatus is an image obtained under the first or second electron beam irradiation conditions. In the review SEM according to any one of claims 1 to 7,
The review SEM, wherein the image processing unit automatically stores the result of classification into an image including a shape defect and an image including an electrical defect in the data storage unit. In the review SEM according to claim 3 or 4,
The review SEM, wherein the review condition file includes image processing conditions for automatically classifying into a shape defect and an electrical defect. In the review SEM according to claim 3,
One of the plurality of detectors is selected by the selection of the review condition file. In the review SEM according to any one of claims 3, 4, and 6,
The review SEM further comprises an operation screen capable of selecting the review condition file. In the review SEM according to any one of claims 4 to 7,
When the inspection apparatus is an electron beam appearance inspection apparatus, the scanning electron microscope automatically irradiates the sample with an electron beam according to the second electron beam irradiation condition. In the review SEM of claim 8,
The review SEM, wherein the image processing unit adds the result to a file in which the position information is stored.

Images