Defect review of samples with un-patterned surface

Abstract

A method for review of a plurality of anomalies on an un-patterned surface of a sample is provided. The method comprises registering the coordinate system of the un-patterned surface of the sample to the positions of a plurality of anomalies.

Description

[0001] Defect review of samples with un-patterned surface

[0002] Field

[0003] The present invention relates to a defect review method of semiconductor samples with an un-patterned surface, such as an un-patterned wafer or a processed wafer covered by an unpatterned surface. With the method, a coordinate system of the un-patterned surface is registered, and positions of possible defects can be found more quickly by a high-resolution measurement tool. The method, computer program product and semiconductor inspection device can be applied to wafers or mask blanks and utilized for various review tasks, such as quantitative metrology, process monitoring, or classification of defects.

[0004] Background

[0005] Semiconductor structures are amongst the finest man-made structures. Semiconductor manufacturing involves precise manipulation, e.g., lithography or etching, of materials such as silicon or oxide at very fine scales in the range of nm. A wafer made of a thin slice of silicon serves as the substrate for microelectronic devices containing semiconductor structures built in and upon the wafer. The semiconductor structures are constructed layer by layer using repeated processing steps that involve repeated chemical, mechanical, thermal and optical processes. Dimensions, shapes and placements of the semiconductor structures and patterns are subject to several influences. For example, during the manufacturing of 3D- memory devices, the critical processes are currently etching and deposition. Other involved process steps such as the lithography exposure or implantation also can have an impact on the properties of the elements of the integrated circuits. Therefore, fabricated semiconductor structures suffer from rare and different imperfections. Devices for quantitative metrology, defect-detection or defect review are looking for these imperfections. These devices are not only required during Wafer fabrication. For example, imperfections may also be present at lithography masks, used for patterning structures onto the wafers by lithography.

[0006] Fabricated semiconductor structures are fabricated by determined processes and are therefore based on prior knowledge. The semiconductor structures are manufactured in a sequence of layers being parallel to a surface of a substrate. For example, in a logic type sample, metal lines are running parallel in metal layers or HAR (high aspect ratio) structures and metal vias run perpendicular to the metal layers. Furthermore, a semiconductor wafer has a diameter of 300 mm and consist of a plurality of several sites, so called dies, each comprising at least one integrated circuit pattern such as for example for a memory chip or for a processor chip. The dies comprise registration or alignment marks, by which a wafer can be aligned or registered very precisely. During fabrication, semiconductor wafers run through about 1000 process steps, and within the semiconductor wafer, about 100 and more parallel layers are formed, comprising the transistor layers, the layers of the middle of the line, and the interconnect layers and, in memory devices, a plurality of 3D arrays of memory cells. The many layers are formed on an un-patterned wafer, and some layers may be formed as un-patterned layers, comprising no access to any alignment or registration marks on a mask or wafer. Since the critical features within semiconductor structures are becoming smaller and reach few nm, a detection and classification of defects becomes more and more challenging. With for example optical inspection, such as for example scatterometry, locations of possible anomalies are detected. After anomalies or defects might be suspected at specific locations, many of those anomalies require a closer review with a high-resolution microscope, for example a charged particle microscope (CPM). Such high-resolution tools usually have a limited field size. This is not an issue if a wafer or mask comprises a patterned surface and can be precisely registered to a coordinate system. However, finding a large number of suspected defects for review on an un-patterned surface is a tedious effort.

[0007] In WO 2020 / 016262 A1 and in US 200510236 569 A1 , methods are disclosed to register unpatterned wafers. However, the methods of WO 2020 / 016262 A1 and US 200510 236 569 A1 are time consuming and of limited accuracy, especially if large numbers of anomalies have to be analyzed.

[0008] It is therefore a task of the invention to provide a method of finding a plurality of locations of previously detected anomalies on an un-patterned substrate more quickly; it is a further task to quickly navigate an object field of a charged particle microscope to the plurality of locations of previously detected anomalies and to facilitate a defect review with high throughput even on un-patterned surfaces of substrates. Such un-patterned surfaces of substrates are present at mask or wafer blanks, or at certain processed wafers with un- patterned surfaces e.g. after deposition or polishing.

[0009] Summary

[0010] The subject of the invention is solved by a method comprising a method step of registering an un-patterned surface of a substrate with respect to a plurality of previously detected anomalies. The plurality of previously detected anomalies is for example a result of a defect inspection task, which determines roughly locations of anomalies for example by optical scatterometry. While optical metrology may not resolve whether an anomaly is a defect or a nuisance, it may provide a list comprising a plurality of anomalies with their respective relative locations with an accuracy of about 1 pm. According to the method, after registering the un-patterned surface of a substrate with respect to a plurality of previously detected anomalies, each location of an anomaly can be aligned within an object field of a charged particle microscope and an image of the anomaly is obtained with high resolution.

[0011] According to the method, the method step of registering comprises a step of an image acquisition at least a first anomaly and a second anomaly, and a step of registering a first review coordinate system to the first and second locations of the first and second anomalies. In an example, the search for the first and second anomalies might require several iterations of moving the un-patterned surface of the subject and acquiring a high-resolution image at a further location before the first or second anomaly is found. After registering the first review coordinate system, further locations of anomalies are moved in the object field of the charged particle microscope and further images of anomalies are obtained. In an example, corrections or modifications are applied to the first reference or review coordinate system. Such corrections or modifications can be based on the actual detected positions of for example a third locations of a third anomalies within the first reference or review coordinate system. Thereby, a reference or review coordinate system is iteratively improved and a plurality of locations of anomalies can be subsequently reviewed by a charged particle microscope with small object field. Thereby, a large plurality of anomalies distributed over large surfaces of for example wafers with 300mm in diameter can be reviewed in an efficient manner.

[0012] In a first embodiment, a method of registering an un-patterned surface of an object to a coordinate system is given. The object is loaded onto a chuck of a stage with a limited loading accuracy of more than 50pm, for example 100pm, and wherein the stage is having a stage coordinate system. The method is comprising selecting at least a first anomaly at a first coordinate [x1 , y1] and a second anomaly at a second coordinate [x2, y2] from a list of a plurality of anomalies. The method is further comprising determining a first search path at the first coordinate [x1 , y 1 ] of the first anomaly within the stage coordinate system, and acquiring an image patch of the un-patterned surface of the object along the first search path, analyzing the image patch whether the first anomaly is detected in the image patch, and continue with acquiring subsequent image patches until the first anomaly is detected on the un-patterned surface of an object. The method is further comprising registering the coordinate system with the first coordinate [x1 , y 1 ] at the detected position of the first anomaly within the stage coordinate system. This registering is however insufficient to completely register the review coordinate system of anomalies to the coordinate system of the stage, since a complete registration requires the registration of at least a second coordinate. Therefore, the method further comprises determining a second search path at a second coordinate of the second anomaly. The method further comprises acquiring an image patch of the un-patterned surface of the object along the second search path, analyzing the image patch whether the second anomaly can be detected in the image patch, and continue with acquiring subsequent image patches until the second anomaly is detected on the un- patterned surface of the object. After the second anomaly is detected, the method comprises completing registering the review coordinate system to the first coordinate [x1 , y1 ] and the second coordinate [x2, y2]. With completed registration of the review coordinate system of the coordinates of the plurality of anomalies to the coordinate system of the wafer stage after only two anomalies have been registered, further anomalies can be found on the un- patterned surface of the object with increased speed and without the need of acquiring and searching in many image patches for each anomaly.

[0013] In an example, the first and second search path are determined according to the limited loading accuracy. For example, the first search path is determined according to a spiral path around a starting point given by the first coordinate [x1, y1] within the coordinate system of the stage. In an example, the second search path is determined according to a segment of a circle around the first coordinate [x1 , y1 ] with radius R corresponding to the distance between first coordinate [x1, y1] and the second coordinate [x2, y2]. The length of the segment of the circle can be determined according to the limited loading accuracy.

[0014] In an example, the step of analyzing an image patch for detection of an anomaly comprises an image processing method selected from a group of methods including thresholding, noise filtering, template matching, morphologic operations, and machine learning.

[0015] In an example, the method is further comprising performing at least one a milling operation with a FIB beam along the first or second search path. Thereby, a buried anomaly can be detected.

[0016] In an example, the step of analyzing an image patch for detection of an anomaly comprises an material analysis of for example a defect, wherein the material analysis comprises a spectrometry of secondary radiation, for example energy dispersive X-ray spectroscopy (EDX).

[0017] In an example, the method is further comprising acquiring at least a third image patch comprising a third anomaly and refining the review coordinate system with the coordinate [x3, y3] of the third anomaly. With repeated refinement, further anomalies can be found on the un-patterned surface of the object with even more increased speed.

[0018] In an example, the selecting the least a first anomaly at the first coordinate [x1 , y 1 ] and a second anomaly at the second coordinate [x2, y2] is comprising selecting the first and the second anomaly as isolated anomalies. Thereby, ambiguities can be avoided. For example, selecting the first or second anomaly can be executed such that no further coordinate [xi, yi] of any further anomaly has a distance to the first or second anomaly of less than a predefined distance. The predefined distance con for example be defined according to the limited loading accuracy.

[0019] In an example, the method is further comprising loading the object onto the chuck of the stage of a review system with the limited loading accuracy. In an example, the is comprising receiving the inspection file comprising the list of a plurality of anomalies. An inspection file is typically generated by a preceding fast inspection with for example an optical inspection using light, which is of limited resolution of about 1 pm and thus not able to identify small defects of nm-sizes. However an optical inspection using light is capable to detect and locate anomalies on a surface with high precision, such that an inspection tool is capable of generating and providing a plurality of anomalies at coordinates [xi, yi] with high precision better than 1 m in an arbitrary coordinate system of the inspection task. The method according to the example is capable to receive the inspection file and to register the arbitrary coordinate system of the inspection task to the coordinate system of a review system, thereby incorporating the limited loading accuracy of an object to the stage of the review system.

[0020] According to a second embodiment, a review method for reviewing a plurality of anomalies at an un-patterned surface of an object is comprising loading the object onto a chuck of a stage of a review system with a limited loading accuracy of exceeding 50pm, for example 100pm, the stage having a stage coordinate system, and receiving an inspection file comprising a list of a plurality of anomalies, and registering a review coordinate system of the list of the plurality of anomalies according to a method described in the first embodiment. The review method further comprises reviewing each of the plurality of anomalies. In an example, the reviewing is comprising classifying an anomaly into a nuisance or wrong detections or into defect classes. In an example, the review method is further comprising a milling operation of the surface area around an anomaly corresponding to a buried defect. In an example, the review method is further comprising repeatedly refining the review coordinate system to an increasing number of positions of anomalies and reducing a size of subsequent image patches. Thereby, a throughput or speed of a review task is even more increased. In an example, a review method is further comprising removing a contamination particle by a movable actuator. The method is further comprising a milling operation with a FIB beam at an expected position of a buried alignment structure or anomaly.

[0021] In a third embodiment, a review system is given. The review system is comprising a charged particle beam imaging system, a stage with a chuck for loading an object, and an operation control unit comprising a memory for storing software instructions and a processing engine for executing the software instructions to cause the wafer review system to execute any of the methods according to the first or second embodiments. In an example, the review system is further comprising a focused ion beam column configured for milling into an object.

[0022] Thereby, buried anomalies, defects or alignment structures can be revealed.

[0023] The invention described by examples and embodiments is not limited to the embodiments and examples but can be implemented by those skilled in the art by various combinations or modifications thereof. The present invention will be even more fully understood with reference to the following drawings:

[0024] Figure 1 shows an example of a dual beam device of an exemplary review system

[0025] Figure 2 is an illustration of a list of anomalies comprised in an inspection file and plurality of anomalies on an un-patterned surface of a wafer.

[0026] Figure 3 illustrates a review method for an un-patterned surface of a sample.

[0027] Figure 4 illustrates the registration method steps

[0028] Figure 5 shows a wafer review system according to an embodiment.

[0029] While the examples and embodiments below are described at the examples of semiconductor wafers, it is understood that the invention is not limited to semiconductor wafers but can for example also be applied to reticles or masks for semiconductor fabrication.

[0030] Throughout the figures and the description, same reference numbers are used to describe same features or components. The coordinate system is selected that the wafer surface 55 coincides with the XY-plane.

[0031] Figure 1 discloses an example for a review system 1000 for wafer defect review.

[0032] The review system 1000 comprises a wafer stage 155 with a wafer support surface or chuck 15 for holding a wafer 8 during wafer defect review. The wafer stage 155 is controlled by a wafer stage control unit 16 and capable of for example for an adjustment with six degrees of freedom. The position of the wafer stage 155 and thereby any wafer 8 fixed to the wafer stage via wafer chuck 15 is controlled by position sensors (not shown), such as laser or grating interferometers, lens arrays, or the like. Actuators and means for precision control for a wafer stage 155 such as Laser interferometers are known in the art. A coordinate system 409 of the stage 155 with stage coordinate [xw, yw] can be selected within the review system 1000.

[0033] The review system 1000 of the present example further comprises a dual beam device 1, comprising a FIB column 50 with a FIB optical axis 48 and a charged particle beam (CPB) imaging system 40 with optical axis 42. The focused ion beam column (50) is arranged at an angle GF to the surface of the wafer support surface 15 of the wafer stage (155). During use, the wafer surface 55 is arranged at the intersection point 43 of both optical axes of FIB and CPB imaging system. FIB axis 48 and CPB imaging system axis 42 include an angle GFE, and the CPB imaging system axis forms an angle GE with the normal to the wafer support surface 15. In the coordinate system of figure 1 , the normal to the wafer support surface 15 is given by the z-axis. The focused ion beam (FIB) 51 is generated by the FIB-column 50 and is impinging under angle GF on the surface of the wafer 8. In the example of figure 1, the slant angle GF is approximately 30°. The actual slant angle of the slanted cross-section surface can deviate from the slant angle GF by up to 1° to 4° due to the beam divergency of the focused ion beam, for example a Gallium-lon beam. The FIB column 50 can for example be a Gallium FIB, or a FIB with a gas field ion source (GFIS) with other kinds of ion species, such as Xenon or Argon ions. With the charged particle beam imaging system 40, inclined under angle GE to the normal to the wafer support surface 15, images of small segments of the wafer surface are acquired. In the example of Figure 1 , the angle GE is about 15°. However, other arrangements are possible as well, for example with GE = 0°, such that the CPB imaging system axis 42 is perpendicular to the wafer support surface 15.

[0034] A wafer 8 might comprise several review sites 6 for performing a review, from which only two (6.1 and 6.2) are shown in figure 1. During a first review operation, a first review site 6.1 is aligned with the stage 155 at the optical axis 42 of a charged particle imaging system 40. Such a charged particle imaging system 40 typically has an object field with field size of up to 10pm and a resolution of below 5nm, for example 3nm, 2nm or even less. A small anomaly within a location given by approximate coordinates with resolution of about 1 m can therefore be found and reviewed within the object field with size of up to 10pm. Typically, the image is obtained by scanning the charged particle imaging beam 44 over a segment of the surface of the wafer 8 and collecting secondary of backscattered charged particles or radiation with sensors 17.1 and 17.2. With the FIB 50, a covering layer above for example a hidden or buried anomaly or defect can be removed, and a defect can be made accessibly for an image acquisition with the charged particle imaging system 40. Thereby, also anomalies normally hidden to the charged particle imaging system 40 can be reviewed. Furthermore, a FIB system 50 allows a sputtering of material from for example a contamination particle and thereby a review of a material composition of a contamination particle. During imaging, a beam of charged particles 44 is scanned by a scanning unit of the charged particle beam imaging system 40 along a scan path over a cross-section surface of the wafer 8 at review site 6.1 , and secondary particles as well as scattered particles are generated. For example, secondary electron particle detector 17.1 collects at least some of the secondary particles and scattered particles and communicates the particle count with a control unit 19. Other detectors for other of interaction products may be present as well, for example in-lens detector 17.2 for collection of backscattered charged particles. Further detectors can be provided, for example x-ray detectors for x-ray spectroscopy. Control unit 19 is in control of the charged particle beam imaging column 40, of FIB column 50 and connected to a stage control unit 16 to control the position of the wafer 8 mounted on the wafer support surface 15 via the wafer stage 155. Control unit 19 communicates with operation control unit 2, which triggers placement and alignment for example of review sites 6 of the wafer 8 at the intersection point 43 via wafer stage movement and triggers repeatedly operations of FIB milling, image acquisition and stage movements.

[0035] In an example, a review system 1000 further comprises a movable actuator or manipulator 81. Movable actuator or manipulator 81 can be configured as a contact pin. Thereby, structures present on the wafer surface can be contacted and electrically connected to control device 19 and local surface charge are provided or removed in the vicinity of a location of an anomaly. Thereby, for example, a voltage contrast image can be detected. In another example, contamination particles can adhere to actuator or manipulator 81 and be removed from the surface of the wafer 8.

[0036] In an example, a review system 1000 further comprises a gas injection system (GIS) 79, with a gas nozzle connected via a valve (not shown) to at least one gas reservoir (not shown).

[0037] Thereby, controlled amounts of precursor gases can be provided during milling or imaging, and for example metal coatings can be generated. For example, alignment marks or fiducials can be generated. For example, a Tungsten metal coating is generated by providing Tungsten Hexacarbonyl. The metal coating can be shaped by ion beam milling and alignment markers or fiducials are formed in proximity to an inspection site. Thereby, a precise registration and image alignment of the plurality of cross section images is enabled. With dedicated precursor gases, a milling operation by FIB 50 can be enhanced. For example, a homogeneity of a milling operation in compositions of different material can be improved and curtaining can be reduced. Compositions of materials in a semiconductor wafer can comprise Silicon, Silicon Dioxide, Silicon Nitride, Copper, Aluminum, Tungsten or other materials. Preferred precursor gases are comprising at least one of Ammonia, Ammonium Hydroxide, Ammonium Carbamate, Bromine, Chlorine, Hydrazine, Hydrogen Peroxide, Hadacidin, Iodine, di-iodo-ethane, Isopropanol, Methy Difluoroacetate, Nitroethane, Nitroethanol, Nitrogen, Nitrogen Tetroxide, Nitrogen Trifluoride, Nitromethane, Nitropropane, Nitrobutane, Oxygen, Ozone, PMCPS, Tungsten Hexacarbonyl, Water, or Xenon Difluoride. Other gases are, however, are possible as well, for example methoxy acetylchloride, methyl acetate, methyl nitroacetate, ethyl acetate, ethyl nitroacetate, propyl acetate, propyl nitroacetate, nitro ethyl acetate, methyl methoxyacetate, and methoxy acetylchloride, Acetic acid or thiolacetic acid, Hexafluoroacetylacetone, silazane, trifluoroacetamide, dicobalt octacarbonyl, molybdenum hexacarbonyl, and combinations thereof.

[0038] The review system 1000 further comprises a control processor unit 19, configured to control the FIB system 50, the charged particle imaging system 40, a gas supply via nozzle 79 or manipulator 81. The control processor unit 19 is connected to stage control unit 16. Control processor unit 19 is further connected to an operation control unit 2.

[0039] For a review task, typically a list of a plurality of locations of anomalies is provided. Figure 2a illustrates an example of such a list from an inspection file 307. Such a list is for example generated by an optical inspection tool, which detects anomalies at certain positions on a surface 55 of a mask or wafer 8, and lists these anomalies in an inspection file 307, for example a KLARF file together in a coordinate system. Such standard file formats are common in semiconductor industry. The list of anomalies 307 can comprise locations of anomalies with the lateral resolution of optical scatterometry of for example up to few pm, with respect to a coordinate system 411 available during optical inspection. In some examples, a list of a plurality of anomalies can comprise a description of a kind of at least some anomalies. Many anomalies 301 can be detected on a large surface of a wafer 8, as shown in figure 2b at some very few examples. The plurality of anomalies 301 can further comprise clusters 303 of anomalies. For example, some of the anomalies can be identified as surface defects 305.1 , 305.2.

[0040] Within a review tool 1000 for 300 mm wafers, it is necessary after loading the wafer 8 to align the wafer 8 for example with respect to the optical axis 42 of the charged particle imaging system 40. This is because of the placement accuracy of a loading of the wafer 8 onto the wafer chuck 15 is much less precise compared to what is needed when moving and navigating to the review sites 6.1 , 6.2 (see figure 1) corresponding to the locations of anomalies 301. The placement accuracy is normally in the range of several 100pm, but the alignment accuracy which is needed for placing the locations of anomalies within the images field of a high resolution charged particle imaging system 40 is better than 10um, for example 5pm.

[0041] On structured wafer it is possible to use structures from which the exact positions are known within the coordinate system 411 used during optical inspection. This can be a die corner or any other unique pattern or alignment mark which is visible by the optical inspection system as well as the charged particle beam imaging system 40. On an un-patterned wafer, however, it is not possible to register a predefined coordinate system 411 of an inspection task and transfer it to the wafer 8 loaded on wafer chuck 15 with the required accuracy. As described above, the coordinates of the plurality of location of anomalies from optical inspection of an un-patterned surface of a substrate is only known with a low precision of more than 100pm. For example, a wafer 8 loaded to the chuck 15 can be displaced with respect to a stage coordinate system by more than 100pm and can be rotated by an angle exceeding 1°. Thereby, a search of anomalies on a wafer surface of 300mm in diameter with small object fields of less than 10pm in diameter is a tedious effort. The problem is not limited to wafer blanks but also present for a wafer where no patterns are visible because of for example a coating is covering the wafer surface.

[0042] A review method according to an embodiment uses known positions of anomalies or defects from a previous wafer inspection tool and collect the actual positions of at least two defects on the tool. Then offset and rotation can be fitted to the dataset and a review coordinate system can be registered on the wafer. An example of the review method according to an embodiment is illustrated in Figure 3.

[0043] The review method according to the example comprises a step S1 of loading a sample 8, for example a specified wafer 8, on a wafer support surface or chuck 15 and positioning the chuck 15 by stage 155 at an arbitrary starting position within a wafer review system 1000 (for reference numbers, reference is made to figure 1, 2 and 4 and the description thereof). During step S1, a corresponding inspection file 307 from a previous inspection of the specified wafer 8 is loaded into the system memory of operation control unit 2. The inspection file is automatically assigned to the specified wafer 8 for example by an identifier label. In an example, an inspection file 307 is assigned to a specified sample 8 by user interaction. The inspection file 307 comprises a list of a plurality of anomalies which have been detected during a preceding inspection, for example optical inspection with a resolution larger than 1 pm. With such low resolution, defects cannot be identified nor classified but are merely detected as anomalies.

[0044] In step S2, a review coordinate system of the plurality of anomalies is registered to the wafer stage coordinate system with coordinates [xw, yw]. As described above, the wafer 8 is loaded onto the chuck 15 of the review system 1000 only with a limited loading accuracy of more than 50|jm, meaning the position of the wafer 8 on the stage 155 is known only to a precision of 50pm or more, for example about 100pm, and the position of the anomalies are thus only known within areas with a diameter of about the limited loading accuracy. The initial wafer stage coordinate system 409 can be an arbitrarily selected coordinate system, for example a reference coordinate system of the wafer stage 155.

[0045] In step S2.1 , a first anomaly 301.1 with first coordinate [x1 , y 1 ] is selected from the inspection file 307. In step S2.2, a first search path 407.1 is determined according to the limited loading accuracy and the object field size of an image patch 401 acquired by the charged particle imaging system 40. In the example, the first search path 407.1 is defined by a spiral path around a starting point corresponding to the first coordinate [x1, y1] within the coordinate system 409 of the stage 155, however with an uncertainty of the limited loading accuracy of the wafer 8 to the wafer chuck 15. Thus, a first search area 403.1 might cover an area of about 10 x 10 image patches, including image patches 401.11, 401.12 to 401.19. An example of steps S2 of determining the registered or review coordinate system 411 of the wafer 8 is further illustrated in figure 4.

[0046] In an example, isolated anomalies 301.1 and further are selected in step S2.1. Within the plurality of anomalies 301 , there might be anomalies with small distances with respect to each other, forming clusters 303 of anomalies 301. A registration of anomalies within such clusters 303 might lead to an ambiguity in the registration. In an example, selecting the first and a second anomaly (301.1, 301.2) is performed by analyzing the distances to other anomalies (301. i) from the inspection file 307, such that no further coordinate [xi, yi] of any further anomaly (301. i) has a distance to the first or second anomaly (301.1, 301.2) of less than a predefined distance. The predefined distance can for example be defined as the limited loading accuracy of the wafer 8 to the chuck 15 or stage 155. In another example, the predefined distance can be defined according to a size of an image patch 401, for example

[0047] 10pm x 10pm.

[0048] In step S2.3, the wafer 8 is positioned by wafer stage 155 at the starting position of the first search path 407.1 under the charged particle beam imaging system 40 and the first image patch 401.11 of the surface of the wafer 8 is acquired by charged particle imaging system 40.

[0049] In Step 2.4, the first image patch 401.11 is analyzed by image processing whether the first anomaly 301.1 is present in the first image patch 401.11. If the answer is no, the method continues with repeating step S2.3, while in a second or any further repetition of step S2.3, the wafer 8 is positioned by stage 155 along the first search path 407.1 such that adjacent image patches 401.12, 401.13 and further can be acquired. The image patches 401.11 to - in this example - 401.19 are sequentially acquired and analyzed by image processing whether the first anomaly 301.1 is present in the any image patch 401.11 to 401.19 until the first anomaly 301.1 is detected. In the example of figure 4, the first anomaly 301.1 is detected in image patch 401.19. Methods of image processing for detection of anomalies within images are well known in the art and may comprise methods including thresholding techniques, noise filtering, template matching techniques, morphologic operations, and machine learning.

[0050] After the first anomaly 301.1 is detected, method continues with step S2.5. In step S2.5, the stage coordinate system of the wafer stage 155 is adjusted to the first coordinates [x1 , y 1 ] according to the inspection file 307. It is determined whether the wafer coordinate system of the wafer stage 155 is determined with sufficient precision. In the example, after the first anomaly 301.1 is detected, a first position of the coordinate system of the wafer stage 155, corresponding to first coordinates [x1, y1], is determined, leaving an angle of rotation of the wafer 8 undetermined. A second anomaly 301.2 at second coordinates [x2, y2] is thus known to be located at a circle around the first coordinate [x1, y1] with radius R corresponding to the distance between second and first coordinates of anomalies 301.1 and 301.2 from inspection file 307. According to the limited loading accuracy, a second anomaly 301.2 is known to be within proximity of a segment of a circle, but however a position of the second anomaly 301.2 on the search path 407.2 along a segment of the circle with radius R is still unknown within the range of the limited loading accuracy of e.g. 100pm. Thus, the coordinate system of the wafer stage 155 is still not determined with sufficient precision, and method continues with a repetition of steps S2, including selecting a second or subsequent anomaly 301.2, determining the second search path 407.2 including a sequence of second image patches 401.2. The repetition of steps S2 is further comprising acquiring a sequence of second image patches 401.21 and following image patches and determining by image processing whether the second anomaly 301.2 is present in any of the image patches 401.21 to following image patches. The search is interrupted in step S2.4 when the second anomaly 301.2 is detected. In step S2.5, the initial coordinate system 409 of the wafer stage 155 is confined with the coordinates [x2, y2] of the second anomaly 301.2 from the inspection file 307. After at least two iterations of the sub-steps of step S2, a position of the anomalies from the inspection file 307 within the coordinate system of the wafer stage 155 is known with sufficient precision, and the wafer review task can be performed with high speed. The positions of the anomalies from the inspection file 307 therefore can precisely registered within the initial coordinate system of the wafer stage 155. For example, the initial wafer coordinate system 409 of the wafer stage 155 is changed into a registered or review coordinate system 411 , and the registered or review coordinate system 411 is used for position control and movement of the wafer stage 155. The registered coordinate system 411 is stored and for example used by stage control unit 16 for movement and alignment of the wafer 8 mounted to the wafer stage 155.

[0051] In an example of step 2.4, the analysis of image patches whether an anomaly is present comprises a user interaction. An image patch can be displayed via a user display to a user and method step 2.4 can comprise requesting a user input. A user may select an anomaly and may mark a position of the anomaly within the image patch displayed on the user display. The user interface is therefore comprising a user display configured to display image patches, images of anomalies, positions of anomalies, and requests for user interaction. The user interface therefore further comprises at least one input device configured to receive user instructions such as a selection of an anomaly from a list of anomalies, an identification of a user anomaly, and setting a position marker of an anomaly. The user display can also be configured to display the result of the image processing whether an anomaly is present in an image patch and be configured to receive user instructions to confirm or discard the result of the image processing. Thereby, for example, ambiguities can be resolved when clusters 303 of anomalies 301 are present.

[0052] In an example, step S2.3 further comprises a milling operation with a FIB system. Thereby, a buried anomaly, a buried defect or a buried alignment marker can be exposed and made accessible for electron-beam imaging. In an example, in step S2.4, a buried anomaly is suspected and a milling step M is inserted to expose the suspected anomaly to enable a scanning electron beam image with higher contrast.

[0053] The wafer review task is then continued in step S3 for the plurality of anomalies 301 from the inspection file 307. In step S3.1 , a subsequent anomaly 301. i is selected from the inspection file 307. The wafer 8 is now accurately positioned according to the registered or review coordinate system 411 , and a high-resolution image patch 401. i comprising the anomaly is 301. i is obtained by charged particle beam imaging system 40. In step S3.2, an image patch 401.i is analyzed whether it comprises an anomaly 301.i. In some examples, some anomalies might correspond to buried defects. In such examples, an anomaly 301. i might not directly be detected within an image patch 401. i, and a milling operation step M of the surface area around the buried defect might be required. Milling step M is for example performed by focused ion beam system 50. After milling step M, image acquisition of the image patch 401. i is repeated. Milling step M an imaging step S3.1 can be repeated until the buried defect is detected in step S3.2.

[0054] In step S3.3, the registered or review coordinate system 411 is reviewed and optionally refined. For example, the position of the detected anomaly 301. i can be at an edge within an image patch 401. i, indicating a registration error. For example, the position of the detected anomaly 301. i within the image patch 401. i is compared to the position [xi, yi] from the inspection file 307, indicating a mismatch or drift in coordinate systems. For example, step S3.3 comprises acquiring a third image patch (401.3) comprising a third anomaly (301.3) and refining the review coordinate system (411) with the coordinate [x3, y3] of the third anomaly (301.3).

[0055] In step S3.3, the registered coordinate system 411 may thus be continuously refined by for example optimization and provided for example to stage control unit 16. Within continuous optimization, the registered coordinate system 411 can be matched to an increasing number of positions of anomalies for example by least squares optimization. Optimization may further comprise a placement of anomalies within centers of image patches 401. If a certain confidence level of a registered coordinate system 411 is achieved, a size of an image patch 401 can be reduced and a review task can be performed with even further increased speed.

[0056] In step S3.4 it is determined whether a review task is completed by completion of the image acquisition of the last anomaly from the inspection file 307. If not, review steps S3 are repeated.

[0057] In step S4 a review of the anomalies is performed. Anomalies are classified into nuisance or wrong detections during optical inspection, and true defects. True defects may be classified into defect classes, such as contamination particles, surface defects, buried defects, material defects, and others. Methods of defect identification and classification are well known in the art and comprise for example pattern matching or machine learning methods. During step S4, further review operations may be included, such as metrology operations to determine a length, area or volume of a defect. Other examples comprise a material composition analysis, for example by EDS (also energy dispersive X-ray spectroscopy or EDX)-methods or mass spectroscopy of removed particles by sputtering. In some examples, a review operation might comprise a removal of a contamination particle by movable actuator 81. After completion of a review task of a wafer 8, the method can be repeated for a next wafer 8 with a corresponding inspection file 307.

[0058] It is understood that the method steps S2 to S4 must not necessarily be performed in sequence but can at least partially be executed in parallel. For example, a first and each further detected anomaly of step S2 and S3 can immediately be subject to a review according to step S4, even before an image of the last anomaly is acquired.

[0059] An example of an improved review system 1000 is illustrated in Figure 5 at the example of a wafer review system (1000). The review system 1000 of this example is comprising a dual beam system 1, but a review system 1000 is not limited to dual beam systems 1. A dual beam system is illustrated in figure 1 and reference is also made to the description of figure 1. Typical features of a dual beam system 1 are for example a first charged particle or FIB column 50 for milling and a second, charged particle beam imaging system 40 for high- resolution imaging. A dual beam system 1 comprises at least one detector 17 for detecting secondary particles, which can be electrons or photons. In the example, a first detector 17.1 is arranged close to the interaction volume of the primary beam 44 with the wafer 8 and configured to attract and collect secondary electrons. A second, in-lens detector 17.2 is arranged within the imaging charged particle beam system 40 and configured to collect backscattered electrons. The dual beam system 1 further comprises a wafer stage 155 configured for holding during use a wafer 8. The wafer stage 155 comprises actuators for lateral and axial displacement or rotation of the wafer stage 155. For example, a wafer stage 155 comprises long stroke actuators for displacements of the wafer 8 from a first site 6.1 to a second review site 6.2 and short stroke actuators of high precision for precision adjustment of the wafer 8 at a review position. The degrees of freedom for position adjustment and movement of the wafer stage 155 can be between three (x,y, rotation around z-axis) and all six degrees of freedom. The dual beam system 1 further comprises a control unit 19. The control unit 19 is configured with memory and logic to control operation of the dual beam system 1 and to store image patches 401.

[0060] The wafer stage 155 is position controlled by a stage control unit 16, which is connected to a high precision position sensor 21 configured for measuring during use the position of the wafer stage 155 relative to the charged particle beam imaging system 40 in at least two degrees of freedom (x,y). In an example, the charged particle beam imaging system 40 and the high precision position sensor 21 are mounted on a rigid support or metrology frame 25, which acts as a reference for the relative position measurement between wafer stage 155 and charged particle imaging beam 44.

[0061] Examples of precision position sensor 21 comprise Laser interferometers, grid interferometers, capacitive sensors or confocal sensors. Precision position sensor 21 is configured for performing during use at least one position measurement 27 of the position of the wafer stage 155 with respect to a metrology frame 25. Precision position sensor 21 is further connected to control unit 19. Generally, the control unit 19 is configured to receive and store a plurality of relative position vectors of the wafer stage 155. Control unit 19 comprises a memory and processor to execute the instruction of the method steps described above and the review coordinate system 411 of the wafer stage 155 can be selected by the method describe above.

[0062] During image acquisition, a charged particle beam source 31 generates charged particles.

[0063] The charged particle imaging system 40 further comprises a deflection scanner 29 for raster scanning the charged particle imaging beam. The charged particle imaging system 40 further comprises an objective lens 33 for focusing the charged particle imaging beam onto a surface of a wafer 8. Deflection scanner 29 and objective lens 33 are connected and controlled by control unit 19. Thereby, a location or position of an anomaly or defect can be determined with high precision within an image patch.

[0064] The wafer review system 1000 is further comprising an operation control unit 2. The operation control unit 2 comprises at least one processing engine 201, which can be formed by multiple parallel processors including GPU processors and a common, unified memory. The operation control unit 2 further comprises an SSD memory or disk memory or storage 203 for storing data, for example including training data and a trained machine learning algorithm, and a plurality of image patches comprising anomalies or defects. The operation control unit 2 further comprises a user interface 205, comprising the user interface display 800 and user command devices 801, configured for receiving input from a user and display quotes or results to a user. The operation control unit 2 further comprises a memory or storage 219 for storing process information of the image generation process of the dual beam device 1 and for storing software instructions, which can be executed by the processing engine 201.

[0065] The operation control unit 2 is further connected to an interface unit 231 , which is configured to receive further commands or data, for example inspection files, from external devices or a network. The interface unit 231 is further configured to exchange information, for example receive instructions from external devices or provide measurement results to external devices or store a set of training data or a trained machine learning algorithm or plurality of cross section images in external storages.

[0066] The operation control unit 2 is connected to dual beam system 1 and configured to receive a plurality of two-dimensional image patches. Operation control unit 2 is further configured to determine position vectors of anomalies or defects within image patches. The wafer review system 1000 is configured to receive user information for execution of a review task. The processing engine 201 is configured to compute and display information via the user display 800 and to receive user input via user interface 801. The wafer review system 1000 is configured to execute a method according to the disclosure.

[0067] The invention described by examples and embodiments is however not limited to the clauses but can be implemented by those skilled in the art by various combinations or modifications.

[0068] A list of reference numbers is provided:

[0069] 1 Dual Beam system

[0070] 2 Operation Control Unit

[0071] 6 review site

[0072] 8 wafer

[0073] 15 support table

[0074] 16 stage control unit

[0075] 17 Electron detector

[0076] 19 Control processor Unit

[0077] 21 Position Sensor

[0078] 23 Environment Sensor

[0079] 25 Metrology Frame

[0080] 27 position measurement

[0081] 29 scanning deflector

[0082] 31 electron source

[0083] 33 objective lens

[0084] 40 charged particle beam (CPB) imaging system

[0085] 42 Optical Axis of imaging system

[0086] 43 Intersection point

[0087] 44 Imaging charged particle beam 48 Fib Optical Axis

[0088] 50 FIB column

[0089] 51 focused ion beam

[0090] 55 wafer top surface

[0091] 79 gas nozzle

[0092] 81 movable actuator

[0093] 155 wafer stage

[0094] 201 processing engine

[0095] 203 memory

[0096] 205 User interface

[0097] 219 memory

[0098] 231 Interface unit

[0099] 301 anomaly

[0100] 303 group of defects; defect cluster

[0101] 305 surface defect

[0102] 307 inspection file

[0103] 401 image patch

[0104] 403 first search area

[0105] 404 second search area

[0106] 407 search path

[0107] 409 stage coordinate system

[0108] 411 registered or review coordinate system

[0109] 800 user interface display

[0110] 801 user command devices

[0111] 1000 review system

Claims

CLAIMS1 . A method of registering an un-patterned surface (55) of an object (8) to a coordinate system (411), wherein the object (8) is loaded onto a chuck (15) of a stage (155) with a limited loading accuracy of more than 50pm, for example 100pm, and wherein the stage (155) is having a stage coordinate system (409), the method comprising:- selecting at least a first anomaly (301.1) at a first coordinate [x1 , y1] and a second anomaly (301 .2) at a second coordinate [x2, y2] from a list of a plurality of anomalies (301 , 301.1 , 303.2, 303.3),- determine a first search path (407.1) at the first coordinate [x1 , y1] of the first anomaly (301.1) within the stage coordinate system (409),- detecting the first anomaly (301.1) on the un-patterned surface (55) of the object (8),- registering the review coordinate system (411) with the first coordinate [x1 , y1] at the detected position of the first anomaly (301.1) within the stage coordinate system (409),- determining a second search path (407.2) at a second coordinate [x2, y2] of the second anomaly (301.2),- acquiring an image patch (401.21) of the un-patterned surface (55) of the object (8) along the second search path (407.2), analyzing the image patch (401.21) whether the second anomaly (301.2) can be detected in the image patch (401.21), and continue with acquiring subsequent image patches (401.22, 404.23, ...) until the second anomaly (302.2) is detected on the un-patterned surface (55) of the object (8),- completing the registering the review coordinate system (411) to the first coordinate [x1 , y1] and the second coordinate [x2, y2], wherein the second search path (407.2) is determined according to a segment of a circle around the first coordinate [x1 , y1 ] with radius R corresponding to the distancebetween first coordinate [x1 , y1 ] and the second coordinate [x2, y2].

2. The method according to claim 1 , wherein the step of detecting the first anomaly(301.1) on the un-patterned surface (55) is comprising acquiring an image patch (401 .11) of the un-patterned surface (55) of an object (8) along the first search path(407.1), analyzing the image patch (401.11) whether the first anomaly (301.1) can be detected in the image patch (401 .11), and continue with acquiring subsequent image patches (401.12, 404.13, ...) until the first anomaly (301.1) is detected on the un- patterned surface (55) of the object (8).

3. The method according to claim 1 or 2, wherein the first or second search path (407.1 , 407.2) are determined according to the limited loading accuracy.

4. The method according to any of the claims 1 to 3, further comprising a milling operation to expose a buried anomaly, defect or alignment mark to the scanning electron-imaging.

5. The method according to any of the claims 1 to 4, wherein the first search path(407.1) is determined according to a spiral path around a starting point given by the first coordinate [x1 , y 1 ] within the coordinate system (409) of the stage (155).

6. The method according to any of the claims 1 to 5, wherein a length of the segment of the circle is determined according to the limited loading accuracy.

7. The method according to any of the claims 1 to 6, further comprising acquiring at least a third image patch (401.3) comprising a third anomaly (301.3) and refining the review coordinate system (411) with the coordinate [x3, y3] of the third anomaly (301.3).

18. The method according to any of the claims 1 to 7, wherein the selecting the least a first anomaly (301.1) at a first coordinate [x1 , y1] and a second anomaly (301.2) at second coordinate [x2, y2] is comprising selecting the first and the second anomaly (301.1 , 301.2) as isolated anomalies (301.1 , 301.2).

9. The method according to claim 8, comprising selecting the first or second anomaly (301.1 , 301.2) such that no further coordinate [xi, yi] of any further anomaly (301. i) has a distance to the first or second anomaly (301 .1 , 301 .2) of less than a predefined distance.

10. The method according to claim 9, further comprising defining the predefined distance according to the limited loading accuracy.11 . The method according to any of the claims 1 to 10, further comprising receiving the inspection file (307) comprising the list of a plurality of anomalies (301 , 301 .1 , 303.2, 303.3).

12. The method according to any of the claims 1 to 11 , wherein analyzing an image patch (401) for detection of anomalies (301) comprises an image processing method selected from a group of methods including thresholding, noise filtering, template matching, morphologic operations, and machine learning.

13. The method according to any of the claims 1-12, further comprising an energy dispersive X-ray spectroscopy (EDX).

14. The method according to any of the claims 1 to 13, further comprising loading the object (8) onto the chuck (15) of the stage (155) of a review system (1000) with the limited loading accuracy.

15. A review method for reviewing a plurality of anomalies at an un-patterned surface of an object (8), comprising- loading the object (8) onto a chuck (15) of a stage (155) of a review system (1000) with a limited loading accuracy of exceeding 50pm, for example 100pm, the stage (155) having a stage coordinate system (409),- receiving an inspection file (307) comprising a list of a plurality of anomalies (301 ,301.1 , 303.2, 303.3, and further),- registering a review coordinate system of the list of the plurality of anomalies (301 ,301.1 , 303.2, 303.3, and further) according to any of the claims 1 to 14,- reviewing each of the plurality of anomalies (301 , 301.1 , 303.2, 303.3, and further).

16. The review method according to claim 15, comprising classifying an anomaly (301 ,301.1 , 303.2, 303.3, or further) into nuisance or wrong detections or into defect classes.

17. The review method according to claim 15 or 16, further comprising a milling operation of the surface area around an anomaly (301 , 301.1 , 303.2, 303.3) corresponding to a buried defect.

18. The review method according to any of the claims 15 to 17, further comprising repeatedly refining the review coordinate system (411) to an increasing number of positions of anomalies (301 , 301.1 , 303.2, 303.3, and further) and reducing a size of subsequent image patches (401).

19. The review method according to any of the claims 15 to 18, further comprising removing a contamination particle by a movable actuator (81).

20. A review system (1000), comprising an charged particle beam imaging system (40), a stage (155) with a chuck(15) for loading an object (8), and an operation control unit (2) comprising a memory (219) for storing software instructions and a processing engine (201) for executing the software instructions to cause the wafer review system (1000) to execute any of the methods according to claims 1 to 19.

21. The review system (1000) according to claim 20, further comprising a focused ion beam column (50) configured for milling into an object (8), thereby revealing a buried defect.

Images