Method and system for image formation using multiple electron beams

Abstract

With a multi-electron beam system in which hundreds of beamlets are formed, it is possible to focus and converge these beamlets, reduce the effect of Coulomb interaction, and improve the resolution of these beamlets. By means of the electrostatic and magnetic deflection fields of a Wien filter, it is possible to separate the secondary electron beam from the primary electron beam and simultaneously correct the spherical aberration and source energy dispersion blur for all beamlets.

Description

Technical Field

[0001] The present disclosure relates to an electron beam system.

Background Art

[0002] As the semiconductor manufacturing industry develops, the demand for yield management, particularly for metrology and inspection systems, is increasing. The critical dimensions continue to shrink, and there is still a need in the industry to reduce the time required to achieve high yield and high value-added production. Reducing the total time from detecting a yield problem to correcting it is a determining factor in the return on investment for semiconductor manufacturers.

[0003] When manufacturing semiconductor devices, such as logic and memory devices, semiconductor wafers are typically processed using a number of manufacturing processes to form various features (outer shape characteristics) and multiple layers of those semiconductor devices. For example, in a semiconductor manufacturing process such as lithography, a pattern is transferred from a reticle to a photoresist disposed on a semiconductor wafer. Further examples of semiconductor manufacturing processes include, but are not limited to, chemical mechanical polishing (CMP), etching, deposition, and ion implantation. An array of a plurality of semiconductor devices may be created on a single semiconductor wafer and separated into individual semiconductor devices.

[0004] The inspection process is used in various steps during semiconductor manufacturing, and by detecting defects on the wafer side, it can promote yield improvement and ultimately profit improvement in the manufacturing process. Inspection has always been an important part of the manufacturing of semiconductor devices such as integrated circuits (ICs). However, as the dimensions of semiconductor devices are scaled down, inspection has become increasingly important as never before for the successful manufacturing of acceptable semiconductor devices; because even small defects may cause malfunctions in those devices. For example, as the dimensions of semiconductor devices are scaled down, the detection of smaller-sized defects has become necessary; because even relatively small defects may cause unwanted errors in those semiconductor devices.

[0005] On the other hand, as design rules are scaled down, semiconductor manufacturing processes tend to operate closer to the limits of their performance capabilities. In addition, as design rules are scaled down, smaller defects can affect the electrical parameters of the device, driving the need for more sensitive inspections. As design rules are scaled down, the population of potential yield-related defects detected by inspection grows dramatically, and the population of nuisance defects detected by inspection also increases significantly. Therefore, more defects tend to be detected on those wafers, and it can be difficult and expensive to correct those processes so that all of those defects are eliminated. By discriminating which defects actually have an effect on the electrical parameters and yield of those devices, a process control method can be adopted that focuses on those defects while generally ignoring other defects. Furthermore, for smaller design rules, process-induced malfunctions tend to be systematic, depending on the case. That is, process-induced malfunctions tend to occur in a given design pattern that is often repeated many times within the design. Eliminating spatially systematic electrical-related defects can have an impact on yield.

[0006] A focused electron beam system is widely used to generate or inspect the fine structure of an object, such as that of a silicon wafer used in the manufacture of integrated circuits. An electron beam formed of electrons emitted from an emitter in an electron gun is made to act as a fine probe, and the fine structure is inspected by causing it to interact with the wafer. Previously, the inspection and review of wafers were performed using a single electron beam, and completed or incomplete IC components with a critical dimension (CD) at the nanometer level were inspected. The throughput of a single electron beam apparatus is quite low. Semiconductor manufacturers are seeking a system with higher throughput.

[0007] What characterizes the throughput of a multi-electron beam apparatus is the number of sub-beams, that is, the number of all electron beamlets. The higher the number of beamlets, the higher the throughput. However, in increasing the number of beamlets, an imaging projection optical system composed of macro optical elements including a macro objective lens and a Wien filter becomes an obstacle. Although there are increasing voices desiring to achieve high throughput by increasing the number of electron beamlets, it is difficult to provide imaging uniformity over a wide field of view (FOV) with an imaging projection system; the reasons are off-axis aberration derived from off-axis beamlets, source energy dispersion blur, optical astigmatism caused by the presence of a Wien filter, and the influence of strong Coulomb interaction caused by an increase in beam current due to a large number of beamlets.

[0008] In a conventional electron beam system, low resolution due to the influence of Coulomb interaction has been a cause for concern. The resolution of a multi-electron beam system has been severely limited by an imaging projection optical system in which there is a beam crossover near the wafer and the resolution exists from the intermediate image plane (IIP) to the wafer. The beamlet imaging resolution has been greatly restricted by the effect of Coulomb interaction near the crossover.

[0009] In conventional electron beam systems, problems related to aberration blur and source energy dispersion blur caused by the presence of Wien filters have also been a source of concern. To remove crosstalk between beamlet signals in a multi-electron beam device, a Wien filter separates a secondary electron beam (SEB) from a primary electron beam (PEB) and increases the SEB deflection angle by increasing the Wien filter intensity. As a result, severe source energy dispersion blur and aberration blur of each beamlet occur over a wide field of view (FOV).

Prior Art Documents

Patent Documents

[0010]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0011] An improved system and method are needed.

Means for Solving the Problems

[0012] In a first embodiment, a system is provided. The system has an electron beam source that generates an electron beam. The electron beam source has a tip, a suppression electrode, and an extraction electrode. The system further includes a stage configured to hold a wafer on the electron beam path, an objective lens on the electron beam path, a Wien filter on the electron beam path between the objective lens and the electron beam source, a transfer lens on the electron beam path between the Wien filter and the electron beam source, and a detection array configured to receive at least one secondary electron beam from the wafer on the stage. The transfer lens has pole pieces and a transfer lens coil. The objective lens has an upper pole piece, a lower pole piece, an objective lens coil disposed on the upper pole piece, a charge control plate disposed on the lower pole piece, an acceleration electrode disposed on the electron beam path between the upper and lower pole pieces, and a scanner disposed on the upper pole piece.

[0013] The Wien filter can be configured to have an electrostatic deflector and a magnetic deflector.

[0014] The system can further include a collimation lens on the electron beam path and a beam limiting aperture on the electron beam path between the collimation lens and the electron beam source.

[0015] The system can further include an aperture array disposed on the electron beam path, a microstigmator array disposed on the electron beam path between the aperture array and the transfer lens, a microdeflector array disposed on the electron beam path between the microstigmator array and the transfer lens, and a microlens array disposed on the electron beam path between the microdeflector array and the transfer lens. The aperture array splits the electron beam into a plurality of beamlets. Include at least 100 beamlets among those plurality of beamlets. The electron beam can be a telecentric beam upstream of the aperture array.

[0016] In one example, the electron beam path is oriented in a first direction when exiting the transfer lens and in a second direction different from the first direction when exiting the Wien filter, and thus forms a non-parallel angle between the first direction and the second direction.

[0017] The system can include a second Wien filter on the electron beam path between the Wien filter and the transfer lens. The second Wien filter can be configured to have a second electrostatic deflector and a second magnetic deflector.

[0018] In the second embodiment, a method is provided. In this method, the electron beams are focused by directing a plurality of beamlets through a transfer lens downstream of the electron beam source. The beamlets are directed through a Wien filter downstream of the transfer lens to separate a secondary electron beam or a beam group from the beamlets. The beamlets are directed through the upper pole piece of the objective lens. The objective lens is downstream of the Wien filter. The beamlets are directed through an acceleration electrode downstream of the upper pole piece. The beamlets are directed through a charge control plate disposed inside the lower pole piece of the objective lens. The charge control plate is disposed on the opposite side of the objective lens as viewed from the upper pole piece. The beamlets are directed onto the wafer. At least one secondary electron beam is received from the wafer by the detection array.

[0019] The plurality of beamlets can include at least 100 beamlets.

[0020] The method can further include generating an electron beam using an electron beam source and converting the electron beam into a plurality of beamlets.

[0021] The method can further include directing the electron beam through a collimation lens and a beam limiting aperture disposed on the electron beam path between the electron beam source and the transfer lens.

[0022] The method can further include scanning with the beamlets using a scanner disposed on the upper pole piece.

[0023] The Wien filter can have an electrostatic deflector and a magnetic deflector.

[0024] The method can further branch an electron beam into beamlets using an aperture array, direct the beamlets through a microstigmator array disposed on the beamlet path between the aperture array and the transfer lens, direct the beamlets through a microdeflector array disposed on the beamlet path between the microstigmator array and the transfer lens, and direct the beamlets through a microlens array disposed on the beamlet path between the microdeflector array and the transfer lens. The aperture array is disposed on the electron beam path between the electron beam source and the transfer lens. At least 100 beamlets can be included in the plurality of beamlets. The electron beam can be a telecentric beam upstream of the aperture array.

[0025] The method can further use a Wien filter to change the direction of the beamlets so that the beamlets are directed at an angle at the exit of the Wien filter with respect to their orientation at the entry of the Wien filter.

[0026] The method can further direct the beamlets through a second Wien filter on the beamlet path between the Wien filter and the transfer lens.

[0027] The acceleration electrode can be configured to be able to change the resolution of the beamlets.

[0028] The Wien filter can be configured to be able to simultaneously adjust dispersion and aberration with respect to the beamlets.

[0029] For a more complete understanding of the nature and objects of the present disclosure, reference should be made to the following accompanying drawings in conjunction with the detailed description set forth below.

Brief Description of the Drawings

[0030]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

DETAILED DESCRIPTION OF THE INVENTION

[0031] The subject matter recited in the claims is described by specific embodiments, but other embodiments are also within the technical scope of the present disclosure, including embodiments in which all of the benefits and features described herein are not provided. Various structural, logical, processing step, and electronic modifications can be made without departing from the technical scope of the present disclosure. Thus, the technical scope of the present disclosure is defined solely by reference to the claims in the separate claims.

[0032] A multi-electron beam system in which hundreds of beamlets are formed is disclosed in the present application. The transfer lens (TL) field focuses those beamlets at an optimal optical magnification. The energy acceleration booster field reduces the effect of Coulomb interaction and improves the resolution of those beamlets. An electron energy retardation (deceleration) and substrate charging field can be used to obtain a desired wafer charge, extraction field, and landing energy. The magnetic objective lens field can image the wafer with a multi-electron beam with minimal optical aberration. The electrostatic and magnetic deflection fields of the Wien filter can separate a secondary electron beam (SEB) or a group of beams from the primary electron beam or a group of beams, and can simultaneously correct the source energy dispersion blur for all of those beamlets. The Wien filter can be made to simultaneously correct the beamlet astigmatism with an electrostatic (or magnetic) stigmator field. The optimal magnification can be selected with the transfer lens and the Coulomb interaction can be reduced with the acceleration electrode. According to one embodiment, the source energy dispersion blur of those beamlets can be corrected by the presence of the Wien filter. According to another embodiment, the astigmatism blur of those beamlets can be corrected by the presence of the Wien filter. According to yet another embodiment, it is possible to compensate for the source energy dispersion between two Wien filters, and thus to remove any energy dispersion blur of those beamlets and simultaneously deflect the secondary electron beam toward the lateral detection array.

[0033] FIG. 1 shows an optical system of an embodiment of a multi-electron beam apparatus. This has four optical modules: an electron gun (“Gun” in the figure), a collimation lens (“CL”), multi-electron beam generation (“MBC”), and projection image formation (“Projection”).

[0034] As shown in FIG. 1, the system 100 has an electron beam source that generates an electron beam 113. This electron beam source can be assumed to have a tip (emission end) 101. The electron beam source can be assumed to also have a suppression electrode and an extraction electrode. The electron beam source can be a thermionic field emission (TFE) or cold field emission (CFE) source that emits electrons from an emitter tip. Those electrons are focused by a gun lens (GL) 102 to form a large-size electron beam 113. The large-current electron beam 113 is collimated (parallelized) by a collimation lens 126 to form a telecentric beam, thereby illuminating an aperture array 103 (AA). The aperture array 103 can also be called a micro-aperture array. An additional electron energy acceleration element, such as an anode, may be used in combination with the electron beam source.

[0035] There is a beam limiting aperture (BLA) 125 downstream of the gun lens 102, by which the total beam current can be selected; the aperture array 103 is illuminated with this total beam current, where the electron beam 113 is branched into beamlets 114. The beam current for each individual beamlet 114 can be selected using the aperture array 103. The aperture array 103 has one hole for each individual beamlet 114. These holes can be circular, hexagonal or other shapes. For simplicity of description, three beamlets 114 are shown in FIG. 1, but other numbers are also possible. For example, at least 100 (e.g., more than 300) beamlets 114 can be provided. Downstream of the aperture array 103, a microlens array (MLA) 106 focuses each beamlet 114 onto the intermediate image plane (IIP). Each microlens in the microlens array 106 can be a magnetic lens or an electrostatic lens. Examples of what can be a magnetic microlens are multiple magnetic pole pieces operated by coil excitation or a permanent magnet. Examples of what can be an electrostatic microlens are an electrostatic Einzel lens and an electrostatic acceleration / deceleration unipotential lens.

[0036] This system 100 has a collimation lens 126 on the path of the electron beam 113. The collimation lens 126 can be an electrostatic lens or a magnetic lens, which focuses the divergent electron beam from the gun into a telecentric beam state and can illuminate the MBC module. The collimation lens 126 can reduce the gun spherical aberration associated with off-axis electron beamlets, and thus help improve the throughput due to an increase in the number of beamlets. The beam limiting aperture 125 is on the path between the collimation lens 126 of the electron beam 113 and the electron beam source.

[0037] The micro stigmer array 104 (MSA) is disposed on the path between the aperture array 103 and the transfer lens 112 of the electron beam 113, whereby the aberration can be corrected for each beamlet 114. The micro deflector array 105 (MDA) is disposed on the path between the micro stigmer array 104 and the transfer lens 112 of the electron beam 113. By the micro deflector array 105, the distortion of each beamlet 114 can be corrected and / or the wafer 107 can be scanned by each beamlet 114 within a given sub-FOV. The micro lens array 106 (MLA) is disposed on the path between the micro deflector array 105 and the transfer lens 112 of the electron beam 113. From the word "micro", not only can the sizes of the members be read, but it can also be understood that those members are used in combination with the beamlets 114. The beamlets 114 are thinner than the electron beam 113.

[0038] The stage 108 is configured to hold the wafer 107 on the path of the beamlets 114 of the electron beam 113. The objective lens 109 and the Wien filter 110 are upstream of the stage 108.

[0039] In the electron source, electrons are emitted from the tip 101, and these electrons are accelerated and focused by the gun lens 102 to form a large-sized electron beam 113. This electron beam 113 has a large beam current, which is collimated by the collimation lens 126 to form a telecentric beam, and the aperture array 103 is illuminated. The tip emission angle α in FIG. 1 when the angular intensity or source luminance is given to characterize the electron beam 113 is provided. Using the beam-limiting aperture 125 located downstream of the gun lens 102, the total beam current for illuminating the aperture array 103 is selected. Using the aperture array 103, the beam current is selected for each individual beamlet 114. Downstream of the aperture array 103, the microlens array 106 focuses each beamlet 114 onto the intermediate image plane. This intermediate image plane is the object plane of the projection imaging optical system within the lower column.

[0040] There is a beam crossover (xo) between the transfer lens 112 and the objective lens 109. The beamlets 114 formed by the upper column and reaching the intermediate image plane are projected onto the wafer (WF) 107 at a desired magnification by the transfer lens 112 and the objective lens 109. This magnification can be configured and set so that any beam blur is minimized for each individual beamlet on the wafer. The optimal magnification is D i ,D o When taking D and D as the multi-electron beam (MB) FOVs in the wafer plane (image plane) and the intermediate image plane (object plane) respectively, D i / D o is given by. The transfer lens 112 can select a desired position of the beam crossover (xo), that is, a position where the axial aberration, off-axis aberration, and electron-electron Coulomb interaction are balanced and the spot size for the total of each beamlet 114 is minimized.

[0041] To inspect and review the wafer, secondary electrons (SE) and / or backscattered electrons (BSE) emitted from the wafer 107 due to the collision of electrons of each primary beamlet 114 can be branched from the optical axis by the Wien filter 110 and deflected toward the detection array 111.

[0042] As shown in FIG. 2, the objective lens 109 is on the path of the beamlet 114 of the electron beam 113. This objective lens 109 has an upper pole piece 115 and a lower pole piece 116. An objective lens coil 117 is disposed on the upper pole piece 115. A charge control plate 118 is disposed on the lower pole piece 116. An acceleration electrode 119 is disposed on the path between the upper pole piece 115 and the lower pole piece 116 of the electron beam 113. A scanner 120 is disposed on the upper pole piece 115. By this scanner 120, scanning by all the beamlets 114 can be performed simultaneously in the same manner (e.g., raster scanning) and in the same scanning FOV. The scanning FOV size is given by the beamlet pitch on the wafer 107.

[0043] The Wien filter 110 is on the path between the objective lens 109 of the beamlet 114 of the electron beam 113 and the electron beam source. The Wien filter 110 according to one embodiment has an electrostatic deflector 121 and a magnetic deflector 122.

[0044] The Wien filter 110 can be removed or made inoperative during operation. This can be beneficial in multi-electron beam lithography or multi-electron beam review and inspection, such as those using an annular detection array (e.g., one having a central detector axis identical to the primary beam optical axis). Lithography is direct writing onto the wafer photoresist, and there is no need to collect secondary electrons or separate primary electrons (PE) from secondary electrons. When using the annular detection array, each SE beamlet directly impinges on the fixed sub-detector, so there is no need to change the direction (deflection). Therefore, a Wien filter is not required for such applications. In simpler applications, for example, the beam energy and landing energy are fixed for a specific use, and the SE trajectory is fixed. When the operating conditions change, the SE trajectory also changes. In this case, the fixed annular detection array is not sufficient to meet all applications.

[0045] The transfer lens 112 (TL) is on the path between the Wien filter 110 and the electron beam source of the beamlet 114 of the electron beam 113. The transfer lens 112 has pole pieces 123 and a transfer lens coil 124. The transfer lens 112 can be shaped as a magnetic lens to improve the off-axis optical performance with a shaped beam or a multi-electron beam.

[0046] The objective lens 109 can have an electrostatic part and a magnetic part. The electrostatic part of the objective lens 109 can have a ground electrode, an acceleration electrode 119 with a voltage Va, a charge control plate 118, and a stage 108. By using one or more of these members, the wafer 107 can be charged, and the electrons can be retarded (decelerated) to convert the beam energy into the landing energy on the wafer 107. For example, when the electron beam in the column is 30 keV, if the wafer 107 is biased at -29 kV, the electron beam landing energy can be made 1 keV. 1 keV can be used in electron beam inspection and review, but other values can also be adopted. To charge the wafer 107 by the extraction field on the wafer surface, it is necessary to bias the charge control plate 118 according to the essential conditions of the application. The magnetic part of the objective lens 109 can have an upper pole piece 115, a lower pole piece 116, and a coil. The upper pole piece 115 and the lower pole piece 116 can be made of a magnetic material. The upper pole piece 115 may be connected to a ground electrode such as a Wien filter shield or a scanner shield. The lower pole piece 116 may be connected to the electrode of the charge control plate 118. The outer gap between the lower pole piece 116 and the upper pole piece 115 may be sealed with an insulator material.

[0047] The detection array 111 (DA) is configured to receive a secondary electron beam from the wafer 107 on the stage 108. A measured value or an image can be generated using the signal from the detection array 111. The detection array 111 can be electronically communicated with a processor to contribute to functions such as image generation, inspection, metrology, and others.

[0048] The multi-electron beam crossover (xo) in FIG. 1 is arranged near the acceleration electrode 119 to reduce the Coulomb interaction (CI); this is because if not reduced, the CI-induced optical blur may become dominant over the resolution of each beamlet 114 on the wafer 107. The Coulomb interaction can mainly occur at the crossover. By arranging the crossover near the acceleration electrode 119, the electrons can be accelerated and the Coulomb effect can be reduced.

[0049] With the scanner 120 in FIG. 2, scanning can be simultaneously performed by all the electron beamlets 114 everywhere in the FOV. With the micro deflector array 105 in FIG. 1 where each micro deflector can be independently controlled, scanning by each beamlet can also be performed individually.

[0050] The Wien filter 110 can have an electrostatic deflector and a magnetic deflector, and the electrostatic field (E field) and the magnetic field (B field) thereof can be orthogonal to each other. This is shown by the electrostatic dipole 121 and the magnetic dipole 122. FIG. 3 shows a Wien filter configured with an octupole electrostatic deflector and an octupole magnetic deflector. Taking the magnetic deflection field as an example, the cross-sectional appearance of the octupole magnetic deflector (MD) is shown in FIG. 3. Eight magnetic pole pieces are arranged as an octupole deflector according to the rotation symmetry principle, and coils with the same number of turns are wound around each pole piece. Those pole pieces are shielded as shown in FIG. 3. By appropriately setting the current in the coil, the distribution of the magnetic deflection field in a wide central area can be made quite uniform, thereby minimizing the coma aberration related to the outermost beams among several hundred beamlets. For example, when supplying the coil current, I y = 1 unit, I x = 0 unit, scaling factor a = 2 -1 / 2 By setting like this, a uniform B field can be obtained along the y-axis direction. According to another example, when supplying the coil current, I x = 1 unit, I y = 0 unit, a = 2 -1 / 2 By setting like this, a uniform B field can be obtained along the x-axis direction.

[0051] The field of the octupole electrostatic deflector (ED) and the one collected from the dashed-line area in FIG. 3 are shown in FIG. 4. When setting the deflection voltage, V x = 1 unit, V y = 0 unit, scaling factor a = 2 -1 / 2Then, it can be seen from the straight equipotential lines that an electrostatic deflection field is generated along the x-axis direction and the field distribution becomes fairly uniform in a wide central area.

[0052] When computer simulations were carried out, as shown in FIG. 5, several hundred electron beamlet image forming portions on the wafer could be identified. The example of FIG. 5 is an example where 331 electron beamlets are in a hexagonal distribution, but other numbers or distributions of beamlets may also be acceptable. The multi-electron beam FOV on the wafer 107 (i.e., D in FIG. 1) i is defined by the farthest corner beam to the farthest corner beam, and this can be about 200 - 300 μm for D on the wafer 107. i Therefore, when the optical reduction ratio from the IIP to the wafer is 8X, D can be about 1600 - 2400 μm at the IIP in FIG. 1. o It can be such.

[0053] When further computer simulations were carried out, as shown in FIG. 6, the relationship between the spot size and the beam current could be identified. The total beam current is the sum of the currents of all the beamlets (e.g., 331 beams in FIG. 5), and the spot size reflects the optical performance (resolution), taking into account all optical blurs (e.g., blurs caused by lens aberrations and blurs caused by electron-electron Coulomb interactions). In the simulation, acceleration voltages Va of 0 kV, 25 kV, 50 kV, and 100 kV were individually applied. For each individual acceleration voltage Va, the beam was focused on the wafer 107 using the excitation (coil current) of the objective lens 109. By setting the location of the crossover (xo) near the acceleration electrode 119 (Va), the beam energy was increased to (BE + Va) near the crossover; where BE is the in-column beam energy before the electrons are accelerated. According to FIG. 6, the acceleration voltage Va on the booster can help improve the multi-electron beam resolution.

[0054] FIG. 5 shows the distribution of the electron beam spots on the wafer 107, and this can be further used to map some xy-plane characteristics provided in the optical system in FIGS. 1 and 2.

[0055] The aperture array 103 in FIG. 1 is shown in FIG. 5. Since the holes of the aperture array 103 are close to rotational symmetry within the optical system in a hexagonal shape, it is preferable to distribute them in a hexagonal pattern. Other shapes can also be adopted. Using the size of each hole in the aperture array 103, the electron current of the beamlets can be selected. The number of holes in the aperture array 103 becomes the number of beamlets.

[0056] The total number of MB (multi-electron beams) in FIG. 1 (MB tot ) can be scaled by Equation (1). MB tot= (1 + 3M x 2 ) / 4 (1)

[0057] M in Equation (1) x is the total number of beamlets along the x-axis in FIG. 5. For example, within 5 rings of the hexagonal distribution beamlet group in FIG. 5, since the total number of beamlets along the x-axis is M x = 11, the total number of beamlets is MB tot = 91. Within 10 rings, M x = 21 and MB tot = 331.

[0058] FIG. 5 can also show the positions and sizes of each micro-stigmator in the micro-stigmator array 104, each micro-deflector in the micro-deflector array 105, and each micro-lens in the micro-lens array 106 in FIG. 1. FIG. 5 can also show the intermediate image plane in FIGS. 1 and 2, that is, the plane where the multi-electron beams form an image to form an intermediate spot size array with a hexagonal distribution.

[0059] FIG. 5 also shows the plane of the specimen (wafer) in FIGS. 1 and 2, i.e., the plane in which the multi-electron beam forms a hexagonal distribution of the final spot size array. The multi-electron beam FOV is reduced by 1 / M from the intermediate image plane to the wafer plane. M is the optical magnification from the IIP to the wafer. This optical magnification is M = D i / D o in FIG. 1, where D i , D o denote the FOVs at the image plane and the object plane of this projection optical system, respectively. Let the number of hexagonal rings be n and the interval between beamlets be p, then D o = 2xn = 2np.

[0060] In FIG. 5, the total number of beamlets is calculated by equation (1), and each beamlet is indicated by the ring number (the n-th ring, n = 0, 1, 2, 3,...) and the polar angle. For example, (the 10th ring, 60°), (the 10th ring, 120°) indicate the outermost corner beamlets at 60° and 120°, respectively.

[0061] Since it is desired to obtain higher throughput with more electron beamlets, it is advisable to separate the secondary electron beams in FIGS. 1 and 2 from the optical axis and deflect them towards the lateral detection array 111 to reduce the crosstalk between the secondary electron beams. This separation can be achieved by the Wien filter in FIGS. 1 and 2. However, in this case, energy dispersion blur may enter the beamlets on the wafer 107; this is because any electron emitted from the electron source exhibits energy dispersion (e.g., about 1 eV in the case of a TFE source).

[0062] According to what is shown in the computer simulation of the optical system in FIGS. 1 and 2, as shown in FIG. 7, the electrons that form the beamlet spots on the wafer spread and distribute in the same direction; this is because the energy dispersion angle of the electrons distributes exclusively along the equilibrium direction of the electrostatic force and magnetic force of the Wien filter (e.g., the x-axis or y-axis direction). Even when the electrons are in a rotating state and are focused by the magnetic objective lens 109, they continue to maintain the energy dispersion distribution in the same direction. FIG. 7 shows seven source energy dispersion blurs of a typical beamlet, namely those at 0°, 60°, 120°, 180°, 240°, 300° in the center (the 0th ring) and the 10th ring respectively.

[0063] The source energy dispersion blur that appears in any beamlet due to the presence of the Wien filter 110 can be corrected with a global tilt optical column as shown in FIG. 8; θ in the figure p is the column tilt angle formed by the primary multi-electron beam with respect to the center of the Wien filter 110, and θ s is the angle of the secondary electron beam (SEB) towards the detection array 111. Since the path of the electron beam 114 faces the first azimuth when exiting the transfer lens 112 and faces a second azimuth different from the first azimuth when exiting the Wien filter 110, the first azimuth will form a non-parallel angle with respect to the second azimuth. The θ s is the secondary electron beam deflection angle by the Wien filter 110. When the angles θ p and θ s meet certain conditions, any energy dispersion blur in FIG. 7 can be corrected. In the case of using a multi-beam, the secondary electron angle θ s can be set to about 10° - 45°, and the primary electron beam angle θ s can be given by various equations.

[0064] The angles θ p and θ sWhen the relationship defined by Equation (2) is satisfied, the source energy dispersion blur generated by the action of the electrostatic and magnetic deflection fields in the Wien filter 110 can be canceled out with each other. Using Equations (2) and (3), the correction of the energy dispersion can be balanced. θ p / θ s =(1 - ρ) / (1 + 2(1 - ρ) 1 / 2 ) (2) ρ = LE / V p =(V p -V s ) / V p =1 - V s / V p (ρ = 0~1) (3)

[0065] In Equations (2) and (3), V p ,V s are the energy voltages of the primary electron beam and the secondary electron beam in the Wien filter region, and LE is the landing energy of the primary electron beam on the wafer. For example, if LE = 1 kV and V p = 30 kV, then ρ = 1 / 30, and thereby θ p / θ s = 0.33 is given. That is, the column tilt angle θ p becomes 1 / 3 of the detection array angle θ s . In a multi-electron beam system having several hundred beamlets, the SEB angle θ s can be made relatively large (e.g., θ s = 15°), and thus the column tilt angle can be made θ p = 5°.

[0066] Equation (2) satisfies not only the cancellation condition of the source energy dispersion but also the alignment condition of the primary beam. If this alignment condition is satisfied, the tilted primary beam forming the angle θ p in FIG. 8 is aligned with the optical axis of the objective lens 109. This alignment condition deflects the primary beam by an angle of θ p in the y direction by the electrostatic field of the Wien filter 110, or deflects it in the -y direction by 2θ p by the magnetic field of the Wien filter 110.It can be required to deflect at a certain angle. Here, it is assumed that the cross-section of FIG. 8 is in the yz plane.

[0067] The electrostatic and magnetic deflection fields of the Wien filter 110 can be generated by octupoles (octapoles) in FIGS. 3 and 4, respectively. From the fact that these deflection fields are fairly uniform within a wide central region (i.e., the equipotential lines in FIG. 4), the cancellation of energy dispersion can also be carried out fairly uniformly within a wide region. Therefore, the energy dispersion blur of several hundred beamlets in FIG. 7 can be simultaneously cancelled by the large-area Wien filter.

[0068] When the primary beam landing energy (LE) is changed within a certain range, the secondary electron beam may have excessive or insufficient deflection compared to the angle θ of a given (i.e., fixed) detection array 111. However, such misaligned secondary electron beams can also be corrected by an aligner (i.e., deflector) within the secondary electron collection optical system (not shown). s Compared with θ, there may be excessive or insufficient deflection. However, such misaligned secondary electron beams can also be corrected by an aligner (i.e., deflector) within the secondary electron collection optical system (not shown).

[0069] Due to the presence of the Wien filter 110 in FIG. 8, spherical aberration blur is introduced into the beamlets on the wafer 107. As shown by computer simulation of the optical system in FIG. 8, as shown in FIG. 9, the electrons of the spherical aberration spots (elliptical spots) on the wafer 107 are distributed along the same direction. Even when the electrons are in a rotating state and are focused by the magnetic objective lens 109, they still continue to maintain the same direction on the wafer 107 under the influence of the Wien filter 110. FIG. 9 shows seven spherical aberration blurs of a typical beamlet, i.e., those at the center (the 0th ring) and at 0°, 60°, 120°, 180°, 240°, and 300° of the 10th ring.

[0070] The spherical aberration blur that appears in every beamlet due to the presence of the Wien filter 110 can be corrected with a global stigmer. As shown in Fig. 10, the eight plates provided in the octupole (electrostatic) deflector in Fig. 10 can be used as two stigmers, and two groups of voltages, namely ±Va and ±Vb, can be applied thereto. For example, the equipotential lines in Fig. 10 are given through computer simulation with voltages Va = 1 unit and Vb = 0 unit applied. In this example, due to the generated electrostatic field distribution, the electron beam is focused along the y-axis direction and defocused along the x-axis direction. By changing Va and Vb, the combined electrostatic force can be changed over the polar angle ranging from 0° to 360°. Therefore, by appropriately selecting the voltages Va and Vb, spherical aberration blur in any direction can be corrected.

[0071] By implementing the correction of source energy dispersion and spherical aberration using the global tilt column in Fig. 8 and the global stigmer in Fig. 10, as shown in Fig. 11, the final spot sizes at various locations in a wide field of view for 331 beamlets can be made to satisfy the desired resolution and image formation uniformity. As further shown by computer simulation, the relationship between the spot size and the beam current can also be characterized by the plot in Fig. 6 here. Therefore, source energy dispersion and spherical aberration will all be removed without affecting the final resolution. Similar to Figs. 7 and 9, (a), (b), (c), (d), (e), (f), (g) in Fig. 11 are the final spots of the beamlets at the center (the 0th ring) and at 0°, 60°, 120°, 180°, 240°, 300° of the 10th ring, respectively.

[0072] As shown in Fig. 12, the second Wien filter 127 can be added along the beamlet path between the Wien filter 110 and the transfer lens 112. This second Wien filter has a second electrostatic deflector 128 and a second magnetic deflector 129.

[0073] The source energy dispersion blur in FIG. 7 caused by the presence of Wien filters can be corrected using two Wien filters. In FIG. 12, a downward Wien filter 110 close to the wafer 107 can be used to deflect the secondary electron beam towards the lateral detection array 111. A second (i.e., upward) Wien filter 127 close to the transfer lens 112 can be used to compensate for the energy dispersion. The energy dispersion generated by the Wien filter 127 can be used to compensate for the energy dispersion generated by the Wien filter 110. Since the use of two Wien filters is also described in Patent Document 1, the entire content thereof will be incorporated by reference.

[0074] In the acceleration magnetic objective lens system in FIGS. 2 and 8, the resolution of the multi-electron beamlets is improved by increasing the acceleration voltage Va. The acceleration voltage Va can be increased to a level without arc generation and such that those electron beamlets are stably focused on the wafer by excitation.

[0075] When correcting the source energy dispersion blur in FIG. 7 using two Wien filters, although the optical column of the primary beam is a straight column, correction performance similar to that provided by the inclined column in FIG. 8 can be provided.

[0076] FIG. 13 is a flowchart of method 200. At 201, the electron beams are focused by directing a plurality of beamlets through a transfer lens downstream of the electron beam source. For example, at least 100 beamlets (e.g., more than 300 beamlets) can be present. At 202, the secondary electron beam is separated from the beamlets by directing the beamlets through a Wien filter downstream of the transfer lens. This Wien filter can have an electrostatic deflector and a magnetic deflector. At 203, the beamlets are directed through the upper pole piece of the objective lens. This objective lens is downstream of the Wien filter. The Wien filter can be configured to simultaneously adjust dispersion and aberration with respect to the beamlets. At 204, the beamlets are directed through an acceleration electrode downstream of the upper pole piece. This acceleration electrode can be made to be able to change the resolution of the beamlets. At 205, the beamlets are directed through a charge control plate disposed inside the lower pole piece of the objective lens. This charge control plate is disposed on the opposite side of the objective lens as seen from the upper pole piece. At 206, the beamlets are directed at the wafer. At 207, the secondary electron beam or beam group from the wafer is received by a detection array.

[0077] Method 200 can further be made to generate an electron beam using an electron beam source and convert the electron beam into a plurality of beamlets. The electron beam can also be directed through a collimation lens and a beam limiting aperture disposed on the electron beam path between the electron beam source and the transfer lens.

[0078] Method 200 can further be made to perform scanning by the beamlets using a scanner (i.e., deflector) disposed on the upper pole piece.

[0079] The method 200 can further be such that an electron beam is branched into beamlets using an aperture array. This aperture array is disposed on the electron beam path between the electron beam source and the transfer lens. Those beamlets are directed through a microstigmator array disposed on the beamlet path between the aperture array and the transfer lens, a microdeflector array disposed on the beamlet path between the microstigmator array and the transfer lens, and a microlens array disposed on the beamlet path between the microdeflector array and the transfer lens. The electron beam can be made a telecentric beam upstream of the aperture array.

[0080] The method 200 can further be such that the Wien filter is used to change the direction of those beamlets so that the beamlets are directed at an angle at the exit of the Wien filter with respect to their orientation when entering the Wien filter.

[0081] The method 200 can be such that those beamlets are directed through a second Wien filter on the beamlet path between the Wien filter and the transfer lens.

[0082] The method 200 can be used when system conditions are changed. For example, the method 200 can be used when parameters such as beam energy, landing energy, beam current, FOV, and others are changed.

[0083] Although the present disclosure has been described in connection with one or more specific embodiments, it will be understood that other embodiments of the present disclosure can be made without departing from the technical scope of the present disclosure. That is, the present disclosure is considered to be limited only by the appended claims and their reasonable interpretation.

Claims

1. A system comprising: an electron beam source that generates an electron beam and has a tip, a suppression electrode, and an extraction electrode; a stage configured to hold a wafer on a path of the electron beam; on the path of the electron beam; an upper pole piece; a lower pole piece; an objective lens coil disposed on the upper pole piece; a charge control plate disposed on the lower pole piece; an acceleration electrode disposed on the path of the electron beam between the upper pole piece and the lower pole piece; and a scanner disposed on the upper pole piece; an objective lens having the same; a Wien filter disposed on the path of the electron beam between the objective lens and the electron beam source; a transfer lens having a pole piece and a transfer lens coil on the path of the electron beam between the Wien filter and the electron beam source; a detection array configured to receive at least one secondary electron beam from the wafer on the stage; a system comprising the same.

2. The system according to claim 1, wherein the Wien filter has an electrostatic deflector and a magnetic deflector.

3. The system according to claim 1, further comprising: a collimation lens on the path of the electron beam; a beam limiting aperture on the path of the electron beam between the collimation lens and the electron beam source; a system comprising the same.

4. The system according to claim 1, further comprising: an aperture array disposed on the path of the electron beam and configured to branch the electron beam into a plurality of beamlets; a microstigmator array disposed on the path of the electron beam between the aperture array and the transfer lens; a microdeflector array disposed on the path of the electron beam between the microstigmator array and the transfer lens; a microlens array disposed on the path of the electron beam between the microdeflector array and the transfer lens; a system comprising the same.

5. The system according to claim 4, wherein each of the plurality of beamlets includes at least 100 beamlets.

6. The system according to claim 4, wherein the electron beam is a telecentric beam upstream of the aperture array.

7. The system according to claim 1, wherein the path of the electron beam is oriented in a first direction when exiting the transfer lens, is oriented in a second direction different from the first direction when exiting the Wien filter, and thus the first direction forms a non-parallel angle with the second direction.

8. The system according to claim 1, further comprising a second Wien filter on the path of the electron beam between the Wien filter and the transfer lens.

9. The system according to claim 8, wherein the second Wien filter has a second electrostatic deflector and a second magnetic deflector.

10. A method comprising: focusing the electron beams by directing a plurality of beamlets through a transfer lens downstream of an electron beam source; separating at least one secondary electron beam from the beamlets by directing the beamlets through a Wien filter downstream of the transfer lens; directing the beamlets through an upper pole piece of an objective lens downstream of the Wien filter; directing the beamlets through an acceleration electrode downstream of the upper pole piece; directing the beamlets through a charge control plate disposed inside a lower pole piece of the objective lens and disposed on the opposite side of the objective lens as viewed from the upper pole piece; directing the beamlets onto a wafer; and receiving the at least one secondary electron beam from the wafer at a detection array. A method.

11. The method according to claim 10, wherein the plurality of beamlets includes at least 100 beamlets.

12. The method according to claim 10, further comprising: generating an electron beam using the electron beam source; and converting the electron beam into the plurality of beamlets. A method.

13. The method according to claim 11, further comprising: directing the electron beam through a collimation lens and a beam limiting aperture disposed on the path of the electron beam between the electron beam source and the transfer lens.

14. The method according to claim 10, further comprising scanning with the beamlets using a scanner disposed on the upper pole piece.

15. The method according to claim 10, wherein the Wien filter has an electrostatic deflector and a magnetic deflector.

16. The method according to claim 10, further comprising: using an aperture array disposed on the path of the electron beam between the electron beam source and the transfer lens to branch the electron beam into the beamlets; directing the beamlets through a microstigmator array disposed on the path of the beamlets between the aperture array and the transfer lens; directing the beamlets through a microdeflector array disposed on the path of the beamlets between the microstigmator array and the transfer lens, and directing the beamlets through a microlens array disposed on the path of the beamlets between the microdeflector array and the transfer lens. Method.

17. The method according to claim 16, wherein the plurality of beamlets include at least 100 beamlets.

18. The method according to claim 16, wherein the electron beam is a telecentric beam upstream of the aperture array.

19. The method according to claim 10, further comprising changing the direction of the beamlets using the Wien filter such that the beamlets are directed at an angle of exit from the Wien filter with respect to their orientation upon entry into the Wien filter.

20. The method according to claim 10, further comprising directing the beamlets through a second Wien filter disposed on the path of the beamlets between the Wien filter and the transfer lens.

21. The method according to claim 10, wherein the acceleration electrode is configured to be able to change the resolution of the beamlets.

22. The method according to claim 10, wherein the Wien filter is configured to be able to simultaneously adjust dispersion and astigmatism with respect to the beamlets.

Images