JP5292132B2 - Image forming apparatus - Google Patents

Description

  The present invention relates to an image forming apparatus that forms an image based on design data of a semiconductor device or the like, and in particular, an image forming apparatus that forms an image or the like used for condition setting at the time of creating a recipe for a charged particle beam apparatus or defect inspection About.

  A charged particle beam apparatus typified by a scanning electron microscope or the like can measure and inspect a fine pattern, and is particularly used for measurement of a semiconductor device. In particular, in an apparatus for inspecting defects and foreign matter, defect extraction is performed by comparing a reference image acquired in advance with a sample image obtained by scanning with an electron beam or the like. Since the reference image is acquired in an area having the same pattern shape as the sample image, the difference between the reference image and the sample image can be recognized as a foreign object or a defect.

  On the other hand, a photomask or the like used to expose a pattern on a semiconductor wafer generally does not have a pattern having the same shape in the vicinity. Japanese Patent Application Laid-Open No. 2004-228561 describes a technique for extracting defects by comparing design data and sample images.

  It is also known that the sample is charged when the electron beam is scanned over the sample. Patent Document 2 discloses that a secondary electron emission efficiency δ indicating a ratio of an incident electron beam amount and an emitted secondary electron has a reaching energy (acceleration voltage) close to 1.0 so as not to cause sample charging. Scanning a sample is described.

JP 2005-277395 A JP-A-7-14537

  As described in Patent Document 1, in a comparative inspection of defects and the like, by using a reference image as design data, it is difficult to acquire a reference image such as a photomask or to inspect a sample before manufacturing the sample. Although it is possible to form a reference image at the design data stage and create a recipe using the reference image, the design data is only layout data indicating the ideal shape of the pattern. Therefore, the data is different from the electron microscope image, and if the design data is used as a reference image for comparative inspection, a portion that is not originally a defect or a foreign substance may be recognized as a defect or a foreign substance. In particular, charging caused by scanning of an electron beam or the like is information that does not exist in the design data. If the sample is charged, the possibility of erroneous recognition becomes higher.

  As described in Patent Document 2, if a beam having a secondary electron emission efficiency δ of 1.0 is always used, an image that is not charged theoretically can be formed. There is a factor, and it is difficult to completely eliminate the influence of charging only by controlling the reached energy.

  Hereinafter, an image forming apparatus for forming an image reflecting the influence of charging or the like without irradiating the sample with a charged particle beam will be described.

  In order to achieve the above object, information on a region including a range irradiated with a charged particle beam is acquired from design data of a semiconductor device, and based on the information, the charged particle beam irradiation range is measured in units of pixels. The present invention proposes an image forming apparatus that calculates the amount of electrons detected in the pixel portion when a charged particle beam is irradiated to the image and arranges luminance information obtained based on the amount of electrons to form an image.

  According to the configuration as described above, since it is possible to reproduce the luminance according to the detected amount of electrons on the same layout as the design data indicating the ideal shape of the pattern, the charged particle beam was irradiated to the ideal shape. It is possible to form images that are sometimes obtained.

It is a figure explaining the process which implements the simulation which considered charging about the design data of a semiconductor pattern, and applies the simulation result to a defect inspection of a device. It is a figure explaining the process which performs the simulation (pre-measurement process) which considered charging, and performs a measurement based on the simulation result concerned. The flowchart which shows the process of simulation. The figure explaining an example of the range which calculates a secondary electron orbit. The flowchart explaining the process of the simulation of the electric charge amount accumulate | stored for every irradiation point of an electron beam. The flowchart explaining the defect extraction process based on an image comparison. The figure explaining an example of the image of a resist hole and a horizontal direction line. The graph explaining the relationship between the irradiation energy of an electron beam, and secondary electron emission efficiency. The schematic block diagram of a scanning electron microscope.

  Optical pattern using comparison (die-to-die comparison) between patterns (die) on a semiconductor device in the pattern inspection of a wafer in the manufacturing process of a semiconductor integrated circuit when the same pattern exists in the vicinity. Inspection equipment is used. In the die-to-die comparison, a defect is detected by comparing a semiconductor device to be inspected with a pattern in the vicinity thereof.

  On the other hand, a method called die-to-database comparison is used for inspection of a photomask or the like in which the same pattern does not exist in the vicinity. In this method, a comparison is made between a die or pattern and a database containing the shape that the die should be. This database is obtained by converting mask data into an image.

  In the production of semiconductor integrated circuits, random defects caused by dust and the like and defects that occur repeatedly (systematic defects) are problematic. Repeated defects (systematic defects) are defined as defects that occur repeatedly on all dies on the wafer due to photomask defects, etc., and repeated defects are both inspected dies and adjacent dies to be compared. Therefore, it cannot be detected by die-to-die comparison. Therefore, wafer inspection by a die-to-database comparison method is required.

  On the other hand, in the SEM, the sample surface is irradiated with an electron beam and the sample is observed, so that the sample itself is charged by the electron irradiation. Due to the charging of the sample, various effects appear on the image such as image distortion and movement (drift), generation of shadows, variation in magnification, and decrease in luminance. As described above, even if the image obtained by measuring the charged sample is compared with the design data not including the influence of charging, the possibility that the patterns match within the required tolerance is low. For this reason, there is a problem that the comparison between the image in which the influence of charging appears significantly and the design data is classified as a defect regardless of the actual shape. In this case, the pattern classified as a defect due to charging and the pattern classified as a defect are handled in the same way even though a normal pattern is originally formed on the wafer. Will be. For this reason, it is necessary to confirm again after the inspection, and the inspection efficiency also decreases due to the decrease in inspection accuracy.

  In a general defect inspection system, since the influence of charging is not considered in the object to be compared with measurement data, the influence of charging cannot be removed at the time of comparison, and the comparison accuracy may be lowered. One of the causes is that it is difficult to distinguish between defects that should be detected originally and image distortion due to charging. If only one of the effects of charging is included, the matching accuracy of the two decreases, and it is unclear whether there is a mismatch due to charging or due to a structural defect. To do. Since the degree of charging differs depending on the image acquisition conditions, it is desirable to perform image formation according to the charging state.

  In this embodiment, in particular, when comparing information based on design data (layout data) and SEM images, measurement conditions (apparatus conditions) at that time are set, or an image to be compared for defect extraction is formed. Next, an image forming apparatus that enables these settings and image formation while suppressing the influence of charging will be described.

  In the present embodiment, a method will be described in which a simulation considering charging is performed on a design pattern before or during measurement, and a comparison with a measurement image is performed based on the simulation. As a result, it is possible to compare between patterns including charge, and an improvement in image comparison accuracy can be expected. In addition, by performing simulation under a plurality of conditions, it is possible to obtain in advance a scanning condition in which charging is difficult to occur for each measurement point, and reflect it in the measurement recipe.

  By preparing an image in consideration of the influence of charging before or during the measurement and comparing it with the measurement image, defect extraction in consideration of the influence of charging becomes possible. Thereby, it is possible to clearly separate the charge distortion and the defect, which are difficult to distinguish, and the defect detection accuracy is improved. In addition, by examining the measurement under a plurality of conditions before each measurement for each pattern, it is possible to obtain a condition in which charging is unlikely to occur, and an improvement in image matching accuracy can be expected.

  Below, the outline | summary of a charging simulation is demonstrated using drawing. FIG. 9 is a diagram for explaining an outline of a scanning electron microscope (hereinafter sometimes referred to as a scanning electron microscope: SEM) which is one type of charged particle beam apparatus. A voltage is applied between the cathode 1 and the first anode 2 by a high-voltage control power source 20 controlled by the control processor 30, and the primary electron beam 4 is drawn from the cathode 1 with a predetermined emission current. An acceleration voltage is applied between the cathode 1 and the second anode 3 by a high-voltage control power source 20 controlled by the control processor 30, and the primary electron beam 4 emitted from the cathode 1 is accelerated and proceeds to the subsequent lens system. To do.

  The primary electron beam 4 is converged by the converging lens 5 that is current-controlled by the lens control power source 21 controlled by the control processor 30, and after the unnecessary region of the primary electron beam 4 is removed by the diaphragm plate 8, the control processor 30. Are converged as a fine spot on the sample 10 by the converging lens 6 whose current is controlled by the lens control power source 22 controlled by the above and the objective lens 7 whose current is controlled by the objective lens control power source 23 which is controlled by the control processor 30. The objective lens 7 can take various forms such as an in-lens system, an out-lens system, and a snorkel system (semi-in-lens system). Further, a retarding method in which a negative voltage is applied to the sample 10 to decelerate the primary electron beam is also possible. Further, each lens may be composed of an electrostatic lens composed of a plurality of electrodes to which a controlled voltage is applied.

  The primary electron beam 4 is scanned two-dimensionally on the sample 10 by a scanning coil 9 that is current-controlled by a scanning coil control power source 24 controlled by a control processor 30. In the example of FIG. 9, a magnetic field type deflector is employed as the scanning deflector, but an electrostatic type deflector may be used. The secondary signal 12 such as secondary electrons generated from the sample 10 by the irradiation of the primary electron beam travels to the upper part of the objective lens 7 and is then separated from the primary electrons by the orthogonal electromagnetic field generator 11 for secondary signal separation. And detected by the secondary signal detector 13. The signal detected by the secondary signal detector 13 is amplified by the signal amplifier 14 and then transferred to the image memory 25 and displayed on the image display device 26 as a sample image. A two-stage deflection coil (objective lens aligner) 16 that is current-controlled by an objective lens aligner control power supply 27 controlled by the control processor 30 is disposed at the same position as the scanning coil 9, and primary electrons for the objective lens 7 are arranged. The four passing positions can be controlled two-dimensionally. The stage 15 can move the sample 10 in at least two directions (X direction and Y direction) in a plane perpendicular to the primary electron beam, thereby changing the scanning region of the primary electrons 4 on the sample 10. it can.

  The pointing device 31 can specify the position of the sample image displayed on the image display device 26 and obtain the information. From the input device 32, it is possible to specify an image capturing condition (scanning speed, total number of images), a visual field correction method, and the like, and output and storage of the image.

  An address signal corresponding to the memory location of the image memory is generated in the control processor 30 or in a control computer installed separately, converted to analog, and then supplied to the scanning coil control power supply 24. The address signal in the X direction is a digital signal that repeats 0 to 512, for example, when the image memory has 512 × 512 pixels (pixels), and the address signal in the Y direction reaches the address signal in the X direction from 0 to 512. It is a digital signal repeated from 0 to 512 that is sometimes incremented by one. This is converted into an analog signal.

  Since the address of the image memory 25 corresponds to the address of the deflection signal for scanning the electron beam, a two-dimensional image of the deflection region of the electron beam by the scanning coil 9 is recorded in the image memory. Note that signals in the image memory 25 can be sequentially read out in time series by a read address generation circuit synchronized with a read clock. The signal read corresponding to the address is converted into an analog signal and becomes a luminance modulation signal of the image display device 26.

  The apparatus described in this example has a function of forming a line profile based on detected secondary electrons or reflected electrons. The line profile is formed based on the amount of detected electrons when the primary electron beam 4 is scanned one-dimensionally or two-dimensionally on the sample 10, or the luminance information of the sample image, etc. For example, it is used for measuring a dimension of a pattern formed on a semiconductor wafer.

  In the description of FIG. 9, the control processor 30 is described as being integrated with or equivalent to the scanning electron microscope. However, the control processor 30 is not limited to this, and is described below with a processor provided separately from the scanning electron microscope body. Such processing may be performed. At that time, the detection signal detected by the secondary signal detector 13 is transmitted to the control processor 30 as an image, or a signal is sent from the control processor 30 to the objective lens control power source 23, the scanning coil control power source 24, etc. of the scanning electron microscope. A transmission medium for transmission and an input / output terminal for inputting / outputting a signal transmitted via the transmission medium are required.

  Furthermore, the apparatus of this example stores in advance, for example, conditions for observing a plurality of points on a semiconductor wafer (measurement locations, optical conditions of a scanning electron microscope, etc.) as a recipe, and measures and observes according to the contents of the recipe. The function to perform.

  Further, a program for performing the processing described below may be registered in a storage medium, and the program may be executed by a processor that supplies necessary signals to a scanning electron microscope or the like. That is, the example described below is also a description as a program or a program product that can be used in a charged particle beam apparatus such as a scanning electron microscope capable of acquiring an image.

  Further, the control processor 30 is connected to a data management device 33 that stores circuit pattern design data of the semiconductor device expressed in the GDS format, OASIS format, etc., and converts it into data necessary for SEM control. The data management device 33 has a function of creating a recipe for controlling the SEM based on the input design data.

  In addition, the edge portion of the obtained SEM image is contoured, and the contour line is also converted into vector data and stored in a predetermined file format such as GDS format. The contour line is formed, for example, by connecting portions having a predetermined luminance in the luminance distribution in a direction perpendicular to the line segment direction of the edge between different positions. This processing may be performed by the control processor 30 or may be performed by the data management device 33. Further, the scanning electron microscope may be controlled by a processor provided in the data management device 33 instead of the control processor 30. Further, the control processor 30 or the data management device 33 includes an interface for receiving SEM images and contour information from the scanning electron microscope main body or the control processor 30, and various types based on SEM images as described later. Measurement and inspection are possible.

  In the description of this example, the data management device 33 is described as a separate unit from the control processor 30, but the present invention is not limited to this. For example, the data management device 33 may be integrated with the control processor 30. .

  The data management device 33 or the like executes a charging simulation described below on the basis of design data of a semiconductor device stored in a storage medium provided in the control processor 30 or an external storage medium (not shown), and performs the simulation. A program for forming an image is provided. It also functions as a processing device for performing measurement and inspection using an image obtained by simulation.

  FIG. 1 is a diagram for explaining a process in which a simulation considering charging is performed on design data of a semiconductor pattern and the simulation result is applied to a defect inspection of a device. The layout data of the semiconductor device included in the design data is vector data indicating the contour of the pattern, and does not include luminance information that varies due to charging or the like provided by a scanning electron microscope or the like.

  If the characteristics of the SEM image acquired by the scanning electron microscope can be added to such layout data, an ideal-shaped SEM image free from pattern deformation or contamination can be reproduced. In the present embodiment, a method for reproducing an ideal SEM image by obtaining luminance information for each pixel of layout data expressed two-dimensionally and adding the luminance information to the layout data will be described. .

  The simulation in consideration of charging in this embodiment is based on irradiation of primary electrons to the sample surface, tracking and detection of secondary electron trajectories generated from the sample surface by irradiation, movement of the surface and internal accumulated charges, etc. It includes calculations such as pixel brightness information. Moreover, the part which acquires the image similar to a scanning electron microscope is included by moving an irradiation point on the same conditions as experiment (scanning), and obtaining the number of detection electrons in each irradiation point (pixel).

  At this time, if simulation is performed under multiple conditions for measurement parameters (electron microscope apparatus conditions) such as magnification, current amount, irradiation energy, and electrode voltage, multiple results are obtained. Accordingly, it is possible to determine the scanning conditions required by the user.

  In the present embodiment, a measurement recipe as described later is created with the optimum condition being a condition in which the influence of charging (deviation from design data, distortion) is small. In addition, an image including the effect of charging obtained at this time is stored in a memory or a hard disk drive, which is one form of the above-described storage medium, for comparison with an image obtained by measurement.

  Next, the sample is measured according to the scanning conditions determined by the simulation considering charging. The obtained measurement image includes the effect that the sample was charged by the measurement, and this was derived before measurement and compared with the simulation image considering the charge stored in the storage medium such as memory or hard disk drive. To do. Here, a comparison is made with the contours of the patterns obtained from both images, and if they do not match within the tolerance specified by the user, they are classified as having a defect at the measurement location.

  FIG. 2 is a diagram illustrating a process for performing a measurement (pre-measurement process) in consideration of charging and performing a measurement based on the simulation result. In this embodiment, as a pre-processing for measurement, the degree of charging at the measurement location is preliminarily evaluated from the design data (step 7) (step 8). In the pre-evaluation here, sample information such as the shape and material of the evaluation portion is extracted from the design data, and the determination is made based on the information. When the surface is uneven and the material is an insulator such as a resist or an oxide film, charging is likely to occur due to electron irradiation. In addition, since the degree of electrification varies depending on the scanning conditions, check the surface unevenness information, the type of surface material (existence of insulators), and whether or not there is a pattern parallel to the scanning direction. To do.

  When the degree of electrification at each measurement location satisfies the conditions specified by the user, a simulation considering the electrification is performed (steps 10 and 12). At this time, if a simulation is performed under a plurality of scanning conditions, conditions under which charging is less likely to occur can be obtained from the simulation results, and the scanning conditions can be reflected in actual measurement.

  FIG. 3 is a flowchart showing a simulation process. First, the calculation range is determined according to the measurement location (step 24). As shown in FIG. 4, the horizontal range includes an observation region (field of view (FOV)) + α in the horizontal direction, a spatial region β above the surface, and a region γ below the surface in the vertical direction. select.

  Here, α, β, and γ are ranges designated by the user, and in this embodiment, α: 5 μm, β: 2 μm, and γ: 2 μm. This is an area necessary for tracking the behavior (such as surface return) of secondary electrons and backscattered electrons emitted from the sample by irradiating primary electrons with α and β.

  In consideration of the behavior of electrons emitted from the sample, it is desirable to secure α about 10 times the observation region. As shown in FIG. 4, the structure includes three-dimensional information. For example, in a semiconductor device, different materials are stacked and another material is embedded inside, and the charged state varies depending on the material, film thickness, and shape.

  In this embodiment, since the structure is defined and simulated based on the three-dimensional information and the pattern configuration outside the FOV, it is possible to take into account the influence of the groundwork and the state inside the sample. Furthermore, since the penetration depth of electrons differs depending on the energy of the primary electrons to be irradiated, it is necessary to set γ in consideration of these. In addition, secondary electrons emitted from the sample diffuse in the space above the sample, but the electric field in the space changes as the sample is charged. For this reason, it is necessary to consider an area of about several μm in the space portion.

  Next, measurement (scanning) conditions (apparatus conditions) are set (step 25). Here, the condition designated by the user includes a magnification at the time of observation. Several types of scanning order, current amount, and primary electron acceleration voltage conditions for observation at a specified magnification are prepared, and a simulation is performed under each condition (step 26).

  For physical properties of each material, if possible, scan only one line or plane before simulation, and obtain the time change (electron and hole mobility) and yield from the experimental results (change in luminance). May be.

  Even if the same material is used, the influence of charging may be different when the supply source is changed. However, by preparing parameters from the above-described experiment, it is possible to reduce variations among materials. As a simulation method, any one of a finite element method (FEM), a boundary element method (BEM), and a Monte Carlo method (MC) is used. FIG. 5 shows the contents of processing performed in this calculation. FIG. 5 shows an example of FEM. First, electric field calculation (step 38) is performed on the range set in step 24 to prepare electric fields in each region. In the obtained electric field, primary electrons are generated (step 39), and the sample surface is irradiated under specified conditions (angle, energy).

  When primary electrons reach the surface of the sample (step 40), 2 in different states depending on the acceleration energy of the primary electrons, the incident angle, and the shape and material (sample information) of the surface on which the electrons have reached. Next electrons are generated (step 41). As the number of secondary electrons to be generated, energy, and angle, values known in literature are used as parameters. Since the secondary electrons generated from the surface are scattered again in the vacuum according to the energy and angle at the time of emission, the secondary electron trajectory is traced in the vacuum (step 43). In addition, since electrons that have entered the sample diffuse in the sample and then jump out into the vacuum, the trajectory is tracked in consideration of this point (step 42).

  When the surface of the sample is reached again due to surface charging or the like (step 44), secondary electrons are emitted, reflected, or attached to the surface depending on the energy, angle, and location (shape, material) of the electrons. Such redeposition causes charge transfer on the sample.

  When attached to the surface (step 45), the surface charge is changed (step 46) and the next primary electron is irradiated. In addition, when it comes out of the calculation structure without reaching the surface, or when it reaches the electrode or the like, it returns to the irradiation of the next primary electron. Here, secondary electrons that satisfy certain conditions (for example, secondary electrons that have reached a height of several μm from the surface) are counted and stored in a memory or a hard disk. Electrons that satisfy this condition can be handled as electrons detected in actual measurement, and the count of electrons can be made to correspond to the luminance of the image. In this case, the electronic count number and the luminance information may be stored in association with each other, and the luminance information corresponding to the calculated count number may be read.

  When all the primary electrons for one pixel have been irradiated, the advection diffusion of charges accumulated on the surface and the sample is simulated (step 47). The charge accumulated on the sample surface and inside moves for each material on the surface and inside depending on the surrounding electric field state. The timing of the advection diffusion calculation here can be changed such as every electron irradiation, every pixel, every 10 pixels, or every line. Next, the irradiation point of the primary electrons is moved, and after the electric field calculation, the primary electrons are irradiated to the next pixel (step 49).

  Although the FEM flow is shown in FIG. 5, it is possible to replace the electric field calculation part with BEM. In addition, when considering the influence of internal material charging in a three-dimensional structure, it is necessary to consider the diffusion of electrons in the sample. By using MC, it is possible to study the diffusion in the sample after the primary electrons reach the surface.

  There are the following methods as a method of calculating the secondary electron amount for each pixel or the luminance information obtained from the secondary electron amount. A description will be given below together with the handling of charging in this embodiment.

First, if the charging phenomenon is classified by the size of the area where the charge adheres, it is considered that it is divided into a global charge of about several tens of millimeters, a local charge of about 100 μm, and a micro charge of 10 μm or less. .
(A) Global electrification Mainly, the charge accumulated in the pre-processes of measurement and inspection by SEM, or secondary electrons are returned to the sample by the electric field formed by the facing electrode placed on the sample. Among the charges formed by reattachment, the charges are formed in a relatively wide range. The spatial change is moderate, and it can be obtained from the voltage value that is in focus by changing the retarding voltage directly measured by the electrostatic potentiometer. The optical magnification can be estimated from the objective lens current after autofocus, and the length measurement value can be corrected.
(B) Local charging The influence of return electrons at the time of observing surrounding devices and the electrons returned by the magnetic field and the electrodes are related, and the astigmatism (the point that is not detected as a result of being detected due to the bending of the electron orbit) is affected. The rate of change of deflection magnification is detected, and the influence of local charging is corrected based on this. Although it can be measured by using an energy filter, it is difficult to specify the spatial distribution due to the following supercharging of the micro charge.
(C) Microcharging Sample charging that occurs on a device structure or pixel scale on a small scale such as an LSI pattern. The sample charging produced by the electrons may cause distortion of the image due to bending of the trajectory of the primary electrons or the secondary electrons escaped from the sample, and contrast anomalies. Cause. Beam scanning further complicates the charging phenomenon. When the pixel is irradiated with electrons, the portion from which the secondary electrons are missing is positively charged, and the secondary electrons return to the sample. While the irradiation position is positively charged, its periphery is negatively charged due to return electrons. When moving to the next pixel, it is affected by the adjacent positive charge and the peripheral negative charge. When moving to another scan line, it is affected by the charging of the already irradiated location. At that time, since the surrounding charging changes with time, it is affected not only by space change but also time change.

  In addition, there are two types of electrification, in which the charge spreads relatively quickly on the surface of the sample and in other cases, the electrons are injected deep into the sample and the charge does not disappear for several minutes. As described above, micro-charging is a complicated phenomenon even by handling a flat sample. To elucidate, it is necessary to use a quantitative analysis tool in addition to the measurement means. In order to calculate the electric field by obtaining the potential distribution, the Poisson equation is discretized using the finite element method.

    Δφ = −ρ / ε (1)

  Here, φ is a potential, ρ is a charge density, and ε is a dielectric constant. First, φ is expressed as follows using a shape function Ni (x, y, z).

    φ = ΣNiφi (2)

  A residual R is defined by combining the following δφ (x) as a weighting function.

δφ (x) = ΣδφiNi (3)
R = εΔφ + ρ (4)

  The integral formula I using the weight function is defined as follows.

    I = ∫Rδφ (x) dV (5)

  Substituting the equations (3) and (4) into the above equation and performing partial integration gives the following equation.

I = ∫ (εΔφδφ (x) + ρδφ (x)) dV
= ∫ (ε ▽ φ) · nδφ (x) dS-∫ε ▽ φi ▽ iδφ (x) dV + ∫ρδφ (x) dV
... (6)

  Further, when the matrix is displayed, the following equation is obtained.

I = {δφ} T∫ [N] T [N] {(ε ▽ φ) n} dS + {δφ} T∫ [B] T [E] [B] dV {φ}
+ {Δφ} T∫ [N] T [N] {ρ} dV (7)

Here, [E] is a unit matrix, [N] is a matrix representation of a shape function, [B] is a derivative of [N],
{dφ (x) / dx} = [B] {φ} (8)
There is a relationship. When ∂I / ∂ {δφ} = 0, ∫ [N] T [N] {(εεφ) n} dS + {δφ} T∫ [B] T [E] [B] dV {φ}
= −∫ [N] T [N] {ρ} dv (9)
It becomes. This is the finite element formula to be solved.

Based on the above formula,
∫ [N] T [N] {(ε ▽ φ) n} dS ′ + {δφ} T∫ [B] T [E] [B] dV ′ {φ}
= −∫ [N] T [N] {ρ} dV ′ (10)
Get. Further, calculation is performed by replacing the integral part with Gauss-Legendre integral.

∫ [N] T [N] {ρ} dV '
= ΣiΣjΣk [N] T [N] ρ (ξi, ηj, ζk) wiwjwk (11)

  Here, ξi, ηj, ζk are the coordinate values of the sampling points in the element, and wiwjwk is a weighting factor. The trajectory of charged particles (electrons, ions) is tracked using the following equation of motion.

    mdv / dt = q (E + v × B), dx / dt = v (12)

Using this, a four-stage fourth-order Runge-Kutta method is used to determine temporal changes in the position and velocity (x, v) of the charged particles. Particles move anywhere in space, but in order to obtain the Coulomb force and Lorentz force acting on them, an electric field or a magnetic field is interpolated with respect to the particle position. For the efficiency of reflected electrons and secondary electrons generated when the electrons collide with the sample, experimental values of the yield versus energy curve are used. The following empirical formula is used for the angle dependence of the incident electrons. If the angle formed with the normal of the sample surface is φ, the reflected electron yield η and the secondary electron yield are η = B (η (φ = 0) / B) cosφ, δ = δ (φ = 0) / cosφ (( 13)
It is expressed. When one electron is incident on the sample, the sample is charged by (1-η−δ) by subtraction. The secondary electron emission angle θ follows the 'cos rule'. The following empirical formula is used for reflected electrons.

    δ (θ) ∝cosθ, η (θ) ∝cos [(π / 2) (φ−θ) / (π / 2−φ)] (14)

  Here, φ is the incident angle of primary electrons. In addition, the charge phenomenon is complicated by the fact that the charge accumulated in the sample changes temporally and spatially. Charging caused by electron irradiation spreads by movement due to an electric field and diffusion due to a density gradient. Therefore, carrier movement is described using a macro model in which carrier density moves by advection and diffusion as follows.

    ∂n / ∂t + ▽ j = 0, j = μnE−D ▽ n (15)

  Here, n is the carrier density (electrons, holes), j is the current density, μ is the mobility, and D is the diffusion coefficient. The first term of the current density j is an advection term, and if it is formulated as it is, instability and vibration will occur. Therefore, it is stabilized using the upwind difference corresponding to the sign of μnE. Since μ and D differ from sample to sample, it is necessary to estimate the parameters for each material from the experimental results.

The program is mainly composed of the following modules (a), (b), and (c).
(A) Pre-processing unit (Pre): Creates a three-dimensional system as a system creation, boundary condition input and beam / scanning condition input calculation system, and inputs boundary conditions and initial conditions. Physical property data and secondary electron emission data are prepared for each material. Input the beam current value, energy value, spot diameter, number of pixels, scanning order, etc. as a file.
(B) Charging calculation part: A tracking part of temporal and spatial change by charge distribution and charge transfer calculation by electric field and orbit calculation. Calculate electric field, trajectory and charge transfer loops for each pixel. It is assumed that the particles are traced to the N-order electrons, and the electric field-orbit calculation is repeated for the number, and then charge transfer is obtained. In the meantime, the charge density of the sample part changes spatially and temporally, and thereafter the trajectory of the irradiated and generated electrons changes. At that time, the arrival position of electrons, potential distribution, and the like are recorded. N = 1 is an incident electron, and N ≧ 2 is an electron from the sample surface.
(C) Post-processing section (Post): When a series of output loops such as image output and potential distribution by contouring the secondary electron distribution is completed, the recorded electron orbit data and the like are output for each condition. For example, an SEM image can be obtained by counting the number of secondary electrons satisfying the condition for each pixel or each region including a plurality of pixels and arranging luminance information reflecting the amount of secondary electrons. Using this program, it is possible to analyze the change in charge distribution and electron trajectory due to the beam, scanning conditions, etc., and investigate the effect of charging on the SEM image.

  Next, in step 27 in FIG. 3, the detected secondary electrons obtained by simulation are converted into an image. This images the secondary electron count stored in each pixel as the luminance of the pixel. Next, the contour of the pattern is extracted from the luminance difference of each pixel (step 28), and matching with the design data is performed (step 29, step 13 in FIG. 2). When the simulation is performed under a plurality of scanning conditions, the matching rate with the design data under each condition is obtained, and the higher the matching rate, the scanning condition is less affected by charging (step 30).

  The influence of charging is not only the distortion of the image currently being observed, but also when measuring the periphery of the charged part, the irradiation point shifts due to the bending of the primary electron trajectory, and the brightness decreases due to the surface return of secondary electrons. It is desirable that charging is not generated as much as possible because it may have an influence. For this reason, when a simulation is performed under a plurality of conditions, it is desirable to select a condition with less charge.

  Next, in step 14 of FIG. 2, a measurement recipe corresponding to the measurement location is created. At this time, the image corresponding to the created measurement recipe is also stored in the memory or the hard disk (steps 15 and 16). The above-described processing is performed for all measurement points.

  Next, measurement is performed using the created measurement recipe. Then, the image data actually obtained using the SEM (steps 17 and 18) is compared with the image formed based on the above simulation (step 19). Here, defect extraction is performed by comparing two images. Step 19 will be described in more detail with reference to FIG. In order to extract the defect, the following two methods are conceivable for comparing the SEM image and the image obtained by the simulation.

  The first method is to compare the simulation image with the measurement image, and the second method is to compare with the design data after removing the influence of charging from the measurement image. The first method is a comparison between images including the influence of charging. Since it is possible to match the images, it is possible to detect defects without charge distortion, but the original shape may not be clearly understood from the images.

  On the other hand, in the second method, before the comparison with the measurement image, the simulation result and the design data are compared, and a mismatched portion (for example, shading, tailing, line disappearance, etc.) that appears due to charging is extracted. . Since the mismatched portion extracted here appears on the image due to charging, the measurement image is corrected (the mismatched portion is removed) based on the mismatch information. As a result, the measurement image and the design data can be compared, and the original shape can also be determined from the image.

  For example, the upper left diagram of FIG. 7 is an image of a sample in which holes (holes) are formed in a resist which is an insulator. A dark shadow (shading) can be confirmed around the original hole. In the contour extraction algorithm, since the contour is extracted from a change in luminance contrast in the image, the surrounding shading may be recognized as a pattern.

  In the above-described first method (comparison with simulation results), shading around the hole can be confirmed in both images, and therefore, coincidence can be obtained as an image including the influence of charging. Further, in the second method (removing the influence of charging from the measurement image), it is found that shading due to charging appears around the hole by comparing the simulation result with the design data. By removing the inconsistent portion from the measurement image (dotted line portion in the lower left diagram in FIG. 7), the design data can be compared with the measurement image, and it can be confirmed that no defect has occurred.

  In addition, the upper right diagram in FIG. 7 is an image when a line formed in the horizontal direction is scanned in parallel to the line. It has been found that when scanning is performed parallel to the line, it is difficult to detect the edge (corner) of the line due to the influence of charging. As can be seen from the upper right diagram in FIG. 7, the edge of the line cannot be clearly read from the obtained image, and it is difficult to extract the contour of the pattern. In such a case, it is determined that the pattern is not formed and is classified as a defect. On the other hand, in the present invention, it is understood from the prior simulation that it is difficult to detect the horizontal line, and therefore it is possible to make settings such as changing the contour extraction parameter of this image.

  However, if the line is completely invisible, it is difficult to determine whether the pattern is formed or whether the edge disappears due to charging, but it is more charged by performing multiple simulations in advance. It is possible to obtain a condition in which the occurrence of (a horizontal line edge is more detectable) is less likely to occur.

  The above comparison method can be selected by the user, and in either case, erroneous determination derived from charging in defect detection can be reduced. In addition, conventional values are used as defect determination criteria such as matching tolerance.

  Whether the pattern is normal or defective is determined from the relationship between the coincidence rate obtained through the above processing and the threshold value designated in advance by the measurer. According to this embodiment, it is possible to detect a defect in consideration of the influence of charging. Further, in the present embodiment, the defect detection has been described, but it may be used for the length measurement SEM such as managing the width of the line or hole.

  Next, the sample surface is charged with an electron source (e.g., a flood gun) other than the electron beam or the electron source that emits the electron beam before the inspection and measurement (precharge), or after the inspection and measurement. A method of applying the above simulation will be described as a charging condition determination method of a method of charging a sample (post-charge).

  Depending on the surface shape of the sample to be observed, the sample surface may be positively charged in advance in preparation for SEM measurement. Such pre-charging conditions can also be determined by the present invention. The shape, material, magnification, etc. of the surface to be observed can be obtained from the simulation results. If the required surface charge value is used as an input parameter, the magnification, current amount, incident energy, etc. for obtaining the value can be obtained by simulation.

  As another method, when the measurement target pattern is a deep hole such as a contact hole, in order to selectively measure the bottom of the deep hole, it is necessary to detect the electrons emitted from the bottom with high efficiency. Therefore, it is preferable to calculate the amount of electrons emitted from the bottom and to select the device conditions (magnification, beam current, beam arrival energy, etc.) at which the value is a predetermined value or more from a plurality of device conditions. Appropriate preliminary charging conditions can be found by evaluating the magnitude of surface charging and the amount of electrons from the bottom of the deep hole.

  Furthermore, it is possible to obtain an image when the surface is scanned in a charged state. In addition, since the potential after observation can be confirmed by simulation, it is possible to perform processing for suppressing charging at the measurement location. For example, when the surface is positively charged, the electron irradiation may be performed with an energy such that the yield of electrons of the charged material is 1 or less. At this time, it is desirable to use Ep1 or less of the primary electron energy and yield distribution shown in FIG. Even at Ep2 or higher, the yield is 1 or less, so that positive charging can be suppressed. However, when irradiated with high energy, not only the surface but also the inside of the sample is negatively charged. For this reason, surface charge is relieved by irradiating with low energy electrons. These irradiation conditions can also be obtained by simulation from the surface potential immediately after observation.

  In the above-described embodiment, the method for calculating the luminance information of each pixel by simulation has been described. However, the luminance information is obtained in advance for each combination of the apparatus conditions, the sample conditions, etc. Depending on the combination of the sample conditions and the like, the luminance information may be read from the database, and the information may be arranged to form an image.

1 Cathode 2 First anode 3 Second anode 4 Primary electron beam (electron beam)
5, 6 Converging lens 7 Objective lens 8 Diaphragm plate 9 Scanning coil 10 Sample 11 Orthogonal electromagnetic field generator (deflector for secondary electrons)
12 Secondary signal (secondary electrons, backscattered electrons)
13 Secondary signal detector 14 Signal amplifier 15 Stage 16 Deflection coil 26 Image display device 30 Control processor 33 Data management device

Claims (10)

A storage medium for storing design data of a semiconductor device, using the design data stored in the storage medium, the measurement of the specimen, or an image forming apparatus having a processor for executing inspection,
The processing device acquires information on a region including a range irradiated with a charged particle beam from design data stored in the storage medium, and based on the information, the irradiation range of the charged particle beam in units of pixels. ddv / dt = q (E + v × B), dx / dt = v (where m is the mass of the electron, v is the velocity of the electron, q is the charge, E is the electric field, and B is the magnetic field). Obtaining the amount of electrons detected in the pixel portion when the sample is irradiated with a charged particle beam based on the trajectory of the electrons, and arranging luminance information obtained based on the amount of electrons, An image forming apparatus for forming an image. In claim 1,
The image forming apparatus, wherein the processing device obtains the amount of electrons based on a potential distribution on the sample. In claim 1,
The image forming apparatus, wherein the processing device obtains the amount of the electrons based on a trajectory calculation of electrons emitted from the sample. In an image forming apparatus comprising a storage medium for storing design data of a semiconductor device, and a processing apparatus for performing measurement or inspection of a sample using the design data stored in the storage medium.
The processing apparatus acquires information on a region including a charged particle beam irradiation range from the design data stored in the storage medium, and based on the information, the charged particle beam irradiation range is a pixel unit and a sample. The amount of electrons detected in the pixel portion when the charged particle beam is irradiated to the image, and luminance information obtained based on the amount of electrons is arranged , and an image is obtained for each device condition of the charged particle beam device. form shape, an image forming apparatus and the plurality of images, and comparing the design data, the apparatus condition matching rate is maximum, and selecting as a device condition of the charged particle beam device. In claim 1,
The image processing apparatus, wherein the processing device performs defect extraction by comparing the formed image with an image obtained from a charged particle beam device. In claim 5,
In the image comparison, an image in which an edge portion of the image is contoured is used. In claim 1,
The processing device extracts a mismatch point between an image obtained from a charged particle beam device and the formed image, and compares the design data with an image obtained from the charged particle beam device in which the mismatch point is corrected. An image forming apparatus characterized in that defect extraction is performed. In claim 7,
In the image comparison, an image in which an edge portion of the image is contoured is used. In claim 1,
The processing apparatus calculates the time change of the charge accumulated in the sample by irradiation of charged particles by ∂n / ∂t + ▽ j = 0, j = μnE-D ▽ n (n is the charge density, j is the current density, μ is The charge mobility, D is the charge diffusion coefficient, E is the electric field), the electric field E ′ is derived again based on the obtained charge distribution, and the amount of the electrons is determined from the electron trajectory based thereon. An image forming apparatus . In an image forming method for performing measurement or inspection of a sample using design data of a semiconductor device,
  From the design data, information on a region including a range in which the charged particle beam is irradiated is acquired, and based on the information, mdv / dt = q (E + v × B), dx / dt = v (m is the mass of the electron, v is the velocity of the electron, q is the electric charge, E is the electric field, and B is the magnetic field) to determine the electron trajectory dx / dt, and based on the electron trajectory, Image formation characterized in that an amount of electrons detected at the pixel portion when a charged particle beam is irradiated onto a sample is obtained, and luminance information obtained based on the amount of electrons is arranged to form an image. Method.

Images