JP2005061998A - Surface potential measuring method and sample observation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method of measuring the surface potential of a sample in a short time and acquiring a sample image with a reduction in the influence of surface potential. <P>SOLUTION: Secondary particles generated from the sample by irradiation with a charged particle beam are detected to acquire a sample image of a measuring area. The peak position of a luminosity histogram of pixels of the sample image is determined. The peak position is compared to the peak position of a reference luminosity histogram for a sample image having a known surface potential. From the amount of shift in the peak position, the difference between the surface potential of the measuring area and the known surface potential is determined. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a method for measuring a surface potential of a sample loaded in a charged particle beam apparatus and an observation method.

  Scanning electron microscopes (SEM) and focused ion beam devices (FIB) are known as charged particle beam devices that perform observation, measurement, and processing of samples using charged particle beams such as electron beams and ion beams. For example, SEM narrows the accelerated electron beam with a focusing lens, raster scans the sample surface, and detects secondary electrons generated from the sample in synchronization with the electron beam scan, thereby reflecting the sample surface shape. Get a statue. The secondary electron generation rate δ when the sample is irradiated with an electron beam varies depending on the type and surface state of the sample and the energy of the irradiated electron beam. Δ = 1 when the amount of irradiated electrons is equal to the amount of secondary electrons generated, δ> 1 when more secondary electrons are generated than the amount of irradiated electrons, and vice versa. When δ> 1 or δ <1, if the sample has a high electrical resistivity such as a semiconductor or an insulator, the surface of the sample is not irradiated with an electron beam and is relative to the potential at which the sample is originally held. May be positive or negative.

  When the potential of the sample surface changes as a result of electron beam irradiation in this way, the originally set electron beam control conditions, such as the energy of the electrons incident on the sample, the raster scanning range, that is, the magnification and focus conditions, become inappropriate. In order to solve this, it is necessary to measure the surface potential of the sample in real time and feed it back to the electron beam control conditions to reset the magnification control and correct the focus.

As a method for measuring the surface potential, there is a method in which a minute probe is directly brought into contact with a measurement location to read the potential. A non-contact type surface electrometer using the Kelvin method is commercially available. In addition, the energy of secondary electrons reflecting the potential information of the wiring of the semiconductor integrated circuit to which the voltage is applied is measured using an energy filter with an electron beam as a probe. There is an EB tester method for analyzing the operation status and failure analysis. Further, an energy filter that can separate the energy of secondary electrons is incorporated into the SEM, and can be used in the same manner as the EB tester. As described in WO01 / 75929, the profile of the amount of secondary electron signals with respect to the energy of secondary electrons emitted from the sample is detected by simultaneously sweeping the potential of the energy filter in the vicinity of the sample surface potential. Is obtained, and compared with a similar profile for a sample set to a known potential obtained in advance, a characteristic amount capable of specifying the sample surface potential, for example, a relative change in the position of the profile peak or gradient is obtained. Thus, the surface potential of the sample can be calculated.
WO01 / 75929

  The area (observation area) on the sample irradiated with an electron beam in SEM is at most several hundred micrometers square at most. Therefore, when using a probe with a very large size for the area where the surface potential is to be measured, such as a surface electrometer, the electron beam is irradiated due to problems with the probe size, measurement sensitivity, and measurement position. It is impossible to measure the surface potential of the area without affecting the electron beam.

  In addition, when the energy separation of the secondary electrons is performed via the energy filter, the detection method is changed to the differential type using the energy dispersion type energy filter, even if the detection method is the integration type using the potential blocking type energy filter. Even in order to optimize the conditions for detecting the secondary electron signal with sufficient S / N and to obtain a profile of the amount of secondary electron signal with respect to the energy of the secondary electron, the electrode of the energy filter It is necessary to sweep the potential or the coil current for generating the magnetic field, and it takes time for the sweep. Furthermore, if the means for detecting the amount of secondary electron signals uses an image constructed by detecting secondary electrons generated from the sample in synchronism with the raster scanning of the electron beam irradiated on the sample, the image is acquired. Time is taken into account, and the acquisition of a profile requires a time that cannot be ignored so as to reduce the throughput of the apparatus.

  An object of this invention is to provide the method which can measure the surface potential of a sample in a short time. Another object of the present invention is to provide a method for obtaining a sample image with reduced influence of surface potential.

  In the present invention, the peak position of the luminance histogram representing the relationship between the luminance gradation and the appearance frequency of the pixels constituting the sample image is used as the feature quantity for specifying the surface potential of the sample. When a voltage is applied to the sample through a holder that holds the sample, secondary particles emitted from the sample by charged particle beam irradiation are detected through an energy filter, and a sample image is formed by the detection signal. To do. As the energy filter, a potential-blocking high-pass filter composed of several electrodes can be used, and the same voltage as that applied to the sample is applied to the electrode that causes the energy filtering action. When no voltage is applied to the sample, the energy filter is not always necessary.

  If the potential of the point on the sample irradiated with the charged particle beam differs from the potential applied to the sample due to charging, the amount of secondary particles detected by the secondary particle detector after energy filtering by the energy filter Is different from the case where the point on the sample is at the same potential as the potential applied to the sample. Furthermore, if the luminance histogram is calculated from the image obtained when the potential on the sample changes, the yield of the secondary particles at the secondary particle detector changes, so the luminance of the image changes and the peak of the luminance histogram The position shifts as the potential on the sample changes. Therefore, if the relationship between the shift amount and the potential on the sample is obtained in advance, the potential on the sample can be obtained from the shift amount of the peak position.

  In the present invention, for calculating the potential of the region on the sample to be measured, as described later, a reference luminance histogram obtained for a conductive sample having a known potential, or a region having a known potential on the sample. A surface potential measured using a luminance histogram and an energy filter is used. The surface potential measured using the energy filter is, for example, a local surface potential, and the secondary electron is blocked with the set voltage of the energy filter as (energy of secondary electrons generated from the sample) + (sample surface potential). Measure the voltage. The reference luminance histogram does not have a surface structure such as wiring like a semiconductor device in order to have versatility for all surface potential measurement objects, in other words, contrast due to the influence of the surface structure and multiple materials. It is preferable to use a sample that is not generated. In this case, the reference luminance histogram has only one peak.

  Note that the inspection SEM used in the manufacturing process of a semiconductor device needs to measure dimensions such as sub-micrometer width wiring, and the observation object is a combination of a plurality of materials, for example, semiconductors, insulators, metals and organic substances. In many cases, they exist within the same field of view, and the contrast difference of the image also increases due to the difference in the generation rate of secondary electrons. In addition, since the normal observation magnification is tens of thousands of times or more, the luminance histogram of the image may have a plurality of peaks at these observation objects and observation magnifications, but the peak of the luminance histogram at the surface potential measurement point is It is desirable to have one. In order to have one peak in the luminance histogram, the magnification of image acquisition at the surface potential measurement point should be sufficiently high so that only a single material structure fits within the field of view, or conversely, a sufficiently low magnification, For example, image acquisition may be performed at about 1 to 2000 times.

  The surface potential measurement method according to the present invention scans and irradiates a measurement region on a sample with a charged particle beam, detects secondary particles generated from the sample by irradiation of the charged particle beam, and acquires a sample image of the measurement region. A step of calculating a luminance histogram of the pixel of the sample image, obtaining a peak position thereof, and obtaining a surface potential of the measurement region based on the information of the peak position.

  More specifically, a step of obtaining a shift amount of the peak position from the peak position of the reference luminance histogram with respect to a sample image having a known surface potential, and a difference between the surface potential of the measurement region and the known surface potential from the shift amount. A step for obtaining. In order to use the relationship between the shift amount of the peak of the luminance histogram and the potential on the sample, secondary particle signals such as the luminance and contrast of the secondary particle detector, which are adjustable parameters, can be used when acquiring the sample image. Amplification, the setting condition of the luminance histogram of the image, and the amount of current of the primary charged particle beam applied to the sample must be the same as when the reference luminance histogram is acquired.

  Further, the surface potential difference between the two regions on the sample can be obtained. In that case, the peak position of the first luminance histogram acquired from the sample image of the first measurement region on the sample and the peak position of the second luminance histogram acquired from the sample image of the second measurement region are obtained. Then, the difference between the surface potentials of the first and second measurement regions is obtained from the shift amount of the two peak positions.

  The sample observation method according to the present invention includes a step of moving a scanning electron microscope field of view to a measurement region on a sample, a step of obtaining a surface potential of the measurement region, a magnification of the scanning electron microscope being set to a high magnification, and a measurement in the measurement region. A step of acquiring a scanning electron microscope image of the long portion and a step of measuring the length measurement portion based on the acquired high-magnification scanning electron microscope image, and the surface potential of the measurement region is the surface potential measuring method described above When the magnification of the scanning electron microscope is set to a high magnification, the magnification and focus conditions are corrected using information on the obtained surface potential. According to this method, it is possible to reduce the magnification error caused by the charging of the sample surface and perform accurate length measurement.

  In the sample observation method according to the present invention, a semiconductor sample including an electric circuit element is irradiated with an electron beam, and a secondary electron generated from the sample by the electron beam irradiation is detected by a secondary electron detector to acquire a sample image. A step of obtaining a surface potential of a predetermined region in the observation region, and a step of setting the obtained surface potential as a voltage of an energy filter disposed between the sample and the secondary electron detector. Is obtained by the surface potential measurement method described above. According to this method, the contrast of the image reflecting the normal part and the abnormal part of the semiconductor circuit based on the charging of the sample surface can be selectively enhanced.

  According to the present invention, the surface potential measurement time in the charged particle beam irradiation region of the sample can be greatly shortened. In addition, the defect detection rate of the defect inspection apparatus can be improved.

  Hereinafter, an example in which an electron beam is used as the primary charged particle beam with which the sample is irradiated will be described. However, it goes without saying that the present invention can be similarly applied to a charged particle beam apparatus that acquires a sample image by irradiating a sample with an ion beam instead of an electron beam.

  FIG. 1 is a schematic diagram showing a configuration example of a scanning electron microscope according to the present invention. By applying a positive high voltage to the extraction electrode 2, the electron beam 3 is extracted from the electron source 1. The extracted electron beam 3 is focused on the sample 8 by the focusing lens system 5 and the objective lens 7 using a magnetic field, and an arbitrary region on the sample 8 is scanned by the deflector 6. A negative retarding voltage for decelerating the primary electron beam 3 is applied to the sample 8 from the sample voltage application unit 15. However, the application of the retarding voltage is not essential for the present invention. Secondary electrons generated from the sample 8 by irradiation with the primary electron beam 3 are accelerated in the direction of the electron source 1, deflected by the deflector 6, transmitted through the energy filter 13 composed of mesh-like electrodes, and then ground potential. It collides with the reflector 9 located in the above, is converted into low-speed secondary electrons 4 and is taken into the secondary electron detector 11.

  A Wien filter 10 is installed between the energy filter 13 and the reflecting plate 9, and the secondary electrons 4 generated by the reflecting plate 9 are removed from the secondary electron detector 11 without affecting the trajectory of the primary electron beam 3. Leads efficiently. An analog signal output from the secondary electron detector 11 is input to the image display unit 12, and a sample image is displayed on the image display unit 12. The image display unit 12 has a built-in computer, can convert an analog signal into a digital signal and perform image processing, and can calculate a luminance histogram. The image displayed on the image display unit 12 or the image processed image is stored in a storage device built in the image display unit or an external storage device.

  The secondary electron detector 11 converts secondary electrons into light with a scintillator, amplifies the light with a photomultiplier tube and an electric circuit-like amplifier, and outputs the gain. The contrast and brightness of the image can be changed by adjusting the bias voltage of the photomultiplier tube and the offset of the amplifier, and the relationship between the gain of the photomultiplier tube and the offset of the amplifier and the magnitude of the amplified signal should be linear. It has been adjusted. As long as the detection system has such characteristics, a detection system that does not use a scintillator or a photomultiplier tube, for example, a semiconductor detector can be used. In addition to the position above the energy filter 13, the secondary electron detector may be installed between the energy filter and the objective lens 7 or below the objective lens, and a plurality of detectors may be used in combination. .

  The energy filter 13 is a potential blocking type integral filter, which is generated from a sample by a energy filter voltage application unit 16 on a mesh-like thin metal plate provided around a pipe through which the primary electron beam 3 passes. By applying a voltage that can block secondary electrons accelerating in the direction of the source 1, energy separation of the secondary electrons is possible. The upper and lower electrodes to which the blocking potential is applied are sandwiched between the ground potential electrodes so that the blocking potential does not leak into the trajectory of the primary electron beam 3.

Next, the method for measuring the surface potential according to the present invention will be described. Here, a case where a bare silicon wafer is used as a sample will be described.
FIG. 2 is a diagram showing a luminance histogram of a sample image obtained by changing the potential of only the sample 8 with the applied voltage of the energy filter 13 being the same as the applied voltage of the sample voltage applying unit 15. The potential of the energy filter 13 was set to -1200V. The sample potential for each luminance histogram is -1165V, -1170V, -1175V, -1180V, -1185V, -1190V from the left of the graph, and the charge on the sample corresponds to + 35V, + 30V, + 25V, + 20V, + 15V, + 10V. .

As can be seen from FIG. 2, the relative change amount of the luminance histogram at each sample potential is almost the same, and the change amount can be calculated from the peak shift amount 17 of the luminance histogram. In this embodiment, the change in the sample potential is 5V. On the other hand, a peak shift of a luminance histogram of 17 gradations was observed. In this embodiment, the luminance per pixel is 256 gradations (8 bits), and when the gradation is set to 12 bits or 16 bits, the voltage weight per gradation is Instead, in this embodiment, the shift is 272 gradations and 4352 gradations for 5 V, respectively. Thus, if the peak position of the luminance histogram (reference luminance histogram) of a sample image using a conductor with a known potential in advance as a sample is stored in the image display unit 12, the reference luminance histogram can be obtained for an arbitrary sample. The peak shift amount 17 of the luminance histogram can be calculated. In the case of this embodiment, assuming that the sample potential corresponding to the reference luminance histogram is R [V], and the shift amount from the reference luminance histogram peak of the luminance histogram peak of the sample image of the sample whose surface potential is to be obtained is A [gradation]. Can calculate the surface potential S [V] of the electron beam irradiation region on the sample by the following equation (1).
S [V] = R [V] + A [gradation] × 5 [V] / 17 [gradation] (1)

  The part of “A [gradation] × 5 [V] / 17 [gradation]” in the above equation (1) can be considered as a device constant. Further, if the reference luminance histogram is calibrated for various samples instead of setting the peak shift amount with respect to the surface potential change of the luminance histogram as a fixed value, more accurate potential measurement can be performed.

  Furthermore, even if a reference luminance histogram measured in advance using a reference sample is not used, a luminance histogram that replaces the reference luminance histogram in a sample for which surface potential measurement is performed before the surface potential measurement is started under the condition that the peak is one. The surface potential can be determined by acquiring the voltage of the energy filter that is obtained and at the same time the yield of secondary electrons that have been energy-filtered using the energy filter is a predetermined threshold value.

  This can be explained as follows. When the potential of the energy filter 13 is constant, the yield of secondary electrons that are transmitted through the energy filter and detected by the secondary electron detector 11 when the potential of the sample 8 changes is the secondary electrons shown in FIG. The yield curve 22 changes. On the contrary, the curve can be obtained even when the voltage of the energy filter is changed by the energy filter voltage application unit 16 while the potential of the sample is kept constant. The horizontal axis of FIG. 3 is the relative energy of the secondary electrons with respect to the blocking potential applied to the energy filter 13, and 0 is a condition where the blocking potential and the energy of the secondary electrons match. Due to the energy resolution of the energy filter and the energy distribution of the secondary electrons, it is not completely cut off at zero. Therefore, the threshold voltage, that is, the secondary electron yield, for example, the secondary electron yield is determined so that the error determined by the resolution of the energy filter and the energy distribution of the secondary electrons is within an allowable value. If the voltage is determined to be 50% of the maximum yield, the sample surface potential can be estimated from the curve.

  By using the luminance histogram at this time as a reference luminance histogram, the peak shift amount at an arbitrary point can be measured, and the surface potential at an arbitrary point can also be measured.

  In addition to this, when repeatedly measuring a pattern with the same surface potential measurement point, if the measurement point of the luminance histogram that is used instead of the reference luminance histogram is selected to be the same pattern as the actual measurement point, the secondary electrons of the material There is no change in the yield of secondary electrons due to the difference in the generation rate, making it difficult to be affected by the difference in material. Therefore, the luminance histogram of the measurement point and the luminance histogram profile used instead of the reference luminance histogram are similar, and the correspondence between the peak shift amount and the sample surface potential is more accurate. What is necessary is just to measure the surface potential of the measurement point of the luminance histogram which is used instead of the reference luminance histogram by the above method.

  Furthermore, even if a luminance histogram is acquired under conditions that cause a plurality of peaks in the luminance histogram, it is clear that the surface potential can be calculated from the peak shift amount of an arbitrary peak using the above method.

  On the other hand, the image acquisition for acquiring the brightness histogram for measuring the surface potential of the sample is the edge enhancement and gamma correction that determine the brightness histogram distribution of the image according to the gain and offset of the secondary electron detection system and these values. When processing such as, the setting of those image processing functions and the setting of the acceleration voltage and irradiation current amount of incident electrons that affect the secondary electron generation rate were set when acquiring the reference luminance histogram It is necessary to be the same as the conditions. In this embodiment, the applied voltage that determines the amplification factor of the photomultiplier tube of the secondary electron detector 11, the gain and offset of the analog signal that is the output of the photomultiplier tube, the signal addition ratio of the plurality of secondary electron detectors, The setting of the number of images to be added and the acceleration voltage and irradiation current amount of incident electrons was the same as the reference luminance histogram acquisition condition.

  Further, the region in which the sample potential and the peak shift amount 17 of the luminance histogram can be considered to change almost linearly as shown in FIG. 2 varies depending on an allowable error, but the secondary electron yield is almost as in the region 23 shown in FIG. This is a region in which the signal output from the secondary electron detector 11 changes linearly with respect to the change in the sample potential. However, even if the peak shift amount of the luminance histogram and the signal output from the secondary electron detector are not linear, the relationship as shown in FIG. 2 is acquired in advance, and the change in the sample potential and the peak change in the luminance histogram are obtained. What is necessary is just to create an approximate expression so that the relationship of quantity can be found uniquely.

That is, the relationship of the peak shift amount A [gradation] of the luminance histogram to the change amount ΔV of the surface potential of the sample is approximately expressed by ΔV [V] = F (A) [V] using the function F (A). In this case, the surface potential S [V] of the electron beam irradiation region on the sample can be calculated using the equation (2) instead of the equation (1). R [V] is the sample potential of the reference sample from which the reference luminance histogram is acquired.
S [V] = R [V] + F (A) [V] (2)

  Further, in this embodiment, the difference from the sample potential that can be measured by comparison with the reference luminance histogram is about 30 V at the maximum from the relationship of FIG. 2, but the difference between the potential of the reference luminance histogram and the sample potential is larger than this. Since the secondary electron yield is a region that does not change as seen in FIG. 3, the peak position of the luminance histogram is at the right end or near the left end, and the peak of the luminance histogram is in a saturated state that does not follow the change in the sample potential. Become. In that case, the voltage applied to the energy filter 13 from the energy filter voltage application unit 16 is changed positively or negatively with respect to the sample potential at a predetermined voltage step, and an image is acquired to calculate a luminance histogram. The difference between the peak position and the peak of the reference luminance histogram may be within a predetermined number of gradations.

As a result, when the voltage of the set energy filter voltage application unit 16 is V ₀ [V] and the peak shift amount of the luminance histogram at the reference luminance histogram and the surface potential measurement point is A [gradation], The surface potential S [V] of the electron beam irradiation region on the sample can be calculated from the following equation (3).
S [V] = R [V] + V ₀ [V] + F (A) [V] (3)

In the case of the present embodiment, Expression (3) can be expressed as Expression (4).
S [V] = R [V] + V ₀ [V] + A [gradation] × 5 [V] / 17 [gradation] (4)

  For example, if the voltage step is 30 V and the increase in surface potential due to electron beam irradiation is 100 V, three or four image acquisitions are performed, and the throughput is reduced. However, once the conditions are known, the measurement conditions can be stored for each lot as long as the same sample, for example, the wafers of the same lot, is used as the measurement target.

  Since the reference luminance histogram is used as an object to be compared with an arbitrary sample, the potential can be known, and a conductor that has a single contrast, for example, a flat surface, is preferably obtained as a sample. In that case, the reference luminance histogram has one peak. However, in the luminance histogram obtained from an actual sample, two peaks are generated, as in the luminance histogram 18 calculated from a sample image (FIG. 4C) acquired from a patterned sample at a magnification of 10,000 times, for example. There is also a case. In such cases, the cause of the appearance of multiple peaks in the luminance histogram is due to the potential on the sample, the difference in the secondary electron generation rate depending on the material, or the edge of the structure to be observed, which is called the edge effect. It is difficult to judge whether this is due to an increase in secondary electron content. Therefore, in order to compare the relative shift of the peak position of the luminance histogram at the actual surface potential measurement point with only one peak of the reference luminance histogram, the peak of the luminance histogram at the surface potential measurement point is 1 Is desirable. In order to have one peak in the luminance histogram, the image acquisition magnification at the surface potential measurement point is sufficiently high so that only the structure of a single material is within the field of view, or a sufficiently low magnification such as 1 What is necessary is just to carry out by about 2,000 times. For example, as shown in FIG. 4B, the peak of the luminance histogram 19 calculated from a sample image acquired from the same sample at a low magnification (1000 times) is one.

  If an image is acquired at a low magnification, the area on the sample per pixel increases, the contrast due to the surface structure is averaged, the influence of a specific structure or material is eliminated, and the luminance histogram has only one peak. . Further, as a merit of setting a low magnification, the surface of the sample is uniformly excited and stabilized by irradiating a wide area on the sample with an electron beam at a low magnification. Therefore, after obtaining an image after electron beam irradiation at a low magnification, a luminance histogram is calculated, and the surface potential on the sample can be quickly determined from the relationship between the relative shift amount of the peak position from the reference luminance histogram and the surface potential. After that, the charge excited at a low magnification is maintained even at a high magnification, and there is almost no change in the surface potential even when the electron beam irradiation is performed at a high magnification. It is possible to measure dimensions more accurately by correcting the focus and focus.

FIG. 5 is a flowchart showing the procedure for measuring the surface potential of a sample and observing the sample according to the present invention.
First, the field of view is moved to the measurement point by moving the sample stage or the like that holds the sample (S11). Next, the magnification is set to a low magnification of about 1 to 2,000 times, the sample surface is scanned, and the sample surface is charged (S12). The sample image of the measurement point is acquired with the magnification as it is or, if necessary, at a high magnification (S13). Thereafter, a luminance histogram of pixels is created from the acquired image (S14), and a peak shift amount is obtained by comparison with the reference luminance histogram (S15). It is determined whether the luminance histogram is saturated based on whether the peak position of the luminance histogram is close to the right end or the left end (S16). If not, the calculated peak shift amount is applied to the above equation (2). Then, the surface potential of the sample is calculated (S17). If the luminance histogram is saturated, the voltage of the energy filter is changed at a predetermined voltage step to cancel the saturation of the luminance histogram (S18), and the peak shift amount calculated from the luminance histogram without saturation is expressed by the above equation (3 ) To calculate the surface potential of the sample (S19). In this way, the surface potential at the measurement point of the sample is measured.

  Next, the SEM magnification and focus are corrected using the measured surface potential information (S20). Specifically, the trajectory of the primary electron beam determined from the setting conditions of the parameters of the electron optical system such as the energy of the primary electron beam 3, the surface potential of the sample, the objective lens current, and the distance between the objective lens and the sample, that is, the magnification or The focus current is obtained in advance by computer simulation or the like, and the magnification and the focus current are calculated by an approximate expression having sufficient accuracy for an arbitrary parameter setting value.

  When contrast enhancement is necessary, the voltage of the surface potential measured is set as the voltage of the energy filter 13 (S22). Then, an image with electrons having energy higher than the potential of the energy filter is obtained, and a sample image reflecting the potential of the sample surface is obtained. Thus, the magnification is set with high accuracy, and the length and observation are performed on the sample image in which the focus is corrected and an appropriate focus state is obtained (S23).

  According to the present invention, the surface potential of the electron beam irradiation region can be measured quickly and easily. Further, based on the measured surface potential, taking into account changes in the energy and orbit of the primary electron beam 3, the focus and magnification control unit 14 calculates the magnification and the optimum value of the focus, and the deflector 6 and the objective lens. 7 can be corrected. Furthermore, when the measured surface potential is fed back to the energy filter voltage application unit 16, a contrast can be given to a desired potential portion, which can be applied to a defect inspection of a semiconductor device.

  For example, as a typical semiconductor defect inspection, there is an inspection of contact hole opening or non-opening which is one of semiconductor circuit components. FIG. 6 is a schematic cross-sectional view of a contact in which a conductor is buried in a contact hole to electrically connect upper and lower layers. FIG. 6A is a schematic cross-sectional view of a normal contact, and FIG. 6B is a cross-sectional view of an abnormal contact. It is a schematic diagram. If a residue remains due to poor etching at the bottom of the contact hole, when the metal is buried in the contact hole, the upper and lower portions of the contact hole to be originally connected are electrically connected as shown in FIG. 6B. It will be defective.

  When the electron beam 26 is incident on the normal contact 24 and the abnormal contact 25 with an accelerating voltage (about 300 to 1500 [V]) that causes the secondary electron generation rate to be 1 or more using a scanning electron microscope, the surface potential due to charging is applied. Are represented by equipotential lines 27 and 28, respectively. The abnormal contact 25 whose bottom is not electrically connected to the substrate wiring 29 is positively charged. Therefore, a potential barrier is formed above the abnormal contact 25, and the secondary electrons 30 having low kinetic energy are drawn back to the sample side, and the amount of electrons detected by the secondary electron detector is reduced. In the normal contact 24, such a phenomenon does not occur with respect to the generated secondary electrons 31, and as a result, the potential contrast image becomes dark at the defective portion and becomes bright at the normal portion.

  Therefore, if the surface potential of a normal contact is measured and the potential of the energy filter 13 is set to the same level as the surface potential, only the secondary electrons generated from the normal part can pass through the energy filter, and the defective part Since secondary electrons from are cut off, it is possible to obtain an image in which the contrast between the dark part and the bright part is further enhanced. Furthermore, since the contrast between the normal part and the defective part is clear and the S / N is high, the binarization process for digitally distinguishing light and dark can be performed reliably even when performing image processing, resulting in defective parts. Discrimination and detection efficiency are improved.

The figure which shows the structural example of the scanning electron microscope by this invention. Explanatory drawing of the brightness | luminance histogram computed from the image acquired with respect to the surface potential of a different sample. Explanatory drawing of the brightness | luminance histogram calculated from the image acquired with respect to different magnifications. Explanatory drawing of the yield change of the secondary electron which permeate | transmits a high pass type energy filter. The flowchart which shows the procedure of the surface potential measurement and sample observation of the sample by this invention. Explanatory drawing about the principle of potential contrast and defect inspection.

Explanation of symbols

1: electron source 2: extraction electrode 3: primary electron beam 4: secondary electron 5: focusing lens system 6: deflector 7: objective lens 8: sample 9: reflector 10: Wien filter 11: secondary electron detector 12 : Image display unit 13: energy filter 14: focus and magnification control unit 15: sample voltage application unit 16: energy filter voltage application unit 17: peak shift amount 24 of luminance histogram: normal contact hole 25: abnormal contact hole 26: Electron beam 27 irradiated to the contact hole 27: equipotential line 28: equipotential line 29: conductor wiring 30 connected to the contact hole 30: orbit of secondary electrons 31: orbit of secondary electrons

Claims (10)

Scanning and irradiating a measurement region on the sample with a charged particle beam, detecting secondary particles generated from the sample by irradiation of the charged particle beam, and obtaining a sample image of the measurement region;
Calculating a luminance histogram of the pixels of the sample image and determining its peak position;
And obtaining a surface potential of the measurement region based on the information on the peak position. In the surface potential measuring method according to claim 1,
Obtaining a shift amount of the peak position from the peak position of the reference luminance histogram for a sample image having a known surface potential;
And obtaining a difference between the surface potential of the measurement region and the known surface potential from the shift amount.   3. The surface potential measurement method according to claim 2, wherein the image acquisition condition when acquiring the luminance histogram of the measurement region is the same as the image acquisition condition when acquiring the reference luminance histogram. Measuring method. 4. The surface potential measurement method according to claim 2, wherein the relationship between the surface potential R [V] corresponding to the reference luminance histogram and the peak shift amount A [gradation] of the luminance histogram with respect to the surface potential variation ΔV. ] = F (A) [V] and the surface potential S [V] of the measurement region is obtained by the following equation.
S [V] = R [V] + F (A) [V]   5. The surface potential measurement method according to claim 1, wherein a negative potential is applied to the sample, and secondary particles generated from the sample are detected through an energy filter set to the same potential as the negative potential. A method for measuring a surface potential. 4. The surface potential measurement method according to claim 2, wherein secondary particles generated from the sample are detected through an energy filter to which a voltage V ₀ [V] is applied to the sample, and the surface corresponding to the reference luminance histogram. Using the relationship ΔV [V] = F (A) [V] of the peak shift amount A [gradation] of the luminance histogram with respect to the potential R [V] and the change amount ΔV of the surface potential, the surface potential S of the measurement region is expressed by A method for measuring a surface potential, wherein [V] is obtained.
S [V] = R [V] + V ₀ [V] + F (A) [V]   The surface potential measurement method according to claim 1, wherein a wide region that does not cause a plurality of peaks in the luminance histogram is set as the measurement region. In the surface potential measuring method according to claim 1,
Obtaining a peak position of a first luminance histogram acquired from a sample image of a first measurement region on the sample;
Obtaining a peak position of a second luminance histogram acquired from a sample image of a second measurement region different from the first measurement region on the sample;
A surface potential measurement method, wherein a difference in surface potential between the first and second measurement regions is obtained from a shift amount of the two peak positions. Moving the scanning electron microscope field of view to the measurement area on the sample;
Obtaining a surface potential of the measurement region;
A step of obtaining a scanning electron microscope image of a length measurement portion in the measurement region, with a high magnification of the scanning electron microscope,
Measuring the length measurement portion based on the acquired high-magnification scanning electron microscope image,
The surface potential of the measurement region is obtained by the surface potential measurement method according to any one of claims 1 to 8, and when the magnification of the scanning electron microscope is set to a high magnification, information on the obtained surface potential is used. A sample observation method, wherein the magnification and focus conditions are corrected. Irradiating a semiconductor sample including an electric circuit element with an electron beam, detecting secondary electrons generated from the sample by electron beam irradiation with a secondary electron detector, and obtaining a sample image;
Obtaining a surface potential of a predetermined region in the observation region;
Setting the determined surface potential as a voltage of an energy filter disposed between a sample and the secondary electron detector,
The sample observation method characterized by calculating | requiring the said surface potential with the surface potential measuring method of any one of Claims 1-8.

Images