JP2006119133A - Semiconductor device tester - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To further improve technology for detecting substrate current generated by irradiation of an electron beam to nondestructively inspect an internal state of a semiconductor device or a detailed shape of a contact hole. <P>SOLUTION: The sample 5 is irradiated with parallel electron beam 2, and the current flowing through a sample 5 is measured by an ammeter 9. Acceleration voltage of electron beam 2 is changed, and the measurement is repeated. Information related to a depth-direction structure of the sample 5 is found from difference of a current value based on difference of a transmittance of the electron beam 2 to the sample 5 by difference of the acceleration voltage, in a data processor 10. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an inspection of a semiconductor device using an electron beam, and more particularly to an inspection of a semiconductor device by measuring a current generated in a sample to be inspected with irradiation of an electron beam.

  In general, a semiconductor device such as a memory has a contact hole or a via hole connecting a lower active element and an upper wiring layer. The contact hole is a through hole that reaches from the surface to the base substrate by digging an insulating film such as an oxide film by reactive ion etching. In order to optimize the etching conditions, it is necessary to detect the appearance shape of the contact hole, the shape inside the contact hole, or the state of the bottom of the contact hole.

  Since the diameter of the contact hole is less than a micron order, visible light cannot enter the bottom of the contact hole, and it is difficult to detect the quality by an optical method. For this reason, a scanning electron microscope (SEM) suitable for analyzing a fine structure is mainly used as an inspection apparatus. In SEM, an electron beam accelerated to several tens of KeV and focused to several nanometers is applied to a contact hole region, and secondary electrons generated in the applied region are detected by a secondary electron detection device to form an image. The sample irradiated with the electron beam generates secondary electrons in an amount corresponding to the constituent atoms. In general, in the SEM, the secondary electron detector is arranged in a specific direction, and all the secondary electrons are detected. is not. For this reason, if the sample has irregularities, even if the same material is used, there are cases where secondary electrons are detected and cases where the secondary electrons are not detected, resulting in a contrast. This is the reason why the inspection object made of the same substance produces contrast, and is a feature of SEM.

  On the other hand, in a contact hole or a through hole, since an electrical contact is taken at the bottom of the contact hole, not only the shape of the contact hole opening but also the shape of the contact hole bottom and its surface state are very important. Etching that forms contact holes with an aspect ratio exceeding 10 due to the recent increase in density and multilayering may result in different internal diameters depending on the process conditions, even if they have the same opening diameter. Such variation in the internal dimensions of the contact hole has a great influence on the characteristics of the device. Therefore, the person in charge of the process needs to control the process so that the entire contact hole has the same dimension. In addition, since these variations should not occur in actual products, it is necessary to make the products subject to inspection, and a technology that can detect both of them non-destructively is very important.

  4 and 5 show examples using the SEM, where (a) shows an example of the inspection method and (b) shows an example of the inspection result, respectively. FIG. 4 shows an example of inspection of a cylindrical contact hole, and FIG. 5 shows an example of inspection of a tapered contact hole. In the inspection by the SEM, the sample to be inspected is irradiated while scanning with the electron beam 31, and the secondary electrons 32 generated from the sample to be inspected are detected by the secondary electron detector 33.

  As shown in FIG. 4A, the structure of the sample to be inspected is provided with an insulating film 41 such as an oxide film on a base substrate 42, and the insulating film 41 is etched perpendicularly from the opening to open the opening. It is assumed that a cylindrical contact hole 43 having the same diameter as the portion is formed. In this case, since the secondary electrons have low energy, it is difficult to reach the detector unless the surroundings are wide open, and the measured value of the amount of secondary electrons is as shown in FIG. That is, the obtained secondary electron image becomes darker corresponding to the opening of the contact hole 46. This shows that there is a contact hole there.

  On the other hand, as shown in FIG. 5A, it is assumed that the contact hole 44 has a tapered shape in which the diameter becomes smaller toward the back of the hole and has a diameter different from the opening diameter. In this case, although secondary electrons from the tapered portion may be observed depending on the position of the detector, the contact hole 44 actually has a high aspect ratio, and the secondary electrons from the inner wall of the hole are shown in FIG. As shown in c), it is almost impossible to observe. Therefore, the shape of the contact hole 44 and the information on the bottom thereof are not reflected in the secondary electron image.

  A tapered contact hole as shown in FIG. 5 (a) is tapered toward the bottom even if the opening is good, resulting in a hole diameter different from the design target, resulting in contact resistance. It may increase and become defective. However, in the inspection by SEM, regardless of whether the shape of the contact hole is cylindrical or tapered, the obtained detection image is darkened rapidly at the opening, and the bottom information is not reflected and is the same. It becomes an image like this. For this reason, the fact is that a normal SEM cannot distinguish between the two.

  As a method for inspecting the inside or the bottom of the contact hole, a method of observing a cross section by vertically cutting the central portion of the contact hole of the sample is employed. This method requires an advanced technology that divides exactly into two at the center of the contact hole. Considering that the current contact hole diameter is on the order of several thousand mm, the center of the contact hole is necessary for determining good products. It is virtually impossible to divide with an accuracy of 10%. Moreover, this is a destructive observation, and the product cannot be seen directly, and it requires a lot of labor and time.

  As means for solving such a problem, it is disclosed that a current generated by an electron beam passing through a contact hole and reaching a substrate is detected to detect the position and size of the bottom of the contact hole (for example, patents). Reference 1). In addition, as another technique, it is disclosed that a secondary electron image is observed by irradiating an ion beam instead of an electron beam, and that a substrate current generated with the ion beam irradiation is measured ( For example, see Patent Document 2.)

  Further, as a similar technique, it is disclosed that a mask alignment misalignment measurement pattern is formed and a mask alignment misalignment amount is obtained from a substrate current generated when an electron beam is irradiated (see, for example, Patent Document 3). . Further, as another technique, it is disclosed that a region including a plurality of contact holes is irradiated with an electron beam, and a ratio of normal contact holes in the region is inspected based on a current value penetrating the holes. (For example, refer to Patent Document 4).

Furthermore, it is also possible to know the film thickness by measuring the substrate current, and the relationship between the waveform on the time axis of the substrate current when the pulsed electron beam is irradiated, the acceleration voltage of the electron beam, and the film thickness. It is disclosed that a film thickness is obtained from a current waveform that is obtained in advance and measured using an electron beam having a certain acceleration voltage (see, for example, Patent Document 5). Further, as another technique, it is disclosed to measure the current penetrating the back surface by irradiating the specimen to be inspected with an electron beam, that is, the current value itself rather than the temporal change of the current (for example, patents). Reference 6). Further, a technique for obtaining a film thickness by measuring the current passing through the thin film and reaching the substrate and comparing it with the value of the standard sample (see, for example, Patent Document 7), and a standard sample suitable for it are known. (For example, refer to Patent Document 8).
JP-A-10-281746 Japanese Patent Laid-Open No. 4-62857 JP-A-11-026343 JP 2000-174077 A Japanese Patent Laid-Open No. 62-19707 JP 2000-124276 A JP 2000-180143 A JP 2000-164715 A

The present invention further improves a technique for detecting a substrate current generated by electron beam irradiation, and provides a semiconductor device inspection apparatus capable of nondestructively inspecting a detailed shape of a contact hole and an internal state of a semiconductor device. With the goal.

  A semiconductor device inspection apparatus according to the present invention includes an electron beam irradiation unit that irradiates a semiconductor device of a sample to be inspected while scanning the electron beam, and a current measurement unit that measures a current generated in the sample to be inspected with the irradiation of the electron beam. In the semiconductor device inspection apparatus including the data processing means for processing the measurement result of the current measuring means, the electron beam irradiation means changes the collimating means for collimating the electron beam and the acceleration voltage of the electron beam. The data processing means includes a depth of the sample to be inspected from a difference in current value generated in the sample based on a difference in transmittance of the electron beam with respect to the sample to be inspected when the electron beam is scanned with different acceleration voltages. It includes means for obtaining information on the structure of the direction.

  The reason why the parallel electron beam is used in the present invention is that the focused electron beam needs to converge the electron beam to the height of the measurement location, which is not preferable for obtaining information in the depth direction of the sample to be inspected. If a parallel electron beam is used, the depth of focus becomes infinite, and it is not necessary to adjust the focus.

  Although the technique disclosed in Patent Document 1 described above can inspect whether the contact hole penetrates, detailed information such as the shape of the contact hole cannot be obtained. The technique described in Patent Document 2 using an ion beam is the same. Patent Document 6 describes that the current amount or acceleration voltage of the electron beam is changed, but this is for reducing noise and for examining the structure in the depth direction of the specimen to be inspected. is not. Japanese Patent Application Laid-Open No. H10-260260 describes the use of a parallel beam. This parallel beam is for irradiating a region including a plurality of contact holes, and the depth of each contact hole or other specimen to be inspected. It does not examine the structure in the vertical direction.

  The electron beam irradiating means includes an electron gun, and the collimating means limits a spot size of the electron beam hitting the semiconductor device to be inspected and a condenser lens for collimating the electron beam emitted from the electron gun. Means for moving the sample to be inspected with respect to the electron beam for scanning the electron beam, including an aperture inserted between the condenser lens and the sample to be inspected at a right angle so that the electron beam hits the opening. It is desirable to provide.

  The electron beam irradiating means includes an electron gun, and the collimating means is arranged to form an afocal system with a first condenser lens that collimates the electron beam emitted from the electron gun. A condenser lens, an objective lens, and an aperture inserted between the first condenser lens and the second condenser lens to limit the spot size of the electron beam, and for scanning the electron beam It is also possible to provide means for moving the sample to be inspected.

  The electron beam irradiation means includes means for vertically irradiating a sample to be inspected along a line segment passing through the center of the measurement area with an electron beam having a spot size smaller than the area of the measurement area. Means may be included for determining the bottom distance of the measurement region from the rise and fall intervals of the current measured along the line segment.

  The present invention can be used in combination with an SEM. That is, a secondary electron detector for detecting secondary electrons emitted from the surface of the processing object to be inspected is provided, and the data processing means calculates the amount of secondary electrons measured by the secondary electron detector as the current measuring means. Means for processing in accordance with the measurement results can be included. Specifically, by the electron beam irradiation means, an electron beam having a spot size smaller than the area of the measurement region is vertically irradiated onto the sample to be inspected along a line segment passing through the center of the measurement region, and the corresponding processing is performed. The bottom distance of the measurement region is obtained from the rising and falling intervals of the current measured along the line segment by the current measuring means, and the amount of secondary electrons detected by the secondary electron detector is obtained. The upper distance of the measurement region can be obtained from the rise and fall intervals.

  Means may be provided for tilting the sample stage on which the sample to be inspected is placed, and the data processing means may include means for taking in and processing the tilt angle of the sample to be inspected with respect to the electron beam generated by the tilting means.

  The electron beam irradiation means includes means for setting the cross-sectional shape of the electron beam so that the entire measurement region can be irradiated at once, and at least one end thereof is linear, and the data processing means Means for determining the distance of the measurement region from the interval from the rising edge to the maximum value can be included.

  The electron beam irradiation means includes means for setting the cross-sectional shape of the electron beam so that the entire measurement region can be irradiated at once, and at least one end thereof is linear, and the data processing means Means for calculating a differential curve with respect to the distance and means for determining the radius of the measurement region from the interval from the rise of the differential curve to the position indicating the apex may be included.

  The data processing means may include means for displaying the measured current value on a map corresponding to the measurement position.

  The data processing means, each of the two areas on the wafer as a sample to be inspected, a comparison means for comparing the measured value obtained in the first area with the measured value obtained in the second area as a reference value, Means for extracting the position coordinates when the comparison result has a predetermined difference or more.

  The comparing means is means for integrating the current flowing from the rising edge to the falling edge of one pulse of the current waveform measured as a change in the current value generated in the sample as the electron beam is irradiated to the electron beam irradiation position; A dividing means for dividing the obtained integrated value by the width from the rising edge to the falling edge of the pulse, and a means for comparing the current value per unit length obtained by the dividing means for the two regions. You can also

  The comparison means may include means for comparing the rising and falling positions of the current waveform pulse of the first inspection sample and the second inspection sample measured as a change in current value with respect to the irradiation position of the electron beam, Means for comparing the center positions may be included.

  The electron beam irradiation means includes a main scanning means for moving the specimen to be inspected with respect to the electron beam, and a sub-scanning means for repeatedly deflecting the electron beam in a direction different from the main scanning direction, superimposed on the main scanning by the main scanning means. Scanning means.

  As described above in detail, the semiconductor device inspection apparatus of the present invention can non-destructive information on the structure in the depth direction of the sample to be inspected, and the quality determination and manufacturing process of the manufactured semiconductor device are optimal. It is particularly effective when used for testing whether or not.

  In addition, by combining the information of the contact hole opening obtained from the secondary electron image, the distance information of the contact hole cross-section, which the conventional SEM has only had to divide and observe the sample, the contact hole bottom shape and In addition to obtaining information on the distance above and below the contact hole, more accurate information can be obtained in combination with information on the structure in the depth direction of the sample to be inspected obtained at different acceleration voltages.

  When a rectangular electron beam is used, the position of the edge of the electron beam can be easily specified, and the area of the region through which the electron beam has passed can be measured easily and with high accuracy. In the embodiments, only application to contact holes has been described, but for example, it can also be used for opening discrimination inspection and shape measurement of through holes, resists, wirings, and grooves having the same structure as contact holes. Similarly, it is also possible to inspect the shape and the state of the bottom after various etching processes or cleaning processes.

  Since the present invention is non-destructive, it is possible to obtain information in the depth direction of the contact hole without using SEM observation of the cross section of the sample as in the prior art, so that the product can be directly measured, a monitor wafer is not required, and the process Costs are reduced.

  In addition, since the area, diameter, or three-dimensional shape of the contact hole bottom during the process can be measured at high speed and in an analog manner, the process can be improved on the spot. For example, when the etching conditions are set, it is necessary to control the shape of the contact hole bottom as well as the shape of the opening. If this method is used, the distribution of the contact hole bottom area in the entire wafer can be measured on the spot.

  Conventionally, since the contact hole opening defect has only been digitally inspected only for the presence or absence of an opening, an abnormality can be detected only when the opening is not made. Since the diameter of the hole in the depth direction can always be monitored, an abnormality can be found as a change in analog quantity such as information on the bottom diameter of the contact hole or depth direction before the contact hole opening defect actually occurs. Therefore, countermeasures against abnormalities can be taken earlier than before. In particular, in combination with measurement of the contact hole bottom diameter by the collective current method, when an abnormality is detected by the collective current method, it is preferable to perform measurement with higher accuracy. The collective current method only needs to be able to adjust the position so that the electron beam hits one contact hole, and can therefore be measured with low alignment accuracy.

  In the case of current measurement, only the current flowing into the wiring contributes to the measurement value, so that it is not necessary to average the inspection results required in the conventional inspection method, and the inspection speed can be improved.

  When measuring and aligning the current flowing in the alignment mark, an expensive secondary electron image acquisition device is not required only for alignment.

  In the measurement of the current waveform in the present invention, information effective for inspection can be acquired regardless of the part of the wiring where the electron beam passes, and it is not always necessary to irradiate the specific position of the wiring with the electron beam. . Conversely, the defect pattern detection sensitivity can be adjusted by changing the electron beam irradiation position. In addition, because the wiring end position information obtained from the rise and fall of the current waveform generated by electron beam irradiation is also used for inspection of the wiring quality, the inspection result is clear only by changing the size of the acquired current waveform Even if it does not come out, inspection may be possible.

  In general, the inspection of wiring arranged in an array can be made faster than the inspection of wiring arranged randomly, but both are mixed in one chip. In that case, the wiring arrangement is examined in advance in the first inspection, the location where the wiring is arranged in an array is estimated from the measured frequency distribution of the current, and the optimal inspection method is based on that information. Since it can be selected, the inspection speed can be increased.

  The measurement of the current waveform in the present invention may be performed not only continuously with the electron beam but also intermittently. In addition, the effective scanning speed can be improved by performing the sub-scanning of the electron beam, and the inspection speed can be improved. The scanned electron beam does not necessarily need to be scanned at a different position, and may be scanned by slightly overlapping the end portion of the inspection region. For the acceleration voltage and the injection current, the most appropriate one is appropriately selected depending on the sample. Even when the pattern defect is partial, a current proportional to the defect area is detected. Therefore, if the current change that causes the defect area is greater than or equal to the SN of the inspection apparatus, it is detected as a defect.

  Measurement of the current waveform in the present invention is effective when the wiring to be inspected has some connection with the substrate, but when the wiring has a large basin area, or when the leakage current is large, Alternatively, inspection can be performed even when the substrate is electrically connected via a large capacity. Since a plurality of wirings can be inspected at the same time, it can be inspected at a higher speed than conventional methods. In addition, the cross-sectional structure of the device can be observed directly from the surface.

  Next, embodiments of the present invention will be described in detail with reference to the drawings. In the figure, equivalent elements are denoted by the same reference numerals, and the description thereof is omitted.

(Generation of parallel electron beam)
FIG. 1 is a block diagram showing a first embodiment of the present invention. This apparatus includes an electron gun 1 that generates an electron beam 2, a condenser lens 3 and an aperture 4 that collimate the electron beam, a movable stage 6 that scans an irradiation position of the electron beam 2 by moving a sample 5 to be inspected, An electrode 7 and an ammeter 9 for measuring a current generated in the sample 5 with irradiation of the electron beam 2, a moving distance measuring device 8 for measuring a moving distance of the movable stage 6, a computer for processing the measurement result of the ammeter 9 and the like Data processing apparatus 10 and a beam control section 11 for performing control such as changing the acceleration voltage of the electron beam, changing the irradiation period, and the like.

  The electron beam 2 that has jumped out of the electron gun 1 is once converted into a parallel electron beam by the condenser lens 3 and irradiated onto the aperture 4 provided with a very small hole. The aperture 4 is made of metal or the like, and is grounded so that electrons irradiated on the aperture 4 do not accumulate in the aperture 4. The electron beam 2 that has passed through the aperture 4 becomes a very thin beam having the same size as that defined by the aperture 4 and is irradiated onto the sample 5. In order to prevent the aperture diameter from changing due to heating, the aperture may be cooled.

  FIG. 3 shows a configuration example of the aperture, in which (a) shows an aperture for making the cross-sectional shape of the beam circular, and (b) shows an aperture for making a square. These apertures have an opening 21 at a substantially central portion, and the periphery thereof is a shielding portion 22. As the material of the shielding part 22, tungsten, molybdenum, silicon, polysilicon, gold, palladium, titanium, or the like, which is a material that hardly generates a gas when an electron beam is applied thereto, is used. The aperture diameter is several hundred to 1,000 mm when scanning the inside of the contact hole to obtain the distance, and several microns when one contact hole is irradiated with an electron beam at a time. The shape of the aperture can be not only a square or a circle but also a rectangle, an ellipse, or other polygons.

  Even if the electron beam uses a beam having a cross-sectional area wider than that of the opening of the aperture or a beam having a cross-sectional area narrower than that of the opening, no problem occurs. When the cross-sectional area of the electron beam is smaller than the area of the aperture opening of the aperture, the same result as that obtained by irradiating with a beam having a cross-sectional area wider than that of the opening can be obtained by scanning the aperture opening.

  The sample 5 is placed on the current collecting electrode 7, and this electrode 7 is placed on the movable stage 6. In the vicinity of the movable stage 6, there is provided a movement distance measuring device 8 for accurately measuring the movement distance of the movable stage 6 in angstrom order by the principle of an interferometer or the like. The moving distance measuring device 8 is generally an optical device, but detects a physical quantity that changes according to the distance, such as a device using electromagnetic waves, electrical resistance or capacitance, or a device using quantum mechanical effects. An apparatus based on this principle can also be used.

  The sample 5 and the electrode 7 may be arranged in contact with each other so that they can be contacted in a direct current. When the electron beam applied to the sample 5 is modulated at a high frequency, the current can be measured by capacitive coupling. The electrode 7 may be merely adjacent to the sample 5. In general, in a semiconductor manufacturing process, a local oxide film for element isolation is often formed on the back surface of a substrate, so that an insulating film is often formed on the back surface of a wafer. When such a wafer is used as the sample 5, it is also effective to use a capacitive coupling stage in order to make contact between the sample 5 and the stage 6. It is also possible to connect using the side surface of the wafer.

  Since the dimension of the contact hole to be measured is fine, the sample 5 needs to be placed flat on the stage 6. For this purpose, it is also effective to press the outer periphery of the sample 5 with, for example, a ring-shaped jig.

  The current collected at the electrodes is measured by an ammeter 9. The measurement result is converted into a digital signal and output to the data processing device 10. It is also effective to adopt a differential amplifier configuration for the ammeter 9 in order to increase resistance to noise.

  The data processing apparatus 10 performs various data processing, and in particular, information on the structure in the depth direction of the sample to be inspected from the difference in transmittance of the electron beam with respect to the sample to be inspected when the electron beam is scanned with different acceleration voltages. Can be requested.

 FIG. 2 is a block diagram showing a second embodiment of the present invention, and particularly shows a configuration in the case of using a thin electron beam of micron order or less. In this apparatus, an afocal system is formed by the second condenser lens 15 and the objective lens 16 as an electron beam generating system, and the diameter of the incident parallel beam is converted to a parallel beam smaller than the width limited by the aperture 14. It is an electron optical system.

 That is, the electron beam 12 emitted from the electron gun 1 is once converted into a parallel beam by the first condenser lens 13, then passes through the aperture 14 and converted into a thin parallel beam. Next, the beam is focused by the second condenser lens 15 and enters the objective lens 16. In this electron beam generator, since an aperture is not used for shaping the final beam irradiated onto the sample 5, it is possible to easily realize a very thin beam of the order of 100 mm, which is difficult to realize by direct microfabrication. Such a thin beam can obtain the same effect as a case where a thick beam is collectively irradiated by scanning a wide area.

 The apparatus shown in FIG. 1 or 2 can be used for the SEM as shown in FIGS. 4A and 5A by providing a secondary electron detector.

 Here, the contact hole which is the main inspection object of the present invention will be described. The contact hole is a hole penetrating from an oxide film or dielectric surface provided on a base substrate such as a silicon substrate to the base substrate. In a good contact hole, the surface of the underlying substrate or the surface of the underlying wiring layer is exposed.

 The contact hole is formed by applying reactive ion etching using a fluorine-containing gas as an etchant to an oxide film portion opened in a hole shape by a resist. Currently used contact holes have a very elongated structure with an aspect ratio of 10 and a hole diameter of 0.15 microns for an oxide film thickness of several microns constituting the contact hole wall. Reactive ion etching is physicochemical etching. The substantial etching rate is determined from the rate at which fluorine-containing gas ions irradiated at high speed perpendicular to the substrate surface etch the oxide film and the rate at which the fluorocarbon polymer film is generated by the etching reaction. In general, a reaction in which an oxide film is etched proceeds in a contact hole irradiated with fluorine-containing gas ions at a high speed, and a fluorocarbon polymer film is deposited on the side wall of the contact hole formed by the etching reaction. Since this mechanism protects the side walls from etching, very deep vertical holes can be formed. On the other hand, when the etching progresses and the contact hole reaches the underlying substrate, the substrate does not have oxygen, so the etching reaction changes from an oxide film etching reaction to a polymer film-forming reaction, and the etching is automatically performed. To prevent it from progressing into

 However, since these reaction balances are delicate, depending on the subtle changes in the conditions of the manufacturing apparatus, the etching may suddenly stop before it reaches the base, or the base may be etched. Since they create defective contact holes or through holes, it is necessary to detect these defects.

(Measurement of bottom diameter of contact hole) A technique for measuring the bottom diameter of such a contact hole will be described.

  6A and 6B are diagrams for explaining a measurement method. FIG. 6A shows the structure of a target contact hole and its measurement system, and FIG. 6B shows an example of measurement results. The contact hole 43 to be measured is formed so as to penetrate the insulating film 41 provided on the base substrate 42. As the insulating film 41, an oxide film or a nitride film is used. In a good contact hole, the surface of the base substrate 42 or the surface of the wiring layer serving as the base is exposed. An electron beam 31 having a beam diameter of about 100 mm obtained by the apparatus shown in FIG. 1 or FIG. 2 is vertically incident on a sample provided with such a contact hole 43. The acceleration voltage of the electron beam 31 is about 0.5 kV to several kV, and the amount of current is about several nA. When the electron beam 31 passes through the contact hole 43 and reaches the base substrate 43, a current flows through the base substrate 43. This current is called “compensation current”. FIG. 6B shows a compensation current generated when the electron beam is scanned left and right along the center line of the contact hole 43 in correspondence with the irradiation position of the electron beam.

  Since the electron beam has a finite beam cross-sectional area, as shown in FIG. 6B, when the beam crosses the end of the insulating film, the compensation current rises and the beam reaches the bottom of the contact hole completely. The compensation current value is saturated. The same is true when the beam leaves the bottom of the contact hole, the current gradually decreases from the saturation current and the compensation current is zero when the beam leaves the bottom of the contact hole completely.

Since contact holes are almost circular, the distance characterizing the contact hole is the diameter or radius of the circle defining the bottom of the contact hole. In order to obtain these, it is necessary to perform measurement so as to pass through the center line of the contact hole. The position where the electron beam is irradiated is determined from the secondary electron image or CAD data which is design information, and the position of the contact hole is controlled by controlling the wafer stage position control motor or the electron beam with a deflector. Let the center line pass.

  The compensation current shown on the vertical axis in FIG. 6B has a property of changing depending on the thickness of the oxide film at the bottom of the contact hole. That is, the compensation current is almost zero in a thick oxide film region such as a contact hole wall, while a large compensation current is observed in a region where the underlying silicon or wiring is exposed. Accordingly, a graph is obtained in which the observed compensation current has a value along the center line that is zero outside the contact hole region and a certain value in the silicon bare region. Since the region where the compensation current is not zero corresponds to the region where the bottom of the contact hole is exposed, by measuring the distance, a substantial width where the silicon is exposed is obtained, and this is the diameter of the bottom of the contact hole. Can be considered.

  FIG. 7 is a diagram for explaining the measurement for a tapered contact hole. FIG. 7A shows the structure of the target contact hole and its measurement system, and FIG. 7B shows an example of the measurement result. The contact hole 44 to be measured has a contact hole bottom smaller than the contact hole opening, and a contact hole wall is tapered. Although it is a taper, its inclination is steep, and the thickness of the insulating film 41 immediately exceeds 1000 mm just by slightly deviating the electron beam irradiation position from the bottom of the contact hole. In such a thick region, almost no compensation current is generated. A compensation current is generated at the bottom of the contact hole, and the compensation current is almost zero at other places. Therefore, the distance at which the compensation current is observed corresponds to the bottom diameter of the contact hole. Thus, even if the contact hole is tapered, the distance at the bottom of the contact hole can be measured.

  Since the diameter of the electron beam is finite, the compensation current has a rising edge and a falling edge. In this case, the diameter of the contact hole can be obtained from various position information such as the rise and fall of the compensation current or the position where the current returns to zero from the position where the current is completely saturated.

  Depending on the acceleration voltage of the electron beam, the electron beam may pass through the insulating film. When the forward taper has a gentle inclination angle, a compensation current flows even when the electron beam is not irradiated on the bottom of the contact hole. Sometimes. By utilizing this and repeating the measurement while changing the acceleration voltage, the three-dimensional structure of the contact hole can be obtained. This will be described in detail later.

  FIG. 8 shows a change in compensation current amount observed when the contact hole bottom area is changed under the condition that a uniform electron beam is incident on the entire bottom of one contact hole. As shown in this figure, it can be seen that the compensation current is directly proportional to the contact hole bottom area.

  FIG. 9 shows a change in compensation current amount corresponding to a diameter assumed when the contact hole bottom is circular (converted contact hole bottom diameter). Since the area is proportional to the square of the contact hole bottom diameter, the compensation current is a curve proportional to the square.

  10 and 11 are diagrams for explaining measurement using an electron beam having a larger cross-sectional shape than the contact hole opening. In each figure, (a) shows the structure of the target contact hole and its measurement system, (B) shows an example of a measurement result. In this measurement, the electron beam generator of FIG. 1 or FIG. 2 is used, the thickness of the electron beam is set to a value sufficiently larger than the contact hole (for example, several microns square), and the electron beam is applied to the entire bottom of one contact hole. In order to irradiate, the compensation current is measured under the condition of collectively irradiating the electron beam vertically to the measurement object. As the electron beam source, it is desirable to use a sufficiently flat electron beam intensity distribution within 1% of the beam diameter.

  When the electron beam 51 is irradiated onto the bottom of the contact hole 43 or 44 at once, all the compensation currents generated at the exposed portions of the base substrate 42 are added together and measured by the ammeter 9. Since the secondary electron emission efficiency is a value inherent to the substance, the compensation current amount indicated by the region where the base per unit area is exposed is constant when the electron beam irradiation conditions are the same. Therefore, when the electron beam 51 is applied to the entire bottom of the contact hole 43 or 44, as shown in FIGS. 10B and 11B, the electron beam 51 is irradiated. A compensation current proportional to the bottom area of the contact hole 43 or 44 is observed.

(Use of standard samples)
The compensation current value thus obtained may change slightly depending on the measurement conditions. Therefore, the standard value of the compensation current obtained when the sample state is clearly known is used to convert the contact hole area. That is, the compensation current amount per unit area of a standard sample with a known contact hole bottom area is measured under a predetermined electron beam irradiation condition, and then the same sample is irradiated to compensate for the unknown sample. Find the amount of current. By dividing the obtained current value by the compensation current amount of the standard sample, the ratio to the standard contact hole area can be obtained, and the contact hole bottom area can be obtained. This is based on the assumption that the compensation current amount per standard sample unit area is equal to the compensation current amount per unknown sample unit area.

  FIG. 12 is a diagram for explaining the same measurement using a beam diameter smaller than the diameter of the contact hole. FIG. 12A shows the structure of the target contact hole and its measurement system, and FIG. When the beam diameter is smaller than the diameter of the contact hole, the compensation current can be obtained only at the position irradiated with the electron beam. Therefore, by integrating the current value that flows as a result of scanning the entire hole region to be inspected, the contact hole diameter can be obtained by the same method as that for obtaining the total sum of the compensation currents. When the time constant of the current measuring device is large, the integral value and the average current value are approximately proportional to each other. Therefore, the diameter of the hole can be estimated using the average current value indicated by the ammeter.

  When the angle of the forward tapered side wall is shallow, naturally, even when the beam is not irradiated to the bottom of the contact hole, the thickness of the insulating film becomes thin, so that a compensation current flows out. Conditions under which the above method can be simply applied can be determined from the thickness of the insulating film, the intensity of the electron beam, the taper angle, and the like.

  The amount of compensation current per unit area varies depending on the material exposed at the bottom, the electron beam acceleration voltage to be injected, and the amount of injection current. Therefore, when performing measurements with other materials or conditions, necessary preliminary experiments are required. To obtain the relationship between the compensation current and the area and convert it into a table or function.

(Contact hole bottom diameter measurement example)
The present inventors verified the calculation of the contact hole bottom diameter using a sample actually manufactured. In the experiment, first, a cylindrical contact hole as shown in FIG. 10A was prepared in advance as a standard sample. The opening diameter and the bottom diameter are each 0.1 micron, the material exposed at the bottom of the hole is silicon, and the insulating film that becomes the wall of the contact hole is a silicon oxide film. When the electron beam acceleration voltage was 1 kV and the electron beam was injected into the contact hole, a compensation current of 100 pA was observed. Next, a contact hole having the same opening diameter and an unknown bottom diameter was measured under the same conditions as shown in FIG. 11A. A current value of approximately 50 pA was detected in the unknown sample, and it was found that the contact hole bottom area of the unknown sample was 0.5 in the standard state.

  On the other hand, the actual contact hole shape can be examined by SEM cross-sectional shape observation (destructive inspection). By associating the contact hole diameter of the standard sample obtained by the cross-sectional observation with the compensation current value in advance, the contact hole bottom area can be obtained from the measured compensation current value detected in the unknown sample. Assuming that the shape of the bottom of the contact hole is similar regardless of its diameter, the diameter of the contact hole to be measured can be obtained from the square root of the area.

  In the above experiment, the bottom diameter of the contact hole to be measured was determined to be 0.07 microns.

  The scanning of the beam may be performed only once for the contact hole to be measured, but it is also possible to scan the same portion a plurality of times for the purpose of improving the measurement accuracy. In that case, the contact hole diameter can be calculated from the average value of the compensation current observed when a certain inspection region is scanned a plurality of times.

(Determination of compensation current amount per unit area when the bottom area is unknown)
A method of determining the compensation current amount per unit area when a standard sample with a known contact hole bottom area cannot be prepared will be described with reference to FIG. In this method, an electron beam 52 that is sufficiently narrower than the contact hole opening and has a known spot size is perpendicularly incident on the sample. Since the spot size of the electron beam produced by the method of FIG. 1 or FIG. 2 is restricted by the size of the opening provided in the aperture, the size of the electron beam can be obtained by calculation. In order to further increase the accuracy, the diameter of the electron beam is directly obtained by the knife edge method or the like to obtain the electron beam diameter. When this electron beam is injected into the standard contact hole, a compensation current as shown in FIG. 12B is measured. By dividing the obtained compensation current by the spot size of the electron beam, the compensation current per unit area of the standard sample is obtained.

(Use in mass production plants)
FIG. 13 shows a flowchart and a pass / fail judgment example when the measurement of the contact hole bottom diameter described above is used in a mass production factory.

  The size of the contact hole is one factor that determines the amount of current flowing through the contact hole. In a high-speed memory or logic device, a very high-speed pulse signal operates at a delicate timing. The variation in the size of the contact hole changes the time constant of the circuit through the contact resistance, so that the pulse transmission time becomes a value different from the design and causes a malfunction of the circuit. Therefore, if there is a variation in contact hole size over a certain range, even if conduction is established, the circuit becomes defective. In order to prevent the occurrence of such defects, mass production factories need to strictly manage variations in the bottom diameter of manufactured contact holes.

  For example, consider a case where the manufacturing tolerance is ± 0.01 microns (± 10%) and the manufacturing is performed with the goal of obtaining a contact hole bottom having a diameter of 0.1 microns. A tolerance of 10% of the diameter is equivalent to a tolerance of ± 20% in terms of area. Using this criterion, the contact hole bottom is managed as shown in the flowchart of FIG.

  First, a contact hole to be measured existing on a wafer is irradiated with an electron beam to measure a compensation current of each contact hole. The result is recorded in a memory or a magnetic disk. There is no problem as long as the result can be recorded even if it is other than the memory and the magnetic disk. Subsequently, the compensation current value indicated by the non-defective contact hole is compared with the compensation current value measured and recorded. Next, the compensation current indicated by each contact hole and the reference value are compared using a comparator. If the difference in size with respect to the reference value is within ± 20%, the measured contact hole is judged to be non-defective, and information indicating that the contact hole is non-defective is stored in the storage device. FIG. 13R> 3 shows the value of the contact hole diameter obtained by scanning the Y coordinate from “1” to “5” with the X coordinate “1” and an example of the quality determination result. If the number of defective contact holes in the entire wafer is below a certain reference value, it is placed in a wafer carrier for transport to the next process. On the other hand, if the reference value range is exceeded, the wafer to be inspected is judged to be defective, the subsequent process processing is stopped, the wafer is separated into a carrier for discarding, and the etching apparatus is maintained. Give instructions.

  Since the measured value of each contact hole bottom diameter is obtained, it is possible to compare with the statistical value when there is no abnormality by calculating the statistic such as dispersion and average value. Since this comparison can be analyzed before the contact hole conduction failure actually occurs, it is possible to grasp the process fluctuation and the change tendency in detail. Furthermore, since defects can be found promptly, occurrence of defects in subsequent products can be prevented by investigating and dealing with the causes of defects.

  In recent semiconductor integrated circuit devices, the circuit scale has become enormous, and the number of contact holes has increased geometrically accordingly. In such a case, it is difficult to measure all the contact holes on the wafer. On the other hand, a plurality of identical elements (chips) are regularly arranged on the wafer. Therefore, it is possible to determine whether each chip is good or bad by scanning the same portion of each chip to be measured from one chip to the next. In this case, it is also possible to collectively measure a plurality of contact holes at the scanning positions of each chip and obtain an average value of the bottom diameters. When obtaining the average value of the bottom diameters of a plurality of contact holes, a single electron beam shaped by an aperture may be irradiated at once, or a thin electron beam may be scanned. When irradiating the electron beam all at once, the average value can be obtained by performing irradiation a plurality of times. Similarly, when scanning with an electron beam, the number of scans may be one or more.

(Map display)
Furthermore, the result obtained from the measurement of the compensation current of the wafer or the diameter of the contact hole can be mapped in correspondence with the measurement position. For example, the distribution state of the contact diameter in the wafer can be known by mapping the compensation current value or the diameter of the contact hole into a contour line. This contour line display can be performed by storing the obtained compensation current value and position information and displaying them on an image display device or recording paper.

  When displaying on an image display device, if it is displayed on the screen based on the compensation current value, aperture diameter, etc., it may be difficult to see due to excessively high brightness or low brightness. Therefore, it is necessary to correct this so that the display screen is always visible. As a correction method, for example, the brightness may be adjusted based on a central value. In addition, since the generation of defective products is more important than non-defective products in the semiconductor manufacturing process, it is desirable to make information relating to defective products easier to see.

(Pass / fail judgment and process evaluation by map display)
The pass / fail judgment can be classified by compensation current, contact diameter, or contact shape. By classifying these classifications on the basis of conditions such as the same etching conditions, the same processing apparatus, or the same processing apparatus in the previous process for each wafer or a plurality of wafers, various information can be collected. It is desirable to display these classification data by the same method as the contour line display. As a result, not only the quality of the wafer can be judged, but also the etching distribution of the etching equipment and other processing conditions can be known, and the amount of etching of the etching equipment can be averaged, etc. It is easy to optimize the conditions.

  For example, contact holes are usually made using a dry etching method, and the device is adjusted so that the etching rate is constant over the entire surface of the device, but still there are places where the etching rate is fast and slow. End up. When comparing the contour display of the measurement results for contact holes for multiple wafers, you can see the trend of the overall distribution of the device, and adjust the device such as changing the tilt of the electrode so that it can be corrected Thus, the uniformity of the etching rate can be improved.

  Various methods for collecting the distribution of a plurality of wafers are conceivable. For example, it is possible to collect only wafers performed under the same working conditions or standardize wafers under different conditions.

  As a cause of the distribution in the etching rate, there is an influence of the thickness distribution of the insulating film, which is the previous process. The thickness distribution of the insulating film is caused by the state of the CVD apparatus. In such a case, it is possible to investigate the cause of the defect by summing up data of wafers in which the previous manufacturing process and etching process were performed with the same model. By collecting such data, it is possible to identify problems in the previous process from the measurement of the contact diameter for determining the quality of etching.

  In addition, when the tendency of the apparatus is known, the measurement time can be shortened by focusing on only the portion where the wafer is likely to be defective instead of inspecting the entire surface of the wafer. For example, it is performed by measuring only portions where the etching speed is high (compensation current value is high, aperture diameter is large) or the etching speed is slow (compensation current value is low, aperture diameter is narrow). Is also possible.

  From the distribution of contact hole diameters, in addition to what has been described above, useful information can be obtained for new start-up of equipment, adjustment after overhaul, replacement after repair, etc., and repairs. Therefore, it is possible to complete these operations in a short time by supplying data necessary for operations such as startup and adjustment. The distribution of the contact hole diameter is also useful as apparatus maintenance information. For example, it is possible to accurately estimate the overhaul time by statistically monitoring the deviation / increase of the defect distribution from the normal distribution, or by utilizing a foreign substance inspection described later. Also, there is an effect that an abnormality of the apparatus can be detected before a failure occurs.

  Regarding the wafer pass / fail judgment, in the normal process, a plurality of wafers are put into semiconductor manufacturing as one set (this is called “batch”), so the first wafer and the last wafer of the batch to be put into the process May be inspected. If it is found as a result of inspection that the state of the first wafer is in a dangerous state, all subsequent wafers that are put into the process are inspected, and the equipment in the process can be adjusted when a defect occurs. Alternatively, the apparatus of the process may be adjusted based on the result of the leading wafer.

(Linkage with SEM)
Since the contact hole is a three-dimensional structure, it is very convenient to obtain inspection results that allow the features of the solid to be understood at a glance. A method for obtaining an accurate three-dimensional structure will be described later. Here, a method for knowing an approximate structure relatively easily will be described.

  In the method described here, since a contact hole or a through hole usually has a circular cross section, the hole opening diameter α, the hole bottom diameter β, and the depth d thereof are specified, and the shape of the contact hole to be measured is determined. Express roughly. That is, the contact hole bottom shape or material information obtained from the compensation current measurement and the contact hole opening shape obtained from the normal scanning electron image are synthesized. The material information is estimated from the amount of compensation current measured using several acceleration voltages using the fact that the amount of compensation current varies depending on the underlying material. The depth of the contact hole can be obtained by electron beam measurement described later, but an actual measurement value of the thickness measured at the time of forming the insulating film in which the contact hole is formed may be used.

  14 and 15 are diagrams for explaining the measurement method. FIG. 14 shows a measurement example of a contact hole having a cylindrical shape, and FIG. 15 shows a measurement example of a contact hole having a forward tapered shape. In each figure, (a) shows the structure of the target contact hole and its measurement system, and (b) shows the amount of secondary electrons and the compensation current measured along the center line of the contact hole. It shows with respect to the position, and (c) shows the restored three-dimensional display figure. Here, for simplicity, it is assumed that scanning is performed only once along the contact hole center line.

  A parallel electron beam obtained by the apparatus shown in FIG. 1 or 2 is used as an electron beam for scanning around or inside the contact hole. When the convergent electron beam is used, it is necessary to converge the electron beam to the respective heights when scanning the peripheral part of the contact hole and when scanning the bottom part. On the other hand, if a parallel electron beam is used, the depth of focus becomes infinite, and it is not necessary to adjust the focus.

  In the contact hole 43 shown in FIG. 14A, the opening diameter α and the bottom diameter β are substantially the same. In this case, as shown in FIG. 14B, the rising and falling positions of the graphs of the secondary electron amount and the compensation current coincide with each other. From this measurement result and the depth d of the contact hole obtained from the process data, a three-dimensional display figure of the cylindrical contact hole as shown in FIG. 14C is obtained. Furthermore, if a number of cross-sectional shapes are measured while changing the scanning direction so as to pass through the center of the contact hole, a more accurate three-dimensional display figure of the contact hole can be obtained. Various methods used in the world of three-dimensional computer graphics can be used as a method of reducing the cross-sectional shape into a three-dimensional image.

  The contact hole 44 shown in FIG. 15A has a contact hole bottom diameter β smaller than the opening diameter α. In this case, as shown in FIG. 15B, there is a difference between the rising and falling shape of the graph obtained from the secondary electrons and the rising and falling shape of the graph obtained from the compensation current. The width of the rectangular region of the graph indicating the amount of secondary electrons corresponding to the opening diameter α is larger than the width of the rectangular region indicated by the compensation current corresponding to the bottom diameter β, and there is a difference between the opening diameter and the contact hole bottom diameter. I understand that. If the contact hole depth d known from the process data is taken into consideration and displayed three-dimensionally, the shape of the contact hole to be inspected is displayed as an inverted triangular pyramid as shown in FIG.

  Note that in the case of a tapered contact hole as shown in FIG. 15A, secondary electrons emitted from the tapered portion are detected depending on the positional relationship between the tapered shape and the secondary electron detector. There is also. However, since the actual contact hole impact is high, the actual situation is that almost no secondary electrons are observed from the inner wall of the hole. In FIG. 15B and the following figures, such secondary electrons are ignored unless otherwise stated.

(Linkage with SEM and oblique incidence of beam)
In the case of a reverse-tapered contact hole where the size of the bottom of the contact hole is larger than the width of the opening, the contact hole having the same diameter as the diameter of the opening is obtained by simply injecting an electron beam vertically. Cannot be distinguished. Accordingly, the shape of the contact hole bottom is measured by tilting the incident angle of the electron beam with respect to the sample to be inspected so that the electron beam reaches the peripheral portion of the bottom of the contact hole. To tilt the electron beam at a small angle, an electron lens or an electron beam scanning deflector is used. To tilt the electron beam at a large angle, the wafer support stage is tilted with the wafer center as the tilt axis. If the stage is tilted, an inclination of ± several tens of degrees can be easily realized, so that an electron beam having an angle substantially equal to the taper angle of the contact hole can be injected.

  FIGS. 16, 17 and 18 show measurement examples in which cylindrical, forward tapered and reverse tapered contact holes are irradiated with respective inclined electron beams. In each figure, (a) shows the structure of the target contact hole and its measurement system, and (b) shows the amount of secondary electrons and the compensation current measured along the center line of the contact hole. Shown against position. The deviation between the measurement position of the amount of secondary electrons due to the tilt of the electron beam and the measurement position of the compensation current is corrected and shown in correspondence with the position of the contact hole. FIG. 18C shows the restored three-dimensional display figure.

  When the columnar contact hole 43 is irradiated with the inclined electron beam 61 while moving along the central axis of the contact hole, the region of the insulating film 41 around the contact hole 43 is irradiated with the electron beam 61 for two times. Secondary electrons 32 are strongly observed. When the electron beam 61 reaches the edge region of the contact hole 43, the secondary electron intensity rapidly decreases. While the bottom of the contact hole 43 is irradiated with the electron beam 61, secondary electrons are not observed. When the electron beam 53 strikes the insulating film 41 on the opposite bank again, secondary electrons are detected. On the other hand, the compensation current is not observed while the electron beam 61 is applied to the insulating film 41, and starts to be detected when the electron beam 61 reaches the edge of the contact hole 43. When the electron beam 61 comes into contact with the bottom of the contact hole 43, the compensation current suddenly increases. When the electron beam 61 comes into contact with the insulating film 41 again, the amount of current suddenly decreases.

  In the tapered contact hole 44, secondary electrons are detected strongly when the insulating film 41 is irradiated with the electron beam 61. When the electron beam 61 reaches the edge of the contact hole 44, the amount of secondary electrons suddenly increases. Decrease. While the bottom of the contact hole 44 is irradiated with the electron beam 61, secondary electrons are hardly detected. When the insulating film 41 outside the contact hole 44 is irradiated again, the secondary electrons are observed. Will come to be. On the other hand, the compensation current is not detected while the electron beam 61 is applied to the insulating film 41 around the contact hole 44, and a large compensation current is applied while the electron beam 61 is applied to the bottom of the contact hole 44. Detected. When the electron beam 61 is again irradiated onto the tapered portion, the compensation current decreases rapidly.

  In the reverse tapered contact hole 45 shown in FIG. 18, secondary electrons are strongly detected while the electron beam 61 irradiates the insulating film 42 around the contact hole 45, and the electron beam 61 is detected at the edge of the contact hole 45. The amount of secondary electrons decreases drastically when approaching. While the electron beam 61 irradiates the bottom of the contact hole 45, almost no secondary electrons are detected, and secondary electrons are observed only after the electron beam 61 is irradiated on the surface of the opposite side of the contact hole 45. become. On the other hand, the compensation current is not detected when the electron beam 61 irradiates the surface of the insulating film 42, and the compensation current is detected only while the electron beam 61 is irradiated on the bottom of the contact hole 45. When the insulating film 42 or its inversely tapered portion starts to be irradiated again, the compensation current is not detected.

  When the inclination of the contact hole taper and the inclination of the incident electron beam coincide with each other, an increase or decrease in the amount of secondary electrons and a decrease or increase in the compensation current occur at the same beam irradiation position. Therefore, in order to obtain the bottom diameter of the inversely tapered contact hole, an experiment is performed at various electron beam incident angles, and the angle at which the compensation current is detected at the outermost side is searched. Since the depth d of the hole is known, the distance from the incident angle and depth of the electron beam to the outermost periphery of the bottom of the contact hole as viewed from the opening can be obtained, and the value is added to the diameter of the opening. Thus, the hole bottom diameter is calculated. If these are added and displayed in three dimensions, the three-dimensional figure of the contact hole shown in FIG. 18C is obtained.

(Foreign object detection)
19 to 21 are diagrams for explaining a method for detecting and specifying a foreign substance generated in a contact hole. In each figure, (a) shows the structure to be inspected and the measurement system, and (b) shows the amount of secondary electrons to be measured and the compensation current corresponding to the irradiation position of the electron beam.

  Various foreign substances such as resist soot, particles generated during etching, and dust generated by other processes may exist at the bottom of the contact hole. If these foreign substances are present, filling of tungsten, aluminum, polysilicon, or the like deposited in the contact hole in order to achieve electrical continuity as a plug becomes insufficient, resulting in an electrical contact failure. Therefore, in a semiconductor process, it is necessary to detect these foreign substances before plug formation.

  The dust in question depends on the material, but generally has a thickness of 500 mm or more, which prevents the irradiated electron beam from reaching the bottom of the contact hole. For this reason, when foreign matter exists at the bottom of the contact hole, the observed compensation current is smaller than the compensation current generated in the normal contact hole.

  In the example of FIGS. 19A and 19B, a small dust 71 exists outside the cylindrical contact hole 43. When the thin electron beam 31 generated by the method of FIG. 1 or 2 is scanned along the contact hole 43 from the left side of FIG. 19A perpendicularly to the sample, the electron beam 31 is insulated around the contact hole 43. While the film 41 is irradiated, no compensation current is observed. When the electron beam 31 reaches the edge of the contact hole 43, the compensation current starts to be detected. While the compensation current is detected while the bottom of the contact hole 43 is irradiated with the electron beam 31, the compensation current is not detected when the dust beam 71 is irradiated with the electron beam 31. In the example of FIG. 19A, the dust 71 is clung to one end of the bottom of the contact hole 43. However, when it exists in the center, the compensation current is observed again when the electron beam 31 exceeds the dust region. It becomes like this. A change in compensation current is measured corresponding to the irradiation position of the electron beam 31, and a result as shown in FIG. 19B is obtained. By comparing this with the measurement result when there is no dust, the size of the dust 71 can be obtained. Thus, the presence or absence or size of dust at the bottom of the contact hole can be obtained by measuring the compensation current.

  FIGS. 20A and 20B show an example where dust 72 is deposited on the bottom of the tapered contact hole 44. When scanning of the electron beam 31 is started from the left side of FIG. 20A, no compensation current is detected while the insulating film 41 around the contact hole 44 is irradiated with the electron beam 31. Even during irradiation of the tapered portion, the compensation current is not detected because the insulator is thick. On the other hand, when the electron beam 31 is applied to the bottom of the contact hole 44, a compensation current is detected. A compensation current of the same magnitude is detected while the bottom is protruding, but the compensation current is not detected while the region where the dust 72 exists is being irradiated with the electron beam 31. By comparing with a graph of a sample without dust, it is possible to check the presence or size of dust.

  FIGS. 21A and 21B show the case where the dust 73 exists at the center of the inverted tapered contact hole 45. When the electron beam 31 is scanned, no compensation current is detected while the electron beam 31 is irradiated on the insulating film 41 around the contact hole 44. When the electron beam 31 reaches the edge of the contact hole 45, a compensation current starts to be observed, and a large compensation current is detected while the bottom of the contact hole 45 is irradiated with the electron beam 31. When it reaches the garbage 73, the compensation current is not detected. When the electron beam 31 irradiates the bottom of the contact hole 45 again after passing through the dust 73, the compensation current is detected again. When the electron beam 31 reaches the edge of the contact hole 45, the compensation current is not detected. The position where the compensation current is not detected corresponds to a region where the dust 73 exists, and the size of the dust 73 can be estimated from this width.

  In the above dust detection method, the secondary electron emission ratio varies depending on the material, and the size of the dust detection method varies depending on the atomic number and the energy dependency of the irradiated electron beam. Therefore, after the presence or absence of dust at the bottom of the contact hole is specified, an electron beam having various acceleration energies is irradiated and the change in compensation current is observed. Measurement is performed in advance by determining the dependence of the compensation current on the acceleration energy by performing the same experiment on the assumed foreign material and calculating the degree of approximation of the dependence using a technique such as a neural network. The object can be specified.

(Measurement by electron beam with large cross-sectional shape)
22A and 22B are diagrams showing a measurement example using an electron beam having a large cross-sectional shape, where FIG. 22A is a plan view showing the positional relationship between the contact hole 81 and the electron beam 82, FIG. 22B is a cross-sectional view, and FIG. Indicates the compensation current obtained with respect to the scanning position of the electron beam and its differential value.

  In this example, an electron beam 82 having a rectangular cross section, an area larger than the size of one contact hole, and perpendicularly incident on the sample is used. As shown in FIGS. 22A and 22B, such an electron beam 82 is maintained with its irradiation axis perpendicular to the sample from one of the regions including one contact hole 81 toward the other. And scanning without rotating the beam axis. In this scanning, the electron beam 82 itself may be scanned, or the irradiation axis of the electron beam 82 may be fixed at a fixed angle with respect to the wafer, and the wafer may be moved horizontally. The electron beam 82 used here is a parallel beam. However, even with such a parallel beam, the beam can be scanned by using two deflectors to translate and shift the beam. The magnitude of the compensation current detected at this time is proportional to the area of the electron beam applied to the bottom of the contact hole 81. Therefore, the differential value indicates the compensation current amount at a position near the beam edge 83 where the rectangular electron beam 82 is about to be irradiated.

  As shown in FIGS. 22A and 22B, in this example, scanning is performed so that the electron beam 82 is gradually irradiated from the region around the contact hole 81 to the bottom of the contact hole 81. While the electron beam 82 irradiates the area around the contact hole 81, no compensation current is detected. When the bottom of the contact hole 81 is irradiated with the electron beam 82, the compensation current increases rapidly. While the electron beam 82 passes through the bottom of the contact hole 81, the compensation current amount gradually increases, and reaches a maximum value when the entire contact hole 81 is included. When the electron beam 82 passes through the contact hole 81 and the beam edge on the other side reaches the contact hole 81, the compensation current starts to decrease, and when the electron beam 82 is no longer irradiated to the region of the contact hole 81, the compensation current is detected. It will not be done.

  The distance between the measured rise position of the compensation current and the top of the peak indicating the maximum value of the compensation current corresponds to the distance from one end to the other end of the bottom of the contact hole 81. Since the distance measured by this method corresponds to the distance obtained when the circle is sandwiched between two parallel lines, the accuracy of the circle can be obtained without accurately aligning the electron beam 82 with the contact hole 81. Can measure any diameter.

  Assuming that the contact hole 81 is circular, the area increase rate of the circle is maximized at the location of the center line of the circle, so the position where the increase rate of the compensation current is maximized is the position of the center line of the circle. Equivalent to. Therefore, the bottom diameter of the contact hole 81 can be obtained by performing the measurement up to the position where the increasing rate of the compensation current is maximized without scanning the entire contact hole 81. That is, the contact hole bottom diameter can be measured in approximately half the time. Moreover, since the vertex of the differential value is clear, the distance can be obtained accurately.

  The use of a thick electron beam has the advantage that the configuration of the electron beam system of the inspection apparatus is easier than when a thin electron beam is used.

(Measurement combining a thin electron beam and a thick electron beam)
FIG. 23 and FIG. 24 are diagrams for explaining a measurement method combining a length measurement mode with excellent measurement accuracy using a thin electron beam and a collective measurement mode for obtaining a contact hole bottom diameter in a short time using a thick electron beam. FIG. 23 is a flowchart, and FIG. 24 shows an example of the positional relationship between the length measurement mode target area 92 and the batch measurement mode target area 93 on the wafer 91.

  In the manufacture of semiconductor devices, it is necessary to accurately measure the contact hole bottom diameter at high speed. In general, in the length measurement mode in which a precise narrow electron beam scan is performed on one contact hole and the contact hole bottom diameter is measured from the interval of the position where the compensation current changes, the relative change in the compensation current is used. There is little influence of variation, and contact hole diameter measurement accuracy is high. However, since a large amount of information is acquired by performing precise electron beam scanning on one contact hole, it takes a long time and a large amount of data processing to inspect many contact holes.

  Therefore, a measurement mode obtained by scanning a thin electron beam with excellent measurement accuracy and a collective measurement of contact hole diameter by measuring the compensation current flowing through the entire contact hole bottom using a relatively thick electron beam. Combine with measurement mode. As a result, the inspection accuracy can be maintained high and the inspection speed can be increased.

  A specific example will be described with reference to FIG. 23 and FIG. 24. First, length measurement is performed on one or a relatively small number of contact holes (in the length measurement mode target region 92) among the contact holes to be measured. Accurately measure contact hole bottom diameter in mode. Next, the batch measurement mode is applied to the same contact hole, the relationship between the compensation current flowing through the contact hole and the contact hole bottom diameter is obtained, and the area in the batch measurement mode is normalized. By this measurement, the relationship between the contact hole diameter and the compensation current in the measurement object is determined. Next, the collective mode is sequentially applied to the other contact holes (in the collective measurement mode target region 93), and the compensation current for each contact hole is measured, and the measurement is performed based on the relationship between the compensation current obtained earlier and the contact hole bottom diameter. The compensated compensation current is converted into the contact hole bottom area or diameter. Thereby, the contact hole bottom diameter can be measured at high speed with high measurement accuracy.

(Comparison inspection of two areas)
FIG. 25 shows an apparatus configuration for performing a comparative inspection using two specimens to be inspected, and FIG. 26 shows an inspection flow thereof. FIG. 27 is a diagram for explaining the principle of comparative inspection, and FIG. 28 is an enlarged view thereof.

  Since the circuit pattern of a semiconductor LSI is produced using an exposure apparatus called a stepper, the distance between adjacent chips or the layout inside the chip is exactly the same. Referring to FIGS. 27 and 28, the internal layout of the chip expressed in relative coordinates with the corner of the first sample 101 as the origin (0, 0) is the second sample on the same wafer. This exactly matches the internal layout of the chip in which the corner of the sample 102 is represented as the origin (0, 0). Therefore, if a change exceeding a certain standard is recognized by comparing the two, the region is considered to contain a defect. By using such a comparative inspection, it is possible to specify the position of the wiring defect regardless of the arrangement state of the wiring without having to know the layout information of the inspection sample from the CAD data. The first and second specimens 101 and 102 are regions formed on the same substrate and finally cut out as one chip.

  As shown in FIG. 25, the apparatus for performing the comparative inspection includes an electron gun 112 that vertically irradiates an inspection sample on the wafer 111 with an electron beam, and a compensation current measurement in which the wafer 111 is disposed in contact with the bottom. An electrode 113, an XY stage 114 on which the electrode 113 is placed and which determines the positional relationship between the wafer 111 and the irradiated electron beam, an electron beam irradiation position detecting device 115 for accurately measuring the position irradiated with the electron beam, the electron An electron beam irradiation position control device 116 that generates a control signal for controlling the irradiation position of the electron beam based on the detection result of the beam irradiation position detection device 115, and based on a control signal output from the electron beam irradiation position control device 116. On the basis of control signals output from the electron gun control device 117 for controlling the electron gun 112 and the electron beam irradiation position control device 116. A stage controller 118 for controlling the XY stage 114, a current amplifier 119 for amplifying the current of the electrode 113, a D / A converter 120 for converting the output of the current amplifier 119 into a digital signal, and a current waveform corresponding to the position of the digital signal. As the first and second storage devices 121 and 122, the waveform comparison device 123 for comparing the stored current waveforms, the failure determination device 124 for determining the quality of the wiring based on the comparison result, and information for determination A recorded determination database 125, a defective position storage device 126 for storing a position opposite to a defect, and a defective position output device 127 for displaying, printing, or outputting the defective value to other processing devices on the network are provided. . As the electron beam irradiation position detecting device 125, for example, an optical precision distance measuring device is used.

  Since the storage devices 121 and 122 store the current waveforms corresponding to each chip separately, they are shown as different in FIG. 25, but these can be implemented by a common storage device. Further, the defective position storage device 126 is also shown as a separate device in FIG. 25, but another storage area of the same device as the storage devices 121 and 122 can be used.

  The defect position storage device 126 can rank defects as necessary and record position information by rank.

  The electron beam scanning is performed by fixing the electron gun 112 at a specific position and moving the XY stage 114 relative to the position of the electron gun 112. By measuring the position of the XY stage 114 by the electron beam irradiation position detector 115, the position where the electron beam is irradiated can be accurately measured with an accuracy of 100 mm. While the electron beam is scanned in a line with respect to the first inspection sample 101 on the wafer 111, the current generated in the sample is measured as a first current waveform by the current amplifier 119 and the D / A converter 120. The first electron beam irradiation position coordinates calculated from the position of the XY stage 114 are recorded in the first storage device 121 as a set. The same measurement is also performed on the second sample 102 to be inspected, which is the same pattern formation location on another chip, a second current waveform is acquired, and the second electron beam irradiation position coordinates are paired with the second electron beam irradiation position coordinates. Record in the storage device 122. The defect determination device 124 determines a pattern defect from the difference between the current waveforms stored in the first and second storage devices 121 and 122 and stores the pattern defect in the defect position storage device 125. If necessary, the data is output from the defective position output device 127 to a display, a printer, or another device on the network so that it can be used for other analysis.

  In the case of measurement using the compensation current, unlike the case of using secondary electrons, the electron beam irradiated to other than the wiring does not generate an effective current, so it was detected as compared with the case of secondary electrons. The noise contained in the signal is small.

  The point at which the comparison between the non-defective product and the defective product is performed depends on the capacity of the storage device used for the waveform storage devices 121 and 122. When comparing a good product and a defective product for each line, a storage capacity capable of recording a waveform for one line is sufficient. When measuring defective chips after measuring all good chips, a storage capacity is required to store all the information for one chip. Since it takes a lot of time to move the electron beam irradiation position between physically separate chips, in order to improve the inspection speed, the measurement of the next chip is completed after the measurement of one chip is completed. It is preferable to use a storage device having a capacity as large as possible.

  29A and 29B are diagrams showing inspection examples, where FIG. 29A shows a measurement example of non-defective chips, and FIG. 29B shows a measurement example of defective chips. The number on the left side of the figure indicates the line number of the electron beam scanned in a strip shape, and the right W indicates the width of the electron beam scanned at one time. In addition, on the lower side of FIGS. 29A and 29B, a compensation current observed during the fourth electron beam scanning related to the pattern defect is shown. In this example, the size of the wiring to be inspected is assumed to be constant (for example, 0.15 microns), as can be seen in a normal semiconductor device. In general, a semiconductor device has a distance between wirings longer than the diameter of the wiring due to restrictions such as an exposure technique and an etching technique. In this example, the wirings are arranged at random and do not have a certain period.

  The chips shown in FIGS. 29A and 29B have the properties described with reference to FIGS. 27 and 28, and are arbitrarily selected from a plurality of chips formed simultaneously on the semiconductor wafer. Which chip is subject to comparative measurement depends on the case, but in general, a specific chip that is easy to pick up next to each other or a non-defective product is designated as the first sample to be tested, and the second sample to be tested is selected. The inspection is performed in turn. By comparing the inspection results of three or more chips, it is possible to determine that a plurality of chips having the same inspection result are good products.

  The determination of whether the wiring is good or bad using an electron beam uses changes in the magnitude and polarity of a current generated when the electron beam is irradiated. Here, for the sake of simplicity, since there is a pattern defect in the defective product wiring, the discussion proceeds assuming that the current observed for the defective product wiring is considerably smaller than that in the case of the non-defective product wiring.

  A specific inspection method will be described. First, the position coordinates of a chip to be an inspection sample are matched with the position coordinates for performing electron beam irradiation. Since the size of the wiring of the cutting-edge device that is the inspection sample is 0.2 microns or less, alignment is performed with a positional accuracy higher than 1000 mm, at which this positional coordinate can be sufficiently reproduced. To do this, alignment marks formed on the wafer are used.

  There are several methods for using alignment marks. One is a method using an alignment mark for mask alignment which is usually used in a semiconductor process. This alignment mark is made of an oxide film, a metal film, and the like, and is on the substrate surface. Therefore, the alignment mark is converted into a secondary electron image using a scanning microscope provided inside the inspection apparatus. Since the position visible in the image is the position where the electron beam is correctly irradiated, the position coordinates of the electron beam scanning system are matched so that the position of the alignment mark is the origin.

  As another method not using the scanning microscope, there is a method of measuring the current flowing through the alignment mark. An alignment mark having the same wiring shape as that of the inspection sample is separately prepared. The size of the wiring may be the same size as the inspection sample, or a smaller size may be used to increase the measurement accuracy. Similar to the measurement principle of the wiring, the observed current is small while the electron beam irradiates a place other than the wiring, and a large amount of current flows when the electron beam is irradiated to the position of the wiring. When the electron beam irradiation position coincides with the mark, the observed current becomes maximum, and this position is used as a position where alignment is in alignment.

  After the alignment is completed, a linear electron beam 131 whose length corresponds to the wiring width is applied to the surface of the inspection sample along the line “1” with respect to the region where the wiring 132 of the first inspection sample is formed. Scan vertically from left to right. When the end of the inspection area is reached, the irradiation position of the electron beam 131 is shifted by the width W in the direction perpendicular to the scanning direction, and the scanning of the electron beam 131 is performed again along the line “2”. The scanning direction may be s-shaped or meandered, or may be scanned from left to right after returning to the initial position. The movement width W in the vertical direction is approximately the size of the wiring width. Similarly, scanning is performed along the lines “3”, “4”, “5”, “6”, “7”, and the entire test sample is scanned without any problem.

  As shown in FIGS. 29A and 29B, in the scanning along the fourth line, the current is observed when the electron beam scanning position comes to the position corresponding to the wiring 132 in the non-defective product. No current is observed in the defective wiring 133. That is, since a difference occurs in the current waveform at the position of the pattern defect 134 between the non-defective product and the defective product, it can be seen that a defect exists there.

  In the above inspection method, the position of the defective wiring can be specified even if the wiring position of the inspection sample is unknown.

(Comparative inspection with a thin electron beam)
FIG. 30 shows an example using a thin electron beam as another example of the comparative inspection. (A) is a measurement example of non-defective chips, and (b) is a measurement example of defective chips. The numbers on the left side of the figure indicate the electron beam line numbers. Further, the lower side of FIG. 30 shows the compensation current observed during the fourth electron beam scanning related to the pattern defect.

  When the compensation current is measured, when the material of the wiring is uniform, the same current flows if converted into a unit area regardless of which part of the wiring is irradiated with the electron beam. Accordingly, it is not always necessary to irradiate the entire wiring uniformly with an electron beam in order to perform a quality inspection of the wiring. Further, since the pattern defect occurs from the periphery of the pattern, the defect detection sensitivity can be increased by setting the electron beam irradiation position in the periphery of the wiring. Conversely, if it is set so that the center of the wiring is irradiated with an electron beam, the defect detection sensitivity is reduced. In the example of FIG. 30, a point-like parallel electron beam 141 that is sufficiently narrower than the wiring width is used, and scanning is performed at a scanning interval L that is about the same as the wiring width.

  Similar to the inspection shown in FIG. 29, this inspection can also be performed according to the inspection flow shown in FIG. 26 using the apparatus shown in FIG. However, it is assumed that the electron gun 112 can generate a point-like parallel electron beam that is sufficiently narrower than the wiring width.

  In this inspection, alignment is performed in the same manner as the inspection described with reference to FIG. 29, and the area where the wiring 142 of the first inspection sample is formed is compared with the wiring width along the line “1”. An electron beam 141 having a sufficiently small size is incident on the surface of the inspection sample perpendicularly and scanned. As described above, since the electron beam irradiation position affects the defect detection sensitivity, it is set so that it passes through the center of the wiring when it is desired to increase the sensitivity, and when it is desired to decrease the sensitivity. When the end of the inspection area is reached, the electron beam irradiation position is shifted in the direction perpendicular to the scanning direction by the scanning interval L, and scanning of the electron beam is performed again along the line “2”. The scanning direction may be s-shaped, or scanning may be performed in the same direction after returning to the initial position. The scanning interval L is approximately the size of the wiring width. Similarly, scanning is performed along the lines “3”, “4”, “5”, “6”, “7”, and the entire test sample is scanned without any problem. The above measurement is performed also on the second inspection sample which is the same pattern formation portion of another chip, and each current waveform is stored corresponding to the irradiation position coordinates of the electron beam.

  In the example of FIGS. 28A and 28B, in the fourth scan, a compensation current obtained by scanning the electron beam 141 at a position corresponding to the wiring 142 is observed in the non-defective product, but the defective product wiring 143 is reached. Sometimes, no compensation current is observed, and by detecting this difference, the wiring pattern defect 143 can be detected.

(Comparative inspection with linear electron beam)
FIG. 31 is a diagram for explaining an example in which a linear electron beam is simultaneously irradiated onto a plurality of randomly arranged wirings as another example of the comparative inspection, (a) is a measurement example of a non-defective chip, and (b). Is an example of measuring a defective chip. The electron beam used here is very short in the scanning direction, for example, has a linear shape with a length of several microns, such as 100 mm, including a plurality of wirings in the direction perpendicular to the scanning direction. The irradiation current amount of the electron beam 165 is several pico to several nanoamperes, and the acceleration voltage is several hundred to several kilovolts.

  In the non-defective sample shown in FIG. 31A, current starts to flow when the electron beam 151 reaches the wiring 152 at the position a. Further, when the electron beam 151 reaches the position b, the range of the wiring 152 to which the electron beam 151 is irradiated becomes large, and thus a current further flows. When the electron beam 151 reaches the position c, the range of the wiring 152 to which the electron beam 151 is irradiated becomes small, and the current decreases. When the electron beam 151 reaches the position d, the current becomes zero because the electron beam 151 is not irradiated. On the other hand, in the defective product sample shown in FIG. 31B, when the current value obtained by the wiring 154 is small and the two current waveforms obtained at the position where the same pattern of the good product and the defective product is formed are compared, The waveform varies depending on the pattern defect 156. On the other hand, at the positions e, f, g, and h where the non-defective wirings 153 and 155 are formed, the same current waveform is obtained in both the non-defective and defective chips.

  As described above, even when a plurality of wirings are irradiated with an electron beam, if the wiring positions are shifted, the current generated by each wiring is measured independently with respect to the measurement position. Since there is a large difference in the current waveform, it is possible to detect a wiring defect by comparing the current waveforms.

  This inspection can be carried out according to the inspection flow shown in FIG. 26 using the apparatus shown in FIG. 25, similarly to the inspections shown in FIGS. However, it is assumed that the electron gun 112 can generate a linear electron beam.

  32A and 32B are diagrams for explaining an inspection example when the wiring has the same shape in the vertical direction. FIG. 32A shows a measurement example of a non-defective chip, and FIG. 32B shows a measurement example of a defective chip. The irradiated electron beam is a linear electron beam having a width of about 100 mm in the scanning direction and a width of several microns in the lateral direction.

  The current value obtained by the current measurement method which is the measurement principle of the present invention is that the total value of the currents generated by the simultaneously irradiated wires is measured. That is, the total amount of current generated when a thin electron beam is irradiated on each wiring portion is measured by the linear electron beam.

  In the example of FIGS. 32A and 32B, when the electron beam 161 passes through the position a to the position b, similar currents are observed in the wirings 162 and 164 regardless of whether the electron beam 161 is good or defective. On the other hand, when the electron beam 161 passes the position c to the position d, a current through the wiring 163 is observed in the non-defective product, whereas only a small current value through the wiring 165 is observed in the defective product. Therefore, a large current waveform difference occurs, and the presence of the pattern defect 1665 is detected. That is, even when the position of the wiring overlaps with the scanning of the electron beam, the defect can be detected and the defect position can be identified using the inspection apparatus of FIG. 25 and the inspection procedure of FIG.

  FIGS. 33A and 33B are diagrams for explaining an inspection example in the case where wirings having different wiring widths exist in an axial symmetry. FIG. 33A shows a measurement example of a non-defective product, and FIG. 33B shows a measurement example of a defective product. In this inspection example, a linear electron beam similar to that in the inspection example shown in FIG. 32 is used, and the total amount of current values generated when each wiring portion is irradiated with a thin electron beam is measured.

  In the non-defective product shown in FIG. 33A, current is obtained by the wiring 172 when the electron beam 171 reaches the position a, whereas in the defective product shown in FIG. 33B, the wiring 173 has a pattern defect 174. Therefore, even when the electron beam 171 starts to scan the position a, no current is observed and a difference occurs in the current waveform. Even at the position b, since the defective product has the pattern defect 174, the observed current waveform is smaller than that of the non-defective product. As described above, when wirings having different wiring widths exist in axial symmetry, even if they are simultaneously inspected, a difference in waveform occurs between a non-defective chip and a defective chip, so that a defect can be detected.

  FIGS. 34A and 34B are diagrams showing an inspection example when wirings having different widths are scattered, where FIG. 34A shows a measurement example of a non-defective product and FIG. 34B shows a measurement example of a defective product. When scanning is performed using the linear electron beam 181, the current change amount at the position b where the pattern defect 184 is present differs between the non-defective chip wiring 182 and the defective chip wiring 183. By detecting this difference, a defect is detected.

(Comparison of current waveforms by integration)
In the above description, the case where the current waveforms obtained by irradiating the sample with the electron beam are directly compared has been described as an example. Various other methods are conceivable for comparing the current waveforms.

  FIG. 35 shows a configuration example of an inspection apparatus that compares current waveforms based on their integrated values, and FIG. 36 shows an inspection flow thereof. This inspection device includes pulse integrators 191 and 192 and an integral value comparison device arrangement 191 instead of the waveform comparison device 123 of the device shown in FIG. 25, and one pulse of a current waveform obtained corresponding to the wiring inspection. The integrated value of the current value that flowed in is obtained by the pulse integrators 191 and 192, and is compared by the integral value comparator 193.

(Comparison by current value per unit area)
In the comparative inspection, since position coordinates such as CAD are not used, the electron beam used for the inspection is not always irradiated onto the wiring completely. Therefore, the current value per unit area can be used as a reference for comparison.

  FIG. 37 shows a configuration example of an inspection apparatus that performs a comparative inspection based on a current value per unit area, and FIG. This inspection apparatus includes a storage device 201, a pulse integrator 202, a pulse width detector 203, a division device 204, and a storage device 205 instead of the storage devices 121 and 122 and the waveform comparison device 123 in the configuration shown in FIG. Prepare. The measured current waveform is stored in the storage device 201, and the pulse integrator 202 integrates the amount of current flowing from the rising edge to the falling edge of one pulse belonging to the stored current waveform and flows between one pulse. Obtain the total current. The pulse width detector 203 obtains the wiring crossing distance from the pulse width of the current waveform stored in the storage device 201. The dividing device 204 divides the total current amount obtained by the pulse integrator 202 by the wiring crossing distance obtained by the pulse width detector 203. By this division, a current value flowing per unit area that is independent of the wiring width is obtained. The defect determination device 124 compares the value obtained by the division device 204 with a reference value stored in the determination database 125 that is obtained in advance, so that a non-defective product and a defective product are determined based on the difference in size. Judging.

  39A and 39B are diagrams for explaining the relationship between the coverage of the electron beam with respect to the wiring and the current waveform. FIG. 39A shows an example of 100% coverage where the electron beam completely passes through one wiring, and FIG. An example of coverage of 50% passing through an area is shown. When the wiring 212 is completely included in the electron beam scanning zone 211 that is the scanning range of the electron beam, the same current waveform is obtained for each scanning. On the other hand, when the wiring 212 is deviated from the electron beam scanning band 211, a current waveform by the wiring 212 is obtained in a certain scan, but cannot be obtained in another scan. However, the amount of current per unit area flowing through the non-defective contact wiring is constant, and the quality can be determined by comparing them.

  The reference value for defect determination used here is the amount of current per unit area indicated by the non-defective wiring. Therefore, as the value, a value indicated by chip wiring of another wafer that has undergone the same process, data obtained from a test pattern, or a value estimated by performing a simulation or the like is used. The use of a reference value derived from another wafer is effective when the rate of non-defective products inside the wafer at the beginning of process development is extremely low.

(Comparison by current value per unit area in mass production factory)
FIG. 40 shows a configuration example of an inspection apparatus for performing a comparative inspection using a plurality of chips on the same substrate, and FIG. 41 shows an inspection flow thereof. This inspection apparatus has two systems equivalent to the storage device 201, pulse integrator 202, pulse width detector 203, division device 204, and storage device 205 shown in FIG. 37, that is, storage devices 221, 231 and pulse integration. Units 222 and 232, pulse width detectors 223 and 233, division units 224 and 234, and storage units 225 and 235, and further a division unit 236 for dividing the values stored in the storage units 225 and 235. This is different from the apparatus configuration shown in FIG.

  This apparatus is used in a mass production factory where production is stabilized to some extent, and performs a comparative inspection using a plurality of chips on the same substrate. That is, the current waveform obtained by irradiating the first specimen to be inspected with the electron beam is stored in the storage device 221, and the amount of current flowing from the rising edge to the falling edge of one pulse belonging to the measured current waveform is pulsed. The total amount of current flowing during one pulse is obtained by integration by the integrator 222, the width from the rising edge to the falling edge of the pulse waveform equal to the wiring crossing distance is obtained by the pulse width detector 223, and obtained by the pulse integrator 222. The total current amount is divided by the pulse width obtained by the pulse width detector 223 to obtain the amount of current flowing per unit area regardless of the wiring width. This is stored in the storage device 225. The same inspection is performed on the second sample to be inspected, and the obtained current amount per unit area is stored in the storage device 235. The values in the storage devices 225 and 235 are divided by the dividing device 236 and compared with the reference value recorded in advance in the determination database 125 by the failure determining device 124. This reference value defines the allowable difference between chips. If the comparison result is large, there is a defect at that position.

(Comparison of rise and fall of current waveform)
FIG. 42 shows an inspection flow for determining the quality of the wiring based on the rise and fall of the acquired current waveform. In this inspection, the quality of the wiring is determined using the fact that the rising and falling edges of the pulse in the current waveform correspond to the edge of the wiring pattern. That is, the current waveform is acquired by scanning the first test sample, and the wiring position of the first test sample is determined. Next, the second inspection sample is scanned to obtain a current waveform, and the wiring position of the second inspection sample is determined. By comparing the difference between the rise and fall of both waveforms, the positional deviation of the wiring can be understood. If this positional deviation is greater than or equal to a certain value, the wiring is determined to be defective, and the position is stored in the storage device.

  FIG. 43 is a diagram showing an inspection example, where (a) shows a non-defective product and (b) shows a defective product. In the non-defective product, the wiring 241 is formed periodically, and correspondingly, rising and falling of the current waveform are measured periodically at the electron beam irradiation positions T1 to T8. In contrast, in the defective product, the rising position of T3 is shifted from the non-defective product.

  FIG. 44 shows an inspection flow for determining the quality of the wiring based on the center positions of the rising and falling edges of the current waveform. In this inspection, the first inspection sample is scanned to obtain a current waveform, and the center coordinates of the rising and falling position coordinates of the pulse in the current waveform are calculated to identify the center position of the wiring. Similarly, the second inspection sample is scanned to obtain a current waveform, and the center coordinates of the rising and falling position coordinates of the pulse in the current waveform are calculated to identify the center position of the wiring. The wiring center positions of the two are compared, and if there is a certain difference or more in the result, at least one wiring center position is stored in the storage device.

(Electron beam sub-scanning)
FIG. 45 shows an example of an apparatus configuration that dramatically improves the inspection speed when an inspection is performed using a thin electron beam. In this configuration example, a sub-scanning deflecting device 251 that deflects an electron beam is provided, and sub-scanning by the sub-scanning deflecting device 251 is simultaneously performed in addition to main scanning in which the wafer position is moved by the XY stage 114.

  Since the main scanning is performed by the movement of the XY stage 114, it is difficult to stably move at a speed exceeding 1 [m / sec] with the current technology. For this reason, even if the processing speed of the current measurement system is very high, the upper limit of the inspection speed is determined by the electron beam scanning speed. Therefore, at the same time as performing the main scanning so as to substantially improve the scanning speed, the sub scanning is performed at a high speed in the direction perpendicular to the main scanning direction. Since sub-scanning is performed by electron beam deflection, the moving speed of the XY stage is tens of thousands of times faster.

  In the case of a small sub-scanning distance, since the incident angle of the electron beam is substantially vertical and does not affect the inspection, a normal electron beam deflector is used. When performing sub scanning with a large amplitude, a beam shifter that translates the beam is used.

  FIG. 46 shows an example of a scanning locus. While the main scanning 262 slowly proceeds in a certain direction with respect to the wiring 251, the sub-scanning 263 reciprocating at a constant width and at a high speed is simultaneously performed. The sub-scans 263 are performed in parallel at intervals corresponding to the width of the measurement target wiring. In this way, scanning is performed at the same speed as the main scanning speed multiplied by the sub-scanning speed, and the inspection speed can be dramatically improved.

(Acceleration of array area inspection)
FIG. 47 shows an inspection flow for speeding up the inspection. In an SOC device or the like, there is an array region in which contact wirings such as a memory are arranged at equal intervals over a long distance together with random logic. Such an array region is automatically extracted from the inspection sample without requiring layout information from CAD or the like, and the portion is inspected by another high-speed method peculiar to the array. For this purpose, first, the first chip is inspected, and current waveforms in all the inspection target areas are acquired. Next, the position of the wiring appearing in the scanning direction is detected from the rise and fall of the current waveform and stored. Next, the frequency distribution of the spatial distribution of the position of the wiring is analyzed for each specific section (for example, several tens to several hundreds of microns).

  FIG. 48 shows an example of a power spectrum obtained by frequency analysis. This power spectrum has position dependency. A region with high power corresponds to a strong correlation between current waveforms, and the presence of an array in that region is detected. Conversely, a region having a small correlation is considered as a random logic region.

  The array portion detected in this manner is irradiated with an electron beam simultaneously on a plurality of wirings, and the defect ratio is obtained collectively. This speeds up the inspection.

(Measurement of three-dimensional shape)
According to the present invention, not only the diameter of the bottom of the hole but also the three-dimensional shape of the hole can be measured. That is, it is utilized that the electron beam intensity and distribution irradiated to the hole bottom are changed by changing the acceleration voltage of the electron beam and further changing the tilt of the wafer. This will be described with reference to FIGS. 49 and 50. FIG. When the acceleration voltage of the electron beam irradiating the hole 510 is low, as shown in FIG. 49, the electrons hardly pass through the insulating film 512, and the bottom part of the hole 510, that is, the part where the wafer 511 is exposed. Others contribute little to the measured current. When the acceleration voltage of the irradiating electron beam is increased, electrons pass through the insulating film 512 around the bottom of the hole 510 as shown in FIG. 50, and the measured current value changes. By utilizing this, the hole edge or the insulating film thickness can be measured.

  The hole shapes shown in FIGS. 49 and 50 are spread upward, but similar measurement results can be obtained when the holes are spread downward as shown in FIGS. In this case, it is not possible to distinguish whether the hole shape is upward or downward by this measurement alone. Therefore, as shown in FIGS. 53 and 54, the measurement is repeated while changing the tilt angle of the wafer. It is possible to distinguish whether the hole shape is upward or downward from the change in the intensity distribution of the wafer current detected by the angle.

  In order to obtain the three-dimensional shape of the hole, the dependence of the electron beam absorption coefficient on the acceleration voltage of the electron beam is obtained in advance for each material constituting the sample to be measured and stored as a library.

As a method for reconstructing a three-dimensional shape as an image from the measured current value, a Fourier transform method, a successive approximation method, and a superposition integration method can be considered. Of these, the successive approximation method will be described with reference to FIGS. 55 to 58. FIG. 55 shows a processing flow, and FIGS. 56 to 58 are diagrams for explaining individual processing.
1) First, as shown in FIG. 56, it is assumed that the two-dimensional image to be measured is decomposed into (M × N) pixels, and a uniform absorption coefficient is given to each pixel with an appropriate initial value.
2) Next, the sum of the absorption coefficients c _mn of the cells in the locus irradiated with the electron beam is taken. However, the total value is the measured substrate current value and I = I ₀ · exp [−Σc _mn ] (1)
The relationship is established. The absorption coefficient c _mn of the corresponding cell is modified so that the formula (1) is satisfied.
3) The operation 2) is sequentially performed by sequentially changing the irradiation angle Θ of the electron beam. That is, the absorption coefficient c _mn of each cell is sequentially modified so that the formula (1) always holds under any measurement condition (irradiation angle Θ, acceleration voltage E).
4) The operations of 2) and 3) are repeated while changing the acceleration voltage E of the electron beam one after another, and an absorption coefficient map for each acceleration voltage is obtained approximately as shown in FIG.
5) As shown in FIG. 58, the acceleration voltage dependence of the absorption coefficient is compared with the data on the library for each cell.
6) Thereby, a qualitative three-dimensional image of the sample to be measured is obtained.

  In the above image reconstruction method, the resolution is determined by the probe diameter of the electron beam, the spot interval of the electron beam, and the cell size in the successive approximation method. The accuracy of qualitative analysis is determined by the electron beam acceleration voltage swing interval, swing width, and substrate ammeter sensitivity.

(Deviation inspection)
In the present invention, it is also possible to detect non-destructive deviations (intermittent deviations) generated between layers by utilizing the fact that an electron beam passes through an insulating layer. That is, the acceleration voltage is sequentially increased, and the structure of the lower layer is obtained by irradiating the diffusion layer and wiring with an electron beam through the insulating film, and information on the upper diffusion layer and wiring is also obtained. The shift between the hole position and the structure of the lower layer can be detected from the information of different layers obtained simultaneously, and the misalignment can be evaluated. Further, by changing the acceleration voltage to change the arrival depth of the electron beam, for example, between the second and third layers from the surface, between the third and fourth layers, or between the second and fourth layers. It can also evaluate the gap between eyes. When measuring the lower layer by increasing the acceleration voltage, the upper layer information is included in the lower layer information, which can be separated by image processing. In addition, when a conductive layer such as a wiring is arranged in the upper layer without electrical connection with the substrate, it can be detected as a negative image when measuring the lower layer.

  FIG. 59 shows an example of misalignment evaluation, (a) shows a cross-sectional view of the device, and (b) shows a current image of a measurement result. In this example, an insulating layer is provided on a wafer provided with a diffusion layer, and a part of the diffusion layer is exposed by a hole provided in the insulating layer. When the acceleration voltage is low and the electron beam cannot pass through the insulating film, the position of the hole can be found from the irradiation position of the electron beam and the measured value of the substrate current. When the acceleration voltage is increased to such an extent that it can pass through the insulating film, the position of the entire diffusion layer can be found in the same manner due to the difference in impurities of the semiconductor substrate. The shift can be evaluated by evaluating the shift of the center position between the hole and the diffusion layer or the distance between the outer periphery of the hole and the diffusion layer.

  FIGS. 60 and 61 are diagrams showing another example of misalignment evaluation. FIG. 60 is a cross-sectional view of a device when there is no misalignment, and FIG. ) Shows the current image of the measurement results. In this example, wiring is provided on the surface of the wafer, an insulating layer is provided thereon, and holes are provided in this insulating layer. The position of the wiring and the position of the hole are originally deviated, but the deviation is large in the example of FIG. Also in this case, as in the example shown in FIG. 59, the position of the wiring can be determined by the electron beam of the low acceleration voltage, and the position of the wiring can be determined by the electron beam of the high acceleration voltage. Can be evaluated.

  In the example of FIG. 62, a diffusion layer is provided on the surface of the wafer, a lower layer wiring is provided thereon via a first insulating film, and an upper layer wiring is provided thereon via a second insulating film, Furthermore, a third insulating film is provided thereon. The position of each layer can be determined by sequentially measuring the substrate current value while changing the acceleration voltage of the electron beam.

  In order to detect misalignment, it is necessary to perform measurement by adjusting the acceleration voltage so that the electron beam reaches a desired layer. FIG. 63 shows a measurement flow when the main insulating layer is made of one kind of material. First, based on CAD data, a place where the wirings or diffusion layers of each layer do not overlap is selected, and a necessary magnification is determined. The overlapping location cannot be observed, the structure cannot be seen if the magnification is too low, and the location where there is no wiring or diffusion layer may be wasted if the magnification is too high. It is desirable to determine from design data. This also facilitates the determination process. Subsequently, the process data of each layer is read, the thickness of the insulating layer up to the lower end of each layer is calculated, the acceleration voltage corresponding to the thickness of the insulating layer is read from the database, and the compensation current is measured with the acceleration voltage. This measurement needs as many as the number of layers.

  If the material of the insulating layer is one type, the behavior with respect to the electron beam is the same. Even when there are a plurality of types of insulating layers, they can be treated as the same material if they have the same physical behavior with respect to the electron beam, such as the amount of secondary electron emission. Each material is stored in a database by prior measurement and is automatically determined when reading process data.

  The data to be prepared before adjusting the acceleration voltage include the current value detected by the compensation current or wiring for the insulation layer type and thickness, and the compensation current or wiring for the acceleration voltage for each insulation layer type and thickness. Current value to be detected. These data are measured in advance and recorded in a database. An example of the compensation current for the film thickness is shown in FIG. 64, and an example of the compensation current for the acceleration voltage is shown in FIG.

  FIG. 66 shows a measurement flow when there are a plurality of insulating layers. Also in this case, first, a place where wirings or diffusion layers of each layer do not overlap is selected based on CAD data, and a magnification suitable for the range of the measurement region is determined. Subsequently, based on the process data, the thicknesses of a plurality of types of insulating layers up to each layer are calculated individually, and a search is made as to whether there is a setting that matches the combination in the database. If there is a matching setting, the acceleration voltage corresponding to the total thickness of the plurality of types of insulating layers is read from the database, and the compensation current is measured with the acceleration voltage. If there is no matching setting, the total insulation layer thickness is calculated based on the process data, and the material that is most difficult to transmit the electron beam is assumed. To find the correct acceleration voltage. Next, the compensation current is measured at a low acceleration of about 500 V and imaged. With such a low acceleration voltage, only the surface layer is visible. Next, the acceleration voltage that can be transmitted to the lowest layer obtained above is divided by “number of layers × n”, and the compensation current is measured and imaged for each acceleration voltage obtained. As the value of n, an optimum value from 1 to 9 is used. The lower layer image obtained at this time includes upper layer information. The obtained images are compared, and if there is a matching image, the acceleration voltage is further finely changed before and after that and remeasured. If the 2nd and 3rd images match, the measurement will be made with an intermediate voltage between the acceleration voltage when the 1st and 2nd images are taken, and with an intermediate voltage when the 3rd and 4th images are taken. Continue to operate until Then, the measurement is terminated when an image different from the process data by the number of layers is obtained.

  FIG. 67 shows a flow of misalignment determination after an image for each layer is obtained. The image of each layer (pattern visible in the current image) is compared with the layout information of the CAD data, and what each image corresponds to on the CAD data, that is, which wiring or which diffusion layer is determined. Next, the coordinate position of the obtained pattern designated at the time of design by CAD data is investigated, and the distance in the projected image from the upper surface is calculated. The ideal value obtained by this calculation is compared with the actually measured value obtained from the image. The shift at this time corresponds to the misalignment.

  Instead of obtaining information by separating the layers by changing the acceleration voltage of the electron beam, it is also possible to obtain the information collectively. The measurement flow in that case is shown in FIG. In this measurement flow, first, similarly to the measurement flow shown in FIGS. 65 and 66, a place where wirings or diffusion layers of each layer do not overlap is selected based on CAD data, and a necessary magnification is determined. Subsequently, the total insulation layer thickness is calculated based on the process data, assuming the material that is most difficult to transmit the electron beam in use, and even in that case, the acceleration voltage that can be transmitted to the lowest layer is obtained. The current image is acquired with the acceleration voltage. The pattern of each layer that seems to contribute to the current image is obtained from CAD data and compared with the measured current image. At this time, the information obtained from the CAD data is used to separate which layer the current image belongs to, and the position coordinates in the current image are compared with the ideal state obtained from the CAD data to obtain a deviation. Although it is necessary to classify current images based on CAD data, it is considered that accuracy can be improved because deviation can be evaluated with a single image.

(Background correction) In the above inspection, the current generated in the substrate by scanning and irradiating the electron beam onto the sample surface is recorded as a function of the electron beam scanning position, and this is used as a luminance signal for image display. A current image corresponding to the substrate surface is formed. Further, when the image is used for the contact hole inspection, the magnitude of the current that flows in a direct current through the contact hole is a criterion for the quality determination. However, in practice, an alternating current component is generated because the electron beam is periodically irradiated in a pulsed manner or the electron beam is scanned. For this reason, the current that is measured includes a component that occurs capacitively in addition to the component that actually flows in a DC manner. When such a component is present, the physical correspondence between the brightness of the image and the object cannot be obtained, the contact hole quality determination becomes inaccurate, and the restoration of the three-dimensional shape becomes difficult.

  Therefore, it is desirable to measure the current by changing the irradiation frequency or scanning frequency of the electron beam and correct the current component flowing through the capacitance of the sample to be inspected. Two processing flow examples for performing such correction are shown in FIGS. 69 and 70, respectively.

  In the processing flow shown in FIG. 69, when the pulsed electron beam is repeatedly irradiated, the measurement is repeated while changing the repetition period, and the value when the electron beam is continuously irradiated from the obtained current waveform. To obtain the DC component. Referring to the apparatus configuration shown in FIG. 1 as an example, the electron gun 1 is configured to repeatedly generate a pulsed electron beam. The beam control unit 11 changes the repetition period of the electron beam, and the data processing apparatus 10 The direct current component is obtained by extrapolating the value when the electron beam is continuously irradiated from the current values measured when the electron beam is irradiated at different repetition periods.

  In the processing flow shown in FIG. 70, the measurement is repeated while switching the scanning speed of the electron beam, and a value when the scanning speed is extrapolated to zero is obtained from the obtained current waveform. Referring to the apparatus configuration shown in FIG. 45 as an example, the electron beam irradiation position control device 116 can switch the scanning speed of the electron beam via the sub-scanning deflection device 251, and the output of the D / A converter 120 is In the supplied data processing apparatus (for example, block 10 in FIG. 1), a value when the scanning speed is extrapolated to zero is obtained from current values respectively measured when the electron beam is scanned at different scanning speeds.

It is a block block diagram which shows 1st embodiment of this invention. It is a block block diagram which shows 2nd embodiment of this invention. It is a figure which shows the structural example of an aperture, (a) shows the aperture for making the cross-sectional shape of a beam circular, (b) shows the aperture for making a square. It is a figure explaining the example of a test | inspection of the column-shaped contact hole using SEM, (a) is an inspection method, (b) shows the example of a test result. It is a figure explaining the example of a test | inspection of the taper-shaped contact hole using SEM, (a) is an inspection method, (b) shows the example of a test result. It is a figure explaining a measuring method, (a) shows the structure of the contact hole used as an object, and its measuring system, and (b) shows the example of a measurement result. It is a figure explaining the measurement with respect to a taper-shaped contact hole, (a) shows the structure and measuring system of the contact hole used as object, (b) shows the example of a measurement result. The change of the compensation current amount with respect to the contact hole bottom area is shown. The change in compensation current with respect to the diameter of the contact hole bottom is shown. It is a figure explaining the measurement using the electron beam which has a larger cross-sectional shape than a hole opening part, (a) shows the structure and measurement system of the contact hole used as object, (b) shows the example of a measurement result. It is a figure explaining the measurement using the electron beam which has a larger cross-sectional shape than a hole opening part, (a) shows the structure and measurement system of the contact hole used as object, (b) shows the example of a measurement result. It is a figure explaining the same measurement using the beam diameter smaller than the diameter of a contact hole, (a) shows the structure of the target contact hole and its measuring system, (b) shows the example of a measurement result. The flowchart in the case of utilizing the measurement of a contact hole bottom diameter in a mass production factory and the example of a quality determination are shown. It is a figure explaining the example of a connection measurement with SEM about the contact hole which has a cylindrical shape, (a) shows the structure of the target contact hole, and its measurement system, (b) is along a contact hole center line. The measured amount of secondary electrons and the compensation current are shown with respect to the irradiation position of the electron beam, and (c) shows the restored three-dimensional display figure. It is a figure explaining the same measurement about the contact hole which has a forward taper shape. It is a figure explaining the same measurement about the cylindrical contact hole by a gradient electron beam. It is a figure explaining the same measurement about a forward taper-shaped contact hole. It is a figure explaining the same measurement about a reverse taper-shaped contact hole. It is a figure explaining the method to detect and identify the foreign material which arose in the contact hole, (a) shows the structure and measuring system of a test object, (b) shows the amount of secondary electrons to be measured and the compensation current. It shows corresponding to the irradiation position of the electron beam. It is the same figure explaining the method to detect and identify the foreign material which arose inside the contact hole. It is the same figure explaining the method to detect and identify the foreign material which arose inside the contact hole. It is a figure which shows the example of a measurement using an electron beam with a big cross-sectional shape, (a) is a top view which shows the positional relationship of a contact hole and an electron beam, (b) is sectional drawing, (c) is scanning of an electron beam. The compensation current obtained with respect to the position and its differential value are shown. The flowchart of the measuring method which combined length measuring mode and batch measurement mode is shown. An example of the positional relationship between the length measurement mode target area and the batch measurement mode target area on the wafer is shown. An apparatus configuration example for performing a comparative inspection using two specimens to be inspected will be shown. The inspection flow is shown. It is a figure explaining the principle of a comparative test. It is an enlarged view. It is a figure which shows the example of a test | inspection, (a) shows the measurement example of a non-defective chip, (b) shows the measurement example of a defective chip. It is a figure explaining the comparative test example using a thin electron beam, (a) shows the measurement example of a non-defective chip, (b) shows the measurement example of a defective chip. It is a figure explaining the comparative test example which irradiates a linear electron beam simultaneously to several wiring arrange | positioned at random, (a) shows the measurement example of a non-defective chip, (b) shows the measurement example of a defective chip. It is a figure explaining the example of a test | inspection in case wiring has the same shape in the vertical direction, (a) shows the measurement example of a non-defective chip, (b) shows the measurement example of a defective chip. It is a figure explaining the example of a test | inspection when the wiring from which wiring width differs exists in an axial symmetry, (a) shows the measurement example of a non-defective product, (b) shows the measurement example of a defective product. It is a figure which shows the example of a test | inspection in case the wiring from which width differs differs, (a) shows the measurement example of a good product, (b) shows the measurement example of a defective product. The structural example of the test | inspection apparatus which compares an electric current waveform with the integral value is shown. The inspection flow is shown. The structural example of the test | inspection apparatus which performs a comparative test | inspection with the electric current value per unit area is shown. The inspection flow is shown. It is a figure explaining the relationship between the coverage of the electron beam with respect to wiring, and a current waveform, (a) Example of coverage 100%, FIG. (B) shows the example of coverage 50%. The structural example of the test | inspection apparatus for performing a comparative test | inspection using the some chip | tip on the same board | substrate is shown. The inspection flow is shown. The inspection flow which determines the quality of wiring by the rising and falling of a current waveform is shown. It is a figure which shows the example of an inspection, and shows a good article and inferior goods. An inspection flow for determining the quality of the wiring based on the center position of the rise and fall of the current waveform is shown. An example of an apparatus configuration for performing sub-scanning of an electron beam is shown. An example of a scanning locus is shown. An inspection flow for speeding up the inspection of the array area is shown. An example of the power spectrum obtained by frequency analysis is shown. It is a figure explaining the measurement of the three-dimensional shape of a hole. It is a figure explaining the measurement of the three-dimensional shape of a hole. It is a figure explaining the measurement of the three-dimensional shape of a hole. It is a figure explaining the measurement of the three-dimensional shape of a hole. It is a figure explaining the measurement of the three-dimensional shape of a hole. It is a figure explaining the measurement of the three-dimensional shape of a hole. The processing flow for calculating | requiring a three-dimensional shape by a successive approximation method is shown. It is a figure explaining a process. It is a figure explaining a process. It is a figure explaining a process. An example of misalignment evaluation is shown, (a) is a sectional view of the device, and (b) shows a measurement result. Another example of misalignment evaluation is shown, (a) is a cross-sectional view of a device when there is no misalignment, and (b) shows a measurement result. 60 shows an example in which there is a misalignment in an element equivalent to FIG. 60, (a) shows a cross-sectional view of the device, and (b) shows a measurement result. Another example of misalignment evaluation is shown, (a) is a cross-sectional view of the device, and (b) shows a measurement result. The measurement flow when the main insulating layer is made of one kind of material is shown. An example of compensation current for film thickness is shown. An example of the compensation current with respect to the acceleration voltage is shown. The measurement flow when there are a plurality of insulating layers is shown. The flow of misalignment judgment after the image for every layer is obtained is shown. The measurement flow for acquiring the information of multiple layers collectively is shown. The processing flow of an example of background correction is shown. The processing flow of another example of background correction is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Electron gun 2, 12 Electron beam 3, 13, 15 Condenser lens 4, 14 Aperture 5 Sample 6 Movable stage 7 Electrode 8 Moving distance measuring device 9 Ammeter 10 Data processing device 11 Beam control part 16 Objective lens

Claims (15)

  The electron beam irradiation means for irradiating the semiconductor device of the sample to be inspected while scanning the electron beam, the current measuring means for measuring the current generated in the sample to be inspected by the electron beam irradiation, and the measurement result of the current measuring means In the semiconductor device inspection apparatus comprising data processing means for processing data, the electron beam irradiation means includes collimating means for collimating the electron beam and means for changing the acceleration voltage of the electron beam, and the data processing means Is a means for obtaining information on the structure in the depth direction of the sample to be inspected from the difference in current value generated in the sample based on the difference in transmittance of the electron beam with respect to the sample to be inspected when the electron beam is scanned with different acceleration voltages. A semiconductor device inspection apparatus comprising:   The electron beam irradiating means includes an electron gun, and the collimating means limits a spot size of the electron beam hitting the semiconductor device to be inspected and a condenser lens for collimating the electron beam emitted from the electron gun. Means for moving the sample to be inspected with respect to the electron beam for scanning the electron beam, including an aperture inserted between the condenser lens and the sample to be inspected at a right angle so that the electron beam hits the opening. The semiconductor device inspection apparatus according to claim 1, further comprising:   The electron beam irradiation means includes an electron gun, and the collimating means includes a first condenser lens that collimates the electron beam emitted from the electron gun and a second condenser arranged to form an afocal system. A lens, an objective lens, and an aperture that is inserted between the first condenser lens and the second condenser lens to limit the spot size of the electron beam, and for scanning the electron beam with respect to the electron beam 2. The semiconductor device inspection apparatus according to claim 1, further comprising means for moving the sample to be inspected.   The electron beam irradiation means includes means for vertically irradiating a sample to be inspected along a line segment passing through the center of the measurement area with an electron beam having a spot size smaller than the area of the measurement area. 2. The semiconductor device inspection apparatus according to claim 1, further comprising means for obtaining a bottom distance of the measurement region from an interval between rising and falling of the current measured along the line segment.   A secondary electron detector for detecting secondary electrons emitted from the surface of the sample to be inspected is provided, and the data processing means measures the amount of secondary electrons measured by the secondary electron detector by the current measuring means. 2. The semiconductor device inspection apparatus according to claim 1, further comprising means for processing corresponding to the result.   The electron beam irradiation means includes means for vertically irradiating a sample to be inspected along a line segment passing through the center of the measurement region with an electron beam having a spot size smaller than the measurement region area. Means for determining the bottom distance of the measurement region from the rising and falling intervals of the current measured along the line segment by the current measuring means, and the amount of secondary electrons detected by the secondary electron detector. 6. The semiconductor device inspection apparatus according to claim 5, further comprising means for obtaining an upper distance of the measurement region from a rising and falling interval.   2. The apparatus according to claim 1, further comprising means for tilting a sample stage on which the specimen to be inspected is mounted, wherein the data processing means includes means for taking in and processing the tilt angle of the specimen to be inspected with respect to the electron beam generated by the tilting means. Semiconductor device inspection equipment.   The electron beam irradiation means includes means for setting the cross-sectional shape of the electron beam so that the entire measurement region can be irradiated at once, and at least one end thereof is linear, and the data processing means 2. The semiconductor device inspection apparatus according to claim 1, further comprising means for obtaining a distance of the measurement region from an interval from the rising edge to the maximum value.   The electron beam irradiation means includes means for setting the cross-sectional shape of the electron beam so that the entire measurement region can be irradiated at once, and at least one end thereof is linear, and the data processing means 2. The semiconductor device inspection apparatus according to claim 1, further comprising: means for calculating a differential curve with respect to the distance; and means for determining a radius of the measurement region from an interval from a rising edge of the differential curve to a position indicating the apex.   2. The semiconductor device inspection apparatus according to claim 1, wherein the data processing means includes means for displaying the measured current value on a map corresponding to the measurement position.   The data processing means, each of the two areas on the wafer as a sample to be inspected, a comparison means for comparing the measured value obtained in the first area with the measured value obtained in the second area as a reference value, 2. The semiconductor device inspection apparatus according to claim 1, further comprising means for extracting the position coordinates when the comparison result has a predetermined difference or more.   The comparing means is means for integrating the current flowing from the rising edge to the falling edge of one pulse of the current waveform measured as a change in the current value generated in the sample as the electron beam is irradiated to the electron beam irradiation position; A dividing means for dividing the obtained integrated value by the width from the rising edge to the falling edge of the pulse, and a means for comparing the current value per unit length obtained by the dividing means for the two regions. The semiconductor device inspection apparatus according to claim 11.   12. The semiconductor device inspection apparatus according to claim 11, wherein the comparison means includes means for comparing the rising and falling positions of a pulse of a current waveform measured as a change in current value with respect to the irradiation position of the electron beam.   The comparison means includes a center position between a rising position and a falling position of the current waveform pulse of the first inspection sample measured as a change in current value with respect to the irradiation position of the electron beam, and a current waveform pulse of the second inspection sample. 12. The semiconductor device inspection apparatus according to claim 11, further comprising means for comparing the center position between the rising position and the falling position of the semiconductor device. The electron beam irradiation means includes a main scanning means for moving the specimen to be inspected with respect to the electron beam, and a sub-scanning means for repeatedly deflecting the electron beam in a direction different from the main scanning direction, superimposed on the main scanning by the main scanning means. The semiconductor device inspection apparatus according to claim 1, further comprising a scanning unit.

Images