JP2014077654A - Inspection device and inspection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an inspection method that can measure a charge collection process of electrons excited in a semiconductor specimen irradiated with radiation.SOLUTION: The inspection method is configured to include: a first step of irradiating a semiconductor specimen M with the radiation; a second step of standing by during a first period of time since the irradiation of the semiconductor specimen M with the radiation is started; a third step of measuring excitation of the electrons in the semiconductor specimen M during a second period of time after the second step; and a fourth step of modifying the first period of time. The inspection method is configured to repeat the first step, the second step and the third step.

Description

  The present invention relates to an inspection apparatus and an inspection method.

  Conventionally, inspection apparatuses and inspection methods for inspecting semiconductor devices and the like have been used.

  As the degree of integration of circuit elements forming a semiconductor device increases, the dimensions of the circuit elements decrease and the operating voltage of the circuit elements decreases. For this reason, radiation such as α rays, electron beams, or electromagnetic waves collides with circuit elements, thereby increasing the possibility of causing soft errors in the semiconductor device.

  A soft error is a phenomenon in which a memory or a logic circuit in a semiconductor device temporarily malfunctions due to electron-hole pairs generated by radiation passing through circuit elements forming the semiconductor device. If bit information is destroyed due to the occurrence of a soft error, the semiconductor device may malfunction.

  The frequency of occurrence of soft errors is low when the size of the semiconductor device is large. Therefore, conventionally, problems due to the occurrence of soft errors were not great except for semiconductor devices used in artificial satellites or space stations. However, with the recent miniaturization of circuit elements, the occurrence of soft errors has become a problem in semiconductor devices used on the ground.

  The causes of soft errors can be broadly divided into the following two types. The first is called single event upset (SEU), a phenomenon in which voltage fluctuation occurs when radiation such as charged particles passes through a data holding cell such as a memory, which causes bit inversion. It is. The second is called single event transient (SET), and voltage pulses generated by radiation incident on a logic cell such as a NOT or NAND circuit propagate through the circuit and enter the flip-flop. It is a phenomenon that latches and stores wrong data. The frequency of occurrence of these phenomena was low in circuit elements with long gate lengths. However, with the recent miniaturization of circuit elements, the threshold of the amount of charge that causes bit inversion relatively decreases, and it also affects the logic cell. The range that affects is increased by relatively expanding.

  In the single event described above, as shown in FIG. 1, radiation such as charged particles enters the semiconductor device, and the kinetic energy of the charged particles is lost due to ionization, so that electrons and holes follow the track. This is a phenomenon where electric charges are generated. This charge is collected by a circuit element such as a transistor or a capacitor by a charge collection process such as diffusion, drift (absorption by a depletion layer electric field) or funneling (collection of electric charge by a funnel-shaped electric field). If the collected pulsed charge amount is larger than a predetermined threshold value, the transistor is turned on, the switch is turned on, and the information in the bit is inverted. Similarly, if a voltage signal generated by the generation of a pulsed charge amount propagates through the circuit, it leads to erroneous bit information storage.

JP 2000-469202 A

  In order to design a semiconductor device that is resistant to soft errors, it is desirable to examine the charge collection process, which is the process of the single event phenomenon described above. This is because, by obtaining information on the charge collection process of the circuit element, it is possible to design a circuit element structure that is resistant to each charge collection process such as drift, funneling, or diffusion.

  Accordingly, the present specification aims to provide an inspection method capable of measuring the charge collection process of electrons excited in a semiconductor sample irradiated with radiation.

  It is another object of the present specification to provide an inspection apparatus capable of measuring the charge collection process of electrons excited in a semiconductor sample irradiated with radiation.

  According to one aspect of the inspection method disclosed in the present specification, a first step of irradiating a semiconductor sample with radiation, and a first step of waiting for a first time from the start of irradiation of the semiconductor sample with radiation. 2 steps after the second step, a third step of measuring excitation of electrons in the semiconductor sample for a second time, and a fourth step of changing the first time, The first step, the second step, and the third step are repeated.

  Moreover, according to one form of the test | inspection apparatus disclosed in this specification, when the 1st time has passed since the radiation | emission part which irradiates a semiconductor sample with radiation, and the said radiation | emission part started irradiation of the radiation A measurement signal generating unit that outputs a measurement signal, a measurement unit that inputs the measurement signal and measures excitation of electrons in the semiconductor sample for a second time, and a control that changes the first time A section.

  According to one embodiment of the inspection method disclosed in this specification, the charge collection process of electrons excited in a semiconductor sample irradiated with radiation can be measured.

  Further, according to one mode of the inspection method disclosed in this specification, it is possible to measure the charge collection process of electrons excited in a semiconductor sample irradiated with radiation.

  The objects and advantages of the invention will be realized and obtained by means of the elements and combinations particularly pointed out in the appended claims.

  Both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention as claimed.

It is a figure explaining that the radiation which injected into the semiconductor excites an electron. It is a figure showing a 1st embodiment of an inspection device indicated to this specification. It is a figure explaining light emission when the excited electron returns to a ground state. It is a flowchart explaining one Embodiment of the test | inspection method disclosed by this specification. It is a figure explaining the temporal relationship between a pulse-shaped electron beam and measurement of light emission. It is a figure explaining the arrangement | sequence where a measurement result is stored. (A) shows a scanning electron microscope image, (B) shows an imaged measurement result image, and (C) shows an image obtained by superimposing the scanning electron microscope image and the measurement result image. Is shown. It is a figure which shows the relationship between the generation frequency of a soft error, and the energy of an electron beam. It is a figure which shows the relationship between the occurrence frequency of a soft error, and an electron dose. It is a figure which shows the formula of stopping power. It is a figure which shows the relationship between an inversion cross-sectional area and LET. It is a figure explaining other temporal relationships between a pulsed electron beam and measurement of light emission. It is a flowchart explaining other embodiment of the test | inspection method disclosed to this specification. It is a figure which shows 2nd Embodiment of the test | inspection apparatus disclosed by this specification. It is a figure which shows 3rd Embodiment of the test | inspection apparatus disclosed by this specification.

  Hereinafter, a preferred first embodiment of an inspection apparatus disclosed in this specification will be described with reference to the drawings. However, the technical scope of the present invention is not limited to these embodiments, but extends to the invention described in the claims and equivalents thereof.

  The present specification proposes an inspection method and an inspection apparatus capable of examining the charge collection process of a single event phenomenon that can cause the soft error described above.

  In the charge collection process described with reference to FIG. 1, it is considered that the charge collection process by drift or funneling occurs from 50 picoseconds to several hundred picoseconds from the start of radiation irradiation. In addition, it is considered that the charge collection process by the diffusion of carriers in the vicinity of the depletion layer occurs in 1 nanosecond to several microseconds from the start of radiation irradiation.

  In view of this, the present specification proposes an inspection method and an inspection apparatus for examining the charge collection process by utilizing the fact that the time during which each charge collection process occurs after irradiation with radiation is different. Specifically, information on the occurrence probability of the charge collection process of the semiconductor sample is obtained by measuring the change over time of the excitation of electrons in the semiconductor sample irradiated with radiation.

  FIG. 2 is a diagram illustrating a first embodiment of the inspection apparatus disclosed in this specification.

  In the inspection apparatus 1 of the present embodiment, the radiation unit 10 that irradiates the semiconductor sample M with radiation, and the first time (hereinafter also referred to as delay time) has elapsed since the radiation unit 10 started irradiating radiation. And a measurement signal generation unit 20 that outputs the measurement signal X1. In addition, the inspection apparatus 10 receives the measurement signal X1 and delays the first electric signal measurement unit 30a that measures excitation of electrons in the semiconductor sample for a second time (hereinafter also referred to as measurement time), and a delay. And a control unit 40 for changing the time.

  The semiconductor sample M may be a silicon semiconductor or a compound semiconductor as long as it has a valence band and a conduction band. The semiconductor sample M is preferably a semiconductor having a pn junction. For example, the semiconductor sample M may be a semiconductor device having circuit elements such as transistors. When a semiconductor device is used as the semiconductor sample M, radiation may be applied from the top surface, the bottom surface, or the cross-sectional direction of the semiconductor device.

  When radiation is emitted to the semiconductor sample M, which is a semiconductor device, light emission or voltage fluctuation occurs due to electron-hole pairs generated by the radiation passing through the circuit elements. Therefore, the excitation of electrons in the semiconductor sample M can be measured by measuring the light emission or voltage fluctuation of the semiconductor sample M.

  Therefore, the inspection apparatus 1 measures light emission based on excitation of electrons in the semiconductor sample M.

  As a mechanism by which a semiconductor device having a pn junction irradiated with radiation emits light, for example, there are two phenomena described below.

  The first is a phenomenon in which excited electrons emit light by emitting photons when they recombine at a pn junction.

  FIG. 3 is a diagram for explaining light emission when excited electrons return to the ground state.

  The semiconductor sample has a pn junction where a p-type semiconductor and an n-type semiconductor are joined. A forward bias is applied to the p-type semiconductor and the n-type semiconductor from the outside. In a semiconductor sample to which a forward bias is applied, the diffusion potential of the depletion layer formed at the pn junction decreases and the diffusion current increases. That is, electrons in the conduction band move from the n region to the p region, and holes in the valence band move from the p region to the n region. As a result, in the vicinity of the pn junction, the probability that electrons and holes are recombined increases, and photons having a wavelength corresponding to the energy gap are emitted. In the case of a silicon semiconductor, photons having an energy of about 1.12 eV are emitted.

  The second is a phenomenon in which a semiconductor device having an inverter formed of a CMOS transistor emits light by a through current flowing at an intermediate potential or a protection circuit at the time of reverse bias.

  In the former case, when the PMOS and NMOS forming the CMOS inverter are both turned on due to a soft error at an intermediate potential that is neither high level nor low level with respect to the power supply voltage, a through current that flows through from the power supply voltage to the ground Occurs. When this through current is generated, majority carriers actively move near the gate of each CMOS transistor to emit light.

  In the latter case, in a state where the transistors are connected in reverse bias (protection circuit), when a voltage higher than the power supply voltage is input due to a soft error, breakdown occurs and heat is generated, and carriers are activated to emit light.

  Hereinafter, the inspection apparatus 10 will be described in more detail.

  The radiation unit 10 includes an electron gun 11, a scanning unit 12 having a scanning coil 13a and a scanning control unit 13b, a stage 14, a light detection probe 15, electric signal probes 16a to 16d, a probe connector 16e, and a housing. 17.

  The radiation unit 10 can use an electron beam, an ion beam, a neutron beam, an electromagnetic wave, or the like as the radiation to be irradiated to the semiconductor sample M. In the present embodiment, the radiation unit 10 includes an electron gun 11 that irradiates an electron beam as radiation.

  The electron gun 11 irradiates the semiconductor sample M with an electron beam R having an incident energy equal to or higher than a predetermined value. Here, the incident energy equal to or higher than a predetermined value is energy that can cause excitation of an amount of electrons that can change a conduction state of a switching element such as a transistor included in the semiconductor sample M that is a semiconductor device. Alternatively, the incident energy equal to or higher than a predetermined value is energy that can excite electrons in the valence band of the semiconductor sample M to the conduction band.

  The scanning unit 12 changes the position where the semiconductor sample M is irradiated with radiation. The scanning control unit 13b controls the scanning coil 13a to change the direction of the electron beam irradiated from the electron gun 11 to the semiconductor sample M.

  The scanning control unit 13b controls the scanning coil 13a so that the electron beam does not face the semiconductor sample M except when the semiconductor sample M is irradiated with the electron beam, and the semiconductor sample M is pulsed. The electron beam is irradiated. The pulse width of the pulsed electron beam can be set as appropriate, and can be set to 20 nanoseconds, for example.

  When a radiation signal described later is input, the scanning control unit 13b controls the scanning coil 13a to irradiate the semiconductor sample M with a pulsed electron beam.

  Note that a radiation signal to be described later may be input to the electron gun 11 and irradiated with a pulsed electron beam.

  The stage 14 places the semiconductor sample M thereon.

  The light detection probe 15 receives light emitted from the semiconductor sample M and outputs the received light to the second light detection unit 31. The light detection probe 15 is arranged so as to receive light emitted from any part of the semiconductor sample M. The light detection probe 15 is formed using a photodiode, for example.

  The second light detection unit 31 converts the light received from the light detection probe 15 into an electric signal, and outputs the converted electric signal to a first electric signal measurement unit 30a described later. The light detection probe 15 and the second light detection unit 31 are connected by an optical fiber.

  The electric signal probes 16a to 16d apply a voltage to the semiconductor sample M that is a semiconductor device. A voltage is supplied to the electric signal probes 16a to 16d from the second electric signal measuring unit 30b described later via the probe connector 16e. Thus, the measurement of the excitation of the electrons in the semiconductor sample M irradiated with the electron beam can be performed in a state where the voltage is applied from the electric signal probes 16a to 16d. The electrical signal probes 16a to 16d and the probe connector 16e are connected by wiring (not shown).

  The second electric signal measuring unit 30b is controlled by the control unit 40 described later and applies a predetermined voltage to the semiconductor sample M. In addition, the second electric signal measurement unit 30b uses the electric signal probes 16a to 16d from the wiring of the semiconductor sample M, which is a semiconductor device, during measurement of excitation of electrons in the semiconductor sample M irradiated with the electron beam. Electrical signals such as current may be measured.

  Note that the excitation of electrons in the semiconductor sample M irradiated with the electron beam may be measured without applying a voltage to the semiconductor sample M.

  The inside of the housing 17 is decompressed, and the semiconductor sample M placed on the stage 14 is arranged.

  The radiating unit 10 preferably includes a secondary electron detector and a backscattered electron detector (not shown) and can detect secondary electrons and backscattered electrons emitted or reflected from the surface of the semiconductor sample M. The secondary electron detector and the backscattered electron detector output detection signals detected in each scan to the control unit 40. The control unit 40 records the input detection signal in the recording unit 41 in association with each scanning position. The control unit 40 creates and displays a scanning electron microscope image of secondary electrons and reflected electrons based on the detection signal recorded in the recording unit 41.

  As the radiation | emission part 10 mentioned above, it can form using a scanning electron microscope, for example.

  The measurement signal generation unit 20 includes a radiation signal generation unit 21, a light source 22, an optical delay unit 23, and a first light detection unit 24.

  The radiation signal generation unit 21 generates a radiation signal H that causes the radiation unit 10 to emit radiation. The radiation signal H is a periodic pulsed electric signal. The radiation signal H generated by the radiation signal generation unit 21 is branched, and one is output to the scanning control unit 13 b described above, and the other is output to the light source 22.

  When the radiation signal H is input from the radiation signal generation unit 21, the scanning control unit 13 b changes the irradiation position on the semiconductor sample M and irradiates a pulsed electron beam.

  For example, when 256 × 256 scans are performed on the semiconductor sample M in one second, the frequency of the radiation signal H generated by the radiation signal generation unit 21 is approximately 65 kHz (≈256 × 256/1 second). can do.

  The light source 22 receives the periodic pulsed radiation signal H, generates a periodic pulsed optical signal in synchronization with the radiation signal H, and outputs the generated optical signal to the optical delay unit 23. The light source 22 may periodically block the continuous light with an optical chopper synchronized with the radiation signal H to generate a periodic pulsed optical signal.

  The optical delay unit 23 receives the light generated by the light source 22, delays the phase of the input light by a delay time, and outputs the delayed light to the first electric signal measurement unit 30a.

  This delay time is provided in order to make the measurement wait until the excited electrons emit light through each charge collection process in the semiconductor sample M irradiated with the electron beam. As shown in FIG. 1, the charge collection process of excited electrons mainly includes drift, funneling, and diffusion. Here, the charge collection process by drift or funneling occurs from 50 picoseconds to several hundred picoseconds from the start of electron beam irradiation, and the charge collection process by diffusion starts from the time of starting electron beam irradiation. It is believed to occur in 1 nanosecond to several microseconds.

  Therefore, the inspection apparatus 1 measures the light emission of the semiconductor sample M after waiting for a delay time from the time when the radiation unit 10 starts the irradiation of the electron beam R.

  In the charge collection process by drift or funneling, the delay time can be, for example, 50 picoseconds.

  The optical delay unit 23 adjusts the phase of the input light to be delayed by the delay time by changing the optical path length of the light input from the light source 22. For example, the optical delay unit 23 can delay the light by 1 picosecond in time by changing the optical path length by 300 μm.

  The first light detection unit 24 receives the light output from the optical delay unit 23, and outputs a measurement signal X1 that is an electrical signal to the first electrical signal measurement unit 30a.

  The first electrical signal measurement unit 30a receives the measurement signal X1 and measures the excitation of electrons in the semiconductor sample M during a measurement time that is a predetermined time. The first electric signal measurement unit 30a starts measuring the electric signal input from the second light detection unit 31 using the measurement signal X1 input from the first light detection unit 24 as a trigger. The measurement of the electric signal input from the second light detection unit 31 is terminated. The first electrical signal measurement unit 30 a outputs an electrical signal obtained by measuring the excitation of electrons in the semiconductor sample M to the control unit 40.

  The control unit 40 records the electrical signal input from the first electrical signal measurement unit 30a in the recording unit 41 as a measurement result obtained by measuring excitation of electrons in the semiconductor sample M. Further, the control unit 40 controls the radiation unit 10 to adjust the incident energy of the electron gun 11 and the scanning unit 12.

  Further, the control unit 40 can display the measurement result recorded in the recording unit 41 as an image.

  The control unit 40 includes a calculation unit (not shown), a storage unit, a display unit, an input unit, an output unit, and a communication unit. The calculation unit implements each function of the control unit 40 described above by executing a predetermined program stored in the storage unit. The recording unit 41 may be a part of this storage unit.

  The control unit 40 can be formed using, for example, a server or a computer such as a personal computer or a state machine.

  Further, the control unit 40 controls the measurement signal generation unit 20. Specifically, the control unit 40 controls the optical path length of the optical delay unit 23 to adjust the delay time. Further, the control unit 40 controls the radiation signal generation unit 21 to adjust the frequency of the radiation signal H. Moreover, you may make it the control part 40 control the 1st electric signal control part 30a.

  The delay time set in the optical delay unit 23 is preferably shorter than the time from the time when the radiation unit 10 starts irradiation of the electron beam to the start time of the charge collection process. Specifically, the delay time can be set to one-tenth of the time from when the radiating unit 10 starts irradiation of the electron beam to the start time of the charge collection process.

  For example, it is considered that the start time of the charge collection process by drift or funneling is about 50 picoseconds at the earliest from the time when the radiating unit 10 starts irradiation of the electron beam. In this case, the delay time is preferably 5 picoseconds.

  In addition, the start time of the charge collection process by diffusion is considered to be about 1 nanosecond to several microseconds from the time when the radiating unit 10 starts irradiation of the electron beam. Charge collection by drift or funneling occurs before it occurs. Therefore, in order to measure light emission from charge collection by drift or funneling to charge collection by diffusion, the delay time is preferably set in accordance with the direction in which charge collection starts earlier. In addition, when measuring light emission by the charge collection process by diffusion, the delay time can be set to 0.1 nanosecond.

  The measurement time for measuring the excitation of electrons in the semiconductor sample M after the first electrical signal measurement unit 30a inputs the measurement signal X1 determines the time resolution of the charge collection process. This measurement time can be appropriately set according to the required time resolution of the charge collection process, but, similar to the delay time described above, from the time when the radiation unit 10 starts irradiation of the electron beam to the start time of the charge collection process. Can be one tenth of the time between. For example, the measurement time can also be 5 picoseconds.

  The end time at which the inspection apparatus 1 finishes measuring the excitation of electrons in the semiconductor sample with respect to each radiation is preferably longer than the time until the start time of the charge collection process. Specifically, the end time can be set to 10 times the start time of the charge collection process from the time when the radiating unit 10 starts irradiation with the electron beam.

  Specifically, the start time of the charge collection process by diffusion is considered to be about 1 nanosecond to several microseconds at the earliest time from when the radiating unit 10 starts irradiation of the electron beam. Therefore, the end time can be set to several tens of microseconds from the time when the radiating unit 10 starts irradiation with the electron beam.

  In the case of measuring light emission in the charge collection process by charge collection by drift or funneling, the end time may be several thousand picoseconds.

  The inspection apparatus 1 performs the irradiation of the electron beam, the waiting for the delay time, and the measurement of the electron excitation while increasing and changing the delay time until the end time after starting the measurement of the electron excitation. repeat.

  The delay time set in the optical delay unit 23 is preferably set in consideration of the time required for the optical signal or the electric signal to be transmitted through the signal path.

  Next, an operation example of the above-described inspection apparatus 10 will be described below with reference to FIGS. 4 and 5.

  FIG. 4 is a flowchart illustrating an embodiment of the inspection method disclosed in this specification.

  First, in step S10, the inspection conditions of the inspection apparatus 1 are set for the semiconductor sample M. Here, the inspection conditions include an electron beam scanning speed, a scanning position and the number of scans (corresponding to the number of pixels), a delay time, a measurement time, and the like. Since the maximum processing time per pixel is determined from the scanning speed and the number of scans, the pulse width of the electron beam is set based on the maximum processing time.

  Next, the electric signal probes 16a to 16d are brought into contact with the semiconductor sample M in step S12. The electric signal probes 16a to 16d are in contact with, for example, a transistor of a semiconductor sample M that is a semiconductor device, or an electrode or wiring of a memory. Then, a voltage is supplied from the second electric signal measurement unit 30b to the electric signal probes 16a to 16d via the probe connector 16e.

  Next, in steps S14 to S30, the processing of steps S16 to S26 is performed in all the set delay times between the start time and the end time of the delay time.

  Next, between steps S16 to S26, the processing of steps S18 to S24 is performed at all scanning points on the semiconductor sample M.

  Next, in step S <b> 18, a pulsed electron beam is irradiated to the semiconductor sample M from the radiation unit 10 to the scanning point on the semiconductor sample M. At this time, the radiating unit 10 generates a detection signal by detecting secondary electrons and reflected electrons emitted or reflected from the surface of the semiconductor sample M using a secondary electron detector and a reflected electron detector (not shown). The generated detection signal is output to the control unit 40.

  Next, in step S <b> 20, the inspection apparatus 1 waits for a delay time from the time when the semiconductor sample M starts to be irradiated with a pulsed electron beam.

  Next, in step S22, the first electrical signal measurement unit 30a measures excitation of electrons in the semiconductor sample M during the measurement time.

  Next, in step S24, the scanning unit 12 changes the position on the semiconductor sample M irradiated with the electron beam.

  Next, the processing of steps S18 to S26 described above will be further described below.

  FIG. 5 is a diagram illustrating a temporal relationship between a pulsed electron beam and light emission measurement.

  FIG. 5A shows a pulsed electron beam that the radiating unit 10 irradiates the semiconductor sample M. FIG. The electron beam has a pulse width D and is irradiated with a period T. Further, the electron beam is scanned, the first irradiation is applied to the position (x0, y0) on the semiconductor sample M, and the second irradiation is applied to the position (x1, y1) on the semiconductor sample M. The third irradiation is applied to the position (x2, y2) on the semiconductor sample M, and all irradiation positions are scanned.

  FIG. 5B shows an electrical signal that the second light detection unit 31 outputs to the first electrical signal measurement unit 30a. The first electrical signal measurement unit 30a inputs the measurement signal X1 from the first light detection unit 24 when the delay time τ elapses from the time when the radiation unit 10 starts irradiating the electron beam. The first electrical signal measurement unit 30a that has input the measurement signal X1 measures the excitation of electrons in the semiconductor sample M by measuring the electrical signal input from the second light detection unit 31 during the measurement time s. . The first electrical signal measuring unit 30a measures the area of the electrical signal during the measurement time s. Then, the first electrical signal measurement unit 30a outputs the measured area of the electrical signal as a measurement result to the control unit 40. The delay time τ may overlap with the electron beam irradiation or may not overlap.

  The control unit 40 arranges the area w of the electric signal as the input measurement result in the recording unit 41 (τ, x, y, w) together with the position (x, y) on the semiconductor sample M and the delay time τ. Record as. Here, the area w of the electrical signal may be recorded in the array as the measurement value, or the area w of the electrical signal may be scored and recorded in the array based on a threshold set by the user.

  In this way, the inspection apparatus 1 repeats the processes in steps S18 to S24 for all scanning points.

  Next, in step S <b> 28, the control unit 40 changes the delay time of the optical delay unit 23. The controller 40 increases the delay time by τ and changes the delay time. Note that the amount of increase in the delay time may not be constant.

  Next, the processing of steps S14 to S30 described above will be further described below.

  After the inspection apparatus 1 performs the measurement using the first delay time τ for each scanning point on the semiconductor sample M, the control unit 40 adjusts the optical path length of the optical delay unit 23 and sets the delay time to τ. To 2τ.

  Next, the inspection apparatus 1 performs measurement using the delay time 2τ for each scanning point on the semiconductor sample M. The first electrical signal measurement unit 30a inputs the measurement signal X1 from the first light detection unit 24 when the delay time 2τ has elapsed from the time when the radiation unit 10 starts irradiation of the electron beam. The first electrical signal measurement unit 30a that has input the measurement signal X1 measures the excitation of electrons in the semiconductor sample M by measuring the electrical signal input from the second light detection unit 31 during the measurement time s. .

  FIG. 5C shows an electrical signal that the second light detection unit 31 outputs to the first electrical signal measurement unit 30a at the delay time 2τ. The first electrical signal measuring unit 30a measures the area of the electrical signal during the measurement time s. Then, the first electrical signal measurement unit 30a outputs the measured area of the electrical signal as a measurement result to the control unit 40.

  The control unit 40 arranges the area w of the electric signal, which is the input measurement result, in the recording unit 41 (2τ, x, y, w) together with the position (x, y) on the semiconductor sample M and the delay time 2τ. Record as.

  Then, the inspection apparatus 1 performs the measurement using the delay time 2τ for each scanning point on the semiconductor sample M, and then repeats the processes of steps S14 to S30 for all the delay times.

  The electric signal waveform shown in FIG. 5B or FIG. 5C is obtained by converting light emitted at each moment in the semiconductor sample M into an electric signal. In the semiconductor sample M, light emitted at each moment is caused by each charge collection process that occurs at each moment. That is, the electric signal waveform shown in FIG. 5B or 5C is a convolution of light emission in the charge collection process having different generation times. Therefore, based on the measurement result recorded in the recording unit 41, the electric signal waveform measured during each measurement time s can be deconvoluted to obtain the emission signal intensity at each occurrence time.

  In the operation of the inspection apparatus 1 described above, the inspection time is shortened by changing the delay time after measuring all the scanning points. If measurement is performed for all scanning points after performing measurement with all delay times changed at one scanning point, the inspection time significantly increases. For example, when an inspection having 256 × 256 scans and 256 delay times is performed, the man-hour associated with the change in the delay time is approximately 16.77 million steps (= 256 × 256 × 256). On the other hand, if the delay time is changed after measuring all the scanning points, the man-hour associated with the change of the delay time can be greatly reduced to 256 steps.

  In step S32, the control unit 40 displays the measurement result.

  Next, measurement results displayed by the control unit 40 will be described below.

  FIG. 6 is a diagram illustrating an array in which measurement results are stored.

  As described above, the array H recorded in the recording unit 41 includes the delay time τ, the position (x, y) on the semiconductor sample M, and the area w of the electric signal that is the input measurement result as array elements. Have. The array Hτ illustrates the array (τ, x, y, w) at the delay time τ. In the delay time τ, each cell corresponding to the position (x, y) has the electric signal area w as a value.

  The array Hτ in which all the cells in the delay time τ are arranged in two dimensions is a two-dimensional measurement result obtained by measuring the amount of light emitted by excitation of electrons generated between the delay time τ and the measurement time s for the semiconductor sample M. This is a visually visualized measurement result.

  Similarly, the array H2τ in which all the cells at the delay time 2τ are two-dimensionally arranged is a measurement result obtained by measuring the light emission amount due to the excitation of electrons generated between the delay time 2τ and the measurement time s with respect to the semiconductor sample M. It is the measurement result which visualized two-dimensionally.

  Therefore, the array H is formed by superposing the arrays Hτ, H2τ,... Arranged in two dimensions for all delay times in the three-dimensional direction (the direction of the delay time).

  Therefore, a cell stack Hc in which cells at a position (x, y) on a certain semiconductor sample M are stacked in the direction of the delay time changes with time in the excitation of electrons at the position (x, y) on the semiconductor sample M. The measurement result is visualized.

  Since the time in which each charge collection process occurs after irradiation is different, the cell stack Hc is considered to represent the probability that each charge collection process of electrons excited in the semiconductor sample occurs. Therefore, based on the cell stack Hc, it is possible to obtain physical property parameters relating to soft errors such as the probability (occurrence cross-sectional area) that each charge collection process occurs. The obtained physical property parameters can be used to perform a soft error simulation.

  7A shows a scanning electron microscope image, FIG. 7B shows an imaged measurement result image, and FIG. 7C shows an SEM image and a measurement result image. A superimposed image is shown.

  Specifically, FIG. 7A shows a scanning electron microscope image G1 of secondary electrons and reflected electrons captured by the radiating unit 10 together with measurement of excitation of electrons of the semiconductor sample M.

  FIG. 7B shows an image G2 created based on the array Hτ at the delay time τ. The position of each pixel corresponds to the position (x, y) on the semiconductor sample M, and the luminance of the pixel corresponds to the area w of the electric signal that is the input measurement result.

  In the example shown in FIG. 7B, a pixel having an electric signal area equal to or larger than a threshold specified by the user is displayed as a singular point E. That is, at the singular point E, it is determined that a soft error has occurred.

  FIG. 7C shows a composite image G3 in which the image G1 and the image G2 are superimposed. The singular point E corresponds to the position of the transistor. By using the composite image G3, the location where the single event phenomenon occurs in the semiconductor sample M is clearly shown.

  FIG. 8 is a diagram showing the relationship between the frequency of occurrence of soft errors and the energy of electron beams.

  FIG. 8 shows the relationship between the frequency of occurrence of soft errors and the energy of the electron beam based on the array H of measurement results. The vertical axis in FIG. 8 represents the frequency of occurrence of soft errors at a certain position (x, y) on the semiconductor sample M. The frequency of occurrence of the soft error is a total of measurement results over a certain delay time or over all delay times.

  A curve C1 shows a measurement result for the semiconductor sample M. A curve C1 indicates that the occurrence frequency has a predetermined threshold with respect to the energy of the electron beam. The reason why there is a threshold for the energy of the electron beam is that the operating voltage of a switching element such as a transistor has a threshold, or because the electrons in the valence band are excited to the conduction band beyond the energy gap. This is because it has a high excitation energy.

  From the curve C1, the relationship between the probability of occurrence of a soft error at a certain position (x, y) on the semiconductor sample M and the energy of the electron beam is obtained.

  FIG. 9 is a diagram showing the relationship between the frequency of occurrence of soft errors and the electron dose.

  FIG. 9 shows a curve C3 showing the relationship between the probability of occurrence of the semiconductor sample M with respect to the soft error and the electron dose when the acceleration voltage is kept constant and the electron dose is changed.

  A curve C3 indicates that the electron dose has a predetermined threshold value. The reason why there is a threshold for the electron dose is that a predetermined amount of excitation of electrons is required to generate an operating voltage of a switching element such as a transistor. The electron dose can be adjusted by a condenser lens, and can be detected by directly irradiating an Faraday cup with an electron beam.

  As shown in FIGS. 8 and 9, a soft error can be generated at a high probability at the same position by irradiating a semiconductor sample with an electron beam energy or an electron dose of a predetermined threshold value or more. . On the other hand, when the semiconductor sample is irradiated with an accelerating voltage or an electron beam having an electron dose less than a predetermined threshold, the frequency of occurrence of soft errors is very small.

  By the way, the electrical signal that can be measured by the first electrical signal measuring unit 30a can also be generated due to defects originally present in the semiconductor sample M such as cracks and crystal defects.

  Therefore, the semiconductor sample M is irradiated with an electron beam having an incident energy equal to or higher than a predetermined value, and the process shown in FIG. 5 is performed to obtain first measurement data. The first measurement data includes a measurement result caused by a defect originally present in the semiconductor sample M, together with a measurement result caused by a soft error.

  Next, an electron beam having an incident energy less than a predetermined value is irradiated onto the semiconductor sample M, and the process shown in FIG. 5 is performed to obtain second measurement data. The second measurement data originally includes a measurement result due to a defect present in the semiconductor sample M, but hardly includes a measurement result due to a soft error.

  Next, the presence or absence of a soft error in the semiconductor sample M is determined based on the first measurement data and the second measurement data. Specifically, a difference array is obtained by obtaining a difference between corresponding array elements between the array based on the first measurement data and the array based on the second measurement data. This difference array includes measurement results mainly due to soft errors. Note that either the first measurement data or the second measurement data may be measured first.

  Next, it will be described below that the inspection apparatus 1 can perform a soft error acceleration test of the semiconductor sample M. The inspection apparatus 1 can be used for performing an accelerated test with charged particles (ion beam or electron beam) to accelerate and quantitatively evaluate a single event phenomenon caused by natural radiation.

  As radiation existing in nature, there is, for example, α rays (nuclei of helium atoms). It will be described below that the frequency of occurrence of soft errors can be estimated using the measurement results of the inspection apparatus 1 when the semiconductor sample M is irradiated with α rays.

  FIG. 10 is a diagram illustrating a formula for stopping power.

  The stopping power Scol of the charged particles in the target substance is represented by the stopping power formula shown in Formula (1). Formula (1) uses the assumption that the energy lost by the charged particles is approximately equal to the energy imparted to the target substance, and is known as the Bethe-Bloch formula.

  The stopping power Scol represented by the formula (1) is an index called linear energy transfer (LET). LET is an index representing the difference in radiation quality, and is the amount of energy (dE / dx) given locally to the target substance per unit length along the track by the ionization action of charged particles. The larger the LET of charged particles, the easier the single event phenomenon occurs. In addition, when electrons are used as charged particles, Equation (2) with relativistic correction is used. LET can be uniquely determined for any charged particle species or energy by the formula (1) or (2).

  The number of occurrences of a single event phenomenon with respect to the number of charged particles per unit area irradiated on a semiconductor sample such as a memory element is called “inverted cross-sectional area”. The “inverted cross-sectional area” is “the number of occurrences of a single event phenomenon with respect to the number of charged particles (alpha rays or electron beams) per unit irradiation area”. The number of occurrences of the single event phenomenon is specifically the number of light emissions equal to or greater than a threshold value or the number of information inversions in the memory. For example, by observing the image (for example, the number of pixels 256 × 256) shown in FIG. 7 and counting the number of light emission or memory inversion in the image, the “inversion cross-sectional area” is obtained.

The inverted cross-sectional area has the relationship shown in FIG. 11 with respect to LET. The inverted cross-sectional area has a predetermined threshold with respect to LET. This threshold is determined based on the critical charge Q _c of the single event phenomenon. The minimum charge to invert the information of the semiconductor sample such as a memory device with the single event phenomenon is called the critical charge Q _c. The critical charge _Qc depends on the material (target) of the memory element, the type of charged particle (radiation type), energy and dose. The minimum LET (threshold) at which the critical charge Q _c is generated is that Q _c is equal to the energy lost by ionization of the charged particles. The minimum LET (threshold) at which the critical charge Q _c is generated corresponds to the threshold values shown in FIGS.

  FIG. 11 shows the relationship between the inverted cross-sectional area (occurrence frequency) and LET. Even if the radiation type (atomic number Zi) or energy (Ei) is different, LET is uniquely determined according to formula (1) or formula (2). Therefore, the acceleration coefficient can be evaluated using the LET ratio of each condition. Even with the same line type, if the energy is high, the LET increases, and the ionization effect increases accordingly, so that the inversion cross section also increases (plot A → plot C).

  According to formula (1) or formula (2), if the charged particle type is determined, LET depends only on energy, so FIG. 11 can be rewritten as shown in FIG. For the single event phenomenon, the greater the energy, the greater the cross section. On the other hand, if the energy is the same, in the case of other than electrons, the inversion cross section increases as the atomic number of the charged particle species increases.

  Similarly, the relationship between the dose and the inversion cross-sectional area can be expressed as shown in FIG. When LET is the same, the inversion cross section increases as the dose increases.

  In FIG. 8, FIG. 9 and FIG. 11, the inversion cross-sectional area increases as the horizontal axis (LET, energy, dose) increases, but gradually becomes saturated. The reason for saturation is unknown, but even under high LET conditions, the region where the single event effect appears is limited to the channel periphery of the transistor, so the difference in LET cannot be seen above a certain value. Conceivable.

  The acceleration rate of the accelerated test is expressed by multiplying the LET ratio and the dose ratio between the comparison objects.

  FIG. 8 shows a curve C2 plotted together with the curve C1 using the equation (1) with the charged particle as α-ray and the target material as silicon based on the relationship between the above-described inversion cross-sectional area and LET. ing. The energy is given as the velocity v of α particles in β.

  The ratio between the occurrence frequency of the curve C1 and the occurrence frequency of the curve C2 at the same acceleration voltage represents the degree of the acceleration test of the measurement result of the inspection apparatus 1.

  Therefore, based on the ratio between the occurrence frequency of the curve C1 obtained using the electron beam and the occurrence frequency of the curve C2, the occurrence frequency of the soft error when the semiconductor sample M is irradiated with α rays is estimated. It is possible.

  When estimating the acceleration of the inspection apparatus 1 using charged particles other than electrons against a soft error due to an electron beam in the natural world, the above-described curve C2 is created using Equation (2).

  In this way, based on the measurement result of the inspection apparatus 1, it is possible to estimate the frequency of occurrence of a soft error with respect to radiation that the semiconductor sample M is irradiated in nature.

  As a conventional measurement method, a field test by continuous operation and an irradiation test by a large accelerator are used, but neither of them is efficient. This is because in the field test, since the occurrence rate of soft errors is low, continuous operation for several months to one year is required for highly reliable evaluation. One irradiation test is because there are few facilities that can use a neutron beam or an ion beam, and the test cannot be performed frequently.

  On the other hand, by using the inspection apparatus 1, it is possible to accelerate the inspection of the soft error of the semiconductor sample.

  In addition, it has been proposed to predict a soft error of a semiconductor sample by performing a simulation. However, since the probability of occurrence in the charge collection process is unknown, a series of mechanisms from a single event phenomenon to a soft error has not been demonstrated, and physical property parameters necessary for simulation have not been obtained with high accuracy.

  On the other hand, since the probability of occurrence of each charge collection process causing a soft error of the semiconductor sample can be obtained by using the inspection apparatus 1, it is considered that a highly accurate simulation can be performed based on the obtained physical property parameters.

  According to the inspection apparatus 1 of the present embodiment described above, it is possible to measure the charge collection process of electrons excited in the semiconductor sample irradiated with radiation.

  In addition, according to the inspection apparatus 1 of the present embodiment, time-resolved observation of a single event phenomenon, identification of a soft error occurrence location, acquisition of a cross-sectional area of occurrence of a single event phenomenon, and radiation quality by LET Comparison is possible.

  Specifically, by using the inspection apparatus 1 of the present embodiment, it is possible to visually evaluate soft errors that occur in a micro LSI device of submicron or less. In addition, with a single transistor, the elementary process of a single event phenomenon can be visually evaluated. In addition, in the inverter or SRAM, the location where the soft error occurs can be observed with high time resolution. Furthermore, since knowledge about the device structure can be obtained from the scanning electron microscope image, quantitative data of the cross-sectional area where the single event phenomenon occurs with respect to the dopant concentration and the distance from the electrode can be obtained from the semiconductor sample.

  This leads to improvement of simulation model accuracy. Furthermore, by comparing electron beam energy dependency and electron dose dependency, such as the single event phenomenon generation cross-section and bit reversal cross-section, with the heavy ion LET dependency, the device's resistance to environmental radiation is accelerated. Can be evaluated.

  Furthermore, it is possible to estimate the influence of the semiconductor sample on the environmental radiation using the measurement result of the inspection apparatus 1. Ambient radiation is generally expressed in units of J / kg, Sv (Sievert) or Gy (Gray). On the other hand, the measurement result of the inspection apparatus 1 is expressed in units of J / cm. Here, since the mass and dimensions of the semiconductor sample are known, the effect of the semiconductor sample on the environmental radiation can be estimated using the measurement result of the frequency of the soft error by converting between the units of both. .

  Next, a modified example of the above-described inspection apparatus will be described below.

  FIG. 12 is a diagram for explaining another temporal relationship between a pulsed electron beam and light emission measurement.

  In this modification, after measuring the excitation of electrons in the semiconductor sample M, the semiconductor sample M was irradiated with a pulsed electron beam and the semiconductor sample M was started to be irradiated with a pulsed electron beam. Waiting for the delay time from the time point and measuring the excitation of electrons in the semiconductor sample M are repeated during the measurement time. After repeating these, the position where the pulsed electron beam is irradiated is changed. Then, the average value of the measurement results is recorded in the array H.

  Thus, by repeating the measurement at the same position with the same delay time, it is possible to improve the S / N of the measurement signal obtained by detecting light emission and converting it into an electrical signal.

  As shown in FIG. 8 or 9, a soft error is repeatedly generated at a high probability even at the same position by irradiating a semiconductor sample with an electron beam having an acceleration voltage or electron dose exceeding a predetermined threshold. Can do.

  FIG. 13 is a flowchart for explaining another embodiment of the inspection method disclosed in this specification.

  In this modification, after performing the processing of steps S46 to S56, the position where the semiconductor sample is irradiated with the electron beam is changed. That is, the semiconductor sample M is irradiated with the pulsed electron beam, the semiconductor sample M is irradiated with the pulsed electron beam, the delay time is started, and the semiconductor sample M After measuring the excitation of electrons in the sample M and changing the delay time, the position where the semiconductor sample is irradiated with the electron beam is changed.

  Next, another embodiment of the above-described inspection apparatus will be described below with reference to FIGS. For points that are not particularly described in the other embodiments, the description in detail regarding the first embodiment is applied as appropriate. Moreover, the same code | symbol is attached | subjected to the same component.

  FIG. 14 is a diagram illustrating a second embodiment of the inspection apparatus disclosed in this specification.

  The inspection apparatus 1 according to the present embodiment is different from the first embodiment described above in that an electrical signal based on excitation of electrons in a semiconductor sample is measured.

  The measurement signal generation unit 20 includes a radiation signal generation unit 21 and a delay circuit 25.

  The radiation signal generation unit 21 generates a radiation signal H that is an electrical signal that causes the radiation unit to emit radiation. The radiation signal H generated by the radiation signal generation unit 21 is branched, and one is output to the scanning control unit 13 b and the other is output to the delay circuit 25.

  The delay circuit 25 receives the radiation signal H, and outputs a measurement signal X2 that is an electrical signal when the delay time has elapsed from the time when the radiation signal H was input.

  The second electrical signal measuring unit 30b receives the measurement signal X2 and measures the excitation of electrons in the semiconductor sample M during a measurement time that is a predetermined time. The second electrical signal measuring unit 30b starts measurement of electrical signals such as currents input from the electrical signal probes 16a to 16d via the probe connector 16e using the measurement signal X2 input from the delay circuit 25 as a trigger, After the measurement time has elapsed, the measurement of the electrical signal is terminated. The second electric signal measurement unit 30 b outputs an electric signal obtained by measuring the excitation of electrons in the semiconductor sample M to the control unit 40.

  The electric signal probes 16a to 16d apply a voltage to the wiring of the semiconductor sample M, which is a semiconductor device, and detect a current signal. In the case of a single transistor, the contact positions of the electric signal probes 16a to 16d are connected to the gate, source, drain and well terminals to monitor voltage fluctuations. When evaluating an SRAM including a plurality of transistors, the contact positions of the electric signal probes 16a to 16d are preferably connected to a wiring in a circuit such as a word line to monitor bit information and switching characteristics.

  The second electric signal measuring unit 30b is controlled by the control unit 40 to apply a voltage to some of the electric signal probes 16a to 16d and measure an electric signal such as a current from the other electric signal probes 16a to 16d. To do. As the second electric signal measuring unit 30b, for example, a semiconductor source measure unit (SMU) can be used.

  In measurement with a delay time of 20 nanoseconds or more, it is preferable to measure an electrical signal based on excitation of electrons in a semiconductor sample from the viewpoint of being able to wait for an accurate delay time using the delay circuit 25.

  According to the inspection apparatus 1 of the present embodiment described above, the apparatus can be formed using simpler components than the first embodiment described above. Further, the same effect as that of the first embodiment described above can be obtained.

  FIG. 15 is a diagram illustrating a third embodiment of the inspection apparatus disclosed in this specification.

  The inspection apparatus 1 of the present embodiment measures light emission and electrical signals based on excitation of electrons in the semiconductor sample M.

  The inspection apparatus 1 of the present embodiment includes the first electric signal measurement unit 30a of the first embodiment and the second electric signal measurement of the second embodiment together with the measurement signal generation unit of the first embodiment and the second embodiment described above. A portion 30b is provided.

  According to the inspection apparatus 1 of the present embodiment described above, light emission based on excitation of electrons in the semiconductor sample M is measured and an electrical signal based on excitation of electrons is measured, so that the reliability of measurement is improved. Further, the same effect as that of the first embodiment described above can be obtained.

  In the present invention, the inspection apparatus and the inspection method of the above-described embodiment can be appropriately changed without departing from the gist of the present invention. In addition, the configuration requirements of one embodiment can be applied to other embodiments as appropriate.

  For example, in the embodiment described above, pulsed radiation is emitted to the semiconductor sample, but continuous radiation may be emitted to the semiconductor sample.

  In the above-described embodiment, the radiation emission and standby and the excitation of electrons are measured while changing the position where the radiation is incident on the semiconductor sample. However, the scanning is not performed on the semiconductor sample. The position where the radiation is incident may be fixed.

  In the above-described embodiment, the light emission or electrical signal based on the excitation of electrons in the semiconductor sample is measured. However, the method for measuring the excitation of electrons is not limited to these.

  All examples and conditional words mentioned herein are intended for educational purposes to help the reader deepen and understand the inventions and concepts contributed by the inventor. All examples and conditional words mentioned herein are to be construed without limitation to such specifically stated examples and conditions. Also, such exemplary mechanisms in the specification are not related to showing the superiority and inferiority of the present invention. While embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions or modifications can be made without departing from the spirit and scope of the invention.

  Regarding the above-described embodiments, the following additional notes are disclosed.

(Appendix 1)
A first step of irradiating a semiconductor sample with radiation;
A second step of waiting for a first time from the start of radiation irradiation on the semiconductor sample;
A third step of measuring excitation of electrons in the semiconductor sample for a second time after the second step;
A fourth step of changing the first time;
With
An inspection method that repeats the first step, the second step, and the third step.

(Appendix 2)
The inspection method according to supplementary note 1, wherein the first step, the second step, and the third step are repeated while changing a position where the semiconductor sample is irradiated with radiation.

(Appendix 3)
The inspection method according to supplementary note 1, wherein the first step, the second step, the third step, and the fourth step are repeated while changing a position at which the semiconductor sample is irradiated with radiation.

(Appendix 4)
The inspection method according to any one of appendices 1 to 3, wherein the first step, the second step, and the third step are further repeated after the third step.

(Appendix 5)
The inspection method according to any one of appendices 1 to 4, wherein in the third step, light emission or an electric signal based on excitation of electrons in the semiconductor sample is measured.

(Appendix 6)
In the first step, the semiconductor sample is irradiated with radiation having an incident energy of a predetermined value or more,
The inspection method according to any one of appendices 1 to 5, wherein in the third step, first measurement data obtained by measuring excitation of electrons in the semiconductor sample is obtained.

(Appendix 7)
In the first step, the semiconductor sample is further irradiated with radiation having an incident energy less than the predetermined value,
The inspection method according to appendix 6, wherein the third step further obtains second measurement data obtained by measuring excitation of electrons in the semiconductor sample.

(Appendix 8)
The inspection method according to appendix 7, wherein the presence or absence of a soft error in the semiconductor sample is determined based on the first measurement data and the second measurement data.

(Appendix 9)
The inspection method according to any one of appendices 1 to 8, wherein in the first step, the semiconductor sample is irradiated with pulsed radiation.

(Appendix 10)
A radiation part for irradiating the semiconductor sample with radiation;
A measurement signal generating unit that outputs a measurement signal when a first time has elapsed from the time when the radiating unit starts irradiation of radiation; and
A measurement unit that inputs the measurement signal and measures excitation of electrons in the semiconductor sample during a second time; and
A control unit for changing the first time;
An inspection apparatus comprising:

(Appendix 11)
The measurement unit measures light emission based on excitation of electrons in a semiconductor sample,
The measurement signal generator is
A radiation signal generating unit for generating a radiation signal for irradiating the radiation unit with radiation; and
A light source that receives the radiation signal and generates light;
An optical delay unit that inputs the light generated by the light source and outputs the phase of the input light delayed by the first time;
A light detection unit that inputs light output from the optical delay unit and outputs the measurement signal;
The inspection apparatus according to appendix 10, wherein

(Appendix 12)
The measurement unit measures an electrical signal based on excitation of electrons in a semiconductor sample,
The measurement signal generator is
A radiation signal generating unit for generating a radiation signal for irradiating the radiation unit with radiation; and
A delay circuit that inputs the radiation signal and outputs the measurement signal when a first time elapses from the time when the radiation signal is input;
The inspection apparatus according to appendix 10, wherein

(Appendix 13)
The inspection apparatus according to appendix 10, wherein the radiation unit includes a scanning unit that changes a position at which the semiconductor sample is irradiated with radiation.

DESCRIPTION OF SYMBOLS 1 Inspection apparatus 10 Radiation part 11 Electron gun 12 Scan part 13a Scan coil 13b Scan control part 14 Stage 15 Photodetection probe 16a, 16b, 16c, 16d Electric signal probe 16e Probe connector 17 Case 20 Measurement signal generation part 21 Radiation signal generation Unit 22 light source 23 optical delay unit 24 first light detection unit 25 delay circuit 30 electric signal measurement unit 30a first electric signal measurement unit 30b second electric signal measurement unit 31 second light detection unit 40 control unit 41 recording unit M semiconductor sample R Electron beam (radiation)
X1, X2 Measurement signal D Pulse time τ Delay time s Measurement time H Array E Array element with soft error G1 SEM image G2 Measurement result image G3 Overlaid image C1, C2 Curved

Claims (7)

A first step of irradiating a semiconductor sample with radiation;
A second step of waiting for a first time from the start of radiation irradiation on the semiconductor sample;
A third step of measuring excitation of electrons in the semiconductor sample for a second time after the second step;
A fourth step of changing the first time;
With
An inspection method that repeats the first step, the second step, and the third step.   The inspection method according to claim 1, wherein the first step, the second step, and the third step are repeated while changing a position where the semiconductor sample is irradiated with radiation.   The inspection method according to claim 1 or 2, wherein in the third step, light emission or an electric signal based on excitation of electrons in the semiconductor sample is measured. In the first step, the semiconductor sample is irradiated with radiation having an incident energy of a predetermined value or more,
The inspection method according to claim 1, wherein in the third step, first measurement data obtained by measuring excitation of electrons in the semiconductor sample is obtained. In the first step, the semiconductor sample is further irradiated with radiation having an incident energy less than the predetermined value,
The inspection method according to claim 4, wherein in the third step, second measurement data obtained by measuring excitation of electrons in the semiconductor sample is further obtained.   The inspection method according to claim 5, wherein the presence or absence of a soft error in the semiconductor sample is determined based on the first measurement data and the second measurement data. A radiation part for irradiating the semiconductor sample with radiation;
A measurement signal generating unit that outputs a measurement signal when a first time has elapsed from the time when the radiating unit starts irradiation of radiation; and
A measurement unit that inputs the measurement signal and measures excitation of electrons in the semiconductor sample during a second time; and
A control unit for changing the first time;
An inspection apparatus comprising:

Images