WO2024100847A1

WO2024100847A1 - Process diagnosis device, and method for determining plasma replacement timing

Info

Publication number: WO2024100847A1
Application number: PCT/JP2022/041945
Authority: WO
Inventors: 貴則近藤; 美臣甲斐; 健一郎米田; 健次岡; 貴志堤; 真史佐藤; 将志川畑; 俊一松本
Original assignee: 株式会社日立ハイテク
Priority date: 2022-11-10
Filing date: 2022-11-10
Publication date: 2024-05-16

Abstract

Provided is a process diagnosis device that diagnoses the state of a semiconductor process device, the process diagnosis device comprising: a sample stage that supports a sample processed by the semiconductor process device; a lighting optical system that radiates an illumination light onto the sample placed on the sample stage; a plurality of detection optical systems that concentrate light from the sample, convert the light to an electrical signal, and output a detection signal; and a signal processing device that processes the detection signals from the plurality of detection optical systems. The signal processing device scans the sample, extracts a haze signal of the sample, compares the haze signal of the sample to a reference haze signal, and outputs an alarm if the difference therebetween exceeds a set value.

Description

Process diagnostic device and method for determining timing for plasma replacement

The present invention relates to a process diagnostic device for diagnosing semiconductor process equipment such as plasma etching equipment, and a method for determining the timing of plasma replacement.

　It is important to manage the condition of process equipment used to process semiconductor wafers, since deterioration of parts and reduced cleanliness can lead to lower semiconductor yields. For example, in the case of plasma etching equipment, consumables such as plasma light sources that deteriorate over time are regularly replaced and cleaned. However, there are individual differences in the service life of plasma light sources, and regular replacement or cleaning cannot necessarily prevent abnormalities in the plasma light source.

Incidentally, Patent Document 1 discloses a technology for an optical inspection device used in defect inspection of semiconductor samples such as wafers, which not only detects the main defect information from the scattered light intensity obtained from the sample, but also identifies information on wafer characteristics other than defects, such as the surface roughness and surface film thickness of the sample. Optical defect inspection devices usually inspect the entire surface of the sample. In Patent Document 1, a correlation is defined in advance between characteristic values such as the surface roughness and surface film thickness of the sample measured at several points on the sample using an atomic force microscope (AFM), spectroscopic ellipsometer, or other measuring device, and the scattered light intensity of the entire surface of the sample obtained when inspected for defects using the optical defect inspection device. Based on this correlation, the characteristic value of the entire surface of the sample is calculated from the scattered light intensity measured in the defect inspection. The surface roughness and surface film thickness can be referenced over the entire surface using the characteristic values of the sample calculated in this way.

JP 2011-013058 A

However, in the technology of Patent Document 1, when determining the correlation between the scattered light intensity and the characteristic value of the sample, it is necessary to measure the characteristic value of the sample using another measuring device such as an AFM. Due to the nature of using a measuring device such as an AFM, the actual measurement of the characteristic value is limited to a portion of the entire surface of the sample. The correlation between the scattered light intensity and the characteristic value determined by the technology of Patent Document 1 is only partially based on the actual measurement value. Therefore, the majority of the rest of the sample surface must be obtained by interpolation. Therefore, the technology of this document has room for improvement in the accuracy of the calculation of the characteristic value of the sample, and ultimately the evaluation of the state of the process equipment.

The object of the present invention is to provide a process diagnostic device and a method for determining the timing of plasma replacement that can accurately diagnose the state of semiconductor process equipment from data obtained by defect inspection of samples after processing, suppress the occurrence of abnormalities in the semiconductor process equipment, and improve semiconductor yields.

In order to achieve the above object, the present invention provides a process diagnostic device for diagnosing the state of a semiconductor process device, the process diagnostic device comprising: a sample stage for supporting a sample processed by the semiconductor process device; an illumination optical system for irradiating illumination light onto the sample placed on the sample stage; a plurality of detection optical systems for collecting light from the sample, converting it into an electrical signal, and outputting a detection signal; and a signal processing device for processing the detection signals from the plurality of detection optical systems, the signal processing device scanning the sample to extract a haze signal of the sample, comparing the haze signal of the sample with a reference haze signal, and outputting an alarm if the difference exceeds a set value.

According to the present invention, the condition of semiconductor processing equipment can be accurately diagnosed from data obtained by defect inspection of samples after processing, the occurrence of abnormalities in the semiconductor processing equipment can be suppressed, and semiconductor yields can be improved.

FIG. 1 is a schematic diagram of a configuration example of a process diagnostic device according to a first embodiment of the present invention; Schematic diagram showing the sample scanning trajectory Schematic diagram showing the sample scanning trajectory Schematic diagram showing the attenuator A schematic diagram showing the positional relationship between the optical axis of illumination light guided obliquely to the surface of a sample and the illumination intensity distribution shape. A schematic diagram showing the positional relationship between the optical axis of illumination light guided obliquely to the surface of a sample and the illumination intensity distribution shape. A diagram showing the area where the detection optics collects scattered light, viewed from above. A schematic diagram showing the zenith angles of low-angle and high-angle detection optical systems. Plan view showing the azimuth angle of the low-angle detection optics Plan view showing the azimuth angle of the high-angle detection optics A schematic diagram showing an example of the configuration of the detection optical system FIG. 1 is a functional block diagram of a main part of a signal processing device provided in a process diagnostic device according to a first embodiment of the present invention; 1 is a flowchart generally showing an example of a series of diagnostics of process equipment in a semiconductor manufacturing process. Schematic diagram showing time-dependent changes in the state of process equipment A schematic diagram showing the effect on the sample due to the change over time in the state of the process equipment, focusing on the XV part in FIG. 14. 1 is a flowchart showing a procedure for diagnostic processing of a process device by a process diagnostic device according to a first embodiment of the present invention. Conceptual diagram of machine learning according to a second embodiment of the present invention FIG. 11 is a schematic diagram showing a main part of a process diagnostic device according to a third embodiment of the present invention; FIG. 13 is a schematic diagram showing a main part of a process diagnostic device according to a fourth embodiment of the present invention;

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
The process diagnostic device described in the following embodiment scans a sample, here a wafer (bare wafer, film-coated wafer, patterned wafer, etc.) during the semiconductor manufacturing process, and diagnoses the process device from the state of the sample processed by the process device. An example of a process diagnostic device is an optical defect inspection device that inspects defects on a sample. The optical defect inspection device outputs the number, coordinates, and type of defects attached to or formed on the wafer, based on a signal based on reflected or scattered light obtained by scanning the sample. It is used to grasp the state of the sample processed based on this information. The optical defect inspection device has a very fast scan speed compared to diagnostic devices that use electron beams, X-rays, etc. as a light source. Therefore, while diagnostic devices that use electron beams, X-rays, etc. as an energy source can only measure a very small number of points on the sample due to time constraints, the optical defect inspection device of this embodiment can scan the entire surface of the sample and diagnose the state of the entire surface of the sample. The light-based signals obtained by scanning the sample with this optical defect inspection device include not only defect signals used for defect detection, but also signals called haze signals. In this embodiment, a method of diagnosing the state of the process device mainly using this haze signal will be described. If a signal with a fluctuation frequency (time fluctuation of signal intensity) higher than a predetermined value is a high frequency, and a signal with a lower frequency is a low frequency, the defect signal corresponds to a high frequency component of the signal based on the light obtained from the sample. Conversely, the haze signal corresponds to a low frequency component. For example, a signal caused by relatively large irregularities attached to the sample, such as foreign matter, is likely to be detected as a high frequency component, while a signal caused by the characteristics of the wafer itself, such as the thickness of the film on the sample and extremely small irregularities (roughness) on the sample surface, is likely to be detected as a low frequency component. The process diagnosis device of this embodiment utilizes a haze signal that is not generally used (removed) in sample defect inspection, and diagnoses the state of the process device that processed the sample based on the haze signal obtained from the sample. Specifically, the device monitors the change over time of the haze signal, and when a predetermined difference (relative change in the haze signal) occurs between the haze signal and the reference haze signal, it detects that there is some change (abnormality) in the sample processed by the process. Such a change in the wafer characteristics based on the haze signal allows an abnormality or a sign of an abnormality in the process device that processed the sample, such as a plasma process device for etching, to be estimated and notified to an operator, etc.

In this specification, an abnormality in a process device refers to a state in which the sample processing quality (surface film thickness or surface roughness of the sample surface in the case of a plasma etching/film formation device) has already fallen below an acceptable level and the process device needs to be maintained immediately. A sign of an abnormality in a process device refers to a state in which the sample processing quality is currently above an acceptable level but has fallen below a set level and is predicted to move into an abnormal state within a specified period of time.

(First embodiment)
- Process diagnostic equipment -
Fig. 1 is a schematic diagram of a configuration example of a process diagnosis device 100 according to a first embodiment of the present invention. An XYZ orthogonal coordinate system with the Z axis extending vertically is defined as shown in Fig. 1. The process diagnosis device 100 inspects a sample W and detects defects such as adhesion of foreign matter and abnormal film formation on the surface of the sample W. The process diagnosis device 100 is a rotary scanning type device that scans the sample W by rotating it in a circumferential direction (θ direction) and moving it in a radial direction (r direction).

The process diagnosis device 100 includes a stage ST, an illumination optical system A, a plurality of detection optical systems Bn (n=1, 2, etc.), sensors Cn, Cn' (n=1, 2, etc.), a signal processing device D, a storage device DB, a control device E1, an input device E2, and a monitor E3.

-stage-
The stage ST is a device including a sample stage ST1 and a scanning device ST2. The sample stage ST1 is a stage that supports a sample W that has been processed in a process device such as a plasma etching device. The scanning device ST2 is a device that drives the sample stage ST1 to change the relative position between the sample W and the illumination optical system A, and is configured to include a translation stage, a rotation stage, and a Z stage, although detailed illustration is omitted. A rotation stage is mounted on the translation stage via the Z stage, and the sample stage ST1 is supported on the rotation stage. The translation stage translates in the horizontal direction together with the rotation stage. The rotation stage rotates (spins) around a rotation axis that extends vertically. The Z stage serves to adjust the height of the surface of the sample W.

FIG. 2 is a schematic diagram showing the scanning trajectory of the sample W by the scanning device ST2. As will be described later, the beam spot BS, which is the incident area of the illumination light emitted from the illumination optical system A on the surface of the sample W, is a minute point with a long illumination intensity distribution in one direction as shown in the figure. The long axis direction of the beam spot BS is s2, and the direction intersecting the long axis (for example, the short axis direction perpendicular to the long axis) is s1. The sample W rotates with the rotation of the rotating stage, and the beam spot BS is scanned in the s1 direction relative to the surface of the sample W, and the sample W moves in the horizontal direction with the translation stage translation, and the beam spot BS is scanned in the s2 direction relative to the surface of the sample W. The beam spot BS moves in the s2 direction by a distance equal to or less than the length of the beam spot BS in the s2 direction during one rotation of the sample W. By such an operation of the scanning device ST2, the sample W rotates and translates, as shown in FIG. 2, and the beam spot BS moves in a spiral trajectory from the center of the sample W to the outer edge or its vicinity, and the entire surface of the sample W is scanned.

The stage ST may be configured to have another translation stage, whose axis of movement extends in a direction intersecting the axis of movement of the translation stage in a horizontal plane, in place of (or in addition to) the rotation stage. In this case, as shown in FIG. 3, the beam spot BS scans the surface of the sample W by folding over a linear trajectory instead of a spiral trajectory. In the example shown in the figure, the first translation stage is driven in translation at a constant speed in the s1 direction, the second translation stage is driven in the s2 direction by a predetermined distance (for example, a distance equal to or less than the length of the beam spot BS in the s2 direction), and then the first translation stage is turned back in the s1 direction and driven in translation again. This causes the beam spot BS to repeat linear scanning in the s1 direction and movement in the s2 direction, scanning the entire surface of the sample W. Compared to this XY scanning method, the rotation scanning method of FIG. 2 does not involve a reciprocating motion that is repeatedly accelerated and decelerated, so the inspection time of the sample W can be shortened.

- Illumination optical system -
The illumination optical system A shown in Fig. 1 includes a group of optical elements for irradiating a desired illumination light onto a sample W placed on a sample stage ST1. As shown in Fig. 1, the illumination optical system A includes a laser light source A1, an attenuator A2, an emitted light adjustment unit A3, a beam expander A4, a polarization control unit A5, a focusing optical unit A6, reflection mirrors A7-A9, etc.

Laser Light Source The laser light source A1 is a unit that emits a laser beam as illumination light. When the process diagnosis device 100 detects minute defects near the surface of the sample W, the laser light source A1 is a unit that emits a high-power laser beam with an output of 2 W or more in ultraviolet or vacuum ultraviolet with a short wavelength (wavelength of 355 nm or less) that does not easily penetrate into the inside of the sample W. When the process diagnosis device 100 detects defects inside the sample W, the laser light source A1 is a unit that emits a visible or infrared laser beam with a long wavelength that easily penetrates into the inside of the sample W.

Attenuator FIG. 4 is a schematic diagram showing the attenuator A2. The attenuator A2 is a unit that attenuates the light intensity of the illumination light from the laser light source A1. In this embodiment, the attenuator A2 is a combination of a first polarizing plate A2a, a half-wave plate A2b, and a second polarizing plate A2c. The half-wave plate A2b is configured to be rotatable around the optical axis of the illumination light. The illumination light incident on the attenuator A2 is converted into linearly polarized light by the first polarizing plate A2a, and then the polarization direction is adjusted to the slow axis azimuth angle of the half-wave plate A2b and passes through the second polarizing plate A2c. The light intensity of the illumination light is attenuated at an arbitrary ratio by adjusting the azimuth angle of the half-wave plate A2b. If the degree of linear polarization of the illumination light incident on the attenuator A2 is sufficiently high, the first polarizing plate A2a can be omitted. For the attenuator A2, an attenuator in which the relationship between the incident illumination light and the light attenuation rate is calibrated in advance is used. It should be noted that the attenuator A2 is not limited to the configuration illustrated in FIG. 4, but can also be configured using an ND filter having a gradation density distribution, and can be configured such that the attenuation effect can be adjusted by combining multiple ND filters having different densities.

Emitted light adjustment unit The emitted light adjustment unit A3 shown in FIG. 1 is a unit that adjusts the angle of the optical axis of the illumination light attenuated by the attenuator A2, and in this embodiment, it is configured to include multiple reflecting mirrors A3a and A3b. The reflecting mirrors A3a and A3b are configured to sequentially reflect the illumination light, but in this embodiment, the incident and exit surfaces of the illumination light to the reflecting mirror A3a are configured to be perpendicular to the incident and exit surfaces of the illumination light to the reflecting mirror A3b. The incident and exit surfaces are surfaces that include the optical axis of the light incident on the reflecting mirror and the optical axis of the light emitted from the reflecting mirror. In a configuration in which the illumination light is incident on the reflecting mirror A3a in the +X direction, which is different from the schematic FIG. 1, for example, the illumination light changes its traveling direction to the +Y direction by the reflecting mirror A3a and then to the +Z direction by the reflecting mirror A3b. In this example, the incident and exit surfaces of the illumination light to the reflecting mirror A3a are the XY plane, and the incident and exit surfaces to the reflecting mirror A3b are the YZ plane. The reflecting mirrors A3a and A3b are provided with a mechanism (not shown) for translating the reflecting mirrors A3a and A3b and a mechanism (not shown) for tilting the reflecting mirrors A3a and A3b. The reflecting mirrors A3a and A3b move, for example, in parallel in the incident or outgoing direction of the illumination light with respect to themselves, and tilt around the normal to the incident and outgoing surfaces. This allows the offset amount and angle in the XZ plane and the offset amount and angle in the YZ plane to be independently adjusted for the optical axis of the illumination light emitted in the +Z direction from the outgoing light adjustment unit A3. In this embodiment, a configuration using two reflecting mirrors A3a and A3b is illustrated, but a configuration using three or more reflecting mirrors may be used.

Beam Expander The beam expander A4 is a unit that expands the diameter of the luminous flux of the incident illumination light, and has a plurality of lenses A4a and A4b. An example of the beam expander A4 is a Galilean type that uses a concave lens as the lens A4a and a convex lens as the lens A4b. The beam expander A4 is provided with a mechanism for adjusting the distance between the lenses A4a and A4b (zoom mechanism), and the expansion rate of the luminous flux diameter changes by adjusting the distance between the lenses A4a and A4b. If the illumination light incident on the beam expander A4 is not a parallel luminous flux, collimation (quasi-parallelization of the luminous flux) is also possible by adjusting the distance between the lenses A4a and A4b. However, the collimation of the luminous flux may be performed by a collimating lens installed upstream of the beam expander A4 separately from the beam expander A4.

Beam expander A4 is installed on a translation stage with two or more axes (two degrees of freedom) and is configured so that its position can be adjusted so that its center coincides with the incident illumination light. Beam expander A4 also has a swing angle adjustment function with two or more axes (two degrees of freedom) so that the incident illumination light coincides with the optical axis.

Although not specifically shown, the state of the illumination light entering the beam expander A4 is measured by a beam monitor midway along the optical path of the illumination optical system A.

Polarization control unit The polarization control unit A5 is an optical system that controls the polarization state of the illumination light, and is configured to include a half-wave plate A5a and a quarter-wave plate A5b. For example, when a reflecting mirror A7 described later is inserted in the optical path to illuminate the sample W obliquely, the amount of scattered light from the surface of the sample W can be increased compared to polarized light other than P-polarized light by making the illumination light P-polarized by the polarization control unit A5. When an oxide film is present on the surface of the sample W, the amount of scattered light from the sample surface can be increased more than P-polarized light by using S-polarized light depending on the material and thickness of the film. By selecting the polarization according to the sample W, it is possible to switch between conditions under which haze light is likely to occur and conditions under which it is difficult to occur, thereby improving the sensitivity of defect inspection and improving the sensitivity of haze light to sample characteristics. For example, when the state of the sample W is evaluated using the output of haze light, it is advantageous to use S-polarized light for the illumination light. It is also possible to use the polarization control unit A5 to make the illumination light circularly polarized or 45-degree polarized light intermediate between P-polarized light and S-polarized light.

Reflecting Mirror As shown in FIG. 1, the reflecting mirror A7 is moved in parallel in the direction of the arrow by a driving mechanism (not shown) and enters and exits the optical path of the illumination light toward the sample W. This switches the incidence path of the illumination light to the sample W. By inserting the reflecting mirror A7 into the optical path, the illumination light emitted from the polarization control unit A5 as described above is reflected by the reflecting mirror A7 and enters the sample W obliquely via the focusing optical unit A6 and the reflecting mirror A8. In this specification, the illumination light is made to enter the sample W from a direction oblique to the normal to the surface of the sample W in this way, which is referred to as "oblique incidence illumination". On the other hand, when the reflecting mirror A7 is removed from the optical path, the illumination light emitted from the polarization control unit A5 is made to enter the sample W perpendicularly via the reflecting mirror A9, the polarizing beam splitter B'3, the polarization control unit B'2, the reflecting mirror B'1, and the detection optical system B3. In this specification, the illumination light is made to enter the sample W perpendicularly to the surface of the sample W in this way, which is referred to as "vertical illumination".

FIGS. 5 and 6 are schematic diagrams showing the positional relationship between the optical axis of the illumination light guided obliquely to the surface of the sample W by the illumination optical system A and the illumination intensity distribution shape. FIG. 5 shows a schematic cross-section of the sample W cut at the plane of incidence of the illumination light incident on the sample W. FIG. 6 shows a schematic cross-section of the sample W cut at a plane that is perpendicular to the plane of incidence of the illumination light incident on the sample W and includes the normal to the surface of the sample W. The plane of incidence is a plane that includes the optical axis OA of the illumination light incident on the sample W and the normal to the surface of the sample W. Note that FIGS. 5 and 6 show only a portion of the illumination optical system A, and for example, the exit light adjustment unit A3 and the reflecting mirrors A7 and A8 are not shown.

As mentioned above, when the reflecting mirror A7 is inserted into the optical path, the illumination light emitted from the laser light source A1 is collected by the collecting optical unit A6, reflected by the reflecting mirror A8, and obliquely incident on the sample W. In this way, the illumination optical system A is configured to make the illumination light obliquely incident on the surface of the sample W. This oblique incidence illumination has its light intensity adjusted by the attenuator A2, its light beam diameter adjusted by the beam expander A4, and its polarization adjusted by the polarization control unit A5, so that the illumination intensity distribution is uniform within the incident surface. As shown in the illumination intensity distribution (illumination profile) LD1 shown in Figure 5, the beam spot formed on the sample W has a Gaussian light intensity distribution in the s2 direction.

In the plane perpendicular to the incident plane and the sample surface, the beam spot has a light intensity distribution with weak intensity at the periphery relative to the center of the optical axis OA, as shown in the illumination intensity distribution (illumination profile) LD2 in Figure 6. This light intensity distribution is, for example, a Gaussian distribution that reflects the intensity distribution of the light incident on the focusing optical unit A6, or an intensity distribution similar to a first-order Bessel function of the first kind or a sinc function that reflects the aperture shape of the focusing optical unit A6.

In addition, the angle of incidence of the oblique incidence illumination on the sample W (the tilt angle of the incident optical axis with respect to the normal to the sample surface) is adjusted to an angle suitable for detecting minute defects by adjusting the positions and angles of the reflecting mirrors A7 and A8. The angle of the reflecting mirror A8 is adjusted by an adjustment mechanism A8a. For example, the greater the angle of incidence of the illumination light on the sample W (the smaller the illumination elevation angle between the sample surface and the incident optical axis), the weaker the haze light that becomes noise in the scattered light from minute defects on the sample surface.

-Detection optical system-
The detection optical systems Bn (n=1, 2, ...) are units that collect scattered light from the sample surface, and are configured to include multiple optical elements including a collecting lens (objective lens). n in the detection optical systems Bn is the number of detection optical systems, and the process diagnosis device 100 of this embodiment is provided with 13 detection optical systems (n=13). However, the number of detection optical systems Bn is not limited to 13 and can be increased or decreased as appropriate. In addition, the layout of the apertures (objective lenses) of the detection optical systems Bn can also be changed as appropriate.

Fig. 7 is a diagram showing the area where the detection optical system Bn collects scattered light as viewed from above, which corresponds to the arrangement of each objective lens of the detection optical system Bn. Fig. 8 is a diagram showing the zenith angles of the low-angle and high-angle optical systems of the detection optical system Bn, Fig. 9 is a plan view showing the azimuth angle of the low-angle detection optical system, and Fig. 10 is a plan view showing the azimuth angle of the high-angle detection optical system.

In the following explanation, the incident direction of the oblique incidence illumination on the sample W is used as a reference, and the traveling direction of the incident light with respect to the beam spot BS on the surface of the sample W when viewed from above (to the right in Figure 7) is referred to as the front, and the opposite direction (to the left in the same figure) is referred to as the rear. The lower side in the figure with respect to the beam spot BS is the right side, and the upper side is the left side. Furthermore, the angle φ2 (Figure 8) that the incident optical axis (center line of the aperture) of each detection optical system Bn makes with the normal N (Figure 8) of the sample W that passes through the beam spot BS is described as the zenith angle. Furthermore, the angle φ1 (Figures 9 and 10) that the incident optical axis (center line of the aperture) of each detection optical system Bn makes with the incident plane of the oblique incidence illumination in a planar view is described as the azimuth angle.

As shown in Figures 7 to 10, the detection optical systems Bn are arranged so that their orientations with respect to the beam spot BS are different. In this embodiment, the objective lenses (apertures α1-α6, β1-β6, γ) of the detection optical system Bn are arranged along the upper hemispherical surface of a sphere (celestial sphere) centered on the beam spot BS on the sample W. The light incident on the apertures α1-α6, β1-β6, γ is focused by the corresponding detection optical system Bn.

Aperture γ overlaps the zenith (intersects with normal N) and is located directly above the beam spot BS formed on the surface of sample W.

　The openings α1-α6 are opened at a low angle so as to equally divide an annular area surrounding 360 degrees around the beam spot BS. The openings α1-α6 are arranged in the order of openings α1, α2, α3, α4, α5, α6 in a counterclockwise direction from the incident direction of the oblique incidence illumination in a plan view. The openings α1-α6 are also laid out to avoid the incident light path of the oblique incidence illumination and the regular reflection light path. The openings α1-α3 are arranged on the right side of the beam spot BS, the opening α1 is located to the right rear of the beam spot BS, the opening α2 is located to the right, and the opening α3 is located to the right front. The openings α4-α6 are arranged on the left side of the beam spot BS, the opening α4 is located to the left front of the beam spot BS, the opening α5 is located to the left, and the opening α6 is located to the left rear. The arrangement of the openings α4, α5, α6 is symmetrical to the openings α3, α2, α1 with respect to the incident plane of the oblique incidence illumination.

The openings β1-β6 are opened so as to equally divide an annular area surrounding 360 degrees around the beam spot BS at a high angle (between the openings α1-α6 and the opening γ). The openings β1-β6 are arranged in the order of the openings β1, β2, β3, β4, β5, and β6 in a counterclockwise direction from the incidence direction of the oblique incidence illumination in a plan view. Among the openings β1-β6, the openings β1 and β4 are laid out at a position intersecting the incidence plane, with the opening β1 located behind the beam spot BS and the opening β4 located in front. The openings β2 and β3 are arranged on the right side of the beam spot BS, with the opening β2 located to the right rear of the beam spot BS and the opening β3 located to the right front. Apertures β5 and β6 are located to the left of the beam spot BS, with aperture β5 located in front of the beam spot BS and aperture β6 located to the rear of the beam spot BS. Scattered light from the beam spot BS in various directions enters apertures α1-α6, β1-β6, and γ, is collected by the detection optical system Bn, and is guided to the corresponding sensors Cn and Cn'.

FIG. 11 is a schematic diagram showing an example of the configuration of the detection optical system. In the process diagnosis device of this embodiment, each detection optical system Bn (or a part of the detection optical system) is configured as shown in FIG. 11, and the polarization direction of the scattered light that is transmitted can be controlled by the polarizing plate Bb. Specifically, the detection optical system Bn includes an objective lens (collecting lens) Ba, a polarizing plate Bb, a polarizing beam splitter Bc, imaging lenses (tube lenses) Bd, Bd', field stops Be, Be', and sensors Cn, Cn'.

The scattered light incident on the detection optical system Bn from the sample W is collected and collimated by the objective lens Ba, and its polarization direction is controlled by the polarizing plate Bb. The polarizing plate Bb is a half-wave plate that can be rotated by a driving mechanism (not shown). The driving mechanism is controlled by the control device E1, and the polarization direction of the scattered light incident on the sensor is controlled by adjusting the rotation angle of the polarizing plate Bb.

The scattered light, whose polarization has been controlled by the polarizing plate Bb, has its optical path split by the polarizing beam splitter Bc according to the polarization direction and enters the imaging lenses Bd and Bd'. The combination of the polarizing plate Bb and the polarizing beam splitter Bc cuts linearly polarized light components in any direction. To cut any polarized light component, including elliptically polarized light, the polarizing plate Bb is composed of a quarter-wave plate and a half-wave plate that can be rotated independently of each other.

The scattered illumination light that passes through the imaging lens Bd and is collected is photoelectrically converted by the sensor Cn via the field diaphragm Be, and the detection signal is input to the signal processing device D. The scattered illumination light that passes through the imaging lens Bd' and is collected is photoelectrically converted by the sensor Cn' via the field diaphragm Be', and the detection signal is input to the signal processing device D. The field diaphragms Be, Be' are installed so that their centers are aligned with the optical axis of the detection optical system Bn, and cut out light generated from positions other than the position to be inspected, such as light generated from positions away from the center of the beam spot BS of the sample W and stray light generated inside the detection optical system Bn. This has the effect of suppressing noise that interferes with defect detection.

The above configuration makes it possible to simultaneously detect two mutually orthogonal polarized components of scattered light generated at the same coordinates, which is effective when detecting multiple types of defects or haze light with different polarization characteristics.

When constructing the objective lens Ba from multiple closely-spaced lenses, in order to reduce loss of detected light due to gaps between the lenses, the outer periphery of the objective lens Ba may be cut out so as not to interfere with the sample W or other objective lenses, as in the example of Figure 11.

-Sensor-
The sensors Cn and Cn' convert the scattered light collected by the corresponding detection optical system into an electric signal and output the detection signal. The sensors C1 (C1'), C2 (C2'), C3 (C3') ... correspond to the detection optical systems B1, B2, B3 .... For these sensors Cn and Cn', single-pixel point sensors such as photomultiplier tubes and SiPM (silicon photomultiplier tubes) that photoelectrically convert weak signals with high gain can be used. In addition, sensors in which multiple pixels are arranged one-dimensionally or two-dimensionally, such as CCD sensors, CMOS sensors, and PSDs (position sensing detectors), may be used for the sensors Cn and Cn'. The detection signals output from the sensors Cn and Cn' are input to the signal processing device D as needed.

-Control device-
The control device E1 is a computer that controls the process diagnosis device 100, and includes a processing device (arithmetic control device) such as a CPU, a GPU, and an FPGA in addition to a ROM, a RAM, and other storage devices. The control device E1 is connected to the input device E2, the monitor E3, and the signal processing device D by wire or wirelessly. The input device E2 is a device through which a user inputs settings of inspection conditions, etc. to the control device E1, and various input devices such as a keyboard, a mouse, and a touch panel can be appropriately adopted. The control device E1 receives the output of the encoder of the rotation stage and the translation stage (rθ coordinate of the beam spot BS on the sample), and the inspection conditions input by the operator via the input device E2. The inspection conditions include the type, size, shape, material, illumination conditions, detection conditions, etc. of the sample W, as well as, for example, the sensitivity settings of each sensor Cn, Cn', and gain values and threshold values used for defect judgment and process diagnosis (state diagnosis of the process device).

The control device E1 also outputs command signals to command the operation of the stage ST, illumination optical system A, etc. according to the inspection conditions, and outputs coordinate data of the beam spot BS synchronized with the defect detection signal to the signal processing device D. The control device E1 also displays and outputs an inspection condition setting screen and sample inspection data (inspection image, etc.) on the monitor E3. In addition to the final inspection result obtained by combining the signals from each sensor Cn, Cn', the inspection data can also display the individual inspection results from these sensors Cn, Cn'.

Also, as shown in Figure 1, the control device E1 may be connected to a Review SEM (Review Scanning Electron Microscope), which is an electron microscope used for defect inspection. In this case, the control device E1 can receive data on the defect inspection results from the Review SEM and transmit it to the signal processing device D.

The control device E1 can be configured as a single computer that forms a unit with the device body (stage, illumination optical system, detection optical system, sensors, etc.) of the process diagnosis device 100, but it can also be configured as multiple computers connected via a network. For example, the inspection conditions can be input to a computer connected via a network, and a computer attached to the device body can be configured to control the device body and the signal processing device D.

- Signal processing device -
The signal processing device D is a computer having a function of processing detection signals input from the sensors Cn, Cn' of the detection optical system Bn to detect defects in the sample W. The signal processing device D is configured to include a memory D1 (FIG. 12) including at least one of RAM, ROM, HDD, SSD, and other storage devices, as well as a processing device such as a CPU, GPU, or FPGA, just like the control device E1. The signal processing device D can be configured as a single computer that forms a unit with the device body (stage, illumination optical system, detection optical system, sensor, etc.) of the process diagnosis device 100, but can also be configured as multiple computers connected by a network. For example, a configuration can be adopted in which a computer attached to the device body acquires defect detection signals from the device body, processes the detection data as necessary and transmits it to a server, and the server executes processes such as defect detection and classification. A configuration in which the signal processing device D and the control device E1 are both performed by a single computer is also conceivable.

FIG. 12 is an example of a functional block diagram of the main parts of a signal processing device D provided in a process diagnosis device according to a first embodiment of the present invention. As shown in FIG. 12, the signal processing device D includes a memory D1, a defect determination circuit D2, a low-pass filter circuit D3, and a process diagnosis circuit D4.

While the process diagnostic device 100 is scanning the sample W, the signal processing device D receives detection signals (scattered light intensity signals) from the sensors Cn and Cn' and the encoder output of the stage ST (rθ coordinate of the beam spot BS on the sample) from the control device E1. In the signal processing device D, these detection signals and encoder outputs are associated with each other and recorded in the memory D1.

The defect judgment circuit D2 reads out the detection signals input from the sensors Cn, Cn' from the memory D1 in chronological order, sequentially judges whether these detection signals are defect signals indicating detected defects, records the judgment results in the memory D1 or the storage device DB, and also outputs them to the control device E1. In the defect judgment circuit D2, for example, high-frequency components of the detection signals are extracted as defect signals relating to defects such as foreign matter. High-frequency components are components with high fluctuating frequencies, specifically components whose time fluctuations exceed a preset value. The control device E1 displays and outputs the judgment results on the monitor E3 automatically, or in response to operation signals from the input device E2 input in conjunction with the operation of the operator.

The low-pass filter circuit D3 reads out the detection signals from the sensors Cn and Cn' from the memory D1 in chronological order, extracts the haze signals excluding the defect signals for each region of the sample W, and creates a haze map, which is the light intensity distribution of the entire surface, by adding coordinate information to the haze signals on the surface of the sample W. The haze signal refers to the low-frequency components of the light signals obtained from the sample, and is a signal that is mainly caused by the characteristics of the sample. Here, for example, the low-frequency components of the detection signal, that is, components whose fluctuation frequency (time fluctuation of the signal intensity) is lower than a preset value, are extracted as the haze signal. Since the optical defect inspection device is capable of high-speed scanning, it is possible to extract the haze signal for the entire surface of the sample, or a haze map based on it. However, there may be cases where it is desired to diagnose the sample partially, rather than the entire surface. In this case, the region from which the haze signal is extracted may be a sampling point, or the haze signal may be extracted by dividing the region into regions partitioned by a mesh of any mesh size.

When extracting haze signals by dividing areas into regions defined by a mesh, multiple haze signals are obtained for each region. The statistical values (average, median, etc.) of these multiple haze signals can be used as the haze signal for that region. One side of the mesh that divides the region can be set to, for example, 1 mm to several mm. Depending on the size of the mesh, for example, with a 1 mm mesh, the sample surface is divided into more than 60,000 regions, and a detailed haze map is generated. However, the smaller the mesh size, the better; by setting the mesh to a large size within a necessary and sufficient range, the calculation load on the signal processing device D can be reduced in accordance with the reduction in the number of regions on the sample surface.

The process diagnostic circuit D4 diagnoses the state of the process equipment that processed the sample W based on the haze signal extracted by the low-pass filter circuit D3. When the process diagnostic circuit D4 predicts an abnormality or a sign of an abnormality in the process equipment, it outputs an alarm related to the state of the process equipment (for example, a screen or a warning sound indicating an abnormal part of the process equipment and recommending its inspection) to the monitor E3 via the control device E1. The abnormality or a sign of an abnormality in the process equipment is predicted by comparing the haze signal of the entire surface or each area of the sample W with the reference haze signal of the corresponding area. When the difference between the measured haze signal and the reference haze signal exceeds a preset value, an abnormality or a sign of an abnormality is predicted. The haze signal used in this comparison may be not only the haze signal itself, but also another output form such as a haze map.

The diagnostic results of the process equipment (including normal judgments in addition to the above alarms) are recorded in memory D1 or storage device DB and output to control device E1. Control device E1 displays and outputs the diagnostic results on monitor E3 in response to an operation signal from input device E2 input in conjunction with an operator's operation or automatically.

Furthermore, from the viewpoint of early detection of abnormalities or signs of abnormalities in the process equipment, it is desirable to perform process diagnosis by sequentially processing the detection signals input from the sensors Cn, Cn' in association with the defect inspection of the sample W by the signal processing device D. However, it is also possible to configure the system so that the detection signals acquired by scanning the entire surface of the sample W are temporarily stored in the storage device DB, and the stored data is post-processed at a desired timing (for example, at a fixed time every day) to perform process diagnosis.

- Reference haze signal -
In this embodiment, the reference haze signal used for the process diagnosis is defined for each of the sensors Cn and Cn′ for the entire surface of the sample W or for each predetermined region (rθ coordinate in this embodiment) and is stored in, for example, a storage device DB. The reference haze signal defined for the same sensor may be different for each region of the sample W.

The reference haze signal is set, for example, by scanning a reference sample. The reference sample is a sample that meets standards in a quality inspection, and is preferably a sample of the same type as sample W and from the same process as sample W. However, the reference haze signal can also be set by daily calculating statistical data (e.g., average value, median value) of haze signals obtained in defect inspections in the semiconductor manufacturing process for samples that are determined to be non-defective, rather than measuring the reference sample. Alternatively, the reference haze signal can be set by simulating the haze signal that can be obtained for each detection optical system Bn based on the design data of sample W.

When the reference haze signal is set by actual measurement of the reference sample, for example, one or more reference samples are prepared and scanned by the process diagnostic device 100. This scan may be performed under the same conditions as the inspection of the sample W, or, for example, when there is a concern about damage to the sensors Cn, Cn', adjustments are allowed as necessary, such as performing the scan under lower sensitivity conditions than the inspection of the sample W. From the detection signal obtained for the reference sample by this scan, a haze signal is extracted by the low-pass filter circuit D3 of the signal processing device D as described in FIG. 12, and a reference haze is set based on this haze signal and stored in the storage device DB. At this time, the reliability of the reference haze signal as a reference value is increased by using statistical data of the reference haze signals obtained for multiple reference samples as the reference haze signal.

-Correlation data-
Further, in addition to the reference haze, the storage device DB prestores a correlation between the detection optical system Bn (in other words, the emission direction of the haze light) and the fluctuation factor of the haze signal.

In the semiconductor manufacturing process, there are several dozen steps in which plasma processing is performed. Each step uses different materials (such as the film quality of the sample and the type of gas used in the chamber) and processing conditions. In some steps, the surface roughness of the sample W after etching is likely to change, while in other steps the surface film thickness is likely to change. In other cases, a characteristic tendency may be observed in accordance with the direction of gas flow in the chamber of the process equipment. The "parameters that are likely to vary" differ for each step. Therefore, it is desirable for the process diagnostic device 100 to diagnose the process equipment using the signal of a detector that is likely to detect changes in the parameters that are likely to vary.

For example, when a sample W is processed in a plasma processing device, the surface roughness of the sample W varies. If this variation is within a certain range, the variation tends to be mainly manifested in the intensity of scattered light incident on the openings α3 and α4 located in the direction of specular reflection of the illumination light relative to the beam spot BS. Similarly, if the surface film thickness of the sample W varies within a certain range, the variation tends to be mainly manifested as a change in the intensity of scattered light incident on the openings α1 and α6 located on the opposite side of the direction of specular reflection relative to the beam spot BS. Based on this relationship, for each process using the plasma processing device, a variation factor is associated with a detection optical system that is highly sensitive to the variation factor, and the correlation data is stored in the storage device DB. A specific detection optical system Bn is selected in advance by the signal processing device D based on this correlation data, and the changes in the surface roughness and surface film thickness are monitored based on the haze signal of the selected detection optical system Bn, thereby detecting abnormalities or signs of abnormalities in related parts of the processing device (e.g., the plasma light source) with high sensitivity.

The above correlation data is merely one example of the correlation between the fluctuation factors of the haze signal and the detection optical system Bn. For example, if it is found that some abnormality or sign of the process device appears in the haze signal acquired by a detection optical system other than the detection optical system Bn corresponding to the openings α3, α4, α1, and α6, the correlation data can be set based on that knowledge. Furthermore, some abnormality or sign of the process device may appear not only in the haze signal detected individually by the detection optical system Bn, but also in the difference or sum of the haze signals detected by multiple detection optical systems Bn. In this case, the correlation data based on the correlation is defined, and the difference or sum of the haze signals detected by multiple detection optical systems Bn can be compared as one form of haze signal with the difference or sum of the reference haze signals related to the same set of detection optical systems Bn, and used to diagnose the process device.

During process diagnosis, the signal processing device D reads the correlation data from the storage device DB and automatically selects the detection optical system that correlates with the state of the process device to be diagnosed. However, the signal processing device D can also select the detection optical system according to the specification made by the operator via the input device E2. Then, the signal processing device D detects the change in the state of the process device based on the haze signal output from the selected detection optical system. For example, following the example described above, the signal processing device D selects the detection optical system Bn corresponding to the openings α3, α4, α1, and α6 located in the direction of regular reflection of the illumination light with respect to the beam spot BS. From the difference between the haze signal detected by the detection optical system Bn corresponding to the openings α3 and α4 and the reference haze signal, a change in the surface roughness of the sample W within a predetermined range, and thus a change in the state of the process device related to the change in the surface roughness within this predetermined range, is detected. From the difference between the haze signal of the detection optical system Bn corresponding to the openings α1 and α6 and the reference haze signal, a change in the surface film thickness of the sample W within a predetermined range, and thus a change in the state of the process device related to the change in the surface film thickness within this predetermined range is detected.

Furthermore, even if the haze light is incident on the same detection optical system Bn, the way in which the fluctuations occur according to the state of the process equipment may differ depending on the polarization direction. In this embodiment, the detection optical system Bn is equipped with a polarizing beam splitter Bc that splits the light according to the polarization direction, and multiple sensors Cn, Cn' that detect the light with different polarization directions split by the polarizing beam splitter Bc (FIG. 11). Therefore, in this embodiment, in each detection optical system Bn, two haze signals with different polarization directions can be obtained for the same coordinates on the sample. Therefore, the polarization direction of the haze light can be added as a parameter of the correlation data between the above-mentioned fluctuation factors of the haze signal and the detection optical system Bn, and the relationship between the haze signal and its fluctuation factors can be specified more precisely and stored in the storage device DB. In this way, the number of parameters of the correlation data is increased, and the state change of the process equipment can be detected more precisely from the haze signal.

There may also be a correlation between deteriorated parts of the process equipment and the position on the sample. On the sample W processed by the plasma process equipment, the haze map can identify areas where there is a large change compared to the sample from which the reference haze signal was obtained (areas where there is a high possibility of some kind of defective workmanship occurring). The storage device DB also stores part information (position in the equipment, material, etc.) and maintenance information (for example, maintenance cycle, date of actual part replacement, etc.) on the process equipment side. The signal processing device D can analyze the correlation between the position of the area identified by the haze map on the sample and the parts of the process equipment, and store the data in advance in the storage device DB. In this case, the signal processing device D can identify not only the presence or absence of an abnormality or a sign of an abnormality in the process equipment, but also the parts of the process equipment where an abnormality or a sign of an abnormality is suspected or the parts recommended for inspection, and notify the operator together with an alarm. By comparing the diagnosis results of the process diagnostic device with the part information and maintenance information on the process equipment side, the user can determine the most optimal maintenance time.

The storage device DB typically stores not only haze maps but also defect information as information output by the process diagnosis device 100. As a defect inspection device, the process diagnosis device 100 can output defect information for the entire surface or specific areas (e.g., coordinates and brightness of each defect, a defect map for the entire surface of the sample, etc.) by scanning the sample W. The signal processing device D of the process diagnosis device 100 can also analyze the correlation with the process device by combining the haze map and defect information.

-Semiconductor manufacturing process-
13 is a flow chart generally showing an example of a series of diagnostics of a process device in a semiconductor manufacturing process. Here, a plasma etching device will be taken as an example for explanation.

In the diagram, a wafer is first plasma etched (step S10) in a plasma etching device (processing device). The plasma etching device monitors fluctuations in its own operating parameters during the plasma etching process. For example, the plasma etching device is equipped with optical emission spectrometry (OES) to monitor the state of the plasma light source. In addition, the plasma etching device has sensors installed in various places inside the chamber to monitor the state inside the plasma chamber. If the plasma etching device detects an abnormality in the behavior of the operating parameters that exceeds the acceptable range based on a specified algorithm, it will output an alarm and stop itself.

The wafers processed in the plasma etching equipment have their CD (Critical Dimension) measured in-line, for example, using a CD-SEM or OCD (Optical Critical Dimension) (Step S20). "In-line" means as one step in semiconductor manufacturing, or in the course of semiconductor research, development, and manufacturing. The OES is installed in the plasma etching equipment. . The etching rate (surface film thickness) of the wafer surface is also inspected in-line, for example, using an AFM or spectroscopic ellipsometer (Step S40). Furthermore, defects such as foreign objects on the wafer surface are also inspected in-line using an optical inspection device (Step S60). The order of inspection of the CD value, etching rate, and defects can be changed as desired.

Anomalies in the plasma etching equipment also affect the condition of the wafers processed by the plasma etching equipment. Typical examples include effects on the CD value, surface film thickness, and number of defects of the processed wafers. This combines wafer quality inspection and plasma etching equipment diagnosis to determine whether the CD value and etching rate inspection results are within the standard range (steps S30 and S50). If the inspection results are outside the standard range, the plasma etching equipment suspected to be abnormal is stopped, and the abnormality in the plasma etching equipment is resolved through processes such as cause identification, countermeasures, and operational tests. Once it is confirmed that the abnormality in the plasma etching equipment has been resolved, the plasma etching equipment is restarted (step S80) and the next wafer processing is started.

In this embodiment, in step S60, the process diagnostic device 100, which is also an optical defect inspection device, is used to diagnose the process device in parallel with the wafer defect inspection, and it is determined whether the diagnosis result is within a standard range (step S70). If the diagnosis result is outside the standard range, the plasma etching device in which the abnormality or symptom has appeared is stopped, and the abnormality in the plasma etching device is resolved through processes such as cause identification, countermeasures, and operational tests. Once it is confirmed that the abnormality in the plasma etching device has been resolved, the plasma etching device is restarted (step S80), and the process moves to the next wafer processing.

The procedure for process diagnosis of the process device by the process diagnosis device 100 in steps S60 and S70 will be described later.

- Changes in the state of process equipment -
Fig. 14 is a schematic diagram showing the change over time in the state of the process equipment, and Fig. 15 is a schematic diagram showing the influence on the wafer caused by the change over time in the state of the process equipment, focusing on part XV in Fig. 14. The horizontal axis of Fig. 14 represents time, and the vertical axis represents the state of the process equipment (here, the plasma discharge state of the plasma etching equipment).

As shown in Figure 14, if the plasma state is in normal state a, which is better than a certain level (the abnormal line in Figure 14), many good wafers that meet the standards for CD value, surface film thickness, etc. are manufactured. However, the plasma state deteriorates as the operation time of the plasma etching equipment passes, and as a result, the quality of the processed wafers also deteriorates. For this reason, plasma etching equipment is regularly maintained (PM: Preventive Maintenance). However, the time until the next regular maintenance PM is not always appropriate, and the deterioration of the plasma light source, etc. may exceed a certain level before the next maintenance time arrives, causing an abnormality. If the plasma state deteriorates and falls into abnormal state c, the CD value and surface film thickness of the processed wafers, which are monitored daily by the user, will no longer meet the standards. The yield will also decrease.

As mentioned above, changes in the plasma state of a plasma etching device are expressed as changes in the state of the wafer. Changes in the wafer state are not necessarily linked one-to-one to specific physical properties such as the CD value, surface film thickness, or number of defects. In the case of changes that are too small to be measured as specific physical properties or changes due to multiple factors, they cannot be grasped by the CD value, etching rate, or number of foreign objects. The process diagnostic device 100 of the present invention is used to measure such changes. In particular, abnormalities due to plasma deterioration may show characteristic distributions in the position of the plasma in the plasma etching device or the direction of gas flow in the chamber. Such characteristic distributions tend to appear locally on the entire surface of the sample processed by the plasma etching device. The part of the entire surface of the sample that is localized varies depending on the type of etching device. Even if there are changes on the sample that cannot be linked one-to-one to the physical properties of the characteristics such as the CD value, surface film thickness, or number of defects, the plasma diagnostic device of this embodiment can measure the entire surface of the sample and detect plasma deterioration from the appearance of the distribution. Moreover, it can be detected earlier than it appears in the physical properties of the characteristics such as the CD value, surface film thickness, or number of defects.

Figure 15 shows examples of haze maps (haze distribution on the sample surface) obtained by the process diagnosis device 100, using a wafer processed in a plasma etching device as sample W when the plasma state is normal state a, transient state b, or abnormal state c in Figure 14. Transient state b is a state in which the yield has not fallen below the tolerance at present, but has fallen to a degree that is expected to fall below the tolerance within a specified period of time. In Figure 14, transient state b is a state in which the plasma discharge state has fallen below (but above) the abnormality line, for example, a warning line set a specified margin higher than the abnormality line. Transient state b corresponds to a state in which warning signs of an abnormality are seen.

Looking at the haze map obtained by the detection optical system Bn, which has high sensitivity to changes in the haze signal, as shown in FIG. 15, a difference appears in the intensity of the haze signal at the edge (outer edge) X1 and local area X2 between the sample W processed in the normal state a and the sample W processed in the transient state b. Assuming that the haze map of the sample W processed in the normal state a is close to the haze map of the reference haze signal, the difference from the haze map of the reference haze signal becomes even larger than in the transient state b when the plasma state further deteriorates and falls into the abnormal state c. In the abnormal state c, the yield falls below the allowable value. Therefore, the signal processing device D timely determines the transient state b in the in-line inspection of the sample W by the process diagnosis device 100, and recommends the operator to perform maintenance at the transient stage b before the plasma state falls into the abnormal state c. In other words, the signal processing device D detects the change in the state of the process device that appears as the microscopic surface shape of the sample W based on the difference between the haze signal and the reference haze signal during daily defect inspection.

- Process diagnosis -
Fig. 16 is a flowchart showing the procedure of diagnostic processing of a process device by the process diagnostic device 100 according to the present embodiment. The processing in Fig. 16 is executed in steps S60 and S70 of the flowchart in Fig. 13. The flow in Fig. 16 is executed in steps S60 and S70 of Fig. 13.

When inspecting a sample W processed by a process device for defects, the signal processing device D of the process diagnosis device 100 reads the reference haze from the storage device DB (step S61). The signal processing device D extracts a haze signal from the detection signal detected by the defect inspection (full surface scan) using the low-pass filter circuit D3, and inputs the haze signal for an arbitrary region of the sample W (step S62). The arbitrary region may be the entire surface of the sample W, or may be limited to a specific region within the sample W. Alternatively, the sample W may be divided into a plurality of regions, and the inspection may be performed for each region in sequence. The signal processing device D compares the haze signal input in step S62 with the reference haze for the relevant region (step S63), and determines whether the difference between the two is within a preset value (step S64). If the difference between the haze signal and the reference haze is within the preset value, the signal processing device D records the relevant region as a normal region (step S65), and if it exceeds the set value, records the relevant region as a defective region (step S66). For ease of explanation, areas that show a large change from the sample from which the reference haze signal was obtained will be referred to as "defective areas."

For example, when an arbitrary region is taken as the entire surface of the sample W, the haze map can be regarded as a pattern, and the form in which the difference from the reference haze is manifested can be used as the judgment criterion. For example, the surface of the sample W can be divided into a circular central portion and an annular outer peripheral portion surrounding it, and a judgment criterion can be applied in which focus is placed on changes in the haze signal in a specific portion of the central portion and outer peripheral portion, for example the outer peripheral portion, and a judgment criterion is applied in which a certain level of change in the outer peripheral portion is judged to be a defective region. In addition, a certain level of change in the haze signal in both the central portion and the outer peripheral portion can also be used as the judgment criterion.

When the sample W is divided into multiple regions, the signal processing device D repeats the processing of steps S61-S66 for each region, and once processing has been performed for all regions of the sample W, it determines whether the diagnosis results are within the standard range, specifically, whether the number of defective regions is below a preset tolerance (step S71).

The processing of steps S62-S66 can also employ an algorithm that creates a haze map of sample W and then compares the haze map of sample W with the intensity distribution of the reference haze signal (haze map related to the reference haze signal). For example, it is determined whether the brightness difference between the haze map of sample W and the reference haze (in this case, this refers to the haze map) exceeds a set value. Alternatively, an algorithm that counts the areas that exceed a set value may be used.

The signal processing device D estimates that the process device is in a normal state if the difference from the reference haze over the entire surface of the sample W is equal to or less than a set value, or if the number of defective areas is equal to or less than a tolerance. In this case, the signal processing device D generates data indicating that the process device is in a normal state, records the data in, for example, the storage device DB, and displays the data on the monitor E3 via the control device E1, and ends the procedure (step S72). On the other hand, if the difference from the reference haze exceeds a set value, or the number of defective areas exceeds a tolerance, an abnormality or a sign of an abnormality is suspected in the process device. In this case, the signal processing device D generates alarm data indicating that an abnormality or a sign of an abnormality is occurring in the process device, records the data in, for example, the storage device DB, and outputs the data to the monitor E3, and ends the procedure (step S72). By recommending maintenance of the process device to the operator through the alarm output to the monitor E3, the procedure proceeds to step S80 in FIG. 13, and maintenance of the process device is performed.

In addition to sound and text messages, the output to monitor E3 in steps S72 and S73 can also display a haze map for sample W. Furthermore, if changes in characteristics that appear in sample W (e.g., changes in surface roughness or surface film thickness), deteriorated parts of the process equipment, or areas that need to be inspected are identified by signal processing device D, this data can be output to monitor E3 as additional information together with an alarm.

-effect-
(1) According to this embodiment, an abnormality or a sign of an abnormality in a process device can be detected by a haze signal obtained by scanning the entire surface of the sample W during defect inspection of the sample W. It is also possible to perform process diagnosis at high speed in parallel with defect inspection by a single scan. The haze signal is not converted into characteristic values of the sample W such as surface roughness and surface film thickness and compared with characteristic values actually measured in advance by AFM or the like, but is evaluated by the difference from a reference haze, which is the same haze signal, that is, the relative change in haze light over time. In order to set a value (reference haze) that serves as a judgment standard, no correlation check with a separate measuring device (ellipsometry, AFM, etc.) is required in addition to the process diagnosis device 100. In addition, since both the reference haze and the haze signal to be compared therewith are obtained by actually measuring the entire surface of the sample W, changes that appear in the sample W according to the state of the semiconductor plasma device can be detected with high precision based on the intensity distribution (haze map) of the haze light on the entire surface of the sample. In particular, since plasma discharge abnormalities in a plasma etching apparatus appear locally within the surface of the sample W, the process diagnosis of this embodiment, which can grasp the in-plane distribution of haze light intensity fluctuations, is particularly effective.

The process diagnostic device 100 acquires a haze signal in conjunction with the daily in-line inspection of samples W for defects, and detects relative changes in the haze signal to identify abnormalities or signs of abnormalities in the process equipment. The results of the diagnosis of the semiconductor manufacturing process based on this haze light are obtained automatically in conjunction with the defect inspection, and do not impose an increased workload on the operator or inspection costs. Since abnormalities in the process equipment are detected early in the defect inspection performed daily during the semiconductor manufacturing process, the occurrence of defective samples W is suppressed, and the occurrence of samples W that are used for defect cause analysis using a TEM or the like is also suppressed. This also leads to avoiding operation stoppages due to failures in the process equipment, and is expected to improve the operating rate of the process equipment.

In addition, because the change in the intensity distribution of the haze light on the sample W can be seen, when an abnormality or a sign of an abnormality is inferred from the correlation between the effects that appear depending on the state of the semiconductor plasma device and the position on the sample, it is also possible to infer the defective part of the process device depending on the conditions.

As described above, according to this embodiment, the state of the process equipment can be accurately diagnosed from the data obtained by defect inspection of the sample W after processing, the occurrence of abnormalities in the process equipment can be suppressed, and the semiconductor yield can be improved.

(2) Multiple detection optical systems Bn are arranged in different directions relative to the beam spot BS, and one or more detection optical systems Bn that are effective in capturing intensity changes in haze light can be selected, and process diagnosis can be performed using only the selected ones of the multiple detection optical systems Bn. If the signals of multiple detection optical systems Bn were merged and output regardless of their sensitivity to haze light, the changes that are detected with high sensitivity by a specific detection optical system Bn would be diluted, and the inspection sensitivity would decrease. In contrast, in this embodiment, a configuration with multiple detection optical systems Bn in different directions is used to perform process diagnosis with high sensitivity.

In this embodiment, in particular, the correlation between the haze signal and its fluctuation factor for each detection optical system Bn is stored in the storage device DB, and the detection optical system Bn with a specific azimuth angle φ1 can be selectively used for process diagnosis based on this correlation. This makes it possible to detect abnormalities or signs of abnormalities in the process equipment according to fluctuation factors of the haze light intensity, such as the surface film thickness and surface roughness of the sample surface. In this embodiment, an example has been described in which a change in the state of the process equipment related to a change in a specified range of the surface roughness of the sample W is detected from the difference between the haze signal incident on the openings α3 and α4 located in the direction of specular reflection of the illumination light with respect to the beam spot BS and the reference haze signal. Also, an example has been described in which a change in the state of the process equipment related to a change in a specified range of the surface film thickness of the sample W is detected from the difference between the haze signal incident on the openings α1 and α6 located in the opposite direction to the specular reflection of the illumination light with respect to the beam spot BS and the reference haze signal.

Furthermore, in this embodiment, the detection optical system Bn separates the haze light according to its polarization direction using the polarizing beam splitter Bc, and can detect two lights with different polarization directions for the haze light emitted in the same direction from the same coordinates on the sample W. With this configuration, it is possible to perform more precise process diagnosis that includes the polarization direction of the haze light as a parameter, based on correlation data that specifies the intensity, polarization direction, and fluctuation factors of the haze signal for each detection optical system Bn (i.e., for each emission direction of the haze light).

Second Example
In the first embodiment, an example has been described in which the process equipment is diagnosed by actually comparing the haze signal of each region of the sample W with the reference haze signal. By accumulating data on the diagnosis performed daily in the semiconductor manufacturing process in this manner and performing machine learning, it is also possible to perform process diagnosis from the haze map of the sample W that is input at any time, based on the learned model.

The trained model is an inference program that incorporates trained parameters through machine learning of training data, and outputs a diagnosis result for the process equipment for input data related to the haze signal. This trained model is created by the signal processing device D or the control device E1, and is stored, for example, in the storage device DB. The signal processing device D uses this trained model to diagnose the process equipment based on data related to the haze signal acquired during defect inspection of the sample W, and outputs an alarm if necessary.

An example of learning data is performance data accumulated in the daily semiconductor manufacturing process, such as the haze map of the sample W, the polarization direction of the haze light, the diagnosis results of the process equipment based on the sample W, the actual maintenance history of the process equipment, and the validity of the diagnosis results. The maintenance history and the validity of the diagnosis results are a type of feedback data, and can be input by, for example, a person who maintained the process equipment using the input device E2 according to a pre-prepared input screen. The validity of the diagnosis results is, for example, the judgment of the person who maintained the process equipment, and can be things such as whether the alarm notified from the process diagnosis device 100 was appropriate, or whether no alarm was notified even though maintenance of the process equipment was performed when necessary.

FIG. 17 is a conceptual diagram of machine learning. Here, an example will be described in which machine learning is executed in the signal processing device D to generate a trained model. The signal processing device D searches for and reads from the storage device DB performance data such as the haze map of the sample W described above, diagnosis results, maintenance history of the process device, and the pros and cons of the diagnosis results, to generate training data. The signal processing device D loads this training data into the neural network D9, and optimizes the weighting of the connections between the neurons in the input layer, intermediate layer, and output layer. As a result, a trained model for diagnosing the process device is generated from the haze signal data obtained for the sample W, such as the scattering direction, light intensity, polarization direction, and coordinates. Note that the trained model is not limited to being generated by the signal processing device D, and may be generated by other computers.

Furthermore, the signal processing device D or the control device E1 can perform machine learning using the haze map and maintenance implementation data as input, and calculate and optimize the maintenance cycle for each part or portion of the process equipment. A trained model for estimating abnormal parts, etc. of the process equipment can be obtained, for example, from the correlation between the parts (or portions) of the process equipment and the variation region (haze map) of the haze signal in the sample W.

In other respects, this embodiment is similar to the first embodiment. In this embodiment, as diagnostic data based on a comparison of the difference between the haze signal and the reference haze signal is accumulated, a trained model is generated that also takes into account feedback data such as the maintenance history and the validity of the diagnostic results, and this is expected to improve the diagnostic accuracy of the process equipment. In addition, as described above, this is also expected to optimize the maintenance cycle of the process equipment.

(Third Example)
Fig. 18 is a schematic diagram of a main part of a process diagnostic device according to a modified example of the present invention. In Fig. 18, elements that are the same as or correspond to those described in the first and second embodiments are given the same reference numerals as those in the previously mentioned drawings, and the description thereof will be omitted.

This embodiment is an example in which data obtained by multiple process diagnostic devices is included in the basic data of the reference haze signal (first embodiment) or the trained model (second embodiment) described above. In this embodiment, the process diagnostic device 100 is connected to a data server DS via a network (not shown) as appropriate. Other process diagnostic devices 100' and 100" different from the process diagnostic device 100 are connected to this data server DS via a network as appropriate. The process

diagnostic devices

100, 100', and 100" are preferably of the same type or similar types (same series, same manufacturer, etc.), but may be devices of different types. Although two other process diagnostic devices 100' and 100" are illustrated in FIG. 18, the number of other process diagnostic devices connected to the data server DS may be one or three or more.

The data server DS receives diagnostic data and the like from the process

diagnostic equipment

100, 100', 100" and stores this data. This stored data can include diagnostic data for the process equipment, including the haze signal and diagnostic results for each process diagnostic equipment, as well as design data for the sample W, maintenance data for the process equipment, the validity of diagnostic results, and inspection data for the sample W. In addition, with regard to defect inspection of the sample W, inspection conditions (inspection recipe), defect review data, defective material analysis data, and the like can also be stored in the data server DS. Defective material analysis data is, for example, information obtained by energy dispersive X-ray analysis. This may be a standalone device, or it may be mounted on a defect review device. It can also be acquired together with defect review information. Based on such accumulated data, the data server DS calculates a reference haze and a learned model to be compared with the sample W for process diagnosis. The calculation of the reference haze signal and the learned model can be performed by the data server DS at regular intervals, or when a certain amount of new data is accumulated. Each

process diagnosis device

100, 100', 100" receives the reference haze signal or the learned model that is updated from time to time by the data server DS, and performs diagnosis of the process device based on the haze signal of the sample W. The haze signal refers mainly to the low-frequency components of the light signal obtained from the sample, and is a signal that is mainly due to the characteristics of the sample. Here, for example, the low-frequency components of the detection signal, i.e.

According to this embodiment, in addition to the process diagnosis device 100's own data, a reference Haze signal or a learned model is calculated using a large amount of data from the other process diagnosis devices 100, 100' as basic data. Therefore, a large amount of basic data can be accumulated in a short period of time, and diagnosis accuracy can be improved quickly.

(Fourth Example)
Fig. 19 is a schematic diagram of the main part of a process diagnostic device according to a fourth embodiment of the present invention. In Fig. 19, the same reference numerals as in the previously mentioned drawings are used for elements that are the same as or correspond to those described in the first to third embodiments, and the description thereof will be omitted.

This embodiment is a variation of the method of acquiring a haze signal. A sample transfer position Pa and an inspection start position Pb are set on the movement axis of the translation stage of the stage ST, and by driving the translation stage, the stage ST moves along a straight line passing through these positions. The inspection start position Pb is the position where the sample W is irradiated with illumination light to start inspection of the sample W, and is the position where the center of the sample W coincides with the beam spot BS of the illumination optical system A. The sample transfer position Pa is the position where the sample W is attached to and detached (loaded and unloaded) from the stage ST by the arm Am, and the stage ST, having received the sample W, moves from the sample transfer position Pa to the inspection start position Pb.

In response to recent demands for even higher sensitivity inspections, the detection optical system Bn is positioned close to the sample W. When the stage ST is directly below the detection optical system Bn, the gap G between the stage ST and the detection optical system Bn is about a few mm or less. Because it is difficult to insert the sample W into the gap G with the arm Am at the inspection start position Pb and place it on the stage ST, a configuration is adopted in which the sample W is transferred at a sample transfer position Pa away from the inspection start position Pb.

In defect inspection, the sample W is generally scanned with P-polarized illumination light while the stage ST moves from the inspection start position Pb, but in this embodiment, a preliminary scan is performed while the stage ST moves from the sample transfer position Pa to the inspection start position Pb. In the preliminary scan, the illumination light is set to S-polarized light, and the sample W is scanned in a spiral trajectory from the outer periphery toward the center. Then, a diagnostic process for the process device is performed based on the Haze signal obtained in this preliminary scan.

In other respects, this embodiment is similar to the first, second, or third embodiment.

Here, in the defect inspection of the sample W, the inspection conditions are set so as to suppress the generation of haze light, which generally becomes noise in defect inspection (for example, the illumination light is set to P-polarized light). Therefore, depending on other conditions, it may be possible that the haze light cannot be sufficiently detected in the defect inspection of the sample W, making it difficult to diagnose the process equipment based on the haze signal.

In contrast, in this embodiment, the sample W can be moved from the sample transfer position Pa to the inspection start position Pb, allowing a preliminary inspection to be performed under conditions different from those for the defect inspection to collect a haze signal. By using the transport operation of the sample W to collect a haze signal in this way, it is possible to achieve both defect inspection and process diagnosis without changing the series of machine operations during defect inspection.

In this embodiment, an example has been described in which a haze signal is acquired when the sample W moves from the sample transfer position Pa to the inspection start position Pb before defect inspection, but a configuration is also conceivable in which a haze signal is acquired when the sample W moves to the sample transfer position Pa after defect inspection.

In addition, separate from the opportunity for the sample W to move from the sample transfer position Pa, after the sample W arrives at the inspection start position Pb, the sample W may be pre-scanned under conditions that are likely to generate haze light prior to the defect inspection to collect a haze signal.

(Modification)
(1) Operation of the process diagnosis apparatus 100 As described above, the process diagnosis apparatus 100 can diagnose the process equipment during defect inspection of the samples W, but defect inspection and process diagnosis can be performed independently, and process diagnosis is not necessarily performed during defect inspection of all samples W. Therefore, for example, during defect inspection of one lot of samples W, it is possible to perform process diagnosis together with defect inspection only for the first sample W in the lot, and perform only defect inspection without process diagnosis for the second and subsequent samples W in the lot. Sufficient opportunities for process diagnosis can be obtained by performing process diagnosis once for each inspection of one lot of samples W.

(2) Application of the Haze Signal In each embodiment, an example has been described in which the state of a process device and a malfunctioning part or part of the process device are estimated based on the fluctuation of the haze signal and the polarization direction or area of the fluctuated haze signal. However, the scope of diagnosis can be expanded by capturing the haze signal from multiple angles.

For example, when scanning the sample W by rotating it as shown in FIG. 2, it is assumed that the intensity of the haze light may change depending on the rotation angle of the sample W even if the detection optical system Bn is the same. In this case, the selection of the detection optical system Bn in the process diagnosis for the same sample W can be configured to switch periodically according to the rotation angle of the sample W. In each of the above embodiments, a bare wafer or a film-covered wafer on which no pattern is formed is typically inspected. For example, when a patterned wafer is rotated and scanned as the sample W, the scattering direction of the haze light may change regularly due to the influence of diffraction caused by fine linear patterns formed periodically vertically and horizontally. In such a case, it may be effective to predefine the relationship data between the rotation angle of the sample W and the selection of the detection optical system Bn and store it in a storage device DB, for example, and to configure the selection of the detection optical system Bn in the process diagnosis to switch according to the rotation angle of the sample W.

In addition to the haze signal, it is also possible to analyze or machine-learn the correlation between the state of the process equipment and the data set of the haze signal and defect signal for the same sample W, and perform process diagnosis based on the haze signal and defect signal. It is possible that a defect in the sample W occurs due to a malfunction of the process equipment, or that the defect affects the haze light, and the accuracy of process diagnosis can be improved by monitoring the defect signal along with the haze signal.

In addition, plasma etching equipment may be equipped with an OES to monitor the plasma state. It is also possible to analyze or learn machine learning the monitor data during plasma etching by this OES together with the haze signal in a signal processing device D or server. If it is possible to identify the correlation between the monitor data of the plasma discharge state and the haze signal, it is expected that the accuracy of process diagnosis will be further improved.

Furthermore, defect inspection is performed after one or several steps in the semiconductor manufacturing process, and haze signals can be obtained during defect inspection before and after the process by the process equipment being diagnosed. It is also possible to calculate the difference in the haze signals before and after the process for the same sample W, and use this difference to evaluate the level of processing by the process equipment. In other words, calculating the difference in the haze signals before and after the process for a reference sample as a reference haze signal related to the level of processing by the process equipment, and comparing a similar difference related to sample W with the reference haze signal, can be considered as one form of process diagnosis.

Also, in the fourth embodiment, an example was described in which the sample W was scanned under conditions in which haze light is likely to occur and a haze signal was acquired separately from the defect inspection. In this case, haze light is also acquired during the defect inspection, and the haze signals sampled under conditions in which haze light is likely to occur and conditions in which it is unlikely to occur are compared and the difference is analyzed, so that a new correlation between the haze signal and the process equipment can be identified.

In addition, in each embodiment, as described above, the haze signals incident on the openings α3, α4, α1, and α6 are used for process diagnosis. However, haze signals incident on other openings can also be used for process diagnosis. For example, a correlation can be found between the state of the process device and the sum or difference signal of the haze signals incident on the openings α2, α5, β2, β3, β5, and β6 located on the left and right of the beam spot BS and the haze signals incident on the openings α3, α4, α1, and α6.

100...Process diagnostic device, A...Illumination optical system, Bc...Polarizing beam splitter, Bn (n=1, 2...)...Detection optical system, BS...Beam spot, Cn (n=1, 2...)...Sensor, D...Signal processing device, DB...Storage device, ST1...Sample stage, ST2...Scanning device, W...Sample

Claims

A process diagnostic device for diagnosing a state of a semiconductor process device, comprising:
a sample stage for supporting a sample processed in the semiconductor processing device;
an illumination optical system that irradiates illumination light onto the sample placed on the sample stage;
a plurality of detection optical systems that collect light from the sample, convert it into an electrical signal, and output a detection signal;
a signal processing device that processes detection signals of the plurality of detection optical systems,
The signal processing device includes:
Scanning the sample to extract a Haze signal of the sample;
The process diagnostic device is characterized in that the haze signal of the sample is compared with a reference haze signal, and an alarm is output when the difference between the haze signal and the reference haze signal exceeds a set value.
2. The process diagnostic device of claim 1,
The signal processing device is a process diagnostic device that detects a change in the state of the semiconductor process device that appears locally in a distribution over the entire surface of the sample as a microscopic surface shape of the sample, based on the difference between the haze signal and the reference haze signal.
2. The process diagnostic device of claim 1,
The signal processing device selects a haze signal of a detection optical system having the strongest correlation with the variation factor from among the plurality of detection optical systems based on correlation data relating to a plurality of detection optical systems and variation factors of a haze signal that is pre-stored in a storage device, and performs the comparison between the selected haze signal and the reference haze signal.
The process diagnostic device according to claim 3,
The process diagnosis device, wherein the plurality of detection optical systems are arranged such that their directions with respect to the beam spot of the illumination light are different from each other.
Process diagnostic equipment.
The process diagnostic device according to claim 4,
the plurality of detection optical systems each include a polarizing beam splitter that splits light in accordance with a polarization direction, and a plurality of sensors that detect the light having different polarization directions split by the polarizing beam splitter,
The process diagnosis device, wherein the correlation is data representing a relationship between the intensity of the haze signal, the polarization direction, and the fluctuation factor, which is set for each of the plurality of detection optical systems.
2. The process diagnostic device of claim 1,
The signal processing device includes:
creating a haze map, which is an intensity distribution of the haze signal, for the sample;
A process diagnostic device that compares the haze map with the intensity distribution of the reference haze signal, and diagnoses the state of the semiconductor process device based on whether or not a difference between the haze signal and the reference haze signal exceeds a set value.
2. The process diagnostic device of claim 1,
The signal processing device is a process diagnosis device that accumulates data related to the diagnosis of the semiconductor process equipment, performs machine learning, and diagnoses the condition of the semiconductor process equipment based on a trained model obtained by machine learning.
A method for determining timing for plasma replacement in a chamber of a semiconductor process apparatus, comprising:
A sample placed in the chamber is subjected to plasma treatment,
a process diagnostic device for irradiating the plasma-treated sample with light and extracting a defect signal from a high-frequency component of the light from the sample and a haze signal from a low-frequency component of the light from the sample, and repeatedly acquiring the defect signal and the haze signal for multiple days;
the process diagnostic device determines a plasma state of the semiconductor process device based on whether a difference between the haze signal and a reference haze signal exceeds a set value;
If the difference exceeds a set value, output the result;
A plasma replacement timing determination method in which a user determines the next timing of plasma replacement based on the results output from the process diagnostic device and maintenance data of the semiconductor process device.
The plasma replacement timing determination method of claim 8, characterized in that the process diagnostic device has a plurality of detection optical systems with different azimuth angles relative to the beam spot that irradiates the light, and the haze signal is obtained repeatedly over a number of days by selecting the detection optical system that exhibits the greatest fluctuation from among the plurality of detection optical systems, and judging the plasma state of the semiconductor process device based on whether the haze signal of the detection optical system and a reference haze signal exceed a set value.