JP7340703B2 - Defect inspection equipment and method - Google Patents

Description

The present invention relates to the field of semiconductor manufacturing technology, and more specifically to a defect inspection apparatus and method.

In semiconductor defect inspection systems, a laser interferometer is usually used to detect the position of the stage on which the wafer (target to be inspected) is mounted, but a time delay integral (TDI) sensor is used to detect the position of the wafer sample. When capturing an image, it is necessary to generate a trigger pulse (also referred to as a "timing pulse") every time the stage moves a certain amount.

FIG. 1 shows a defect inspection device used for detecting pattern defects on a wafer, according to Japanese Patent (JP-A-2003-177101). This device is a device for detecting the position and image of defects on the wafer surface.

As shown, FIG. 1 shows a timing generation module designated at 29. As shown in FIG. The timing generation module generates a timing pulse based on the relationship between the output data of the laser interferometer and the target (object to be inspected) position, that is, compares the output data with the target position, and determines whether the position information is the target position. If the location exceeds the limit, location information will be sent at regular intervals.

However, when generating timing pulses using the above method, an error equivalent to the data output interval of the position measurement circuit is generated, and even if the fastest position measurement circuit is used, this error is insufficient to meet the required specifications of the system. There may be no. In particular, when the moving speed of the stage on which the wafer is mounted increases, the output interval of a normal laser interferometer cannot generate timing pulses with sufficient accuracy.

In order to overcome the above-mentioned deficiencies in the prior art, the present invention generates timing pulses with sufficient precision in the output interval of a typical laser interferometer even when the speed of movement of the stage on which the wafer is mounted increases. The present invention provides a wafer defect inspection device and a detection method thereof.

In order to realize the above objectives, the technical solution of the present invention provides a defect inspection device.

The defect inspection apparatus includes a main control interface module used to input/output data including user information data, inspection result data with digital and/or image data, and operational flow control data;
a stage on which an object to be inspected is mounted, and the object to be inspected moves as the stage moves;
a lighting module that illuminates the object to be inspected mounted on the stage;
an imaging module that images the surface of the object to be inspected illuminated by the illumination module and outputs an image signal;
a position measurement module that measures position information of the object to be inspected;
A timing pulse is generated when the object to be inspected moves by one pixel of the imaging module based on the position information, and by correcting an error in the timing pulse, the timing pulse is delayed by one clock and output; a timing generation module that images the surface of the object to be inspected;
and an image processing module that inspects defects in the object to be inspected based on the image signal.

Further, the position measurement module includes a linear image sensor.

Furthermore, the timing generation module includes:
an adder that adds the pixel size (ΔX) of the linear image sensor and outputs a trigger coordinate X _T ;
a differentiator that calculates the velocity v based on the position information _XL of the object to be inspected;
a comparator that compares the magnitude relationship between the position information _XL and the output value of the adder and generates a trigger signal Trg;
An operation that takes the first trigger coordinates X _T , trigger signal Trg, position information X _L and speed v as input, executes a predetermined operation (X _T +ΔX-X _L )/v, and obtains a one-clock delay time. The vessel and
a delay unit that outputs a delay pulse that is delayed by one clock based on the clock delay time;
The fall of the delay pulse is detected, the imaging module is started to image the surface of the object to be inspected, the update frequency of the position information _XL is greater than the timing pulse frequency of the linear image sensor, and the delay pulse is The operating frequency of the unit includes a fall detection unit that is higher than the update frequency of the position information _XL .

Further, an update frequency of the position information _XL is 10 MHz, a timing pulse frequency of the linear image sensor is 3 MHz, and an operating frequency of the delay unit is 300 MHz.

Furthermore, the linear image sensor is a time delay integration type image sensor.

Furthermore, the object to be inspected is a semiconductor wafer.

Further, the illumination module includes a deep UV light source forming an illumination optical path using a broadband illumination method or a narrowband illumination method with bright lines.

Further, the position measurement module includes a laser length measurement device.

In order to achieve the above objectives, the technical solution of the present invention provides a defect inspection method. The method is
Step S1 of measuring position information of the object to be inspected using a position measurement module;
Based on the position information, a timing pulse is generated when the object to be inspected moves by one pixel of the imaging module, and the timing generation module delays the timing pulse by one clock by correcting an error in the timing pulse. step S2 of outputting the
Step S3 of imaging the surface of the object to be inspected based on the timing pulse delayed by one clock and outputting an image signal;
and step S4 of inspecting the object to be inspected for defects based on the image signal.

Furthermore, in step S2,
step S21 of adding the pixel size (ΔX) of the linear image sensor and outputting the trigger coordinate X _T ;
a step S22 of comparing the magnitude relationship between the position information _XL and the output value of the adder to generate a trigger signal Trg;
Step S23 of inputting the trigger coordinates X _T , trigger signal Trg, position information X _L and speed v, and executing a predetermined calculation (X _T +ΔX-X _L )/v to obtain a one-clock delay time;
step S24 of outputting a delay pulse that is delayed by one clock based on the clock delay time;
The fall of the delay pulse is detected, the imaging module is started to image the surface of the object to be inspected, the update frequency of the position information _XL is greater than the timing pulse frequency of the linear image sensor, and the delay pulse is The operating frequency of the unit is greater than the update frequency of the location information _XL (step S25).

As can be seen from the above technical solution, the present invention corrects timing pulse errors by adding a counter circuit that operates at a speed that is faster than the output rate of the position measurement circuit and satisfies the required specifications of the system. In other words, the timing pulse generation unit outputs a pulse after waiting for a time period obtained by subtracting the timing delay time from the average interval.

By improving the accuracy of the timing pulse, the present invention improves image resolution and can generate accurate trigger pulses even with slow output rate position information.

1 is a schematic structural diagram of a defect inspection device according to the prior art; FIG. FIG. 1 is a schematic structural diagram of a defect inspection device according to an embodiment of the present invention. 1 is a schematic diagram of a timing generation module according to an embodiment of the invention; FIG. FIG. 4 is a time series schematic diagram illustrating each pulse according to an embodiment of the present invention.

Hereinafter, specific embodiments of the present invention will be described in more detail with reference to the drawings.

In the following specific embodiment of the present invention, FIG. 2 is a schematic structural diagram of a semiconductor wafer defect inspection apparatus 100 according to the present invention. The defect inspection apparatus 100 typically mainly includes a main control interface module, a stage, an illumination module, an imaging module, an image processing module, a position measurement module, and a timing generation module.

In an embodiment of the invention, as shown in FIG. 2, the main control interface module 9 is used to control the operation of the entire device, i.e. the test results with user information data, digital and/or image data. Used to input/output data including data and operational flow control data. Specifically, the main control interface module 9 is connected to a display device 10, an input module 11, a storage module 12, and an external network 13. The display device 10 is used to display test information. The input device 11 captures user information. The storage module 12 stores and manages test result data and operation flow control data. The external network 13 transmits and receives inspection data and operation flow control data to and from a host computer and other inspection devices.

In an embodiment of the present invention, a lighting module is used to illuminate the object to be inspected mounted on the stage. As shown in FIG. 2, the illumination module includes a xenon light source 14, a wavelength filter 15a, a polarizing beam splitter (PBS) 15, a λ/4 plate 16, and an objective lens 17.

In one preferred embodiment, the semiconductor wafer defect inspection apparatus 100 is an optical pattern defect inspection apparatus in which a lighting module uses deep ultraviolet (DUV) light. The DUV light may be broadband illumination such as 200 to 400 nm, or narrowband illumination having bright lines. As shown in FIG. 2, the light beam emitted from the xenon light source or the mercury-xenon light source 14 is band-limited by the wavelength filter 15a to form a DUV light beam of narrowband illumination having bright lines.

Broadband lighting has the effect of reducing color unevenness caused by thin film interference on the wafer. On the other hand, narrowband illumination with bright lines can accurately correct the chromatic aberration of the optical system and has the effect of improving resolution.

In this embodiment, the DUV light source is used as a lamp light source, but in other embodiments of the present invention, a high-intensity laser light source may be used. For example, the second harmonic of 266 nm can be used by combining a YAG laser (λ=532 nm) and a nonlinear optical element.

In an embodiment of the present invention, the image signal of the surface of the object to be inspected is completed by an imaging module, which images the surface of the object to be inspected illuminated by an illumination module and outputs an image signal, and the image signal is Processing and determining whether there are any defects on the surface of the object to be inspected completed during the manufacturing process. Similar to the prior art, the imaging module may include an objective lens 17, a λ/4 plate 16, a polarizing beam splitter 15, an imaging lens 19, a linear image sensor 20, and the like.

The optical system of the imaging module will be explained below. As shown in FIG. 2, S-polarized light is formed after being band-limited by the wavelength filter 15a, and the S-polarized light is reflected downward at the polarizing beam splitter 15, passes through the λ/4 plate 16, and becomes circularly polarized light. Become. The light beam illuminated on the wafer 18 (also referred to as the "object to be inspected") by the objective lens 17 is reflected on the wafer surface, passes through the λ/4 plate 16 again, becomes P-polarized light, and is sent to the polarizing beam splitter 15. reflected. By arranging the optical system as described above, it is possible to prevent a drastic decrease in the amount of detected light.

Next, the reflected light is received by the imaging lens 19, and an enlarged optical image of the wafer surface is formed on the linear image sensor 20. The enlarged optical image of the wafer surface captured by the linear image sensor 20 is input as a detection image signal to the image processing device 32 and inspected for defects. The inspection imaging optical path further has one branch, and another image of reflected light separated by the imaging lens 19 and the polarizing beam splitter 15 is provided to the TV imaging unit 21. The imaging unit 21 is connected to another image processing device 22 that performs alignment and defect inspection on the inspection image signal.

In an embodiment of the present invention, a stage is used to mount an object to be inspected, and the object to be inspected moves as the stage moves. Specifically, the stages are stacked from top to bottom. It includes a wafer chuck 23, a Z stage 30, a Θ stage 24, an X stage 25, and a Y stage 26. The object to be inspected (wafer) 18 is vacuum-adsorbed and flattened by the wafer chuck 23, thereby preventing the wafer 18 waiting to be inspected and the wafer chuck 23 from moving relative to each other. The wafer chuck 23 is mounted on a Z stage 30, a Θ stage 24, an X stage 25, and a Y stage 26, which are stacked. After the wafer is attached to the wafer chuck 23, the Θ stage 24 is aligned so that the direction in which the aligned wafer array is aligned matches the scanning direction of the X stage 25, and is fixed during the inspection operation. There is a need.

During the inspection operation, the X stage 25 moves left and right toward the plane of the paper, and when turning back, the Y stage 26 moves stepwise in a direction perpendicular to the plane of the paper. Here, a wafer (singular form of "Die" or "Dice") refers to a semiconductor chip obtained by printing a circuit pattern on the wafer 18 and cutting it into dice.

In the embodiment of the present invention, the stage control computer 27 controls the movement of the Θ stage 24, the X stage 25, and the Y stage 26 using control signals. The position measurement module is used to measure position information of the object to be inspected. Preferably, the position measurement module may be a laser length measurement device. For example, the laser length measurement device 28 measures the position information of the X stage 25 (that is, the position information of the wafer 18 in the scanning direction of the X stage 25). can do.

The timing generation circuit 200 (also referred to as "timing generation device") indicates the pixels of the linear image sensor 20 based on position information (X _L ) output from the laser length measuring device 28 (also referred to as "position measurement circuit"). Generates a start timing signal (also referred to as a "timing pulse"). That is, the timing pulse is transmitted to the linear image sensor 20 and the image signal is read out.

In this embodiment, a linearly moving stage (X stage 25) is taken as an example, and the timing generation module 200 will be explained as an example. Complete the output of location information. The X stage 25 can be similarly applied to the Z stage 30, Θ stage 24, and Y stage 26, and will not be repeatedly described here.

Furthermore, the stage control computer 27 is connected to the overall control section 9 using the internal network 33, and is used to download wafer layout information and the like and identify inspection positions.

The imaging process will be explained below. As shown in FIG. 2, a timing signal is generated every fixed amount of movement of the X stage to drive the linear image sensor 20. In order to synchronize the reading of the linear image sensor 20 with the amount of movement of the X stage, two-dimensional images can be sequentially captured. Further, by using a time delay integration type (TDI-CCD) image sensor as the linear image sensor 20, the S/N in high-speed scanning can be improved.

A TDI sensor (Time Delayed and Integration) has a structure in which multiple one-dimensional image sensors are arranged in two dimensions, and the output of each one-dimensional image sensor is delayed by a predetermined time to image the same location on the object to be inspected. The amount of detected light is increased by adding images output from adjacent one-dimensional image sensors.

The autofocus module includes a detection optical unit 47 and an autofocus computer 39. The detection optical unit 47 detects the heights of multiple points near the imaging position and sends a detection signal to the autofocus computer 39. The autofocus computer 39 calculates a control amount 48 for the Z stage 30 based on the deviation between the detected height and a preset control target, and controls the movement of the Z stage 30.

The autofocus computer 39 is connected to the main control interface module 9 by an internal network 33, and switches operation modes such as inspection mode, review mode, etc., and transmits and receives autofocus operation flow control data. Further, a wafer start signal 40 and a stage return signal 41 are input from the stage control computer 27. Furthermore, a highly real-time control signal 43 such as an autofocus ON/OFF signal during inspection is transmitted from the stage control computer 27.

FIG. 3 is a schematic layout diagram of a timing generation module 200 according to an embodiment of the invention. The timing generation module 200 receives the position information _XL output from the position measurement circuit 28, generates a timing pulse to be described later, and outputs it to the linear image sensor 20.

Unlike the prior art, in the embodiment of the present invention, the timing generation module 200 includes an adder 210 that adds the pixel size (ΔX), etc., and a velocity (dX _L /dt=v) from the position information (X _L ). A differentiator 220 to seek, a comparator 230 to compare the magnitude relationship between the position information (X _L ) and the output value of the adder 210, an arithmetic unit 240 to perform a predetermined operation, and a delay (Delay) to be calculated. It includes a delay unit 250 and a fall detection circuit 260 that receives a delay pulse as input and outputs a timing pulse.

Here, the update frequency (10 MHz = 100 ns) of the position information _XL is faster than the timing pulse frequency (3 MHz = approximately 333 ns) of the linear image sensor 20. Furthermore, the delay unit 250 operates at 300 MHz (approximately 3 nsec).

That is, the timing generation module outputs the timing pulse delayed by one clock by correcting the error in the timing pulse, and images the surface of the object to be inspected.

Adder 210, comparator 230, and arithmetic unit 240 will be explained in detail.

The output of the adder _XT represents the pixel's trigger coordinate (also called "conventional timing pulse"), and is compared with the time when the position information _XL is read, and when X _T < _XL , the timing pulse ("conventional timing pulse") is This pulse is transmitted to the linear image sensor 20 and used to read out image signals.

In the embodiment of the present invention, the adder 210 calculates the size of the pixel (pixel size, ΔX) of the linear image sensor 20, the previous output (X _T ) of the adder 210, and the output from the comparator 230. Based on the trigger signal (Trg), new trigger coordinates _XT are output.

In the embodiment of the present invention, the comparator 230 compares the output value (X _T ) of the adder 210 and the position information (X _L ) in real time, and if X _L is larger than X _T (X A trigger signal (Trg) is output at _T <X _L ). The calculator 240 holds the value of ΔX in advance, receives the trigger coordinates X _T , position information X _L and velocity v as input, and calculates (X _T +ΔX−X _L )/v based on the trigger signal Trg. Execute.

<First embodiment>
FIG. 4 is a time series schematic diagram illustrating each pulse according to an embodiment of the present invention. In this embodiment, the update frequency of the location information _XL is 10 MHz. Therefore, the acquisition times of the position information _XL are at equal intervals of 100 nsec. The stage (X stage in this embodiment) is set to move by one pixel at intervals of about 333 nsec.

Ideally, the timing pulse should be output when the image moves by one pixel. This is shown in FIG. 4 as a conventional timing pulse output time.

In the conventional technology, since the output time of the comparator is used as a timing pulse, an error M occurs between it and the original timing pulse. This error M corresponds to a blur of 1/3 pixel. In this embodiment, the goal is to reduce this to about 1/10 to 1/100.

Therefore, in this embodiment, the conventional timing pulse is delayed by one clock and used. What is needed for this is a delay pulse, which can be generated by the delay unit 250.

In order to obtain the output time of the delay pulse, first, in order to calculate, a predetermined calculation (X _T +ΔX−X _L )/v is obtained by the arithmetic unit 240 to obtain the one clock delay time. Next, delay unit 250 further delays the error (M) by (X _T +ΔX−X _L )/v. In this way, the pulse width of the delay pulse is calculated.

The falling detection path 260 detects the falling edge of the delay pulse and outputs the timing pulse of this embodiment. In this way, it is possible to use a conventional timing pulse delayed by one clock.

According to an embodiment of the present invention, the operating frequency of the delay unit 250 is 300 MHz, the update frequency of the output position information _XL of the position measurement circuit 28 is 10 MHz, and the delay unit 250 operates at 300 MHz (approximately 3 nsec). , the timing pulse error is corrected by delaying it by one clock. A specific example is to output a timing pulse after waiting a time (X _T +ΔX−X _L )/v obtained by subtracting the timing delay time (error M) from the average interval.

The defect inspection method will be briefly explained below. In an embodiment of the present invention, a step S1 of measuring position information of an object to be inspected using a position measurement module, and a timing pulse when the object to be inspected moves by one pixel of an imaging module based on the position information. and the timing generation module outputs the timing pulse after delaying it by one clock by correcting the error in the timing pulse, and detecting the surface of the object to be inspected based on the timing pulse delayed by one clock. The method can specifically include a step S3 of capturing an image of the object and outputting an image signal, and a step S4 of inspecting the object to be inspected for defects based on the image signal.

Step S2 of the defect inspection method includes step S21 of adding the pixel size (ΔX) of the linear image sensor and outputting a trigger coordinate X _T ;
a step S22 of comparing the magnitude relationship between the position information _XL and the output value of the adder to generate a trigger signal Trg;
Step S23 of inputting the trigger coordinates X _T , trigger signal Trg, position information X _L and speed v, and executing a predetermined calculation (X _T +ΔX-X _L )/v to obtain a one-clock delay time;
step S24 of outputting a delay pulse that is delayed by one clock based on the clock delay time;
The frequency at which the falling edge of the delay pulse is detected, the imaging module is started to image the surface of the object to be inspected, and the positional information _XL is updated is greater than the timing pulse frequency of the linear image sensor; The operation frequency of the delay unit may specifically include step S25, where the operating frequency of the delay unit is higher than the update frequency of the location information _XL .

What has been described above are only preferred embodiments of the present invention. These embodiments do not limit the scope of the claimed invention. For the same reason, all equivalent structural changes made using the contents of the specification and drawings of the invention fall within the scope of the claims of the invention.

Claims (6)

a main control interface module used to input/output data including user information data, test result data with digital data and/or image data, and operational flow control data;
a stage on which an object to be inspected is mounted and the object to be inspected moves as the stage moves;
a lighting module that illuminates the object to be inspected mounted on the stage;
an imaging module that images the surface of the object to be inspected illuminated by the illumination module and outputs an image signal;
a position measurement module that measures position information of the object to be inspected;
A timing pulse is generated when the object to be inspected moves by one pixel of the imaging module based on the position information, and by correcting an error in the timing pulse, the timing pulse is delayed by one clock and output; a timing generation module that images the surface of the object to be inspected;
an image processing module that inspects defects in the object to be inspected based on the image signal ,
The position measurement module includes a linear image sensor,
The timing generation module includes:
an adder that adds the pixel size (ΔX) of the linear image sensor and outputs a trigger coordinate X _T ;
a differentiator that calculates the velocity v based on the position information _XL of the object to be inspected ;
a comparator that compares the magnitude relationship between the position information XL _and the output value of the adder and generates a trigger signal Trg;
an arithmetic unit that receives the trigger coordinates X _T , trigger signal Trg, position information X _L and speed v as input, executes a predetermined calculation (X _T +ΔX-X _L )/v, and obtains a one-clock delay time;
a delay unit that outputs a delay pulse that is delayed by one clock based on the clock delay time;
The fall of the delay pulse is detected, and the imaging module is started to image the surface of the object to be inspected, and the update frequency of the position information XL is greater than the timing pulse frequency of the linear image sensor, and _the The operating frequency of the delay unit is higher than the update frequency of the position information XL _and the fall detection unit.
A defect inspection device characterized by: The defect inspection apparatus according to claim 1 , wherein the linear image sensor is a time delay integration type image sensor. 2. The defect inspection apparatus according to claim 1, wherein the object to be inspected is a semiconductor wafer. The defect inspection apparatus according to claim 1, wherein the illumination module includes a deep ultraviolet light source that forms an illumination optical path using a broadband illumination method or a narrowband illumination method with bright lines. The defect inspection apparatus according to claim 1, wherein the position measurement module includes a laser length measuring device. Step S1 of measuring position information of the object to be inspected using a position measurement module;
Based on the position information, a timing pulse is generated when the object to be inspected moves by one pixel of the imaging module, and the timing generation module delays the timing pulse by one clock by correcting an error in the timing pulse. Step S2 of outputting;
Step S3 of imaging the surface of the object to be inspected based on the timing pulse delayed by one clock and outputting an image signal;
step S4 of inspecting the defect of the object to be inspected based on the image signal ,
The step S2 is
step S21 of adding the pixel size (ΔX) of the linear image sensor and outputting the trigger coordinate X _T ;
step S22 of comparing the magnitude relationship between the position information XL _and the output value of the adder and generating a trigger signal Trg;
Step S23 of inputting the trigger coordinates X _T , trigger signal Trg, position information X _L and speed v, and executing a predetermined calculation (X _T +ΔX-X _L )/v to obtain a one-clock delay time;
step S24 of outputting a delay pulse that is delayed by one clock based on the clock delay time;
The fall of the delay pulse is detected, the imaging module is started to image the surface of the object to be inspected, and the update frequency of the position information XL is greater than the timing pulse frequency of the linear image sensor, and the delay unit operates _. The frequency is greater than the update frequency of the location information XL (step S25) _.
A defect inspection method characterized by: