METHOD TO DETECT A DEFECT ON A LITHOGRAPHIC SAMPLE AND METROLOGY SYSTEM TO PERFORM SUCH A METHOD

Abstract

A method to detect a defect on a lithographic sample includes the following steps: detection light and a detector having at least one sensor pixel are provided. Further, a detection pattern is provided causing a light structure of the detection light being structured at least along one dimension (1D, x). The detection pattern is aligned such that the detector is aligned normal to an extension (xy) of the light structure. Further, a complimentary pattern is provided having a 1D structure which is complimentary to that of the detection pattern. The sample is moved relative to the detection pattern while gathering the detection light on the detector. Further, a reference sample without defects or with negligible defects is provided. The reference sample also is moved relative to the detection pattern while gathering the detection light on the detector. A defect (s1 to s4) localization on the sample is decoded by correlation using the complementary pattern. Using such defect detection method improves a signal-to-noise ratio.

Description

TECHNICAL FIELD

The invention relates to a method to detect a defect on a lithographic sample, i.e. a mask blank or a structured wafer. Further, the invention relates to a metrology system to perform such a method.

BACKGROUND

The detection of nanoscale imperfections or defects on a lithographic sample is of critical importance to high-volume semi-conductor manufacturing. A defect buried in a nanopattern can substantially affect the functionality of the whole device. Thus, detecting defects during nanofabrication is critical for maintaining a high yield. Current optical inspection technologies are challenged as critical dimensions (CDs) of the structures of the lithographic sample drop below e.g. 20 nm. A scattered intensity from such diminished defect dimensions is low and it becomes difficult to discern between the defect signal and the instrument noise.

A method of defect detection on a nanoscale is disclosed in U.S. Pat. No. 7,812,943 B2 and in B. M. Barnes et al., Optics Express 21 (22), pp. 26219-26226 (2013). These documents disclose techniques which are known as through-focus scanning optical microscopy and scattered-field optical microscopy. Another technique for defect localization is diffractive phase optical interferometric microscopy, disclosed in R. Zhou et al., Nano Lett. 13, pp. 3716-3721 (2013).

SUMMARY

An aspect of the present invention is to further improve a signal to noise ratio of a method to detect a defect on a lithographic sample, for example, by using the method according to claim 1.

According to the invention, it has been recognized that using a reference sample on the one hand and further using a decoding step by correlation using a complementary pattern helps to amplify the desired defect signal with respect to a noise signal. A good signal to noise ratio with respect to the desired defect signal results.

The detector may be embodied as a single pixel sensor or alternatively as a sensor pixel array extending at least in one dimension (1D) or in two dimensions (2D). The detector may be a CCD array or a CMOS array.

Alignment of the detection pattern such that the detector is aligned normal to an extension of the light structure ensures that light emanating from the detection pattern is received by the detector. The detection pattern may be a one-dimensional (1D) or a two-dimensional (2D) detection pattern. In cases of a 2D detection pattern, a 2D defect distribution on the sample may be detected requiring a movement of the sample relative to the detection pattern in two dimensions, in particular a two-dimensional scanning movement. Uniformly redundant arrays which may be used as 1D/2D detection patterns are known from A. Busboom et al. Experimental Astronomy 8: 97-123, 1998.

The method can be performed using coherent and/or partially coherent detection light.

A matched filtering according to claim 2 is done by correlating a known signal, i.e. the complementary pattern, with an unknown signal, i.e. the defect signal to be detected coded with a detection pattern, to detect the presence of a pattern, i.e. the defect signal, in the unknown signal.

Such filtering algorithm allows a high signal to noise ratio depending on the quality of the approximation of a convolution of the detection pattern with the complementary pattern to a delta distribution.

Providing the detection pattern according to claim 3 via a structured mask can be done with sufficient accuracy. The structured mask may result in a binary detection pattern causing a detection light structure by changing the detection light intensity between “no detection light” and “full detection light intensity”.

Alternatively, the detection pattern may be not binary and may include further e.g. staggered detection light levels, i.e. more than two different intensities, e.g. 50%, 25%, 75% of the maximum intensity, and/or may include intensity grey levels continuously changing between no intensity and full detection light intensity.

Providing the detection pattern via structured illumination according to claim 4 is an alternative allowing an omission of a structured mask. Such structure illumination may be realized via an interference pattern, in particular via diffraction.

A periodical shift of the sample according to claim 5 gives a favorable control of correlation properties.

Aperiodic shifting of the sample according to claim 6 in certain cases gives sufficient accuracy.

The advantages of a metrology system according to claim 7 correspond to those discussed above with respect to the detection method. Depending on the metrology system setup, a 1D and/or a 2D detection pattern may be utilized. The sample stage may have drives to allow a 1D movement and/or a 2D movement of the sample in a respective sample stage plane. The sensor pixel may be realized as a photodiode or as a CMOS or CCD sensor.

A sensor pixel array according to claim 8 allows the utilization of a 2D detection pattern and a 2D measurement while moving the sample relative to the detection pattern only along one dimension. Alternatively, also a 2D scanning movement is possible.

The sensor pixel array may extend in two dimensions and may be realized as a CMOS or as a CCD sensor array. The metrology system may be embodied such that the sample or a sample region is imaged to the detector.

The advantages of a metrology system according to claims 9 and 10 correspond to those of the method according to claims 3 and 4.

BRIEF DESCRIPTION OF DRAWINGS

Exemplary embodiments are described hereinafter with respect to the enclosed drawings. It shows:

FIG. 1 a schematical layout of a metrology system for defect detection on lithographic samples, shown in a meridional section view;

FIG. 2 a diagram showing the dependence of an ideal difference signal from a scanned position showing four defects of an exemplified defect pattern on a lithographic mask sample;

FIG. 3 in a diagram similar to FIG. 2 the dependence of a coded difference signal on the scan position which would be measured by a single pixel detector of the metrology system without noise, a detection pattern being used for signal coding;

FIG. 4 in a diagram similar to FIG. 2 the dependence of an added noise signal on the scan position, the added noise signal being larger than the coded difference signal shown in FIG. 3;

FIG. 5 in a diagram similar to FIG. 2 the dependence of the added signals according to FIGS. 3 and 4 on the scan position;

FIG. 6 in a diagram similar to FIG. 2 the dependence of a recovered/decoded signal on the scan position, the recovered signal being generated by convolving the signal according to FIG. 5 with a complimentary pattern using a defect detection method which can be performed with the metrology system;

FIG. 7 an example for a 1D pattern structure used to provide the detection pattern, shown as an intensity I dependence along a coordinate x;

FIG. 8 a 1D pattern structure which is complimentary to that of FIG. 7 which can be used as the complementary pattern, also shown in an I/x diagram;

FIG. 9 the outcome of a convolution between the structures shown in FIGS. 7 and 8; and

FIGS. 10 to 12 similar to FIGS. 7 to 9 a 2D pattern structure (FIG. 10), a 2D complimentary pattern structure (FIG. 11) and a convolution result (FIG. 12), each in a top view.

DETAILED DESCRIPTION

FIG. 1 shows in a meridional view a layout and a beam path of a metrology system 1. The metrology system 1 is capable to perform a defect detection method to detect at least one defect or multiple defects on a lithographic sample, e.g. a mask blank or a structured reticle or wafer.

In the following, a Cartesian coordinate system xyz is used to facilitate the description of orientations and directions of the components of the metrology system 1. In FIG. 1, the x-direction runs horizontally to the right. The y-direction is along the direction of view perpendicular to the drawing plane. The z-direction runs vertically in FIG. 1.

The metrology system 1 includes a detection light source 2 producing detection light 3. A wavelength of the detection light 3 may be an EUV wavelength in the range e.g. between 5 nm and 30 nm or may be a larger VUV, DUV, UV or VIS wavelength, e.g. 193 nm. The detection light 3 may be coherent and/or may be partially coherent.

The metrology system 1 further includes a structured mask 4 with a mask pattern 5. The mask pattern 5 has absorbing pattern regions 5a which absorb the detection light 3 and transmitting pattern regions 5b which transmit the detection light 3. Together with the light source 2, the structured mask 4 produces a detection pattern 6 which causes a light structure of the detection light 3 being structured at least along one dimension (1D). In FIG. 1, such dimension along which the detection light 3 is structured via the mask pattern 5 is the x-direction.

In the embodiment of FIG. 1, the detection pattern 6 is provided via the mask pattern 5 of the structured mask 4. In that respect, the mask pattern 5 of the mask 4 is a model of the detection pattern 6. In an alternative embodiment of the metrology system 1, the detection pattern 6 is produced via a structured illumination of the sample with the detection light 3. To achieve such structured illumination, a condenser lens may be provided in the beam path between the lights source 2 and the detection pattern 6.

Further, the metrology system 1 includes a sample stage 7 to move a sample 8, i.e. the lithographic sample to be inspected, which is held with the sample stage 7 relative to the detection pattern 6. The sample stage 7 includes a stage drive to scan the sample along the x-direction which is visualized in FIG. 1 by a scanning arrow S.

Further, the metrology system 1 has a detector 9. Such detector 9 may be a single pixel detector, a sensor pixel array extending at least in one dimension, i.e. a sensor pixel column, or a two-dimensional sensor pixel array. The detector 9 may be realized by at least one photodiode or by a CMOS or CCD sensor array.

A detection optics 10 which in FIG. 1 exemplified is shown as a single condenser lens integrates a detection light intensity resulting from illuminating the sample 8 through the detection pattern 6. The detection optics 10 collects the detection light 3 which is reflected from the sample 8.

In an alternative embodiment, the detection optics 10 is designed to image the sample 8 to the detector 9, i.e. in an image plane 12.

The detection light path between the detection pattern 6 and the detector 9 is aligned such that the detector 9 is aligned normal, i.e. distant in z-direction to an x- and/or to an xy-extension of the light structure caused by the detection pattern 6.

Further, the metrology system 1 has a decoding module 11 to detect a defect localization on the sample 8 by a correlation using the detection result and further using a complimentary pattern. Such decoding module 11 is in signal connection to the detector 9 and also to the sample stage 7.

The detection pattern 6 has a good correlation property with the complimentary pattern which is used during the detection by the decoding module 11.

The detection method performed by the metrology system 1 now is described with use of FIG. 2 et seqq.

During the detection method, the detection pattern 6 is provided and aligned as shown e.g. in FIG. 1 and a complementary pattern is provided which has a 1D structure which is complementary to that of the detection pattern 6. Examples of such detection patterns 6, P and of complimentary patterns P are described below with reference to FIGS. 7 to 12.

The sample 8 is scanned, i.e. is moved relative to the detection pattern 6 in the x-direction while gathering the detection light 3 on the detector 9. The resulting intensity I during such scanning operation can be represented by the following equation:

I(τ)={P(x),O(x−τ)}+N(τ)=P*O+N  (1)

Here:

τ represents the scanning shift, i.e. the respective relative position between the sample 8 and mask 4 along the scan direction x, i.e. the scanning shift between the structured mask 4 and the sample 8,

P(x) represents the detection pattern 6,

O(x−τ) represents the sample pattern which may include at least one defect,

N(τ) represents a noise value.

P*O represents the scanning convolution between the detection pattern 6 and the sample 8.

FIG. 2 shows an example for a defect distribution s to be detected on the sample 8. Shown is the respective scan position τ of the respective defect s_i (i=1 to 4). In total, the defect distribution s shown in FIG. 2 has four defects s₁, s₂, s₃ and s₄.

The scanning shift τ can be realized as an aperiodic shift, i.e. the detection pattern 6 is scanned over the sample 8 once and outside the detection pattern 6 there is zero intensity. Alternatively, the scanning shift may be realized as a periodic shift, i.e. during the scan intensities outside the sample 8 are used as the values that are obtained during the scan of the sample 8 giving a periodic detection result. Such periodic shift is advantageous since periodic correlation properties of the illumination pattern I(τ) are better controllable than aperiodic properties.

A difference between I(τ) measured as mentioned above and a further measured reference scan intensity I₀(τ) measured with a reference sample showing a sparse defect distribution, i.e. no defects or negligible defects, gives:

I−I ₀ =P*(O−O₀)+(N−N₀)=P*s+Ñ  (2)

Here:

O₀ denotes the reference sample pattern,

N₀ denotes the noise during the reference scan,

s is the distribution of defects, i.e. the difference between the sample 8 to be measured, O and the reference sample O₀ without or with negligible defects, Ñ is the difference between the noise measured in the sample scan (equation (1)) and the noise N₀ measured in the reference scan.

In effect, after having scanned the sample 8 relative to the detection pattern 6 giving I(τ), the reference sample is scanned relative to the detection pattern 6 while again gathering the detection light 3 on the detector 9. During this, the sample 8 is replaced by the reference sample on the sample stage 7. The reference sample may have no defects or negligible defects.

FIG. 3 shows the coded difference signal I−I₀, i.e. P*(O−O₀) of equation (2) above.

FIG. 4 shows the added noise signal, i.e. the term N−N₀ of the equation (2) above.

FIG. 5 shows the addition of the signals of FIGS. 3 and 4, i.e. the signal I−I₀ according to equation (2) above.

By help of the measurement results represented by equations (1) and (2) above, a defect localization on the sample 8 can be decoded during the defect detection method by a correlation using the complimentary pattern. This is done by convolving the intensity difference result given above in equation (2) with the complimentary pattern:

I _Det ={{tilde over (P)},P*s+Ñ}={tilde over (P)}*P*s+{tilde over (P)}*Ñ≅s+ε  (3)

Here:

I_Det is the decoded signal obtained by convolution of the measured data

P*s+Ñ with the complimentary pattern P.

FIG. 6 shows the result after the decoding step, i.e. after convolving the signal according to FIG. 5 with the complementary pattern P according to equation (3) above. Such signal I_Det is an addition of the defect distribution s with the convolved noise ε={tilde over (P)}*Ñ.

The four defects s₁ to s₄ are clearly visible and show a good signal to noise ration above the residual noise ε.

FIG. 7 shows an example for a detection pattern P, i.e. an example for the detection pattern 6 given by the mask pattern 5 of the structured mask 4. Shown is the pattern structure as intensity I over the x-coordinate. The pattern P switches in an almost digital manner between values 1 (ideal absorption of the detection lights 3) and 0 (ideal transmission of the detection light 3). The overall structure of the detection pattern 6 shown in FIG. 7 is similar to a barcode.

FIG. 8 shows the complementary pattern {tilde over (P)} which is used in the decoding step according to equation (3) above. The complementary pattern {tilde over (P)} when seen along the coordinate x is exactly mirror symmetrical to the pattern P. Further embodiments of the complementary pattern are possible, which are not mirror symmetrical to the pattern P.

FIG. 9 shows the result of the convolution P*{tilde over (P)}. Due to the good correlation property of {tilde over (P)} and P, their convolution gives at least approximately a delta function leaving the defect distribution s as a result as given in equation (3) above.

As an addition in equation (3), the convolution of the complementary pattern with the noise results, i.e. {tilde over (P)}*Ñ. Since the complementary pattern {tilde over (P)} and the noise signal Ñ are not correlated, the noise signal ε is significantly smaller than the defect distribution signal s. In other words, the defect distribution signal s is amplified with respect to the noise in the decoding step.

Such defect detection method using a detection pattern and a complimentary pattern to localize defects on a sample also is possible in two dimensions (2D).

FIG. 10 shows an example for a detection pattern P, i.e. for a detection pattern 6 resulting from respective mask pattern 5 of a 2D structured mask 4 in two dimensions xy.

FIG. 11 shows the complimentary pattern {tilde over (P)} complimentary to the detection pattern P of FIG. 10.

FIG. 12 shows the convolution result P*{tilde over (P)}, i.e. the convolution of the two-dimensional detection pattern P of FIG. 10 and its complimentary pattern {tilde over (P)} of FIG. 11.

Further examples for such two-dimensional patterns can be found in A. Busboom et al. Experimental Astronomy 8: 97-123, 1998.

Performing the defect detection method in two dimensions requires scanning/moving the sample 8 relative to the detection pattern 6 of FIG. 10 in the two dimensions x and y, e.g. linewise along x with increments in y or vice versa, while gathering the detection light I(τx, τy) assigned to the respective scanning coordinates in x and y.

Alternatively, such two-dimensional scanning can be performed in a single x scanning step by using an array detector 9 having a sensor pixel array extending in the y-direction, i.e. at least one pixel row.

During the signal processing of the defect detection method, a matched filter is obtained by correlating the complementary pattern {tilde over (P)} with an unknown signal, i.e. the scanned intensity measured in equation (1) above, to detect the presence of a detection pattern P in the unknown signal. In that sense, the decoding step uses a filtering algorithm. This is equivalent to convolving the unknown signal with a conjugate time-reversed version of the pattern. {tilde over (P)} and P need to be assigned such that their convolution approximates a delta distribution as close as possible. Such approximation is possible via a ratio between a peak signal and a background signal of such convolution. Such ration should approach infinity, i. e. should be larger than 10³ or larger than 10⁴ or larger than 10⁵ or even larger.

The detector defect distribution s on the sample 8 then can be further evaluated to qualify the sample 8. Samples that pass such qualification step then can be further used in the lithography process to produce semiconductor structures, in particular memory chips or ASIC chips.

In some implementations, the decoding module, or a computer for performing various computations described in this document, such as the evaluation of the detector defect distribution, the decoding of the defect localization, the implementation of the filtering algorithm, can include one or more data processors for processing data, one or more storage devices for storing data, such as one or more databases, and/or one or more computer programs including instructions that when executed by the decoding module or the computer causes the decoding module or the computer to carry out the processes. The decoding module or the computer can include one or more input devices, such as a keyboard, a mouse, a touchpad, and/or a voice command input module, and one or more output devices, such as a display, and/or an audio speaker.

In some implementations, the decoding module or the computer can include digital electronic circuitry, computer hardware, firmware, software, or any combination of the above. The features related to processing of data can be implemented in a computer program product tangibly embodied in an information carrier, e.g., in a machine-readable storage device, for execution by a programmable processor; and method steps can be performed by a programmable processor executing a program of instructions to perform functions of the described implementations by operating on input data and generating output. Alternatively or addition, the program instructions can be encoded on a propagated signal that is an artificially generated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus for execution by a programmable processor.

In some implementations, the operations associated with processing of data described in this document can be performed by one or more programmable processors executing one or more computer programs to perform the functions described in this document. A computer program can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.

For example, the decoding module or the computer can be configured to be suitable for the execution of a computer program and can include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only storage area or a random access storage area or both. Elements of a computer include one or more processors for executing instructions and one or more storage area devices for storing instructions and data. Generally, a computer will also include, or be operatively coupled to receive data from, or transfer data to, or both, one or more machine-readable storage media, such as hard drives, magnetic disks, magneto-optical disks, or optical disks. Machine-readable storage media suitable for embodying computer program instructions and data include various forms of non-volatile storage area, including by way of example, semiconductor storage devices, e.g., EPROM, EEPROM, and flash storage devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM discs.

Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. The separation of various system components in the embodiments described above should not be understood as requiring such separation in all embodiments. The dependent claims listed below recite various features of the invention. Various combinations of the features recited in multiple dependent claims are also within the scope of the invention.

Claims

1. A method to detect a defect on a lithographic sample with the following steps: providing detection light and a detector with at least one sensor pixel,providing a detection pattern (P) causing a light structure of the detection light being structured at least along one dimension (1D, x),aligning the detection pattern (P) such that the detector is aligned normal to an extension (xy) of the light structure,providing a complimentary pattern ({tilde over (P)}) which has a 1D structure which is complementary to that of the detection pattern (P),moving the sample relative to the detection pattern (P) while gathering the detection light on the detector,providing a reference sample without defects or with negligible defects,moving the reference sample relative to the detection pattern (P) while gathering the detection light on the detector,decoding a defect (s1 to s4) localization on the sample by correlation using the complementary pattern ({tilde over (P)}).

2. The method according to claim 1, wherein the decoding step is done using a filtering algorithm.

3. The method according to claim 1, wherein the provision of the detection pattern (P) is done via a structured mask showing a model of the detection pattern as a mask pattern having pattern areas transmitting the detection light and pattern areas blocking the detection light.

4. The method according to claim 1, wherein the provision of the detection pattern (P) is done via a structured illumination of the sample with the detection light.

5. The method according to claim 1, wherein during the moving, the sample is periodically shifted.

6. The method according to claim 1, wherein during the moving, the sample is aperiodically shifted.

7. A metrology system to perform the method according to claim 1, having a detection light source,having a detector with at least one sensor pixel,having means to produce a detection pattern (P) causing a light structure of the detection light being structured at least along 1D,having a sample stage to move a sample held with the sample stage relative to the detection pattern andhaving a decoding module to detect a defect (s1 to s4) localization on the sample by a correlation using a complimentary pattern ({tilde over (P)}).

8. The metrology system according to claim 7, wherein the detector has a sensor pixel array extending at least in 1D.

9. The metrology system according to claim 7, wherein the detection pattern is produced via the mask pattern of a structured mask.

10. The metrology system according to claim 7, wherein the detection pattern (P) is produced via a structured illumination of a sample with the detection light.