JPH1194847A - Scanning type probe microscope - Google Patents

Abstract

PROBLEM TO BE SOLVED: To highly accurately detect actual information of a sample surface, by providing a fetch means for a plurality of line data, a frequency-analyzing means, a noise-filtering means and an imaging means, and completely removing only a noise component to be removed. SOLUTION: For instance, the same line of the same object to be measured is scanned by one line under measurement conditions of scan speeds of 1 Hz and 2 Hz. Respective line data 2a, 2b are obtained because of influences of superposed noise components on scan signals. The line data 2a, 2b are fetched and temporarily preserved in a line buffer 14 via an A/D converter 12. These two line data 2a, 2b are processed in a predetermined signal process on the basis of a preliminarily set algorithm via a subtraction circuit 16, a D/A converter 18 and a fast Fourier converter 20, whereby only noise components included in the data are extracted. The noise components are filter by a noise-filtering means 8. Thereafter, the data are fetched into a line buffer 28 via an A/D converter 26, and an addition value of the data via an addition circuit 30 is stored in an image buffer 32. The addition value stored in the image buffer 32 is a line data without noises which accurately represents a shape of the object to be measured.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a scanning probe microscope for measuring surface information of a sample with, for example, atomic order resolution.

[0002]

2. Description of the Related Art Conventionally, a scanning probe microscope (SPM:
Scanning Probe Microscope)
(STM: Scanning Tunneling Microscope) with a (Binnig) or roller (Rohrer)
Has been invented. However, in STM, observable samples are limited to conductive samples. Therefore, an atomic force microscope (AFM) has been proposed as a device capable of observing an insulating sample at an atomic order resolution by using elemental technologies of the STM such as a servo technology (Japanese Patent Application Laid-Open No. 62-130302).

The AFM structure is positioned as one of the SPMs, and includes a cantilever having a sharpened probe at a free end, and a scanner for relatively moving the probe and the sample.

In such a configuration, when the probe is brought close to the sample in opposition, the free end of the cantilever is caused by an interaction (atomic force) acting between atoms at the tip of the probe and atoms on the surface of the sample. Displace. Then, the probe and the sample are relatively scanned in the XY directions along the sample surface while electrically or optically detecting the displacement amount generated at the free end, thereby obtaining surface information of the sample (for example, Unevenness information)
Is measured three-dimensionally.

In order to realize high resolution in such SPM measurement, a scanner capable of controlling the relative distance between the probe and the sample with high accuracy is required. Therefore, as a scanner, a tri-pot or tube-type piezoelectric scanner using a piezoelectric material is generally used (see Japanese Patent Application Laid-Open No. 63-236992).

[0006]

However, even when a high-precision scanner is used, noise may be added to SPM measurement data due to thermal drift, electrical noise, mechanical vibration, or the like of the SPM device.

Specifically, considering the case where the surface of the sample on which the mesh pattern is formed is scanned at a predetermined scanning speed and the surface information of the sample is measured, the scanning speed is several tens to several hundred times the scanning speed. Vibration noise of frequency and several tens
Electric noise having a frequency of 0 or more, noise due to thermal drift of several tenths or less, and the like are included in the SPM measurement data. As a result, as shown in FIG.
The measurement data is in a state where noise is superimposed on the actual surface information, and the measurement accuracy is reduced. In addition, FIG.
Is line data along the line bb in FIG.
FIG. 7C shows line data along the line cc in FIG.

In this case, for example, a frequency analysis is performed on the SPM measurement data (see Japanese Patent Application Laid-Open No. H5-248813), and the frequency component of the noise is removed based on the analysis result. Is also conceivable.

However, the noise component removing method has the following problems. That is, as shown in FIG. 9, when noise is included in the SPM measurement data (see FIGS. 9A and 9B), a frequency analysis is performed to determine the frequency component of the noise (the symbol P in FIG. 9D). When the frequency components within the range indicated by are removed, the same frequency components (actual surface information components) as the noise components originally included in the surface shape of the sample
Is also removed (see FIG. 4D). As a result, in the SPM measurement data after the frequency analysis processing, the same frequency component as the removed actual surface information component becomes noise and rides on the SPM measurement data (see FIG. 3C).

As described above, in the conventional noise component removing method, it is difficult to extract the frequency component to be removed because the actual SPM measurement data and the noise component cannot be distinguished. Therefore, it becomes difficult to detect SPM measurement data corresponding to the actual sample surface shape with high accuracy.

The present invention has been made to solve such a problem, and an object of the present invention is to accurately detect actual sample surface information by completely removing only noise components to be removed. It is an object of the present invention to provide a scanning probe microscope.

[0012]

In order to achieve the above object, a scanning probe microscope according to the present invention moves a measuring object and a probe along a predetermined scanning line defined by the measuring object. When scanning is performed a plurality of times relatively, capturing means for capturing a plurality of line data having different measurement conditions, and performing a frequency analysis on the plurality of line data to reduce noise components included in the plurality of line data, respectively. Frequency analysis means for extracting, noise removal means for removing noise components from the plurality of line data, and an image for imaging the object to be measured by performing predetermined arithmetic processing on the plurality of line data from which the noise components have been removed Conversion means.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A scanning probe microscope according to a first embodiment of the present invention will be described below with reference to FIGS. Although not particularly shown, the scanning probe microscope (SPM) of the present embodiment includes a cantilever having a sharpened tip at the tip, a displacement sensor for optically detecting the displacement of the cantilever, a probe and a measuring object. A scanner for relatively moving the object. Then, by driving the scanner based on the output of the displacement sensor, the relative distance between the probe tip and the surface of the measurement object can be feedback-controlled, and the scanner is driven. Thus, the scanning speed and scanning direction of the probe with respect to the surface of the measurement target can be controlled.

In such a configuration, as shown in FIG. 2, the scanning probe microscope according to the present embodiment includes a measuring object and a probe along a predetermined scanning line defined by the measuring object. A plurality of line data 2 having different measurement conditions such as a scanning speed when scanning is performed a plurality of times relatively, and a frequency analysis is performed on the plurality of line data 2 so as to be included in each line data 2. Frequency analysis means 6 for extracting the noise component, noise removal means 8 for removing the noise component from the plurality of line data 2, and performing predetermined arithmetic processing on the plurality of line data 2 from which the noise component has been removed. And an imaging means 10 for imaging the object to be measured.

As the measurement object, for example, a sample having a mesh pattern formed on the surface can be used. FIG. 2A shows a configuration for extracting a noise component, and FIG. 2B shows a configuration for extracting a noise component.
2 shows a configuration for removing the noise component extracted by the above configuration.

As shown in FIG. 2A, as an example, an acquisition unit 4 applied to this embodiment includes an A / D converter 12 for digitally converting each line data 2 and an A / D converter And a line buffer 14 for storing each line data 2 output via the line 12.

Frequency analysis means 6 applied to this embodiment
Is a subtraction circuit 16 for performing a subtraction process on each line data 2 stored in the line buffer 14, a D / A converter 18 for converting the subtraction value of the subtraction circuit 16 into an analog signal, and a predetermined algorithm ( For example, by performing predetermined signal processing on the output of the D / A converter 18 based on a noise component extraction program), only noise components (noise frequency components) included in each line data 2 are extracted. Fast Fourier transform (F
FT (Fast Fourier Transform) analyzer 20 (hereinafter referred to as FFT analyzer). If the FFT function is incorporated in the software of the SPM, the noise component included in each line data 2 can be extracted by software processing without using the frequency analysis unit 6.

As shown in FIG. 2B, the noise removing means 8 applied to the present embodiment includes a plurality of frequency filters 22 and 24 for removing only noise components from each line data 2 as an example. It is configured.

The imaging means 10 applied to the present embodiment
Is, as an example, an A / D converter 26 for digitally converting each line data 2 from which a noise component has been removed;
A line buffer 28 for storing each line data 2 output through the A / D converter 26, an addition circuit 30 for performing an addition process on each line data 2 stored in the line buffer 28, and an addition circuit 30 And an image buffer 32 for imaging the object to be measured based on the added value of.

Next, the operation of the present embodiment (noise removal scanning processing) will be described. Now, for example, at a scanning speed of 1H
Under the measurement condition of z, when one line of the object to be measured (for example, a sample having a mesh pattern formed on the surface) is scanned,
A noise component superimposed on the scanning signal (hereinafter, noise component Sn
1 (refer to FIG. 1B)), FIG.
Line data as shown in (a) (hereinafter referred to as line data 2a) is obtained.

At this time, the line data 2a is fetched into the line buffer 14 via the A / D converter 12, and then temporarily stored in the line buffer 14. continue,
For example, when one line is scanned on the same line of the same object under the measurement condition of a scanning speed of 2 Hz, a noise component superimposed on the scanning signal (hereinafter, noise component Snb (FIG. 1)
(D), line data (hereinafter, referred to as line data 2b) as shown in FIG. 1C is obtained. As a result, two line data 2 having different measurement conditions (scan speeds) along the same scan line.
a and 2b are obtained.

At this time, the line data 2b is fetched into the line buffer 14 via the A / D converter 12, and then temporarily stored in the line buffer 14. Next, the two line data 2 stored in the line buffer 14
a and 2b are subjected to a subtraction process via a subtraction circuit 16. Then, the subtracted value is subjected to FFT via the D / A converter 18.
The signal is output to the analyzer 20 and subjected to predetermined signal processing based on a preset algorithm. As a result, FIG.
As shown in (e) and (f), each line data 2a, 2
Only the noise components Sna and Snb included in b can be extracted.

Here, the noise removing means 8 is initialized so as to remove the noise components Sna and Snb, respectively. Specifically, the filtering value of the frequency filter 22 is initialized so as to cut the frequency band W1 (see FIG. 1 (h)) including the noise component Sna.
Frequency band W2 including noise component Snb (FIG. 1 (j)
) Is initialized so that the filtering value of the frequency filter 24 is cut. In this case, by adding W1 and W2, each filtering value is initialized so that all frequencies in one-line scanning are covered.

After the completion of the initial setting, the same scanning process as described above is performed again to obtain line data 2a and 2b. Subsequently, a filtering process is performed on the line data 2a by the frequency filter 22 (see FIG. 1H). As a result, the line data 2a becomes line data 2a 'as shown in FIG. At the same time, the line data 2b is filtered by the frequency filter 24 (see FIG. 1 (j)). As a result, the line data 2b is
The line data 2b 'as shown in FIG.

The noise component Sn by the noise removing means 8
line data 2a ', 2b' from which a and Snb have been removed
Are both taken into the line buffer 28 via the A / D converter 26 and then temporarily stored in the line buffer 28.

Next, the 2 stored in the line buffer 28 is
After performing an addition process on the two line data 2a 'and 2b' via the addition circuit 30, the added value is stored in the image buffer 32.
To save. In this state, since the noise components Sna and Snb have been completely removed from all the frequencies during one-line scanning (see FIG. 1 (l)), the added value stored in the image buffer 32 is as shown in FIG. As shown in ()), the data is noise-free line data that accurately represents the shape of the measurement object.

Therefore, by performing the above-described noise removal scanning processing over all the scanning lines defined for the measurement object, the actual surface information of the measurement object from which noise has been completely removed (for example, Surface irregularities) can be measured.

As described above, according to the present embodiment, by completely removing only noise components to be removed, it is possible to detect measurement data corresponding to the actual surface shape of the object to be measured with high accuracy. A simple scanning probe microscope can be provided.

Next, a scanning probe microscope according to a second embodiment of the present invention will be described with reference to FIGS. The scanning probe microscope according to the present embodiment has the same configuration as that of the scanning probe microscope except that an offset processing system (not shown) is provided between the addition circuit 30 and the image buffer 32 (see FIG. 2B). Since the configuration is the same as that shown in FIGS. 2A and 2B, description thereof is omitted.

As in the first embodiment, even if the noise component of the line data along the scanning direction (tentatively, the X direction) is removed, the direction crossing the scanning direction (X direction) That is, in the line data in the orthogonal direction (Y direction), a noise component due to thermal drift in the Y direction or the like remains without being removed.

For this reason, as shown in FIG. 3, the surface information (measurement data) of the measuring object 34 is
4 is distorted with respect to the surface shape. In this case, in order to obtain the actual surface shape of the measurement object 34, offset processing is performed on the measurement data so that the distortion location matches the actual surface 34a (see FIGS. 3 and 5). (A process of displacing in the direction of arrow F).

Therefore, the scanning probe microscope according to the present embodiment, when scanning the same measurement range in directions orthogonal to each other, as in the first embodiment, sets a plurality of line data having different measurement conditions such as scanning speed. Then, by performing offset processing on each of these line data, actual surface information of the measurement object can be measured.

Next, the operation (noise removal scanning process) of this embodiment will be described with reference to FIGS.
First, a plurality of line data in a scanning direction orthogonal to each other in the same measurement range are captured by the capturing unit 4 (see FIG. 2A) (step S1).

Specifically, the scanning directions orthogonal to each other are represented by X
When the scanning direction (0 ° scanning direction) and the Y scanning direction (90 ° scanning direction) are defined, a plurality of line data X (1) to X (n) along the X scanning direction are taken in, and along the Y scanning direction. A plurality of line data Y (1) to Y (m) are fetched (see FIG. 4). Note that n and m are positive integers (natural numbers), respectively.

The noise removal process in the X scanning direction and the noise removal process in the Y scanning direction will be described separately. First, the noise removal process in the X scanning direction will be described.

For example, the acquisition time of the first line data X (1) in the X scanning direction is the time required for scanning one line. The acquisition time of the line data in the X direction formed by connecting the intersection points T1 with the line data Y (1) to Y (m) is the time required for scanning a plurality of lines over the entire measurement range.

Next, the fetch time of the second line data X (2) in the X scanning direction is the time required for scanning one line. The acquisition time of the line data in the X direction formed by connecting the intersection points T2 with the line data Y (1) to Y (m) is the time required when scanning a plurality of lines over the entire measurement range.

The same data fetching process as described above is performed for the third and subsequent lines, and the fetching time of the n-th line data X (n) in the X scanning direction is determined when one line is scanned. , The line data X (n)
The acquisition time of the line data in the X direction formed by connecting the intersection points Tn of the line data Y (1) to Y (m) in the Y scanning direction is required when a plurality of lines are scanned over the entire measurement range. Time.

By performing such a data fetching process from the first to the n-th, the same operation as in the first embodiment (noise removal scanning process) is performed on the same line along the X scanning direction. , Two n-line data having different measurement conditions (data acquisition time) can be obtained.

Subsequently, n along the X scanning direction (scanning direction 0 °) is detected by the frequency analysis means 6 (see FIG. 2A).
Frequency analysis is performed on the line data (step S2),
At the same time, X obtained by Y scanning (scanning direction 90 °)
Frequency analysis is performed on the n-line data in the direction (step S3). As a result, only the noise components included in each of the two n-line data are extracted.

Next, as in the first embodiment, after initializing the noise removing means 8 (see FIG. 2 (b)), two frequency filters 22 and 24 (see FIG. 2 (b)) A filtering process is performed on the n-line data (steps S4 and S5). As a result, the noise component in the X scanning direction is completely removed from the two n-line data.

Thereafter, the two n-line data from which the noise components have been removed are added to the adder circuit 30 (see FIG. 2B).
Performs an addition process (step S6). Then, the processes in steps S2 to S6 are performed for all scan lines over the entire measurement range (step S7).

As a result, as shown in FIG. 6A, measurement data (surface information of the object to be measured) from which the noise component in the X scanning direction has been removed is obtained. Note that a noise component in the Y scanning direction remains in the measurement data.

Thereafter, an average of the data for n lines is obtained, the average value is temporarily stored as average line data in the X direction, and an average of the data for m lines is obtained, and the average value is averaged in the Y direction. The data is temporarily stored as line data (step S8). In the present embodiment, a memory circuit (not shown) built in the offset processing system can be used for this storage processing.

Next, the noise removal process in the Y scanning direction will be described. For example, the acquisition time of the first line data Y (1) in the Y scanning direction is the time required for scanning one line, but the line data Y (1) and X
Intersection T with line data X (1) to X (n) in the scanning direction
The acquisition time of the line data in the Y direction formed by connecting 1 to Tn is the time required when scanning a plurality of lines over the entire measurement range.

Next, the fetching time of the second line data Y (2) in the Y scanning direction is the time required for scanning one line. Of intersections T1 to T with line data X (1) to X (n)
The acquisition time of the line data in the Y direction formed by connecting n is the time required when scanning a plurality of lines over the entire measurement range.

The data acquisition process similar to that described above is performed for the third and subsequent data, and the acquisition time of the m-th line data Y (m) in the Y scanning direction is determined when one line is scanned. The line data Y (m)
The acquisition time of the line data in the Y direction formed by connecting the intersections T1 to Tn of the line data X (1) to X (n) in the X scanning direction is obtained when a plurality of lines are scanned over the entire measurement range. It is the time required.

By performing such a data fetching process from the first to the m-th data, similar to the operation of the first embodiment (the noise removal scanning process), the data fetch process is performed on the same line along the Y scanning direction. , Two n-line data having different measurement conditions (data acquisition time) can be obtained.

Subsequently, by the frequency analysis means 6 (see FIG. 2A), n along the Y scanning direction (scanning direction 0 °)
Frequency analysis is performed on the line data (step S9),
At the same time, Y obtained by X scanning (scanning direction 90 °)
Frequency analysis is performed on the n-line data in the direction (step S10). As a result, only the noise components included in each of the two n-line data are extracted.

Next, as in the first embodiment, after initializing the noise removing means 8 (see FIG. 2B), two frequency filters 22 and 24 (see FIG. 2B) are used. A filtering process is performed on the n-line data (steps S11 and S12). As a result, the noise component in the Y scanning direction is completely removed from the two n-line data.

Thereafter, the two n-line data from which the noise components have been removed are added to the adder circuit 30 (see FIG. 2B).
Is added (step S13). Then, the processes of steps S9 to S13 are performed for all scan lines over the entire measurement range (step S14).

As a result, as shown in FIG. 6B, measurement data (surface information of the measurement object) from which the noise component in the Y scanning direction has been removed is obtained. Note that a noise component in the X scanning direction remains in the measurement data.

Thereafter, an average of the data for m lines is obtained, the average value is temporarily stored as average line data in the Y direction, and an average of the data for n lines is obtained. The data is temporarily stored as line data (step S15). In the present embodiment,
For this storage processing, a memory circuit (not shown) built in the offset processing system can be used.

Next, in the offset processing system, the average line data in the X direction for the n lines obtained in step S8 and the average line data in the X direction for the n lines obtained in step S15 mutually match. As shown in FIG.
An offset process is performed on the measurement data (n-line data in the X scanning direction) shown in (a) (step S16).

In the offset processing system, the average line data in the Y direction on the m line obtained in step S8 and the average line data in the Y direction on the m line obtained in step S15 match each other. FIG. 6
The same effect can be realized by performing an offset process on the measurement data (m-line data in the Y scanning direction) shown in (b).

Then, in the image buffer 32 (see FIG. 2B), the surface information (surface irregularities) of the measuring object is reconstructed based on the offset-processed n-line data (step S17). . At this time, measurement data from which noise components in the X scanning direction and the Y scanning direction are completely removed is obtained. As a result, as shown in FIG. 5, it is possible to measure the actual surface information (surface irregularities) of the measurement object 34 from which the distortion has been completely removed.

As described above, according to the present embodiment, by completely removing only the noise component to be removed, it is possible to detect the measurement data corresponding to the actual surface shape of the object to be measured with high accuracy. A simple scanning probe microscope can be provided.

In the present embodiment, steps S8,
Measurement data without any distortion as shown in FIG. 5 can be obtained without performing S15 and S16. For example, X
Measurement data from which noise components in the scanning direction have been removed (FIG. 6
(See (a)), the (m-3) th line data in the Y direction is defined as Ly (m-3). In this case, the line data Ly (m-3) is distorted due to the influence of a noise component in the Y scanning direction. Therefore, this distortion line is set to the (m−m) -th in the measurement data (see FIG.
3) An offset value over the entire distortion line is set so as to match the Y-th line data Ly (m−3). Then, based on this offset value, all the line data X (1) to X (n) along the X scanning direction
Is subjected to an offset process. As a result, as shown in FIG. 5, it is possible to measure actual surface information (for example, surface unevenness) of the measurement object 34 from which distortion has been completely removed.

For example, in the measurement data from which the noise component in the Y scanning direction has been removed (see FIG. 6B), the (n-2) th line data in the X direction is represented by Lx (n-
2). In this case, the line data Lx (n
-2) is distorted due to the influence of noise components in the X scanning direction. Therefore, this distortion line is measured data (FIG. 6) from which noise components in the X scanning direction have been removed.
(See (a)), an offset value is set over the entire distortion line so as to match the (n-2) th line data Lx (n-2) in the X direction. Then, based on this offset value, an offset process is performed on all the line data Y (1) to Y (m) along the Y scanning direction. As a result, as shown in FIG. 5, it is possible to measure actual surface information (for example, surface unevenness) of the measurement object 34 from which distortion has been completely removed.

[0060]

According to the present invention, it is possible to provide a scanning probe microscope capable of accurately detecting actual sample surface information by completely removing only noise components to be removed. .

[Brief description of the drawings]

FIGS. 1A to 1L are diagrams showing a noise removal scanning process applied to a first embodiment of the present invention.

FIG. 2A is a block diagram illustrating a configuration of a capturing unit and a frequency analyzing unit applied to the first embodiment of the present invention, and FIG. 2B is a block diagram illustrating the configuration of the first embodiment of the present invention; FIG. 2 is a block diagram showing a configuration of an applied noise removing unit and an imaging unit.

FIG. 3 is a perspective view showing a state where distortion occurs in measurement data.

FIG. 4 is a view showing a process of taking in line data in a noise removal scanning process applied to the second embodiment of the present invention;

FIG. 5 is a perspective view showing measurement data without distortion.

FIG. 6A is a perspective view showing measurement data obtained by a noise removal process in the X scanning direction applied to the second embodiment of the present invention, and FIG. 6B is a perspective view showing the second embodiment of the present invention. FIG. 17 is a perspective view showing measurement data obtained by a noise removal process in the Y scanning direction applied to the embodiment.

FIG. 7 is a flowchart illustrating a noise removal scanning process applied to the second embodiment of the present invention.

FIG. 8A is a perspective view showing a state in which a noise component is superimposed on measurement data measured by a conventional scanning probe microscope, and FIG. 8B is along a line bb in FIG. 8A. Sectional drawing, (c) is sectional drawing which follows the cc line | wire of the same figure (a).

FIG. 9 is a diagram showing a conventional noise removal process,
(A) is a diagram showing measurement data before noise removal, (b)
FIG. 4A is a diagram showing a result of frequency analysis of measurement data before noise removal, FIG. 4C is a diagram showing measurement data after noise removal, and FIG. 4D is a diagram showing a result of frequency analysis of measurement data after noise removal.

[Explanation of symbols]

 2 line data 4 acquisition means 6 frequency analysis means 8 noise removal means 10 imaging means

Claims (3)

[Claims] An acquisition means for acquiring a plurality of line data having different measurement conditions when the measurement object and the probe are relatively scanned a plurality of times along a predetermined scanning line defined for the measurement object. Frequency analysis means for extracting a noise component included in each of the plurality of line data by performing frequency analysis on the plurality of line data; and noise elimination for removing the noise component from the plurality of line data. A scanning probe microscope comprising: means for performing predetermined arithmetic processing on the plurality of line data from which the noise component has been removed to thereby image the object to be measured. . 2. The method according to claim 1, wherein the scanning unit scans the probe a plurality of times along the same scan line defined on the measurement target.
2. The scanning probe microscope according to claim 1, wherein a plurality of line data based on different scanning speeds can be captured. 3. The capturing means can capture a plurality of line data having different measurement conditions when the probe is scanned a plurality of times along scanning lines defined in the same measurement range of the measurement object and orthogonal to each other. The scanning probe microscope according to claim 1, wherein the scanning probe microscope is configured as follows.