JP2001050919A - Observation method for internal structure - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an observation method, in which a fine structure buried at the inside of a sample ca be observed clearly without destroying the sample. SOLUTION: In this observation method for an internal structure, an electron beam 3 is scanned, and a fine structure 1 which is embedded at the inside of a sample is observed and measured. The mean free path of electrons in the electron beam 3 is set at about the same extent as the depth to which the fine structure 1 as an object to be observed is embedded. That is to say, when the observation conditions of a scanning electron microscope and the surface state of the sample are controlled properly, fine structures which is embedded at the inside of the sample can be obserbed, and the observation method is effective in observing and measuring the three-dimensional shape of the fine structure.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an internal structure observation method for observing and measuring a fine structure embedded in a sample by scanning an electron beam.

[0002]

2. Description of the Related Art Microfabrication technology has been developed with the aim of achieving higher integration and higher performance of semiconductor integrated circuits. Currently 1
Processing of a structure of about 00 nm is possible, and further miniaturization is expected to continue in the future.

[0003] With the miniaturization, complicated three-dimensional structures are often used. Accordingly, the importance of observing a three-dimensional internal structure of a complicated fine structure has been increasing.

[0004] The methods of observing the internal structure include cross-sectional observation of a sample cross section with a scanning electron microscope (SEM), thinning of the sample cross section with a transmission electron microscope (TEM), and other than the target structure. In general, the shape is observed with an SEM or the like after removing the particles.

That is, in the conventional method, a method is employed in which the internal structure is exposed on the surface by a treatment before observation, and the internal structure is directly observed.

This is due to the nature of the electron beam used for observation. Since the electron beam can be narrowed down (1 nm or less), it is very suitable for observation of a fine structure (this is a property mainly used in SEM). Further, even when viewed as a wave, an image having a short wavelength and high resolution can be obtained (this is a property mainly used in TEM).

However, since an electron beam has a large interaction with a substance, it cannot generally penetrate a material (X
The inside of the structure can be seen through a line or the like because the interaction is moderately small.) Nevertheless, since the interaction with the substance is finite, it is possible to penetrate into the material to some extent. For example, in the case of an electron beam with an electron acceleration voltage of 200 kV, which is often used in a TEM, transmission is possible with a thin film of about 10 μm, depending on the material. TE
It is for this reason that the sample must be thinned for observation of the sample at M.

[0008] In the case of a TEM, transmitted electrons are used as a carrier of information. Even in a microscope using the same electron beam, the SEM uses secondary electrons emitted from the surface of a sample as information.
The main component of this secondary electron is relatively slow (around 10 eV)
And information on secondary electrons generated at a depth of about several nm to 10 nm from the sample surface. That is, in general, only the surface structure of the sample becomes an image in the SEM. For this reason, to observe the internal structure with a normal SEM, either a method of observing a fractured surface or removing a structure other than the observation target structure and exposing it to the surface is used.

As described above, in the conventional method for observing the internal structure using an electron microscope, it is necessary to expose the object to be observed to the surface in some form. However, with the miniaturization of the observation target (especially, a semiconductor integrated circuit), it has become very difficult to prepare a cross-sectional sample. Usually, the simplest method of preparing a cross-sectional sample is to break and divide the sample, but it is necessary to scribe and break at a desired position while visually observing the sample. Have difficulty.
In the preparation of a TEM cross-sectional sample, it is very difficult to reduce the thickness of a desired location because the thickness is reduced by polishing. In recent years, a cross-sectional sample preparation method using a focused ion beam (FIB) has been devised, but even with this method, positioning on the order of submicrons is difficult. Also,
Since the structure itself becomes brittle due to the miniaturization, processing for peeling of the film is also extremely difficult.

[0010]

The problem to be solved by the present invention is to obtain a clear internal structure image in a scanning electron microscope image.

With the miniaturization of the structure, it has become extremely difficult to prepare a sample for grasping the three-dimensional structure. That is, in cross-sectional observation, it is very difficult to accurately cut out a cross-sectional sample at a specific location. It is also very difficult to remove other structures without damaging the target structure. In view of these points, the method according to the present invention is an internal structure observation method that does not require pretreatment of a sample such as exposure of the internal structure.

Normally, in order to observe the internal structure of the structure in a non-destructive manner, a photon-based radiation source having a high transmittance such as X-rays is used.
However, an observation method using such a radiation source cannot obtain a high resolution in principle, and thus cannot be applied to observation of a fine structure used in a semiconductor circuit.
A charged particle beam represented by an electron beam can be easily converged and a thin beam can be obtained. Therefore, it is often used for observation of a fine structure. However, particle beams have much greater interaction with substances than photons, and are inferior in permeability.
Among various particle beams, electrons have a small interaction with a substance and a relatively high permeability. However, in the SEM, which is the most frequently used method of observing a fine structure, secondary electrons emitted from the surface of the sample are used as information, so that usually only the surface structure is observed. Also,
Until now, the main focus was on observing the surface structure,
This feature was useful. However, there is a need for nondestructive observation of the internal structure, and this problem was addressed using an SEM-like method.

An electron beam having a relatively high accelerating voltage penetrates into a material, and can be used as a probe for obtaining internal structure information. However, normal SEM
The low-speed secondary electrons used as the image information have a shallow escape depth of about several nanometers, and cannot be directly captured as internal structure information. For example, if reflected electrons are used as image information as one method, it is possible in principle to obtain information from a depth substantially equal to the penetration depth of incident electrons. However, in order to obtain a backscattered electron image, the atomic weight (molecular weight) of the constituent materials needs to be greatly different, and the S / N ratio is lower than that of a normal secondary electron image. Can not get. In particular, it is very difficult to obtain a clear image in a silicon / insulating film system, which is the mainstream of semiconductor circuits.

As another method, there is a so-called EB tester method. In this method, a voltage is applied to the internal structure, and the potential contrast is applied to the internal structure.
By converting into the amount of secondary electrons, an internal structure image can be obtained. However, in this method, the resolution is not so high because of the potential contrast.

Since the internal structure image is less intense than the surface shape image in both the backscattered electron image and the potential contrast image,
It is not possible to separate them by a normal observation method or to obtain only an internal structure image.

As described above, the conventional SEM method cannot clearly observe the internal structure of the sample without destruction.
This is due to the fact that many conventional methods originally used observation conditions set with a focus on surface observation. Accordingly, the present invention provides an observation condition and method for obtaining a clear image of the internal structure without destruction of the sample using a scanning electron beam suitable for the internal structure.

[0017]

When a material is irradiated with an electron beam, it penetrates into the material according to its energy, is scattered by atoms constituting the material, and is eventually absorbed. Among secondary electrons generated in the scattering process, a scanning electron microscope (SEM) uses low-energy secondary electrons generated near the sample surface (shallower than about 10 nm in depth) as main image information. . For this reason, ordinary SEM is mainly used to observe the surface structure. The present invention is a method for obtaining internal structure information using an SEM that can normally obtain only surface structure information, and this will be described below.

When obtaining information on the internal structure using an electron beam as a probe, it is necessary to first have sufficient energy so that the probe reaches the object to be observed. The penetration depth of the electron beam used for the probe (the depth at which the incident electrons can reach) must be larger than the internal structure depth. In addition, for high resolution observation, it is important that scattering is small at the position of the observation target and that the probe has little blur. This is referred to as the scattering radius. Both the penetration depth and the scattering radius depend on the target material and the accelerating voltage. From these two parameters, the higher the accelerating voltage, the more advantageous the observation of the internal structure. However, when the interaction between the electron beam and the material is considered, the higher the energy, the relatively smaller the amount of energy lost per unit length (the traveling direction of the electron beam). Therefore, when obtaining information on the internal structure of a specific size, the lower the energy, the more advantageous. Considering the above, it is predicted that an optimum acceleration voltage exists for the internal structure observation. Although the optimum acceleration voltage is considered to be different depending on the material system, the acceleration voltage is such that the mean free path of electrons in the material and the depth of the structure to be observed are substantially the same.

The mean free path of an electron refers to an average distance traveled by an incident electron beam until it is first scattered inside a material. This physical quantity is conceptual, and it is generally impossible to measure it effectively. It is known that this amount generally increases with increasing acceleration voltage. For example, in Si, the mean free path of electrons is 1.7 nm at an acceleration voltage of 1 keV, 9.4 nm at 5 keV, and 58 at 30 keV.
There is a simulation result of 370 nm at 200 keV. Although no general relation is known, S
Within the range of the acceleration voltage used in the EM, the mean free path of the electrons is approximately proportional to the acceleration voltage.

For example, for a Si structure buried in a Si oxide film, a structure having a depth of about 10 nm is approximately 5 kV, and a structure having a depth of about 100 nm is several tens kV, which is on the order of microns. Several hundred k in depth
By using a V electron beam as a probe, an image with high contrast and high resolution can be obtained.

In the case of a material composed of a so-called light element such as a Si-based material, the beam blur (scattering radius) at a depth of about the mean free path of electrons is estimated to be about 1 nm, so that a sufficiently thin probe is used. Allows observation. (Set the acceleration voltage to about the mean free path of electrons)
Is the first condition of the internal observation, which is the condition for the probe (this condition must be satisfied).

Next, there is a problem of separation between the internal structure information and the surface structure information. Usually, in the case of shape observation by SEM,
The difference in secondary electron emission amount resulting from the unevenness of the surface structure and the difference in material is used as image information. For this reason, it is common that irregularities and dissimilar material distributions usually present on the surface of the observation target appear as images. SEM can also image minor differences in surface structure (a useful property for normal SEM methods).

However, since image information derived from the internal structure is known to be weaker than surface structure information, images derived from such various surface structures are required to observe the internal structure. It is an obstacle. Therefore, it is desirable that the material be made of a uniform material with as little unevenness as possible. If the sample surface can be composed of a uniform material with few irregularities (a state in which an SEM image in a normal meaning cannot be obtained), delicate internal information can be effectively used. That is, the second requirement of the internal observation is that the unevenness of the surface structure is reduced and made uniform (not an essential requirement).

By setting the probe conditions and sample surface conditions so as to satisfy the two requirements, the internal structure can be clearly observed as a secondary electron image. Relatively deep (10 nm or more) internal structure information is stored on the sample surface (about 10
is converted into the amount of secondary electrons emitted from the electron beam (within nm), and this can be visualized. That is, by making the electron beam probe interact with the internal structure in a sufficiently thin state (first requirement) and homogenizing the surface structure (second requirement), the delicate interior that could not be used as information in the past was obtained. Image information derived from the structure can be used.

That is, the present invention provides an internal structure observation method for observing and measuring a fine structure embedded in a sample by scanning an electron beam. It is characterized in that the mean free path of electrons of the electron beam is substantially the same. That is, the electron beam is scanned by adjusting the accelerating voltage of the electron beam so that the mean free path of the electrons of the electron beam is substantially the same as the depth of the structure to be observed.

[0026]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiment 1 FIG. 1A is a view showing an internal structure observation method according to Embodiment 1 of the present invention, and FIG. 1B is a diagram showing a secondary electron intensity distribution derived from the internal structure from above. It is a figure which shows the secondary electron intensity distribution derived from a surface structure, and the secondary electron intensity distribution of the total amount. In FIG. 1A, reference numeral 2 denotes an Si oxide film, 1 denotes an internal Si structure embedded in the Si oxide film 2, and 3 denotes an electron beam having the same depth as the embedded internal Si structure 1 to be observed. This is an electron beam whose mean free path has been adjusted.

FIG. 3A shows a conventional surface observation method, and FIG. 3B shows, from above, a secondary electron intensity distribution derived from the internal structure, a secondary electron intensity distribution derived from the surface structure, and a secondary of the total amount. It is a figure showing an electron intensity distribution. In FIG. 3A, the same components as those in FIG. 1A indicate the same components. Reference numeral 5 denotes an electron beam whose mean free path of electrons of the electron beam is shallow near the surface.

As shown in FIG. 1A, consider the observation of a structure in which a Si structure 1 is embedded in a Si oxide film 2. S
The oxide film thickness on the i-structure 1 is about 100 nm. It is important to select the accelerating voltage for observing the internal structure,
Here, the structure (depth of the internal Si structure 1) and the material (Si
The voltage is set to 30 kV in consideration of the oxide film 2).

As a result, the observation target depth and the mean free path of the electrons in the oxide film substantially coincide with each other, so that the internal structure information can be easily obtained. Simulates the mean free path). Further, under these conditions, the spread due to the scattering of the electron beam near the internal structure is estimated to be about 2 nm, so that observation with sufficiently high resolution is possible.

In this state, the secondary electron image is shown in FIG.
As shown in (b), the secondary electron component derived from the surface structure and the secondary electron component derived from the internal structure are superimposed and observed, but information on the internal structure can be obtained. .

In normal SEM observation, since the purpose is to observe only the surface structure, a relatively low acceleration voltage of, for example, 1 kV is used. In this case, FIG.
As shown in the figure, the electron beam is absorbed only on the surface,
The secondary electron signal intensity is as shown in FIG.
It reflects only the surface structure.

When the acceleration voltage is increased to about 5 kV, the electron beam can reach the internal Si structure 1. However, at this level of acceleration voltage, the effective beam diameter becomes about 100 nm due to scattering in the material, so that the internal structure cannot be observed (effectively, the acceleration voltage near this (several kV)). In this case, observation of the surface structure becomes difficult due to charge-up of the oxide film 2). If the acceleration voltage is further increased to about 30 kV, an internal structure image can be obtained as described above. If the accelerating voltage is further increased to, for example, about 200 kV, the effective beam diameter is slightly reduced, but the interaction with the Si structure is reduced, so that the contrast is reduced and the effective resolution is not significantly changed by this influence. Or, in some cases, deteriorate. In short, it is important to set the observation conditions to the accelerating voltage that gives the mean free path of the electrons in the material through which the electron beam passes at the same depth as the buried depth of the structure to be observed. Observations can be made.

In the present embodiment, an internal structure having a depth of about 100 nm is assumed, but, of course, the optimum acceleration voltage condition changes depending on the structure depth and the material. For example, in the case of a Si-based material (Si, Si oxide film, various Si oxide-based materials), if the structure has a depth of 20 to 30 nm, 10 k is used.
When observing a structure having a depth of about V and about 100 nm, about 30 to 50 kV and several hundreds in the order of 1 μm.
An acceleration voltage of kV or more is desirable. In addition, a heavier (lower electron beam transmission) material requires a higher acceleration voltage than a lighter material.

Embodiment 2 Secondary electrons derived from the surface structure and secondary electrons derived from the internal structure
Since it is impossible to select the next electron, the two components always overlap. If there is little change in the surface shape on the internal structure of the observation target, and if there is no large change in the secondary electron emission efficiency depending on the location, 2
The change in the amount of secondary electrons can be clearly observed as an image. However, in general, the surface structure has irregularities to some extent corresponding to the internal structure, and an image derived from this surface structure overlaps the internal structure image. In such a state, it is difficult to accurately grasp the internal structure. In order to prevent this, it is necessary to reduce surface irregularities and to make the material uniform.

FIG. 2 (a) is a diagram showing an internal structure observation method according to the second embodiment of the present invention, and FIG. 2 (b) is a secondary electron intensity distribution derived from the internal structure from above, and secondary electrons derived from the surface structure. It is an intensity distribution, and a secondary electron intensity distribution of the total amount. In FIG. 2A, the same reference numerals as those in FIG. 1A indicate the same components. Reference numeral 3 'denotes an electron beam whose mean free path of electrons of the electron beam is adjusted to the same depth as the depth in which the internal Si structure 1 to be observed is buried, and 4 denotes a light element mainly deposited on the sample surface. It is a coating.

As shown in FIG. 2A, for example, a carbon-based coating 4 (so-called,
The purpose can be achieved by applying a contamination film to the observation target region of the sample. The present inventors have confirmed that when the coating film 4 is applied on the Si oxide film, the secondary electron image derived from the surface irregularities of the oxide film 2 may be almost completely erased. Since the surface irregularities of the coating 4 are considerably smaller than the irregularities before the application, the image derived therefrom is considerably weakened, and the internal structure image can be effectively observed clearly (FIG. 2B). ). Coating 4 to be applied
May have a thickness of about 10 nm to several tens of nm. For example, if an acceleration voltage of several tens of kV or more is used, there is almost no effect of the presence or absence of the film on the transmittance.

In this embodiment, a contamination film which can be applied in the apparatus at the time of observation by SEM is used as the coating 4 to be applied to the surface of the sample. A system film may be applied. In short, it is only necessary to apply a film having a thickness that does not hinder the electron beam transmission ability and that alleviates surface irregularities. In order not to hinder the electron beam transmission ability, a light element is preferably used, and a carbon-based material is most suitable for this purpose. Since the contamination film is generally not a dense film, the surface is unclear and the effect of eliminating irregularities on the sample surface is very strong.However, if it is desired to remove the film later, for example, by spin coating, A thin resist material formed from a solution may be applied. From the viewpoint of alleviation of unevenness, a film such as a resist material formed from a solution by spin coating is preferable. However, there is no problem if a desired effect can be obtained even with a carbon film applied by sputtering or vapor deposition.

Although the present invention has been specifically described based on the embodiments, the present invention is not limited to the above-described embodiments, and it is needless to say that various modifications can be made without departing from the scope of the invention. It is.

[0039]

As described above, according to the present invention,
By appropriately controlling the observation conditions and the sample surface state of the scanning electron microscope, a fine structure embedded in the sample material can be observed, which is effective in observing and measuring the three-dimensional shape of the fine structure.

[Brief description of the drawings]

FIG. 1 (a) is a diagram showing an internal structure observation method according to Embodiment 1 of the present invention, FIG. 1 (b) is a secondary electron intensity distribution derived from an internal structure, a secondary electron intensity distribution derived from a surface structure, It is a figure which shows the secondary electron intensity distribution of the total amount.

2A is a diagram showing an internal structure observation method according to Embodiment 2 of the present invention, FIG. 2B is a secondary electron intensity distribution derived from an internal structure, a secondary electron intensity distribution derived from a surface structure, It is a secondary electron intensity distribution of the total amount.

FIG. 3A is a diagram showing a conventional surface observation method, and FIG.
FIG. 3 is a diagram showing a secondary electron intensity distribution derived from an internal structure, a secondary electron intensity distribution derived from a surface structure, and a total secondary electron intensity distribution.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Internal Si structure, 2 ... Si oxide film, 3, 3 '... Si
An electron beam whose mean free path of electrons is adjusted to the same depth as the depth at which the structure is embedded, a coating mainly composed of light elements,
5. An electron beam whose electron mean free path is shallow.

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] December 2, 1999 (1999.12.
2)

[Procedure amendment 1]

[Document name to be amended] Drawing

[Correction target item name] Figure 2

[Correction method] Change

[Correction contents]

FIG. 2

[Procedure amendment 2]

[Document name to be amended] Drawing

[Correction target item name] Figure 3

[Correction method] Change

[Correction contents]

FIG. 3

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2G001 AA03 BA07 CA03 GA06 GA09 KA20 LA11 NA07 NA20 RA08 RA20 4M106 AA20 BA02 CA70 DH24 DH60 5C033 UU03

Claims (2)

[Claims] 1. An internal structure observation method for observing and measuring a fine structure embedded in a sample by scanning an electron beam, wherein the mean free path of electrons of the electron beam is determined by observing the structure. A method for observing an internal structure, wherein the depth is approximately the same as the embedded depth. 2. The method for observing an internal structure according to claim 1, wherein a thin film mainly composed of a light element is deposited on the surface of the sample.