JPH05152390A - Electron beam diagnosis device - Google Patents

Abstract

PURPOSE:To obtain a device which is able to quickly optimize an S-curve so as to enhance utilization factor and measurement accuracy by a method wherein an electron optical system is reset in control parameters by a parameter setting means until an S-curve becomes adequate in shape when it is judged by a shape judging means that an S-curve is not adequate. CONSTITUTION:An electron optical system 21 equipped with an electron beam irradiating means and a secondary electron detecting means, a parameter setting means 22, and a secondary electron signal processing means 23 are provided. Furthermore, an electron beam diagnosis device is equipped with a shape judging means 24 which judge whether a secondary electron energy analysis curve (S-curve) is adequate in shape as prescribed or not, where the S-curve is obtained by the secondary electron signal processing means 23 basing on the mount of secondary electron signal at analysis voltages set by the parameter setting means 22. When it is judged by the shape judging means 24 basing on its measurement result that the S-curve is not adequate in shape, the control parameters of the electron optical system 21 continue to be reset by the parameter setting means 22 until the S-curve becomes adequate in shape.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an electron beam diagnostic apparatus, and more specifically, it optimizes a secondary electron energy distribution curve suitable for use in the field of design verification and failure analysis of semiconductor integrated circuits, for example. The present invention relates to an electron beam diagnostic device. In recent years, semiconductor integrated circuits, especially LSI (Large Scale Inte
A large number of electron beam diagnostic devices using electron beams are used for design verification and failure analysis of large scale integrated circuits such as grated circuits.

An electron beam diagnostic apparatus irradiates a semiconductor integrated circuit to be diagnosed with an electron beam to detect secondary electrons reflected from the semiconductor integrated circuit, thereby performing design verification and failure analysis of the semiconductor integrated circuit. It is a thing. However, for example, unless the energy analysis curve of secondary electrons (hereinafter referred to as S curve) necessary for measuring the wiring voltage is obtained correctly, the measurement cannot be performed accurately.

Therefore, in order to perform accurate measurement, a means for optimizing the S curve in the electron beam diagnostic apparatus is required.

[0004]

2. Description of the Related Art A conventional electron beam diagnostic apparatus of this type is shown in FIG. FIG. 6 shows an electron optical system 1 which is a main part for voltage measurement in the electron beam diagnostic apparatus, 2 is an objective lens, 3 is a sample LSI, 4 is a drawing grid, 5 is a buffer grid, 6
Is an analysis grid, 7 is a shield, 8 is a collector, 9 is a reflector, 10 is a detector, and 11 is an amplifier. In addition,
EB indicates an electron beam, and e indicates a secondary electron.

In the above structure, the electron beam EB narrowed down by the objective lens 2 irradiates the fine wiring in the LSI 3 and the secondary electrons e generated from the wiring of the LSI 3 are extracted by the extraction grid 4 and the buffer grid 5. , The analysis grid 6 and the shield 7, respectively, and guided to the detector 10 by the action of the collector 8 and the reflector 9,
After being converted into an electric signal by the detector 10, the amplifier 1
It is amplified by 1 and detected as a secondary electron signal.

At this time, the analysis voltage Vr of the analysis grid 6
Is changed, the output of the electronic signal changes twice as shown in FIG. The energy analysis curve obtained at this time is the S curve. That is, when the wiring voltage increases by a volt, the S curve moves by the increased amount. Therefore, the wiring voltage is measured by detecting the movement amount of the S curve.

However, the shape of the S-curve depends on various factors such as the wiring width, the material of the wiring, the presence or absence of a protective film on the wiring, the surface condition of the wiring, the axial position of the LSI, and the like, as shown in FIG. As shown in, the theoretically predicted proper S
It often has a distorted S-curve shape that is different from the shape of the curve. In this case, of course, the accuracy of voltage measurement deteriorates, and in some cases, measurement becomes impossible. Therefore, making the S curve into a proper shape is a very important problem for the electron beam diagnostic apparatus.

Therefore, for example, while observing the S curve displayed on the display device or the like, the voltages (Ve, Vb, Vc, Vf) applied to the extraction grid 4, the buffer grid 5, the collector 8 and the reflector 9 can be measured. By adjusting the adjustment parameter, the shape of the S curve is optimized.

[0009]

However, in such a conventional electron beam diagnostic apparatus, in order to optimize the shape of the S-curve, while observing the S-curve displayed on the display device or the like, Voltage applied to the extraction grid 4, the buffer grid 5, the collector 8 and the reflector 9 (V
e, Vb, Vc, Vf) and other adjustment parameters are adjusted, which causes the following problems.

That is, since there are four change parameters in the adjustment for optimization, the combination amount is large,
Moreover, since the shape of the S-curve changes subtly depending on each parameter, even an expert requires a working time of about 1 hour for this adjustment, and there is a limit to the adjustment performed by a normal operator. Therefore, inadequate optimization of the S curve by adjustment has been a serious obstacle to improving the utilization efficiency and measurement accuracy of the electron beam diagnostic apparatus.

[Purpose] Therefore, an object of the present invention is to provide an electron beam diagnostic apparatus in which the S-curve is promptly optimized to improve the utilization efficiency and the measurement accuracy.

[0012]

In order to achieve the above object, an electron beam diagnostic apparatus according to the present invention has an electron beam irradiating means for irradiating an object to be diagnosed with an electron beam as shown in the principle diagram of FIG. An electron optical system 21 having secondary electron detection means for detecting secondary electrons emitted from the diagnosis target by the electron beam emitted from the beam irradiation means,
Parameter setting means 22 for setting a control parameter including each analysis voltage in the electron optical system 21 and signal processing for obtaining a secondary electron energy analysis curve from secondary electrons obtained from the electron optical system 21 The shape of the secondary electron energy analysis curve obtained by the secondary electron signal processing means 23 is predetermined based on the secondary electron signal processing means 23 and the secondary electron signal amount at each analysis voltage set by the parameter setting means 22. Shape determining means 24 for determining whether or not the shape is an appropriate shape, and when the shape determining means 24 determines that the shape of the energy analysis curve of the secondary electrons is not appropriate, the shape is determined as an appropriate shape. Until then, the control parameters of the electron optical system 21 are reset by the parameter setting means 22.

When the secondary electron signal amount for each analysis voltage is input, the shape determining means 24 outputs a high appropriate value when the shape of the analysis curve is appropriate, while the shape of the analysis curve is determined. It is preferable to have a neural network that outputs a low proper value when it is incorrect.

[0014]

According to the present invention, when the shape is judged to be not proper based on the result of the shape determination of the secondary electron energy analysis curve by the shape determining means, the electron optical system is set by the parameter setting means until the shape becomes proper. Each control parameter of is reset. That is, the S-curve is promptly optimized, and the utilization efficiency and measurement accuracy of the electron beam diagnostic apparatus are improved.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the drawings. FIG. 2 is a diagram showing an embodiment of the electron beam diagnostic apparatus according to the present invention, and is a block diagram showing a main configuration of the present embodiment. First, the configuration will be described.

In FIG. 2, the same numbers as the numbers given to the principle diagram shown in FIG. 1 indicate the same parts. The electron beam diagnostic apparatus of this embodiment is roughly classified into an electron optical system 21,
Parameter setting means 22, secondary electronic signal processing means 23,
It is composed of the shape determination means 24.

The electron optical system 21 has the same main parts as the conventional example shown in FIG. 6, and detects an electron beam irradiating means 25 for irradiating an LSI to be diagnosed with an electron beam and a secondary electron emitted from the LSI. The secondary electron detection means 26 is provided. The parameter setting means 22 comprises a parameter setting circuit 27 and an analysis voltage scanning circuit 28, and the analysis voltage scanning circuit 28 sets the analysis voltage value in the electron optical system 21 according to the parameter set by the parameter setting circuit 27.

The secondary electronic signal processing means 23 comprises a secondary electronic signal processing circuit 29, an AD conversion circuit 30 and a memory 31, and the analog signal processed by the secondary electronic signal processing circuit 29 is processed by the AD conversion circuit 30. It is converted into a digital signal and stored in the memory 31. Shape determination means 2
4 includes a neural network 32, a switching circuit 33, an adequacy determination system 34, a comparison circuit 35, an S-curve appropriateness degree calculation system 36, and a designated parameter calculation system 37. In particular, the neural network 32 is, as shown in FIG. It comprises an input layer 38 having a plurality of input terminals, an intermediate layer 39, and an output layer 40 having a plurality of output terminals for inputs.

The neural network 32 receives the input y ₁
The output data z _j (j = 1, 2, ..., M) is given by performing weighted addition of (i = 1, 2, ..., N) and threshold processing, and the output data z _j (j = 1, 2 ,. A corresponding output z _j is obtained. Next, the operation will be described.

First, with respect to the electron optical system 21, the parameter setting circuit 27 adjusts the voltages (Ve, Vb, Vc, Vf) and the like applied to the extraction grid 4, buffer grid 5, collector 8 and reflector 9 described above. Parameters are adjusted arbitrarily. Subsequently, the secondary electron signal when the analysis grid voltage Vr is continuously changed by the analysis voltage scanning circuit 28 is processed by the secondary electron signal processing means 23 by the secondary electron signal processing circuit 29 to be an analog signal. Is converted into a digital signal by the AD conversion circuit 30 and the memory 3
1 is stored and stored.

The S-curve information stored in the memory 31 is
It is input to the neural network 32 shown in FIG. 3, and the output from the suitability determination system 34 is obtained. The suitability is judged to be proper when the first output z ₁ of the neural network 32 is substantially equal to 1, that is, when (z ₁ −1) ≦ ε (ε is a small number), and the output is proper. If so, the adjustment is completed, and the work is shifted to the next operation such as measurement and diagnosis. On the other hand, if the conditions are not satisfied, the above process is repeated until the output becomes appropriate.

Here, in order for the neural network 32 to judge the suitability of the S curve, it is necessary to teach the neural network 32 prior to this suitability judgment. This will be described in detail with reference to FIGS. First, the voltages (Ve, V) applied to the extraction grid 4, the buffer grid 5, the collector 8 and the reflector 9 described above.
b, Vc, Vf) and other adjustment parameters are set (step 1), and the analysis voltage scanning circuit 28 scans the analysis grid voltage Vr, that is, the analysis voltage (step 2), and the secondary electron signal processing means 23. S curve data y _i (i = 1, 2, ..., N) is acquired (step 3).

Next, the deviation amount Σ (s _i −y _i ) ^{2 from} the ideal S curve data s _i is calculated, and the appropriate degree Z _j (j =
1, 2, ..., M) are calculated (step 4). This Z
_j is teaching data, and for example, if the appropriateness degree is k, Z _k = 1 and Z _j = 0 (j ≠ k). On the other hand, similarly, the S curve data y _i (i = 1, 2, ..., N) is input to the neural network 32 (step 5),
Output data z _j is obtained (step 6).

Then, the teaching data Z _j and the output data z _j
And the shift amount Σ (Z _j −z _j ) ² is calculated, and Σ (Z _j −
The weight of the neural network 32 and the threshold value are adjusted until z _j ) ² ≤ρ (ρ is a small amount) (steps 7 and 8). The teaching process ends when the above process is performed for all parameters (step 9).

That is, when teaching is performed, the changeover switch 33 is turned on, the S-curve information from the memory 31 is input to the S-curve appropriateness degree calculation system 36, and the comparison circuit 3 is inputted.
5 is compared with the output from the neural network 32, and as a result of the comparison, the weight and threshold value of the neural network 32 are reduced until the difference between the teaching data from the S-curve appropriateness degree calculation system 36 and the output data from the neural network 32 becomes sufficiently small. Are adjusted, and appropriate setting parameter values are calculated. That is, the output z _j of the neural network 32 at this time is z of | z ₁ −1 | ≦ ε (ε is a small number) when proper S curve data is input.
₁ will be output.

As described above, in this embodiment, the suitability of the S curve can be easily determined, and the proper S curve can be obtained quickly. Therefore, the measurement accuracy and use efficiency of the electron beam diagnostic apparatus can be improved, and the efficiency and reliability of LSI design verification and failure analysis can be improved. In the above embodiment, the shift amount Σ (s
_{Although the} appropriateness degree is calculated by _i− y _i ) ² , the present invention is not limited to this, and the appropriateness degree may be calculated by another method such as a correlation method or a pattern matching method.

In the above embodiment, the teaching data is calculated from the deviation amount from the ideal S curve, but this calculation may be done manually. That is, the S curve may be displayed, and the operator may observe the displayed S curve to determine the suitability.

[0028]

According to the present invention, each control parameter of the electron optical system is set until the proper shape is obtained when it is determined that the shape is not proper based on the determination result of the shape of the secondary electron energy analysis curve. Can be redone. Therefore, the S curve can be promptly optimized, and the utilization efficiency and measurement accuracy of the electron beam diagnostic apparatus can be improved.

[Brief description of drawings]

FIG. 1 is a principle diagram of an electron beam diagnostic apparatus of the present invention.

FIG. 2 is a block diagram showing a main configuration of the present embodiment.

FIG. 3 is a diagram showing a schematic configuration of a neural network.

FIG. 4 is a flowchart showing a teaching process for determining whether or not an S curve is appropriate according to the present embodiment.

FIG. 5 is a flowchart showing a teaching process for determining whether or not an S curve is appropriate according to the present embodiment.

FIG. 6 is a schematic sectional view showing an electron optical system of a conventional electron beam diagnostic apparatus.

FIG. 7 is a diagram showing a relationship between a secondary electron signal and an analysis grid voltage.

FIG. 8 is a diagram for comparing an appropriate S curve and a distorted S curve.

[Explanation of symbols]

 1 Electron Optical System 2 Objective Lens 3 LSI 4 Extraction Grid 5 Buffer Grid 6 Analysis Grid 7 Shield 8 Collector 9 Reflector 10 Detector 11 Amplifier 21 Electron Optical System 22 Parameter Setting Means 23 Secondary Electron Signal Processing Means 24 Shape Determining Means 25 Electrons Beam irradiation means 26 Secondary electron detection means 27 Parameter setting circuit 28 Analytical voltage scanning circuit 29 Secondary electron signal processing circuit 30 AD conversion circuit 31 Memory 32 Neural network 33 Switching circuit 34 Appropriate judgment system 35 Comparison circuit 36 S curve adequacy calculation System 37 Designated Parameter Calculation System 38 Input Layer 39 Intermediate Layer 40 Output Layer EB Electron Beam e Secondary Electron

Claims (2)

[Claims] 1. An electron beam irradiating means for irradiating an electron beam to a diagnostic object, and a secondary electron detecting means for detecting a secondary electron emitted from the diagnostic object by the electron beam emitted from the electron beam irradiating means. An electron optical system, parameter setting means for setting a control parameter including each analysis voltage in the electron optical system, and signal processing for obtaining an energy analysis curve of secondary electrons from secondary electrons obtained from the electron optical system. The secondary electron signal processing means to perform, and the shape of the secondary electron energy analysis curve obtained by the secondary electron signal processing means based on the secondary electron signal amount at each analysis voltage set by the parameter setting means has a predetermined shape. A shape determining means for determining whether or not the shape is proper, and the energy determining curve of the secondary electron by the shape determining means Jo determination result, when it is determined not to be appropriate, the electron beam diagnostic device, characterized in that the said parameter setting means until the proper shape reset the control parameters of the electron optical system. 2. The shape determining means, when the secondary electron signal amount for each analysis voltage is input, outputs a high appropriate value when the shape of the analysis curve is appropriate, while the shape of the analysis curve is not correct. 3. A neural network which outputs a low proper value when it is proper.
The electron beam diagnostic apparatus described.