JPS5827662B2 - Insulating film quality evaluation method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method for evaluating the quality of an insulating film formed on a semiconductor substrate or a conductive substrate.

The quality of the insulating film on the semiconductor substrate has an extremely large effect on the manufacturing yield and reliability of semiconductor devices.

It is difficult to manufacture semiconductor devices with high manufacturing yield and high reliability using an insulating film of poor quality that contains mobile ions, pinhole defects, and defects that locally weaken the electrical withstand voltage. I can't hope.

Therefore, in the semiconductor industry, there is a demand for film formation technology that is free of these defects, but in order to improve the film formation technology, it is necessary to evaluate film quality in detail.

As a typical conventional method for detecting mobile ions, pinhole defects, and voltage resistance defects and evaluating film quality, metal (M
), a so-called MIS capacitor method consisting of a sandwich structure of an insulating film (1) semiconductor (S).

Although this method is excellent in evaluating film quality regarding mobile ions, it has the following drawbacks.

In other words, when evaluating an insulating film with a relatively wide area, or when performing a dense evaluation on an insulating film with a constant area, a large number of conductive electrode groups are formed on these insulating films, and the semiconductor substrate is A voltage is applied between the conductive electrode and each conductive electrode, and the capacitance, withstand voltage, and leakage current between them are measured across all the conductive electrodes, and the results are adjusted to correspond to the geometrical arrangement of the conductive electrode group. These measurements must then be mapped.

Although measurements using individual conductive electrodes can provide detailed electrical measurement results, in order to observe defect distribution in detail, it is difficult to measure deviations by measuring individual conductive electrodes formed on an insulating film. This has the disadvantage that rapid measurement across all conductive electrodes requires expensive automated measuring and mapping equipment.

Furthermore, since a probe is required to make an electrical connection to each conductive electrode, there is a drawback that the dimensions of the conductive electrode cannot be made very small.

In this electrical connection using a probe, the tip of the probe may not only damage the conductive electrode, but also the insulating film under the conductive electrode.

Furthermore, the results of gradual measurement of the defect distribution in the insulating film are realized only by a mapping device, and there is also the drawback that they cannot be directly visualized by the observer's eyes on the semiconductor substrate.

This invention relates to the evaluation of pinhole defects and defective voltage resistance defects, excluding the evaluation of mobile ions, compared to conventional MIS.
The aim is to eliminate the drawbacks of the capacitor method by applying the dynamic scattering phenomenon of nematic liquid crystals, and to perform detailed film quality evaluation by varying the electric field strength applied to the insulating film.

The present invention will be described in detail below with reference to the drawings.

In the description, a case will be described in which a silicon (Si) substrate is used as the semiconductor substrate, a silicon dioxide (S102) film is used as the insulating film, and an aluminum (Al) electrode is used as the conductive electrode.

FIG. 1 is a cross-sectional view for explaining the principle by which pinhole defects and breakdown voltage defects can be detected by the dynamic scattering phenomenon of liquid crystal.

In the figure, a sample to be evaluated 100 is placed on a Si substrate 1.
An in2 film 2 is formed.

It is assumed that the S t 02 film 2 includes a defect 3.

This defect 3 is a defect that existed before the electric field was applied to the SiO2 film 2 or a defect that occurred due to the application of the electric field, and indicates a pinhole defect or a defect with poor breakdown voltage.

A large number of Ad electrodes 4 are formed over the entire surface of the Si02 film 2 using known AA vapor deposition techniques and photolithography techniques.

Next, the Ad electrode 4 is coated with a nematic liquid crystal film 5.

In this case, the nematic liquid crystal film 5 may be coated on the S i 02 film 2 in the portion where the Al electrode 4 is not present.

Next, the nematic liquid crystal film 5 is placed on the nematic liquid crystal film 5.
The transparent conductive film 6 is placed in contact with the transparent conductive film 6.

Since the transparent conductive film 6 is a thin film, it is normally adhered to the glass plate 1.

In addition, in order to prevent electrical short-circuiting between the transparent conductive film 6 and the Al electrode 4 and to ensure a nematic liquid crystal film 5 with a sufficient and uniform thickness to cause a dynamic scattering phenomenon, the Al electrode 4 is placed around the sample 100 to be evaluated. An insulating spacer 8 such as a polyethylene film having openings is interposed between the transparent conductive film 6 and the transparent conductive film 6 .

Next, with the Si substrate 1 at a positive potential and the transparent conductive film 6 at a negative potential, a DC voltage is applied between the Si substrate 1 and the transparent conductive film 6 using a DC power supply 9 whose DC voltage value can be changed.

The DC voltage is above the threshold value that causes a dynamic scattering phenomenon in the nematic liquid crystal film 5, and the defect 3 is connected to the Al electrode 4 and Si
If it is a pinhole defect that causes electrical leakage between the substrates 1, the nematic liquid crystal film 5 on the Al electrode 41 with the defect 3 will generate a dynamic scattering phenomenon and appear cloudy.

Further, if the DC voltage is higher than the threshold value and is a DC voltage that breaks the insulation at a location where the dielectric strength is low in the SiO2 film 2, the defect 3 is considered to be a voltage failure defect in the A7 electrode 4.
Since electrical leakage occurs between the Si substrate 1 and the Si substrate 1, clouding due to the dynamic scattering phenomenon is visible as described above.

If an MBBA film with a thickness of about 10 μm is used as the nematic liquid crystal film 5, the resistance perpendicular to the film surface is 108Ω/
cm2 or less, and if the film thickness of the 5102 film 2 is 1oooX, the resistance perpendicular to the film surface is 101)Ω/α2 or more.

Further, the threshold voltage of a nematic liquid crystal is generally around IOV.

Furthermore, the dielectric breakdown voltage of the S102 film 2 used in the semiconductor industry corresponds to an electric field strength near 107v/Crl1.

FIG. 2 shows the top surface of a sample to be evaluated 100 in which both an Al electrode 41 in which the nematic liquid crystal film 5 appears cloudy due to defects 3 and an Al electrode 42 in which the nematic liquid crystal film 5 does not appear cloudy due to the absence of defects 3 coexist. It is a diagram.

Although the transparent conductive film 6, DC power supply 9, etc. are not shown in the figure, the presence distribution of defects 3 can be seen over the entire surface of the sample 100 to be evaluated, and the A7 electrode 4 of the nematic liquid crystal film 5 appears cloudy.
1 and the distribution of the Al electrode 42 visible without cloudiness can be clearly seen, and the uniformity of the film quality can be evaluated.

Further, since the A7 electrode 41 is observed as cloudy, a probe is not required.

The present invention performs the operations described below using the apparatus configuration, defect detection, and display described in FIGS. 1 and 2.

FIG. 3 shows the DC voltage between the Si substrate 1 and the transparent conductive film 6.
An example of a change in the number of Al electrodes 41 that appear cloudy due to a dynamic scattering phenomenon due to the inclusion of defects 3 when changing is shown.

In the figure, vth is a threshold value that causes the nematic liquid crystal film 5 to enter a dynamic scattering state.

First, set the DC voltage to V, which is slightly larger than Vth.

At this time, the number of Al electrodes 41 that appeared cloudy was N.
.

Count. Next, 1M current voltage ■.

Set ■1=■o+ΔV, which is larger by ΔV, and count N1, the number of 41 Al electrodes that appear cloudy at this time.

Next, look at the DC voltage■. The number N10 of the 41 AA electrodes that appear cloudy is counted.

Next, the DC voltage is set to ■2=■o+2Δ■, and N2 of the number of 41 Al electrodes that appear cloudy at this time is counted.

Subsequently, the DC voltage was returned to (2), and the Ni on the Al electrode 41, which appeared cloudy, was counted.

Similarly below ■.

-Vo + nΔ■ Al electrode 41 appears cloudy
After making the numbers Nn and vn+1-vo ten(n+1)△■■.

Count the number N' of the 41 A and L electrodes that appear cloudy when returned to normal.

Through these operations, the Nn of the 41 Al electrodes appear cloudy due to pinhole defects, and the NnB of the 41 Al electrodes appear cloudy due to defective withstand voltage at DC voltage vn.
It consists of.

In addition, N'n is A that appears cloudy due to pinhole defects.
It consists of N, which is the number of 1 electrodes, and N'nB, which is the number of 41 AA electrodes, which appear cloudy due to a defect in voltage resistance that caused permanent breakdown due to the application of the voltage increase ΔV.

(2) If o is a value slightly larger than vth, No is approximately equal to N, which is the number of AA electrodes 41 that appear cloudy due to pinhole defects.

For example, in the case of the above-mentioned nematic liquid crystal film 5 and 5i02 film 2, even if o is a few tens of volts, which is slightly larger than vth, the SiO2 film 2 hardly causes dielectric breakdown, so N.

-N. becomes.

Therefore, [1] No is the number of Al electrodes 41 that appear cloudy due to pinhole defects, [2] (Nn-No) is the number of Al electrodes 4 that appear cloudy due to defective breakdown voltage at DC voltage ■.
1 number, (3,:] (]N, +1-Nn is the number of 41 Al electrodes that appear cloudy due to a newly generated breakdown voltage defect due to the DC voltage increase △■, [4] (N'-N ) is the number of times that occurred when the DC voltage increased by △■ n The number of 41 Al electrodes that appeared cloudy due to an irreversible and permanent breakdown voltage failure defect, C51 (Nn + 1 - N'n) is the number of times that occurred when the DC voltage was increased by △■ The number of Al electrodes 41 that appear cloudy due to defective breakdown voltage defects that can be recovered is also shown.

Here, the size of the Al electrode 4 is calculated from the defect density value of the normal 8102 film, and is set as a value smaller than the size where one or less defects 3 exist under one Al electrode 4. If you draw it, it will look cloudy kl' = ki, pole 41
The number now indicates 3 defects.

In normal cases, it is sufficient that the size of the Al electrode 4 is 1 to 2 squares.

It is preferable that the distance d between the four All electrodes be as small as possible, and this distance d can be made extremely small using current photolithography techniques.

Therefore, by determining the above-mentioned [1] to [5], it is possible to determine not only the number of defects 3 such as pinhole defects and defective withstand voltage defects in the SiO2 film 2 and the distribution of defects 3 within the plane of the 5IO2 film 2, but also the DC Voltage V to V.

By increasing the voltage stepwise by △■, it is possible to know the breakdown voltage distribution characteristics of defects with breakdown voltage defects, and to perform detailed film quality evaluation.

In addition, in the explanation so far, a film quality evaluation method in which an A7 electrode was provided on an 8102 film on a Si substrate was exemplified, but germanium (Ge), gallium arsenide (GaAs)
semiconductor substrates such as aluminum (A), conductive substrates such as gold (Au), and silicon nitride (s) other than 5i02 films.
13N4, ), alumina (A1203), etc., and electrodes other than AA electrodes such as polycrystalline silicon (Si) and molybdenum (Mo). It is.

As explained in detail above, the method for evaluating the film quality of an insulating film according to the present invention evaluates changes in the number of defects and their positional distribution by changing the electric field strength applied to the insulating film, without the need for a probe, and by It is possible to know directly visually over the whole area, and the evaluation equipment is not expensive and the work is simple.

This evaluation method can be expected to improve the technology for forming defect-free insulating films and, ultimately, to produce semiconductor devices with high production yields and high reliability.

[Brief explanation of drawings]

FIG. 1 is a cross-sectional view for explaining the principle by which defects can be detected by the insulating film quality evaluation method of the present invention, and FIG. FIG. 3, a top view of the evaluation sample, is a diagram showing the relationship between DC applied voltage and the number of defects for explaining in detail the method for evaluating the film quality of an insulating film according to the present invention. 1 is a Si substrate, 2 is a 5IO2 film, 3 is a defect, 4 is an Al electrode, 5 is a nematic liquid crystal film, 6 is a transparent conductive film, 9 is a DC power supply, 41 is cloudy due to the dynamic scattering phenomenon of the nematic liquid crystal film 5 42 is Al electrode visible without cloudiness.
It is an electrode. Note that the same reference numerals in the figures indicate the same or corresponding parts.

Claims (1)

[Claims] 1. A conductive electrode group is formed on an insulating film formed on a semiconductor substrate or a conductive substrate, a nematic liquid crystal film is coated on at least the conductive electrode group, and the liquid crystal film and A transparent conductive film is placed in contact with the transparent conductive film, and a DC voltage is applied between the semiconductor substrate or conductive substrate and the transparent conductive film, with the semiconductor substrate or conductive substrate at a positive potential and the transparent conductive film at a negative potential. When counting the number of conductive electrodes under the liquid crystal film that is causing the scattering phenomenon, add an increment to the DC voltage value, then subtract the increment,
A method for evaluating film quality of an insulating film, characterized in that the operation of counting the number of conductive electrodes accompanied by a dynamic scattering phenomenon that decreases before adding an increment and after adding an increment is performed while increasing a DC voltage value. 2. The film quality of the insulating film according to claim 1, characterized in that the DC voltage value before adding the increment is set to a DC voltage value slightly larger than the threshold value for the occurrence of the dynamic scattering phenomenon of the nematic liquid crystal. Evaluation method.