JPH04107840A - Etching depth measuring instrument - Google Patents

Abstract

PURPOSE:To obtain a measuring instrument by which the etching depth can be measured accurately by a method wherein a luminous intensity of the reflected and diffracted light from a substrate which is to be etched is detected and the spaces between peaks of the detected signals are corrected for measurement of the etching depth. CONSTITUTION:An electrode 2, having a substrate to be etched 3 on its upper surface, is connected to a high-frequency source. On a part of a wall of an etching chamber 1 which faces the upper surface of the electrode 2, an observation window 4 is installed. A laser oscillator 5 is located outside and above the etching chamber 1 whereas a reflection mirror 6 is located to cast the laser light Q on the substrate 3. A luminous intensity of the light Q' which is reflected and diffracted from the substrate 3 is detected by a detector 7 and the detected signals are sent to a correction arithmetic processing section 8.

Description

[Detailed Description of the Invention] [Object of the Invention] (Industrial Application Field) The present invention provides an etching depth measuring device having a mechanism for measuring the etching depth of a substrate to be etched housed in an etching chamber in real time. Regarding equipment.

(Conventional technology) In general, etching in the semiconductor manufacturing process includes wet etching and dry etching, but for microfabrication of ultra-precision semiconductors, a fine photoresist pattern is faithfully transferred onto a silicon substrate to be etched. Dry etching is widely used. In a dry etching apparatus that performs ultrafine processing, it is required not only to simply transfer a photoresist pattern onto a substrate to be etched, but also to perform accurate etching while measuring the etching depth in real time.

As such an etching depth measuring device for measuring the etching depth of a substrate to be etched in real time, one having a structure disclosed in Japanese Patent Application Laid-Open No. 61-290308 is known.

This etching depth measuring device consists of a grounded vacuum container with a quartz window member, a high-frequency electrode placed inside the vacuum container for holding a sample, and a high-frequency power applied to this electrode. a high frequency power source for generating plasma between the vacuum container and a means provided outside the vacuum container for irradiating the surface of the sample undergoing reactive ion etching with coherent light having a wavelength shorter than 633 nm through the observation window; It has a structure comprising: a means for detecting the intensity of the reflected diffracted light from the sample; and a means for converting the detected signal into an electric signal and measuring the etching depth of the sample through processing calculations.

The etching depth measuring device having the above configuration is configured to:
The etching depth is determined by utilizing the fact that etching progresses by an amount corresponding to 1/2 of the wavelength λ of the coherent light.

However, in reality, the etching mask is etched along with the etching substrate, and the intensity of the reflected and diffracted light from the bottom of the trench decreases as the etching depth increases, so the time t from the maximum point to the maximum point increases. There may be cases where the etching progress time does not correspond to λ/2. If the above-mentioned time t is adopted as the time corresponding to the etching depth λ/2, the etching depth finally obtained will include a considerable error.

(Problem to be Solved by the Invention) In this way, the conventional etching depth measuring device uses only the maximum points of the intensity of the reflected diffracted light from the substrate to be etched, the time interval between the maximum points, and the wavelength of the laser beam. I was looking for etching depth. With such a measurement method, errors will be included in the etching depth obtained due to reduction in mask thickness and signal attenuation.

This invention was made to solve the above problems,
It is an object of the present invention to provide an etching depth measuring device which can obtain accurate etching depth by correcting the influence of mask thickness and signal attenuation in reflected diffraction light of a substrate to be etched.

[Structure of the Invention (Means for Solving Problems)] The present invention includes an etching chamber that accommodates a substrate to be etched therein, a light-transmissive window member provided on a wall of the etching chamber, and a measurement light irradiation means disposed outside the etching chamber for passing the measurement light through the window member and irradiating the substrate to be etched;
The method is characterized by comprising a detection means for detecting the intensity of reflected diffracted light from the substrate to be etched, and a correction arithmetic processing section for measuring the etching depth by correcting the peak interval in the detection signal from the detection means. .

With this configuration, the etching depth can be determined by correcting the influence of the thickness of the mask on the reflected diffracted light from the substrate to be etched.

(Example) Hereinafter, an example of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a schematic diagram of an etching depth measuring device showing an embodiment of the present invention.

In the figure, 1 is an etching chamber. In this etching chamber 1, one electrode (not shown) and the other electrode 2 are arranged below. The substrate to be etched 3 is placed on the upper surface of the other electrode 2, and is connected to a high frequency power source (not shown). On the wall (soil wall) of the etching chamber 1 corresponding to the other electrode 2, an observation window 4 made of a light-transmitting window member made of quartz is arranged. The etching chamber 1 is provided with a gas supply pipe (not shown) for supplying a reactive gas into the etching chamber 1 and an exhaust pipe (not shown) for exhausting the gas inside the etching chamber 1.

Further, a laser oscillator 5 is arranged outside and above the etching chamber 1. In the optical path of the laser beam Q emitted from the laser oscillator 5, there is a reflection mirror 6 for irradiating the laser beam Q through the observation window 4 almost perpendicularly to the substrate 3 to be etched on the other electrode 2. is located. The intensity of some of the reflected diffraction lights Q' out of the plurality of reflected diffraction lights Q' obtained from the substrate 3 to be etched is detected by a detector 7, and the detection signal is sent to a correction calculation processing section 8.

The correction calculation processing section 8 consists of a correction section 9 and an etching depth calculation section 10'. The detection signal from the detector 7 is first corrected by a correction method described later in a correction section 9, and then the etching depth is determined in an etching depth calculation section 101.

Next, the correction method performed in the correction calculation processing section 8 will be explained with reference to FIGS. 2 to 4.

FIG. 2 shows a cross section of the etched portion of the substrate 3 to be etched. The etching substrate 3 is formed from a silicon wafer 11 and an etching mask 13 having a certain pattern on its upper surface. The silicon wafer 11 is etched according to the pattern of the etching mask 13, and a trench 12, which is an etching hole, is formed substantially perpendicularly from the surface of the silicon wafer 11. Here, for convenience in explaining the correction method described later, T is the mask thickness and D is the etching depth (trench depth).

The first assumption is that the influence of the mask on etching,
Consider the case where signal attenuation is not considered.

In this case, the change in light intensity ■ of the detection signal over time is (
1) Make it appear like the formula.

1-C0+cos (4yr ・D(t)/λ)・
(1) 2〇 is a constant, π is pi, and λ is the wavelength of the measurement laser beam. In Fig. 3(a), the vertical axis is the light intensity of 1,
It is a graph in which the horizontal axis is time t. This signal of light intensity l is a cosine curve of wavelength λ/2 that takes constant maximum and minimum values.

In this cosine curve, any first maximum point A1
If we take the second maximum point A2, we get the maximum point A1.
, A2 corresponds to the period of this detection signal, and corresponds to the wavelength λ/2 of the detection signal.

This indicates that the etching became deeper by λ/2 at time tA.

On the other hand, the minimum point 81B2 corresponding to the maximum point A and A2 is
Similarly, the peak interval tB corresponds to the etching depth of λ/2, and the etching rate ERA (etching amount per unit time) obtained from the maximum points A and A2 and the etching rate ERB obtained from the minimum point are both ( 2) Equation, (
3) Make it rain with formula.

ER8-(λ/2)/tA...
...(2) Formula ERB-(λ/2)/ta
・・・・・・・・・The changes in the peak interval tA obtained at the maximum point of equation (3) and the peak interval tB obtained at the minimum point are plotted as shown in Figure 3(a), with the vertical axis representing the peak interval and the horizontal axis representing the peak interval tB. Similarly, a graph of time t is shown in FIG. 3(b). The change in the peak-to-peak distance itA at the maximum point is shown by a solid line A, and the change in the peak-to-peak interval tB at the minimum point is shown by a chain line B. From this graph, if we do not consider the influence of the mask and the influence of signal attenuation, the peak interval remains 1A regardless of whether we calculate it from the maximum point or the minimum point! 1. , and is constant over the entire etching time. That is, (
According to equations 2) and 3, in the case of the first assumption, the etching rate ER is (ERA=ERB), which takes the same value whether it is determined from the maximum point or the minimum point. In this case, the etching depth D is increased by D-ERX (etching time).

Next, as a second assumption, consider a case in which changes in mask thickness due to etching are taken into account, but signal attenuation is not taken into account.

In this case, the change in the light intensity ■ of the detection signal over time is (4
) expression.

1-C, 10c2cos2(4πn, T(t)/λ)-e
os(4yr (D (t) + T (t)) lλ
) (4) Equations C and C2 are constants, and n+ is the refractive index of the mask.

In equation (4), the mask thickness T is 1 of the etching depth D.
Assuming a function T (t) of time t decreasing at a rate of /100, the light intensity I will appear as a function of time t.

FIG. 4(a) shows a graph of the light intensity I in relation to time t, similar to FIG. 3(a). The vibrations at short wavelengths are due to changes in the intensity of the reflected diffracted light from the bottom surface of the trench 12, and the vibrations at long wavelengths are due to changes in the intensity of the diffracted light reflected from the surface of the mask 13. In this case, the peak interval tA obtained from the maximum point and the peak interval 1.A obtained from the minimum point. FIG. 4(b) shows the change over time in the same way as FIG. 3(b). In the case of the second assumption, unlike the case of the first assumption, the solid line A representing the change in the peak interval tA between the maximum points and the dashed line B representing the change in the peak interval tB between the minimum points do not match. In this case, since the etching rate differs depending on which peak interval is used, it is not possible to determine the exact etching depth.

Next, as a third assumption, a case will be considered in which the influence of the mask 13 and the reduction in signal caused by deeper etching are taken into consideration, similar to the signals actually obtained. In this case, the optical intensity I of the detection signal is expressed by equation (5).

1- C, +C2cos2(4πn IT (t)/λ
)−7(D, λ, d) ・cos(4yr (D
(t) + T (t))/λ)...(5) In formula (5), C, C2 are constants, and γ(D.

λ, d) is the attenuation coefficient. γ(D, λ, d) is a function determined by the etching depth D, laser wavelength λ, and trench diameter d, but here it is assumed that it is a function of the etching depth D given by equation (6). do.

r (D, λ, d)−7(D)−exp(−
aD) ... (α is a constant) Equation (6) In this way, the intensity can be expressed as a function of time, so as with the first and second assumptions, the graph of the light intensity ■ of the detected signal, the peak The interval graphs are shown in Figure 5(a).
, shown in (b). In FIG. 5(a), the vibrations at short wavelengths are due to changes in the intensity of the reflected diffracted light from the bottom surface of the trench 12, and the vibrations at long wavelengths are due to changes in the intensity of the reflected diffracted light from the surface of the mask 13. Furthermore, the short wavelength signals are attenuated over time. In this case, Fig. 5 (b
), the deviation between the solid line A showing the change in the peak interval tA obtained from the maximum point of the detection signal and the dashed line B showing the change in the peak interval tB obtained from the minimum point is quite large compared to the second assumption. It's getting bigger. Moreover, the above-mentioned deviation becomes larger as the etching time t elapses and the trench 12 becomes deeper.

The first assumption is an ideal situation, in which case the local maximum point A
Accurate etching depth D can be determined only by the peak interval tA obtained from the minimum point or the peak interval tB obtained from the minimum point, but in reality, as in the second and third assumptions, changes in the mask thickness The peak interval tA determined from the maximum point of the detection signal and the peak interval tB determined from the minimum point of the detection signal do not match due to factors such as signal attenuation. This means (
Etching rate ERA determined from equation 2) and equation (3),
Indicates that the ERRs do not have the same value.

In other words, even though the etching rate is the same at the same time, the values are different, so it is not possible to determine the exact etching depth.

Originally, the etching rate is determined by the specifications of the etching apparatus and the surface condition of the substrate to be etched, and is a substantially constant value unrelated to the etching depth D. Furthermore, the wavelength λ of the laser beam for measurement is also a constant value determined by the performance of the laser oscillation device.

In other words, in order to obtain a constant etching rate ER using equation (2) or (3), it is necessary to correct the peak interval 1A due to the maximum point and the peak interval tB due to the minimum point of the detection signal.

It can be seen from FIG. 4(b) and FIG. 5(b) that the solid line A and the chain line B have shapes that are substantially inverted from each other. Therefore, if we take the average value of the peak interval 1A obtained from the maximum point in Fig. 4(b) and Fig. 5(b) and the peak interval to obtained from the minimum point, the - point in Fig. 5(b) The peak interval tc takes a substantially constant value as shown by the dashed line C, and this dashed-dotted line C coincides with the solid line A and the dashed line B in the first assumption shown in FIG. 3(b). That is, by correcting the peak interval in this way, the same condition as the first assumption, that is, the effects of changes in mask thickness and signal attenuation are removed. Then, by substituting the peak interval tc obtained in this way into equation (2) or (3), a constant etching rate can be obtained as in the first assumption. That is, if the accurate etching rate is set as ERc, then this etching rate ERc can be determined by the following formula: ERc=(λ/2)/tc-(7). By using this etching rate E Rc, it is possible to obtain an accurate etching depth D regardless of the deviation between the maximum and minimum points of the detection signal.

The above correction method is only one example, and other correction methods are also possible.

First, there is a method of directly using equation (5) to directly determine the deviation of the maximum point or minimum point. That is, in equation (5), the constants 01 and C2 are determined from the measured values. In addition, the mask thickness D Ct) is obtained in the same way as the etching rate from the peak interval of long wavelength vibrations in Fig. 5, and the attenuation coefficient γ
(D, λ, d) can be calculated in advance since the trench diameter d and the wavelength λ of the laser light are known values.

In this case, since the etching depth D (t) is a function proportional to time, the deviation of the maximum point or minimum point can be directly determined by analyzing equation (5) by differential analysis.

There is also a method of correcting the deviation of the maximum point and minimum point by frequency-analyzing the actually obtained signal and removing components due to changes in mask thickness. That is, the actually obtained signal is subjected to Fourier analysis, and the frequency components of the signal due to changes in mask thickness are removed from the Fourier spectrum obtained by this using a band pass filter, and then this Fourier spectrum is subjected to inverse Fourier transform. Then, a signal as shown in FIG. 3(a) from which the influence of the mask has been removed is obtained. Since this signal has only an attenuation component, the etching rate can be easily determined by calculating the attenuation coefficient γ in advance.

Note that when etching deeply, trench 1
It is conceivable that the intensity of the light reflected and diffracted from the two bottom surfaces is attenuated and cannot be detected. This state is shown in FIG. FIG. 6 shows the detection signal with the horizontal axis representing the etching time t1 and the vertical axis representing the light intensity I. After time t1, the reflected and diffracted light from the bottom of the trench 12 is attenuated and the peak position cannot be confirmed.

In such a case, the time t is adjusted using the above correction method.

The etching gray (ER) is calculated using the peak interval tc determined up to now and the ratio m of the etching amounts of the trench 12 and the mask 13 calculated from the measurements up to that point, and the etching depth after t1 is determined from this. The etching rate ER is determined by equation (7).

ER-((λ/2)/(11/m))/lc...
(7) In other words, the etching depth D at time t and subsequent time t2 can be determined by equation (8) using the etching rate ER determined by equation (7).

D- (etching depth up to time t1) 1ERx (
t2t+ )...Equation (8) [Effect of the Invention] This invention corrects the etching depth by the interval between the maximum points and the interval between the minimum points of the intensity signal of the reflected diffracted light, and adjusts the etching depth based on the correction value. Measure the quality.

According to the above configuration, it is possible to eliminate the error caused to the intensity signal due to a change in the mask thickness, so that it is possible to accurately measure the etching depth.

[Brief explanation of the drawing]

FIG. 1 is a schematic configuration diagram of an embodiment of the present invention, FIG. 2 is a cross-sectional view of an etched portion of an etching substrate, and FIG.
a) and (b) are graphs showing the intensity and peak interval of the detected signal in relation to time under the first assumption;
Figures (a) and (b) are graphs for the second assumption,
FIGS. 5(a) and 5(b) are graphs for the third assumption, and FIG. 6 is a graph showing the strength of the detection signal as a function of time when etching is deep. DESCRIPTION OF SYMBOLS 1... Etching chamber, 3... Substrate to be etched, 4... Observation window, 5... Laser oscillator, 8...
Correction Calculation Processing Department Applicant's Representative Patent Attorney Takehiko Suzue Figure (b) A (a) Figure (b) (a) Figure 4 (b)

Claims (1)

[Claims] an etching chamber that accommodates a substrate to be etched therein; a light-transmissive window member provided on a wall of the etching chamber; and a light-transmitting window member disposed outside the etching chamber, the measurement light passing through the window member. A measurement light irradiation means for irradiating the substrate to be etched, a detection means for detecting the intensity of the reflected diffracted light from the substrate to be etched, and a peak interval in the detection signal from the detection means is corrected to determine the etching depth. An etching depth measuring device characterized by having a correction calculation processing section for measuring.