JPH0820224B2 - Method of measuring photoresist dissolution rate - Google Patents

Description

TECHNICAL FIELD The present invention relates to a method for measuring a dissolution rate of a photoresist in a photolithography process, which is included in a process for manufacturing a semiconductor integrated circuit device.

(Prior Art) Photoresists have been widely used in semiconductor manufacturing processes, but in recent years, in order to meet the demand for further improvement in the degree of integration, the resolution of development is improved, that is, the line width of the obtained circuit is further improved. There has been a demand for narrowing. However, when manufacturing a 16-megabit DRAM chip, the line width of the mask pattern used during development is about 0.
It becomes as narrow as 5μ. On the other hand, the wavelength of the light beam used for development is about 0.436 μm or 0.365 μm, and both are at almost the same level. In this way, if an attempt is made to obtain a development pattern having a very narrow width, the following new problems will occur.

That is, the light used for exposure is diffracted by the slit of the mask pattern. This is the first
It is shown schematically in the figure. Therefore, the distribution of the total exposure amount around the slit is shown in FIG. If the dissolution rate of the photoresist after exposure in a developing solution is proportional to the exposure dose as shown by the dotted line in FIG. 5, the cross section of the photoresist obtained by development is as shown in FIG. It becomes a shape. On the other hand, if the dissolution rate of the photoresist after exposure in the developing solution rises sharply as the exposure amount reaches a certain level, as shown by the solid line in FIG. It becomes like the figure. Naturally, the cross section shown in FIG. 4 is preferable, and conversely, a photoresist having the relationship between the exposure amount and the dissolution rate as shown by the solid line in FIG. 5 is developed. Is required.

For this purpose, it is necessary to determine what kind of dissolution properties the photoresist has from the relationship with the exposure amount and other conditions, and it is desired to develop an accurate measuring method of the dissolution rate. .

As disclosed by the present inventors in JP-A-2-63061, conventionally, in the resist developing step,
Light is irradiated from the top surface of the resist, the change in the intensity of the reflected light that is detected is monitored, and the time (CPT) to the final minimum point C of the reflected light intensity is detected, and the total development time is determined as a polynomial of this time. It is known how to do it. At each peak of the reflected light intensity change obtained at this time, the resist film thickness is λ
The resist dissolution rate can be determined by measuring the change in reflected light intensity and determining the peak-to-peak period.

However, even in that case, it was not possible to know the dissolution rate of the photoresist at an arbitrary time after the start of development. Furthermore, it was not possible to know the dissolution rate at any film thickness.

(Problems to be Solved by the Invention) An object of the present invention is to solve the above problems and provide a method for measuring the dissolution rate of a photoresist at an arbitrary time and an arbitrary film thickness.

(Means for Solving the Problems) In the resist developing step, the present invention irradiates light from the upper surface of the resist and measures the change in the dissolution rate with time by monitoring the change in reflected light intensity. This is a method for measuring the dissolution rate, in which the intensity change curve between the two poles of the reflected light that was monitored is standardized, the two poles are divided into multiple sections, and the time corresponding to each is obtained, and the reflection intensity-time data is obtained. Create a table and divide the two-pole points corresponding to the above-mentioned two-pole points into the same number of sections in the model calculation curve of the change in the reflected light intensity with the change in the film thickness, find the corresponding film thickness, and reflect The intensity-film thickness data table is created, the reflection intensity is deleted from the above two types of data tables, and the time-film thickness data table is created. By calculating the rate of decrease of the film thickness of or relates to a method for measuring the dissolution rate of the photoresist.

A method of measuring the dissolution rate of a photoresist according to the present invention will be described by taking a case of a single layer photoresist film on a bare Si wafer as an example.

(1) Measurement of dissolution rate of single layer photoresist film on bare Si wafer (interference light analysis of single layer film considering multiple reflection) Calculation of standing wave intensity generated in resist film at monitor wavelength λ FIG. 6 shows a schematic diagram in the case where a single-layered photo resist film is present on a bare Si wafer.

When a transparent film with a refractive index n ₁ and a film thickness d ₁ is present on a substrate (refractive index n ₀ ), and when a parallel light flux is vertically incident from a transparent medium with a refractive index n ₂ , the reflectance intensity R _2.0 is Indicated by.

ρ _A and ρ _B are medium-thin film (A-th surface), thin film-substrate (B-th surface).
It is indicated by the Fresnel coefficient in (plane).

Then Here, the spectral index of refraction of n ₀ (the index of refraction of silicon as a function of wavelength) can be obtained by the following approximation formula, where n (real number part) of n ^* = n−ik is n = n ₀ . Since the monitor wavelength is normally 0.6 to 0.8 μm, it is treated as an extinction coefficient k≈0.

n ₀ = A + BL + CL ² + Dλ ² + Eλ ⁴ (1,8) A = 3.41696 B = 0.138497 C = 0.013924 D = −2.09 × 10 ⁻⁵ E = 1.48 × 10 ⁻⁷ L = (λ ² −0.028) ⁻¹ ( However, the unit of λ is μm) n ₀ is calculated by the formula (1,8), and the formulas (1,3), (1,4),
The reflectance intensity (standing wave) at the monitor wavelength λ generated in the resist film can be obtained from the equations (1,5), (1,6) and (1,7).

Calculation of peak position of standing wave R can be considered as a function of γ ₁ . The obtained interference waveform is a periodic sine wave. Differentiating equation (1,1) as a function of γ ₁ , When R = 0, γ ₁ = 0, π.

The maximum reflection intensity R is when γ ₁ = 0, so
Therefore, γ ₁ = 2Nπ (N = 0,1,2,3 ...). Expression (1,2)
Than The reflection intensity R is minimized when γ ₁ = π, so
Therefore, γ ₁ = (2N + 1) π (N = 1,2,3 ...). From formula (1,2) Determination of Dissolution Rate from Measured Interference Waveform i) FIG. 7 shows an example of an apparatus for actually measuring the interference waveform. Reference numeral 1 is a chuck of a spin device, which is rotated by a rotating shaft 2. Reference numeral 13 is a motor drive circuit for rotating the motor of the spin device. The chuck 1 is provided with a suction mechanism, and the base 3 is sucked and mounted. Above the base 3, a developer supply mechanism 4 and a rinse solution supply mechanism 5 are provided.
And are provided. From the developer supply mechanism 4, the developer is sprayed in a band shape having a width of about 1 cm and a length of about 10 cm in the radial direction of the base 3, and the base 3 is rotated by a spin device so that the developer is evenly distributed on the base 3. Is puddle. Similarly, the rinse liquid supply mechanism 5 can spray the rinse liquid onto the substrate 3.

Reference numeral 6 is a light source having a long wavelength of an optical system for measuring the film thickness of the resist on the underlayer 3, and, for example, a Dangsten lamp or a mercury lamp is used. Reference numeral 7 is a filter for passing light of a wavelength which does not expose the resist among light from the light source 6, and for example, a filter which shields light of a short wavelength of 460 nm or less is used. Reference numeral 8 denotes an optical fiber bundle that guides the light passing through the filter 7 onto the base 3, and irradiates it so that the diameter of the light on the base 3 is, for example, about 10 mm.

The optical fiber bundle 8 also serves to guide the reflected light from the substrate 3. The reflected light guided by the optical fiber 8 passes through the filter 11 and is received by the phototransistor 10. The filter 11 is a narrow-band bandpass filter that transmits only light in a specific wavelength range, and is the most suitable depending on the type of the base 3.
Use one that can select a wavelength that increases the S / N ratio. The detection signal of the phototransistor 10 is converted into a digital signal through the A / D converter 15 and is taken into the computer 12 to perform a thorn-shaped analysis, and the rinse liquid supply mechanism 5 and the motor drive circuit 13 are analyzed based on the waveform analysis result. To control. A commercially available personal computer can be used as the computer 12.

The relationship between the reflected light intensity obtained by the measurement and the time is typically shown in a graph as shown in FIG. Further, FIG. 9 shows a developing state corresponding to points A to E in FIG.

The point C, that is, the last minimum point is called a control point (CP), and the point D, that is, the point where the base is exposed is called a breakthrough point (BT).

The obtained actual measurement data is weighted and averaged.

(Ii) The breakthrough point is detected using the no-fringe algorithm whose flow chart is shown in FIG.

(Iii) The number of data up to the breakthrough, Tri: Decimate to the initial resist film thickness (Å).

(Iv) Detect peak & valley inflection points in the above data.

(V) The number of data between peaks is adjusted by the following formula.

N _p1 : Number of data between peak and lever λ: Monitor wavelength n ₁ : Resist refractive index CL: Compress Level (waveform thinning rate) (vi) Number peak points from breakthrough. The resist film thickness d _1n at the peak number is _calculated by the equations (1,12) and (1,13).

As shown in FIG. 11, the first peak P ₁ has a minimum point,
The next peak P ₂ has a maximum point. Since this is repeated thereafter, the resist film thickness d _1n at the peak point (n is the peak point number) is Resist film thickness at the peak P _n peak points number n
d _1n is used in the following formula.

You can ask for it.

Now, let the peak points with actual data be P and Q, respectively (Fig. 12, when V _p > V _Q ). At this time, the voltage at P point is V _P and the voltage at Q point is V _Q And the times at that time are t _p and t _Q , respectively.

Considering this data as a function of cosγ ₁ , the intensity R (V _Q = R _Q =
1, V _P = R _P = −1). Then R _P ~ R _Q (−
1 to 1) are divided by ΔR, and these are defined as R ₁ , R ₂ , R ₃ , ... R _n (Fig. 13).

From Fig. 13, the times t ₁ , t ₂ , t _{3 at} R ₁ , R ₂ , R ₃ ... R _n ...
Calculate t _n . This is the (R, t) data table.

On the other hand, the resist film thickness at the point P is calculated by the formula (1,20), Q
The resist film thickness at the point is obtained by the equation (1,22). Assuming that the resist film thickness at the P point is d _1p and the resist film thickness at the Q point is d _1Q , from equation (1,2) Is obtained.

So in equation (1,7) Equations 1 and 7 are applied between R _{P and} R _Q to create a function of γ ₁ of the intensity R in the film thickness d _{1P to} d _1Q (Fig. 14).

 This model is shown by the following equation.

Then, R _{P to} R _Q (−1 to 1) in FIG. 14 is divided by ΔR, and R ₁ , R ₂ , R ₃ , ... R _n are respectively directed toward R _P → P _Q. At this time, ΔR has the same resolution as the division ΔR of R in FIG.

From Fig. 14, the resist film thickness d at R ₁ , R ₂ , R ₃ , ... R _{n
It is possible} to calculate ₁₁ , d ₁₂ , d _13, ... D _1n . This is the (R, d) data table.

The intensity R is erased from the (R, d) data table and the (R, d) data table, and the relationship (d, t) data table between the resist film thickness d and the time t is obtained.

The developing speed can be obtained as a function of the resist film thickness and its time by using the equation (1,27).

V _DEVn : Resist Dissolution Rate Next, a method for measuring the dissolution rate of a photoresist film in a multilayer film on bare Si will be described.

The case of a three-layer film (including a photoresist film) will be described below. However, the measurement method according to the present invention can be applied to the two-layer film by appropriately modifying it even when it includes more layers. It is possible.

Also, if the interpolation formula of the refractive index of the substrate shown in the formula (1, 8) is replaced with the interpolation formula of the wavelength vs. the refractive index of the other substrate such as GaAs, GaP, SiO ₂ , Cr, Al, the bare formula is used. Not limited to Si, Ga
It can also be applied to the measurement of photoresist dissolution rate on a multilayer film of As, GaP, SiO ₂ , Cr, Al, etc. on a substrate.

(2) Measurement of dissolution rate of photoresist film in three-layer film (including photoresist film) on bare Si.
Shown in the figure.

Calculation of the position of the standing wave at the monitor wavelength λ generated in the resist film. The refractive index n ₁ , the film thickness d _1, and the refractive index on the substrate (refractive index n ₀ ).
There is a base film of n ₂ and film thickness d ₂ on which the refractive index n ₃ and film thickness d ₃
Photoresist is present. Consider the reflection intensity R ₄ .0 when a parallel light beam is vertically incident on this from a transparent medium having a refractive index n ₄ .

The refractive index n ₀ of the Si wafer at the monitor wavelength λ is calculated by the equation (1,8). First, substrate (Si), underlayer 1
Determining the reflectance intensity R _{2 · 0} at.

ρ _A = R _A … (1,5) ρ _B = R _B … (1,6) Next, the reflectance intensity R on the substrate (Si), undercoat film 1, undercoat film 2
_Ask for _3.0 .

ρ _A2 = R _A2 … (2,4) Finally, the reflection intensity R _4.0 at the substrate (Si), the base film 1, the base film 2, and the resist film is This can be obtained by calculating.

Calculation of peak position of standing wave (i) Resist film thickness d ₃ Initial film thickness Tri is divided into Δd.

Tri: Initial film thickness (Å) Toward Tri _{⇒ 0} , d 3 _{Δ 1} , d for each Δd
_{Let 3Δ2} , ... D _3Δn , 0. This value is substituted into the equation (3,2) to obtain γ _3Δ1 , γ _3Δ2 , ... γ _3Δn . By substituting this into the equation (3,1), the reflectance intensities R _4Δ1 , R _4Δ2 , ... R _4Δn at d _3Δ1 , d _3Δ2 , ... D _3Δn can be obtained. Therefore, the film thickness d _3Δ1 , d _3Δ2 , ... D _3Δn
The reflectance intensities R _4Δ1 , R _4Δ2 , ... R _4Δn in the above can be obtained as a (d, R) data table.

(Ii) Rri, R _4Δ1 , R in the (d, R) data table
_4Δ2 , ... R _4Δn data is searched to detect the film thicknesses d _3maxn and d _3minn that give the maximum value and the minimum value of the periodic change of the intensity R. Resist film thickness 0 ⇒ toward _Tri d _3maxn , d
Number each 3 _minn , and use the following (d _3max , d
_3min ) Create a data table.

(Iii) _{Recalculation} of peak position _{Recalculate the} peak position around the resist film thicknesses d _3maxn and d _3minn that give each extreme value.

Recalculation of resist film thickness at which reflectance intensity reaches a maximum value d
_{The 3maxn} ± Δd is further divided by Δd / 10, and the resist film thickness having the maximum value of the intensity R is newly set as d _3maxn .

Recalculation of the resist film thickness at which the reflectance intensity has a minimum value d
_{The interval} between _3maxn ± Δd is further divided by Δd / 10, and the resist film thickness having the minimum value of the intensity R is newly set as d _3minn .

The following data table shows the relationship between d _3maxn and d _3minn after _{recalculation} and the peak position.

(D _3max, d _3min) data _table; d 3max1> d at _3Min1 (Figure 16).

(D _3max , d _3min ) data table; when d _3max1 > d _3min1 (Fig. 17).

Determination of dissolution rate from measured interference waveform (i) Calculation method of dissolution rate of single layer photoresist film on bare Si wafer from breakthrough to where peak point number is assigned Calculation method of dissolution rate from measured waveform (i) Is the same as ~ (vi).

(Ii) The relationship between the peak number and the resist film thickness at that time can be obtained using the following data table.

When the peak number 1 (P ₁ ) is extremely small (Fig. 18) When peak number 1 (P ₁ ) is maximum (Fig. 19) Now, let the peak points with actual data be P and Q, respectively. (When V _Q > V _p ) At this time, the voltage at point P is
The voltages at points V _p and Q are V _Q, and the times at that time are t _p and t _Q , respectively (Fig. 20).

Considering this waveform data as a function of cosγ ₃ , the intensity R (V _p
= R _p = −1, V _Q = R _Q = 1).

Then, divide between R _{p and} R _Q (−1 to 1) by ΔR, and R _p →
Let R ₁ , R ₂ , R ₃ , ... R _n towards R _Q. From Figure 21, R ₁ ,
R _2, R _3, ... it calculates the R _n in the time _{t 1, t 2, t 3} ... t n.
This is the (R, t) data table.

(R, t) data table On the other hand, the resist film thicknesses at the points P and Q are obtained from the data table of the above-mentioned peak number and resist film thickness. Assuming that the resist film thickness at the P point is d _3p and the resist film thickness at the Q point is d _3Q , from the formula (3, 2), Is obtained. So in equation (3,1) Then, applying equation (3,1) between R _{p and} R _Q , the film thickness d _{3p to} d
Create a function of γ ₃ of intensity R during _3Q . This model is shown by the following equation.

Next, divide between R _{p and} R _Q (-1 to 1) in FIG. 22 by ΔR,
R ₁ , R ₂ , R ₃ , ... R _n are set for R _p → R _Q , respectively. At this time, ΔR has the same resolution as the divided ΔR in FIG.

From FIG. 22, the resist film d ₃₁ , R _{3 for} R ₁ , R ₂ , R ₃ ... R _n ,
Calculate d ₃₂ , d ₃₃ , ... D _3n . This is the (R, d) data table.

(R, t) data table (R, d) Data table (R, d) The intensity R is erased and the relationship between the resist film thickness d and the time t (d,
t) Get the data table.

The developing speed can be obtained as a function of the resist film thickness and its time by using the equation (3,12).

V _DEVn : Resist Dissolution Rate Hereinafter, the present invention will be described in more detail with reference to examples.

Example A resist was applied to a base having a SiO ₂ film of 0.1800 μm on a Si substrate and used as an example.

OFPR-800 (made by Tokyo Ohka Kogyo Co., Ltd.) is used as a resist,
This was applied to a thickness of 1.0970 μm and baked at 100 ° C./60 seconds. The exposure conditions were NA; 0.38, g-line stepper (70 mJ / cm ² ). The developer is NMD-3 (Tokyo Ohka Kogyo Co., Ltd.)
It was done by dipping at 2 ° C. The film thickness was measured using Nanospec AFT.

Figure 23 shows the measured waveforms of the interference waves. Table 1 shows the normalized intensity of the interference wave actually measured.

A graph of the interference model waveform is shown in FIG. Table 2 shows the calculated resist film thickness at each peak.
Table 3 shows the data of the film thickness and the normalized strength between the peak 2 and the peak 1.

FIG. 25 shows the relationship between the residual film thickness of the resist and the development time obtained from the above values.

The relationship between the residual film thickness of the resist and the development time determined by the conventional method is shown at the same time, but it is understood that the method according to the present invention can provide more excellent results.

[Brief description of drawings]

FIG. 1 is a schematic diagram of how light is diffracted at the time of exposure by a mask having a minute width slit, and FIG. 2 is a diagram showing distribution of total exposure amount around the slit when diffraction occurs at the time of exposure, FIG. Is a sectional view when developing a photoresist whose dissolution rate is proportional to the exposure dose, FIG. 4 is a sectional view when developing using a photoresist having a preferable exposure dose and dissolution rate relationship, and FIG. FIG. 6 is a diagram showing a preferable relationship between exposure dose and dissolution rate, FIG. 6 is a schematic diagram when a single-layer photoresist film is present on a bare wafer, and FIG. 7 is an example of an apparatus for actually measuring an interference waveform in the present invention. , FIG. 8 is a diagram showing an example of the change over time of the reflected light intensity, FIG. 9 is a diagram showing the development state at each point shown in FIG. 8, FIG. 10 is a flowchart of the no-fringe algorithm, and FIG. The figure shows the time variation of the reflected light intensity. FIG. 12 and FIG. 20 are diagrams showing a part of the diagram showing the time change of the reflected light intensity, FIG.
Figures and 21 are figures for obtaining the (R, t) data table, Figures 14 and 22 are figures for obtaining the (R, d) data table, and Figure 15 is a three-layer film on a bare wafer. Schematic diagram when there is (including photoresist film), FIGS. 16 and 18 are diagrams when peak number 1 is the minimum point,
FIG. 17 and FIG. 19 are diagrams when peak number 1 is the maximum point, FIG. 23 is a diagram showing a temporal change of actually measured reflected light intensity in the example, and FIG. 24 is a calculated peak in the example. FIG. 25 is a diagram showing the relationship between the number and the film thickness, and FIG. 25 is a diagram showing the results in the examples. Explanation of reference numerals 1 ... Chuck of spin device, 2 ... Rotating shaft 3 ... Underground, 4 ... Developer supply mechanism 5 ... Rinse solution supply mechanism, 6 ... Light source 7, 11 ... Filter, 8 ... Optical fiber bundle 10 …… Photo transistor, 12 …… Computer 13 …… Motor drive circuit 15 …… A / D converter

Claims (1)

[Claims] 1. A method for measuring a dissolution rate of a photoresist, which comprises irradiating light from the upper surface of the resist in a resist developing step and monitoring a change in reflected light intensity to measure a change in dissolution rate with time. , Normalize the intensity change curve between the two poles of the reflected light that was monitored, divide the two poles into multiple sections, find the time corresponding to each, create a data table of reflection intensity-time, and change the film thickness. In the model calculation curve of the change of the reflected light intensity with the above, the two pole points corresponding to the above two pole points are divided into the same number of sections as described above, the corresponding film thickness is obtained, and the reflection intensity-film thickness data table is obtained. Created, delete the reflection intensity from the above two data tables to create a time-film thickness data table, and calculate the rate of decrease of film thickness per hour from this data table. By measuring the dissolution rate of the photoresist.