JP2013115354A - Simulation method, simulation program, semiconductor manufacturing apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a simulation method which allows for accurate prediction of the amount of damage of ultraviolet ray on a film.SOLUTION: The simulation method for predicting the amount of damage by ultraviolet ray is carried out in the manufacturing of a semiconductor device from a predicted emission spectrum in an ultraviolet region by calculating the light emission intensity of each wavelength in a visible region on the basis of a particle density calculated by performing simulation based on a differential equation of particle density, determining an electronic energy distribution function by comparing the light emission intensity of each wavelength in the visible region, calculated with reference to the information of luminescent species and emission wavelength in a manufacturing process of simulation object, with the emission spectrum of a visible region detected actually, and then predicting the emission spectrum of the ultraviolet region by using the electronic energy distribution function and a reaction cross sectional area related to the luminescent species.

Description

  The present technology relates to a simulation method and a simulation program for predicting an amount of damage caused by ultraviolet rays in manufacturing a semiconductor device. The present technology also relates to a semiconductor manufacturing apparatus that predicts the amount of damage caused by ultraviolet rays and controls process conditions.

  In the manufacture of semiconductor devices, processing using plasma generated by applying a high frequency bias is widely performed.

In plasma, depending on the structure of the processing chamber and the process conditions, collisions between particles and interaction reaction with the walls of the chamber occur, and ions, radicals, and light are generated.
Various technologies have been proposed, ranging from simulations to process methods and device structures, for the development of technology for predicting and controlling damage (crystal defects) to the film caused by the irradiation of ions. ing.

  In addition to the damage caused by ions, in recent years, there is a concern about damage to the film due to light irradiated from plasma, particularly ultraviolet rays, that is, ultraviolet (UV) or vacuum-ultraviolet (VUV). And basic research and monitoring technology about damage to the film by ultraviolet rays irradiated from plasma have been activated in Japan and overseas.

As semiconductor devices become thinner and the development of organic and compound semiconductor devices expands in recent years, prediction of emission intensity in the ultraviolet region, quantitative prediction of damage caused by ultraviolet rays, and development of control technology are desired. .
Conventionally, several methods have been proposed for the monitoring method of ultraviolet rays and the prediction of damage caused by ultraviolet rays in the manufacture of semiconductor devices.

For example, a method in which a sensor made of a laminated silicon film and oxide film is installed in a wafer or on the top plate of a plasma chamber, and a hole generated by ultraviolet light incident on the sensor is detected as an induced current under negative bias application. Has been proposed (see Patent Document 1 and Patent Document 2).
By measuring the variation of the induced current during the plasma process, the total (non-spectral) ultraviolet intensity can be monitored in real time. Further, the amount of damage is predicted using this ultraviolet intensity as an index.

In addition, a method for monitoring ultraviolet rays emitted from an ultraviolet lamp has been proposed in an ultraviolet cleaning apparatus for removing organic substances (see Patent Document 3).
Ultraviolet light having a wavelength of 254 nm is used for cleaning, but due to its strong emission intensity, there is a problem that the sensor deteriorates quickly when directly monitored.
Therefore, the proposed method does not directly monitor the wavelength of 254 nm, but monitors the UV intensity of 313 nm, which has a correlation with this wavelength and has a small intensity, as a reference, so that the intensity of the wavelength of 254 nm originally desired is known. Predict changes.

JP 2009-283838 A JP 2009-59879 A JP 62-122130 A

By the way, damage to the film due to ultraviolet irradiation is related to the wavelength of the irradiated ultraviolet ray and the band gap of the target film.
For example, in the case of silicon (with a band gap of 1.12 eV), when ultraviolet rays (wavelengths of 4 eV or more and shorter than 310 nm) are incident, absorption occurs in the entire ultraviolet wavelength range, causing crystal defects.
In the case of SiO ₂ (8.95 eV: 139 nm) and SiN (5.1 eV: 243 nm), the band gap is larger than that of silicon. Therefore, even at a wavelength shorter than 310 nm, the absorbance and the amount of damage vary greatly depending on the wavelength band.
Therefore, the amount of damage to silicon varies greatly depending on the structure and thickness of the upper layer of silicon and the irradiation wavelength.
In addition, since the emission spectrum distribution varies depending on the process conditions to be used, spectroscopy from ultraviolet (UV) to vacuum ultraviolet (VUV) is important in terms of practical use for manufacturing semiconductor devices.
And, it is considered that various ultraviolet wavelengths are involved in the damage during the plasma process in the actual manufacture of the semiconductor device.

However, in the methods described in Patent Document 1 and Patent Document 2, the total amount of ultraviolet rays is monitored, but the intensity for each wavelength cannot be monitored (spectrometric). As a result, the amount of damage cannot be accurately predicted or controlled.
Further, in the method described in Patent Document 3, since attention is focused on only two specific wavelengths, the damage amount cannot be accurately predicted and controlled.

  An object of the present technology is to provide a simulation method and a simulation capable of accurately predicting the amount of damage to an ultraviolet ray film. It is another object of the present invention to provide a semiconductor manufacturing apparatus capable of controlling the amount of damage by predicting the amount of damage to an ultraviolet film.

The simulation method of the present technology is a simulation method for predicting an amount of damage caused by ultraviolet rays in manufacturing a semiconductor device.
First, a simulation based on a differential equation of particle density is performed to calculate the particle density, and the emission intensity of each wavelength in the visible region is calculated based on the calculated particle density.
Then, referring to the information on the emission type and emission wavelength in the target manufacturing process, the calculated emission intensity of each wavelength in the visible region and the actually detected emission spectrum in the visible region are compared, and the electron energy is compared. Find the distribution function.
Furthermore, by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species, the emission spectrum in the ultraviolet region is predicted, and the predicted emission spectrum in the ultraviolet region is calculated. From this, the amount of damage caused by ultraviolet rays is predicted.

The simulation program of the present technology is a simulation program that predicts the amount of damage caused by ultraviolet rays in manufacturing a semiconductor device.
First, a simulation based on a differential equation of particle density is performed to calculate the particle density, and the emission intensity of each wavelength in the visible region is calculated based on the calculated particle density.
Then, referring to the information on the emission type and emission wavelength in the target manufacturing process, the calculated emission intensity of each wavelength in the visible region and the actually detected emission spectrum in the visible region are compared, and the electron energy is compared. Find the distribution function.
Furthermore, by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species, the emission spectrum in the ultraviolet region is predicted, and the predicted emission spectrum in the ultraviolet region is calculated. From this, the amount of damage caused by ultraviolet rays is predicted.
Then, these functions are implemented in the information processing apparatus and executed.

A semiconductor manufacturing apparatus according to an embodiment of the present technology controls a chamber in which a wafer is disposed, a sensor that detects a light emission spectrum in a visible region in the chamber, a calculation unit that calculates a process condition in the chamber, and a process condition in the chamber. A control unit is provided.
The calculation unit performs a simulation based on a differential equation of particle density, calculates the particle density, and calculates the emission intensity of each wavelength in the visible region based on the calculated particle density. Then, referring to the information of the emission type and emission wavelength in the target manufacturing process, the emission intensity of each wavelength in the visible region calculated is compared with the emission spectrum of the visible region actually detected by the sensor, Obtain the electron energy distribution function. Furthermore, the emission spectrum in the ultraviolet region is predicted using this electron energy distribution function and the reaction cross section for the luminescent species. Then, the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region, and the process conditions in the chamber are calculated so that the amount of damage is minimized.
The control unit controls the process condition in the chamber so as to be the process condition calculated by the calculation unit.

According to the simulation method of the present technology described above, the emission intensity of each wavelength in the visible region calculated based on the particle density is compared with the actually detected emission spectrum of the visible region, and the electron energy distribution function is calculated. Ask. Furthermore, the emission spectrum of the ultraviolet region is predicted by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species. Thereby, the emission spectrum in the ultraviolet region can be accurately predicted.
Furthermore, since the damage amount due to the ultraviolet rays is predicted from the predicted emission spectrum in the ultraviolet region, the emission intensity and the damage amount due to the ultraviolet rays can be accurately predicted according to the wavelength of the emitted ultraviolet light.

According to the above-described simulation program of the present technology, the emission intensity of each wavelength in the visible region calculated based on the particle density is compared with the actually detected emission spectrum of the visible region, and the electron energy distribution function is calculated. Ask. Furthermore, the emission spectrum of the ultraviolet region is predicted by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species. Further, the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region.
Thereby, the emission spectrum in the ultraviolet region can be accurately predicted, and the emission intensity and the amount of damage due to the ultraviolet rays can be accurately predicted in accordance with the wavelength of emitted ultraviolet light.

According to the configuration of the semiconductor manufacturing apparatus of the present technology described above, the calculation unit compares the emission intensity of each wavelength in the visible region calculated based on the particle density with the emission spectrum of the visible region actually detected. To obtain the electron energy distribution function. Furthermore, the emission spectrum of the ultraviolet region is predicted by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species. This makes it possible to accurately predict the emission spectrum in the ultraviolet region.
And a calculation part estimates the amount of damage by ultraviolet rays from the emission spectrum of the estimated ultraviolet region. This makes it possible to accurately predict the light emission intensity and the amount of damage caused by ultraviolet rays according to the wavelength of emitted ultraviolet rays.
Further, the calculation unit calculates the process conditions in the chamber so that the amount of damage is as small as possible, and the control unit controls the process conditions in the chamber so as to be the process conditions calculated by the calculation unit. As a result, the process conditions in the chamber are controlled so as to reduce the damage amount as much as possible, so that it is possible to reduce the damage amount due to ultraviolet rays when manufacturing the semiconductor device.

According to the above-described present technology, it is possible to accurately predict the light emission intensity and the amount of damage caused by ultraviolet rays according to the ultraviolet wavelength, so that it is possible to efficiently reduce deterioration (damage) of each film by ultraviolet rays. become.
Therefore, according to the present technology, it becomes possible to manufacture a semiconductor device having less quality (damage) due to ultraviolet rays and having good characteristics.

It is a basic flow figure of the simulation method concerning this art. It is a flowchart of simulation of quantitative prediction of the emission spectrum of an ultraviolet region. It is a schematic diagram of a part of a plasma chamber of an apparatus to be a target of plasma emission simulation. In FIG. 2, it is a schematic diagram explaining the process of calculating | requiring EEDF from the light emission intensity prediction of a plasma light emission simulation. It is an outline of a part of damage amount database. It is a conceptual diagram of the semiconductor manufacturing apparatus by this technique. 7 is a flowchart for correcting a manufacturing process condition in the semiconductor manufacturing apparatus of FIG. 6. It is a pattern structure figure used with the simulation method of a 1st embodiment. It is a conceptual diagram of the simulator of 2nd Embodiment. It is a schematic block diagram of the dry etching apparatus of 3rd Embodiment. FIG. 11 is a flowchart for correcting manufacturing process conditions in the dry etching apparatus of FIG. 10. FIG. It is a schematic block diagram (block diagram) of the dry etching apparatus of 4th Embodiment. 13 is a flowchart for correcting manufacturing process conditions in the dry etching apparatus of FIG. 12. 9 is a flowchart for detecting an end point of overetching in a dry etching apparatus according to a fifth embodiment.

The best mode for carrying out the present technology (hereinafter referred to as an embodiment) will be described below.
The description will be given in the following order.
1. 1. Overview of this technology First embodiment (simulation method)
3. Second embodiment (simulator)
4). Third embodiment (dry etching apparatus)
5. Fourth embodiment (dry etching apparatus)
6). Fifth embodiment (dry etching apparatus and etching processing method)

<1. Overview of this technology>
First, prior to description of specific embodiments, an outline of the present technology will be described.

In the present technology, the ultraviolet emission spectrum is predicted by simulation based on the visible emission spectrum data without directly monitoring the ultraviolet emission.
In addition, the amount of damage to the film is predicted by simulation using the result of prediction of the ultraviolet emission spectrum by simulation as input information.
Furthermore, for the purpose of reducing ultraviolet damage in real time, a semiconductor manufacturing apparatus is installed that is equipped with software based on the above-described prediction method.

  In the present technology, a simulation method, a simulation program, and a semiconductor manufacturing apparatus are configured as follows.

The simulation method of the present technology is a simulation method for predicting an amount of damage caused by ultraviolet rays in manufacturing a semiconductor device.
First, a simulation based on a differential equation of particle density is performed to calculate the particle density, and the emission intensity of each wavelength in the visible region is calculated based on the calculated particle density.
Then, referring to the information on the emission type and emission wavelength in the target manufacturing process, the calculated emission intensity of each wavelength in the visible region is compared with the actually detected emission spectrum in the visible region, and the electron energy is compared. Find the distribution function.
The emission spectrum of the ultraviolet region is predicted by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species. Further, the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region.

The simulation program of the present technology is a simulation program that predicts the amount of damage caused by ultraviolet rays in manufacturing a semiconductor device.
First, a simulation based on a differential equation of particle density is performed to calculate the particle density, and the emission intensity of each wavelength in the visible region is calculated based on the calculated particle density.
Then, referring to the information on the emission type and emission wavelength in the target manufacturing process, the calculated emission intensity of each wavelength in the visible region and the actually detected emission spectrum in the visible region are compared, and the electron energy is compared. Find the distribution function.
The emission spectrum of the ultraviolet region is predicted by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species. Further, the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region.
Then, these functions are executed by the computer.

A semiconductor manufacturing apparatus according to an embodiment of the present technology controls a chamber in which a wafer is disposed, a sensor that detects a light emission spectrum in a visible region in the chamber, a calculation unit that calculates a process condition in the chamber, and a process condition in the chamber. A control unit is provided.
The calculation unit performs a simulation based on a differential equation of particle density, calculates the particle density, and calculates the emission intensity of each wavelength in the visible region based on the calculated particle density. Then, referring to the information of the emission type and emission wavelength in the target manufacturing process, the emission intensity of each wavelength in the visible region calculated is compared with the emission spectrum of the visible region actually detected by the sensor, Obtain the electron energy distribution function. Furthermore, the emission spectrum in the ultraviolet region is predicted using this electron energy distribution function and the reaction cross section for the luminescent species. Then, the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region, and the process conditions in the chamber are calculated so that the amount of damage is minimized.
The control unit controls the process condition in the chamber so as to be the process condition calculated by the calculation unit.

(Simulation method)
A simulation method according to the present technology will be described in detail.
A basic flow diagram of a simulation method according to the present technology is shown in FIG.
The flow shown in FIG. 1 is executed by a simulator (information processing apparatus, calculation unit), which will be described in detail later.
As shown in FIG. 1, first, in step S <b> 1, the simulator reads emission spectrum data in a visible region (for example, 300 nm-800 nm: referred to as region A) as input data. The visible spectrum emission spectrum data is data of a visible spectrum emission spectrum that is actually emitted in a chamber of a semiconductor manufacturing apparatus (processing apparatus or the like) and detected by a sensor such as OES.
Next, in step S2, the simulator performs quantitative prediction of the emission spectrum in the ultraviolet region (for example, 10 nm-300 nm: referred to as region B) from the plasma emission simulation and the recipe information. Details of the plasma emission simulation will be described later. The recipe information is information indicating a light emission type and a light emission wavelength group in a target process (processing or film formation).
Next, in step S <b> 3, the simulator reads pattern structure data, that is, information on the pattern structure that is a target irradiated with ultraviolet rays. As pattern structure data, mask information (for example, GDS file information) and film thickness information are read.
Next, in step S4, the simulator predicts the damage amount. Here, the simulator compares the damage spectrum with the emission spectrum in the ultraviolet region predicted in step S2 and the read pattern structure data, and predicts the damage amount.

Furthermore, each step of the flow of FIG. 1 will be described in detail.
The quantitative prediction of the emission spectrum in the ultraviolet region in step S2 in FIG. 1 can be performed, for example, as described below.

A flow chart of a simulation for quantitative prediction of the emission spectrum in the ultraviolet region is shown in FIG.
The flow shown in FIG. 2 is also executed by a simulator (information processing apparatus, calculation unit), which will be described in detail later.

First, in the leftmost process of FIG. 2, the spectral data 101 in the visible region and the emission intensity prediction 102 of the plasma emission simulation detected by a sensor such as OES during processing or film formation using plasma are used as recipe information. Compare with the representative wavelength λ obtained from
As a result, an EEDF (Electron Energy Distribution Function) 103 is derived as will be described in detail later.
Prediction of light emission intensity by plasma light emission simulation is performed by a simulation method described in the literature (see Japanese Patent Application Laid-Open No. 2011-134927 and Jpn. J. Appl. Phys., 49, (2010), 08JD01). It can be carried out.

Next, the cross-sectional area (σ _λ ) of the reaction between particles relating to the luminescent species that emits the representative wavelength λ described above is collated from the cross-sectional area database, and at each wavelength λ, <σ _{λ '} Find v> 104.
Here, f (E) is EEDF, v is electron velocity, and E is electron energy.

Next, using the obtained <σ _{λ ′} v> 104, calculation 105 of the ultraviolet emission intensity of each wavelength λ is performed.
That is, by considering the related transition λ ′, from the following detailed equilibrium equation, it is the population of excitation levels related to light emission due to electronic transition, fine structure transition, hyperfine structure transition accompanying vibration / rotational transition, [X] _{Find λ} .

Here, _ne is the electron density, σ _λ is the reaction cross section, and A _λ is Einstein's A coefficient at each transition.

Further, using [x] _λ obtained by this equation, the emission intensity I _λ of light having a wavelength λ in the ultraviolet region is expressed as follows.

Here, h is a Planck coefficient, ν _λ is a corresponding frequency, β is an escape rate, S is a source function, and P is a background function.

Here, if the background component is ignored, P = 0. In general, since the optical thickness in the plasma process used in semiconductor manufacturing is thin, β≈1 holds approximately. Then, the above formula can be expressed as follows.

In this way, ultraviolet emission intensity at each wavelength λ is obtained, and a spectral distribution in the ultraviolet region (region B) is obtained.
The obtained ultraviolet region (region B) spectral distribution and the visible region (region A) spectrum data 101 detected by a sensor such as OES are combined to obtain an emission spectral distribution 106 in the ultraviolet region and the visible region. It is done.

Next, an outline of a specific method for predicting the emission intensity by the plasma emission simulation will be described below.
In order to explain the method and conditions of the plasma emission simulation, FIG. 3 shows a schematic view of a part of the plasma chamber of the apparatus (apparatus that performs processing and film formation using plasma) that is the object of the plasma emission simulation.

As shown in FIG. 3, the apparatus 1 includes a plasma chamber 2 composed of a chamber base 3, a space formed in a chamber wall 9 of the plasma chamber 2, a lower electrode 5B provided in the space, a chamber base. 3 and an upper electrode 5 </ b> A provided in 3.
In this plasma light emission simulation, a plurality of mesh spaces 2m are set by partitioning the space in the plasma chamber 2. The particle density n is calculated by simulation for every 2 m of mesh space.

A plurality of mesh walls 9 m are set in the chamber wall 9 by dividing the chamber wall 9. And the adhesion probability S of a particle | grain is set for every mesh wall 9m.
By setting the adhesion probability S for each mesh wall 9m, the adhesion probability S can be appropriately set according to the circumstances of each part of the chamber wall 9. For example, in the chamber wall 9, different adhesion probabilities S can be set for a portion (SiO ₂ ) constituted by the chamber base 3 and a portion (Si) constituted by the upper electrode 5A.

The shapes and sizes of the mesh space 2m and the mesh wall 9m may be set as appropriate.
The shape and size of the mesh space 2m (at the end face) and the shape and size of the mesh wall 9m may be the same as or different from each other. The size of the mesh space 2m and the mesh wall 9m is, for example, on the order of several mm.

  The simulation may be a two-dimensional simulation or a three-dimensional simulation. In other words, the mesh space 2m may be set two-dimensionally or may be set three-dimensionally, and the mesh wall 9m may be set one-dimensionally or two-dimensionally. It may be set.

Below, the identification number of the mesh space 2m is set to p, and (p) may be attached | subjected about the parameter (function of p) from which a value changes for every mesh space 2m.
In addition, the identification number of the mesh wall 9m is set to w, and the parameter (function of w) whose value changes for each mesh wall 9m may be given (w).

  In the simulation, a parameter indicating the distance between the mesh space 2 m and the chamber wall 9 is also considered. Specifically, the (shortest) distance L (p) between the mesh space 2m and the side wall 2a, the (shortest) distance R (p) between the mesh space 2m and the top plate 2b, and the distance between the mesh space 2m and the mesh wall 9m. d (w) is taken into account. Note that d (w) is also a function of p, but the addition of p is omitted. The distances L (p), R (p), and d (w) are distances from the center position of the mesh space 2m, for example. Further, d (w) is a distance from the center position of the mesh wall 9m, for example.

  Thus, the particle density n is appropriately calculated by considering the distance from the chamber wall 9. In particular, the particle density n is appropriately calculated by setting the adhesion probability S (w) and the distance d (w) for each mesh wall 9m.

  For example, the following equation is used as the differential equation of the particle density n.

As described above, this mathematical formula is used for each mesh space 2m. n (i, t) and n (j, t) are the particle densities at time t of particles of types i and j. For example, when the gas is hydrogen, the particle type i (j) is H, H ₂ , H ⁺ , H ₂ ⁺ , H ₃ ⁺ , or e. Hereinafter, instead of i and j, H, H _{2 and the} like may be indicated in parentheses.
k _m is the reaction rate, tau _r exhaust characteristic time, the tau _n is wall loss characteristic time. v _r mobility particle velocity, D is the diffusion constant, P is the pressure, mu is, k is the Boltzmann constant, T _e is the electron temperature, e is the elementary charge.

  In the equation on the first line of this equation, the first term represents the density change due to the generation and disappearance of particles due to collision reaction between particles. The second term shows the density change due to the exhaust of particles from the plasma chamber 2. The third term shows the density change due to the interaction between the particles and the chamber wall 9 and the disappearance of the particles at the wafer.

  As shown in the formulas in the first and second lines of the mathematical formula, the distance d (so that the disappearance of particles increases as the value of the distance d (w) between the mesh space 2m and the mesh wall 9m decreases. w) is incorporated.

Next, based on the calculated particle density n (i), the particle density n (H ^* ) of the particle H ^* contributing to light emission is calculated. H ^* is a concept different from the type of particles such as H and H ₂ (types i and j in the above formula), and the particle density n (H ^* ) cannot be obtained from the above formula.
Specifically, for each mesh space 2m, a particle density n (H ^* ) is calculated by solving a differential equation of the particle density n (H ^* ) by simulation, including the particle density n (i) obtained by a mathematical expression as a parameter ^. ) Is calculated.

Next, the emission intensity I (p) for each mesh space 2m is calculated based on the particle density n (H ^* ) for each mesh space 2m.
Then, as shown by hatching in FIG. 3, the emission intensity I (p) for each mesh space 2m is integrated in a predetermined viewing direction y1, and the emission intensity I when the plasma chamber 2 is viewed in the viewing direction y1 is calculated. To do.
Thus, by calculating the light emission intensity I in the line-of-sight direction y1, the light emission intensity I in the line-of-sight direction y1 detected by the OES 10 can be reproduced.
The line-of-sight direction y1 may be set to an appropriate direction such as a horizontal direction, a vertical direction, a direction inclined to these, a direction in which the mesh space 2m is arranged, or a direction inclined in the arrangement direction. Moreover, the cross-sectional area in the line-of-sight direction y1 of the space to be integrated may be set as appropriate. The position of the integrated space in the direction orthogonal to the line-of-sight direction y1 may be set as appropriate, and the light emission intensity I in the line-of-sight direction y1 may be calculated at a plurality of positions.

The simulation emission intensity I _{λ at} each wavelength is obtained by the plasma emission simulation of the above-described method. That is, the emission intensity prediction 102 of the plasma emission simulation of FIG. 2 is obtained.
FIG. 4 is a schematic diagram for explaining the process of obtaining the EEDF 103 from the emission intensity prediction 102 of the plasma emission simulation in FIG.
Simulation emission intensity I _lambda at each wavelength of the emission intensity prediction 102, as shown in FIG. 4, the device structure parameters alpha, the process condition parameters beta, expressed as a function of the electron temperature T _e.
Here, for example, from a representative wavelength lambda (recipe information measured values IO _lambda of the emission intensity I _lambda emission intensity obtained by the electron temperature T _e varied between 0～10EV, the luminescent species to be compared Judge). Then, an electron temperature _Te is searched for such that the emission intensity satisfies a desired error range (for example, ± 10%). The electron temperature _{T e} assuming a Maxwell distribution or the like using, derives the EEDF (electron energy distribution function) 103.
In this manner, the EEDF 103 shown in FIG. 2 is derived, so that the spectral distribution in the ultraviolet region can be obtained using the EEDF 103 through the process shown in FIG.

The damage amount prediction in step S4 in FIG. 1 uses a damage amount database. This damage amount database includes each of the pattern structure, dose amount, and process conditions (gas type, flow rate, pressure, power) for each specific wavelength with respect to the film type (Si, SiO ₂ , SiN, organic film, compound). It consists of a combination of data and damage amount (crystal defect) data.
An outline of a part of the damage amount database is shown in FIG. FIG. 5 shows combinations of pattern structure, dose amount, process condition data, and damage amount data for each of the specific wavelength 1, wavelength 2, and wavelength 3 for Si and the organic film, respectively.
In FIG. 5, a combination of pattern structure data, dose amount, process condition data and damage amount data is described for wavelength 1, but in an actual damage amount database, wavelength 1, wavelength 2, There are many sets of combinations for each of the wavelengths 3.
The damage amount database may be created in advance by actual measurement, or may be created by numerical simulation using MD (Molecular Dynamics) or the like.

By using the damage amount database configured as shown in FIG. 5, the damage amount can be predicted by associating the wavelength and intensity of ultraviolet rays with the process conditions, pattern structure, and aperture ratio.
This makes it possible to quantitatively optimize process conditions and device structure (or pattern layout) that suppress damage due to ultraviolet rays as much as possible based on the predicted damage amount.

Then, the simulator uses the emission spectrum in the ultraviolet region obtained by the calculation and the information on the actual pattern structure irradiated with ultraviolet rays read in step S3 in FIG. Check the quantity database and perform spline interpolation. Thereby, the amount of ultraviolet damage is predicted.
In this way, the amount of ultraviolet damage is predicted, and the process shown in FIG. 1 ends.

According to the simulation method of the present technology described above, the particle density is calculated by performing a simulation based on the differential equation of the particle density, and the emission intensity of each wavelength in the visible region is calculated based on the calculated particle density. Then, referring to the information on the emission type and emission wavelength (recipe information) in the target manufacturing process, the calculated emission intensity of each wavelength in the visible region is compared with the emission spectrum of the visible region actually detected by the sensor. Then, an electron energy distribution function is obtained. Furthermore, the emission spectrum of the ultraviolet region is predicted by calculating the emission intensity of each wavelength in the ultraviolet region using this electron energy distribution function and the reaction cross section for the luminescent species. Thereby, the emission spectrum in the ultraviolet region can be accurately predicted.
Furthermore, since the amount of damage caused by ultraviolet rays is predicted from the emission spectrum of the ultraviolet region obtained by prediction, the emission intensity and the amount of damage caused by ultraviolet rays can be accurately predicted according to the wavelength of ultraviolet rays emitted during the process. .

(Semiconductor manufacturing equipment)
A conceptual diagram of a semiconductor manufacturing apparatus according to the present technology is shown in FIG. FIG. 7 shows a flowchart of the process correction operation of the semiconductor manufacturing apparatus according to the present technology.
The semiconductor manufacturing apparatus 11 shown in FIG. 6 includes a plasma chamber 12, an OES that is a visible light emission monitor, a simulator 14 that implements a simulation program (software) for executing a simulation, and a control unit 15. Yes. The simulator 14 is an information processing apparatus such as a computer, and corresponds to a “calculation unit” according to the present technology.

An electrode 13 is provided in the lower part of the plasma chamber 12. A wafer 16 is placed on the electrode 13, and film formation or film processing is performed on the wafer 16.
Although not shown, an electrode (upper electrode) is also provided in the upper part of the plasma chamber 12, and an electric field is applied between the electrode (lower electrode) 13 and the upper electrode to generate plasma. Film formation or film processing is performed on the wafer 16.

The simulation program (software) to be installed in the simulator 14 is configured to be able to operate online and / or offline.
Then, as shown in the diagram on the right side of FIG. 6, the simulator 14 predicts the spectrum in the ultraviolet region from the monitoring data acquired by OES, predicts the amount of damage, and finds the optimized process condition. Execute.
Note that a simulation program (software) to be installed in the simulator 14 may be configured so as to search for an optimized process condition by predicting a shape variation surrounded by a broken line after predicting the damage amount.

The control unit 15 performs control so as to feed back the correction of the process conditions of the plasma chamber 12 obtained by the simulation executed in the simulator 14 to the plasma chamber 12.
Note that a configuration other than the OES shown in FIG. 6 can be adopted as the visible region light emission monitor.

In the semiconductor manufacturing apparatus 11 shown in FIG. 6, for example, the manufacturing process conditions are corrected according to the flowchart shown in FIG.
As shown in FIG. 7, first, in step S31, emission spectrum data (for example, wavelength 300 to 800 nm) in the visible region is sequentially acquired by OES provided in the plasma chamber 12 during plasma processing.
Next, in step S32, the simulator 14 predicts a light emission spectrum in the ultraviolet region from the above-described plasma light emission simulation and recipe information.
Next, in step S <b> 33, the simulator 14 reads pattern structure data (mask information, film thickness information, and the like), which is a target to be irradiated with ultraviolet rays.
Next, in step S34, the simulator 14 predicts the amount of ultraviolet damage by the above-described simulation method. Here, the simulator 14 collates the damage amount database from the emission spectrum in the ultraviolet region predicted in step S32 and the read pattern structure data, and predicts the damage amount.
Next, in step S35, the simulator 14 selects process conditions (pressure, flow rate, power) for reducing damage within a range close to the recipe conditions (for example, ± 50%) from the damage amount database.
Next, when the shape variation prediction enclosed by a broken line in FIGS. 6 and 7 is not performed, the process proceeds to step S37 as it is, and the simulator 14 transfers the correction parameter of the selected process condition to the control unit 15. .

On the other hand, in the case of performing shape variation prediction surrounded by a broken line in FIGS. 6 and 7, the process proceeds from step S35 to step S36, and the simulator 14 performs shape variation prediction. Specifically, the simulator 14 predicts the shape variation by performing a shape simulation using the selected process conditions and the wafer aperture ratio / Semi-Local aperture ratio / pattern structure. For the shape simulation, for example, a method described in JP2009-152269A can be employed.
As a result of the prediction, if the fluctuation is within an allowable range (for example, within ± 10%), the process proceeds to step S37, and the simulator 14 transfers the correction parameter of the selected process condition to the control unit 15.
On the other hand, when the fluctuation exceeds the allowable range, the process returns to step S35, and the simulator 14 selects a process condition that can reduce damage and is different from the process condition selected last time. And it progresses to step S36 and the simulator 14 performs shape variation prediction with respect to the selected process conditions.
By repeating this, the simulator 14 searches for the optimum process condition that can reduce the ultraviolet ray damage amount and the shape satisfying the specifications.
When the process condition is determined, the simulator 14 sends the determined process condition to the plasma chamber 12 side through the control unit 15 and corrects the process condition in real time.

As described above, the semiconductor manufacturing apparatus according to the present technology predicts the ultraviolet spectrum using the information of visible region luminescence that reflects the plasma state and the chamber wall state during processing, and thus more accurately than the conventional methods. It is possible to control the amount of ultraviolet damage. Thereby, it becomes possible to improve the electrical characteristics of the semiconductor device, that is, the transistor IV characteristics, the white spot / dark current of the imager, the conversion efficiency of the solar cell, the output characteristics of the laser, and the like.
In addition, when shape variation prediction is also performed, the process conditions can be further optimized.

  Note that the semiconductor manufacturing apparatus according to the present technology can be applied to any type of processing such as dry etching, CVD (Chemical Vapor Deposition), PVD (Physical Vapor Deposition), etc., as long as the processing uses plasma.

<2. First Embodiment (Simulation Method)>
Next, a specific embodiment will be described.
First, the simulation method of the first embodiment will be described.
In the simulation method of the present embodiment, the ultraviolet emission spectrum and the damage amount are simulated using the monitoring data by OES.

FIG. 8 shows a pattern structure diagram used in the simulation method of the present embodiment.
The pattern structure shown in FIG. 8 is each layer from the photodiode PD to the OCL (On Chip Lens) of the backside illuminated image sensor (pixel size 1.5 μm).
The photodiode PD has a thickness of 3000 nm, the interlayer film 21 thereon has a thickness of 1500 nm, the color filters 22 of the three colors R, G, and B thereon have a thickness of 700 nm, and the OCL 23 thereon has a thickness of 700 nm.
FIG. 8 shows that the vacuum ultraviolet VUV generated during the dry etching process of the OCL 23 during the production of this pattern structure attenuates and transmits through the OCL 23, the color filter 22, and the interlayer film 21, and damages the photodiode PD. It is a conceptual diagram. Here, the vacuum ultraviolet VUV has a wavelength of 10 to 200 nm.
In the present embodiment, the amount of damage that the vacuum ultraviolet VUV gives to the photodiode PD in this situation is predicted.
The structure information and process conditions of the processing apparatus used for processing the OCL 23 are shown below.

(Processing equipment)
CCP (Capacitive Coupled Plasma) dry etching equipment Frequency: Upper = 60 MHz, Lower = 2 MHz
Chamber diameter: 100cm
Upper and lower electrode gap: 40 mm
Visible light monitoring device: OES

(Process conditions)
Gas type: CF ₄
Gas flow rate: 150sccm
Pressure: 60mTorr
Upper applied power: 1000W
Etching time: 15 minutes

An OES (wavelength band: 200-800 nm, wavelength resolution: 0.5 nm) which is a visible light monitoring device is installed over a port (a window material is, for example, sapphire) of a chamber wall of the processing device.
By this OES, the emission spectrum emitted from the plasma during etching is sequentially acquired.
In the present embodiment, information on the above-described processing apparatus, structure information (the thickness of each layer of the OCL 23, the color filter 22, and the interlayer film 21), information on the above-described process conditions, and visible light emission spectrum information acquired by OES are input. Information.
For example, 321 nm (caused by the CF ₂ radical), 240 nm (caused by the CF radical), and 686 nm (caused by the F radical) can be used as the spectral wavelength necessary for deriving the electron energy distribution. However, the wavelength is not limited as long as the intensity is high.
The input information and the damage amount prediction method according to the present technology shown in FIG. 1 are used.
Thereby, in the present embodiment, it is possible to predict the simulation of the spectrum in the ultraviolet region having a wavelength of 200 nm or less and the crystal defect amount of silicon formed on the surface layer of the photodiode PD.

  According to the simulation method of the present embodiment described above, the particle density is calculated by performing a simulation based on the differential equation of the particle density, and the emission intensity of each wavelength in the visible region is calculated based on the calculated particle density. . Then, with reference to recipe information (information on the emission type and emission wavelength in the target manufacturing process), the calculated emission intensity of each wavelength in the visible region and the emission spectrum of the visible region actually obtained by OES are input. It is used as information. From the input information, the damage amount prediction method according to the present technology is used to predict the emission spectrum in the ultraviolet region, and the damage amount due to ultraviolet rays is predicted from the emission spectrum in the ultraviolet region obtained by the prediction. Thereby, according to the ultraviolet wavelength which light-emits during a process, the emitted light intensity and the damage amount by an ultraviolet-ray can be estimated correctly.

Although the present technology is applied to the CCP dry etching apparatus in the present embodiment, the processing apparatus to which the simulation method according to the present technique is applied is an ICP (Inductive Coupled Plasma) apparatus, an ECR (Electron Cyclotron Resonance) apparatus, or the like. Does not matter.
In addition to the dry etching of the OCL 23 having the pattern structure shown in FIG. 8, the types of processing steps to which the simulation method according to the present technology is applied include transistors, STI (Shallow Trench Isolation), sidewalls, contact holes, and the like. It doesn't matter.
Furthermore, the process condition to which the simulation method according to the present technology is applied is not particularly limited as long as the target film can be processed.

<3. Second Embodiment (Simulator)>
Next, the simulator according to the second embodiment will be described.
The simulator (information processing apparatus, calculation unit) according to the present embodiment has a computer program (software) that executes a simulation function for predicting the amount of damage according to the present technology, and executes the simulation method according to the present technology. To do.

A conceptual diagram of the simulator of the second embodiment is shown in FIG.
The simulator 30 shown in FIG. 9 includes an input unit 31, an ultraviolet spectrum calculation unit 32, a light emission wavelength database 33 for each process condition, a damage amount calculation unit 34, a damage amount database 35, and an output unit 36.

The simulator 30 exchanges signals and data with the processing apparatus 40 such as an etching apparatus.
In the simulator 30, the process proceeds according to the arrow in FIG.

Various data DATA is input to the input unit 31 as input information.
Examples of the various data DATA include pattern structure data (for example, GDS data), film thickness data, apparatus parameters, process parameters / visible emission spectrum data, and the like.

In the ultraviolet spectrum calculation unit 32, the data DATA input to the input unit 31 is used, and based on the data DATA, the corresponding wavelength data is obtained by referring to the wavelength database 33 of light emission for each process condition. The calculation for obtaining the emission spectrum is performed. The emission wavelength database 33 for each process condition has a wavelength group as data for each process condition.
The damage amount calculation unit 34 uses the emission spectrum in the ultraviolet region obtained by the ultraviolet spectrum calculation unit 32 and refers to the damage amount database 35 based on the emission spectrum in the ultraviolet region and the input data DATA to obtain the damage amount. Perform the operation.
The output unit 36 outputs the obtained damage amount from the simulator 30 to the outside.

The computer program (software) to be implemented in the simulator according to the present embodiment preferably has a configuration in which calculation is performed from data input, calculation history, and results are displayed by operating from a GUI (graphical user interface). To do. As the GUI, for example, a GUI constructed by tcl / tk, Motif, or the like can be used, but is not limited thereto.
The operation platform of the computer program (software) to be implemented in the simulator according to the present embodiment can be applied to various OSs regardless of the type of OS (Operation System).
Furthermore, the computer program (software) to be implemented in the simulator according to the present embodiment may be an offline operation or may be connected to the simulator 30 online.

<4. Third Embodiment (Dry Etching Apparatus)>
Next, a dry etching apparatus according to a third embodiment will be described.
The dry etching apparatus of the present embodiment is configured to automatically control process conditions so as to optimize the amount of damage caused by ultraviolet rays.
FIG. 10 shows a schematic configuration diagram (block diagram) of a dry etching apparatus according to the third embodiment.

  A dry etching apparatus 41 shown in FIG. 10 includes a plasma chamber 42, an OES, a simulator 44 on which a simulation program (software) for executing a simulation is mounted, and a control unit 45. The simulator 44 is an information processing apparatus such as a computer, and corresponds to a “calculation unit” according to the present technology.

In the plasma chamber 42, an electrode 43 is provided at the bottom. A wafer 46 is placed on the electrode 43 and the wafer 46 is etched.
Although not shown, an electrode (upper electrode) is also provided in the upper part of the plasma chamber 42, and an electric field is applied between the electrode (lower electrode) 43 and the upper electrode to generate plasma. Etching is performed on the wafer 46.
The OES is installed through a port window on the wall of the plasma chamber 42.
For example, the OES can have a waveform resolution of 0.5 nm and a wavelength band of 300 nm to 800 nm.

  In the dry etching apparatus 41 of the present embodiment, the etching conditions include, for example, an upper electrode frequency of 60 MHz, a lower electrode frequency of 13.56 MHz, a plasma chamber 42 diameter of 100 cm, and upper and lower electrode gaps. Is 40 mm.

The simulation program (software) to be installed in the simulator 44 is configured to be able to operate online and / or offline.
Then, as shown in the diagram on the right side of FIG. 10, the simulator 44 predicts the spectrum in the ultraviolet region from the monitoring data acquired by OES, predicts the amount of damage, and finds the optimized process conditions. Execute.

The control unit 45 performs control so that the correction of the process conditions of the plasma chamber 42 obtained by the simulation is fed back to the plasma chamber 42.
Note that a configuration other than the OES shown in FIG. 10 can be employed as the visible region light emission monitor.

In the dry etching apparatus 41 shown in FIG. 10, for example, the manufacturing process conditions are corrected according to the flowchart shown in FIG.
As shown in FIG. 11, first, in step S41, during dry etching, emission spectrum data (for example, wavelength 300-800 nm) in the visible region is obtained, for example, every second by the OES provided in the plasma chamber 42. get.
Next, in step S42, the simulator 44 predicts an emission spectrum in the ultraviolet region from the above-described plasma emission simulation and recipe information.
Next, in step S43, the simulator 44 reads pattern structure data (mask information, film thickness information, etc.), which is a target irradiated with ultraviolet rays.
Next, in step S44, the simulator 44 predicts the amount of ultraviolet damage by the above-described simulation method. Here, the simulator 44 collates with the damage amount database from the predicted emission spectrum in the ultraviolet region and the read pattern structure data, and predicts the damage amount.
Next, in step S45, the simulator 44 selects process conditions (pressure, flow rate, power) from the damage amount database so as to reduce damage within a range close to the recipe conditions (for example, ± 50%).
Next, in step S <b> 46, the simulator 44 delivers the correction parameter for the selected process condition to the control unit 45.
By sequentially repeating these steps until the end of etching, dry etching with reduced UV damage can be achieved.

In step S47, the simulator 44 determines whether the etching time has ended.
If the etching time has expired, the etching process is terminated.
If the etching time has not ended, the process returns to step S41, and emission spectrum data in the visible region is acquired by OES.
In this way, dry etching with reduced UV damage is possible.

In the present embodiment, any of a CCP device, an ICP device, and an ECR device can be used as the dry etching device.
The present embodiment can also be applied to a semiconductor processing apparatus using plasma such as a CVD (chemical vapor deposition) apparatus or a PVD (physical vapor deposition) apparatus.

According to the dry etching apparatus 41 of the present embodiment described above, the simulator 44 predicts the emission spectrum in the ultraviolet region and the damage amount, and selects the process conditions that optimize the damage amount. Then, the control unit 45 performs control so that the correction of the process condition of the plasma chamber 42 obtained by the simulation is fed back to the plasma chamber 42.
As a result, the process conditions of the process chamber 42 are controlled so as to optimize the damage amount (as much as possible), so that it is possible to reduce the damage amount due to ultraviolet rays during dry etching.
Therefore, it is possible to manufacture a semiconductor device having good characteristics with little damage caused by ultraviolet rays during dry etching.

<5. Fourth Embodiment (Dry Etching Apparatus)>
Next, a dry etching apparatus according to a fourth embodiment will be described.
The dry etching apparatus of the present embodiment is configured to automatically control process conditions so as to optimize the amount of damage caused by ultraviolet rays in consideration of shape variation.
FIG. 12 shows a schematic configuration diagram (block diagram) of a dry etching apparatus according to the fourth embodiment.

  A dry etching apparatus 51 shown in FIG. 12 includes a plasma chamber 52, an OES, a simulator 54 on which a simulation program (software) for executing a simulation is mounted, and a control unit 55. The simulator 54 is an information processing apparatus such as a computer, and corresponds to a “calculation unit” according to the present technology.

In the plasma chamber 52, an electrode 53 is provided in the lower part. A wafer 56 is placed on the electrode 53 and the wafer 56 is etched.
Although not shown, an electrode (upper electrode) is also provided in the upper part of the plasma chamber 52, and an electric field is applied between the electrode (lower electrode) 53 and the upper electrode to generate plasma. Etching is performed on the wafer 56.
The OES is installed through a port window on the wall of the plasma chamber 52.
For example, the OES can have a waveform resolution of 0.5 nm and a wavelength band of 300 nm to 800 nm.

  In the dry etching apparatus 51 of the present embodiment, the etching conditions include, for example, an upper electrode frequency of 60 MHz, a lower electrode frequency of 13.56 MHz, a plasma chamber 52 diameter of 100 cm, and upper and lower electrode gaps. Is 40 mm.

The simulation program (software) to be installed in the simulator 54 is configured to be able to operate online and / or offline.
Then, as shown in the diagram on the right side of FIG. 12, the simulator 54 performs an operation of predicting the spectrum in the ultraviolet region and predicting the damage amount from the monitoring data acquired by OES.
After predicting the amount of damage, the simulator 54 further performs an operation of predicting the shape variation and searching for an optimized process condition.

The control unit 55 performs control so that the correction of the process condition of the plasma chamber 52 obtained by the simulation is fed back to the plasma chamber 52.
Note that a configuration other than the OES shown in FIG. 12 can be employed as the visible region light emission monitor.

In the dry etching apparatus 51 shown in FIG. 12, for example, the manufacturing process conditions are corrected according to the flowchart shown in FIG.
As shown in FIG. 13, first, in step S51, emission spectrum data (for example, wavelength 300-800 nm) in the visible region is obtained, for example, every second by OES provided in the plasma chamber 52 during dry etching. get.
Next, in step S52, the simulator 54 predicts an emission spectrum in the ultraviolet region from the above-described plasma emission simulation and recipe information.
Next, in step S <b> 53, the simulator 54 reads pattern structure data (mask information, film thickness information, and the like), which is a target to be irradiated with ultraviolet rays.
Next, in step S54, the simulator 54 predicts the amount of ultraviolet damage by the above-described simulation method. Here, the simulator 54 collates the damage amount database from the predicted emission spectrum of the ultraviolet region and the read pattern structure data, and predicts the damage amount.
Next, in step S55, the simulator 54 selects process conditions (pressure, flow rate, power) from the damage amount database so as to reduce damage within a range close to the recipe conditions (for example, ± 50%).

Then, it progresses to step S56 and the simulator 54 performs shape fluctuation | variation prediction. Specifically, the simulator 54 predicts the shape variation by performing a shape simulation using the selected process conditions and the wafer aperture ratio / Semi-Local aperture ratio / pattern structure. Note that the Semi-Local aperture ratio is a resist mask aperture ratio in a region of several tens of mm around the pattern of interest (for example, 30 mm, which varies depending on the process conditions and is a value about several times the average free path of particles). It is.
For the shape simulation, for example, a method described in JP2009-152269A can be employed.
If the shape variation is within the allowable range as a result of the prediction, the process proceeds to step S57, and the simulator 54 passes the correction parameter of the selected process condition to the control unit 55.
On the other hand, when the shape variation exceeds the allowable range, the process returns to step S55, and the simulator 54 can reduce the damage and selects a process condition different from the process condition selected last time. And it progresses to step S56 and the simulator 54 performs shape fluctuation | variation prediction with respect to the selected process conditions.
When the process conditions are determined, the simulator 54 sends the determined process conditions to the plasma chamber 52 side through the controller 55, and corrects the process conditions in real time.
By sequentially repeating these steps until the end of etching, it becomes possible to control the shape satisfying the specifications and reduce the amount of ultraviolet damage.

Next, in step S58, the simulator 54 determines whether or not the etching time has ended.
If the etching time has expired, the etching process is terminated.
If the etching time has not ended, the process returns to step S51, and emission spectrum data in the visible region is acquired by OES.
In this way, it is possible to perform dry etching with suppressed shape variation and reduced ultraviolet damage.

In the present embodiment, any of a CCP device, an ICP device, and an ECR device can be used as the dry etching device.
The present embodiment can also be applied to a semiconductor processing apparatus using plasma such as a CVD (chemical vapor deposition) apparatus or a PVD (physical vapor deposition) apparatus.

According to the dry etching apparatus 51 of the present embodiment described above, the simulator 54 predicts the emission spectrum in the ultraviolet region and the damage amount, and selects the process conditions that optimize the damage amount. Then, the control unit 55 performs control so that the correction of the process conditions of the plasma chamber 52 obtained by the simulation is fed back to the plasma chamber 52.
As a result, the process conditions of the process chamber 52 are controlled so as to optimize the damage amount (as much as possible), so that it is possible to reduce the damage amount due to ultraviolet rays during dry etching.
Therefore, it is possible to manufacture a semiconductor device having good characteristics with little damage caused by ultraviolet rays during dry etching.

Further, according to the dry etching apparatus 51 of the present embodiment described above, the simulator 54 predicts the shape variation between the prediction of the emission spectrum in the ultraviolet region and the prediction of the damage amount.
Therefore, it is possible to stably manufacture a semiconductor device having good characteristics with a good yield with little variation in shape during dry etching.

<6. Fifth Embodiment (Dry Etching Apparatus and Etching Method)>
Next, a dry etching apparatus according to a fifth embodiment will be described. An etching method using this dry etching apparatus will be described.
The dry etching apparatus of the present embodiment is configured to control the detection of the end point of overetching by simulation of the damage amount.

  The configuration of each part of the dry etching apparatus can be the same as the apparatus shown in FIGS. 6, 10, and 12.

In the dry etching apparatus of this embodiment, for example, the end point of overetching is detected according to the flowchart shown in FIG.
As shown in FIG. 14, in step S61, the dry etching apparatus starts main etching.
As the main etching proceeds, in step S62, the dry etching apparatus detects the end point of the main etching.
The end point of the main etching can be detected, for example, by monitoring light emission in the visible region by the above-described OES.

After detecting the end point of the main etching, in step S63, the dry etching apparatus starts over-etching.
Next, in step S64, the dry etching apparatus predicts the amount of damage due to ultraviolet rays. The amount of damage is predicted according to the simulation flow shown in FIG.

Subsequently, in step S65, the dry etching apparatus determines whether or not a preset amount of damage has been reached.
If the set amount of damage has been reached, the process proceeds to step S66, and the dry etching apparatus detects the end point of overetching.
If the set amount of damage has not been reached, the process returns to step S64, and the dry etching apparatus continues overetching. Then, the damage amount is predicted again.
In this way, the end point of overetching can be detected.
The flowchart for detecting the end point of over-etching shown in FIG. 14 can be executed by a computer program (software) installed in the simulators 14, 44, 54 of the apparatus shown in FIGS. .

Here, a specific example in the case where the etching end point detection according to the present embodiment is applied to etching in a manufacturing process of a semiconductor device will be described.
In this specific example, the SiO ₂ on the silicon substrate is 2 nm thick, the polysilicon thereon is 150 nm thick, and the resist mask formed thereon is 300 nm thick, and the polysilicon is etched by etching using the resist mask. Is patterned into a gate pattern.
The line width of the initial resist mask is 100 nm.
In this specific example, the apparatus and process conditions used for processing are shown below.

(Processing equipment)
CCP (Capacitive Coupled Plasma) dry etching system Frequency: Upper = 60 MHz, Lower = 13.56 MHz
Chamber diameter: 100cm
Upper and lower electrode gap: 60mm
Visible light monitoring device: OES

(Process conditions)
Gas type: HBr / O ₂
Gas flow rate: 500sccm / 5sccm
Pressure: 30mTorr
Upper applied power: 150W
Lower applied power: 60W
Etching time: Main etching 100 seconds + Over etching

Detection of the end point of the main etching of polysilicon is performed using OES (wavelength band: 300-800 nm, wavelength resolution: 0.5 nm) as an index, and the emission intensity of the visible region as an index. Do by variation. In this specific example, for example, light emission with a wavelength of 526 nm of SiBr is used as an index.
For over-etching, the end point is detected when the amount of damage to the silicon substrate predicted by the ultraviolet damage simulation exceeds an allowable value (for example, the silicon crystal defect density is 10 ¹⁷ cm ³ ).
For example, 484 nm (cause of CO radical), 777 nm (cause of O radical), and 656 nm (cause of H radical) can be used as representative spectral wavelengths for deriving the electron energy distribution. However, the wavelength is not limited as long as the intensity is high.
As a result, over-etching of polysilicon can be performed while visualizing the amount of damage to the silicon substrate, and damage to the silicon substrate can be reduced.

  According to the etching processing method of the present embodiment described above, the end point of over-etching is detected using the amount of damage due to ultraviolet rays as an index, so that plasma processing with reduced damage can be performed while checking the amount of damage in real time. .

Note that the etching method of the fifth embodiment described above is not limited to gate etching by optimizing the processing conditions and visible region light emission as a reference at that time, so that etching of STI, sidewalls, contact holes, etc. is performed. Is also applicable.
Further, the etching method of the fifth embodiment can be applied to overetching endpoint detection of other films such as organic films and compound semiconductors (GaN, SiGe, etc.).
Further, the end point detection of the fifth embodiment may be provided in the etching apparatus of the third embodiment or the etching apparatus of the fourth embodiment.

As mentioned above, the above-mentioned embodiment shows the desirable mode of this art.
The technical scope of the present technology is not limited to the above-described embodiments, and application to wavelength regions other than ultraviolet rays is possible as well. Furthermore, the present invention is not limited to dry etching, but can be widely applied to semiconductor manufacturing processes and semiconductor manufacturing apparatuses that use plasma.

In addition, this indication can also take the following structures.
(1) A simulation method for predicting the amount of damage caused by ultraviolet rays in the manufacture of a semiconductor device, performing a simulation based on a differential equation of particle density, calculating a particle density, and visible based on the calculated particle density. Calculate the emission intensity of each wavelength in the region, and refer to the information on the emission type and emission wavelength in the target manufacturing process, and calculate the emission intensity of each wavelength in the visible region and the actually detected emission in the visible region. By comparing the spectra and obtaining an electron energy distribution function, and using the electron energy distribution function and the reaction cross section for the luminescent species, the emission intensity of each wavelength in the ultraviolet region is calculated. A simulation method that predicts the emission spectrum and predicts the amount of damage caused by ultraviolet rays from the predicted emission spectrum in the ultraviolet region.
(2) The simulation method according to (1), wherein the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region and data of the pattern structure to be produced.
(3) A damage amount database in which the film material, the wavelength of ultraviolet rays, the pattern structure, the dose amount of ultraviolet rays, the process conditions, and the damage amount are associated in advance and the damage amount due to the ultraviolet ray is predicted. The simulation method according to (2), wherein collation with a quantity database is performed.
(4) A simulation program for predicting the amount of damage caused by ultraviolet rays in the manufacture of a semiconductor device, performing a simulation based on a differential equation of particle density, calculating a particle density, and visible based on the calculated particle density. Calculate the emission intensity of each wavelength in the region, and refer to the information on the emission type and emission wavelength in the target manufacturing process, and calculate the emission intensity of each wavelength in the visible region and the actually detected emission in the visible region. By comparing the spectra and obtaining an electron energy distribution function, and using the electron energy distribution function and the reaction cross section for the luminescent species, the emission intensity of each wavelength in the ultraviolet region is calculated. Predicts the emission spectrum, predicts the amount of damage caused by ultraviolet rays from the predicted emission spectrum in the ultraviolet region, and processes information Simulation program to be executed to implement the location.
(5) Perform a simulation based on a differential equation of a particle density by a chamber in which the wafer is placed, a sensor for detecting a light emission spectrum in the visible region in the chamber, and calculate the particle density. Based on the calculated emission intensity of each wavelength in the visible region, referring to the information on the emission type and emission wavelength in the target manufacturing process, Comparing the detected emission spectrum of the visible region to obtain an electron energy distribution function, and using the electron energy distribution function and the reaction cross section for the luminescent species, the emission spectrum of the ultraviolet region was predicted and predicted. The amount of damage due to ultraviolet rays is predicted from the emission spectrum of the ultraviolet region, and the inside of the chamber is reduced so that the amount of damage is as small as possible. A calculation unit for calculating a process condition, so that the process conditions calculated by the calculation unit, a semiconductor manufacturing apparatus having a control unit for controlling the process conditions in the chamber.
(6) The calculation unit predicts the amount of damage due to ultraviolet rays, calculates the process conditions in the chamber, predicts the shape variation during the process by shape simulation, and when the shape variation exceeds the allowable range, The semiconductor manufacturing apparatus according to (5), wherein a process condition different from the previously calculated process condition is calculated.
(7) The calculation unit further detects the end point of over-etching after main etching when the predicted damage amount due to ultraviolet rays reaches a preset damage amount, (5) or (6) The semiconductor manufacturing apparatus described in 1.

11 Semiconductor manufacturing equipment, 12, 42, 52 Plasma chamber, 13, 43, 53 Electrode, 14, 44, 54 Simulator, 15, 45, 55 Control unit, 16, 46, 56 Wafer, 21 Interlayer film, 22 Color filter, 23 OCL, 30 simulator, 31 input unit, 32 ultraviolet spectrum calculation unit, 34 damage amount calculation unit, 35 damage amount database, 36 output unit, 40 processing device, 41, 51 dry etching device, PD photodiode, VUV vacuum ultraviolet ray

Claims (7)

A simulation method for predicting the amount of damage caused by ultraviolet rays in the manufacture of semiconductor devices,
Perform a simulation based on the particle density differential equation to calculate the particle density,
Based on the calculated particle density, calculate the emission intensity of each wavelength in the visible region,
Compare the calculated emission intensity of each wavelength in the visible region with the emission spectrum of the visible region actually detected by referring to the information of the emission species and emission wavelength in the target manufacturing process, and the electron energy distribution function Seeking
By calculating the emission intensity of each wavelength in the ultraviolet region using the electron energy distribution function and the reaction cross section for the luminescent species, the emission spectrum in the ultraviolet region is predicted,
A simulation method that predicts the amount of damage caused by ultraviolet rays from the predicted emission spectrum in the ultraviolet region.   The simulation method according to claim 1, wherein the damage amount due to ultraviolet rays is predicted from the predicted emission spectrum of the ultraviolet region and data of a pattern structure to be produced.   A damage amount database in which a film material, an ultraviolet wavelength, a pattern structure, an ultraviolet dose, a process condition, and a damage amount are associated in advance is created, and when the damage amount due to ultraviolet rays is predicted, the damage amount database and The simulation method according to claim 2, wherein collation is performed. A simulation program for predicting the amount of damage caused by ultraviolet rays in the manufacture of semiconductor devices,
Perform a simulation based on the differential equation of particle density, calculate the particle density, calculate the emission intensity of each wavelength in the visible region based on the calculated particle density, and the emission type and emission wavelength in the target manufacturing process The calculated emission intensity of each wavelength in the visible region and the actually detected emission spectrum of the visible region are compared to obtain an electron energy distribution function, and the electron energy distribution function, By calculating the emission intensity of each wavelength in the ultraviolet region using the reaction cross section for the luminescent species, the emission spectrum in the ultraviolet region is predicted, and the amount of damage due to ultraviolet rays is predicted from the predicted emission spectrum in the ultraviolet region. A simulation program that implements and executes functions on an information processing device. A chamber in which the wafer is placed;
A sensor for detecting an emission spectrum in a visible region in the chamber;
Perform a simulation based on the differential equation of particle density, calculate the particle density, calculate the emission intensity of each wavelength in the visible region based on the calculated particle density, and the emission type and emission wavelength in the target manufacturing process By comparing the calculated emission intensity of each wavelength in the visible region with the emission spectrum of the visible region actually detected by the sensor, an electron energy distribution function is obtained, and the electron energy distribution function is obtained. And the reaction cross-section for the luminescent species, the emission spectrum in the ultraviolet region is predicted, the amount of damage caused by ultraviolet rays is predicted from the predicted emission spectrum in the ultraviolet region, and the damage amount in the chamber is reduced as much as possible. A calculation unit for calculating the process conditions of
A semiconductor manufacturing apparatus comprising a control unit that controls process conditions in the chamber so that the process conditions calculated by the calculation unit are obtained.   The calculation unit predicts the amount of damage caused by ultraviolet rays, calculates the process conditions in the chamber, and then predicts the shape variation during the process by shape simulation. If the shape variation exceeds the allowable range, calculate it first. The semiconductor manufacturing apparatus according to claim 5, wherein a process condition different from the processed process condition is calculated.   The semiconductor manufacturing apparatus according to claim 5, wherein the calculation unit further detects an end point of overetching after main etching when the predicted damage amount due to ultraviolet rays reaches a preset damage amount.