Plasma Etching Apparatus

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
9Forward
Citations 
0
Petitions 
0
Assignments
First Claim
1. A plasma etching apparatus comprising:
 a vacuum processing chamber for performing plasma processing on a workpiece;
a gas introduction means for introducing an etching gas to the vacuum processing chamber;
a high frequency feeding means for feeding highfrequency power for producing plasma in the vacuum processing chamber;
a spectroscope for measuring a spectrum of light of a specific wavelength emitted by the plasma; and
an arithmetic unit for determining an endpoint of etching of the workpiece,wherein the arithmetic unit is provided with;
a regression line computing unit for computing a regression line on the basis of timesequential data of luminous intensity of the specific wavelength sampled by the spectroscope;
a distance computing unit for computing a distance from the timesequential data to the regression line;
a computing unit for calculating a distance in a timebase direction by computing a slope of the regression line, and the distance from the timesequential data to the regression line, computed by the distance computing unit; and
an endpoint determiner for outputting an endpoint determination signal on the basis of the distance in the timebase direction computed by the computing unit.
0 Assignments
0 Petitions
Accused Products
Abstract
A plasma etching apparatus includes a vacuum processing chamber for performing plasma processing on a workpiece, a gas introducer, a high frequency power feeder, a spectroscope, and an arithmetic unit for determining an endpoint of etching of the workpiece. The arithmetic unit includes a regression line computing unit for computing a regression line on the basis of timesequential data of luminous intensity of a specific wavelength sampled by the spectroscope, a distance computing unit for computing a distance from the timesequential data to the regression line, a computing unit for calculating a distance in a timebase direction by computing a slope of the regression line, and the distance from the timesequential data to the regression line, computed by the distance computing unit, and an endpoint determiner for outputting an endpoint determination signal on the basis of the distance in the timebase direction computed by the computing unit.
18 Citations
View as Search Results
METHOD OF ENDPOINT DETECTION OF PLASMA ETCHING PROCESS USING MULTIVARIATE ANALYSIS  
Patent #
US 20140106477A1
Filed 10/17/2013

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Method of endpoint detection of plasma etching process using multivariate analysis  
Patent #
US 9,330,990 B2
Filed 10/17/2013

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

METHOD OF ENDPOINT DETECTION OF PLASMA ETCHING PROCESS USING MULTIVARIATE ANALYSIS  
Patent #
US 20160172258A1
Filed 02/25/2016

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Data analysis method for plasma processing apparatus, plasma processing method and plasma processing apparatus  
Patent #
US 9,443,704 B2
Filed 03/18/2016

Current Assignee
Hitachi HighTech Company Limited

Sponsoring Entity
Hitachi HighTechnologies Corporation

Method of endpoint detection of plasma etching process using multivariate analysis  
Patent #
US 10,002,804 B2
Filed 02/25/2016

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Compositional optical emission spectroscopy for detection of particle induced arcs in a fabrication process  
Patent #
US 10,436,717 B2
Filed 11/17/2017

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Surface modification control for etch metric enhancement  
Patent #
US 10,446,453 B2
Filed 03/13/2018

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Endpoint detection algorithm for atomic layer etching (ALE)  
Patent #
US 10,453,653 B2
Filed 03/08/2017

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Advanced optical sensor and method for detecting an optical event in a light emission signal in a plasma chamber  
Patent #
US 10,692,705 B2
Filed 11/15/2016

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Endpoint Detection Device For Realizing RealTime Control Of Plasma Reactor, Plasma Reactor With Endpoint Detection Device, And Endpoint Detection Method  
Patent #
US 20090029489A1
Filed 02/25/2008

Current Assignee
DMS CO. LTD

Sponsoring Entity
DMS CO. LTD

PLASMA PROCESSING APPARATUS  
Patent #
US 20080216956A1
Filed 09/06/2007

Current Assignee
Hitachi HighTechnologies Corporation

Sponsoring Entity
Hitachi HighTechnologies Corporation

Plasma processing apparatus  
Patent #
US 20070202613A1
Filed 03/10/2006

Current Assignee
Hitachi HighTech Company Limited

Sponsoring Entity
Hitachi HighTechnologies Corporation

Monitoring method of processing state and processing unit  
Patent #
US 20040197938A1
Filed 06/01/2004

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Disturbancefree, recipecontrolled plasma processing system and method  
Patent #
US 20030003607A1
Filed 09/06/2001

Current Assignee
Hitachi America Limited

Sponsoring Entity
Hitachi America Limited

Method and devices for detecting the end point of plasma process  
Patent #
US 5,728,253 A
Filed 09/30/1994

Current Assignee
Tokyo Electron Limited

Sponsoring Entity
Tokyo Electron Limited

Apparatus for monitoring the dry etching of a dielectric film to a given thickness in an integrated circuit  
Patent #
US 5,658,418 A
Filed 09/29/1995

Current Assignee
SOFIE INSTRUMENTS

Sponsoring Entity
SOFIE INSTRUMENTS

Etching endpoint determination method  
Patent #
US 8,083,960 B2
Filed 08/12/2008

Current Assignee
Hitachi HighTechnologies Corporation

Sponsoring Entity
Hitachi HighTechnologies Corporation

Single particle film etching mask and production method of single particle film etching mask, production method of micro structure with use of single particle film etching mask and micro structure produced by micro structure production method  
Patent #
US 8,202,582 B2
Filed 06/21/2007

Current Assignee
OJI Paper Company Limited

Sponsoring Entity
OJI Paper Company Limited

3 Claims
 1. A plasma etching apparatus comprising:
a vacuum processing chamber for performing plasma processing on a workpiece; a gas introduction means for introducing an etching gas to the vacuum processing chamber; a high frequency feeding means for feeding highfrequency power for producing plasma in the vacuum processing chamber; a spectroscope for measuring a spectrum of light of a specific wavelength emitted by the plasma; and an arithmetic unit for determining an endpoint of etching of the workpiece, wherein the arithmetic unit is provided with; a regression line computing unit for computing a regression line on the basis of timesequential data of luminous intensity of the specific wavelength sampled by the spectroscope; a distance computing unit for computing a distance from the timesequential data to the regression line; a computing unit for calculating a distance in a timebase direction by computing a slope of the regression line, and the distance from the timesequential data to the regression line, computed by the distance computing unit; and an endpoint determiner for outputting an endpoint determination signal on the basis of the distance in the timebase direction computed by the computing unit.
 2. A plasma etching apparatus comprising:
a vacuum processing chamber for performing plasma processing on a workpiece; a gas introduction means for introducing an etching gas to the vacuum processing chamber; a high frequency feeding means for feeding highfrequency power for producing plasma in the vacuum processing chamber; a spectroscope for measuring a spectrum of light of a specific wavelength emitted by the plasma; and an arithmetic unit for determining an endpoint of etching of the workpiece, wherein the arithmetic unit is provided with; a regression line computing unit for computing a regression line on the basis of timesequential data of luminous intensity of the specific wavelength sampled by the spectroscope; a distance computing unit for computing a distance from the timesequential data to the regression line; a computing unit for calculating a distance in a timebase direction by computing a slope of the regression line, and the distance from the timesequential data to the regression line, computed by the distance computing unit; and an endpoint determiner for outputting an endpoint determination signal on the basis of the product of the maximum value in the distance in the timebase direction computed by the computing unit by the slop of the regression line.
 3. A plasma etching apparatus comprising:
a vacuum processing chamber performing plasma processing on a workpiece; a gas introduction means for introducing an etching gas to the vacuum processing chamber; a high frequency feeding means for feeding highfrequency power for producing plasma in the vacuum processing chamber; a spectroscope for measuring a spectrum of light of a specific wavelength emitted by the plasma; and an arithmetic unit for determining an endpoint of etching of the workpiece, wherein the arithmetic unit is provided with; a regression line computing unit for computing a regression line on the basis of timesequential data of luminous intensity of the specific wavelength sampled by the spectroscope; and an endpoint determiner for outputting an endpoint determination signal on the basis of the product of the maximum value of an area, which is surrounded by the regression line and the timesequential data, by a difference between an area formed above the regression line and an area formed below the regression line.
1 Specification
This application is a continuation application of U.S. application Ser. No. 12/189,883, filed Aug. 12, 2008, the contents of which are incorporated herein by reference.
1. Field of the Invention
The present invention relates to an etching endpoint determination method and apparatus, or more particularly, to an etching endpoint determination method or apparatus making it possible to highly precisely determine an endpoint of etching processing to be performed by utilizing plasma discharge.
2. Description of the Related Art
In dry etching processing of a semiconductor wafer, the luminous intensity of light of a specific wavelength under plasma light emission changes along with the progress of etching of a specific film. The change in the luminous intensity of the light of the specific wavelength under plasma light emission is detected during the etching processing of the semiconductor wafer. Based on the result of the detection, an etching endpoint for the specific film can be detected.
For example, Japanese Patent Application LaidOpen Publication No. 200479727 reads that: a plasmaemitted light signal to be generated in an etching apparatus during etching of a material to be etched is acquired; and a linear model or a nonlinear model expressed by a polynomial expression is obtained for the acquired signal. Thereafter, an error of the signal from a calculated value represented by the obtained model is worked out. When the temporal change rate of the error exceeds a predetermined threshold value, the time point is detected as an endpoint of etching.
In recent years, an aperture ratio (a ratio of an area to be etched to the area of a semiconductor wafer) has diminished along with a tendency to a microscopic working dimension and highdensity integration in a semiconductor wafer. Consequently, a change in a luminous intensity produced by plasma near an etching endpoint is microscopic. Moreover, the precision in a working dimension to be attained through etching is more and more severely requested.
For example, in determination of an endpoint of etching, in case an endpoint determination time lags only several seconds, a film underlying a film that should be etched is excessively etched. A disorder in a working dimension takes place or side etching (excessive etching in a lateral direction) is carried out. Consequently, a workpiece shape deteriorates.
In order to avoid the dimensional disorder or the deterioration of a shape, a microscopic change in a luminous intensity occurring near an etching endpoint has to be accurately detected in an early stage, and processing has to proceed to the next step of removing the residue of a film to be etched which is called an overetching step.
When a change in a luminous intensity becomes microscopic, if the change in a luminous intensity is detected using, for example, a linear model expressed by a polynomial expression, the change in a luminous intensity is detected with an error occurring in a direction of a signal component. Therefore, when the aperture ratio is low, the detection is susceptible to a noise component contained in a signal. It is hard to detect an etching endpoint accurately in an early stage.
The present invention is intended to address the foregoing problem, and to provide an etching endpoint determination technology making it possible to determine an etching endpoint by detecting the etching endpoint accurately in an early stage.
In order to solve the foregoing problem, the present invention adopts a means described below.
An etching endpoint determination method or apparatus for determining an endpoint of etching processing which is implemented in a plasma etching apparatus that introduces a processing gas to a vacuum processing chamber via a gas introduction means, produces plasma by feeding highfrequency energy to the introduced processing gas, and uses the produced plasma to perform plasma processing on a workpiece stored in the processing chamber includes: a step of sampling light of a preset wavelength from light emitted by the plasma produced in the vacuum processing chamber, acquiring the luminous intensity of the sampled light of the specific wavelength as timesequential data, and computing a regression line on the basis of the acquired timesequential data; and a step of computing distances in a timebase direction between the regression line and the timesequential data which are obtained at the first step. An endpoint of etching processing is determined based on the distances in the timebase direction obtained at the second step.
The present invention has the foregoing constitution, and can therefore provide an etching endpoint determination technology making it possible to determine an etching endpoint by detecting the etching endpoint accurately in an early stage.
Embodiments of the present invention will be described in detail with reference to the following figures, wherein:
Referring to the appended drawings, the best embodiment will be described below.
A spectroscope is used to measure the spectrum of light of a specific wavelength emitted by plasma generated in the vacuum chamber during etching processing, and the signal component of the specific wavelength acquired through the spectral measurement is fetched into an arithmetic unit 105. The arithmetic unit 105 is formed with, for example, a personal computer (PC).
The fetched signal is used to deduce a regression line expression by a regression line computing unit 106. Preferably, the fetched signal has the gain thereof processed (has the signal intensity thereof amplified or decayed) for the purpose of enhancing a change in a timebase direction.
Thereafter, a distance computing unit 107 computes a distance between each of signals, which range from the latest signal to a signal having been obtained at a certain past time point, and a regression line obtained through computation (expressed by the regression line expression).
The number of signals to be employed in the computation can be set to an arbitrary value. Moreover, a line slope computing holding unit 108 that holds the value of the slope of the calculated regression line and a computing unit 109 are included. The computing unit 109 computes the value (distance) obtained by the distance computing unit 107 and the value (slope) held in the line slope computing holding unit so as to work out a component in a temporal direction of the distance. An endpoint determiner 110 monitors the result of the computation by the computing unit 109. When the result of the computation exceeds a preset criterial level, the endpoint determiner 110 outputs an endpoint criterial signal. The endpoint criterial signal is displayed on a display device 111.
The inputted timesequential data is subjected to gain processing by a gain computing unit. In the gain processing, a signal waveform is reduced, that is, multiplied by, for example, 0.001 so that the signal waveform can markedly represent a change in a component in a timebase direction. Incidentally, the number of signal points for the data has to be larger than the number of signal points to be used to obtain a regression line and the number of distance computation signals needed to obtain distances from the regression line (step S201).
After the number of obtained signals becomes equal to or larger than a required number of signals, when a predetermined determination beginning time has elapsed, computing processing of obtaining a regression line is performed. The computation of obtaining the regression line is performed according to, for example, a least squares method. Moreover, the signals to be used for the computation of obtaining the regression line range from the latest signal to a signal acquired at an arbitrary past time point. Data of the slope of the obtained regression line is held in the regression line slope computing holding unit 108 (steps S202, S203, S204, and S205).
Thereafter, the obtained regression line is used to calculate the distance from each of the points of the signals having been acquired in the past. The signals to be used to calculate the distance range from the latest signal to a signal acquired at an arbitrary past time point. The number of signal points to be employed, M, is larger than, for example, the number of signal points used to compute the regression line.
Moreover, the distance d from the regression line to a certain past point can be obtained according to an equation 2 on the assumption that the regression line is expressed by an equation 1. In the equation 2, x1 and y1 denote an xcoordinate and a ycoordinate representing the certain past point.
Assuming that the axis of abscissas is a time base and the axis of ordinates represents signal intensities, the distance given by the equation 2 contains both of components of a time and a signal intensity (step S206).
A signal acquired from plasma stemming from etching processing contains a noise component, and the noise component is expressed as a signalintensity component on the axis of ordinates. Consequently, when only components dx in the timebase direction are sampled from a signal obtained through distance computation, an adverse effect by the noise component can be minimized. Incidentally, each of the components dx in the timebase direction of distances from the regression line can be calculated according to an equation 3.
The calculating processing of a component in the timebase direction, dx, of a distance from the regression line is performed on all of arbitrarily designated distance calculation signal points M (step S207).
Thereafter, power computation is performed on a maximum value (distance maximum value) among components in the timebase direction of distances of all signal points from the regression line. As for the computing method, for example, an inputted signal is squared. If the computed value exceeds an arbitrarily designated threshold, the signal point is determined as an endpoint (steps S209 and S210). By performing the power computation, a signal component can be further amplified and noise contained in a signal can be reduced.
The waveform shown in
Incidentally, assuming that the distance between a regression line and each of signal points, that is, the length d of a perpendicular extended from each of the signal points to the regression line is designated as a signal to be used for determination, the signal representing the distance contains an intensity component. This makes it hard to lower the threshold to be used for the determination.
For detection of an endpoint, determination may be performed late using, for example, the trailing part of a waveform. For example, at a step of detecting an endpoint in the course of etching, even when a residue or the like remains, if the etching should be continuously performed, it is effective to perform endpoint determination late. In this case, there is a fear that erroneous determination may be invited due to a signal component observed at the trailing end of the waveform. In contrast, when only components in the timebase direction are used, after a waveform has fallen, a noise component is hardly observed. The erroneous determination can be prevented.
As described in
As described so far, according to the embodiments of the present invention, components in the timebase direction of distances between a regression line and respective signal points is used to determine an endpoint. Consequently, even when a signal produced from plasma stemming from etching processing contains a noise component, since the noise component is expressed as a signalintensity component on the axis of ordinates, an adverse effect of the noise component can be minimized. Consequently, even when a change in the luminous intensity of light emitted by plasma is microscopic, the microscopic change in the luminous intensity occurring near an etching endpoint can be accurately detected in an early stage. Consequently, the endpoint of etching can be quickly and accurately determined. Moreover, when a waveform of a signal varies abruptly, since the slope of the waveform gets large, a criterial waveform little affected by a noise component can be obtained according to the second embodiment. Moreover, when determination is requested to be performed rather late, a point of inflection can be readily grasped by utilizing the third embodiment. Consequently, determination can be achieved stably. Thus, an appropriate one of the determination methods can be selected. Moreover, the precision in determination largely varies depending of whether the number of data items to be used to plot a regression line or the number of data items to be used to calculate distances is optimized. The precision in determination will drastically improve by adopting appropriate values in every determination.