KR20160019749A - Method of monitoring process performed by fabrication apparatus of semiconductor device and monitoring system using the same - Google Patents

Abstract

The present invention relates to a method for monitoring repeated processes performed by an apparatus for fabricating a semiconductor device, and a monitoring system using the same. According to an embodiment of the present invention includes the steps of: defining a reference pattern by receiving a first intermediate signal from a normally performed reference process among the repeated processes by at least one sensor installed in the apparatus for manufacturing a semiconductor device; defining a target analysis pattern by receiving a second intermediate signal in real time by the at least one sensor from the same post process after the reference process; and comparing the reference pattern and the target analysis pattern to determine whether there is an error in the same post process.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a method of monitoring a semiconductor device and a monitoring system using the same,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a monitoring technology for a semiconductor device manufacturing apparatus, and more particularly, to a progress monitoring method and a monitoring system for repeated processes performed in the semiconductor device manufacturing apparatus.

Generally, a semiconductor device includes a FAB (fabrication) process for forming an electrical circuit on a substrate such as a silicon wafer, an electrical dicing (EDS) process for inspecting electrical characteristics of the semiconductor devices formed in the FAB process, Can be manufactured through a package assembly process for encapsulating and individualizing each with an epoxy resin. In the FAB process, various layers such as a silicon oxide layer, a polysilicon layer, an aluminum layer, or a copper layer may be formed on a substrate by chemical vapor deposition, physical vapor deposition, thermal oxidation, Such as ion implantation, or ion diffusion. In addition, the layers thus formed can be formed of reactive materials in a plasma state or patterns having electrical characteristics through a dry and / or wet etching process.

In this FAB process, it is possible to control the process such as the normal progress of the process in the semiconductor device manufacturing apparatus such as the reaction chamber, the leakage of the reactant in the reaction chamber, or the end of the process, In order to make operational policy decisions, various sensors are operated or the user visually observes through a window installed in the reaction chamber. For example, an apparatus for monitoring the state of the gas in the reaction chamber for management of the process may be used. The progress of the process and the state of the process equipment can be inferred by monitoring the change in gas type or pressure in the reaction chamber, and the process equipment can be controlled or maintained so that the optimum process can be performed according to the result.

In the conventional monitoring method, a limit setting value such as an average value, a lower limit value, or an upper limit value of a specific parameter such as temperature and pressure in the reaction chamber is determined and whether or not the process is normally progressed by monitoring whether the parameters are deviated from the actual process, . However, since the monitoring method monitors only the coarse process variation due to the average value setting, or monitors only the process parameters exceeding the upper limit value and below the lower limit value, It exacerbates or omits changes, resulting in inaccurate monitoring results.

SUMMARY OF THE INVENTION The present invention provides a method for monitoring progressive processes of a semiconductor device manufacturing apparatus, the method comprising the steps of: So as to minimize the errors of the user's process management and manufacturing facility operation policy through accurate monitoring.

It is another object of the present invention to provide a computer readable storage medium storing computer program codes for performing the process progress monitoring method having the above-described advantages.

It is another object of the present invention to provide a monitoring system capable of monitoring the progress of the process having the above-described advantages.

According to an aspect of the present invention, there is provided a method for monitoring progress of a process according to an embodiment of the present invention, the method comprising the steps of: Receiving an intermediate signal to define a reference pattern; Receiving a second intermediate signal from the same post process after the reference process in real time by the one or more sensors to define a pattern to be analyzed; And comparing the reference pattern and the analyzed pattern by pattern recognition to determine whether or not the same post-process is abnormal.

Wherein the step of determining whether or not the process abnormality is determined includes the steps of: when the pattern to be analyzed is recognized as the same pattern as the reference pattern by the pattern recognition, it is determined that the after- If it is recognized as a pattern different from the reference pattern, the post-same process may be judged to have been performed abnormally.

If it is determined that the post-same process has been performed abnormally, the process may be stopped or a combination of at least one of a warning sound, a warning light, and a warning screen for notifying the user of the process stop time and / Process interruption may be indicated.

If the analyzed pattern is recognized as a pattern different from the reference pattern, it may be determined whether the process is conformable to the same post-process. If it is determined that the post-identical process is appropriate at the step of determining whether or not the process conformity is determined, the learning process is performed to include the analyzed pattern in the reference pattern to update the reference pattern, The process may be terminated.

The learning step may be performed in accordance with a pattern recognition approach by updating one or more components of the reference pattern and combining the template matching method, the statistical approach, the structural approach, and the neural network approach.

In one embodiment, the at least one sensor is a plurality of sensors, and the first intermediate signal and the second intermediate signal are selected from among arithmetic operations, differentiation, integration, scaling, and weighting among individual signals received from the plurality of sensors, A signal computed by one or a combination of two or more, or individual signals not computed. The first intermediate signal and the second intermediate signal are further defined such that individual signals received from the at least one sensor are converted into digital signals and defined through statistical methods such as Fourier transform or sequential probability test. In some embodiments, the signal received from the one or more sensors may be a digital signal or an analog signal.

The at least one sensor may be a quartz crystal microbalance (QCM), a Langmuir probe, a magnetic probe, an optical sensor, a vibration sensor, a temperature sensor, a pressure sensor, a gas sensor, a spectroscopy, and an ion species mass spectrometer.

In one embodiment, the starting point and the ending point of the first intermediate signal may be a starting point and a finishing point of a normally progressing process of the semiconductor device manufacturing apparatus, or a specific section of the process progress. The reference pattern and the pattern to be analyzed are time, position, frequency, amplitude, frequency, inflection point, maximum value, minimum value, differential value (or slope), average value, variance value , The integral value, the real part, the imaginary part, the peak, the lowest point, and the maximum value, the minimum value and the average of the absolute values of the second derivative of the signal.

The reference pattern may include information on a marginal error range. The reference pattern may be all or part of the signal from the start point to the end point of the reference process. In addition, the reference pattern and the analyzed pattern may include a discrete form. The reference pattern and the analyzed pattern may include repetitive waveforms. In another embodiment, the reference pattern and the analyzed pattern may include a table shape or a matrix shape.

The step of determining whether or not an abnormality of the same post process is determined may be determined by a pattern recognition approach according to any one or two or more combinations of the template pattern matching method, the statistical approach, the structural approach, and the neural network approach . The step of determining whether there is an abnormality in the same post-process may include comparing the analyzed pattern and the reference pattern by a one-to-one comparison, a many-to-one comparison, and a many-to-many comparison. In addition, the step of determining whether or not the abnormality is present may be performed based on whether a real-time trend of the analyzed pattern follows the reference pattern. In one embodiment, the step of determining the presence or absence of the abnormality includes: setting a time interval of the plurality of reference patterns continuously or discontinuously; comparing the time variation of the plurality of reference patterns with reference to whether the real- .

According to another aspect of the present invention, there is provided a computer-readable storage medium for storing semiconductor devices, the semiconductor device comprising: Receiving a first signal from a normally performed reference process and defining a reference pattern; Receiving a second signal in real time from the same post process after the reference process by the at least one sensor to generate a analyzed pattern; And comparing the reference pattern and the analyzed pattern by pattern recognition to determine whether there is an abnormality in the same after-process.

If it is determined that the analyzed pattern of the same post-process is not the same as the reference pattern but the post-process is appropriate, the learning step of including the analyzed pattern in the reference pattern is performed to update the reference pattern , And if the process is judged as an inappropriate process, the process may be terminated.

According to another aspect of the present invention, there is provided a monitoring system comprising: at least one sensor unit mounted in a semiconductor device manufacturing apparatus; A first pattern generator for receiving a first signal from a reference process normally performed during repetitive processes of the apparatus for manufacturing a semiconductor device by the one or more sensors to define a reference pattern; A second pattern generator for receiving a second signal from the same post process after the reference process by the one or more sensors in real time to generate a pattern to be analyzed; And comparing the reference pattern and the analyzed pattern by pattern recognition to determine whether there is an abnormality in the same after-process.

A method for monitoring the progress of a manufacturing process of a semiconductor device according to an embodiment of the present invention includes the steps of receiving individual signals generated by one or more sensors mounted on the manufacturing facility to generate a reference pattern and an analyzed pattern, It is possible to provide a process progress monitoring method that minimizes a user's process management and manufacturing facility operation policy decision error by precise process progress monitoring that does not exaggerate or omit changes in temporal information or parameters.

According to another embodiment of the present invention, there is also provided a computer readable storage medium for storing computer program codes for performing process monitoring and process progress monitoring that minimizes user process management and manufacturing facility operation policy determination errors through accurate process progress monitoring, System can be provided.

1A is a graph showing individual signals received by various sensors mounted on an apparatus for manufacturing a semiconductor device according to an embodiment of the present invention.
FIG. 1B is a graph showing an intermediate signal obtained through arithmetic processing of the individual signals according to an embodiment of the present invention. FIG.
1C is a graph illustrating a reference pattern and an analyzed pattern defined from the intermediate signal according to an embodiment of the present invention.
2 is a flow chart illustrating a method for monitoring the progress of processes in an apparatus for manufacturing a semiconductor device according to an embodiment of the present invention.
3 illustrates a computer readable storage medium storing computer program code for performing a monitoring method according to an embodiment of the present invention.
4 is a schematic configuration diagram of a monitoring system according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The embodiments of the present invention are described in order to more fully explain the present invention to those skilled in the art, and the following embodiments may be modified in various other forms, The present invention is not limited to the embodiment. Rather, these embodiments are provided so that this disclosure will be more faithful and complete, and will fully convey the scope of the invention to those skilled in the art.

In the following drawings, thickness and size of each layer are exaggerated for convenience and clarity of description, and the same reference numerals denote the same elements in the drawings. As used herein, the term "and / or" includes any and all combinations of any of the listed items.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the singular forms "a," "an," and "the" include singular forms unless the context clearly dictates otherwise. Also, " comprise "and / or" comprising "when used herein should be interpreted as specifying the presence of stated shapes, numbers, steps, operations, elements, elements, and / And does not preclude the presence or addition of one or more other features, integers, operations, elements, elements, and / or groups.

Although the terms first, second, etc. are used herein to describe various elements, components, regions, layers and / or portions, these members, components, regions, layers and / It is obvious that no. These terms are only used to distinguish one member, component, region, layer or section from another region, layer or section. Thus, a first member, component, region, layer or section described below may refer to a second member, component, region, layer or section without departing from the teachings of the present invention.

As used herein, the term "sensor" is used to sense, distinguish, and measure physical quantities such as heat, light, temperature, pressure, and sound of a process being performed, Refers to a measuring device capable of indicating an electrical signal. For example, the sensor may be a quartz crystal microbalance (QCM), a Langmuir probe, a magnetic probe, a light sensor, a vibration sensor, a temperature sensor, a pressure sensor, a gas sensor, It may be a spectroscopy or mass spectrometer. In an embodiment of the invention, these sensors may be used singly or in combination of two or more.

The apparatus for manufacturing a semiconductor device in this specification may be a physical thin film deposition apparatus such as a sputtering or a laser ablation, a chemical thin film deposition apparatus such as a chemical vapor deposition apparatus and an atomic layer deposition apparatus, Or a dry etching apparatus. However, the apparatus for manufacturing a semiconductor device is not limited to the thin film deposition apparatus and the etching apparatus described above. For example, the semiconductor device manufacturing apparatus may be a device for manufacturing a semiconductor device such as a spin coater, a heat treatment apparatus, an aligning apparatus, a moving apparatus, a rotation apparatus, a measuring apparatus, a gas moving and controlling apparatus, . ≪ / RTI > Also, the facilities and devices may be in the form of a chamber, a bath, a tube or a plate, but the present invention is not limited thereto.

1A is a discrete signal (10, 11, 12, 13) received respectively by various sensors mounted on a semiconductor device manufacturing apparatus, and FIG. 1B is a circuit diagram of a semiconductor device according to an embodiment of the present invention, FIG. 1C is a graph showing a reference pattern 30 and an analyzed pattern 31 defined from the intermediate signal according to an embodiment of the present invention. Referring to FIG. The measured results of the graph are related to a plasma deposition apparatus in a semiconductor device manufacturing apparatus.

Referring to FIG. 1A, individual signals 10, 11, 12, and 13 received from sensors mounted on the reaction chamber and line of the plasma deposition apparatus may be generated. The sensors are for observing the magnetic field characteristics in the reaction chamber of the magnetron for controlling the electrical and chemical properties of the plasma or the plasma density in the reaction chamber. For example, the discrete signals may each include a signal 10 of a neutral beam intensity measurement sensor, a signal 11 of a neutral beam pressure measurement sensor, a signal 12 of a sensor measuring a change in charge amount of Thompson scattering, The signal 13 of FIG.

In an embodiment of the present invention, the individual signals 10, 11, 12, 13 may be used alone or in combination with two or more separate signals as is or in combination and used as an intermediate signal for monitoring the process. These individual signals 10, 11, 12, 13 may be subjected to noise removal processing to reduce measurement errors.

In order to convert into the intermediate signal, the arithmetic processing can be performed by a single discrete signal or a combination of any one or more of arithmetic operation, differentiation, integration, and scaling between these discrete signals. In another embodiment, in the arithmetic processing between the plurality of individual signals for converting into the intermediate signal, there may be arithmetic processing for setting the weight of the plurality of individual signals and giving the weight. The weighting calculation process monitors the main signal for the process so that the optimal process can be monitored. As another example, the discrete signals may be a digital signal, or the analog signal may be converted to a digital signal and defined as the intermediate signal through statistical methods such as Fourier transform or sequential probability test.

The intermediate signal can be used to indicate that the individual signals 10, 11, 12, 13 received from the respective sensors as described in FIG. 1A are used alone or in combination with two or more separate signals 10, 11, 12, And is a signal obtained by combination with the above-described arithmetic processing, and is a base signal for defining a reference pattern 30 and an analysis pattern 31 to be described later. Referring to FIG. 1B, the abscissa of the graph represents the time flow, and the ordinate axis represents the intermediate signal 20 according to the change of time. The intermediate signal 20 corresponds to the individual signals 10 , 11, 12, 13) are summed or multiplied. The intermediate signal 20 shown in FIG. 1B is illustrative, and the present invention is not limited thereto. As another example, in the above graph showing the intermediate signal 20, the abscissa indicates the other process parameters such as pressure and flow rate, not the change in time, and the ordinate indicates the value for the other process parameters different from the process parameters on the abscissa It is possible.

In one embodiment, the intermediate signal 20 may have a starting point and a terminating point. The starting point and the ending point may be the same as the starting point and finishing point of the process of the semiconductor device manufacturing apparatus. In another embodiment, the starting and ending points may be defined within any interval in the middle of the process.

The intermediate signal 20 may be defined as a first intermediate signal and a second intermediate signal, as follows, for monitoring the semiconductor device manufacturing apparatus. The first intermediate signal is an intermediate signal when a process performed through the semiconductor device manufacturing apparatus is normally performed. The first intermediate signal becomes a base signal defining the reference pattern 30 as shown in FIG. 1C. The second intermediate signal is an intermediate signal of the same post-process in the same semiconductor device manufacturing apparatus after the reference pattern 30 is defined. The second intermediate signal is a base signal for defining a pattern 31 to be analyzed, which is compared with the reference pattern 30 in FIG. 1C.

The reference pattern 30 and the analyzed pattern 31 shown in Fig. 1C are linear graphs, but this is merely an example, and the present invention is not limited thereto. As another example, the reference pattern and the analyzed pattern may include at least one of a discrete form, a repetitive waveform, a table, and a matrix.

Wherein the reference pattern includes at least one of a time, a position, a frequency, an amplitude, a frequency, an inflection point, a maximum value, a minimum value, a differential value (or slope), an average value, a variance value, an integral value, A minimum value, and a maximum value, a minimum value, and an average value of higher order differential values of a signal.

When the first intermediate signal includes information on the entire signal from the start point to the end point of the process performed through the semiconductor device manufacturing apparatus, the reference pattern is also obtained from the start point to the end point of the process And may include information of intermediate signals. Similarly, when the first intermediate signal has a start point and an end point within an arbitrary section of the middle of the process performed through the apparatus for manufacturing semiconductor elements, the reference pattern is also obtained within any section of the middle of the process And may include information of intermediate signals.

In one embodiment, the reference pattern may include information regarding a marginal error range. The limit error range may be defined by any one of a mean, a standard deviation, a minimum value, and a maximum value of a plurality of first intermediate signals obtained after repeating a plurality of processes determined to be normal processes, or a combination thereof.

Similarly, the analyzed pattern may be defined from the second intermediate signal in the same manner as the reference pattern is defined from the first intermediate signal. Wherein the analysis pattern includes at least one of a time, a position, a frequency, an amplitude, a frequency, an inflection point, a maximum value, a minimum value, a differential value (or slope), an average value, a variance value, an integral value, , The lowest point, and the maximum value, the minimum value and the average value of the absolute values of the second derivative of the signal, or two or more pieces of information.

If the reference pattern includes information on the entire signal from the start point to the end point of the process performed through the apparatus for manufacturing a semiconductor device, the analyzed pattern is also terminated from the start point of the process Based on the information of the obtained intermediate signals up to the point. Similarly, when the reference pattern has a start point and a termination point within an arbitrary section of the middle of the process performed through the apparatus for manufacturing a semiconductor device, the analyzed pattern is also a middle point obtained within an arbitrary section of the process intermediate Information of the signals.

2 is a flow chart illustrating a method for monitoring the progress of processes in an apparatus for manufacturing a semiconductor device according to an embodiment of the present invention.

Referring to FIG. 2, the first intermediate signal is received from a normally performed process among processes repeatedly performed by one or more sensors attached to the apparatus for manufacturing a semiconductor device, and the reference pattern is defined (S10). In step S20, the second intermediate signal is received in real time by the at least one sensor in the same post process after the normally performed process, and the analyzed pattern is generated.

Subsequently, the pattern to be analyzed and the reference pattern are compared by pattern recognition to determine whether there is a process abnormality (S30). The reference pattern and the analyzed pattern may be compared according to any one or more combinations of a template matching approach, a statistical approach, a structural approach, and a neural network approach, which are approaches of pattern recognition.

The template matching method is a method of recognizing whether or not the reference pattern and the pattern to be analyzed are identical to each other based on a degree of similarity such as a cross-correlation or a distance between a component of the reference pattern and a component of the pattern to be analyzed. The template matching method may be performed by setting a template serving as a reference for comparing the analyzed pattern using the elements of the reference pattern. The template matching method is advantageous in that the calculation speed is fast and the sensitivity of the pattern to be analyzed is sensitive because the patterns are compared based on the template.

The statistical approach is a method of recognizing whether or not the reference pattern and the analyzed pattern are identical based on a decision boundary generated from a statistical distribution of a set of constituent elements of a reference pattern defined from normal processes . The statistical model of the reference patterns generates a probability density function using, for example, Bayes' rule. A range of components for discriminating the identity between the reference pattern and the analyzed pattern can be selected by adjusting the range of the value of the crystal boundary. In addition, it is possible to add an operation other than a general recognition procedure between the reference pattern and the analyzed pattern, or to increase the accuracy of recognition of identities by increasing the number or types of elements of the reference pattern and the analyzed pattern.

The structural approach utilizes a structure formed between components that form a reference pattern defined from normally processed processes. Since the analysis is performed according to the relationship between the prototypes constituting the pattern without analyzing the reference pattern and the quantitative characteristics of the components of the analyzed pattern, the same can be recognized by the relationship on the graph . That is, the components of the reference pattern and the pattern to be analyzed are normalized to omit the analysis of the quantitative characteristics of the components, and the basic prototypes among the components are compared. The structural approach may improve or reduce the recognition efficiency of the analyzed pattern and the reference pattern while reducing or reducing the influence of the noise.

The neural network approach is a modeling of the connective-coupling relationship of biological neurons and is an approach to compare the reference pattern with the response process of the network consisting of processing elements for the components of the analyzed pattern. The information for comparing the reference pattern with the analyzed pattern is stored as a weight of a connection weight connecting the processing unit. In the neural network approach, as the above-described first intermediate signal for defining the reference pattern, that is, the signal information for the normally processed process, is more diversified, the connections of the processing units are formed to enable accurate comparison. The neural network approach may be distributed parallel processing, and one or more dedicated processors may be required to perform the computation of the neural network.

The method for monitoring the progress of the processes according to the embodiment of the present invention may further comprise the steps of: when the analyzed pattern is recognized as the same pattern as the reference pattern through pattern recognition of the reference pattern and the analyzed pattern, It is determined that the process is normally performed (YES) (S30). On the contrary, if the analyzed pattern and the reference pattern are recognized as different patterns, it is determined that the same post-process has been performed abnormally (No) (S30).

In the above-described embodiment, a single reference pattern and a single analyzed pattern are used to monitor the presence or absence of a process abnormality, but this is merely an example, and the present invention is not limited thereto. For example, the compared reference pattern and the analyzed pattern may include a plurality of patterns, and the plurality of patterns may be one-to-one, many-to-one, and one-to-many Most can be compared.

In another embodiment, in order to monitor the presence or absence of a process abnormality, the analyzed pattern is generated over time to determine whether the analyzed pattern generated in real time follows the reference pattern along with the flow of time Pattern recognition may be performed. Also, the time interval of the plurality of reference patterns may be set continuously or discontinuously, and a comparison may be made based on whether a real-time trend of the analyzed pattern follows the reference pattern.

If the analyzed pattern is recognized as the same pattern as the reference pattern in the step S30 of comparing the analyzed pattern and the reference pattern by pattern recognition to determine whether there is a process anomaly, the corresponding post-process is a normal process The same subsequent process will be repeated again in the semiconductor device manufacturing apparatus. A step of stopping the same post-process if the analyzed pattern is recognized as a pattern different from the reference pattern, or displaying an abnormality of the same post-process to a user by using one or more combinations of a warning sound, . ≪ / RTI >

In one embodiment, if it is determined that the analyzed pattern and the reference pattern are recognized as different patterns and the post-identical process is performed abnormally (NO), it is further determined whether the process is conformable to the same post process (S40). If it is determined in step S40 that the process conformity is determined to be suitable, the process proceeds to a learning step of including the analyzed pattern in the reference pattern to update the reference pattern, If it is determined to be a process, the process may be terminated. The updating of the reference pattern may include modifying the normal operating range of the equipment by reflecting changes in process parameters due to aging of the manufacturing equipment of the semiconductor device and / or interaction between equipment, do. The learning step may be performed according to the pattern recognition approach described above by updating the elements of the reference pattern.

Figure 3 illustrates a computer-readable storage medium 300 that stores computer program code for performing a monitoring method in accordance with an embodiment of the invention.

3, storage medium 300 may include information 310 about encoded program codes from instructions 320 for implementing the monitoring method 330 described with reference to FIG. 2 have. Examples of such instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter.

Such computer readable storage media 300 may include, for example, non-volatile memory devices such as RAM, ROM, EEPROM and flash memory, CD-ROMs, DVD ROMs, hard disks, Optical or holographic media, a magnetic tape, a solid state disk device (SSD), and the present invention is not limited thereto.

The above-described embodiments of the present invention may be implemented as a method, system or system for producing software, firmware, hardware, products or combinations thereof, such as, for example, a cloud computing environment, using programming and / Or storage media, and the article of manufacture may comprise a computer program, a carrier wave, or a media.

4 is a schematic block diagram of a monitoring system 400 according to an embodiment of the present invention.

4, the monitoring system 400 may include one or more sensors 41, one or more sensors 41a, 41b mounted on the semiconductor device manufacturing apparatus 40, A pattern for receiving the first intermediate signal from the performed process to define the reference pattern and receiving the second intermediate signal in real time by the one or more sensors in the same post process to generate the analyzed pattern And a control unit 43 for comparing the reference pattern with the analyzed pattern to determine whether there is an abnormality in the same after-process. The sensor unit 41, the pattern generating unit 42, and the control unit 43 are connected to each other via a network 46. 4, the network 46 is a wired network. However, this is exemplary and the network 46 may be a combination of one or more of a local wired / wireless communication network, a wired / wireless communication network, an internet network, and a mobile communication network, There may be relay devices such as hubs, gateways, and routers.

The control unit 43 may include an output unit 44 for displaying the determination result in addition to the function for determining the abnormality of the process and an input unit 45 for receiving the action information for the process from the user. The output unit 44 may be one or more of a screen display device such as a computer monitor, a portable terminal device, a warning light device, and a warning sound device, but the present invention is not limited thereto.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention. Will be clear to those who have knowledge of.

Claims (24)

A method for monitoring progress of repetitive processes of a semiconductor device manufacturing apparatus,
Defining a reference pattern by receiving a first intermediate signal from a reference process performed normally among the repetitive processes by one or more sensors mounted on the manufacturing apparatus;
Receiving a second intermediate signal from the same post process after the reference process in real time by the one or more sensors to define a pattern to be analyzed; And
And comparing the reference pattern and the analyzed pattern by pattern recognition to determine whether or not there is an abnormality in the same after-process. The method according to claim 1,
Wherein the step of determining whether or not the process abnormality is determined includes the steps of: when the pattern to be analyzed is recognized as the same pattern as the reference pattern by the pattern recognition, it is determined that the after- And if it is recognized as a pattern different from the reference pattern, it is determined that the same after-process has been performed abnormally. 3. The method of claim 2,
If it is determined that the post-same process has been performed abnormally, the process may be stopped or a combination of at least one of a warning sound, a warning light, and a warning screen for notifying the user of the process stop time and / Wherein the process interruption is indicated. The method according to claim 1,
And if the analyzed pattern is recognized as a pattern different from the reference pattern, whether or not process conformity is determined for the same post-process. 5. The method of claim 4,
If it is determined that the post-identical process is appropriate at the step of determining whether or not the process conformity is determined, the learning process is performed to include the analyzed pattern in the reference pattern to update the reference pattern, And if so, the process is terminated. 6. The method of claim 5,
Wherein the learning step is performed according to a pattern recognition approach by updating one or more components of the reference pattern and combining the template matching method, the statistical approach, the structural approach, and the neural network approach. The method according to claim 1,
Wherein the at least one sensor is a plurality of sensors and wherein the first intermediate signal and the second intermediate signal are selected from the group consisting of any one or a combination of two or more of arithmetic operation, differentiation, integration, scaling, A signal computed by the computation unit, or individual signals not computed. The method according to claim 1,
Wherein the first intermediate signal and the second intermediate signal are individual signals received from the one or more sensors are converted into digital signals and defined through statistical methods such as Fourier transform or sequential probability test. The method according to claim 1,
Wherein the signal received from the one or more sensors is a digital signal or an analog signal. The method according to claim 1,
The at least one sensor may be a quartz crystal microbalance (QCM), a Langmuir probe, a magnetic probe, an optical sensor, a vibration sensor, a temperature sensor, a pressure sensor, a gas sensor, a mass spectrometer, a spectroscopy, and an ion species mass spectrometer. The method according to claim 1,
Wherein the starting point and the ending point of the first intermediate signal are a starting point and a finishing point of a normally proceeding process of the semiconductor manufacturing facility, or a specific period of a process progress. The method according to claim 1,
The reference pattern and the pattern to be analyzed are time, position, frequency, amplitude, frequency, inflection point, maximum value, minimum value, differential value (or slope), average value, variance value Wherein at least one of a maximum value, a minimum value and an average of absolute values of an integral value, a real part, an imaginary part, a peak, a minimum point, and an absolute value of a second derivative of a signal is defined as a component. The method according to claim 1,
Wherein the reference pattern includes information about a marginal error range. The method according to claim 1,
Wherein the reference pattern is all or part of the signal from the start point to the end point of the reference process. The method according to claim 1,
Wherein the reference pattern and the analyzed pattern include a discrete form. The method according to claim 1,
Wherein the reference pattern and the analyzed pattern include repetitive waveforms. The method according to claim 1,
Wherein the reference pattern and the analyzed pattern include a form of a table or a matrix. The method according to claim 1,
The step of determining whether or not an abnormality of the same post process is determined may be determined by a pattern recognition approach according to any one or two or more combinations of the template pattern matching method, the statistical approach, the structural approach, and the neural network approach A method of monitoring the progress of a process being performed. The method according to claim 1,
Wherein the step of determining whether there is an abnormality in the same post-process includes comparing the analyzed pattern and the reference pattern by one-to-one comparison, by-one comparison, and by many-to-many comparison. The method according to claim 1,
Wherein the step of determining whether or not the abnormality is present is performed based on whether a real-time trend of the analyzed pattern follows the reference pattern. The method according to claim 1,
Wherein the step of determining whether or not the abnormality is present comprises the steps of setting a time interval of a plurality of reference patterns consecutively or discontinuously and comparing based on whether or not a real time trend of the analyzed pattern follows the reference pattern . A computer readable storage medium storing computer program code for performing a method of monitoring a manufacturing device of a semiconductor device,
The monitoring method includes:
Receiving a first signal from a reference process performed normally among repetitive processes of the semiconductor device manufacturing apparatus by one or more sensors mounted on the manufacturing apparatus to define a reference pattern;
Receiving a second signal in real time from the same post process after the reference process by the at least one sensor to generate a analyzed pattern; And
And comparing the reference pattern with the analyzed pattern by pattern recognition to determine whether or not there is an abnormality in the after-the-same process. 23. The method of claim 22,
If it is determined that the analyzed pattern of the same post-process is not the same as the reference pattern but the post-process is appropriate, the learning step of including the analyzed pattern in the reference pattern is performed to update the reference pattern And terminating the process if it is determined that the process is not suitable. At least one sensor unit mounted on a semiconductor device manufacturing apparatus;
A first pattern generator for receiving a first signal from a reference process normally performed during repetitive processes of the apparatus for manufacturing a semiconductor device by the one or more sensors to define a reference pattern;
A second pattern generator for receiving a second signal in real time by the one or more sensors from the same post process after the reference process and generating an analysis pattern; And
And comparing the reference pattern with the analyzed pattern by pattern recognition to determine whether or not there is an abnormality in the same after-process.

Images