JP4643392B2 - Operating state determination method, operating state determination apparatus, program, and storage medium for plasma processing apparatus - Google Patents

Description

  The present invention relates to a method for determining an operating state of a plasma processing apparatus, an operating state determining apparatus, a program, and a storage medium, and in particular, when the processing conditions of plasma processing applied to a certain plasma processing apparatus are applied to another plasma processing apparatus. It is related with the driving | running state determination method in.

  A plasma processing apparatus that performs plasma processing on a semiconductor wafer (hereinafter simply referred to as “wafer”) as a substrate includes a chamber that accommodates the wafer, an electrode disposed in the chamber, and a shower head that supplies a processing gas into the chamber. In addition to the above, various sensors for measuring output values of the respective constituent elements as processing parameter values are provided. This plasma processing apparatus controls each component according to the applied processing conditions (recipe) and performs various plasma processes on the wafer. At this time, in the plasma processing apparatus, output values and the like of each component to be controlled are measured by various sensors as processing parameter values, and the measured processing parameter values are recorded as log data.

  Usually, when a predetermined plasma process that has not been executed in a certain plasma processing apparatus (hereinafter referred to as “applied apparatus”) is executed, another plasma processing apparatus that has a track record of executing the predetermined plasma process ( Hereinafter, the processing conditions applied in the “reference apparatus” (hereinafter referred to as “applied processing conditions”) are applied to the plasma processing apparatus. However, since the plasma processing is a delicate process, even if the apparatus to be applied performs plasma processing on the wafer according to the application processing conditions, normal processing is performed on the wafer due to individual differences between the reference apparatus and the apparatus to be applied. It may not be done. Therefore, it is necessary to determine whether the apparatus to be applied performs normal processing on the wafer.

  Whether or not the apparatus to be applied performs normal processing on the wafer according to the application processing conditions, that is, whether or not the plasma processing process in the apparatus to be applied is normally performed is normal according to the application processing conditions in the reference apparatus. It is determined by comparing the data of the processing results of a large number of wafers processed in a step with the data of the processing results of a large number of wafers processed in accordance with the application processing conditions in the application apparatus. Specifically, when the data of the processing result in the reference apparatus and the data of the processing result in the application target apparatus substantially coincide with each other, it is determined that the wafer is normally processed, and the processing result in the reference apparatus is If the data does not match the processing result data in the apparatus to be applied, it is determined that the wafer has not been processed normally.

  A processing parameter value (hereinafter referred to as a “normal processing parameter value”) measured by the application apparatus when the data of the processing result in the reference apparatus and the data of the processing result in the application apparatus are substantially the same. This is a criterion for determining whether or not normal processing is performed on the wafer in the application apparatus (operating state of the apparatus to be applied). For example, in the subsequent plasma processing, if the processing parameter value measured by the application apparatus is substantially the same as the normal processing parameter value, it can be determined that the wafer is normally processed (normal operating state). If the measured processing parameter value does not match the normal processing parameter value, it can be determined that the wafer is not normally processed (abnormal operating state). Note that many normal processing parameter values are required for accurate determination.

  By the way, the data items of the wafer processing results (for example, the width and depth of the trench formed in the wafer) and the types of processing parameter values (for example, the chamber wall temperature and the applied voltage of the electrodes) are diverse. Multivariate analysis, especially principal component analysis, is often used to compare the processing result data of the reference device with the processing result data of the application device, and the measured processing parameter value and the normal processing parameter value. The By using principal component analysis, it is possible to convert processing result data consisting of a large number of variables (items, types) and processing parameter values into data consisting of one or two variables, for example. Comparison can be performed easily and easily.

As a method using principal component analysis, principal component analysis (first principal component value, second principal component value) is performed by principal component analysis of the values of a plurality of monitored process data, and the principal component values are There is known a method for evaluating the operating status of a semiconductor manufacturing apparatus (plasma processing apparatus) by monitoring in series (for example, see Patent Document 1).
JP 2004-47885 A

  However, in order to be able to determine whether or not normal processing has been performed on the wafer in accordance with the application processing conditions in the application apparatus (operating state of the application apparatus), processing result data of a large number of wafers in the application apparatus Or many normal processing parameter values are required. Therefore, there is a problem in that the operating state of the apparatus to be applied cannot be determined unless plasma processing is performed in advance in the apparatus to be applied according to the application processing conditions.

  The object of the present invention is to perform the first plasma in the second plasma processing apparatus without executing the plasma processing in advance in the second plasma processing apparatus according to the processing conditions applied to the first plasma processing apparatus. An object of the present invention is to provide an operation state determination method, an operation state determination device, a program, and a storage medium of a plasma processing apparatus that can determine an operation state when processing conditions applied to a processing apparatus are applied.

  In order to achieve the above object, an operating state determination method for a plasma processing apparatus according to claim 1 is the second case where the processing conditions applied to the first plasma processing apparatus are applied to the second plasma processing apparatus. An operation state determination method for a plasma processing apparatus for determining an operation state of the plasma processing apparatus, measured in a plasma process executed according to a first processing condition applied to the first plasma processing apparatus A first operating state data group acquisition step for acquiring a first operating state data group consisting of processing parameter values, and a plasma process executed in accordance with a second processing condition applied to the first plasma processing apparatus. A second operating state data group acquiring step for acquiring a second operating state data group consisting of the processing parameter values measured in step 2, and the second plasma processing A third operating state data group acquiring step for acquiring a third operating state data group consisting of processing parameter values measured in the plasma processing performed in accordance with the first processing condition applied to a device; A principal component analysis step of performing principal component analysis using at least one of the acquired first operating state data group, second operating state data group, and third operating state data group; and Main corresponding to two principal components selected from a plurality of principal components obtained by the principal component analysis using the driving state data group, the second driving state data group, and the third driving state data group Principal component score plotting step for plotting component scores in a two-dimensional coordinate system with the two principal components as axes, and the first operating state data group and second operating state data in the two-dimensional coordinate system A distribution center calculating step for calculating a center of a principal component score distribution corresponding to each of the group and the third operating state data group; and a center of the principal component score distribution corresponding to the first operating state data group. A movement vector calculation step for calculating a movement vector to the center of the principal component score distribution corresponding to the three driving state data groups; and a principal component score distribution corresponding to the second driving state data group in the two-dimensional coordinate system. A distribution area setting step for setting an area; and a region of a principal component score distribution corresponding to the set second operating state data group is moved according to the movement vector, thereby causing the second plasma processing apparatus to The principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing performed when the processing condition of 2 is applied A predicted distribution area setting step for setting a predicted distribution area in a two-dimensional coordinate system, and the second processing condition is actually applied to the second plasma processing apparatus, and is executed according to the second processing condition. A principal component score inclusion determination step for determining whether or not principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing are included in the set prediction distribution region. It is characterized by that.

  The operation state determination method of the plasma processing apparatus according to claim 2 is the operation state determination method of the plasma processing apparatus according to claim 1, wherein the center of the principal component score distribution is the first operation state data group, the second It is an average value or median value of the principal component score distribution corresponding to the driving state data group and the third driving state data group.

  The operation state determination method for the plasma processing apparatus according to claim 3 is the operation state determination method for the plasma processing apparatus according to claim 1 or 2, wherein the two principal components are a first principal component and a second principal component. It is characterized by.

  The plasma processing apparatus operating state determination method according to claim 4 is the plasma processing apparatus operating state determination method according to claim 1 or 2, wherein the two main components are the first operation in the two-dimensional coordinate system. The principal component score distribution corresponding to each of the state data group, the second operation state data group, and the third operation state data group is selected from a plurality of principal components so that the centers are arranged on a straight line. It is characterized by that.

  The operation state determination method for a plasma processing apparatus according to claim 5 is the operation state determination method for a plasma processing apparatus according to any one of claims 1 to 4, wherein the main operation state data group corresponds to the second operation state data group. The component score distribution region is set based on a variance value of principal component scores corresponding to the second operating state data group.

  The operating state determination method of the plasma processing apparatus according to claim 6 is the operating state determination method of the plasma processing apparatus according to any one of claims 1 to 5, wherein the main processing state data corresponding to the second operating state data group is set. The component score distribution region is an ellipse having the two principal components as axes.

  The operating state determination method for a plasma processing apparatus according to claim 7 is the operating state determination method for a plasma processing apparatus according to any one of claims 1 to 6, wherein the operating state determination method is applied to the first plasma processing apparatus. Processing parameter values measured in the plasma processing executed according to the first processing condition, processing measured in the plasma processing executed according to the second processing condition applied to the first plasma processing apparatus Both the parameter value and the processing parameter value measured in the plasma processing executed according to the first processing condition applied to the second plasma processing apparatus are measured in the plasma processing performed normally. It is characterized by that.

  The plasma processing apparatus operating state determination method according to claim 8 is the plasma processing apparatus operating state determination method according to any one of claims 1 to 7, wherein in the principal component analysis step, the acquired first state is obtained. Principal component analysis is performed using all of the first operation state data group, the second operation state data group, and the third operation state data group.

  In order to achieve the above object, the operation state determination apparatus according to claim 9 is the second plasma processing apparatus when the processing conditions applied to the first plasma processing apparatus are applied to the second plasma processing apparatus. The operation state determination device for determining the operation state of the first, comprising a processing parameter value measured in the plasma processing executed in accordance with the first processing condition applied to the first plasma processing apparatus From the first operating state data group acquisition means for acquiring the operating state data group and the processing parameter values measured in the plasma processing executed in accordance with the second processing condition applied to the first plasma processing apparatus. A second operating state data group acquiring means for acquiring the second operating state data group, and the first processing condition applied to the second plasma processing apparatus. Third operating state data group acquiring means for acquiring a third operating state data group consisting of processing parameter values measured in the plasma processing to be performed, the acquired first operating state data group, and second operation A principal component analysis means for performing principal component analysis using at least one of the state data group and the third operation state data group; the first operation state data group; the second operation state data group; A two-dimensional principal component score corresponding to two principal components selected from a plurality of principal components obtained by the principal component analysis using the three driving state data groups, with the two principal components as axes. Principal component score plotting means for plotting in the coordinate system, and main components corresponding to each of the first operating state data group, the second operating state data group, and the third operating state data group in the two-dimensional coordinate system. Distribution center calculation means for calculating the center of the score distribution, and from the center of the principal component score distribution corresponding to the first operating state data group to the center of the principal component score distribution corresponding to the third operating state data group A movement vector calculation means for calculating a movement vector; a distribution area setting means for setting a principal component score distribution area corresponding to the second operating state data group in the two-dimensional coordinate system; and the set second It is measured in the plasma processing executed when the second processing condition is applied to the second plasma processing apparatus by moving the principal component score distribution region corresponding to the operating state data group according to the movement vector. Predictive distribution area setting means for setting a predictive distribution area in the two-dimensional coordinate system of principal component scores corresponding to the two principal components of the processing parameter values The second processing condition is actually applied to the second plasma processing apparatus, and corresponds to the two principal components of the processing parameter values measured in the plasma processing executed according to the second processing condition. And a principal component score inclusion judging means for judging whether or not a principal component score is included in the set prediction distribution region.

  In order to achieve the above object, the program according to claim 10 is an operating state of the second plasma processing apparatus when the processing conditions applied to the first plasma processing apparatus are applied to the second plasma processing apparatus. Is a program for causing a computer to execute an operating state determination method for a plasma processing apparatus, and is a process measured in plasma processing executed according to a first processing condition applied to the first plasma processing apparatus In a first operation state data group acquisition module for acquiring a first operation state data group consisting of parameter values, and in plasma processing executed in accordance with a second processing condition applied to the first plasma processing apparatus A second operating state data group acquisition module for acquiring a second operating state data group consisting of the measured processing parameter values; 3rd operation state data group acquisition which acquires the 3rd operation state data group which consists of a processing parameter value measured in plasma processing performed according to the 1st processing conditions applied to the 2nd plasma processing apparatus A principal component analysis module that performs principal component analysis using at least one of the module and the acquired first operating state data group, second operating state data group, and third operating state data group; Two principal components selected from a plurality of principal components obtained by the principal component analysis using the first operating state data group, the second operating state data group, and the third operating state data group A principal component score plotting module for plotting a principal component score corresponding to 1 in the two-dimensional coordinate system with the two principal components as axes, and the first operating state data in the two-dimensional coordinate system. A distribution center calculation module for calculating the center of the principal component score distribution corresponding to each of the group, the second operation state data group, and the third operation state data group, and a main corresponding to the first operation state data group A movement vector calculation module for calculating a movement vector from the center of the component score distribution to the center of the principal component score distribution corresponding to the third operation state data group; and the second operation state data group in the two-dimensional coordinate system. A distribution area setting module for setting a principal component score distribution area corresponding to the second operating state data group and moving the principal component score distribution area corresponding to the set second operating state data group according to the movement vector. The two main parameter values measured in the plasma processing executed when the second processing condition is applied to the second plasma processing apparatus. Applying the second processing condition to the predicted distribution region setting module for setting the predicted distribution region in the two-dimensional coordinate system of the principal component score corresponding to the component, and actually applying the second processing condition to the second plasma processing apparatus; It is determined whether or not principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing executed according to the processing condition 2 are included in the set prediction distribution region. And a principal component score inclusion determination module.

  In order to achieve the above object, the storage medium according to claim 11 is an operation of the second plasma processing apparatus when the processing conditions applied to the first plasma processing apparatus are applied to the second plasma processing apparatus. A computer-readable storage medium storing a program for causing a computer to execute an operation state determination method for a plasma processing apparatus for determining a state, wherein the program is a first process applied to the first plasma processing apparatus. A first operating state data group acquisition module for acquiring a first operating state data group consisting of processing parameter values measured in plasma processing executed according to conditions, and the first plasma processing apparatus. A second operating state data consisting of processing parameter values measured in the plasma processing executed in accordance with the second processing condition. A second operating state data group acquisition module for acquiring a data group, and a processing parameter value measured in the plasma processing executed in accordance with the first processing condition applied to the second plasma processing apparatus. A third operation state data group acquisition module for acquiring a third operation state data group, and the acquired first operation state data group, second operation state data group, and third operation state data group; The principal component analysis using a principal component analysis module that performs principal component analysis using at least one of them, and the first operation state data group, the second operation state data group, and the third operation state data group Principal component score plot for plotting principal component scores corresponding to two principal components selected from among a plurality of principal components obtained by the above in a two-dimensional coordinate system having the two principal components as axes. Joule and a distribution center for calculating a center of a principal component score distribution corresponding to each of the first operation state data group, the second operation state data group, and the third operation state data group in the two-dimensional coordinate system And a movement vector calculation module for calculating a movement vector from the center of the principal component score distribution corresponding to the first driving state data group to the center of the principal component score distribution corresponding to the third driving state data group. A distribution area setting module for setting a principal component score distribution area corresponding to the second operating state data group in the two-dimensional coordinate system, and a principal component corresponding to the set second operating state data group Executed when the second processing condition is applied to the second plasma processing apparatus by moving the score distribution area according to the movement vector. A predicted distribution region setting module for setting a predicted distribution region in the two-dimensional coordinate system of principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing; and actually the second plasma processing Applying the second processing condition to the apparatus, the principal component score corresponding to the two principal components of the processing parameter values measured in the plasma processing executed according to the second processing condition is set. And a principal component score inclusion determination module for determining whether or not it is included in the region of the predicted distribution.

  According to the operating state determination method of the plasma processing apparatus according to claim 1, the operating state determination apparatus according to claim 9, the program according to claim 10, and the storage medium according to claim 11, the first operating state data group The principal component score corresponding to two principal components selected from a plurality of principal components obtained by principal component analysis using the second operational state data group and the third operational state data group is 2 Plotted in a two-dimensional coordinate system with two principal components as axes, from the center of the principal component score distribution corresponding to the first operating state data group to the center of the principal component score distribution corresponding to the third operating state data group A movement vector is calculated, and the second plasma is moved by moving the principal component score distribution region corresponding to the second operating state data group set in the two-dimensional coordinate system according to the calculated movement vector. A region of a predicted distribution in a two-dimensional coordinate system of principal component scores corresponding to two principal components of processing parameter values measured in the plasma processing executed when the second processing condition is applied to the processing apparatus; Principal component scores corresponding to the two principal components of the processing parameter values actually applied in the second plasma processing apparatus and measured in the plasma processing executed according to the second processing condition Is determined in the second plasma processing apparatus without performing plasma processing according to the second processing condition in the second plasma processing apparatus. It is possible to set a determination criterion for an operation state in the case where the second processing condition is applied, and therefore, it is applied in advance to the first plasma processing apparatus in the second plasma processing apparatus. Without running the plasma process in accordance with the management conditions, it is possible to determine the operating conditions in the case of applying the processing conditions to be applied to the first plasma processing apparatus to the second plasma processing apparatus.

  According to the operating state determination method of the plasma processing apparatus according to claim 2, the center of the principal component score distribution is in the first operating state data group, the second operating state data group, and the third operating state data group. Since it is the average value or median value of the corresponding principal component score distribution, the center of the principal component score distribution can be easily calculated, and the operating state of the second plasma processing apparatus can be determined in a short time. .

  According to the operating state determination method of the plasma processing apparatus of claim 3, since the two principal components are the first principal component and the second principal component, it corresponds to each operating state data group plotted in the two-dimensional coordinate system. The statistical reliability of the distribution of the main component scores to be improved can be improved, so that the operating state of the second plasma processing apparatus can be accurately determined.

  According to the operating state determination method of the plasma processing apparatus of claim 4, the two main components are the first operating state data group, the second operating state data group, and the third operating state in the two-dimensional coordinate system. Since the principal component score distribution corresponding to each of the data groups is selected from a plurality of principal components so that the center is arranged on a straight line, it is easy to grasp the relative positional relationship of each principal component score distribution Therefore, the operating state of the second plasma processing apparatus can be easily determined.

  According to the operating state determination method of the plasma processing apparatus of claim 5, the principal component score distribution region corresponding to the second operating state data group is a variance of the principal component scores corresponding to the second operating state data group. Since it is set based on the value, the statistical reliability of the set principal component score distribution region can be improved, and the operating state of the second plasma processing apparatus can be determined more accurately. .

  According to the operating state determination method of the plasma processing apparatus of the sixth aspect, since the principal component score distribution region is an ellipse having two principal components as axes, it can be set easily and easily.

  According to the operating state determination method for the plasma processing apparatus according to claim 7, since each processing parameter value is measured in the plasma processing performed normally, each operating state data plotted in the two-dimensional coordinate system. The reliability of the main component score corresponding to the group can be improved, and thus the operating state of the second plasma processing apparatus can be determined more accurately.

  According to the operation state determination method of the plasma processing apparatus of claim 8, the main operation is performed using all of the acquired first operation state data group, second operation state data group, and third operation state data group. Since component analysis is performed, the statistical reliability of the two-dimensional coordinate system can be improved.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  First, a plasma processing apparatus and an operation state determination apparatus to which the operation state determination method according to the first embodiment of the present invention is applied will be described.

  FIG. 1 is a cross-sectional view illustrating a schematic configuration of a plasma processing apparatus and an operation state determination apparatus to which the operation state determination method according to the present embodiment is applied.

  In FIG. 1, an etching processing apparatus 1 configured as a plasma processing apparatus places, for example, an aluminum cylindrical chamber 2 and a semiconductor wafer W having a diameter of 200 mm, for example, disposed in the chamber 2. A lower electrode 3, a support 5 that can be moved up and down supported via an insulating material 4, and a shower head 6 as an upper electrode that is disposed above the inside of the chamber 2 so as to face the lower electrode 3.

  The upper portion of the chamber 2 is formed as an upper chamber 7 having a small diameter, and the lower portion is formed as a lower chamber 8 having a large diameter. A dipole ring magnet 9 is disposed around the upper chamber 7, and the dipole ring magnet 9 forms a uniform horizontal magnetic field directed in one direction in the upper chamber 7. A gate valve 10 for opening and closing the loading / unloading port of the semiconductor wafer W is attached to the upper side of the lower chamber 8, and the etching processing apparatus 1 is connected to an adjacent load lock chamber (not shown) or the like via the gate valve 10. Yes. A heater unit (not shown) is embedded in the side wall of the upper chamber 7, and the heater unit controls the atmospheric temperature in the chamber 2, particularly in the upper chamber 7.

  A high frequency power source 11 is connected to the lower electrode 3 via a matching unit 12, and the high frequency power source 11 applies predetermined high frequency power to the lower electrode 3. A high frequency power source 30 is connected to the shower head 6 disposed on the ceiling portion of the chamber 2 via a matching unit 31, and the high frequency power source 30 applies a predetermined high frequency power to the shower head 6. The lower electrode 3 and the shower head 6 cooperate to generate a high-frequency electric field in the processing space S between the lower electrode 3 and the shower head 6.

  On the upper surface of the lower electrode 3, an electrostatic chuck (ESC) 13 for adsorbing the semiconductor wafer W with an electrostatic attraction force is disposed. A disk-shaped electrode plate 14 made of a conductive film is disposed inside the electrostatic chuck 13, and a DC power supply 15 is electrically connected to the electrode plate 14. The semiconductor wafer W is attracted and held on the upper surface of the electrostatic chuck 13 by a Coulomb force or a Johnson-Rahbek force generated by a DC voltage applied to the electrode plate 14 from the DC power supply 15.

  An annular focus ring 16 is disposed around the electrostatic chuck 13, and the focus ring 16 surrounds the periphery of the semiconductor wafer W attracted to the electrostatic chuck 13. In addition, since the focus ring 16 is made of silicon carbide or quartz, the focus ring 16 has substantially the same conductivity as that of the semiconductor wafer W. Thereby, the focus ring 16 efficiently guides plasma, which will be described later, generated in the chamber 2 to the semiconductor wafer W.

  An exhaust path is formed between the side wall of the upper chamber 7 and the lower electrode 3 to exhaust the gas above the lower electrode 3 to the outside of the chamber 2, and an annular baffle plate 17 is disposed in the middle of the exhaust path. Has been. The space downstream of the baffle plate 17 in the exhaust passage (the internal space of the lower chamber 8) communicates with an automatic pressure control valve (hereinafter referred to as “APC valve”) 18 that is a variable butterfly valve. . The APC valve 18 is connected to a turbo molecular pump (hereinafter referred to as “TMP”) (not shown) which is an exhaust pump for evacuation, and TMP is a dry pump (hereinafter referred to as “TMP”). DP ") (not shown). The exhaust flow path constituted by the APC valve 18, TMP, and DP controls the pressure in the chamber 2 by the APC valve 18, and further reduces the pressure in the chamber 2 by the TMP and DP until the inside of the chamber 2 is almost in a vacuum state.

  Below the lower electrode 3, a lower electrode lifting mechanism comprising a ball screw 19 extending downward from the lower portion of the support 5 is disposed. The lower electrode raising / lowering mechanism supports the lower electrode 3 via the support 5 and raises / lowers the lower electrode 3 by rotating the ball screw 19 with a motor or the like (not shown). This lower electrode raising / lowering mechanism is shielded from the atmosphere in the chamber 2 by a bellows 20 arranged around the lower electrode and a bellows cover 21 arranged around the bellows 20.

  The lower electrode 3 is provided with a plurality of pusher pins 22 that can protrude from the upper surface of the electrostatic chuck 13. These pusher pins 22 move in the vertical direction in the figure. In the etching processing apparatus 1, when the semiconductor wafer W is loaded / unloaded, the lower electrode 3 is lowered to the loading / unloading position of the semiconductor wafer W, and the pusher pins 22 protrude from the upper surface of the electrostatic chuck 13 to move the semiconductor wafer W to the lower position. Separated from the electrode 3 and lifted upward. Further, during the etching process of the semiconductor wafer W, the lower electrode 3 rises to the processing position of the semiconductor wafer W, and the pusher pins 22 are stored in the lower electrode 3 so that the electrostatic chuck 13 holds the semiconductor wafer W by suction. To do.

  Further, for example, an annular refrigerant chamber 23 extending in the circumferential direction is provided inside the lower electrode 3. A coolant having a predetermined temperature, for example, cooling water, is circulated and supplied to the coolant chamber 23 through a pipe 24 from a chiller unit (not shown), and the semiconductor wafer W placed on the lower electrode 3 according to the temperature of the coolant. The processing temperature is controlled.

  A plurality of heat transfer gas supply holes and heat transfer gas supply grooves (not shown) are arranged on the upper surface of the electrostatic chuck 13. These heat transfer gas supply holes and the like are connected to a heat transfer gas supply unit 26 via a heat transfer gas supply line 25 disposed inside the lower electrode 3, and the heat transfer gas supply unit 26 includes a heat transfer gas, For example, He gas is supplied to the gap between the electrostatic chuck 13 and the semiconductor wafer W. Thereby, the heat transfer between the semiconductor wafer W and the electrostatic chuck 13 is improved. The heat transfer gas supply unit 26 is also configured to be able to evacuate the gap between the electrostatic chuck 13 and the semiconductor wafer W.

A buffer chamber 27 is provided on the upper surface of the shower head 6, and a processing gas introduction pipe 28 from a processing gas supply unit (not shown) is connected to the buffer chamber 27. An MFC (Mass Flow Controller) 29 is disposed in the middle of the processing gas introduction pipe 28. The MFC 29 supplies a predetermined gas, for example, processing gas or N ₂ gas into the chamber 2 through the buffer chamber 27 and the shower head 6 and controls the flow rate of the gas to control the pressure in the chamber 2. The desired value is controlled in cooperation with the exhaust passage.

  In the chamber 2 of the etching processing apparatus 1, as described above, high-frequency power is applied to the lower electrode 3 and the shower head 6, and high-density plasma is generated from the processing gas in the processing space S by the applied high-frequency power. Is generated and ions and radicals are generated.

In the etching processing apparatus 1, during the etching process, first, the gate valve 10 is opened, and the semiconductor wafer W to be processed is loaded into the chamber 2. Then, a processing gas (for example, a mixed gas composed of at least one of carbon tetrafluoride gas (CF ₄ ) gas and oxygen (O ₂ ) gas having a predetermined flow rate ratio) is supplied from the shower head 6 at a predetermined flow rate and flow rate ratio. The pressure is introduced into the chamber 2 and the pressure in the chamber 2 is set to a predetermined value by an exhaust passage or the like. Further, a high frequency power is applied to the lower electrode 3 from the high frequency power source 11, a high frequency power is applied to the shower head 6 from the high frequency power source 30, and a DC voltage is applied to the electrode plate 14 from the DC power source 15. Adsorbed on the lower electrode 3. Then, the processing gas discharged from the shower head 6 is turned into plasma as described above. The plasma is focused on the surface of the semiconductor wafer W by the focus ring 16, and ions generated by the plasma, such as fluorine ions and oxygen ions, physically etch the surface of the semiconductor wafer W.

  The etching processing apparatus 1 includes a plurality of sensors (not shown) that measure the output values of the above-described components, for example, the wall temperature of the upper chamber 7, the applied voltage of the lower electrode 3, and the opening degree of the APC valve 18 in real time. Have. The plurality of sensors measure an output value of each component as a processing parameter value during the etching process, and transmits the measured processing parameter to a system controller 32 described later.

  The system controller 32 (operating state determination device) is an integrated server that controls the plasma processing system including the etching processing device 1 and the load lock chamber, and is connected to the etching processing device 1 and the load lock chamber. A RAM 35, an HDD 36 as a recording device, a display unit 37, and a bus 38 for connecting the respective components to each other are provided. Further, the system controller 32 controls the operation of each component of the etching processing apparatus 1.

  The ROM 34 is a program for executing an etching process in the etching processing apparatus 1, a program for executing an operating state determination method for determining an operating state of the etching processing apparatus 1 to be described later, and various statistical processes such as many Stores programs for executing variable analysis (principal component analysis, regression analysis). The RAM 35 is a writable memory, and each program is loaded from the ROM 34. The RAM 35 is also used as a work area when the CPU 33 executes each process according to the loaded program.

  The HDD 36 stores processing parameters transmitted by the sensors of the etching processing apparatus 1 as device log data of the etching processing apparatus 1. Since the apparatus log data is stored every time the semiconductor wafer W is processed, a large amount of apparatus log data exists in the HDD 36. Of these apparatus log data, the apparatus log data corresponding to the same etching process is in the same operation state. Form a data group. The HDD 36 may store each program. The CPU 33 communicates addresses and data with the ROM 34, the RAM 35, and the HDD 36 via the bus 38, and executes each process by executing a program loaded in the RAM 35.

  The display unit 37 is composed of, for example, an LCD (Liquid Crystal Display), and displays the operation status of each component of the etching processing apparatus 1. The display unit 37 also functions as an input unit that receives input from the operator, and the operator can call a program corresponding to a desired process via the display unit 37 and input an execution instruction for the program.

  Next, the driving | running state determination method which concerns on the 1st Embodiment of this invention is demonstrated.

  Normally, an etching processing apparatus can perform various etching processes on a semiconductor wafer. However, when performing a desired etching process that has not been performed in the etching processing apparatus, the etching processing apparatus has the same specifications as the etching processing apparatus. An other etching processing apparatus (hereinafter referred to as “reference apparatus”) that has already successfully performed a desired etching process on a semiconductor wafer is searched for, and the desired etching process in the reference apparatus is performed. (For example, the flow ratio of processing gas and the application time of high-frequency power) (hereinafter referred to as “recipe”) are applied to the etching processing apparatus. Hereinafter, an etching processing apparatus to which a recipe applied to the reference apparatus (hereinafter referred to as “applied recipe”) is applied is referred to as “applied apparatus”.

  However, even with an etching processing apparatus of the same specification, there are various variations in the components in each etching processing apparatus, such as variations in dimensional accuracy and mounting accuracy of the focus ring, differences in response performance of the matching unit, and slight gas Differences in flow rate control capability (deviation), temperature control capability differences due to heater unit performance differences and circulating cooling water supply capability differences, TMP and DP exhaust capability differences, and consumption amount differences of consumable parts placed in the chamber Therefore, individual differences occur between the respective etching processing apparatuses, and even if the same recipe is applied to each etching processing apparatus, the result of the etching process of the semiconductor wafer (for example, the width and depth of the trench formed in the semiconductor wafer) May be different. Therefore, even if the application recipe is applied to the apparatus to be applied, it is not always possible to normally perform a desired etching process on the semiconductor wafer. Therefore, when the application recipe is applied to the device to be applied, it is determined whether or not the desired etching process has been normally performed on the semiconductor wafer, thereby determining whether the operation state of the device to be applied is normal or abnormal. There is a need to.

  By the way, as described above, the system controller connected to each etching processing apparatus stores a large amount of processing parameter values in various etching processes as apparatus log data of the etching processing apparatus. A large amount of stored apparatus log data is classified for each identical recipe, and forms the same operating state data group corresponding to each recipe. Therefore, in determining whether the operation state of the device to be applied is normal or abnormal, it is conceivable to use the operation state data group of the device to be applied or the reference device.

  Here, in the operation state data group, apparatus log data (processing parameter values) is grouped for each identical semiconductor wafer. Hereinafter, the grouped process parameter values are referred to as “single wafer process parameter groups”. For example, if K types of processing parameter values (for example, chamber wall temperature and applied voltage of electrodes) are measured for each semiconductor wafer, one single wafer processing parameter group includes K processing parameter values. .

  Usually, since there are various types of processing parameter values in the operating state data group (for example, chamber wall temperature and applied voltage of electrodes), it is difficult to use the operating state data group as it is. Therefore, the driving state determination method according to the present embodiment uses principal component analysis which is multivariate analysis.

  Principal component analysis is a method of converting many variables (each processing parameter value in the present embodiment) into principal components, which are a small number of comprehensive indicators, with as little information loss as possible. The principal component is a linear transformation of the original variable. The principal component that reflects the most information of the original variable (that is, the variance is maximum) is the first principal component, and is orthogonal to the first principal component. In addition, the main component having the second largest dispersion after the first main component is defined as the second main component.

  For example, in the present embodiment, when N semiconductor wafers are etched according to the first recipe in the reference apparatus, N single wafer processing parameter sets exist in one operating state data group. To do. Each processing parameter value may change with time in the etching process, but in this embodiment, an average value of the processing parameter value in the etching process is used. Here, each processing parameter value is x, and the number of processing parameter values measured for each semiconductor wafer is K (that is, K kinds of processing parameter values are measured for each semiconductor wafer). The state data group is represented by a matrix X shown in the following equation (1-1).

  Next, an average value, maximum value, minimum value, and dispersion value (for example, average value, maximum value, minimum value, and dispersion value of N electrode applied voltages) for each processing parameter value of the same type in the operation state data group X are obtained. The eigenvalues and eigenvectors of the principal components of the K kinds of processing parameter values are calculated using a variance-covariance matrix based on these calculated values. Here, as many as K principal components are obtained, which is the same as the number of types of processing parameter values in one single wafer processing parameter set.

  The eigenvalue represents the variance of each principal component, and each principal component is defined as a first principal component, a second principal component, a-th principal component, and a K-th principal component in the order of the eigenvalues. ing. Each eigenvalue has an eigenvector (loading vector) corresponding to the eigenvalue. In general, the higher the order of the principal component, the lower the contribution rate indicating the degree of information reflection of the processing parameter value, and the less the utility value. Therefore, in order to perform statistically reliable analysis, it is preferable to use a low-order principal component (for example, the first principal component or the second principal component).

Further, a value in each principal component of each single wafer processing parameter group in the driving state data group X is referred to as a principal component score (score), and corresponds to, for example, the nth single wafer processing parameter group in the driving state data group X. The a-th principal component score is expressed by the following equation (1-2).
t _na = x _n1 p _1a + x _n2 p _2a +... + x _nK p _Ka (1-2)
Here, p _na (n = 1, 2,..., K) represents each component of the eigenvector (loading vector) corresponding to the a-th principal component.

  When each single wafer processing parameter set of the operating state data group X is converted into each principal component score, the converted vector (score vector) ta of the a-th principal component score and the matrix (score matrix) Ta of all principal component scores are It is represented by the following formula (1-3).

  In the driving state determination method according to the present embodiment, any two principal component score vectors are selected and used from the converted principal component score vectors. In the operating state determination method according to the present embodiment, a large amount of semiconductor wafers are normally subjected to etching processing according to each of the first recipe and the second recipe in the reference device, and The operating state of the device to be applied when the second recipe is applied to the device to be applied when the etching process is normally applied to a large amount of semiconductor wafers according to the recipe of 1 is determined.

  Therefore, before the operation state determination method according to the present embodiment is executed, the measurement is performed when the reference device performs an etching process on a large number of semiconductor wafers according to the first recipe in the HDD 36 of the system controller 32. An operation state data group (hereinafter referred to as “reference device first recipe operation state data group”) (first operation state data group) consisting of the processing parameter values (device log data), and the reference device according to the second recipe Operation state data group (hereinafter referred to as “reference device second recipe operation state data group”) (second operation state data group) consisting of processing parameter values measured when a large number of semiconductor wafers are etched. And an operation state data comprising processing parameter values measured when the apparatus to be applied performs an etching process on a large number of semiconductor wafers according to the first recipe. Group (hereinafter, referred to as "the application device a first recipe operation status data group") (third operation status data group) is stored.

  In each etching process described above, values for 16 types of processing parameters are measured for each semiconductor wafer, and thus each single wafer processing parameter group in each operating state data group includes 16 processing parameter values. The processing parameters to be measured include, for example, upper electrode high frequency power reflection amount (absolute value), lower electrode high frequency power reflection amount (absolute value), lower electrode applied voltage, upper electrode temperature, lower electrode temperature, and APC valve opening. Temperature, upper chamber wall temperature, ESC applied current, matching device capacitance, and matching device inductance. Needless to say, the number of types of processing parameters measured for each semiconductor wafer is not limited to 16, and the types of processing parameters measured for each semiconductor wafer are not limited to those described above.

  FIG. 2 is a flowchart of the operation state determination method according to the present embodiment. Processing corresponding to this method is executed by the CPU 33 in accordance with a program loaded in the RAM 35 in response to an input via the display unit 37 of the operator.

  In FIG. 2, first, the reference device first recipe operation state data group stored in the HDD 36 is acquired (step S201), and subsequently, the reference device second recipe operation state data group and the applied device first recipe operation state data. A group is acquired (steps S202 and S203).

  Next, a principal component analysis is performed using all of the acquired reference device first recipe operation state data group, reference device second recipe operation state data group, and applied device first recipe operation state data group, and eigenvalues of each principal component and An eigenvector is calculated (step S204).

  Next, as a result of the principal component analysis in step S204, the first principal component and the second principal component are selected from the 16 principal components obtained, and the eigenvectors corresponding to the first principal component and the second principal component are used. Thus, as in the above equation (1-2), each sheet processing parameter set in each operating state data group is converted into a first principal component score and a second principal component score, and the converted first principal component score and As shown in FIG. 3, the second principal component score is plotted on a two-dimensional coordinate system having the first principal component on the horizontal axis and the second principal component on the vertical axis (step S205). In FIG. 3, the principal component score corresponding to the reference device first recipe operation state data group is represented by “◯”, and the principal component score corresponding to the reference device second recipe operation state data group is represented by “●”. The principal component score corresponding to the application target device first recipe operation state data group is represented by “◇”.

  Next, an average value of the first principal component score and the second principal component score in the distribution of the principal component scores corresponding to the reference device first recipe operation state data group plotted in the two-dimensional coordinate system is calculated as the center of the distribution. Similarly, the average value of the first principal component score and the second principal component score in the distribution of the principal component scores corresponding to the reference device second recipe operation state data group, and the application target device first recipe operation state data group The average value of the first principal component score and the second principal component score in the distribution of the principal component scores is calculated as the center of each distribution.

Then, a movement vector P from the center of the principal component score distribution corresponding to the reference device first recipe operation state data group to the center of the principal component score distribution corresponding to the application target device first recipe operation state data group is calculated (step) In step S206, a principal component score distribution area A2 (hereinafter referred to as “measured normal area A2”) corresponding to the reference device second recipe operation state data group is set in the two-dimensional coordinate system (step S207). The actual measurement normal area A2 is an elliptical area, and the major axis and minor axis of the ellipse are set based on the variance values of the first principal component score and the second principal component score in the reference device second recipe operation state data group. Is done. Specifically, when the standard deviation of the first principal component score in the reference device second recipe operation state data group is σ ₁ and the standard deviation of the second principal component score is σ ₂ , the ellipse of the measured normal region A2 is the reference An ellipse having a major axis length of 6σ ₁ and a minor axis length of 6σ ₂ around the center of the principal component score distribution corresponding to the apparatus second recipe operation state data group.

  Next, the measured normal area A2 is moved in accordance with the movement vector P in the two-dimensional coordinate system, and the moved measured normal area A2 is set as the predicted normal area B2 (step S208). Here, since the movement vector P is considered to represent the transition amount of each processing parameter value when the apparatus to which the same recipe is applied is changed from the reference apparatus to the apparatus to be applied, the predicted normal area B2 is the second It is considered that the distribution of principal component scores (first principal component score, second principal component score) is predicted when the device to which the recipe is applied is changed from the reference device to the device to be applied.

  Next, in practice, the second recipe is applied to the apparatus to be applied, and the semiconductor wafer is subjected to the etching process according to the second recipe, and 16 kinds of processing parameter values are measured during the etching process to set the single wafer processing parameter group. Get. Further, the obtained single wafer processing parameter set is converted into a first principal component score and a second principal component score using eigenvectors corresponding to the first principal component and the second principal component, and in the two-dimensional coordinate system of FIG. Then, it is determined whether or not points (hereinafter referred to as “determined points”) corresponding to the converted first principal component score and second principal component score are included in the predicted normal region B2 (step S209).

  As a result of the determination in step S209, if the determined point is included in the predicted normal area B2, the target apparatus normally performs the etching process on the semiconductor wafer in accordance with the second recipe, and the target apparatus is subjected to the second process. When the recipe is applied, it is determined that the operation state of the apparatus to be applied is normal (step S210), and the process is terminated.

  As a result of the determination in step S209, if the determined point is not included in the predicted normal area B2, the target apparatus does not normally perform the etching process on the semiconductor wafer according to the second recipe, and the target apparatus It determines with the driving | running state of the said to-be-applied apparatus when 2 recipes are applied being abnormal (step S211), and complete | finishes this process.

  According to the operation state determination method according to the present embodiment, the main component using the reference device first recipe operation state data group, the reference device second recipe operation state data group, and the application target device first recipe operation state data group. The principal component scores (first principal component score and second principal component score) corresponding to the first principal component and the second principal component obtained by the analysis (step S204) are the axes of the first principal component and the second principal component. Is plotted in the two-dimensional coordinate system (step S205), and the principal component score distribution corresponding to the application target device first recipe operation state data group from the center of the principal component score distribution corresponding to the reference device first recipe operation state data group is plotted. The movement vector P up to the center of is calculated (step S206), and the measured normal area A2 set in the two-dimensional coordinate system is moved according to the calculated movement vector P. The predicted normal region B2 is set (step S208), and the second recipe is actually applied to the apparatus to be applied, and the single wafer processing parameter composed of the processing parameter values measured in the etching processing executed in accordance with the second recipe. It is determined whether the first principal component score and the second principal component score of the set are included in the predicted normal area B2 (step S209). Since the movement vector P represents the transition amount of each processing parameter value when the device to which the same recipe is applied is changed from the reference device to the device to be applied, the prediction normal region B2 is a device to which the second recipe is applied. The distribution of principal component scores when the reference device is changed to the device to be applied is predicted. Therefore, by setting the predicted normal region B2, the determination criterion for the operation state when the second recipe is applied to the application target device without setting the etching process according to the second recipe in the application target device is set. Therefore, it is possible to determine the operating state when the second recipe is applied to the apparatus to be applied without executing the etching process according to the second recipe in the apparatus to be applied in advance.

  In addition, in order to confirm the accuracy of the operation state determination method according to the present embodiment, the present inventor applied the second recipe to the apparatus to be applied and performed a normal etching process on a large amount of semiconductor wafers. The operation state data group (henceforth "applied apparatus 2nd recipe operation state data group") which consists of a lot of single wafer processing parameter groups in was obtained. Furthermore, a large number of single wafer processing parameter sets in the obtained application target device second recipe operation state data group are converted into first principal component scores and second principal component scores, and these first principal component scores and second component scores are converted. When the principal component scores are plotted in a two-dimensional coordinate system using “♦” as shown in FIG. 4, it is confirmed that the second recipe operation state data group to be applied is almost included in the predicted normal area B2. did. Thereby, it turned out that the driving | running state in the case of applying a 2nd recipe to a to-be-applied apparatus using prediction normal area | region B2 can be determined correctly.

  The operation state determination method according to the present embodiment described above corresponds to each of the reference device first recipe operation state data group, the reference device second recipe operation state data group, and the application target device first recipe operation state data group. Since the center of the principal component score distribution is an average value of the respective principal component score distributions, the center of the principal component score distribution can be easily calculated, so that the operating state of the device to be applied can be determined in a short time. it can. In addition, the median value of each principal component score distribution may be calculated as the center of the principal component score distribution, whereby the same effect can be obtained.

  In the driving state determination method according to the above-described embodiment, since the two-dimensional coordinate system has the first principal component and the second principal component as axes, it corresponds to each driving state data group plotted in the two-dimensional coordinate system. Statistical reliability of the distribution of principal component scores can be improved, and thus the operating state of the device to be applied can be accurately determined.

  In the driving state determination method according to the present embodiment described above, the measured normal area A2 is set based on the first principal component score and the variance value of the second principal component score corresponding to the reference device second recipe driving state data group. Therefore, the statistical reliability of the measured normal area A2 can be improved, and the operating state of the application target apparatus can be determined more accurately. In addition, since the measured normal area A2 is an ellipse having the first principal component and the second principal component as axes, it can be set easily and easily.

  Moreover, in the driving | running state determination method which concerns on this Embodiment mentioned above, each process in a reference | standard apparatus 1st recipe driving | running state data group, a reference | standard apparatus 2nd recipe driving | running state data group, and a to-be-applied apparatus 1st recipe driving | running state data group Since all parameter values are measured when the semiconductor wafer is normally etched, it is possible to improve the reliability of the principal component score corresponding to each operation state data group, and thus the apparatus to be applied. Can be determined more accurately.

  Furthermore, in the operation state determination method according to the present embodiment described above, the acquired reference device first recipe operation state data group, reference device second recipe operation state data group, and applied device first recipe operation state data group. Since the principal component analysis is performed using all of the above, the statistical reliability of the two-dimensional coordinate system can be improved.

  Further, in the operation state determination method according to the above-described embodiment, the single wafer obtained by actually applying the second recipe to the apparatus to be applied and subjecting the semiconductor wafer to the etching process according to the second recipe. When the first principal component score and the second principal component score converted from the processing parameter set are not included in the predicted normal region B2, it is immediately determined that the operating state of the applied apparatus is abnormal. When a single wafer processing parameter set is obtained, only when a percentage of the first principal component score and the second principal component score converted from the plurality of single wafer processing parameter sets is not included in the predicted normal region B2. The operating state of the device to be applied may be determined to be abnormal.

  In the operation state determination method according to the present embodiment described above, the movement vector P corresponds to the center of the principal component score distribution corresponding to the reference device first recipe operation state data group and the application target device first recipe operation state data group. However, the method of calculating the movement vector P is not limited to this. For example, as shown in FIG. 5, a principal component score distribution region A1 (hereinafter referred to as “measured normal region A1”) corresponding to the reference device first recipe operation state data group is set in the two-dimensional coordinate system, and A principal component score distribution region B1 (hereinafter referred to as “measured normal region B1”) corresponding to the first recipe operation state data group to be applied is set, and the areas of the measured normal region A1 and the measured normal region B1 are the same. Thus, the measured normal area A1 and the measured normal area B1 may be adjusted, and the movement vector P may be calculated based on the centers (center of gravity) of the adjusted measured normal area A1 and measured normal area B1. Thereby, from the movement vector P, the difference between the form of the principal component score distribution corresponding to the reference apparatus first recipe operation state data group and the form of the principal component score distribution corresponding to the application target first recipe operation state data group Therefore, the statistical reliability of the predicted normal region B2 set using the movement vector P can be improved.

  Also, a method for removing the influence of the difference between the form of the principal component score distribution corresponding to the reference device first recipe operation state data group and the form of the principal component score distribution corresponding to the application target first recipe operation state data group The method is not limited to the method described above. For example, as shown in FIG. 6, the movement vector P is calculated by the same method as the method in step S206, the measured normal area A1 and the measured normal area B1 are set in the two-dimensional coordinate system, and the measured normal area A1 and the measured normal area are set. When calculating the correlation of B1 (for example, area ratio, ellipse major axis ratio, minor axis ratio) and moving the measured normal region A2 according to the movement vector P to set the predicted normal region B2, the measured normal region A2 and The correlation between the actual measurement normal area A1 and the actual measurement normal area B1 calculated above may be applied to the correlation between the predicted normal area B2.

  Next, the driving | running state determination method which concerns on the 2nd Embodiment of this invention is demonstrated.

  This embodiment is basically the same in configuration and operation as the above-described first embodiment, and the main component to be selected is different from that in the above-described first embodiment. Therefore, the description of the duplicated configuration and operation is omitted, and the description of the different configuration and operation is given below.

  In the driving state determination method according to the present embodiment, in step S205 in FIG. 2, two principal components satisfying a predetermined condition are selected from the 16 principal components obtained as a result of the principal component analysis in step S204. To do.

  The two main components that satisfy the predetermined condition in the present embodiment are the reference device first recipe operation state data group, the reference device second recipe operation state data group, and the application target device first recipe operation state data group. When the single wafer processing parameter set is converted into the principal component scores of the two principal components, and the converted principal component scores are plotted in a two-dimensional coordinate system with the two principal components as axes, the reference device first recipe The principal component is such that the center of the principal component score distribution corresponding to each of the operation state data group, the reference device second recipe operation state data group, and the application target device first recipe operation state data group is arranged on a straight line. In this embodiment, the third principal component and the fifth principal component correspond.

  Therefore, in the driving state determination method according to the present embodiment, each single wafer processing parameter set in each driving state data group is determined using the eigenvectors corresponding to the third principal component and the fifth principal component. Converted into five principal component scores, and the converted third principal component score and fifth principal component score are set with the third principal component on the horizontal axis and the fifth principal component on the vertical axis, as shown in FIG. Plot in a two-dimensional coordinate system. Also in FIG. 7, the principal component score corresponding to the reference device first recipe operation state data group is represented by “◯”, and the principal component score corresponding to the reference device second recipe operation state data group is represented by “●”. The principal component score corresponding to the application target first recipe operation state data group is represented by “◇”. Further, the center of the principal component score distribution corresponding to each operating state data group is arranged on a one-dot chain line shown in the figure.

  According to the driving state determination method according to the present embodiment, when two principal components are selected from the 16 principal components obtained as a result of the principal component analysis, each single wafer processing parameter in each driving state data group is selected. When a set is converted into a principal component score of the two principal components and plotted in a two-dimensional coordinate system with the two principal components as axes, the center of the principal component score distribution corresponding to each of the operating state data groups Are selected so that the relative positional relationship of each principal component score distribution corresponding to each operating state data group can be easily grasped. The operating state when the second recipe is applied can be easily determined.

  Moreover, the driving | running state determination method which concerns on this Embodiment can show effects other than the effect by having selected the 1st main component and the 2nd main component among the effects in 1st Embodiment mentioned above. Needless to say. Furthermore, in the operation state determination method according to the present embodiment, the main component score distribution form corresponding to the reference device first recipe operation state data group shown in FIG. 5 and FIG. It goes without saying that a method for removing the influence of the difference from the form of the principal component score distribution corresponding to the data group is also applicable.

  In order to confirm the accuracy of the operation state determination method according to the present embodiment, the inventor has obtained a large number of single wafer processing parameter sets in the second recipe operation state data group to be applied as the third principal component score and When converted into the fifth principal component score, the third principal component score and the fifth principal component score are plotted in a two-dimensional coordinate system using “♦” as shown in FIG. It was confirmed that the two recipe operation state data group was almost included in the predicted normal region B2. Thereby, also in the driving | running state determination method which concerns on this Embodiment, it turned out that the driving | running state at the time of applying a 2nd recipe to an applicable apparatus using prediction normal area | region B2 can be determined correctly.

  Next, the driving | running state determination method which concerns on the 3rd Embodiment of this invention is demonstrated.

  This embodiment is basically the same in configuration and operation as the first embodiment described above, and the number of operating state data groups used in the principal component analysis and the selected principal components are as described above. Different from the first embodiment. Therefore, the description of the duplicated configuration and operation is omitted, and the description of the different configuration and operation is given below.

  In the operation state determination method according to the present embodiment, in step S204 of FIG. 2, principal component analysis is performed using only the reference device first recipe operation state data group and the application target device first recipe operation state data group. Compute the eigenvalues and eigenvectors of the components.

  Next, in step S205 in FIG. 2, two principal components satisfying a predetermined condition are selected from the 16 principal components obtained as a result of the principal component analysis in step S204.

  The two main components that satisfy the predetermined condition in the present embodiment are the reference device first recipe operation state data group, the reference device second recipe operation state data group, and the application target device first recipe operation state data group. When the single wafer processing parameter set is converted into the principal component scores of the two principal components, and the converted principal component scores are plotted in a two-dimensional coordinate system with the two principal components as axes, the reference device first recipe The principal component score distribution corresponding to each of the operation state data group, the reference device second recipe operation state data group, and the application target device first recipe operation state data group is a principal component that is appropriately separated and arranged. In the embodiment, the first principal component and the fifth principal component correspond.

  Therefore, in the driving state determination method according to the present embodiment, each single wafer processing parameter set in each driving state data group is obtained using the eigenvectors corresponding to the first principal component and the fifth principal component. Converted into five principal component scores, and the converted first principal component score and fifth principal component score are represented by the first principal component on the horizontal axis and the fifth principal component on the vertical axis, as shown in FIG. Plot in a two-dimensional coordinate system. In FIG. 9, the principal component score corresponding to the reference device first recipe operation state data group is represented by “◯”, and the principal component score corresponding to the reference device second recipe operation state data group is represented by “●”. The principal component score corresponding to the application target first recipe operation state data group is represented by “◇”.

  According to the operation state determination method according to the present embodiment, the principal component analysis is performed using only the reference device first recipe operation state data group and the application target device first recipe operation state data group. The reference device first recipe operation state data group and the application target device first recipe operation state data group are both operation state data groups when the first recipe is applied, and the reference device second recipe operation state data group is the second. It is a driving | running state data group when a recipe is applied. Here, since the tendency of the processing parameter value when the first recipe is applied is different from the tendency of the processing parameter value when the second recipe is applied, the reference device first recipe operation state data group and the application target device are different. When the principal component score corresponding to the reference device second recipe operation state data group is plotted on the two-dimensional coordinate system based on the principal component obtained by the principal component analysis performed using only the first recipe operation state data group, The distribution of principal component scores corresponding to the reference device first recipe operation state data group and the application target first recipe operation state data group and the distribution of principal component scores corresponding to the reference device second recipe operation state data group are clear. To separate. Thereby, it is possible to easily grasp the form of each principal component score distribution.

  Moreover, the driving | running state determination method which concerns on this Embodiment is a reference | standard apparatus 1st recipe driving | running state data group, a reference | standard apparatus 2nd recipe driving | running state data group, and to-be-applied among the effects in 1st Embodiment mentioned above. It goes without saying that effects other than the effect of performing the principal component analysis using all of the apparatus first recipe operation state data group and the effect of selecting the first principal component and the second principal component can be achieved. . Furthermore, in the operation state determination method according to the present embodiment, the main component score distribution form corresponding to the reference device first recipe operation state data group shown in FIG. 5 and FIG. It goes without saying that a method for removing the influence of the difference from the form of the principal component score distribution corresponding to the data group is also applicable.

  In order to confirm the accuracy of the operation state determination method according to the present embodiment, the present inventor selects a large number of single wafer processing parameter sets in the second recipe operation state data group to be applied as the first principal component score and the first component score. When converted into five principal component scores, the first principal component score and the fifth principal component score are plotted in a two-dimensional coordinate system using “♦” as shown in FIG. It was confirmed that the recipe operation state data group was almost included in the predicted normal area B2. Thereby, also in the driving | running state determination method which concerns on this Embodiment, it turned out that the driving | running state at the time of applying a 2nd recipe to an applicable apparatus using prediction normal area | region B2 can be determined correctly.

  Next, the driving | running state determination method which concerns on the 4th Embodiment of this invention is demonstrated.

  This embodiment is basically the same in configuration and operation as the first embodiment described above, and the number of operating state data groups used in the principal component analysis and the selected principal components are as described above. Different from the first embodiment. Therefore, the description of the duplicated configuration and operation is omitted, and the description of the different configuration and operation is given below.

  In the operation state determination method according to the present embodiment, in step S204 of FIG. 2, principal component analysis is performed using only the reference device first recipe operation state data group to calculate eigenvalues and eigenvectors of each principal component.

  Next, in step S205 in FIG. 2, two principal components satisfying a predetermined condition are selected from the 16 principal components obtained as a result of the principal component analysis in step S204.

  The two principal components satisfying the predetermined condition in the present embodiment are principal components having properties similar to those of the principal component selected in the third embodiment described above, and in this embodiment, the second principal component. The component and the fifth main component are applicable.

  Therefore, in the driving state determination method according to the present embodiment, each single wafer processing parameter set in each driving state data group is obtained by using the eigenvectors corresponding to the second principal component and the fifth principal component. Converted into five principal component scores, and the converted second principal component score and fifth principal component score are represented by the second principal component on the horizontal axis and the fifth principal component on the vertical axis, as shown in FIG. Plot in a two-dimensional coordinate system. Also in FIG. 11, the principal component score corresponding to the reference device first recipe operation state data group is represented by “◯”, and the principal component score corresponding to the reference device second recipe operation state data group is represented by “●”. The principal component score corresponding to the application target first recipe operation state data group is represented by “◇”.

  According to the driving state determination method according to the present embodiment, since the principal component analysis is performed using only the reference device first recipe driving state data group, the individual difference between the reference device and the application target device in the principal component analysis. , And the processing condition difference between the first recipe and the second recipe is not considered. Therefore, the application target device first recipe operation state data group and the reference device second are applied to the two-dimensional coordinate system based on the principal component obtained by the principal component analysis performed using only the reference device first recipe operation state data group. When the principal component scores corresponding to the recipe operation state data group are plotted, the distribution of the principal component scores corresponding to each operation state data group is clearly separated from each other. Thereby, it is possible to easily grasp the form of each principal component score distribution.

  Moreover, the driving | running state determination method which concerns on this Embodiment is also the reference | standard apparatus 1st recipe driving | running state among the effects in 1st Embodiment mentioned above similarly to the driving | running state determination method which concerns on 3rd Embodiment. Effect by performing principal component analysis using all of data group, reference device second recipe operation state data group, and applied device first recipe operation state data group, and selection of first principal component and second principal component Needless to say, effects other than those obtained can be achieved. Furthermore, in the operation state determination method according to the present embodiment, the main component score distribution form corresponding to the reference device first recipe operation state data group shown in FIG. 5 and FIG. It goes without saying that a method for removing the influence of the difference from the form of the principal component score distribution corresponding to the data group is also applicable.

  In order to confirm the accuracy of the operation state determination method according to the present embodiment, the present inventor has obtained a large number of single wafer processing parameter sets in the second recipe operation state data group to be applied as the second principal component score and the second component score. When converted into five principal component scores, the second principal component score and the fifth principal component score are plotted in a two-dimensional coordinate system using “♦” as shown in FIG. It was confirmed that the recipe operation state data group was almost included in the predicted normal area B2. Thereby, also in the driving | running state determination method which concerns on this Embodiment, it turned out that the driving | running state at the time of applying a 2nd recipe to an applicable apparatus using prediction normal area | region B2 can be determined correctly.

  In the operation state determination method in each embodiment described above, two principal components are selected from the plurality of principal components obtained as a result of the principal component analysis, and the two-dimensional coordinate system having the two principal components as axes is selected. Although the principal component score is plotted, the number of selected principal components is not limited to two, and may be three or more. For example, three principal components may be selected from the obtained plurality of principal components, and the principal component scores may be plotted in a three-dimensional coordinate system having the three principal components as axes. In this case, since the three principal component scores are reflected in the points plotted in the three-dimensional coordinate system, it is possible to reduce the loss of information of the single wafer processing parameter set, and thus statistically reliable operation. The state can be determined.

  Further, the system controller 32 as the operating state determination device in each of the above-described embodiments is provided separately from the etching processing apparatus 1, but may be integrated with the etching processing apparatus. Thereby, space saving can be performed. Further, the system controller 32 may be configured by a portable PC or the like instead of the integrated server. Thereby, the convenience of the operator can be improved.

  Furthermore, the substrate subjected to the etching process in the etching processing apparatus to which the operation state determination method in each embodiment described above is applied is not limited to a semiconductor wafer, but may be an LCD (Liquid Crystal Display), an FPD (Flat Panel Display), or the like. Various substrates to be used, a photomask, a CD substrate, a printed substrate, and the like may be used.

  Moreover, although the operating state determination method in each embodiment described above is applied to the etching processing apparatus, the plasma processing apparatus to which the present operating state determination method can be applied is not limited to the etching processing apparatus, and is applied to, for example, a CVD processing apparatus. May be.

  In addition, an object of the present invention is to supply a computer, for example, the system controller 32, a storage medium that records software program codes that implement the functions of the above-described embodiment, and the CPU of the computer is stored in the storage medium. It is also achieved by reading and executing the program code.

  In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiment, and the program code and the storage medium storing the program code constitute the present invention.

  Examples of the storage medium for supplying the program code include RAM, NV-RAM, floppy (registered trademark) disk, hard disk, magneto-optical disk, CD-ROM, CD-R, CD-RW, DVD (DVD). -ROM, DVD-RAM, DVD-RW, DVD + RW) and other optical disks, magnetic tapes, non-volatile memory cards, other ROMs, etc., as long as they can store the program code. Alternatively, the program code may be supplied to the computer by downloading from another computer or database (not shown) connected to the Internet, a commercial network, a local area network, or the like.

  Further, by executing the program code read by the computer, not only the functions of the present embodiment are realized, but also an OS (operating system) running on the CPU based on the instruction of the program code, etc. Includes a case where part or all of the actual processing is performed and the above-described functions of the present embodiment are realized by the processing.

  Further, after the program code read from the storage medium is written in a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. This includes a case where the CPU or the like provided in the board or the function expansion unit performs part or all of the actual processing, and the above-described functions of the present embodiment are realized by the processing.

  The form of the program code may include an object code, a program code executed by an interpreter, script data supplied to the OS, and the like.

It is sectional drawing which shows schematic structure of the plasma processing apparatus and the driving | running state determination apparatus to which the driving | running state determination method which concerns on the 1st Embodiment of this invention is applied. It is a flowchart of the driving | running state determination method which concerns on this Embodiment. It is a figure for demonstrating the setting method of the prediction normal area | region B2 in the driving | running state determination method which concerns on this Embodiment. It is a figure which shows the relationship between the prediction normal area | region B2 set in the driving | running state determination method which concerns on this Embodiment, and the main component score distribution corresponding to a to-be-applied apparatus 2nd recipe driving | running state data group. It is a figure for demonstrating the 1st modification of the setting method of the prediction normal area | region B2 in the driving | running state determination method which concerns on this Embodiment. It is a figure for demonstrating the 2nd modification of the setting method of the prediction normal area | region B2 in the driving | running state determination method which concerns on this Embodiment. It is a figure for demonstrating the setting method of the prediction normal area | region B2 in the driving | running state determination method which concerns on the 2nd Embodiment of this invention. It is a figure which shows the relationship between the prediction normal area | region B2 set in the driving | running state determination method which concerns on this Embodiment, and the main component score distribution corresponding to a to-be-applied apparatus 2nd recipe driving | running state data group. It is a figure for demonstrating the setting method of the prediction normal area | region B2 in the driving | running state determination method which concerns on the 3rd Embodiment of this invention. It is a figure which shows the relationship between the prediction normal area | region B2 set in the driving | running state determination method which concerns on this Embodiment, and the main component score distribution corresponding to a to-be-applied apparatus 2nd recipe driving | running state data group. It is a figure for demonstrating the setting method of the prediction normal area | region B2 in the driving | running state determination method which concerns on the 4th Embodiment of this invention. It is a figure which shows the relationship between the prediction normal area | region B2 set in the driving | running state determination method which concerns on this Embodiment, and the main component score distribution corresponding to a to-be-applied apparatus 2nd recipe driving | running state data group.

Explanation of symbols

W Wafer S Processing space 1 Etching apparatus 2 Chamber 3 Lower electrode 6 Shower head 7 Upper chamber 11, 30 High frequency power supply 12, 31 Matching device 13 Electrostatic chuck 15 DC power supply 32 System controller 33 CPU
36 HDD
37 Display

Claims (11)

An operating state determination method for a plasma processing apparatus for determining an operating state of the second plasma processing apparatus when the processing conditions applied to the first plasma processing apparatus are applied to a second plasma processing apparatus,
First operating state data group for acquiring a first operating state data group composed of processing parameter values measured in plasma processing executed in accordance with a first processing condition applied to the first plasma processing apparatus. An acquisition step;
A second operating state data group for acquiring a second operating state data group composed of processing parameter values measured in the plasma processing executed in accordance with the second processing condition applied to the first plasma processing apparatus. An acquisition step;
Third operating state data for acquiring a third operating state data group composed of processing parameter values measured in the plasma processing executed according to the first processing condition applied to the second plasma processing apparatus. A group acquisition step;
A principal component analysis step of performing principal component analysis using at least one of the acquired first operating state data group, second operating state data group, and third operating state data group;
Two principal components selected from a plurality of principal components obtained by the principal component analysis using the first operating state data group, the second operating state data group, and the third operating state data group A principal component score plotting step for plotting a principal component score corresponding to the above in a two-dimensional coordinate system with the two principal components as axes;
A distribution center calculation step of calculating a center of a principal component score distribution corresponding to each of the first operation state data group, the second operation state data group, and the third operation state data group in the two-dimensional coordinate system; ,
A movement vector calculation step for calculating a movement vector from the center of the principal component score distribution corresponding to the first driving state data group to the center of the principal component score distribution corresponding to the third driving state data group;
A distribution area setting step for setting a principal component score distribution area corresponding to the second operating state data group in the two-dimensional coordinate system;
Executed when the second processing condition is applied to the second plasma processing apparatus by moving the principal component score distribution region corresponding to the set second operating state data group according to the movement vector. A predicted distribution area setting step for setting a predicted distribution area in the two-dimensional coordinate system of principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing to be performed;
The second processing condition is actually applied to the second plasma processing apparatus, and corresponds to the two principal components of the processing parameter values measured in the plasma processing executed according to the second processing condition. And a principal component score inclusion determination step for determining whether or not a principal component score is included in the set prediction distribution region.   The center of the principal component score distribution is an average value or median value of principal component score distributions corresponding to the first operating state data group, the second operating state data group, and the third operating state data group. The method for determining an operating state of a plasma processing apparatus according to claim 1.   The method of claim 1 or 2, wherein the two principal components are a first principal component and a second principal component.   The two principal components have a principal component score distribution center corresponding to each of the first operating state data group, the second operating state data group, and the third operating state data group in the two-dimensional coordinate system. 3. The operating state determination method for a plasma processing apparatus according to claim 1, wherein the operating state is selected from a plurality of main components so as to be arranged on a straight line.   The region of the principal component score distribution corresponding to the second operating state data group is set based on a variance value of the principal component scores corresponding to the second operating state data group. 5. The operating state determination method for a plasma processing apparatus according to any one of 4 above.   6. The plasma processing according to claim 1, wherein a region of the principal component score distribution corresponding to the second operating state data group is an ellipse having the two principal components as axes. Method for determining the operating state of the device.   The processing parameter value measured in the plasma processing executed according to the first processing condition applied to the first plasma processing apparatus, and the second processing condition applied to the first plasma processing apparatus. The processing parameter value measured in the plasma processing executed according to the above, and the processing parameter value measured in the plasma processing executed according to the first processing condition applied to the second plasma processing apparatus, 7. The method for determining an operating state of a plasma processing apparatus according to any one of claims 1 to 6, wherein both are measured in a plasma processing performed normally.   In the principal component analysis step, principal component analysis is performed using all of the acquired first operation state data group, second operation state data group, and third operation state data group. The operating state determination method of the plasma processing apparatus of any one of Claims 1 thru | or 7. An operating state determination device that determines an operating state of the second plasma processing apparatus when the processing conditions applied to the first plasma processing apparatus are applied to the second plasma processing apparatus,
First operating state data group for acquiring a first operating state data group composed of processing parameter values measured in plasma processing executed in accordance with a first processing condition applied to the first plasma processing apparatus. Acquisition means;
A second operating state data group for acquiring a second operating state data group composed of processing parameter values measured in the plasma processing executed in accordance with the second processing condition applied to the first plasma processing apparatus. Acquisition means;
Third operating state data for acquiring a third operating state data group composed of processing parameter values measured in the plasma processing executed according to the first processing condition applied to the second plasma processing apparatus. Group acquisition means;
Principal component analysis means for performing principal component analysis using at least one of the acquired first operating state data group, second operating state data group, and third operating state data group;
Two principal components selected from a plurality of principal components obtained by the principal component analysis using the first operating state data group, the second operating state data group, and the third operating state data group Principal component score plotting means for plotting the principal component score corresponding to 1 in a two-dimensional coordinate system with the two principal components as axes;
A distribution center calculation means for calculating a center of a principal component score distribution corresponding to each of the first operation state data group, the second operation state data group, and the third operation state data group in the two-dimensional coordinate system; ,
Movement vector calculation means for calculating a movement vector from the center of the principal component score distribution corresponding to the first driving state data group to the center of the principal component score distribution corresponding to the third driving state data group;
A distribution area setting means for setting an area of a principal component score distribution corresponding to the second operating state data group in the two-dimensional coordinate system;
Executed when the second processing condition is applied to the second plasma processing apparatus by moving the principal component score distribution region corresponding to the set second operating state data group according to the movement vector. Predicted distribution area setting means for setting a predicted distribution area in the two-dimensional coordinate system of principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing to be performed;
The second processing condition is actually applied to the second plasma processing apparatus, and corresponds to the two principal components of the processing parameter values measured in the plasma processing executed according to the second processing condition. A driving state determination device comprising: a principal component score inclusion determination unit that determines whether or not a principal component score is included in the set prediction distribution region. When a processing condition applied to the first plasma processing apparatus is applied to the second plasma processing apparatus, a computer is caused to execute an operation state determination method of the plasma processing apparatus that determines an operation state of the second plasma processing apparatus. A program,
First operating state data group for acquiring a first operating state data group composed of processing parameter values measured in plasma processing executed in accordance with a first processing condition applied to the first plasma processing apparatus. An acquisition module;
A second operating state data group for acquiring a second operating state data group composed of processing parameter values measured in the plasma processing executed in accordance with the second processing condition applied to the first plasma processing apparatus. An acquisition module;
Third operating state data for acquiring a third operating state data group composed of processing parameter values measured in the plasma processing executed according to the first processing condition applied to the second plasma processing apparatus. A group acquisition module;
A principal component analysis module that performs principal component analysis using at least one of the acquired first operating state data group, second operating state data group, and third operating state data group;
Two principal components selected from a plurality of principal components obtained by the principal component analysis using the first operating state data group, the second operating state data group, and the third operating state data group A principal component score plotting module for plotting the principal component scores corresponding to the above in a two-dimensional coordinate system with the two principal components as axes;
A distribution center calculation module for calculating a center of a principal component score distribution corresponding to each of the first operation state data group, the second operation state data group, and the third operation state data group in the two-dimensional coordinate system; ,
A movement vector calculation module for calculating a movement vector from the center of the principal component score distribution corresponding to the first driving state data group to the center of the principal component score distribution corresponding to the third driving state data group;
A distribution area setting module for setting an area of a principal component score distribution corresponding to the second operating state data group in the two-dimensional coordinate system;
Executed when the second processing condition is applied to the second plasma processing apparatus by moving the principal component score distribution region corresponding to the set second operating state data group according to the movement vector. A predicted distribution area setting module for setting a predicted distribution area in the two-dimensional coordinate system of principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing to be performed;
The second processing condition is actually applied to the second plasma processing apparatus, and corresponds to the two principal components of the processing parameter values measured in the plasma processing executed according to the second processing condition. A program comprising: a principal component score inclusion determination module for determining whether or not a principal component score is included in the set prediction distribution region. When a processing condition applied to the first plasma processing apparatus is applied to the second plasma processing apparatus, a computer is caused to execute an operation state determination method of the plasma processing apparatus that determines an operation state of the second plasma processing apparatus. A computer-readable storage medium for storing a program, wherein the program is
First operating state data group for acquiring a first operating state data group composed of processing parameter values measured in plasma processing executed in accordance with a first processing condition applied to the first plasma processing apparatus. An acquisition module;
A second operating state data group for acquiring a second operating state data group composed of processing parameter values measured in the plasma processing executed in accordance with the second processing condition applied to the first plasma processing apparatus. An acquisition module;
Third operating state data for acquiring a third operating state data group composed of processing parameter values measured in the plasma processing executed according to the first processing condition applied to the second plasma processing apparatus. A group acquisition module;
A principal component analysis module that performs principal component analysis using at least one of the acquired first operating state data group, second operating state data group, and third operating state data group;
Two principal components selected from a plurality of principal components obtained by the principal component analysis using the first operating state data group, the second operating state data group, and the third operating state data group A principal component score plotting module for plotting the principal component scores corresponding to the above in a two-dimensional coordinate system with the two principal components as axes;
A distribution center calculation module for calculating a center of a principal component score distribution corresponding to each of the first operation state data group, the second operation state data group, and the third operation state data group in the two-dimensional coordinate system; ,
A movement vector calculation module for calculating a movement vector from the center of the principal component score distribution corresponding to the first driving state data group to the center of the principal component score distribution corresponding to the third driving state data group;
A distribution area setting module for setting an area of a principal component score distribution corresponding to the second operating state data group in the two-dimensional coordinate system;
Executed when the second processing condition is applied to the second plasma processing apparatus by moving the principal component score distribution region corresponding to the set second operating state data group according to the movement vector. A predicted distribution area setting module for setting a predicted distribution area in the two-dimensional coordinate system of principal component scores corresponding to the two principal components of the processing parameter values measured in the plasma processing to be performed;
The second processing condition is actually applied to the second plasma processing apparatus, and corresponds to the two principal components of the processing parameter values measured in the plasma processing executed according to the second processing condition. A storage medium comprising: a principal component score inclusion determination module that determines whether or not a principal component score is included in the set prediction distribution region.

Images