• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    According to the present invention, in the processing chamber 102 of the etching apparatus 100, a plasma P fluctuating by the magnet 128 is formed. The arithmetic controller 120 samples the fluctuating plasma light signal detected by the light receiving section 136 through the detection window 134 at a predetermined sampling period to obtain a data train. Assuming a plurality of hypothesis fluctuation cycles, the process of calculating the moving average value in each hypothesis fluctuation cycle from the corresponding data row is repeated. Then, the moving average value obtained for each repetition timing is statistically processed as moving average value data for each assumed fluctuation period to obtain an approximate equation corresponding to each assumed fluctuation period. Further, the data sequence of each moving average value and the deviation amount of each approximation equation in the predetermined period are obtained, and the hypothesis variation period corresponding to the minimum deviation amount is determined as the variation period of the plasma P. Based on the determined fluctuation period, the data string of the moving average value is obtained from the data string of the sampling signal, and the end point of the processing is determined.
    公开号:KR19990063203A
    申请号:KR1019980056094
    申请日:1998-12-18
    公开日:1999-07-26
    发明作者:스스무 사이토
    申请人:히가시 데쓰로；동경엘렉트론 주식회사；
    IPC主号:
    专利说明:

    Plasma processing method
    The present invention relates to a plasma processing method.
    Conventionally, in the field of semiconductor manufacturing apparatuses, a configuration of a plasma processing apparatus using various kinds of plasma sources has been proposed. As one of such methods, an electric field is formed between the upper electrode and the lower electrode arranged in opposition to the processing chamber, the processing gas introduced into the processing chamber by the electric field is converted into plasma, and a predetermined plasma process is performed on the object to be processed There is a plasma processing apparatus capable of performing plasma processing.
    In the plasma processing apparatus, there is a so-called magnetic field assisted plasma processing apparatus having a magnet capable of forming a rotating magnetic field in the processing chamber. By adopting such a configuration, electrons in the plasma are trapped by the magnetic field formed in the processing chamber, and the number of collision with the processing gas particles is increased, whereby the high-density plasma can be excited. Further, by rotating the magnetic field, the plasma density can be made uniform, and high-speed and uniform plasma processing can be realized.
    However, in the plasma processing step such as etching, it is important to accurately determine the end point of the plasma processing and end the plasma processing without delay. As a method of detecting an end point of plasma processing, a method has been proposed in which a change in spectral light of a specific substance contained in a plasma in a conventional processing chamber is detected and the end point detection is performed based on the change. In this method, attention is paid to the fact that the components contained in the plasma are changed at the same time as the etching progresses with respect to the object to be treated, and the change in the intensity of the spectrum light of an arbitrary substance is observed to detect the end point of the etching process accurately in real time will be.
    However, in the above-described magnetic field assisted plasma processing apparatus, regions having different densities in the plasma are formed along the magnetic field direction formed in the processing chamber. Then, as the magnetic field rotates, the density distribution of the plasma also fluctuates. Therefore, in the case of observing the plasma light at the vertex from a detection window or the like provided on a wall of the processing chamber, it is necessary to take into account the variation of the plasma depending on the rotating magnetic field.
    For this reason, Japanese Patent Laid-Open No. 4-338663, for example, discloses a rotary encoder for generating a pulse in synchronization with the rotation of a magnet in an etching apparatus, sampling the plasma light in accordance with the pulse, Discloses a technique for eliminating noise components following a rotation period and performing accurate end point detection.
    However, as described above, in the configuration in which the rotation period of the magnet is determined in hardware and the sampling is performed in accordance with the rotation period, an apparatus such as a rotary encoder must be added to the processing apparatus, There is a problem that the initial cost of the apparatus is increased.
    Further, in the configuration in which the rotation period of the magnet is obtained by hardware as described above and sampling is performed in accordance with the rotation period, the sampling period is a fluctuation period which fluctuates every time. However, there is a case where the fixed period is used as the sampling period by the signal processing software for end point detection, and the case where the efficient processing can be executed and the fixed period is required. However, the above configuration has a problem that it can not flexibly cope with the demand of the software side.
    SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems of the conventional plasma processing method, and it is an object of the present invention to provide a plasma processing method and a plasma processing method, The present invention provides a new and improved plasma processing method capable of accurately obtaining a fluctuation period, that is, a rotation period of a magnet by a software method.
    It is a further object of the present invention to provide a method and apparatus for generating a similar sampling signal of a relatively narrow sampling period from a sampling signal with a relatively wide sampling period, The present invention provides a new and improved plasma processing method capable of obtaining a more accurate moving average value by increasing the number of actual sampling times by a small amount of computation.
    1 is a schematic sectional view showing an etching apparatus to which the present invention can be applied,
    FIG. 2 is a schematic explanatory view for explaining a rotation period determination process applied to the etching apparatus shown in FIG. 1;
    3 is a schematic explanatory view for explaining a rotation period determination process applied to the etching apparatus shown in FIG. 1;
    FIG. 4 is a schematic explanatory view for explaining a rotation period determination process applied to the etching apparatus shown in FIG. 1;
    5 is a schematic explanatory view for explaining an end point determination process applied to the etching apparatus shown in Fig.
    DESCRIPTION OF THE REFERENCE NUMERALS
    100: etching apparatus 102: processing chamber
    104: Vacuum container 106: Lower electrode
    108: upper electrode 116: matching device
    118: High frequency power supply 120: Operation controller
    124: gas supply pipe 126: exhaust pipe
    128: Magnet 134: Detection window
    136:
    According to a first aspect of the present invention, there is provided a plasma processing method comprising: forming an electric field in a vacuum container into which a process gas is introduced to generate a plasma that fluctuates at a predetermined fluctuation period; A plasma processing method comprising:
    When the fluctuation period of the plasma is found,
    (a) sampling the plasma light of the plasma at a predetermined sampling period to obtain sampling data;
    (b) calculating a moving average value over a period of each of the plurality of assumed fluctuation periods based on the sampling data, assuming a plurality of assumed fluctuation periods, and obtaining moving average value data for each of the assumed fluctuation periods and,
    (c) obtaining each approximate expression corresponding to each hypothesis variation cycle from moving average value data for each hypothesis variation cycle;
    (d) obtaining a deviation amount of the approximate equation corresponding to the moving average value data at one or two or more points of time for each of the hypothetical variation periods;
    (e) determining the hypothetical variation period having the smallest deviation amount among the deviation amounts, and determining the hypothetical variation period as the variation period of the plasma
    Is provided.
    When the end point of the plasma process is determined based on the obtained fluctuation period of the plasma,
    (f) calculating moving average value data for a period of the fluctuation period of the plasma obtained in the step (e) from the sampling data;
    (g) a step of determining an end point of the plasma process based on the moving average value data obtained in the step (f) is executed.
    With this configuration, even when sampling of the plasma light is performed at a constant sampling period, the fluctuation period of the plasma, which is likely to change for each process, can be calculated only by software statistical calculation processing. Therefore, even if a special hardware device is not added, the fluctuation period of the plasma can be obtained accurately for each process.
    According to a second aspect of the present invention, there is provided a plasma processing apparatus for forming an electric field in a vacuum container into which a process gas is introduced to generate a plasma that fluctuates at a predetermined variation period, The plasma processing method comprising the steps of:
    When detecting the end point of the process,
    (a) sampling the plasma light of the plasma at a predetermined sampling period to obtain sampling data (the fluctuation period is not limited to an integer multiple of the sampling period)
    (b) calculating a moving average value for each of the fluctuation periods from the sampling data;
    (c) a step of determining the end point of the plasma process based on the change of the moving average value is executed.
    According to such a configuration, even if a special hardware device is not added, it is possible to accurately obtain the end point of the plasma processing by compensating the variation of the fluctuation period of the different plasma for each process only by software processing.
    These and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description, which is given with reference to the accompanying drawings.
    Example
    Hereinafter, preferred embodiments in which the plasma processing method according to the present invention is applied to the end point determination method of etching processing will be described in detail with reference to the accompanying drawings.
    First, an apparatus configuration of an etching apparatus 100 to which such an end point determination method can be applied will be described with reference to FIG.
    A treatment chamber 102 of the etching apparatus 100 is formed in a conductive vacuum vessel 104. In the processing chamber 102, a conductive lower electrode 106 constituting a susceptor and a grounded conductive upper electrode 108 are disposed opposite to each other. A semiconductor wafer (hereinafter, referred to as a " wafer ") W can be mounted on the lower electrode 106 and fixed thereon during processing. A high-frequency power source 118 is connected to the lower electrode 106 via a matching unit 116. High-frequency power of a predetermined output can be applied according to a control signal from the arithmetic controller 120 during processing.
    A gas supply pipe 124 for supplying a predetermined process gas into the process chamber 102 and an exhaust pipe 126 for exhausting the atmosphere in the process chamber 102 are connected to the process chamber 102. A magnet 128 such as a permanent magnet capable of forming a predetermined rotating magnetic field is disposed in the processing chamber 102 above the vacuum container 104.
    A detection window 134 is formed on the side wall of the processing chamber 102. Plasma light transmitted through the detection window 134 is detected by the light receiving section 136. [ The light-receiving unit 136 includes a spectroscope (not shown) for spectrally separating only the emission spectrum of a specific material from the plasma light, a photoelectric converter (not shown) for detecting the emission intensity of the emission spectrum, / D converter and the like. The sampled data in the light receiving section 136 is appropriately outputted to the arithmetic controller 120. In the illustrated example, the light receiving section 136 is configured to directly detect the plasma light. However, the configuration in which the light transmitting section such as optical fiber is interposed between the detection window 134 and the light receiving section 136 .
    Next, a case where the etching process is performed on the wafer W in the etching apparatus 100 will be described. The wafer W is mounted on the lower electrode 106 to introduce the process gas into the process chamber 102 from the gas supply pipe 124 and exhaust the atmosphere from the process chamber 102 through the exhaust pipe 126, 102 are maintained in a predetermined reduced pressure atmosphere. Subsequently, the magnet 128 above the vacuum chamber 104 is rotated at a predetermined number of revolutions to form a rotating magnetic field in the processing chamber 102.
    Then, when high-frequency power is applied to the lower electrode 106 from the high-frequency power source 118, the process gas is dissociated by the electric field formed in the process chamber 102 to generate a high-density plasma P, . Further, by the rotating magnetic field formed in the processing chamber 102, the plasma P fluctuates at a predetermined rotation cycle, and the non-uniform density is averaged, and the wafer W can be uniformly processed.
    Next, a method of determining the end point of the etching process according to the present embodiment, which is applied to the etching apparatus 100 described above, will be described with reference to Figs. 2 to 5. Fig. The end point determination method includes a plasma fluctuation period determination step and an etching end point determination step.
    The fluctuation period determining step is a step of determining a fluctuation period of the magnetic field that changes in accordance with the rotation of the magnet 128, that is, a fluctuation period of the plasma. The end point determination step is a step of performing accurate end point determination after removing a noise component in a sampling signal generated according to a fluctuation of the plasma based on the determined fluctuation period of the plasma.
    Hereinafter, the fluctuation period determination step (A) and the end point determination step (B) will be described separately.
    (A) Variable Period Determination Step
    First, the fluctuation period determining step will be described with reference to Figs. 2 to 4. Fig. 2 shows a time-series variation of the sampling signal P (t) sampled in the light receiving unit 136 of the etching apparatus 100. As shown in FIG. As shown in the figure, the sampling signal also fluctuates unstably at the time of plasma rising, but the sampling signal is stabilized at the same time as the plasma is stabilized. Thus, the fluctuation period of the plasma is obtained based on the sampling signal in the period A after the sampling signal becomes stable (in the example shown, time a after the time a).
    FIG. 3 is an enlarged view of the transition of the sampling signal over the period A of FIG. In the illustrated example, it is assumed that sampling is performed at intervals of 0.1 second. Here, since the period of fluctuation of the magnetic field in the treatment chamber 102 is substantially interlocked with the period of rotation of the magnet 128, the approximate period can be estimated from the global. For example, in the method according to the present embodiment, it can be assumed that the plasma fluctuates at a period of 3.0 to 3.1 seconds according to the rotation period of the magnet 128.
    Therefore, when it is desired to obtain an accurate fluctuation period of the plasma, which is assumed to be in the period of 3.0 to 3.1 seconds, to be more precisely, for example, on the order of 0.01 second, it is necessary to perform sampling at 0.01 second have. However, simply shortening the sampling period has a problem that the amount of sampling data increases unnecessarily and time is required for data processing. In this regard, according to the method of this embodiment, it is possible to obtain the fluctuation period of the plasma with the accuracy equivalent to that in the case where the sampling is performed at short intervals of 0.01 second or more even by sampling at intervals of 0.1 second It becomes possible. Hereinafter, the method will be described in detail.
    First, it is assumed that the plasma fluctuates at a cycle of 3.0 seconds. Then, similarly to the conventional method, the moving average value H 3.00 (t) of the sampling data is obtained on the basis of the following expression (1) in order to remove the noise component contained in the sampling data. In this case, the moving average calculation period for a 3.0 second assumption cycle is 3.0 seconds.
    Then, it is assumed that the plasma fluctuates at a cycle of 3.1 seconds. Then, similarly to the previous processing, the moving average value H 3.10 (t) of the sampling data is obtained on the basis of the following expression (2) in order to remove the noise component contained in the sampling data. In this case, the moving average calculation period for a 3.1 second assumption period is 3.1 seconds.
    Here, in the above case, for a sampling period of 0.1 second, the assumed variation period is set to 3.0 seconds and 3.1 seconds which are integral multiples of the sampling period, respectively. However, the fluctuation period of the actual plasma is not limited to an integral multiple of the sampling period. For example, assuming that the fluctuation period of the actual plasma is 3.03 seconds, the moving average calculation period is 3.03 seconds, and measurement is impossible with a sampling period of 0.1 second. If the moving average is obtained, the sampling period is executed in units of 0.01 second There is a need. However, by finely setting the sampling period in this manner, the amount of data increases unnecessarily, and the processing efficiency can not be improved.
    In this respect, in this embodiment, data sampled at a sampling period of 0.1 second is sampled at a sampling period of 0.01 second by the procedure described below, Lt; RTI ID = 0.0 > a < / RTI >
    For example, the moving average value H 3.03 (t) at time t when the plasma is assumed to fluctuate at a cycle of 3.03 seconds is expressed by
    + S (t)) 10 + S (t-2.9-0.01) + S (t-2.9-0.02)
    + S (t-2.9-0.03)} / 303
    (3) " (3) "
    The data indicated by S (t-2.9-0.01), S (t-2.9-0.02) and S (t-2.9-0.03) in the equation (3) are not actually sampled data but are actually sampled data Since it is software-wise data by numerical processing, it is referred to as pseudo-sampling data in the present specification. These pseudo sampling data are obtained by connecting the data of the adjacent sampling signals by a straight line or an approximate curve and calculating the time t-2.9-0.01, the time t-2.9-0.02, the time t- Can be obtained as the corresponding value at the pseudo sampling timing such as 2.9-0.03. As described above, in the present specification, the assumed time interval for obtaining pseudo-sampling data is called pseudo-sampling timing.
    In order to obtain an approximate curve connecting data of adjacent sampling signals, a method of approximating a first-order or higher-order polynomial by a least squares method, a method of using a Lagrangian interpolation method, a spline interpolation method, Can be obtained from the point data.
    Further, for example, when the data of the adjacent sampling signals are connected by a straight line, S (t-2.9-0.01)
    , And S (t-2.9-0.02) can be obtained by the following equation
    , And S (t-2.9-0.03) can be obtained by the following equation
    (6) " (6) "
    As described above, in the fluctuation period determining step according to the present embodiment, the moving average value H (t), the pseudo sampling data S (t-2.9-0.01) 4 to (6). With this configuration, even a moving average value in a period that is not an integer multiple of the sampling period, that is, a moving average value in a period shorter than the actual sampling period, can be obtained from less sampling data.
    For example, when the rotation period of the magnet 128 is assumed to be 3.03 seconds as described above, 11 sampling data S (t-3.0), ..., ... , S (t), the cycle of fluctuation of the plasma in the order of 1/100 second can be obtained. According to the present embodiment, the moving average value of 303 pieces of data including the pseudo-sampling data can be calculated by a small amount of calculation processing as a result of the weighted average method.
    By the same method, the moving average H (t) at time t when the rotation period of the magnet 128 is assumed to be 3.01 seconds to 3.09 seconds can be obtained.
    3.30 seconds, 3.02 seconds, 3.03 seconds, 3.05 seconds, 3.06 seconds, 3.07 seconds, 3.08 seconds, 3.05 seconds, 3.05 seconds, 3.09 seconds, and 3.10 seconds in parallel, that is, at the same time, the moving average value for each moving average value calculating period. In this way, each moving average value, H 3.00 (t), H 3.01 (t), H 3.02 (t), H 3.03 (t), H 3.04 (t), H 3.05 (t), H 3.06 (t), H 3.07 (t), H 3.08 (t), H 3.09 (t) and H 3.10 (t).
    Further, by executing a plurality of times of sampling, it is possible to obtain a data string of each moving average value H 3.00 (t) to H 3.10 (t) corresponding to each sampling time. In FIG. 3, among these moving average values, plotting H 3.03 (t) over a predetermined period is exemplified as H (t).
    As described above, the method of obtaining the moving average value at a precision of ten times the actual sampling period has been described. However, the present embodiment is not limited to such a structure. For example, the moving average value may be set to be 100 times the real sampling period Can be obtained. That is, for example, the moving average value H 3.031 (t) at time t when the fluctuation period of the plasma is assumed to be 3.031 seconds,
    ... + S (t-0.1) + S (t)) 100 + (S (t-2.9-0.01)
    + S (t-2.9-0.02) + S (t-2.9-0.03)) x10
    + S (t-2.9-0.03-0.001)} / 3031
    Can be obtained by the following equation (7).
    (T-2.9-0.01), S (t-2.9-0.02), and S (t-2.9-0.03) in Equation 7 correspond to Equation 4, Equation 5, Can be obtained by the following equation (6). Also, S (t-2.9-0.03-0.001) in the same equation is S
    Can be obtained by the following equation (8).
    Next, with reference to FIG. 4, a step of obtaining an approximate expression, for example, a first approximate expression, from the data string of each moving average value determined by the above-described steps is described. Fig. 4 is an enlarged view of H (t) shown in Fig. In addition, K (t) shown in FIG. 4 is a first order approximation expression obtained by, for example, a least squares method from the value of H (t) which is a predetermined period.
    Here, in the case of the present embodiment, since it is known in advance that the fluctuation period of the plasma, that is, the rotation period of the magnet 128 is in the range of 3.0 seconds to 3.1 seconds, the predetermined period is set to 3.1 seconds. Then, a first approximate expression corresponding to each of the data strings of H 3.00 (t) to H 3.10 (t) described above is obtained, and each of the first approximate expressions is set to K 3.00 (t) to K 3.10 (t) corresponding to each.
    Next, a description will be given of a step of obtaining an average of the absolute values of the data sequences of the respective moving average values obtained by the above-described steps and the deviation amounts of the respective linear approximate expressions corresponding to the data strings of the respective moving average values. The average A of the absolute values of such deviation amounts is expressed by
    + | H (t-2.8) -K (t-2.8) | + ... ... + | H (t-0.1) -K (t-0.1) |
    + | H (t) -K (t) |} / 31
    Can be obtained by the following equation (9). The time t in Equation (9) represents a time different from the time t in Equations (1) to (8) used for obtaining the moving average value.
    The above equation (9) can be used as a mathematical expression for obtaining the absolute value of the deviation amount, that is, a period for obtaining the moving average value, that is, whichever of the rotation period of the supposed magnet 128 is between 3.00 and 3.10 seconds.
    In addition, in order to increase the accuracy of obtaining the fluctuation period, it is preferable to perform a comparison between the data string of the moving average value and the deviation amount of the approximate expression over at least the period of the expected fluctuation period or more.
    In this way, by the equation (9), moving average value H 3.00 (t) with respect to the ~H 3.10 (t) data columns, their primary approximate formula K 3.00 (t) corresponding to each data column ~K 3.10 (t), respectively , The average of the absolute values of the deviation amounts is obtained and is set to A 3.00 to A 3.10 , respectively.
    Further, as long as mubang indicating the degree of displacement of the piece to obtain the average A of the absolute value of the expression vehicle, H 3.00 (t) ~H 3.10 (t) of the data string, K 3.00 (t) ~K 3.10 (t) , It is not limited to the above-mentioned Equation (9). The time to be the target of the total average A may not be common to all of A 3.00 to A 3.10 as described above. For example, when A 3.03 is obtained, time t-2.9, t-2.8, t -2.7, ... ... , the data values at t-0.1, t, t-2.9-0.01, t-2.9-0.02 and t-2.9-0.03, that is, at the time of the calculation of the moving average value .
    Next, the process of calculating the fluctuation period of plasma, that is, the rotation period of the magnet 128, from the average of the absolute values of the deviation amounts obtained by the above-described process will be described.
    A minimum value is found out of the average A 3.00 to A 3.10 of the absolute value of the deviation amount obtained by the above equation (9), and the assumed variation period for the value is set as the variation period T of the actual plasma. That is, when A 3.03 is the minimum value, the fluctuation cycle T of the plasma is nearest to 3.03 seconds. For this reason, the following are considered.
    That is, in the period A of FIG. 2, the state of the plasma P in the processing chamber 102 becomes relatively stable, and the original waveform based on the plasma light not including the rotation fluctuation becomes a relatively smooth curve. In addition, since the period during which calculation is performed is minute, the original waveform can approximate a straight line. Further, when the target period for calculating the moving average value completely coincides with the rotation period of the magnet 128, the moving average value coincides with the original waveform not including the rotation fluctuation. Therefore, from the point that the displacement (deviation) of the data sequence of the moving average value obtained in the period close to the fluctuation period is smaller than that of the first approximation formula obtained from the data sequence, the minimum value obtained by averaging the absolute values of the above- As shown in FIG.
    The fluctuation period determining step according to the present embodiment is configured as described above. Even when the fluctuating plasma light is sampled at regular intervals, the fluctuation period of the plasma corresponding to the turn period of the magnets 128 can be accurately determined. In addition, since the weighted average is used by the above-described calculation, it is possible to obtain a predetermined moving average value necessary for determination of the fluctuation period of the plasma without increasing the number of sampling data and by relatively small amount of calculation processing.
    (B) End point determination step
    Next, the end point determination process of the plasma process will be described with reference to Fig. In the end point determination step, the period of fluctuation of the plasma is obtained and the signal P (t) of the plasma light is sampled based on the rotation period of the magnet 128, And storing the data string and determining the end point of the etching process by analyzing the change of the data string of the moving average value by software, for example.
    FIG. 5 shows a data sequence of moving average values,
    + S (t-0.1) x (m-1) -0.01) + ... ... + S (t-0.1 占 (m-1) -0.01 占 n)} / (10 占 m + n)
    (10). &Quot; (10) " Equation (10) is the same as Equation (3). Further, m in Equation (10) is a share of (T / sampling period), n is ten times the remainder of (T / sampling period), and 0.1 is a sampling period.
    S (t-0.1 × (m-1) -0.01) to S (t-0.1 × (m-1) -0.01 × n) which are pseudo sampling data in Equation (10) . Therefore, by calculating using Equation (10), it is possible to obtain the moving average value of the period in which the sampling period is not an integral multiple of the sampling period, as shown in Equations (4) to (6) The moving average value of a plurality of data can be calculated by a relatively small amount of calculation processing. Further, when the moving average value is calculated, more accurate end point determination can be performed by weighting the sampling data with respect to the pseudo-sampling data.
    Further, the end point determination process of this etching process is not limited to the case of being performed in combination with the above-described fluctuation period determination process. For example, even when the rotation period of the magnet 128 is known in advance, when the rotation period is not an integer multiple of the sampling interval, the moving average value is calculated by the same method as that of the end point determination process, It is possible to output a signal of plasma light that is not affected by the rotation fluctuation of the magnet 128. [ The determination of the end point of the etching process is performed by a commonly used method, for example, by differentiating the data sequence of the moving average value and determining that the differential value is the end point of the etching process if the differential value is equal to or larger than the predetermined value.
    The end point determination process according to the present embodiment is configured as described above. Based on the rotation period of the correct magnet 128 determined in the rotation period determination process described above, Since the end point of the process is determined by using the obtained signal, accurate end point determination can be performed, and the etching process can be accurately and reliably terminated.
    The preferred embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to such a configuration. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims and their equivalents .
    For example, in the above embodiment, the method of determining the fluctuation period of the fluctuation of the signal of the plasma light generated when the magnet 128 made of the permanent magnet is rotated is described, but the present invention is not limited to this configuration Alternatively, the present invention can be applied to the case where an electromagnet is used instead of the permanent magnet, or when the magnet is reciprocated instead of being rotated. Also, the present invention can be applied to a case where a plasma fluctuates even when a magnet is not used, for example, when a position where an electric field is applied on an electrode is sequentially and periodically changed.
    According to the present invention, when the plasma fluctuates, the fluctuation cycle of the plasma or the end point of the plasma treatment can be accurately determined only by software-based calculation processing, so that there is no need to form hardware for executing the determination in the processing apparatus . As a result, the apparatus configuration of the processing apparatus can be facilitated, and the present invention can be easily implemented in existing apparatuses. Further, in the present invention, since the fluctuation period and the end point of the process can be determined based on the sampling signal sampled by the sampling period of the constant period, it is preferable to use the fixed period as the sampling period It can be applied to a case where it is good or a case where it is necessary to use a fixed period.
    权利要求:
    Claims (13)
    [1" claim-type="Currently amended] A plasma processing method for forming an electric field in a vacuum container into which a process gas is introduced to generate a plasma which fluctuates at a predetermined fluctuation period and subjecting the object to be processed placed in the vacuum container to a plasma process,
    (a) sampling the plasma light of the plasma at a predetermined sampling period to obtain sampling data;
    (b) calculating a moving average value over a period of each of the plurality of assumed fluctuation periods based on the sampling data, assuming a plurality of assumed fluctuation periods, and obtaining moving average value data for each of the assumed fluctuation periods ,
    (c) obtaining each approximate expression corresponding to each hypothesis variation cycle from moving average value data for each hypothesis variation cycle;
    (d) obtaining a deviation amount of the approximate equation corresponding to the moving average value data at one or two or more points of time for each of the hypothetical variation periods;
    (e) determining the hypothetical variation period having the smallest deviation amount among the deviation amounts, and determining the hypothetical variation period as the variation period of the plasma
    The plasma processing method comprising the steps of:
    [2" claim-type="Currently amended] The method according to claim 1,
    (f) calculating moving average value data for a period of the fluctuation period of the plasma obtained in the step (e) from the sampling data;
    (g) determining an end point of the plasma process based on the moving average value data obtained in the step (f).
    [3" claim-type="Currently amended] The method according to claim 1,
    Wherein the maximum value of the hypothetical variation period is an upper limit value of the expected variation period and the minimum value of the hypothetical variation period is a lower limit value of the expected variation period.
    [4" claim-type="Currently amended] The method according to claim 1,
    Wherein the hypothesis fluctuation period is an integral multiple of the sampling period.
    [5" claim-type="Currently amended] The method according to claim 1,
    Wherein the hypothesis fluctuation period is not an integer multiple of the sampling period.
    [6" claim-type="Currently amended] 6. The method of claim 5,
    Wherein the step (b) comprises numerically processing the sampling data to obtain similar sampling data corresponding to the hypothesis fluctuation period, and calculating the moving average value based on the sampling data and the pseudo sampling data. Processing method.
    [7" claim-type="Currently amended] The method according to claim 6,
    Wherein in the step (b), the weighting is performed on the sampling data and the pseudo-sampling data when the moving average value is calculated.
    [8" claim-type="Currently amended] The method according to claim 1,
    Wherein in the step (e), a deviation amount of the approximate equation corresponding to the moving average value data is obtained for a period longer than an expected upper limit value of the variation period for each hypothetical variation period .
    [9" claim-type="Currently amended] The method according to claim 1,
    Wherein a rotating magnetic field for changing the plasma at a predetermined rotation cycle is formed in the vacuum chamber.
    [10" claim-type="Currently amended] A plasma processing method for forming an electric field in a vacuum container into which a process gas is introduced to generate a plasma which fluctuates at a predetermined fluctuation period and subjecting the object to be processed placed in the vacuum container to a plasma process,
    (a) sampling the plasma light of the plasma at a predetermined sampling period to obtain sampling data (the fluctuation period is not limited to an integer multiple of the sampling period)
    (b) calculating a moving average value for each of the fluctuation periods from the sampling data;
    (c) determining an end point of the plasma process based on the change in the moving average value.
    [11" claim-type="Currently amended] 11. The method of claim 10,
    Wherein the step (b) includes numerically processing the sampling data to obtain similar sampling data corresponding to the fluctuation period, and calculating the moving average value based on the sampling data and the pseudo-sampling data. Way.
    [12" claim-type="Currently amended] 12. The method of claim 11,
    Wherein in the step (b), the weighting is performed on the sampling data and the pseudo-sampling data when the moving average value is calculated.
    [13" claim-type="Currently amended] 11. The method of claim 10,
    Wherein a rotating magnetic field for changing the plasma at a predetermined rotation cycle is formed in the vacuum chamber.
    类似技术:
    公开号 | 公开日 | 专利标题
    EP1593141B1|2007-12-19|End point detection in time division multiplexed etch processes
    US6828165B2|2004-12-07|Semiconductor plasma processing apparatus with first and second processing state monitoring units
    KR100912478B1|2009-08-17|Etching processing apparatus and etching processing method
    US5308414A|1994-05-03|Method and apparatus for optical emission end point detection in plasma etching processes
    KR100560886B1|2006-03-13|System and method for monitoring and controlling gas plasma processes
    US6771481B2|2004-08-03|Plasma processing apparatus for processing semiconductor wafer using plasma
    US6745133B2|2004-06-01|Mass spectral peak identification
    JP5366413B2|2013-12-11|Dose measurement using emission spectroscopy / residual gas analyzer related to ion current
    US6665576B2|2003-12-16|Method and system for managing semiconductor manufacturing equipment
    Malyshev et al.1999|Trace rare gases optical emission spectroscopy: nonintrusive method for measuring electron temperatures in low-pressure, low-temperature plasmas
    US5928532A|1999-07-27|Method of detecting end point of plasma processing and apparatus for the same
    KR960004555B1|1996-04-09|System and method for detecting partial discharge of gas-insulated switch gear
    CN101490816B|2011-06-22|Plasma abnormal discharge diagnosis method, plasma abnormal discharge diagnosis system
    US6680476B1|2004-01-20|Summed time-of-flight mass spectrometry utilizing thresholding to reduce noise
    KR100798648B1|2008-01-28|Method and apparatus for monitoring a process by emploing principal component analysis
    US6985215B2|2006-01-10|Plasma processing method and plasma processing apparatus
    JP3766991B2|2006-04-19|End point detection method and apparatus for plasma processing, and semiconductor manufacturing method and apparatus using the detection method and apparatus
    JP3429137B2|2003-07-22|Method for real-time in-situ monitoring of trench formation process
    US6881352B2|2005-04-19|Disturbance-free, recipe-controlled plasma processing method
    KR100748288B1|2007-08-09|Film thickness control using spectral interferometry
    US4289188A|1981-09-15|Method and apparatus for monitoring etching
    JP5583603B2|2014-09-03|Arc detection system and method
    US6448094B2|2002-09-10|Method of detecting etching depth
    US6847037B2|2005-01-25|Ion trap mass spectrometer
    EP0535700B1|1997-03-26|Method and apparatus for comparing spectra
    同族专利:
    公开号 | 公开日
    JPH11186239A|1999-07-09|
    US6231774B1|2001-05-15|
    KR100423195B1|2004-05-17|
    JP3563949B2|2004-09-08|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    1997-12-19|Priority to JP36522297A
    1997-12-19|Priority to JP97-365222
    1998-12-18|Application filed by 히가시 데쓰로, 동경엘렉트론 주식회사
    1999-07-26|Publication of KR19990063203A
    2004-05-17|Application granted
    2004-05-17|Publication of KR100423195B1
    优先权:
    申请号 | 申请日 | 专利标题
    JP36522297A|JP3563949B2|1997-12-19|1997-12-19|Plasma processing method|
    JP97-365222|1997-12-19|
    [返回顶部]