JP2018053297A - Control method for plasma generation device, plasma generation device and film deposition device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a technique allowing for stable generation of uniform plasma by allowing for proper control of the amount of gas supplied to a plasma space.SOLUTION: A plurality of pieces of gas ejection means 63a-63c and 64a-64c eject gas as plasma seeds toward partial regions Za, Zb and Zc differing from one another in a plasma space PL where plasma is generated. A plurality of detecting means 191a, 191b and 191c each detect plasma light emission intensity in a partial region to which one corresponding gas ejection means ejects gas, and control means 190 sets the ejection amounts of gases of the plurality of gas ejection means based upon a predetermined multi-variable function including variables of respective detection results of the plurality of detecting means. The multi-variable function is found through multi-variable analysis which includes plasma light emission intensities that the respective detecting means detect as target variables, and includes ejection amounts of gases that the gas ejection means eject respectively as explanatory variables.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a plasma generator, and more particularly to a technique for controlling the amount of gas supplied to a plasma space.

  As a method of forming a thin film such as aluminum oxide on the substrate surface, there is a method using a plasma sputtering technique. For example, in the sputtering apparatus described in Patent Document 1 previously disclosed by the applicant of the present application, the target is sputtered by generating plasma in the vicinity of the surface of the target extending elongated with one axial direction as the longitudinal direction, and orthogonal to the axial direction. An apparatus for forming a thin film containing a target material on the surface of a substrate that is scanned and moved relative to the direction.

  In an apparatus using plasma, in order to stably generate a uniform plasma in a space region (plasma space) where plasma is generated, the density of gas as a plasma species supplied to the plasma space should be appropriately controlled. Is required. In particular, when a plasma space elongated in the axial direction is formed as in the above-described apparatus, the uniformity of gas density in the axial direction can be a problem.

  In order to cope with this problem, in the technique described in Patent Document 1, a nozzle that supplies oxygen gas or nitrogen gas as a reactive gas to the plasma space is divided into a plurality of portions in the longitudinal direction of the plasma space. Yes. A plurality of probes for detecting plasma emission intensity are provided corresponding to each of the divided nozzles. Based on the detection result of each probe, the gas supply amount from each nozzle is controlled. That is, in this technique, the plasma space is divided into a plurality of virtual zones in the longitudinal direction, and the gas supply amount is adjusted for each zone.

JP2015-193863A

  The plasma space is a continuous space, and the zones as described above are not physically clearly defined in the apparatus. For this reason, if the gas supply amount changes in one zone, the influence may reach the plasma state in the other zone. Therefore, in order to generate uniform plasma in the plasma space, it is necessary to control the gas supply amount in consideration of such interaction between zones. However, Patent Document 1 does not describe in detail such problems and specific solutions for solving them.

  The present invention has been made in view of the above problems, and an object thereof is to provide a technique capable of stably generating uniform plasma by appropriately controlling the amount of gas supplied to the plasma space.

  According to one aspect of the present invention, a plurality of gas discharge means discharge gases that become plasma seeds toward different partial areas of a plasma space in which plasma is generated, and the discharge amount of the gas is the gas volume. In the control method of the plasma generator that can be set for each discharge means, in order to achieve the above object, each of the plurality of detection means provided corresponding to each of the plurality of gas discharge means corresponds to the corresponding one of the above-mentioned ones. The plasma emission intensity is detected in the partial region where the gas is discharged by the gas discharge means, and the control means is based on a predetermined multivariable function having a detection result of each of the plurality of detection means as a variable. The gas discharge amount of each of a plurality of gas discharge means is set, and the multivariable function uses the plasma emission intensity detected by each of the detection means as a target variable, and the gas Each detecting means is determined by multivariate analysis as explanatory variables the discharge amount of the gas to be discharged.

  According to another aspect of the present invention, a plasma generating means for generating plasma and a gas serving as a plasma species are discharged toward partial areas different from each other in the plasma space in which the plasma is generated. A plurality of gas discharge means capable of independently changing the discharge amount of the reactive gas, and a plurality of gas discharge means corresponding to each of the plurality of gas discharge means are provided, and the gas is discharged by one corresponding gas discharge means. The gas of each of the plurality of gas discharge means is based on a detection means for detecting the plasma emission intensity in the partial region and a predetermined multivariable function having the detection results of the plurality of detection means as variables. Control means for setting the discharge amount of each of the gas discharge means, wherein the multi-variable function uses the plasma emission intensity detected by each of the detection means as a target variable. Determined by multivariate analysis as explanatory variables the discharge amount of the gas to be discharged, a plasma generator.

  In the invention thus configured, a plurality of detection means are provided corresponding to each of the plurality of gas discharge means for discharging gas into the plasma space. Each of the gas discharge means discharges gas to different partial areas in the plasma space. Each of the detection means detects the plasma emission intensity in a partial region in the plasma space where the gas discharge means corresponding to the detection means discharges the gas. That is, the plasma space is virtually divided into a plurality of zones, and a pair of gas discharge means and detection means is provided in each zone.

  Here, when the gas discharge amount from one gas discharge means is controlled based only on the detection result of the corresponding one detection means, the detection means is affected by the gas discharged from the other gas discharge means. And the influence which the said gas discharge means has on the detection result of the other detection means is not taken into consideration. For this reason, the control capability to each plasma space is inferior, and as a result, problems such as process fluctuations not converging or taking a long time to converge may occur.

  In the present invention, the gas discharge amount from each gas discharge means is set based on a multi-variable function having the detection results of each of the plurality of detection means as variables. The multivariable function is obtained in advance by multivariate analysis in which the plasma emission intensity detected by each of the detection means is an objective variable and the discharge amount of the gas discharged by each gas discharge means is an explanatory variable. For this reason, the gas discharge amounts from the plurality of gas discharge means are set in consideration of the interaction with other gas discharge means and detection means. As a result, in the present invention, instability caused by the interaction between the plurality of gas ejection means and the plurality of detection means is eliminated, and the uniform plasma is stabilized by appropriately controlling the gas supply amount to the plasma space. Can be generated.

  According to still another aspect of the present invention, there is provided a target holding means for holding a target in a processing chamber, a plasma generator configured as described above that generates plasma near the surface of the target in the processing chamber, and the plasma generation The film forming apparatus includes substrate holding means for holding the substrate so that the apparatus faces a plasma space in which plasma is generated. In the invention configured as described above, sputtering film formation using plasma uniformly controlled by the control method described above is possible, and a uniform film can be formed on the substrate.

  As described above, according to the present invention, since the multivariable function for setting the gas discharge amount of the plurality of gas discharge means from the detection results of the plurality of detection means is determined based on the result of the multivariate analysis, the plasma space It is possible to stably generate a uniform plasma by appropriately controlling the gas supply amount.

1 is a side view showing a schematic configuration of an embodiment of a film forming apparatus according to the present invention. It is a perspective view which shows the three-dimensional structure of a sputtering source. It is a figure which shows the control system for adjusting the density of reactive gas. It is a flowchart which shows the film-forming process in this film-forming apparatus. It is a figure showing the change of a probe output when a variation is added to a gas flow rate from a steady state. 10 is a flowchart illustrating a processing example for specifying a matrix G. It is a flowchart which shows the example of the gas flow control during film-forming. It is a figure which shows the 1st modification of gas flow control. It is a figure which shows the 2nd modification of gas flow control.

  FIG. 1 is a side view showing a schematic configuration of an embodiment of a film forming apparatus according to the present invention. In order to uniformly indicate directions in the following description, XYZ orthogonal coordinate axes are set as shown in FIG. The XY plane represents a horizontal plane. Further, the Z axis represents the vertical axis, and more specifically, the (−Z) direction represents the vertical downward direction.

The film forming apparatus 1 is an apparatus that forms a film on the surface of a substrate S to be processed by reactive sputtering. For example, the film forming apparatus 1 can be applied for the purpose of forming a metal oxide film such as aluminum oxide (Al ₂ O ₃ ) on a glass substrate as the substrate S. Here, a case where film formation is performed on a rectangular or single-wafer type substrate S will be described as an example, but the substrate S may have an arbitrary shape.

  The film forming apparatus 1 includes a vacuum chamber 10, a transport mechanism 3 and a sputter source 5 for transporting a substrate S disposed therein, and a control unit 19 that performs overall control of the film forming apparatus 1. The vacuum chamber 10 is a hollow box-shaped member having a substantially rectangular parallelepiped outer shape, and is arranged so that the upper surface of the bottom plate is in a horizontal posture. The vacuum chamber 10 is composed mainly of a metal such as stainless steel or aluminum, but a transparent window made of, for example, quartz glass may be partially provided in order to make the inside of the chamber visible.

  The transport mechanism 3 includes a carrier 31 that holds the substrate S with its lower surface open, a plurality of transport rollers 32 that contact the lower surface of the carrier 31 and support the carrier 31, and a carrier roller 32 by rotating the transport roller 32. And a transport driving unit (not shown) for moving the head 31 in the X direction. The transport driving unit is controlled by the control unit 19. The carrier 31 holds the substrate S with its lower surface exposed by sucking and holding the upper surface of the substrate S.

  The transport mechanism 3 configured as described above transports the substrate S in the vacuum chamber 10 while maintaining the horizontal posture, and moves the substrate S in the X direction. The movement of the substrate S by the transport mechanism 3 may be a reciprocating movement as indicated by a dotted arrow in FIG. 1, or may be either the (+ X) direction or the (−X) direction.

  A sputtering source 5 is provided below the substrate S that is transported in the vacuum chamber 10 by the transport mechanism 3. The sputter source 5 includes a cathode plate 51, a magnet unit 53 provided below the cathode plate 51, and a pair of inductively coupled antennas 56 and 57 for generating a high frequency electric field in the vacuum chamber 10. .

  The cathode plate 51 and the magnet unit 53 together constitute a magnetron type cathode. A flat target T is placed on the upper surface of the cathode plate 51. The target T includes all or a part of the material of the film formed on the substrate S. In an aspect in which a part of the film material is supplied as a reactive gas, the target material needs to be included in the target material. Not necessarily. When the film to be formed is aluminum oxide, for example, an aluminum plate can be used as a target. When the target material is conductive, the base member can be omitted.

  The magnet unit 53 disposed below the cathode plate 51 includes a yoke 531 formed of a magnetic material such as magnetically permeable steel, a plurality of magnets provided on the yoke 531, that is, a central magnet 532 and surrounding the magnet. And a peripheral magnet 533 provided. The yoke 531 is a flat plate member extending in the Y direction. The yoke 531 is fixed to the bottom surface of the vacuum chamber 10 by a fixing member (not shown).

  A central magnet 532 extending in the Y direction is disposed on the center line along the longitudinal direction (Y direction) on the upper surface of the yoke 531. Further, an annular (endless) peripheral magnet 533 surrounding the periphery of the central magnet 532 is provided on the outer edge portion of the upper surface of the yoke 531. The central magnet 532 and the peripheral magnet 533 are, for example, permanent magnets. The polarities of the central magnet 532 and the peripheral magnet 533 on the side facing the lower surface of the cathode plate 51 are different from each other. Accordingly, a static magnetic field is formed around the target T by the magnet unit 53.

  A pair of inductively coupled antennas 56 and 57 are provided on the bottom surface of the vacuum chamber 10 so as to protrude in the vacuum chamber 10 across the cathode plate 51. The inductively coupled antennas 56 and 57 are also referred to as LIA (Low Inductance Antenna: a registered trademark of EM Corporation), and the surfaces of the conductors 561 and 571 formed in a substantially U shape are, for example, quartz or the like The dielectric 562 and 572 are covered. The inductively coupled antennas 56 and 57 are extended in the Y direction through the bottom surface of the vacuum chamber 10 with the U-shape turned upside down. A plurality of inductively coupled antennas 56 and 57 are arranged side by side in different positions in the Y direction. The structure in which the surfaces of the conductors 561 and 571 are covered with the dielectrics 562 and 572 prevents the conductors 561 and 571 from being exposed to plasma. Thereby, it is avoided that the constituent elements of the conductors 561 and 571 are mixed into the film on the substrate S.

  The inductively coupled antennas 56 and 57 configured as described above can be regarded as loop antennas in which the X direction is the winding axis direction and the number of turns is less than one. Therefore, the inductance is low. By arranging a plurality of such small antennas side by side in a direction perpendicular to the winding axis direction, it is possible to form an induction electric field for generating plasma, which will be described later, in a wide range while suppressing an increase in inductance. is there.

  The film forming apparatus 1 also includes a chimney 11 that is a cylindrical or box-shaped shielding member that is disposed in the vacuum chamber 10 so as to surround the sputter source 5 and has an upper portion that is elongated in the Y direction. I have. The chimney 11 has a function as a shield that limits the scattering range of the plasma generated in the sputtering source 5 and the sputtered particles sputtered from the target.

  The upper surface of the cathode plate 51 and the lower surface of the substrate S held by the carrier 31 are opposed to each other through the opening 111 above the chimney 11. As will be described later, the space PS surrounded by these becomes a processing space for generating a plasma and sputtering the target to form a film on the substrate S.

  A reactive gas is introduced into the processing space PS from the reactive gas supply unit 6, and an inert gas as a sputtering gas is introduced from the sputtering gas supply unit 7. Specifically, the reactive gas supply unit 6 includes a reactive gas supply source 61 that supplies the reactive gas, a pipe 62 that guides the reactive gas to the processing space PS, and a pipe 62 that is provided inside the chimney 11. And a pair of nozzles 63 and 64 communicating with the reactive gas supply source 61 and a flow rate adjusting unit 65 provided in the middle of the pipe 62. For the purpose of forming an aluminum oxide film on the substrate S, the reactive gas contains oxygen gas.

  The nozzles 63 and 64 are disposed above the inductively coupled antennas 56 and 57, and discharge reactive gas into the processing space PS. The flow rate adjusting unit 65 has a function of controlling the flow rate of the reactive gas supplied to the processing space PS in accordance with a control command from the control unit 19. The flow rate adjusting unit 65 is preferably capable of automatically controlling the flow rate of the reactive gas, and may be provided with a mass flow controller, for example.

  On the other hand, the sputtering gas supply unit 7 discharges the sputtering gas into the vacuum chamber 10 and a sputtering gas supply source 71 that supplies an inert gas, for example, argon gas or xenon gas, introduced into the processing space PS as the sputtering gas. A pair of nozzles 73 and 74, a pipe 72 connecting the sputtering gas supply source 71 and the nozzles 73 and 74, and a flow rate adjusting unit 75 provided in the middle of the pipe 72 are provided.

  The pair of nozzles 73 and 74 are provided on the bottom surface of the vacuum chamber 10 so as to sandwich the cathode plate 51, and the processing space PS between the cathode plate 51 and the inductively coupled antennas 56 and 57 is not sputtered gas. The active gas is discharged. The flow rate adjusting unit 75 has a function of controlling the flow rate of the sputtering gas supplied to the processing space PS in accordance with a control command from the control unit 19. The flow rate adjusting unit 75 is preferably capable of automatically controlling the flow rate of the sputtering gas, and may be provided with, for example, a mass flow controller.

  An appropriate voltage is applied from the power supply unit 8 between the cathode plate 51 and the inductively coupled antennas 56 and 57. Specifically, the cathode plate 51 is connected to a magnetron power supply 81 provided in the power supply unit 8, and an appropriate potential is applied from the magnetron power supply 81. On the other hand, a high frequency power supply 82 provided in the power supply unit 8 is connected to the inductive coupling antennas 56 and 57 via a matching circuit 83, and an appropriate high frequency voltage is applied from the high frequency power supply 82. The voltage value and the waveform output from each of the magnetron power supply 81 and the high frequency power supply 82 are controlled by the control unit 19.

  By applying an appropriate potential from the magnetron power supply 81 to the cathode plate 51, an electric field is formed on the upper surface of the cathode plate 51 and in the vicinity of the surface of the target T held on the cathode plate 51, thereby generating a plasma of sputtering gas (magnetron plasma). ) Is generated. That is, the magnetron power supply 81 applies a voltage necessary for generating magnetron plasma in the processing space PS by the static magnetic field formed by the magnet unit 53 to the cathode plate 51. The output voltage of the magnetron power supply 81 can be various, such as a DC voltage, an AC voltage, and a voltage having a waveform in which an AC voltage or a pulse voltage is superimposed on the DC voltage.

  Further, high frequency power (for example, high frequency power of 13.56 MHz) is supplied from the high frequency power supply 82 to the inductive coupling antennas 56 and 57, so that a high frequency induction electric field is generated between the inductive coupling antennas 56 and 57 and the cathode plate 51. As a result, inductivity coupled plasma (ICP) of sputtering gas and reactive gas supplied to the processing space PS is generated. The plasma generated in this way is also attracted to the static magnetic field formed in the processing space PS by the magnet unit 53. As a result, high-density plasma is generated in the plasma space PL near the surface of the target T in the processing space PS surrounded by the surface of the target T held by the cathode plate 51 and the lower surface of the substrate S.

  The cathode plate 51, the magnet unit 53, and the inductively coupled antennas 56 and 57 all extend long along the Y direction perpendicular to the paper surface of FIG. Therefore, the plasma space PL is also a space region having a shape extending in the Y direction along the surface of the cathode plate 51.

  Thus, the surface of the target T is sputtered by the plasma generated in the plasma space PL, and fine target material particles adhere to the lower surface of the substrate S together with the reactive gas, thereby forming a film on the surface (lower surface) of the substrate S. . Specifically, film formation by plasma sputtering is performed on a band-shaped region along the Y direction on the lower surface of the substrate S, and the substrate S is scanned in a direction parallel to the main surface and perpendicular to the Y direction, that is, the X direction. By being moved, film formation is performed two-dimensionally over the entire film formation target region.

  In order to realize the atmosphere control in the vacuum chamber 10, the film forming apparatus 1 includes a mechanism for detecting plasma emission in the processing space PS and a mechanism for detecting the atmospheric pressure in the processing space PS. Specifically, a probe 191 made of, for example, an optical fiber is disposed in the vicinity of the nozzle 63 that supplies the reactive gas to the processing space PS, and a part of the plasma emission generated in the processing space PS enters the probe 191. . The probe 191 is connected to the spectroscope 192, and the output signal of the spectroscope 192 is input to the control unit 19.

  The control unit 19 detects the plasma emission intensity in the processing space PS by the plasma emission method (PEM) based on the output signal of the spectroscope 192, controls the flow rate adjusting unit 65 as necessary, and is supplied to the processing space PS. Control the flow rate of reactive gas. When the detection target of the probe 191 is, for example, oxygen gas, light intensity at a wavelength of 777 nm, which is a spectral component unique to oxygen, is effective control information in the output signal of the spectrometer 192.

  The vacuum chamber 10 is connected to an exhaust pump 194 and a pressure sensor 193 that outputs a signal corresponding to the atmospheric pressure in the processing space PS to the control unit 19 is provided in the vacuum chamber 10. The control unit 19 controls the exhaust pump 194 according to the signal, whereby the atmospheric pressure in the processing space PS during film formation, that is, the film formation pressure is controlled to a predetermined value.

  The control unit 19 includes a CPU for performing various arithmetic processes, a memory and storage for storing programs executed by the CPU and various data, an interface for exchanging information with external devices and users, and the like. For example, a general-purpose computer device can be used as the control unit 19.

  The film forming apparatus 1 configured as described above arranges a plurality of inductively coupled antennas 56 and 57, which are small antennas having low inductance, in the Y direction, and supplies high frequency power thereto, thereby generating a high frequency induction electric field for generating plasma. Form. According to such a configuration, it is possible to generate a high-density plasma at a low potential and a low temperature. Moreover, uniform plasma generation having an arbitrary shape and size is possible by appropriately setting the antenna arrangement.

  Furthermore, by using together the cathode plate 51 that functions as a magnetron type cathode, it is possible to intensively generate plasma with a particularly high density in a limited region (plasma space PL) in the processing space PS.

  FIG. 2 is a perspective view showing a three-dimensional structure of the sputtering source. In FIG. 2, the chamber 11 covering the sputter source 5 is not shown to clearly show the structure of the sputter source 5. As shown in FIG. 2, the cathode plate 51 and the target T placed thereon are formed in an elongated flat plate shape with the Y direction as the longitudinal direction. The inductively coupled antennas 56 and 57 provided so as to sandwich them from the X direction are also extended with the Y direction as the longitudinal direction. Although not shown in the drawing, a magnet unit 53 extending in the Y direction is provided below the cathode plate 51.

  More specifically, on the (−X) side of the target T, a plurality of inductively coupled antennas 56 having the same shape are arranged in the Y direction. Each inductive coupling antenna 56 is a small one having a conductor 561 formed in a U-shape, but a plurality of the inductive coupling antennas 56 are arranged side by side in the Y direction, so that the entire range is longer than the length of the target T in the Y direction A high frequency electric field can be formed. On the other hand, also on the (+ X) side of the target T, a plurality of inductively coupled antennas 57 are arranged side by side in the Y direction. By arranging the plurality of inductively coupled antennas 56 and 57 so as to sandwich the target T in this way, a uniform high-frequency electric field is formed in the vicinity of the target T in the Y direction.

  Above the inductive coupling antennas 56 and 57, nozzles 63 and 64 for supplying reactive gas to the space above the target T are provided. The nozzles 63 and 64 also have an elongated shape along the inductively coupled antennas 56 and 57, but are divided into a plurality (in this example, 3) in the Y direction. That is, the nozzle 63 includes three nozzle blocks 63a, 63b, and 63c arranged side by side in the Y direction. Similarly, the nozzle 64 includes three nozzle blocks 64a, 64b, and 64c arranged side by side in the Y direction.

  The nozzle blocks 63a and 64a are arranged to face each other at the same position in the Y direction, and a plurality of discharge ports for discharging reactive gas are provided on the surfaces facing each other. Similarly, the nozzle block 63b is arranged so as to face the nozzle block 64b, and the nozzle block 63c is arranged so as to face the nozzle block 64c.

  Corresponding to the fact that the nozzle 63 is divided into three in this way, three sets of probes 191 for detecting the plasma emission intensity are also provided. That is, the probe 191a is provided at a position directly below the center of the nozzle block 63a in the Y direction. Further, a probe 191b is provided at a position directly below the center of the nozzle block 63b in the Y direction. Furthermore, a probe 191c is provided at a position directly below the center of the nozzle block 63c in the Y direction.

  In the sputter source 5 configured in this way, plasma is generated near the upper surface of the target T, and the surface of the target T is sputtered. The substrate S is scanned and moved in the X direction above the sputter source 5, whereby film formation is performed on the lower surface of the substrate S. For example, when aluminum is used as the target T and a gas containing oxygen is used as the reactive gas, an aluminum oxide film is formed on the lower surface of the substrate S.

  By forming an electric field uniformly along the Y direction in the vicinity of the surface of the target T that extends long along the Y direction, plasma having a uniform density in the Y direction can be generated. While the target T is sputtered by such plasma, the substrate S is scanned and moved in the X direction orthogonal to the Y direction, which is the longitudinal direction of the plasma space, so that the coating film formed on the substrate S is made homogeneous. can do.

  In the process of forming a film by reactive sputtering, in addition to the uniform plasma as described above, the density of the reactive gas (in this case, oxygen gas) as the film material in the plasma space PL is uniform in the Y direction. It is necessary to be. It is relatively easy to obtain a uniform gas density at the central portion in the Y direction of the plasma space PL. However, there may be a difference in gas density from the central portion at both ends in the Y direction. In order to cope with this problem, in the sputtering source 5, the nozzles 63 and 64 for discharging the reactive gas to the plasma space PL are divided into three nozzle blocks in the Y direction. Further, as will be described below, the gas discharge amount can be individually adjusted between nozzle blocks having different positions in the Y direction.

  FIG. 3 is a diagram showing a control system for adjusting the density of the reactive gas. When the sputter source 5 is viewed from above, the plasma space PL is virtually divided into three zones as shown in FIG. That is, the plasma space PL is divided into a zone Za sandwiched between the nozzle blocks 63a and 64a facing each other, a zone Zb sandwiched between the nozzle blocks 63b and 64b, and a zone Zc sandwiched between the nozzle blocks 63c and 64c.

  In the reactive gas supply path, the nozzle blocks facing each other are connected in parallel. That is, the output of the flow rate adjusting unit 65a connected to the reactive gas supply source 61 is branched into two, one connected to the nozzle block 63a and the other connected to the nozzle block 64a. Similarly, the nozzle blocks 63b and 64b facing each other across the zone Zb are connected in parallel to the output of the flow rate adjusting unit 65b connected to the reactive gas supply source 61. The nozzle blocks 63c and 64c facing each other across the zone Zc are connected in parallel to the output of the flow rate adjusting unit 65c connected to the reactive gas supply source 61.

  Therefore, the gas discharge amount from the pair of nozzle blocks facing each other across one zone is the same, but the gas discharge amount can be individually changed between the nozzle block pairs having different positions in the Y direction. It is. In other words, in this sputter source 5, it is possible to individually adjust the supply amount of the reactive gas to the plasma space PL for each zone.

  The gas flow rate from the flow rate adjusting unit 65a is controlled by a PID controller 195a that operates according to a control signal from a CPU (Central Processing Unit) 190 of the control unit 19. The gas flow rate from the flow rate adjusting unit 65b is controlled by a PID controller 195b that operates in response to a control signal from the CPU 190. Further, the gas flow rate from the flow rate adjusting unit 65c is controlled by a PID controller 195c that operates in response to a control signal from the CPU 190. As the PID controllers 195a, 195b, and 195c, known configurations used for this kind of flow rate control can be applied.

  The CPU 190 outputs a signal indicating the plasma emission intensity in the zone Za output from the probe 191a via the spectroscope 192a, a signal indicating the plasma emission intensity in the zone Zb output from the probe 191b via the spectroscope 192b, and the probe 191c. Control signals individually for each of the PID controllers 195a, 195b, and 195c based on a signal indicating the plasma emission intensity in the zone Zc output from the A through the spectroscope 192c and a control target value given from a user or a control recipe. give. Thereby, the gas supply amount to each zone Za, Zb, Zc is controlled individually. Specific modes of control will be described later.

  FIG. 4 is a flowchart showing a film forming process in this film forming apparatus. This process is realized by the CPU 190 of the control unit 19 executing a predetermined control program to control each part of the apparatus. First, a flow control function corresponding to the process conditions of the film forming process to be executed is set (step S101). The flow rate control function is a function used for control to maintain the gas supply amount to the plasma space PL at an amount corresponding to the process during the film formation process, and will be described in detail later.

  When the target T containing the film forming material is set on the cathode plate 51 (step S102), the supply of the sputtering gas (for example, argon gas) from the sputtering gas supply unit 7 to the processing space PS is started (step S103). A predetermined voltage is applied from the power supply unit 8 to the inductive coupling antennas 56 and 57 and the cathode plate 51 in a state in which the pressure in the vacuum chamber 10 is controlled to a predetermined value by the exhaust pump 194, and thereby plasma is generated in the processing space PS. It occurs (step S104). Reactive gas is introduced into the processing space PS from the reactive gas supply unit 6 (step S105), and the substrate S is scanned and moved above the sputter source 5 by the transport mechanism 3, thereby forming a film on the lower surface of the substrate S. Is executed (step S106).

  The substrate S may be carried into the vacuum chamber 10 in advance. For example, the substrate S stored in a load lock chamber (not shown) connected to the vacuum chamber 10 is carried into the vacuum chamber 10 after plasma generation. An aspect may be sufficient. If there are a plurality of substrates S to be deposited, these substrates S may pass over the sputtering source 5 sequentially while the plasma is stably generated in the plasma space PL. When film formation is completed for all the film formation target substrates, voltage application and gas supply to each part are stopped, and the film formation process ends (step S107).

  While the reactive gas is supplied to the processing space PS and film formation is performed, the CPU 190 adjusts the discharge amount of the reactive gas from each nozzle, and the density of the reactive gas in the plasma space PL depends on the position and time. To keep it constant. Specifically, the CPU 190 gives a control signal to the PID controllers 195a, 195b, and 195c in accordance with the detection outputs from the probes 191a, 191b, and 191c, and from these, the flow rate adjusters 65a, 65b, and 65c Control the amount of reactive gas delivered.

  The control signal given from the CPU 190 to the PID controller 195a reflects the detection results of the probes 191a, 191b, and 191c. Similarly, the control signal given to the PID controller 195b and the control signal given to the controller 195c also reflect the detection results of the probes 191a, 191b, 191c.

  By reflecting not only the output of the probe that detects the plasma emission intensity in the zone but also the output of the probe provided in the other zone, the gas discharge amount of the nozzle block that supplies gas to one zone is reflected between the zones. It is possible to achieve a uniform and stable gas density by suppressing non-uniformity in gas density and time-dependent fluctuations in gas density due to the interaction. The specific contents of the control for enabling this will be described below.

  The purpose of the control is to control the amount of gas discharged from each nozzle so that the plasma state in each place in the plasma space PL indicated by the output of each probe is preferable for executing a good film forming process. Is to adjust. The “preferred” condition is predetermined as a control target value or given by the user in a timely manner. Therefore, the required control system outputs the gas flow rate to be sent by each flow rate adjusting unit 65a, 65b, 65c in response to the detection signals of the probes 191a, 191b, 191c. Control system.

Such a control system is
Fa = Ga (Ia, Ib, Ic)
Fb = Gb (Ia, Ib, Ic)
Fc = Gc (Ia, Ib, Ic)
Can be described. Here, symbols Fa, Fb, and Fc indicate gas flow rates to be sent by the flow rate adjusting units 65a, 65b, and 65c, respectively, and symbols Ia, Ib, and Ic indicate detection output values of the probes 191a, 191b, and 191c, respectively. Yes, index of plasma emission intensity in each zone. Symbols Ga, Gb, and Gc are multivariable functions described using variables Ia, Ib, and Ic, respectively.

  If these multivariable functions Ga, Gb, and Gc can be specified, the gas to be sent from each flow rate adjusting unit 65a, 65b, and 65c from the detection outputs Ia, Ib, and Ic of the lobes 191a, 191b, and 191c at a certain point in time. The flow rates Fa, Fb, and Fc can be obtained. A preferable condition is realized by providing a control signal for realizing such a flow rate to each PID controller 195a, 195b, 195c.

  The multivariable functions Ga, Gb, Gc are generally nonlinear functions. In order to specify this precisely, for example, a large number of cases of the output Ia, Ib, Ic of each probe when the values of the gas flow rates Fa, Fb, Fc are variously set are collected, and the output Ia of each probe is collected. , Ib, Ic are objective variables, and each gas flow rate Fa, Fb, Fc is used as an explanatory variable to perform multivariate analysis.

From the viewpoint of performing uniform film formation, the plasma state during the film formation process may be constant. Therefore, it is sufficient to correct the plasma when it starts to deviate from the steady state due to some disturbance during the film formation. Thus, as long as it is limited to the vicinity of a certain state, the control system can be approximately expressed by a linear multivariable function. Specifically, it can be expressed as follows.

Here, ΔFa, ΔFb, ΔFc are the amounts of change of the gas flow rates Fa, Fb, Fc from the steady state, and ΔIa, ΔIb, ΔIc are the amounts of change of the probe outputs Ia, Ib, Ic from the steady state at this time. is there. G is a matrix representing the relationship between the gas flow rates Fa, Fb, Fc and the probe outputs Ia, Ib, Ic. More specifically, the matrix G is an expression matrix that linearly maps a three-dimensional vector having the probe outputs Ia, Ib, and Ic as elements to a three-dimensional vector having the gas flow rates Fa, Fb, and Fc as elements.

と表すと、
Ｆ＝ＧＩ … （２）
である。 If such a matrix G is specified, the control amount of the gas flow rate to be set by each PID controller 195a, 195b, 195c is obtained from the change in the output of each probe 191a, 191b, 191c. here,

And
F = GI (2)
It is.

により表すとすると、

と表すことができる。定数ｋは行列Ａの各要素を正規化するための係数である。これを展開して、

が得られる。式（５）の関係は、ガス流量Ｆａ，Ｆｂ，Ｆｃにそれぞれ小さな変分ΔＦａ，ΔＦｂ，ΔＦｃを与えたときの、各プローブ１９１ａ，１９１ｂ，１９１ｃの出力の変化量を示している。 If the inverse matrix of matrix G is matrix A,
I = AF
And the matrix A is

Is represented by

It can be expressed as. The constant k is a coefficient for normalizing each element of the matrix A. Expand this,

Is obtained. The relationship of Expression (5) indicates the amount of change in the output of each probe 191a, 191b, 191c when small variations ΔFa, ΔFb, ΔFc are given to the gas flow rates Fa, Fb, Fc, respectively.

  Therefore, a minute change is given to these gas flow rates from the steady state where the gas flow rates Fa, Fb, and Fc are kept constant, and the amount of change in the output of each probe 191a, 191b, 191c at this time is detected. Thus, each element of the matrix A can be specified. Specifically, it can be considered as follows. This is a simple example of multivariate analysis.

  FIG. 5 is a diagram showing a change in probe output when a variation is added to the gas flow rate from the steady state. As shown in FIG. 5 (a), the probes 191a, 191b, and 191c when the gases with the flow rates Fa, Fb, and Fc are supplied to the sputtering source 5 from the flow rate adjustment units 65a, 65b, and 65c, respectively. Outputs are represented by symbols Ia, Ib, and Ic, respectively. Consider a case where the gas flow rate from the flow rate adjusting unit 65a is changed by ΔFa from this state as shown in FIG. At this time, the output of each probe 191a, 191b, 191c changes due to a change in the amount of gas flowing into the plasma space PL. The amounts of change at this time are denoted by symbols ΔIa1, ΔIb1, and ΔIc1, respectively.

This state corresponds to a state in which ΔFb and ΔFc are set to 0 in Equation (5). As is clear from equation (5),
ΔIa1 = ka ₁₁ ΔFa
ΔIb1 = ka ₂₁ ΔFa
ΔIc1 = ka ₃₁ ΔFa
It is. Therefore, by measuring the output change amounts ΔIa1, ΔIb1, ΔIc1 of the probes 191a, 191b, 191c when the gas flow rate from the flow rate adjusting unit 65a is changed by ΔFa, the element ka ₁₁ among the elements of the matrix A is measured. , Ka ₂₁ , ka ₃₁ can be specified.

Similarly, as shown in FIG. 5C, the elements of the matrix A are calculated from the output change amounts ΔIa2, ΔIb2, ΔIc2 of the probes 191a, 191b, 191c when the gas flow rate from the flow rate adjusting unit 65b is changed by ΔFb. As shown in FIG. 5 (d), ka ₁₂ , ka ₂₂ , and ka ₃₂ change the output change amount ΔIa3 of the probes 191a, 191b, and 191c when the gas flow rate from the flow rate adjustment unit 65c is changed by ΔFc. Elements ka ₁₃ , ka ₂₃ , and ka _{33 of} matrix A are specified from ΔIb3 and ΔIc3, respectively. If each element of the matrix A is specified in this way, a coefficient k for normalizing this and each element a _{11 to} a ₃₃ after normalization are specified. Then, a matrix G is obtained as an inverse matrix of the matrix A. Specifically, the matrix G can be obtained by the following processing.

  FIG. 6 is a flowchart showing a processing example for specifying the matrix G. This process is realized by the CPU 190 executing a control program prepared in advance. First, the gas flow rates Fa, Fb, Fc from the flow rate adjusting units 65a, 65b, 65c are set to a predetermined standard state (step S201). This standard state is a state in which the gas flow rates Fa, Fb, and Fc are known and constant. Since the relationship represented by the matrix G is established in the vicinity of the standard state, it is desirable that the preferable film formation conditions in the actual film formation process be the standard state. In this state, the probe 191a, 191b, 191c detects the plasma emission intensity in each zone (step S202), thereby obtaining the values Ia, Ib, Ic shown in FIG. 5A.

  Next, the gas flow rate from one flow rate adjusting unit 65a is changed by ΔFa (step S203). The values ΔIa1, ΔIb1, and ΔIc1 shown in FIG. 5B are obtained from the outputs of the probes 191a, 191b, and 191c at this time (step S204). After the gas flow rate from the flow rate adjustment unit 65a is returned to the standard state (step S205), the gas flow rate from the other flow rate adjustment unit 65b is changed by ΔFb (step S206), and each probe 191a, Each value ΔIa2, ΔIb2, ΔIc2 shown in FIG. 5C is obtained from the outputs of 191b and 191c (step S207).

  Further, after the gas flow rate from the flow rate adjustment unit 65b is returned to the standard state (step S208), the gas flow rate from the other flow rate adjustment unit 65c is changed by ΔFc (step S209), and each probe at that time Each value ΔIa3, ΔIb3, ΔIc3 shown in FIG. 5D is obtained from the outputs of 191a, 191b, 191c (step S210). In addition, the change order of the flow volume in the three flow volume adjustment parts 65a, 65b, 65c is not limited to the above, but is arbitrary.

  As described above, when the relationship between the flow rate change amount in each flow rate adjustment unit 65a, 65b, 65c and the output change amount of each probe 191a, 191b, 191c is obtained in this way, the matrix A is specified from the result. (Step S211). Then, a matrix G as an inverse matrix of the matrix A is obtained (step S212). The obtained matrix G is stored in a storage unit (not shown) of the control unit 19.

  The process of FIG. 6 can be applied as a process of determining the flow rate control function in step S101 of FIG. That is, by obtaining the matrix G, the equation (2) is specified. Expression (2) is an expression that represents the amount of change in the gas flow rate of each of the flow rate adjusters 65a, 65b, and 65c as a function of the amount of change in the output of each probe 191a, 191b, and 191c. From this relationship, in order to cancel the fluctuation of the plasma emission intensity in the plasma space PL indicated by the output of each probe 191a, 191b, 191c, how should the gas flow rate of each flow rate adjustment unit 65a, 65b, 65c be changed? You are shown how. That is, the relationship represented by the obtained matrix G and equation (2) is applied as a flow rate control function in the control of the gas flow rate for continuously generating a uniform and stable plasma during the film forming process. be able to.

  The timing at which the process of determining the flow rate control function in step S101 should be executed will be described. The relationship represented by the matrix G and the equation (2) is established only in the vicinity of the standard state. Therefore, the standard state should be set according to preferable film forming conditions. For this reason, it is preferable that the flow rate control function is determined under a new standard state if the film formation conditions are changed. On the other hand, as long as the same film forming conditions are applied in the same apparatus, the flow rate control function should basically be the same. However, it is preferable to obtain a flow control function again when a part or target involved in the film forming process is exchanged. In addition, since the plasma state may change over time depending on the consumption state of the target, the flow control function should be periodically reviewed even when the film formation process is continued under the same film formation conditions. Is preferred.

  During the film formation process, the reactive gas (oxygen gas) to the plasma space PL is determined based on the plasma emission intensity of each zone detected by each probe and the flow rate control function obtained as described above. ) Is feedback controlled. The details of the control are as follows.

  FIG. 7 is a flowchart showing an example of gas flow rate control during film formation. This process is realized by the CPU 190 executing a control program prepared in advance. The detection outputs of the probes 191a, 191b, 191c are always input to the control unit 19, and from this, the actual measurement results of the plasma emission intensity in each zone Za, Zb, Zc are acquired (step S301). These are compared with a target value of a preferable plasma emission intensity under ideal film formation conditions, and the difference is calculated (step S302). The “preferred plasma emission intensity” is given in advance as a control target value or acquired in step S202 of FIG. Differences between the target value of the plasma emission intensity and the actual measurement value are variables ΔIa, ΔIb, ΔIc in the flow rate control function.

  By substituting these difference values ΔIa, ΔIb, ΔIc into the flow rate control function represented by the matrix G, the control target values Fa, Fb, Fc of the gas flow rate from the flow rate adjustment units 65a, 65b, 65c and Differences ΔFa, ΔFb, ΔFc from the actual flow rate are estimated (step S303). The difference values ΔFa, ΔFb, ΔFc are respectively given as control amounts to the PID controllers 195a, 195b, 195c (step S304), and the PID controllers 195a, 195b, 195c control the flow rate adjustment units 65a, 65b, 65c, respectively. The gas flow rate from each flow rate adjustment unit 65a, 65b, 65c is brought close to the control target value. By constantly executing this process, the plasma state in the plasma space PL is stably maintained.

  The fluctuation of the gas supply amount to one zone affects not only the zone but also the plasma state of other zones. Therefore, the control of merely feeding back the plasma emission intensity detected in one zone to the gas supply amount to the zone does not necessarily provide the uniformity and stability of the plasma in the entire plasma space PL. In this embodiment, the gas supply amount to each zone is comprehensively controlled from the plasma emission intensities detected in each of the three zones Za, Zb, and Zc based on a flow control function acquired in advance. Therefore, uniform and stable plasma can be maintained in the entire plasma space PL. As a result, the film forming apparatus 1 of this embodiment can form a uniform film on the substrate S.

  Next, a modified example of the gas flow rate control in the film forming apparatus 1 will be described. In the gas flow rate control of the above embodiment, the supply amount of the reactive gas to each zone is determined based on the plasma emission intensity detected in each zone and the flow control function obtained with the preferable film forming conditions as the standard state. Be controlled. For this reason, it has a favorable responsiveness with respect to the comparatively small fluctuation | variation produced focusing on a standard state. However, for some unexpected reason, the plasma state does not always deviate greatly from the standard state during film formation. In such a case, the above-described control may not be able to follow the fluctuation, or it may take time to converge the vibration due to an excessive response.

  The modification described below is for dealing with such a problem. In the following description, the same components as those in the above embodiment are denoted by the same reference numerals, and detailed description thereof is omitted. In addition, each configuration of the above-described embodiment that is not mentioned in the following description is assumed to function in the same manner in the modified example.

  FIG. 8 is a diagram showing a first modification of the gas flow rate control. In this modification, a cooperative correction circuit 196 is added to the control signal path from the CPU 190 to the PID controllers 195a, 195b, and 195c. The cooperative correction circuit 196 may be configured by either software or hardware, but its function is that the control signal output from the CPU 190 is a response by the PID controllers 195a, 195b, 195c, that is, the flow rate adjustment units 65a, 65b, 65c. It is to change and adjust the time constant until it is reflected in the output.

  When the plasma state changes in the vicinity of the standard state, feedback control with quick response is performed in order to correct the deviation from the optimum film formation condition at an early stage. That is, the cooperative correction circuit 196 has a relatively small time constant. On the other hand, when there is a large deviation in the film forming conditions, such a fast response feedback causes a large vibration in a transient response state. When such a phenomenon occurs in one zone, the effect may spread to other zones, and the plasma state may fluctuate greatly in the plasma space PL.

  In the modification shown in FIG. 8, the CPU 190 monitors the magnitude of the signal output from each probe 191a, 191b, 191c, the rate of change thereof, the amount of change in the gas flow rate obtained from the equation (2), and the like. If it is estimated from these values that the film forming conditions have changed significantly, the time constant of the cooperative correction circuit 196 is changed to a larger one. Thereby, the responsiveness of the feedback control becomes more gradual, and the vibration of the entire film forming condition is suppressed. By doing so, it is possible to achieve both highly accurate process control at the time of stability and stability against irregular disturbances. Note that the time constant setting in the cooperative correction circuit 196 may be three or more stages.

  FIG. 9 is a diagram showing a second modification of the gas flow rate control. In this modification, the CPU 190 has a flow rate control function table 197. In this example, a plurality of flow rate control functions are obtained in advance by partially changing the film forming conditions, and these are stored in the flow rate control function table 197.

  For example, the film formation rate and film quality differ between the film formation using the new target T and the film formation using the target T whose surface has advanced oxidation, and the preferable film formation conditions differ accordingly. When film formation is continuously performed on a large number of substrates S, the surface state of the target T may be different between the initial stage and the latter stage. In order to cope with such a problem, in this modification, a plurality of flow control functions are prepared in advance. Then, the flow rate control function is changed as necessary from the process state information indicating the process state such as the film formation rate and the film quality measurement result on the substrate S carried out of the vacuum chamber 10 after film formation. This makes it possible to continue good film formation.

  Note that the CPU 190 may be configured to set an optimal flow control function according to the process state at that time by fuzzy control that performs multi-stage determination based on a plurality of pieces of information representing the state of the film forming process. Good.

  As described above, in the film forming apparatus 1 of the above embodiment, the pair of nozzles 63a and 64a, the pair of nozzles 63b and 64b, and the pair of nozzles 63c and 64c each function as the “gas discharge means” of the present invention. ing. In addition, the probes 191a, 191b, and 191c each function as “detection means” of the present invention. The CPU 190 functions as a “control unit” of the present invention.

  In the above embodiment, the power supply unit 8, the cathode plate 51, the magnet unit 53, and the inductively coupled antennas 56 and 57 function as a “plasma generating means” of the present invention. Moreover, the sputter source 5 including these corresponds to the “plasma generator” of the present invention as a whole. The cathode plate 51 also functions as “target holding means” of the present invention, and the carrier 31 of the transport mechanism 3 functions as “substrate holding means” of the present invention. Each of the zones Za, Zb, and Zc corresponds to a “partial region” of the present invention.

  The present invention is not limited to the above-described embodiment, and various modifications other than those described above can be made without departing from the spirit of the present invention. For example, in the above embodiment, the plasma space PL is divided into three zones Za, Zb, and Zc, but the number of zone divisions is not limited to this and is arbitrary. However, as the number of zone sections increases, the interaction between them becomes complicated and stable control becomes difficult. As long as the uniformity of plasma in each zone is ensured, it is preferable that the number of zone sections is small.

  In addition, for example, in the above embodiment, the flat cathode plate 51 and the target T are used. However, a film forming technique using a cylindrical rotary cathode whose surface is coated with a target material is also known. The present invention can also be applied to an apparatus using a target.

  Moreover, in the said embodiment, this invention is applied to the flow control of oxygen gas as material gas for forming an aluminum oxide film. However, the application target of the present invention is not limited to oxygen gas, and the present invention can also be applied to flow control of other reactive gases such as nitrogen gas and sputtering gas such as argon. Further, the flow rate control of the present invention may be applied to each of a plurality of types of gases introduced into the plasma space PL.

  In the above embodiment, the gas flow rate is calculated using a linear flow rate control function that approximately describes the behavior in the vicinity of the standard state on the assumption that stable control should be performed in the vicinity of the standard state. Is controlled. However, in an actual film forming process, more factors influence each other in a complicated manner, and the process state is determined. In general, the behavior is described by a higher-order nonlinear function. If such a behavior can be modeled into an appropriate model, control may be realized by such a model and its optimization method.

  In the above description, a glass substrate is exemplified as the substrate S and an aluminum oxide film is exemplified as the film forming material. However, the substrate and the film forming material are not limited thereto. In particular, when the film is formed by various reactive sputtering, the film forming apparatus 1 of the present embodiment functions effectively.

  Moreover, although the said embodiment is a film-forming apparatus by reactive plasma sputtering, this invention is not limited to what uses plasma for the purpose of film-forming in this way, For example, surface processing and modification | reformation processing of material are performed. The present invention can be applied to various applications that require uniform plasma over a wide range, such as the intended one.

  As described above, the specific embodiment has been described by way of example. In the present invention, the multivariable function is a vector having elements of the detection values of the plurality of detection units as gas elements from the plurality of gas discharge units. It may be expressed by an expression matrix that linearly maps to a vector having the discharge amount as an element. According to such a configuration, since the gas discharge amount from the gas discharge means is represented by a linear function with the detection value of each detection means as a variable, it is relatively easy to specify a multivariable function, and Since the control used is also linear, it is easy to implement.

In this case, the control means acquires the detection value of each of the plurality of detection means in a standard state in which the gas discharge amounts from the plurality of gas discharge means are constant, and the gas for one selected from the plurality of gas discharge means. The step of obtaining the detection value of each of the plurality of detection means by setting the discharge amount to a value different from the standard state is performed while selecting each of the plurality of gas discharge means one by one, from the plurality of gas discharge means Assuming that a vector having the change amount of the gas discharge amount as an element is F and a vector having the change amount of the detection value of each of the plurality of detection means as an element is I,
I = AF
The matrix A satisfying the above relationship may be obtained, and the inverse matrix of the matrix A may be used as the expression matrix.

  According to such a configuration, each element of the matrix A can be reliably specified one by one, so that the expression matrix can be uniquely specified regardless of the dimension of the matrix.

  Further, the control means changes the time constant until the setting of the gas discharge amount of each of the plurality of gas discharge means represented by a multivariable function is reflected in the gas discharge amount from each of the plurality of gas discharge means. It may be configured to be possible. In a control system with a small response time constant, it is possible to quickly correct slight fluctuations in the state, while large deviations may cause state vibration. By making it possible to change the response time constant, stable control is possible even for large fluctuations in the state.

  The control means may be configured to set the gas discharge amount of each of the plurality of gas discharge means based on one multivariable function selected from a plurality of prepared in advance. Preferred process conditions can change over time due to changes in the state of each part of the apparatus. By preparing a plurality of types of multivariable functions and applying one selected from them, it is possible to stably maintain the plasma state in response to such a state change.

  Further, in the film forming apparatus of the present invention, the plasma generator forms a plasma space extending long in one axial direction along the surface of the target, and the substrate holding means is relative to the target and in the plasma space. The substrate may be scanned and moved in a direction orthogonal to the longitudinal direction. According to such a configuration, it is possible to perform uniform film formation with uniform plasma on an elongated strip-shaped region along the longitudinal direction of the plasma space in the substrate surface. Then, by moving the substrate in a direction orthogonal to the longitudinal direction of the plasma space relative to the plasma space, a film having good in-plane uniformity can be formed two-dimensionally on the substrate. .

  The present invention is suitable for applications that require a uniform plasma space in a wide range, and can be particularly suitably used for, for example, plasma film formation.

DESCRIPTION OF SYMBOLS 1 Film-forming apparatus 3 Transfer mechanism 5 Sputter source (plasma generator)
8 Power supply unit (plasma generating means)
31 Carrier (substrate holding means)
51 Cathode plate (plasma generating means, target holding means)
53 Magnet unit (plasma generating means)
56, 57 Inductively coupled antenna (plasma generating means)
63, 63a-63c, 64, 64a-64c Nozzle (gas discharge means)
190 CPU (control means)
191, 191a, 191b, 191c Probe (detection means)
S substrate

Claims (8)

A plurality of gas discharge means discharge gas serving as plasma seeds toward different regions in the plasma space for generating plasma, and the discharge amount of the gas can be set for each gas discharge means. In the control method of the plasma generator,
Each of a plurality of detection means provided corresponding to each of the plurality of gas discharge means detects a plasma emission intensity in the partial region where the gas is discharged by the corresponding one of the gas discharge means,
The control means sets the gas discharge amount of each of the plurality of gas discharge means based on a predetermined multi-variable function having the detection results of the plurality of detection means as variables.
The multivariable function is obtained by multivariate analysis using the plasma emission intensity detected by each of the detection means as an objective variable and the discharge amount of the gas discharged from each of the gas discharge means as explanatory variables. Control method of the device.   The multivariable function is represented by an expression matrix that linearly maps a vector whose element is a detection value of each of the plurality of detection means into a vector whose element is the discharge amount of the gas from the plurality of gas discharge means. The control method of the plasma generator of Claim 1. The control means includes
Obtaining a detection value of each of the plurality of detection means in a standard state in which the discharge amount of the gas from the plurality of gas discharge means is fixed, respectively;
The step of setting the discharge amount of the gas for one selected from the plurality of gas discharge means to a value different from the standard state, and obtaining the detection value of each of the plurality of detection means, Run while selecting each one,
Assuming that a vector whose element is a change amount of the gas discharge amount from the plurality of gas discharge means is F and a vector whose element is a change amount of the detection value of each of the plurality of detection means is I, the following formula:
I = AF
The control method of the plasma generator of Claim 2 which calculates | requires the matrix A which satisfy | fills this relationship, and makes the inverse matrix of the matrix A the said expression matrix.   The control means until the setting of the gas discharge amount of each of the plurality of gas discharge means represented by the multivariable function is reflected in the gas discharge amount from each of the plurality of gas discharge means. 4. The method for controlling a plasma generating apparatus according to claim 1, wherein the time constant can be changed.   5. The control unit according to claim 1, wherein the control unit sets a discharge amount of the gas of each of the plurality of gas discharge units based on one multivariable function selected from a plurality of prepared in advance. Control method of plasma generator. Plasma generating means for generating plasma;
A plurality of gas ejection means for ejecting gases serving as plasma seeds toward different regions in the plasma space where the plasma is generated, and capable of independently changing the amount of the reactive gas discharged When,
A plurality of gas discharge units corresponding to each of the plurality of gas discharge units, and detecting means for detecting plasma emission intensity in the partial region in which the gas is discharged by the corresponding one of the gas discharge units;
Control means for setting the gas discharge amount of each of the plurality of gas discharge means based on a predetermined multi-variable function having the detection results of the plurality of detection means as variables;
The multivariable function is a plasma generating apparatus that is obtained by multivariate analysis using the plasma emission intensity detected by each of the detection means as an objective variable and the discharge amount of the gas discharged from each of the gas discharge means as explanatory variables. . Target holding means for holding the target in the processing chamber;
The plasma generator according to claim 6, wherein plasma is generated near the surface of the target in the processing chamber;
A film forming apparatus comprising substrate holding means for holding the substrate so that the plasma generating apparatus faces a plasma space where plasma is generated. The plasma generator forms the plasma space extending long in one axial direction along the surface of the target,
The film forming apparatus according to claim 7, wherein the substrate holding unit scans and moves the substrate relative to the target and in a direction orthogonal to a longitudinal direction of the plasma space.

Images