Exhaust gas decomposition apparatus

Abstract

The present disclosure discloses an exhaust gas decomposition apparatus that uses plasma to decompose metal precursors emitted from a semiconductor process chamber. In particular, the exhaust gas decomposition apparatus may comprise: a discharging tube that provides an inner space in which the plasma is induced; an antenna that is disposed so as to surround the discharging tube and induces the plasma in the inner space of the discharging tube by receiving RF power; and a radio frequency generator (RFG) that applies either a first voltage or a second voltage to the antenna.

Description

Exhaust gas decomposition device

[0001] The present disclosure is intended to disclose an exhaust gas decomposition device. More specifically, the invention relates to a device for decomposing and oxidizing unreacted metal precursors between a process chamber and a pump in order to prevent unreacted metal precursors discharged from a process chamber from being deposited on a pump and piping, thereby causing abnormalities in the semiconductor process or failure of the pump.

[0002] Recently, as the miniaturization of semiconductors has become required, the thickness of the oxide film (SiO2) deposited on the wafer has also been required to become thinner. However, if the oxide film becomes too thin, leakage current increases, which has the disadvantage of degrading semiconductor performance. Accordingly, High-K insulating films with a higher dielectric constant than conventional oxide films have been used as dielectric materials for transistor gates, and as metal materials capable of lowering the threshold voltage when deposited on High-K insulating films are used as gate electrodes, the use of metal precursors in semiconductor processes is increasing.

[0003] However, some of the metal precursors introduced into the semiconductor process do not react during the process and are discharged out of the semiconductor process chamber. These unreacted metal precursors are deposited on pipes or pumps connected to the semiconductor process chamber, shortening the lifespan of the pipes or pumps and causing frequent replacement cycles. Additionally, metal precursors deposited on the pipes or pumps can affect the pressure in the semiconductor process chamber, potentially leading to a degradation in the quality of semiconductor products.

[0004] Accordingly, there is an increasing demand for exhaust gas decomposition devices that maximize the decomposition of unreacted metal precursors to prevent their deposition on pipes or pumps, thereby extending the replacement cycle of pipes or pumps and preventing semiconductor quality degradation.

[0005] The present disclosure aims to provide an apparatus for decomposing exhaust gas discharged from a semiconductor process chamber.

[0006] The problem that the present disclosure aims to solve is to provide an apparatus that increases the lifespan of pipes or pumps by oxidizing a metal material discharged from a semiconductor process chamber and converting it into a powder form, thereby minimizing deposition on the pipes or pumps.

[0007] The problem that the present disclosure aims to solve is to provide an apparatus for igniting plasma using a gas already used in a semiconductor process.

[0008] The problems that the present disclosure aims to solve are not limited to those described above, and problems not mentioned will be clearly understood by those skilled in the art corresponding to the present disclosure from the present specification and the attached drawings.

[0009] According to one embodiment of the present disclosure, an exhaust gas decomposition apparatus for igniting a plasma for decomposing a metal precursor discharged from a semiconductor process chamber using nitrogen gas comprises: a load; and an RF Generator (RFG) electrically connected to the load and supplying RF power to the load; wherein the load comprises a discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet where the plasma is induced—wherein the nitrogen gas and the metal precursor are introduced into the internal space of the discharging tube through the inlet. and an antenna disposed to surround the discharge tube and receiving RF power from the RFG to induce the plasma in the internal space of the discharge tube - the antenna comprises a plurality of turn antenna segments and at least one capacitor, and has a source resonant frequency defined by the inductance of each of the plurality of turn antenna segments and the capacitance of each of the at least one capacitor -; wherein the RFG comprises an inverter that applies an AC voltage to the antenna so as to supply the RF power to the load - the AC voltage has a driving frequency determined by a control signal applied to the inverter -; and a controller that inputs the control signal to the inverter so that the plasma can be ignited using the nitrogen gas to decompose the metal precursor;The controller comprises: inputting the control signal to the inverter such that the AC voltage having the source resonant frequency as the driving frequency is applied to the antenna; periodically performing a frequency tracking operation that causes the driving frequency of the AC voltage to track the resonant frequency of the load, which varies according to the change in impedance of the load caused by the inflow of the nitrogen gas, at least until the plasma is ignited from the nitrogen gas; and inputting the control signal to the inverter such that the metal precursor is decomposed using the plasma ignited from the nitrogen gas according to the frequency tracking operation, wherein the frequency tracking operation is characterized by the controller assuming a virtual signal obtained by shifting the control signal by a predetermined time—wherein the predetermined time is the time from a first time point when the control signal is applied to the inverter to a second time point when an AC voltage corresponding to the control signal is expected to be output from the inverter—and the output current output from the inverter to the antenna and the virtual signal A delay time between is calculated, and based on the delay time, a next control signal is determined such that the driving frequency of the next AC voltage to be output to the inverter in the next cycle corresponds to the resonant frequency of the load, and the determined next control signal can be input to the inverter such that the next AC voltage having a driving frequency corresponding to the resonant frequency of the load is applied to the antenna.;

[0010] According to one embodiment of the present disclosure, a plasma ignition device for igniting plasma using nitrogen gas comprises: a load; and an RFG (RF Generator; RFG) electrically connected to the load and supplying RF power to the load; wherein the load comprises a discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet where the plasma is induced—wherein the nitrogen gas is introduced into the internal space of the discharging tube through the inlet. and an antenna disposed to surround the discharge tube and receiving RF power from the RFG to induce the plasma in the internal space of the discharge tube - the antenna comprises a plurality of turn antenna segments and at least one capacitor, and has a source resonant frequency defined by the inductance of each of the plurality of turn antenna segments and the capacitance of each of the at least one capacitor -; wherein the RFG comprises an inverter that applies an AC voltage to the antenna so that the RF power is supplied to the load - the AC voltage has a driving frequency determined by a control signal applied to the inverter -; and a controller that inputs the control signal to the inverter so that the plasma can be ignited using the nitrogen gas;The controller comprises, wherein the controller inputs the control signal to the inverter so that the AC voltage having the source resonant frequency as the driving frequency is applied to the antenna, and periodically performs a frequency tracking operation that causes the driving frequency of the AC voltage to track the resonant frequency of the load, which varies according to the change in impedance of the load caused by the inflow of the nitrogen gas, at least until the plasma is ignited from the nitrogen gas; wherein the frequency tracking operation is characterized in that the controller assumes a virtual signal obtained by shifting the control signal by a predetermined time—wherein the predetermined time is the time from a first time point when the control signal is applied to the inverter to a second time point when the AC voltage corresponding to the control signal is expected to be output from the inverter—calculates the delay time between the output current output from the inverter to the antenna and the virtual signal, and based on the delay time, the driving frequency of the next AC voltage to be output to the inverter in the next cycle is of the load A next control signal is determined to correspond to a resonant frequency, and the determined next control signal can be input to the inverter so that the next AC voltage having a driving frequency corresponding to the resonant frequency of the load is applied to the antenna.;

[0011] According to one embodiment of the present disclosure, an exhaust gas decomposition apparatus for decomposing a metal precursor discharged from a semiconductor process chamber using plasma comprises: a discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet where the plasma is induced—wherein the metal precursor is introduced into the internal space through the inlet—; an antenna disposed to surround the discharging tube and receiving RF power to induce the plasma into the internal space of the discharging tube—wherein the antenna has an inductance less than or equal to a predetermined value such that, after the plasma is ignited, an electromotive force is provided to the internal space of the discharging tube to cause the metal precursor to be decomposed at a rate greater than the expected decomposition rate—; and having a first output terminal and a second output terminal, and applying a first voltage or a second voltage to the antenna through the first output terminal and the second output terminal, and comprising an RFG (Radio Frequency Generator; RFG) - wherein the first voltage is defined as the first potential of the first output terminal being higher than the second potential of the second output terminal, and the second voltage is defined as the second potential being higher than the first potential -; and the RFG comprises an inverter having a plurality of switches and applying the first voltage or the second voltage to the antenna according to the operation of the plurality of switches; and a controller that controls the operation of the plurality of switches;The controller includes, wherein the controller senses an output current provided from the inverter to the antenna, and based on the sensed output current, selects one of a plurality of output voltage patterns in which the first voltage and the second voltage are output from the inverter, and controls the operation of the plurality of switches so that the first voltage or the second voltage is applied to the antenna according to the selected output voltage pattern, - wherein each of the plurality of output voltage patterns includes at least one powering section that outputs the first voltage or the second voltage and a freewheeling section that does not output both the first voltage and the second voltage - The controller selects a first output voltage pattern in which the total length of the freewheeling section within the output voltage pattern is greater than or equal to a first length in response to the sensed output current indicating that the plasma is ignited, and controls the operation of the plurality of switches according to the selected first output voltage pattern, - wherein the selected first output voltage pattern is such that the current flowing through each of the plurality of switches before the plasma is ignited, despite an inductance less than or equal to a predetermined value By making each of the plurality of switches below the maximum allowable current, and selecting a second output voltage pattern in which the total length of the freewheeling section within the output voltage pattern is less than the first length from the point in time when the sensed output current indicates that the plasma has been ignited, the operation of the plurality of switches can be controlled according to the selected second output voltage pattern.

[0012] According to one embodiment of the present disclosure, exhaust gas discharged from a semiconductor process chamber is decomposed and oxidized to prevent deposition on pipes or pumps, thereby increasing the lifespan of pipes or pumps.

[0013] According to one embodiment of the present disclosure, by igniting plasma using a gas already used in a semiconductor process, changes in the semiconductor process environment caused by the inflow of ignition gas can be minimized and semiconductor quality can be maintained.

[0014] The effects according to the present disclosure are not limited to the effects described above, and unmentioned effects will be clearly understood by those skilled in the art from the present disclosure and the attached drawings.

[0015] FIG. 1 is a drawing showing a semiconductor process system according to the present disclosure.

[0016] FIGS. 2 and 3 are drawings for explaining the configuration of an exhaust gas decomposition device according to the present disclosure.

[0017] FIGS. 4 and 5 are drawings for illustrating a portion of the load of an exhaust gas decomposition device according to the present disclosure.

[0018] FIGS. 6 and 7 are drawings for explaining an RFG (Radio Frequency Generator) according to the present disclosure.

[0019] FIGS. 8 to 10 are drawings for explaining a method of igniting plasma using nitrogen gas.

[0020] FIGS. 11 and 12 are drawings for explaining an antenna structure according to one embodiment of the present disclosure.

[0021] FIGS. 13 to 20 are drawings for explaining a method for decomposing exhaust gas according to one embodiment of the present disclosure.

[0022] FIG. 21 is a drawing for explaining the configuration of an RFG according to another embodiment of the present disclosure.

[0023] FIGS. 22 to 25 are drawings for explaining a method for decomposing exhaust gas according to another embodiment of the present disclosure.

[0024] FIGS. 26 and 27 are drawings for illustrating various experimental examples according to embodiments of the present disclosure.

[0025] The aforementioned purposes, features, and advantages of this specification will become more apparent from the following detailed description in conjunction with the accompanying drawings. However, as this specification is subject to various modifications and may have various embodiments, specific embodiments are illustrated in the drawings and described in detail below.

[0026] In the drawings, the thicknesses of layers and regions are exaggerated for clarity, and when an element or layer is referred to as being "on" or "on" another element or layer, this includes not only being directly on top of another element or layer but also cases where another layer or element is interposed in between. Throughout the specification, identical reference numerals generally represent identical elements. Additionally, elements with identical functions within the same scope of concept appearing in the drawings of each embodiment are described using the same reference numeral, and redundant descriptions thereof are omitted.

[0027] Numbers used in the description of this specification (e.g., 1st, 2nd, etc.) are merely identifiers to distinguish one component from another.

[0028] Furthermore, the suffixes "module" and "part" for components used in the following embodiments are assigned or used interchangeably solely for the ease of drafting the specification, and do not inherently possess distinct meanings or roles.

[0029] In the following examples, singular expressions include plural expressions unless the context clearly indicates otherwise.

[0030] In the following embodiments, terms such as "comprising" or "having" mean that the features or components described in the specification are present, and do not preclude the possibility that one or more other features or components may be added.

[0031] In the drawings, the size of components may be exaggerated or reduced for convenience of explanation. For example, the size and thickness of each component shown in the drawings are arbitrarily depicted for convenience of explanation, and the present invention is not necessarily limited to what is illustrated.

[0032] Where an embodiment can be implemented differently, a specific process sequence may be performed differently from the order described. For example, two processes described consecutively may be performed substantially simultaneously or proceed in the reverse order of the description.

[0033] In the following embodiments, when membranes, regions, components, etc. are described as being connected, the cases include not only those where membranes, regions, and components are directly connected, but also those where other membranes, regions, and components are interposed between them to be indirectly connected.

[0034] For example, when a membrane, region, component, etc. is described as being electrically connected in this specification, it includes not only cases where the membrane, region, component, etc. are directly electrically connected, but also cases where another membrane, region, component, etc. is interposed therein and they are indirectly electrically connected.

[0035] An exhaust gas decomposition apparatus according to the present disclosure for igniting a plasma for decomposing a metal precursor discharged from a semiconductor process chamber using nitrogen gas comprises: a load; and an RF Generator (RFG) electrically connected to the load and supplying RF power to the load; wherein the load comprises a discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet where the plasma is induced—wherein the nitrogen gas and the metal precursor are introduced into the internal space of the discharging tube through the inlet. and an antenna disposed to surround the discharge tube and receiving RF power from the RFG to induce the plasma in the internal space of the discharge tube - the antenna comprises a plurality of turn antenna segments and at least one capacitor, and has a source resonant frequency defined by the inductance of each of the plurality of turn antenna segments and the capacitance of each of the at least one capacitor -; wherein the RFG comprises an inverter that applies an AC voltage to the antenna so as to supply the RF power to the load - the AC voltage has a driving frequency determined by a control signal applied to the inverter -; and a controller that inputs the control signal to the inverter so that the plasma can be ignited using the nitrogen gas to decompose the metal precursor;The controller comprises: inputting the control signal to the inverter such that the AC voltage having the source resonant frequency as the driving frequency is applied to the antenna; periodically performing a frequency tracking operation that causes the driving frequency of the AC voltage to track the resonant frequency of the load, which varies according to the change in impedance of the load caused by the inflow of the nitrogen gas, at least until the plasma is ignited from the nitrogen gas; and inputting the control signal to the inverter such that the metal precursor is decomposed using the plasma ignited from the nitrogen gas according to the frequency tracking operation, wherein the frequency tracking operation is characterized by the controller assuming a virtual signal obtained by shifting the control signal by a predetermined time—wherein the predetermined time is the time from a first time point when the control signal is applied to the inverter to a second time point when an AC voltage corresponding to the control signal is expected to be output from the inverter—and the output current output from the inverter to the antenna and the virtual signal A delay time between is calculated, and based on the delay time, a next control signal is determined such that the driving frequency of the next AC voltage to be output to the inverter in the next cycle corresponds to the resonant frequency of the load, and the determined next control signal can be input to the inverter such that the next AC voltage having a driving frequency corresponding to the resonant frequency of the load is applied to the antenna.;

[0036] In addition, the antenna is composed of a plurality of layers corresponding to each of a plurality of planes perpendicular to the virtual central axis of the discharge tube, and each of the plurality of layers includes a turn antenna segment having a certain radius from the virtual central axis, and the at least one capacitor can electrically connect the turn antenna segments included in adjacent layers.

[0037] Additionally, the frequency tracking operation may be such that the controller determines the next control signal so that the driving frequency of the next AC voltage maintains the driving frequency of the current AC voltage in response to the delay time being included within a preset delay time range, and determines the next control signal so that the driving frequency of the next AC voltage is higher or lower than the driving frequency of the current AC voltage in response to the delay time not being included within the preset delay time range.

[0038] Additionally, the fact that the above delay time is included within the above-preset delay time range indicates that the driving frequency of the current AC voltage is sufficiently corresponding to the resonant frequency of the load, and the fact that the above delay time is not included within the above-preset delay time range may indicate that the driving frequency of the current AC voltage is not sufficiently corresponding to the resonant frequency of the load.

[0039] In addition, the change in impedance of the load may be caused by parasitic capacitance and parasitic inductance that occur as the RF power is supplied to the antenna and change due to the inflow of the nitrogen gas.

[0040] In addition, the controller can measure the output current and determine that the plasma is ignited based on the fact that the value of the output current is less than or equal to a preset value.

[0041] In addition, the controller may continue to perform the frequency tracking operation to decompose the metal precursor even after determining that the plasma has been ignited.

[0042] In addition, the amount of change in the resonant frequency of the load that changes over time before the plasma is ignited may be smaller than the amount of change in the resonant frequency of the load that changes over time after the plasma is ignited.

[0043] A plasma ignition device for igniting plasma using nitrogen gas according to the present disclosure comprises: a load; and an RFG (RF Generator; RFG) electrically connected to the load and supplying RF power to the load; wherein the load comprises a discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet in which the plasma is induced—wherein the nitrogen gas is introduced into the internal space of the discharging tube through the inlet. and an antenna disposed to surround the discharge tube and receiving RF power from the RFG to induce the plasma in the internal space of the discharge tube - the antenna comprises a plurality of turn antenna segments and at least one capacitor, and has a source resonant frequency defined by the inductance of each of the plurality of turn antenna segments and the capacitance of each of the at least one capacitor -; wherein the RFG comprises an inverter that applies an AC voltage to the antenna so that the RF power is supplied to the load - the AC voltage has a driving frequency determined by a control signal applied to the inverter -; and a controller that inputs the control signal to the inverter so that the plasma can be ignited using the nitrogen gas;The controller comprises, wherein the controller inputs the control signal to the inverter so that the AC voltage having the source resonant frequency as the driving frequency is applied to the antenna, and periodically performs a frequency tracking operation that causes the driving frequency of the AC voltage to track the resonant frequency of the load, which varies according to the change in impedance of the load caused by the inflow of the nitrogen gas, at least until the plasma is ignited from the nitrogen gas; wherein the frequency tracking operation is characterized in that the controller assumes a virtual signal obtained by shifting the control signal by a predetermined time—wherein the predetermined time is the time from a first time point when the control signal is applied to the inverter to a second time point when the AC voltage corresponding to the control signal is expected to be output from the inverter—calculates the delay time between the output current output from the inverter to the antenna and the virtual signal, and based on the delay time, the driving frequency of the next AC voltage to be output to the inverter in the next cycle is of the load A next control signal is determined to correspond to a resonant frequency, and the determined next control signal can be input to the inverter so that the next AC voltage having a driving frequency corresponding to the resonant frequency of the load is applied to the antenna.;

[0044] In addition, the controller may continue to perform the frequency tracking operation to process the process gas introduced into the internal space of the discharge tube through the inlet even after determining that the plasma has been ignited.

[0045] An exhaust gas decomposition apparatus according to the present disclosure for decomposing a metal precursor discharged from a semiconductor process chamber using plasma comprises: a discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet where the plasma is induced—wherein the metal precursor is introduced into the internal space through the inlet—; an antenna disposed to surround the discharging tube and receiving RF power to induce the plasma into the internal space of the discharging tube—wherein the antenna has an inductance less than or equal to a predetermined value such that, after the plasma is ignited, an electromotive force is provided to the internal space of the discharging tube to cause the metal precursor to be decomposed at a rate greater than the expected decomposition rate—; and having a first output terminal and a second output terminal, and applying a first voltage or a second voltage to the antenna through the first output terminal and the second output terminal, and comprising an RFG (Radio Frequency Generator; RFG) - wherein the first voltage is defined as the first potential of the first output terminal being higher than the second potential of the second output terminal, and the second voltage is defined as the second potential being higher than the first potential -; and the RFG comprises an inverter having a plurality of switches and applying the first voltage or the second voltage to the antenna according to the operation of the plurality of switches; and a controller that controls the operation of the plurality of switches;The controller includes, wherein the controller senses an output current provided from the inverter to the antenna, and based on the sensed output current, selects one of a plurality of output voltage patterns in which the first voltage and the second voltage are output from the inverter, and controls the operation of the plurality of switches so that the first voltage or the second voltage is applied to the antenna according to the selected output voltage pattern, - wherein each of the plurality of output voltage patterns includes at least one powering section that outputs the first voltage or the second voltage and a freewheeling section that does not output both the first voltage and the second voltage - The controller selects a first output voltage pattern in which the total length of the freewheeling section within the output voltage pattern is greater than or equal to a first length in response to the sensed output current indicating that the plasma is ignited, and controls the operation of the plurality of switches according to the selected first output voltage pattern, - wherein the selected first output voltage pattern is such that the current flowing through each of the plurality of switches before the plasma is ignited, despite an inductance less than or equal to a predetermined value By making each of the plurality of switches below the maximum allowable current, and selecting a second output voltage pattern in which the total length of the freewheeling section within the output voltage pattern is less than the first length from the point in time when the sensed output current indicates that the plasma has been ignited, the operation of the plurality of switches can be controlled according to the selected second output voltage pattern.

[0046] Additionally, if the sensed output current is greater than or equal to a predetermined value, it may indicate that the plasma has not yet been ignited, and if the sensed output current is less than the predetermined value, it may indicate that the plasma has been ignited.

[0047] In addition, the controller can control the operation of the plurality of switches according to the replaced second output voltage pattern by repeatedly replacing the second output voltage pattern such that the total length of the freewheeling section included in the second output voltage pattern gradually decreases from the point in time when the sensed output current indicates that the plasma has been ignited.

[0048] In addition, the controller can control the operation of the plurality of switches according to the second output voltage pattern being replaced, while repeatedly replacing the second output voltage pattern until it is replaced with the second output voltage pattern consisting of the power ring section without the freewheeling section.

[0049] In addition, after the second output voltage pattern without the freewheeling section is selected, the controller can control the operation of the plurality of switches according to the second output voltage pattern without the freewheeling section so that the metal precursor is decomposed at a rate greater than the expected decomposition rate.

[0050] In addition, even if the second output voltage pattern without the freewheeling section is selected, the current flowing through each of the plurality of switches may be less than or equal to the maximum allowable current due to the increase in impedance caused by the ignited plasma.

[0051] In addition, the above expected decomposition rate may be 90% or more.

[0052] Additionally, the exhaust gas decomposition device may further include a gas injection tube for injecting oxygen into the internal space of the discharge tube to oxidize the metal precursor; and a trap fluidically connected to the outlet to collect the metal precursor oxidized by the oxygen.

[0053] In addition, the inlet of the discharge tube is fluidically connected to the piping of the semiconductor process chamber through which the metal precursor is discharged, and the diameter of the discharge tube may correspond to the diameter of the piping.

[0054] Additionally, the antenna is composed of a plurality of layers corresponding to each of a plurality of planes perpendicular to the virtual central axis of the discharge tube, and each of the plurality of layers includes a turn antenna segment having a certain radius from the virtual central axis, wherein the combined inductance of the turn antenna segments included in each of the plurality of layers may be less than or equal to the predetermined value.

[0055] Additionally, the controller controls the operation of the plurality of switches by inputting a control signal that controls the operation of the plurality of switches to the inverter, wherein the controller assumes a virtual signal that is shifted by a predetermined time for the control signal, calculates a delay time between the virtual signal and the output current, determines a driving frequency corresponding to a resonant frequency that varies according to at least one of the antenna, the discharge tube, and the plasma based on the delay time, and controls the operation of the plurality of switches so that the first voltage or the second voltage is output based on the driving frequency.

[0056] [The Necessity of Exhaust Gas Decomposition]

[0057] As explained earlier in [Technical Background of the Invention], when exhaust gases such as unreacted metal precursors are discharged from a semiconductor process chamber, they deposit on pipes or pumps. This deposition of exhaust gases prevents the pipes and pumps from being used for an extended period, as it can clog them or cause operational malfunctions. Consequently, the replacement cycle of pipes or pumps is shortened, inevitably leading to a continuous increase in maintenance costs for the semiconductor process system.

[0058] Furthermore, semiconductor processes are highly sensitive to even minute environmental changes, making precise control of the process environment crucial. However, exhaust gases deposited on pipes or pumps affect the pressure within the piping and the connected semiconductor process chambers, preventing the semiconductor process environment from being controlled as intended and consequently causing a degradation in the quality of the produced semiconductors.

[0059] As explained earlier in [Technology Forming the Background of the Invention], as the production of semiconductors using the HKMG (High-K Metal Gate) process increases, the use of metal precursors will continue to grow. Accordingly, the necessity of decomposing exhaust gases, such as unreacted metal precursors, to prevent deposition on pipes or pumps is bound to increase, as this extends the replacement cycle of pipes or pumps, thereby reducing maintenance costs for the semiconductor process system and maintaining the quality of the produced semiconductors.

[0060]

[0061] [Method of decomposing exhaust gases using plasma]

[0062] There are various methods for decomposing exhaust gases. For example, (1) exhaust gases can be decomposed by applying high temperature to the pipes. However, in order to decompose exhaust gases, very high heat of at least 1500 degrees Celsius to 3000 degrees or more must be applied to the pipes, and such ultra-high temperatures as described above can cause the pipes to melt or the very high heat to be transferred to the semiconductor process chamber, which can affect the semiconductor process.

[0063] Meanwhile, although there is a method to decompose exhaust gases using catalysts, this requires replacing the catalyst placed within the piping to match the type of exhaust gas whenever the type changes. This entails the inconvenience of continuously disassembling and reassembling semiconductor process equipment whenever the type of exhaust gas changes. Furthermore, as semiconductor process equipment is highly delicate, continuous disassembly and reassembly can cause misalignment between the devices, potentially affecting the semiconductor process environment.

[0064] Meanwhile, methods to dissolve or neutralize exhaust gases using liquid solvents can also be considered. However, removing secondary liquid contaminants resulting from the dissolution of exhaust gases is a very difficult task, and using liquid solvents has the disadvantage of requiring very large equipment.

[0065] In contrast, decomposing exhaust gases using plasma is a highly efficient method. When exhaust gases are decomposed using plasma, it is possible to decompose them at relatively low temperatures of approximately 500 to 1,000 degrees Celsius. Furthermore, once a plasma generator is installed, its replacement cycle is long, eliminating the need to frequently disassemble and reassemble semiconductor process equipment. Additionally, when exhaust gases are oxidized, they undergo a phase change into a solid oxide powder form, making them much easier to remove compared to liquid contaminants.

[0066] Accordingly, development of exhaust gas decomposition devices using plasma generators has already been carried out. However, as can be seen in FIGS. 1 and FIGS. 8 of Korean Patent Publication No. 10-2022-0039128 (KR 2022-0039128 A), the discharge tube of the existing exhaust gas decomposition device (100) has an inlet (170a) fluidically connected to a semiconductor process chamber, but has a toroidal structure that branches into two plasma channels (133) from the inlet (170a). That is, plasma is generated in a ring shape within the two plasma channels (133) branched from a single inlet (170a). However, the two plasma channels (133) branched as described above can have a significant impact on the process environment of the semiconductor process chamber (20). Specifically, exhaust gas discharged from the semiconductor process chamber (20) flows into the exhaust gas decomposition device through the inlet (170a) and branches into two plasma channels (133). When branching, the direction of movement of the exhaust gas changes, causing a change in the flow velocity of the exhaust gas. Additionally, since the toroidal structure disclosed in the above Korean published patent must generally branch into two plasma channels (1300), the diameter of the plasma channel (133) is smaller than the diameter of the inlet (170a). In this case, the flow of the exhaust gas passing through the inlet (170a) and entering the plasma channel (133) is restricted due to a bottleneck effect. The change in the flow velocity of the exhaust gas and the restriction of flow due to the bottleneck effect described above cause changes in the hydraulic pressure inside the passage through which the exhaust gas is discharged and the semiconductor process chamber (20), and consequently, cause changes in the semiconductor process environment, which may have a negative impact on the quality of the semiconductor product being produced.

[0067]

[0068] [Structure of an exhaust gas decomposition device according to the present disclosure]

[0069] The applicant manufactures and sells a Remote Plasma Source (RPS) that generates radicals and ions using plasma at a distance from a semiconductor process chamber and transfers the generated radicals and ions to the semiconductor process chamber for use in various semiconductor processes. The applicant's RPS is a device comprising a cylindrical discharge tube and an antenna surrounding the cylindrical discharge tube, and is an Inductive Coupled Plasma (ICP) device that generates plasma in the internal space of the cylindrical discharge tube by applying RF power to the antenna.

[0070] Meanwhile, for linear discharge tubes, such as cylindrical discharge tubes, if their shape is identical to that of the piping, the discharge tube itself can function as part of the piping, thereby enabling the creation of a system that decomposes exhaust gases without significantly altering the structure and environment of the existing semiconductor process system. In other words, the linear discharge tube structure provides an environment where exhaust gases discharged from the semiconductor process chamber can flow smoothly without changes in flow velocity or bottlenecks, thereby providing a highly effective structure that decomposes exhaust gases while maintaining the quality of semiconductor products by minimizing the impact on the existing semiconductor process environment.

[0071] Accordingly, the applicant intends to develop an exhaust gas decomposition device by utilizing the structure of the existing RPS device manufactured and sold by the applicant. However, considering the environment in which the exhaust gas decomposition device is installed, the characteristics to be taken into account, and the required performance, it is difficult to utilize the existing RPS device as is as an exhaust gas decomposition device, and various problems are encountered in developing an exhaust gas decomposition device utilizing an ICP plasma device with a cylindrical discharge tube.

[0072] Therefore, below, we will examine the various problems faced by the applicant while developing an exhaust gas decomposition device utilizing an ICP plasma device having a linear discharge tube, and the solutions to the problems faced.

[0073] However, prior to that, in order to explain the various problems encountered while developing the ICP plasma device with the aforementioned linear discharge tube into an exhaust gas decomposition device and to facilitate understanding of the solutions, we will examine the exhaust gas decomposition system including the exhaust gas decomposition device and explain the configuration of the exhaust gas decomposition device.

[0074]

[0075] [Exhaust gas decomposition system using plasma according to the present disclosure]

[0076] FIG. 1 is a drawing for illustrating an exhaust gas decomposition system including an exhaust gas decomposition device (100) according to the present disclosure as described above. Referring to FIG. 1, the exhaust gas decomposition system includes an exhaust gas decomposition device (100), a semiconductor process chamber (200), an additional gas supply line (300), a trapper (400), a pump (500) and a scrubber (600), at least one pipe (700), at least one gas supply line (10), and at least one valve (20).

[0077] A semiconductor process chamber (200) is a device for providing a space where various processes for producing semiconductors are performed. Inside the semiconductor process chamber (200), a substrate support for loading a substrate may be included. Inside the semiconductor process chamber (200), at least one semiconductor process among various semiconductor processes such as deposition, ion implantation, heat treatment, etching, cleaning, and plasma treatment may be performed. Meanwhile, the semiconductor process is highly sensitive to the internal environment of the semiconductor process chamber (200), such as the type and flow rate of gas introduced into the semiconductor process chamber (200), and the temperature and pressure inside the semiconductor process chamber; accordingly, the internal environment of the semiconductor process chamber (200) needs to be controlled very precisely.

[0078] Process gas required for each semiconductor process can be supplied into the semiconductor process chamber (200) through at least one gas supply pipe (10). For example, a metal precursor for depositing a semiconductor gate insulating film and a metal gate can be supplied to the semiconductor process chamber (200) through the at least one gas supply pipe (10). Additionally, a purge gas for removing byproducts or residual gas generated during the deposition process and for controlling the reaction of the metal precursor can be supplied to the semiconductor process chamber (200) through the at least one gas supply pipe (10). For example, nitrogen gas can be supplied to the semiconductor process chamber (200) through the at least one gas supply pipe (10) as the purge gas.

[0079] At least one valve (20) can control the flow rate of the processing gas and purge gas supplied to the semiconductor process chamber (200) through the at least one gas supply pipe (10). For example, the at least one valve (20) can control the flow rate of the processing gas and purge gas to control the pressure and gas concentration, etc., required by the semiconductor process performed within the semiconductor process chamber (200).

[0080] The pump (500) is a device for controlling the pressure inside the semiconductor process chamber (200) to a constant level. For example, the pump (500) enables the pressure inside the semiconductor process chamber to be maintained at a vacuum or a low pressure close to a vacuum. Maintaining the pressure inside the semiconductor process chamber (200) to a vacuum or a low pressure close to a vacuum enables the gas density and gas concentration inside the semiconductor process chamber (200) to be maintained at a constant level, thereby increasing the efficiency of the semiconductor process and maintaining the quality of the semiconductor. Additionally, the pump (500) enables exhaust gas to be discharged from the semiconductor process chamber (200) and allows the exhaust gas to exit to the outside of the exhaust gas decomposition system. For example, the pump (500) may be one of a turbo molecular pump, a Roots pump, a dry pump, a liquid ring pump, or a scroll pump, but is not limited thereto, and any pump capable of maintaining the internal pressure of the semiconductor process chamber (200) at a vacuum or a low pressure close to a vacuum may be used as the pump (500) according to the present disclosure.

[0081] A scrubber (600) is positioned downstream of a pump (500) and is a device for purifying and treating exhaust gas discharged from the pump (500) before it escapes to the outside, thereby preventing environmental pollution (e.g., air pollution) caused by the exhaust gas. For example, the scrubber (600) may purify the exhaust gas using water or a chemical solution, or purify the exhaust gas by collecting dust using an adsorbent or a solid chemical substance. As another example, the scrubber (600) may also treat the exhaust gas using pyrolysis or thermal oxidation phenomena.

[0082] The exhaust gas decomposition device (100) is positioned between the semiconductor process chamber (200) and the pump (500) and is a device for decomposing and oxidizing exhaust gas discharged from the semiconductor process chamber (200). When exhaust gas is introduced into the exhaust gas decomposition device (100), the exhaust gas decomposition device (100) decomposes the exhaust gas using ignited plasma. At this time, the decomposed exhaust gas reacts with oxygen supplied to the exhaust gas decomposition device (100) and is oxidized into a solid form (e.g., powder). Hereinafter, the exhaust gas compound oxidized into a solid form is referred to as a solid oxide.

[0083] The solid oxide generated from the exhaust gas decomposition device (100) is transferred to the collector (400) by the pulling force of the pump (500) and / or gravity, and the unoxidized gas among the decomposed exhaust gas is transferred to the pump (500) by the pulling force of the pump (500). For example, when the exhaust gas decomposition device (100) receives an unreacted metal precursor or process by-product, it decomposes the unreacted metal precursor or process by-product using plasma. At this time, the metal ions decomposed from the unreacted metal precursor or process by-product are oxidized into solid metal oxides within the plasma and transferred to the collector (400), and the unoxidized gas is transferred to the pump (500). Accordingly, the unoxidized gas does not contain metal precursor components and is not deposited on the pipe (700) and pump (500). Accordingly, the lifespan of the pipe (700) and pump (500) can be extended and the replacement cycle can be extended.

[0084] The method of igniting plasma by the exhaust gas decomposition device (100) described above and the method of effectively decomposing exhaust gas using plasma will be described in detail later.

[0085] The additional gas supply pipe (300) is a passage for supplying additional gas to the exhaust gas decomposition device (100). Here, the additional gas may be a gas that is required to ignite the plasma or process the exhaust gas but is not supplied from the semiconductor process chamber (200). Alternatively, the additional gas may be a gas that is supplied from the semiconductor process chamber (200) but requires a higher flow rate for igniting the plasma or processing the exhaust gas. For example, the additional gas may be nitrogen for igniting the plasma or oxygen for oxidizing the exhaust gas.

[0086] The collector (400) is a device that is fluidically connected to an outlet of the exhaust gas decomposition device (100) and collects solid oxides from the exhaust gas decomposition device (100). Since the solid oxides collected in the collector (400) are in the form of a solid (e.g., powder), the user can easily remove the solid oxides collected from the collector.

[0087] Meanwhile, as shown in FIG. 1, the collector (400) may be placed between the exhaust gas decomposition device (100) and the pump (500). In this case, the unoxidized gas may be transferred to the pump (500) via the collector (400). However, the collector (400) does not necessarily have to be placed between the exhaust gas decomposition device (100) and the pump. For example, the collector (400) may be fluidly connected to the outlet of the exhaust gas decomposition device (100) but not connected to the pump (500). In this case, the unoxidized gas may be transferred to the pump (500) without passing through the collector (400) via a pipe connected between the exhaust gas decomposition device (100) and the pump (500).

[0088] Meanwhile, the exhaust gas decomposition device (100), semiconductor process chamber (200), collector (400), pump (500), and scrubber (600) can be fluidly connected through at least one pipe (700). For example, the semiconductor process chamber (200) and the exhaust gas decomposition device (100) may be connected through a first pipe, the exhaust gas decomposition device (100) and the collector (400) may be connected through a second pipe, the collector (400) and the pump (500) may be connected through a third pipe, and the pump (500) and the scrubber (600) may be connected through a fourth pipe. Meanwhile, as described above, the exhaust gas decomposition device (100) and the collector (400) may be connected through a second pipe, and the exhaust gas decomposition device (100) and the pump (500) may be connected through a third pipe.

[0089]

[0090] [Composition of the exhaust gas decomposition device]

[0091] Below, we will specifically examine an exhaust gas decomposition device (100) that is a plasma device having a linear discharge tube as described above and is included in the exhaust gas decomposition system described above to decompose exhaust gas.

[0092] FIG. 2 is a drawing showing an exhaust gas decomposition device (100) according to the present disclosure. The exhaust gas decomposition device (100) is composed of an RFG (Radio Frequency Generator, 1000) and a load (2000). The RFG (1000) is electrically connected to an antenna structure (2100) of the load (2000) and can provide voltage, current, or power to the antenna structure (2100). Meanwhile, the current provided to the antenna structure (2100) can supply power or electromotive force to the load (2000) to induce the generation of plasma in the internal space of a discharge tube (2200).

[0093] The load (2000) includes an antenna structure (2100) and a discharge tube (2200). In addition, when the RFG (1000) supplies power to the load (2000), the load (2000) may further include a plasma generated inside the discharge tube (2200), a parasitic capacitor, and a parasitic inductor. The parasitic capacitor and the parasitic inductor may be formed within the antenna structure (2100) when power is supplied to the load (2000). Furthermore, the parasitic capacitor and the parasitic inductor may be formed between the antenna structure (2100) and the discharge tube (2200). Additionally, the parasitic capacitor and the parasitic inductor may be formed between the antenna structure (2100) and the plasma. Accordingly, the impedance of the load (2000) in this disclosure may be defined as (i) a composite impedance due to the inductor and capacitor included in the antenna structure (2100), (ii) an impedance due to the plasma generated inside the discharge tube (2200), and (iii) a total composite impedance due to the composite impedance due to the parasitic capacitor and parasitic inductor described above. Meanwhile, since the state of the plasma, the impedance of the plasma, and the parasitic capacitor and parasitic inductor may vary over time, the impedance of the load may also vary over time accordingly.

[0094] Let us examine in detail the configuration of the antenna structure (2100), the discharge tube (2200), and the RFG (1000).

[0095] [1] Antenna structure (2100) and discharge tube (2200)

[0096] FIGS. 3 to 5 are drawings for explaining the configuration of an antenna structure (2100) and a discharge tube (2200) according to the present disclosure. FIG. 3 is a drawing for explaining the arrangement of the antenna structure (2100) and the discharge tube (2000) and the connection relationship with other components.

[0097] Referring to FIG. 3, the antenna structure (2100) may be arranged to wrap around the outer wall of the discharge tube (2200). The discharge tube (2200) may have an inlet (2210) and an outlet (2220). As described above, the discharge tube (2200) may be a straight tube extending from the inlet (2210) to the outlet (2220). That is, in the present disclosure, the inlet (2210) and the outlet (2220) refer to the two ends of the discharge tube (2200). The inlet (2210) and the outlet (2220) are merely descriptions to distinguish the end where gas is introduced and the end where gas is discharged, respectively, for the convenience of explaining the examples according to the present disclosure, and do not imply that the inlet (2210) and the outlet (2220) are configured as separate devices from the discharge tube (2200). Furthermore, since the discharge tube (2200) is a straight tube, the diameter of the discharge tube (2200), the diameter of the inlet (2210), and the diameter of the outlet (2220) are all the same. That is, a straight tube may mean a tube in which all diameters of the discharge tube (2200) are substantially the same, with an inlet (2210) and an outlet (2200) formed between the inlet (2210) and the outlet (2220). Accordingly, in the following description, the diameter of the discharge tube (2200) should be understood to include the diameter of the inlet (2210) and the diameter of the outlet (2220).

[0098] The inlet (2210) of the discharge tube (2200) is connected to the semiconductor process chamber (200) through the first pipe (700a), and the outlet (2220) of the discharge tube (2200) can be connected to the collector (400) and / or pump (500) through the second pipe (700b). Exhaust gas from the semiconductor process chamber (200) is introduced into the internal space of the discharge tube (2200) through the inlet (2210) via the first pipe (700a), and solid oxides and unoxidized gases generated by decomposition in the internal space can be discharged to the second pipe (700b) via the outlet (2210).

[0099] Meanwhile, plasma can be generated in the internal space of the discharge tube (2200). Additionally, the internal space of the discharge tube (2200) can form part of a gas passage connecting the semiconductor process chamber (200) to the scrubber (600).

[0100] Meanwhile, as described above, the semiconductor process can be highly sensitive to pressure. However, if the diameter of the discharge tube (2200) is much smaller than the diameter of the pipe (700), or if the internal space of the discharge tube (2200) through which gas flows is branched or has a very severe curvature, then due to this structure of the discharge tube (2200), when the exhaust gas passes through the discharge tube (2200), it causes a large change in pressure, and the caused change in pressure affects the pressure of the semiconductor process chamber (200), which can affect the quality of the semiconductor process. Therefore, if the discharge tube (2200) has a straight structure without curvature and its diameter and shape are the same as or at least similar to the pipe (700), it can avoid affecting the pressure of the semiconductor process chamber (200), which is advantageous for maintaining the quality of the semiconductor process. Accordingly, the discharge tube (2200) may have a cylindrical structure extending from the inlet (2210) to the outlet (2220), and it is preferable that the diameter of the cylinder be similar to the diameter of the pipe (700). For example, the diameter of the discharge tube (2000) may be the same as the diameter of the pipe (700). Or, the diameter of the discharge tube (2000) may be smaller within a 1% range than the diameter of the pipe (700). Or, the diameter of the discharge tube (2000) may be smaller within a 3% range than the diameter of the pipe (700). Or, the diameter of the discharge tube (2000) may be smaller within a 5% range than the diameter of the pipe (700). Or, the diameter of the discharge tube (2000) may be smaller within a 7% range than the diameter of the pipe (700). Alternatively, the diameter of the discharge tube (2000) may be within a 10% range smaller than the diameter of the pipe (700). Alternatively, the diameter of the discharge tube (2000) may be within a 13% range smaller than the diameter of the pipe (700). Alternatively, the diameter of the discharge tube (2000) may be within a 15% range smaller than the diameter of the pipe (700). Alternatively, the diameter of the discharge tube (2000) may be within a 20% range smaller than the diameter of the pipe (700).

[0101] Meanwhile, in the above description and the description below, the discharge tube (2200) is described as being cylindrical, but is not limited thereto and may have various shapes depending on the shape of the pipe (700). For example, if the pipe (700) is a square pipe, the discharge tube (2200) may also be a square pipe, and if the pipe (700) is a triangular pipe, the discharge tube (2200) may also be a triangular pipe. Furthermore, in the case where the shape of the pipe (700) and the discharge tube (2000) are polygonal pipes such as a square pipe or a triangular pipe, the diameter of the discharge tube (2000) and the diameter of the pipe (700) in the above description or the description below may be interpreted as the inradius diameter of the cross-section of the pipe (700) or the discharge tube (2000).

[0102] Meanwhile, the discharge tube (2200) can be made of various materials. For example, the discharge tube (2200) can be manufactured from a non-conductive or highly thermally conductive material. For example, the discharge tube (2200) can be manufactured from ceramics such as aluminum nitride (AlN), aluminum oxide (Al2O3), silicon nitride (SiN), silicon nitride (Si3N4), silicon dioxide (SiO2), yttrium oxide (Y2O3), or silicon carbide (SiC). Furthermore, the discharge tube (3000) can be manufactured from a material that does not generate impurities (particles) by reacting with the gas introduced into the discharge tube (3000) for plasma induction.

[0103] Additionally, the additional gas supply pipe (300) may be inserted into the first pipe (700a) so that additional gas is supplied through the inlet (2210) of the discharge tube (2200) as in FIG. 3 (a). Alternatively, as in FIG. 3 (b), it may be inserted into the discharge tube (2200) so that additional gas is supplied directly into the internal space of the discharge tube (2200). For example, the additional gas supply pipe (300) may be inserted into the discharge tube (2200) between the inlet (2210) and the antenna structure (2100).

[0104] Referring to FIGS. 4 and 5, let us examine an antenna structure (2100) arranged to surround a discharge tube (2200). Referring to FIG. 4, a layer antenna (2110) is arranged on each of a plurality of different virtual planes perpendicular to a virtual central axis extending from the inlet (2210) to the outlet (2220) of the discharge tube (2200), and the antenna structure (2100) is formed by connecting the layer antennas (2110) arranged on each plane. Meanwhile, the layer antenna (2110) may be configured to have two or more turns. For example, the layer antenna (2110) may include a first turn and a second turn having different radii, the first turn being arranged closer to the virtual central axis than the second turn, and the radius of the first turn may be smaller than the radius of the second turn.

[0105] Similarly to the above, the layer antenna (2110) may be composed of three or more turns. In this case, each turn has a different radius, and the radius of each turn may be smaller as each turn is positioned closer to a virtual central axis. Meanwhile, each turn constituting the layer antenna (2110) in the present disclosure may be referred to as a Turn Antenna Segment.

[0106] Meanwhile, at least one inter-layer capacitor (2131 to 2134) may be electrically connected between layer antennas. That is, at least one inter-layer capacitor (2131 to 2134) may be electrically interposed between layer antennas, which means that at least one inter-layer capacitor (2131 to 2134) electrically connects layer antennas to layer antennas. For example, the inter-layer capacitor (2131 to 2134) may be placed between layer antennas. These inter-layer capacitors (2131 to 2134) lower the potential of the antenna structure (2000) and distribute the voltage applied to the antenna structure (2100). This prevents damage to the discharge tube (2200) by causing the potential of the layer antenna placed closest to the inlet (2210) (or outlet (2220)) of the discharge tube (2200) to become much higher than the potential of the layer antenna placed closest to the outlet (2220) (or inlet (2210)) of the discharge tube (2200). Meanwhile, to further enhance the voltage distribution effect of these interlayer capacitors (2131 to 2134), an interlayer capacitor (not shown) may be further placed between the turn antenna segments within the layer antenna (2110). Meanwhile, FIG. 4 shows, for example, a layer antenna composed of two turns placed on each of five different planes, but the configuration of the antenna structure (2100) according to the present disclosure is not limited thereto. For example, the antenna structure (2100) may be composed of layer antennas placed on each of N different planes (N is a natural number), and N may be any one of 6, 7, 8, 9, 10, 11 and 12, but is not limited thereto, and may be 5 or less, or 12 or more.Meanwhile, FIG. 4 is a side view of the discharge tube (2200), and although the interlayer capacitor (2132) is not visually visible, it should be kept in mind that the interlayer capacitor (2132) is obscured by the discharge tube (2220) and is not absent. Additionally, the interlayer capacitors (2131 to 2134) can be configured to electrically connect the outermost turn of one layer antenna and the innermost turn of another layer antenna, as described in FIG. 5. An example of how these interlayer capacitors (2131 to 2134) electrically connect the layer antennas will be examined with reference to FIG. 5.

[0107] FIG. 5 is intended to illustrate an example of a connection in which interlayer capacitors (2131 to 2134) are electrically connected between layer antennas (2110) arranged on different planes. In FIG. 5, for convenience of explanation, the layer antennas (2110) are arranged in a single drawing, but it should be understood that they are arranged to surround the discharge tube (2200) as described in FIG. 4. Referring to FIG. 5, the layer antennas (2110) have ends of outermost turn antenna segments (hereinafter, outermost end, 2112) and ends of innermost turn antenna segments (hereinafter, innermost end 2111), said ends (2111, 2112) are connected to the RFG (1000) or are electrically connected to adjacent layer antennas (2110). Meanwhile, when arranging the layer antennas (2110), the ends (2111, 2112) that are connected to the RFG (1000) or electrically connected to the adjacent layer antenna (2110) can be arranged so as not to overlap each other when viewed from the inlet (2210) (or outlet (2220)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)).

[0108] For example, assuming a virtual first line connecting the innermost end (2111) (or outermost end (2112)) of a layer antenna (2110) placed on a first plane and the virtual central axis shown in FIG. 4, and assuming a virtual second line connecting the innermost end (2111) (or outermost end (2112)) of a layer antenna (2110) placed on a second plane adjacent to the first plane and the virtual central axis shown in FIG. 4, the layer antennas (2110) can be placed on each plane such that when viewed from the inlet (2210) (or outlet (2220)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)), the angle between the virtual first line and the virtual second line becomes a predetermined angle. Additionally, assuming a virtual N line connecting the innermost end (2111) (or outermost end (2112)) of a layer antenna (2110) placed on the Nth plane as described above and the virtual central axis shown in FIG. 4, and assuming a virtual N+1 line connecting the innermost end (2111) (or outermost end (2112)) of a layer antenna (2110) placed on the N+1 plane adjacent to the Nth plane and the virtual central axis shown in FIG. 4, the layer antennas (2110) can be placed on each plane such that when viewed from the inlet (2210) (or outlet (2220)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)), the angle between the virtual N line and the virtual N+1 line becomes a predetermined angle.

[0109] In addition, as described above, with the layer antennas (2110) arranged, the interlayer capacitors (2131 to 2134) can electrically connect the outermost end (2112) of the Nth layer and the innermost end (2111) of the N+1th layer.

[0110] The rotational arrangement of the layer antennas (2110) disclosed in FIGS. 4 and 5 may be useful when the space in which the discharge tube (2200) and the antenna structure (2100) are placed is not sufficiently large and the layer antennas (2110) are composed of multiple turns. When the layer antennas (2110) are composed of multiple turns, each interlayer capacitor (2131 to 2134) must be placed between turns of different radii of adjacent layer antennas (2110). In this case, because the connection between the interlayer capacitors (2131 to 2134) and the layer antennas (2110) is complex, alignment between the interlayer capacitors (2131 to 2134) is not easy, and a relatively large space may be required to easily align the interlayer capacitors (2131 to 2134).

[0111] In addition, although the interlayer capacitors (2131 to 2134) are depicted as small box shapes in FIGS. 4 and 5, in reality, the interlayer capacitors (2131 to 2134) may be structures that occupy a very large volume. If the interlayer antennas (2110) are not rotated in a situation where the volume of the interlayer capacitors (2131 to 2134) is very large, the spacing between the interlayer antennas (2110) will inevitably become very wide due to the interlayer capacitors (2131 to 2134) having a very large volume.

[0112] However, as the spacing between the layer antennas (2110) widens, the number of layer antennas (2110) covering the surface of the discharge tube (2200) in a certain area (i.e., the density of the layer antennas (2110)) inevitably decreases, so the induced electromotive force transmitted inside the discharge tube (2200) decreases.

[0113] Accordingly, in order to increase the induced electromotive force transmitted inside the discharge tube (2200), the interlayer capacitors (2131 to 2134) are positioned so as to be pulled out in a direction away from the discharge tube (2200), and the layer antennas (2110) are rotated so that the interlayer capacitors (2131 to 2134) are distributed along the outer surface perimeter of the discharge tube (2200), thereby preventing the spacing between the layer antennas (2110) from narrowing due to the volume of each interlayer capacitor (2131 to 2134) as much as possible. In other words, rotating the layer antennas (2110) so that the interlayer capacitors (2131 to 2134) are distributed along the outer surface perimeter of the discharge tube (2200) makes the spacing between the layer antennas (2110) as dense as possible and increases the number of layer antennas (2110) covering the surface of the discharge tube (2200) in a certain area (i.e., the density of the layer antennas (2110)), thereby allowing the induced electromotive force transmitted into the discharge tube (2200) to be increased to the maximum.

[0114] Additionally, the rotational arrangement between the layer antennas (2110), as shown in FIGS. 4 and 5, can minimize arcing caused by potential difference. For example, if the antenna (2110) has multiple turns, when viewed from the inlet (2210) (or outlet (2220)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)), if the angle between the virtual N-th line and the virtual N+1-th line is 0 degrees, the distance between the innermost end of the N-th layer and the outermost end of the N+1-th layer is arranged very close, while the potential difference is maximized, which may increase the likelihood of arcing occurring. Similarly, arcing may occur due to the potential difference between the M-th turn of the N+1-th layer and a turn other than the M-th turn of the N-th layer. In particular, a significant arcing phenomenon may occur between a point in the M turn of the N+1th layer and a point in a turn other than the M turn of the Nth layer where the straight-line distance from the said point is minimized.

[0115] Accordingly, the layer antenna (2110) and the interlayer capacitors (2131 to 2134) are arranged so that this potential difference is minimized, thereby minimizing the arcing phenomenon. For example, when the antenna (2110) between the layer antennas (2110) has multiple turns, when looking from the inlet (2210) (or outlet (2220)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)), the angle between the virtual Nth line and the virtual N+1th line can be determined as an angle that minimizes the arcing phenomenon. For example, the layer antennas (2110) and interlayer capacitors (2131 to 2134) can be arranged by determining the angle between the virtual lines so that the potential difference between different turns between the layer antennas (2110) included in the antenna structure (2100) is minimized.

[0116]

[0117] [2] Composition of RFG(1000)

[0118] FIGS. 6 and 7 are drawings for explaining the configuration of RFG (1000) according to the present disclosure.

[0119] Referring to FIG. 6, the RFG (1000) according to the present disclosure may include an AC power source (1100), a rectifier (1200), an inverter (1300), a controller (1500), and a sensor module (1400). The RFG (1000) may convert the AC supplied from the AC power source (1100) into another AC and supply it to a load (2000). For example, the RFG (1000) may convert the AC used in a typical household or industrial setting into another AC having a frequency of several hundred kHz to several tens of MHz and a power of several kW or more, and supply it to the load. Hereinafter, the power supplied from the RFG (1000) to the load (2000) refers to RF power unless otherwise specified.

[0120] Meanwhile, the RFG (1000) and the load (2000) may be electrically and / or physically connected by one or more nodes. For example, the RFG (1000) and the antenna structure (2100) may be physically and / or electrically connected so that the RFG (1000) can supply power to the load (2000). Referring to FIG. 6, two terminals (1610, 1630) of the RFG (1000) and two terminals (1710, 1730) of the antenna structure (2100) may be electrically and / or physically connected. For example, terminal (1610) may be connected to terminal (1710), and terminal (1630) may be connected to terminal (1730). For example, terminal (1610) may be connected to terminal (1710) by one wire, and terminal (1630) and terminal (1730) may be connected by another wire.

[0121] Hereinafter, for convenience of explanation, it is assumed that the RFG (1000) and the antenna structure (2100) are connected through terminals (1610, 1630, 1710, 1730), and that power supplied by the RFG (1000) is delivered to the load (2000). That is, for convenience of explanation, the configurations of the terminals (1610, 1630, 1710, 1730) may be omitted from the drawings or from the specific description in the specification, but it can be easily inferred by the above description and by a person skilled in the art that the RFG (1000) and the antenna structure (2100) are connected through the terminals (1610, 1630, 1710, 1730).

[0122] The rectifier (1200) can convert the output of the AC power source (1100) into DC. The rectifier (1200) can convert the AC supplied from the AC power source (1100) into DC and apply it to both ends of the inverter (1300) (e.g., the input ends of the inverter (1300)).

[0123] The inverter (1300) can receive direct current from the rectifier (1200) and supply alternating current to the antenna structure (2100). For example, the inverter (1300) can receive a control signal from the controller (1500) and use the received control signal to supply alternating current to the load. Here, the inverter (1300) may include at least one switch element controlled by the control signal, and the alternating current supplied from the inverter (1300) to the load may have a specific frequency (e.g., a driving frequency) set based on the control signal provided to the inverter (1300) by the controller (1500). According to some embodiments disclosed by this application, the inverter (1300) may be a full bridge inverter. Accordingly, the inverter (1300) may include first to fourth switches (S1, S2, S3, S4), as shown in FIG. 7.

[0124] Here, each of the first to fourth switches (S1, S2, S3, S4) can be turned on or turned off by receiving a control signal from the controller (1500). At this time, when the first and third switches (S1, S3) are turned on and the second and fourth switches (S2, S4) are turned off, a positive voltage is applied to the load, and when the first and third switches (S1, S3) are turned off and the second and fourth switches (S2, S4) are turned on, a negative voltage can be applied to the load. In this way, the inverter (1300) can apply an alternating voltage having a specific frequency by alternately applying positive and negative voltages to the load. In this specification, for convenience of explanation, when the electric potential applied to terminals (1610) and (1710) shown in FIG. 6 is higher than the electric potential applied to terminals (1630) and (1730) shown in FIG. 6, the voltage applied to the antenna structure (2100) is described as a "positive voltage," and when the electric potential applied to terminals (1610) and (1710) shown in FIG. 6 is lower than the electric potential applied to terminals (1630) and (1730) shown in FIG. 6, the voltage applied to the antenna structure (2100) is described as a "negative voltage." Additionally, the terminal (1610) shown in FIG. 6 is connected to a node (1310) located between the first switch (S1) and the switch (S2) shown in FIG. 7, and the terminal (1630) shown in FIG. 6 is connected to a node (1320) located between the third switch (S3) and the fourth switch (S4) shown in FIG. 7.

[0125] Meanwhile, a capacitor may be placed between the rectifier (1200) and the inverter (1300). For example, the RFG (1000) includes a capacitor connected in parallel with the rectifier (1200) and the inverter (1300), and the capacitor can discharge the alternating current component of the voltage or current applied to the inverter (1300) to a ground node (GND). Here, the ground node is a node that provides a reference voltage, and the reference voltage may be 0, but may have other values.

[0126] The controller (1500) can generate a control signal by receiving data sensed from the sensor module (1400) described later. For example, the controller (1500) may be implemented to generate a control signal by acquiring data related to the resonant frequency, such as the current and voltage of the load, from the sensor module (1400). The controller (1500) may be implemented using Field Programmable Gate Arrays (FPGA) technology. The specific configuration and structure of the controller (1500) will be described later.

[0127] The sensor module (1400) can obtain data regarding the resonant frequency of the load or information regarding the current supplied to the load from the controller (1500). Referring again to FIG. 6, the sensor module (1400) may include a current transformer (1410), a filter (1420), and a comparator (1430). The sensor module (1400) can receive a current or voltage signal flowing to the load through the current transformer (1410), convert it into a current or voltage signal of a different magnitude, filter the converted current or voltage signal using the filter (1420), and output information regarding the output current to the controller (1500) through the comparator (1430). For example, the information regarding the output current may be current phase information of the output current.

[0128] Here, the current transformer (1410) can be inductively coupled to the wiring between the inverter (1300) and the antenna structure (2100) and can convert a voltage or current signal applied to the antenna structure (2100) and provide it to the filter (1420). Specifically, the current transformer (1410) can convert the current flowing through the wire connected to the antenna structure (2100) into a voltage signal.

[0129] Here, the filter (1420) can remove high-frequency noise from the input current or voltage signal and output it to the comparator (1430). To do this, the filter (1420) can perform high-pass filtering or low-pass filtering.

[0130] Here, the comparator (1430) can obtain information regarding the phase of the output current. For example, the comparator (1430) can obtain information regarding the phase of the output current by comparing a voltage signal obtained from the current transformer (1410) or filter (1420) with a preset value. At this time, the information regarding the phase of the output current may refer to data regarding the phase and direction of the current applied to the load (i.e., output from the inverter (1300)). At least one of the components included in the above-described sensor module (1400) may be omitted.

[0131] Meanwhile, although not shown in FIG. 6, the RFG (1000) may include memory. Here, the memory may store various types of data. Various types of data may be stored in the memory temporarily or semi-permanently. Examples of memory may include a hard disk drive (HDD), a solid state drive (SSD), flash memory, ROM (Read-Only Memory), and RAM (Random Access Memory). The memory may be provided in a form embedded in the RFG (1000) or in a detachable form.

[0132] As described above, the RFG (1000) can control the frequency of the alternating current supplied to the antenna structure (2100) based on data regarding the resonant frequency of the load. At least one of the components of the RFG (1000) described above may be omitted. For example, the RFG (1000) may acquire electrical data regarding the load from an external sensor without including a sensor module (1400). As another example, the RFG (1000) may receive direct current or rectified direct current from an external source without including an alternating current power source (1100) and a rectifier (1200).

[0133]

[0134] Meanwhile, the exhaust gas decomposition device (100) described above must be able to decompose the supplied exhaust gas at a rate greater than the expected decomposition rate. This is because if the exhaust gas is not decomposed at a rate greater than the expected decomposition rate, the problem of metal material being deposited on the pump (500) and pipe (700), which shortens the lifespan of the pump (500) and pipe (700), cannot be effectively solved.

[0135] In addition, the exhaust gas decomposition device (100) should be designed to minimize the impact on the semiconductor process performed within the semiconductor process chamber (200) when decomposing the supplied exhaust gas, thereby maintaining the quality of the semiconductor process.

[0136] Below, we will examine in detail an exhaust gas decomposition device capable of efficiently decomposing the exhaust gas supplied as described above while minimizing its impact on the semiconductor process.

[0137]

[0138] [Plasma Ignition Method in Exhaust Gas Decomposition Device]

[0139] [1] Reasons why plasma ignition is difficult in the exhaust gas decomposition device (100)

[0140] Inert gases (noble gases), such as helium or argon, are primarily used as ignition gases for plasma ignition. Since inert gases are inherently stable atoms, they do not exist in molecular form and, consequently, do not possess 'covalent bonds.' Furthermore, the atoms of inert gases like helium and argon have their electron shells fully saturated with electrons. Therefore, inert gases such as helium and argon are primarily used as ignition gases because it is easy to obtain the 'electrons' necessary to form plasma from the aforementioned inert gases. In particular, argon is the most widely used ignition gas due to its low cost and stable supply compared to other inert gases.

[0141] In addition to inert gases such as argon, other gas molecules like oxygen or nitrogen can also be used as ignition gases; however, because these other gases have many electrons bound by covalent bonds between molecules, more energy is required to detach electrons from the molecules compared to inert gases. Therefore, using gases other than inert gases like argon for plasma ignition is disadvantageous in terms of energy efficiency. Consequently, unless there are special circumstances, it is more advantageous to use inert gases like argon as ignition gases.

[0142] However, the semiconductor process used for integrating semiconductors is highly sensitive to process conditions (i.e., the environment in which the process is carried out), so even if process conditions such as pressure and temperature differ slightly from those required for each process, it becomes difficult to guarantee that the semiconductor of the desired result will be integrated. For this reason, the gas supply / discharge 'environment,' such as the pressure and temperature of the gas supply pipe (10) that supplies processing gas to the semiconductor process chamber (200) through the gas supply pipe (10), and the pressure and temperature of the pipe (700) through which the exhaust gas discharged from the semiconductor process chamber (200) is transported, must be controlled very carefully.

[0143] However, if argon is supplied to the exhaust gas decomposition device (100) through an additional gas supply pipe (300) to ignite the plasma, the argon will have an adverse effect on the process conditions of the semiconductor process chamber (200). This is because argon is not included in the exhaust gas discharged from the semiconductor process chamber (200), and the inflow of a gas other than the exhaust gas (i.e., argon) into the pipe (700) and the exhaust gas decomposition device (100) can cause changes in the pressure or temperature of the pipe (700), and the changed pressure and temperature of the pipe (700) can affect the semiconductor process chamber (200). Furthermore, if argon, which is not included in the exhaust gas, is used as an ignition gas, a significant amount of argon must be supplied to the exhaust gas decomposition device (100), so the impact on the semiconductor process chamber (200) caused by the supply of argon becomes significant. Therefore, it is not desirable to use argon as the ignition gas in the exhaust gas decomposition device (100).

[0144] As described above, in order to maintain the process conditions of each process performed in the semiconductor process chamber (200) and to minimize the impact on the semiconductor process chamber (200), it may be most desirable to use the gas that is already supplied to the semiconductor process chamber (200) through the gas supply pipe (10) and discharged from the semiconductor process chamber (200). However, among the gases supplied to the semiconductor process chamber (200), there are not many gases that can be used for plasma ignition, and considering the absolute amount and chemical / physical properties, nitrogen gas is almost the only gas that can be used as an ignition gas.

[0145] However, since nitrogen gas is fundamentally composed of molecules with strong triple bonds, very high energy must be supplied to it for plasma ignition compared to argon gas, which is a monatomic gas.

[0146] In addition, as described above, if there is a large difference between the diameter of the piping (700) of the semiconductor process chamber that discharges the exhaust gas and the diameter of the discharge tube (2200), a pressure change occurs when the exhaust gas moves from the semiconductor process chamber (200) to the discharge tube (2200), and such a pressure change can make it difficult for the semiconductor to be integrated as desired. Therefore, as seen in the [Exhaust Gas Decomposition System], in order to minimize the pressure change, the diameter of the discharge tube (2200) needs to be very similar to the diameter of the piping (700).

[0147] However, while the diameter of the pipe (700) is generally 10 cm to 25 cm, the diameter of the discharge tube (2200) used in the RPS as described above is 4 cm to 8 cm, which is much smaller than the diameter of the pipe (700). Therefore, using such a discharge tube (2200) as is in the exhaust gas decomposition device (100) can lead to a degradation of semiconductor quality. Accordingly, it was necessary to make the diameter of the discharge tube (2200) larger than the existing diameter of the discharge tube (2200) so that the diameter of the pipe (700) and the diameter of the discharge tube (2200) are similar.

[0148] Meanwhile, as the diameter of the discharge tube (2200) increases, the energy required for plasma ignition increases. When an alternating current is applied to the antenna structure (2100) surrounding the discharge tube (2200), an induced electromotive force is generated in the internal space of the discharge tube (2200), and plasma is induced by this generated induced electromotive force. However, since the induced electromotive force attenuates with distance, if the diameter of the discharge tube (2200) increases, the induced electromotive force cannot affect the center of the discharge tube (2200), making it difficult for the plasma to ignite. Accordingly, as the diameter of the discharge tube (2200) increases, it is required to apply greater energy to increase the magnitude of the induced electromotive force.

[0149] To summarize the above, the exhaust gas decomposition device (100) needs to provide a discharge tube (2200) with a larger diameter than the existing plasma generation device and a much higher energy than the existing plasma generation device to ignite the plasma with nitrogen gas, but there was a great difficulty in delivering a very high energy to the discharge tube (2200) that meets the above description using the energy application method of the existing Induced Coupled Plasma (ICP) device.

[0150]

[0151] [2] Method of generating plasma in other ICP plasma generation devices (e.g., the RPS device described above or an exhaust gas decomposition device having a toroidal discharge tube)

[0152] In other ICP plasma generation devices that used argon gas as the ignition gas and had a relatively small diameter of the discharge tube, after introducing argon gas, the RFG applied an alternating voltage to the antenna structure to ignite the plasma. At this time, the RFG supplied an alternating voltage to the antenna structure having a driving frequency that is equal to or slightly higher than the source resonance frequency. Here, the source resonance frequency is the resonance frequency of the antenna structure, which refers to the resonance frequency determined by the inductances of the turn antenna segments included in the antenna structure and the capacitances of the capacitors (inter-layer and / or inter-turn).

[0153] In addition, other ICP plasma generation devices apply an alternating voltage to the antenna structure in a fixed state without changing the driving frequency until the plasma is ignited, and after the plasma is ignited, change the frequency of the applied alternating voltage periodically or non-periodically to track the resonant frequency of the load fluctuating due to the plasma, thereby enabling energy to be supplied to the ignited plasma while minimizing power loss supplied to the load. This is because, since the impedance of the load changes rapidly due to the plasma after the plasma is ignited, if the driving frequency is not changed, power loss occurs due to the time-varying impedance of the load, and a large amount of energy is not transferred to the plasma, which results in the failure to generate the desired type or amount of active species (e.g., radicals or ions) or the plasma being extinguished.

[0154] On the other hand, since there is no change in load impedance due to the plasma before plasma ignition, there was no need to change the driving frequency, and even if an AC voltage was applied with the driving frequency fixed at a frequency equal to or slightly higher than the source resonance frequency, sufficient energy to ignite the plasma could be delivered inside the discharge tube. In particular, if argon was used as the ignition gas, plasma ignition was possible with relatively little energy, so there was no problem with plasma ignition even if an AC voltage was applied with the driving frequency fixed.

[0155]

[0156] [3] Method to control driving frequency even before plasma ignition

[0157] However, when nitrogen gas is used as the ignition gas and the diameter of the discharge tube (2200) increases, sufficient energy is not transferred to the discharge tube (2200) using the conventional plasma generation method, so the plasma is not ignited.

[0158] Accordingly, while considering various methods to supply sufficient energy to the discharge tube (2200), the applicant realized that even if the plasma is not ignited, the parasitic inductor and parasitic capacitance generated by the voltage applied to the antenna structure (2100) continuously change as nitrogen gas is continuously introduced into the discharge tube (2200), and the impedance of the load (2000) continuously changes, even if it is not large.

[0159] Accordingly, the applicant thought that if the RFG (1000) is controlled so that the driving frequency of the alternating voltage applied to the load (2000) before plasma ignition follows the resonant frequency of the load (2000) that changes slightly over time, then even a little more energy can be delivered to the load (2000), and thus plasma ignition using nitrogen gas may be possible. As a result, the applicant succeeded in igniting plasma using nitrogen gas in the exhaust gas decomposition device (100) according to the present disclosure.

[0160] Accordingly, the present disclosure describes a frequency tracking method for igniting a plasma using nitrogen gas. Here, frequency tracking refers to a technique in which the RFG (1000) controls the driving frequency of the AC voltage applied to the load (2000) to track the resonant frequency of the time-varying load (2000). At this time, the meaning of "tracking" is that the driving frequency of the AC voltage matches the resonant frequency of the load, or the difference from the resonant frequency of the load is controlled to be within a predetermined range.

[0161] FIG. 8 is a drawing for explaining a method for igniting plasma using nitrogen gas in an exhaust gas decomposition device (100) according to the present disclosure.

[0162] Referring to FIG. 8, nitrogen gas is supplied to the internal space of the discharge tube (2200) (S810). At this time, the nitrogen gas may be nitrogen gas discharged from the semiconductor process chamber (200), or nitrogen gas supplied to the internal space of the discharge tube (2200) through an additional gas supply pipe (300). When nitrogen gas is supplied to the discharge tube (2200), the controller (1500) of the RFG (1000) controls the inverter (1300) so that the AC voltage having the source resonance frequency of the load (2000) as the driving frequency is applied to the load (2000) (S830). Here, the definition of the source resonance frequency is omitted as it has been previously described.

[0163] The controller (1500) periodically determines whether the resonant frequency of the load (2000) corresponds to the driving frequency of the AC voltage applied to the load (2000) (S850). The method by which the controller (1500) determines whether the resonant frequency of the load (2000) corresponds to the driving frequency of the AC voltage will be described in detail later through FIG. 9.

[0164] If the controller (1500) determines that the resonant frequency of the load (2000) corresponds to the driving frequency of the AC voltage, the controller (1500) can maintain the driving frequency of the AC voltage without changing it (S870). In other words, the controller (1500) can control the next driving frequency of the AC voltage output in the next cycle to be the same as the current driving frequency of the AC voltage output in the current cycle.

[0165] If the controller (1500) determines that the resonant frequency of the load (2000) and the driving frequency of the AC voltage do not correspond, the controller (1500) may change the driving frequency of the AC voltage and apply an AC voltage having the changed driving frequency to the load (2000) in the next cycle (S890). For example, the controller (1500) may increase or decrease the current driving frequency of the AC voltage output in the current cycle so that the next driving frequency of the AC voltage output in the next cycle follows or corresponds to the resonant frequency of the load (2000).

[0166] Subsequently, the controller (1500) may repeat steps S850 to S890 until the plasma is ignited through at least nitrogen gas. For example, the controller (1500) may repeat steps S850 to S890 until the plasma is ignited, or continue repeating steps S850 to S890 until the exhaust gas is decomposed even after the plasma is ignited.

[0167] Hereinafter, with reference to FIG. 9, we will examine the method by which the controller (1500) determines whether the driving frequency and the resonance frequency correspond in step S850. With reference to FIG. 9, the controller (1500) can obtain voltage phase information of the AC voltage in the current cycle (S851). For example, the controller (1500) can estimate and obtain the voltage phase information of the AC voltage by using one of the control signals applied to the four switches of the inverter (1300) to control the switch operation of the controller (1500). For example, the controller (1500) may assume a virtual signal obtained by shifting a predetermined control signal among the control signals by a predetermined time, and estimate the virtual signal as the voltage phase information of the AC voltage. However, the method of obtaining the voltage phase information of the AC voltage is not limited to the method of estimating by assuming a virtual signal. The controller (1500) may also obtain the voltage phase information of the AC voltage using a separate voltage sensor.

[0168] Additionally, the controller (1500) can obtain current phase information of the output current of the inverter (1300) (S853). A sensor module (1400) may be used to obtain the current phase information of the output current, and since this has been explained in detail in "[2] Configuration of RFG (1000)", a redundant explanation is omitted.

[0169] The controller (1500) determines whether a change in the driving frequency is required in the next cycle based on the acquired voltage phase information and current phase information (S855). For example, the controller (1500) calculates the delay time between the voltage phase information and the current phase information, and if the delay time falls within a predetermined range, the controller (1500) determines that the driving frequency of the AC voltage output in the current cycle is sufficiently following the resonance frequency of the load (2000), and can control the inverter (1300) to output the AC voltage by maintaining the driving frequency of the next cycle as the driving frequency of the current cycle in the next cycle.

[0170] On the other hand, if the above delay time is not included within the above-determined range, the controller (1500) determines that the driving frequency of the AC voltage output in the current cycle is not sufficiently following the resonant frequency of the load (2000), determines a frequency that is higher or lower than the driving frequency of the current cycle by a change unit as the next driving frequency of the next cycle, and can control the inverter (1300) to output an AC voltage having the next driving frequency.

[0171] Here, the aforementioned predetermined range may be a range with a lower limit and an upper limit. For example, the predetermined range may be set to "-R(ns) ≤ delay time ≤ R(ns)", "0 ≤ delay time ≤ R(ns)", or "-R(ns) ≤ delay time ≤ 0". Alternatively, it may be set to "-R1(ns) ≤ delay time ≤ R2(ns)". In this case, R, R1, and R2 can be arbitrarily set by the user based on the difference between voltage phase information and current phase information, which determines that plasma can be ignited even using nitrogen. Additionally, R, R1, and R2 may be set based on the system clock period, which is the inverse of the system clock frequency. For example, R, R1, and R2 may be determined as integer multiples of the system clock period. As a specific example, if the system clock frequency of the controller (1500) is 200 MHz, the system clock period may be 5 ns, so R, R1, and R2 can be set as integer multiples of 5 ns. This is because, as described below, since the controller (1500) outputs the value of the control signal in units of the system clock period, the period of the AC voltage is also controlled in units of the system clock period, and accordingly, the driving frequency of the AC voltage can also be controlled in units of the system clock frequency.

[0172] Meanwhile, the method for determining whether a frequency change is required in S855 is not limited to the method described above. For example, the controller (1500) can determine the driving frequency of the next cycle by decomposing the voltage phase information and current phase information into a DC component and an AC component according to the DQ transformation method based on the voltage phase information and current phase information, reflecting the DC component on the D-axis (Direct Axis) and the AC component on the Q-axis (Quadrature Axis), and estimating the resonance frequency based on the rate of change of the Q-axis.

[0173] Meanwhile, the method for determining frequency change in S850 is not limited to the methods described in FIG. 9, and the controller (1500) can determine whether there is a frequency change using various methods such as phase tracking, frequency estimation, filtering techniques, digital filtering and tracking filters, and determine the driving frequency of the next cycle.

[0174] [4] Method to control the driving frequency so that the driving frequency and the resonance frequency are in phase

[0175] Meanwhile, according to the above [3] method for controlling the driving frequency even before plasma ignition, the applicant was able to ignite the plasma using nitrogen gas, but in order to deliver maximum power to the load (2000) while minimizing power loss, it is most effective to control the driving frequency so that the driving frequency and the resonance frequency are in phase. Accordingly, in the additional solution, the driving frequency control method described above will be examined in more detail based on FIGS. 8 and FIGS. 9.

[0176] To explain additional solutions, refer again to FIGS. 8 and 9. S810 and S830 are the same as described in [3] Method for controlling the driving frequency before plasma ignition, so the description is omitted. In S850, the controller (1500) determines whether the resonant frequency of the load corresponds to the driving frequency of the AC voltage. Referring to FIG. 10, the controller (1500) may assume the aforementioned virtual signal to obtain voltage phase information of the AC voltage. At this time, as described above, the controller (1500) shifts the predetermined control signal by a predetermined time, and the predetermined time is as (1) in FIG. 10. (1) in FIG. 10 may be the time required from the time when the control signal is applied to the inverter (1300) until the time when the AC voltage is output from the inverter (1300) due to the applied control signal. This is because, in an ideal situation, the phase of the control signal differs from the phase of the output voltage by the aforementioned time, so the phase of the output voltage can be determined through the control signal using the method described above. In other words, it is equivalent to assuming a virtual signal, which is the output voltage of an ideal inverter, in Fig. 10.

[0177] The controller (1500) calculates a delay time between the virtual signal and the current phase information based on the current phase information obtained according to S853 and the virtual signal. For example, the controller (1500) calculates a delay time which is the difference between a first time interval from the rising edge of the virtual signal to the falling edge of the virtual signal that appears for the first time after the rising edge, and a second time interval from the rising edge of the virtual signal to the falling edge of the current phase information that appears for the first time after the rising edge. Additionally, the controller (1500) determines whether the calculated delay time falls within a predetermined range. Since the method of maintaining / changing the driving frequency based on whether the delay time falls within the predetermined range has been described above in [3] Method of controlling the driving frequency even before plasma ignition, a redundant description is omitted. Meanwhile, the controller (1500) controls the driving frequency of the AC voltage so that it is in phase with the resonant frequency of the load (2000), where "in phase" means that the absolute value of the delay time is less than or equal to the system clock period, which is the inverse of the system clock frequency of the controller (1500). In other words, in [4] Method of controlling the driving frequency so that the driving frequency and the resonant frequency are in phase, the controller (1500) can set the predetermined range based on the system clock period. For example, if the system clock period is Xns, the controller (1500) can set the predetermined range to "-X(ns) ≤ delay time ≤ X(ns)". For example, if the system clock frequency of the controller (1500) is 200MHz, the system clock period may be 5ns.

[0178] This means that the controller (1500) outputs the value of the control signal in units of the system clock period, and therefore, the period of the AC voltage is also controlled in units of the system clock period. That is, the driving frequency of the AC voltage is controlled in units of the system clock frequency. Therefore, since the minimum unit in which the controller (1500) can change the driving frequency is determined by the system clock frequency, if a delay time less than the minimum unit in which the controller (1500) can change the driving frequency is calculated, it can be considered that there is a sufficient degree of correspondence to determine that the driving frequency of the AC voltage and the resonance frequency of the load (2000) are in phase.

[0179]

[0180] Meanwhile, according to the [Plasma Ignition Method in Exhaust Gas Decomposition Device] described above, the applicant succeeded in igniting plasma using nitrogen gas in an exhaust gas decomposition device (100) in which the diameter of the discharge tube (2100) corresponds to the diameter of the pipe (700), but there was difficulty in sufficiently decomposing the exhaust gas. As described above, in order to extend the lifespan of the pump (500) and the pipe (700), it is important to sufficiently decompose the exhaust gas. At this time, the exhaust gas decomposition rate at which it can be judged that the exhaust gas has been sufficiently decomposed is called the expected decomposition rate. For example, the expected decomposition rate may be 90% or more.

[0181] In order for the exhaust gas decomposition device (100) to sufficiently extend the lifespan of the pump (500) and piping (700), the decomposition rate must be at least equal to the expected decomposition rate, and since a higher decomposition rate proves that the performance of the exhaust gas decomposition device (100) is superior, it is necessary to develop an exhaust gas decomposition device capable of maximizing the exhaust gas decomposition rate.

[0182] However, as described below, there are various difficulties in developing an exhaust gas decomposition device (100) that has decomposition performance exceeding the expected decomposition rate. Below, we will examine a method for decomposing exhaust gas exceeding the expected decomposition rate despite these difficulties, the design of an exhaust gas decomposition device (100), and an exhaust gas decomposition device (100) that performs such a method.

[0183]

[0184] [Exhaust gas decomposition device that maximizes exhaust gas decomposition rate]

[0185] [1] Emission gas decomposition rate below the expected decomposition rate

[0186] According to the [Plasma Ignition Method in Exhaust Gas Decomposition Device] described above, the applicant succeeded in igniting plasma using nitrogen gas in the exhaust gas decomposition device (100). However, even after the successful ignition of nitrogen plasma, ZAC (Zirconium Ammonium Carbonate) was decomposed while controlling the driving frequency of the AC voltage to be in phase with the resonance frequency of the load (2000), but the ZAC decomposition rate remained at the 60% level, which was far short of the expected decomposition rate of 90%. As mentioned earlier, the exhaust gas decomposition rate must be 90% or higher to significantly solve the problem of metal materials being deposited on the pump (500) and pipe (700), so the exhaust gas decomposition device (100) having a decomposition rate of 60% is not effective.

[0187] Accordingly, in order to find the factor that has the greatest influence on the exhaust gas decomposition rate, the applicant observed changes in the exhaust gas decomposition rate while changing the exhaust gas decomposition device (100) and various factors. Specifically, the ZAC decomposition rate was measured while changing the RF power, the pressure inside the discharge tube (2200), the flow rate of oxygen supplied to the discharge tube (2200), and the flow rate of ZAC supplied to the discharge tube (2200). In this case, even when changing various factors, there is no occurrence where the ZAC decomposition rate is 90% or higher.

[0188] In addition, when the flow rates of oxygen and ZAC are changed, the degree of change in the ZAC decomposition rate does not change significantly. Also, since changing the flow rates of oxygen and ZAC can affect the environment of the semiconductor process chamber (200) by changing the hydraulic pressure, it is not desirable to increase the ZAC decomposition rate by changing the flow rates of oxygen and ZAC.

[0189] Meanwhile, the applicant found that changing the RF power and pressure results in a greater change in the ZAC decomposition rate. Accordingly, it can be seen that controlling the RF power supplied to the load (2000) or the pressure inside the discharge tube (2200) is effective in increasing the ZAC decomposition rate.

[0190] However, the pressure inside the discharge tube (2200) may vary depending on the amount of exhaust gas and purge gas discharged from the semiconductor process chamber (200), and since the amount of exhaust gas and purge gas discharged from the semiconductor process chamber (200) cannot be arbitrarily controlled by the exhaust gas decomposition device (100), it is very difficult to accurately control the pressure inside the discharge tube (2200).

[0191] On the other hand, RF power can be controlled by the exhaust gas decomposition device (100). However, the exhaust gas is decomposed after the plasma is ignited. When the plasma is ignited, the impedance of the load (2000) increases rapidly, and the current flowing through the antenna structure (2100) included in the load (2000) decreases rapidly. As the current decreases, the RF power supplied to the load (2000) also decreases, making it difficult to increase the RF power after the plasma is ignited. In particular, in order to minimize power loss when igniting plasma using nitrogen gas in a discharge tube (2200) having a larger diameter than the RPS or toroidal structure exhaust gas decomposition device described above, the RFG (1000) controls the driving frequency of the AC voltage so that the driving frequency follows the resonant frequency of the load (200) even before plasma ignition. Therefore, it is very difficult to increase the power supplied to the load (2000) and increase the exhaust gas decomposition rate by controlling the driving frequency alone, in a situation where the current flowing to the load (2000) is reduced after plasma ignition.

[0192] [2] Design of an antenna structure (2100) to reduce the inductance of the antenna structure (2100)

[0193] As described above, in order to increase the power supplied to the load (2000), the magnitude of the current flowing through the antenna structure (2100) must be increased. To increase the magnitude of the current, the impedance of the load (2000) must be reduced. To this end, the inductance of the antenna structure (2100) can be reduced by reducing the number of turns and / or layers of the antenna structure (2100).

[0194] However, if the number of turns and / or layers of the antenna structure (2100) is reduced to decrease the inductance of the antenna structure (2100), the coupling coefficient between the plasma and the antenna structure (2100) is reduced. That is, if the number of times the antenna structure (2100) is wound around the discharge tube (2200) is reduced, the coupling coefficient between the plasma and the antenna structure (2100) is reduced, so that the electromagnetic field generated by the alternating current flowing through the antenna structure (2100) is not transmitted to the plasma, making it very difficult to ignite and maintain the plasma.

[0195] In other words, for plasma ignition and maintenance to be successful, the induced electromotive force must be high, and the induced electromotive force increases in proportion to the frequency and current applied to the antenna structure (2100) and the inductance of the antenna structure (2100). Accordingly, if the inductance of the antenna structure (2100) is lowered, the impedance of the load (2000) is reduced, and the current of the load (200) increases, thereby increasing the induced electromotive force. However, conversely, if the inductance of the antenna structure (2100) is lowered too much, the induced electromotive force may be weakened, causing the plasma to turn off or not ignite.

[0196] That is, if the inductance of the antenna structure (2100) is appropriately reduced, the energy (i.e., power and induced electromotive force) supplied to the load (2000) can be increased. However, if the inductance of the antenna structure (2100) is reduced too much, the energy (i.e., power) supplied to the antenna structure (2100) is not properly transferred to the plasma due to a decrease in the coupling coefficient between the plasma and the antenna structure (2100) and a decrease in the induced electromotive force transmitted to the plasma. Consequently, the plasma is extinguished, and the exhaust gas is not decomposed. Therefore, it is important to find an appropriate number of turns and / or layers and inductance of the antenna structure (2100) that can supply high power to the load (2000) without extinguishing the plasma.

[0197] As can be seen in the experimental example described below, the applicant found the number of turns and layers of the antenna structure (2100) and the corresponding inductance that allow high power to be delivered to the load (2100) while keeping the plasma from being turned off, by varying the number of turns and layers of the antenna structure (2100).

[0198] This will be examined with reference to FIGS. 11 and 12. FIG. 11 shows an example in which an antenna structure (2100) having an inductance below a predetermined upper limit value is placed in a discharge tube (2200) so that an induced electromotive force can be provided in the internal space of the discharge tube to cause the exhaust gas to decompose at an expected decomposition rate (e.g., 90%) or higher even after the plasma is ignited inside the discharge tube (2200). At this time, the antenna structure (2100) described through FIG. 11 has an inductance above a predetermined lower limit value to have a coupling coefficient with the plasma and an induced electromotive force that prevents the ignited plasma from being extinguished. That is, the antenna structure (2100) described through FIG. 11 may have an inductance below the predetermined upper limit value and above the predetermined lower limit value.

[0199] When comparing FIG. 11 and FIG. 4, in FIG. 4, turn antenna segments are arranged so that one layer antenna (2110) has two or more turns to increase the coupling coefficient between the antenna structure (2100) and the plasma. In addition, the antenna structure (2100) is formed by arranging turn antenna segments so that the layer antenna (2110) has five or more (i.e., five layers or more), preferably ten or more (i.e., ten layers or more).

[0200] Meanwhile, looking at FIG. 11, a turn antenna segment is arranged so that only one turn is formed in a single layer antenna (2110). That is, in the antenna structure (2100) of FIG. 11, a single layer antenna (2110) has one turn antenna segment. Also, in the antenna structure (2100) of FIG. 11, the turn antenna segment is arranged so that the layer antenna (2110) is seven or fewer (i.e., seven layers or fewer). Preferably, the turn antenna segment is arranged so that the antenna structure (2100) has five layer antennas (2110) (i.e., five layers). However, if the antenna structure (2100) has fewer than five layer antennas (2110), the inductance of the antenna structure (2100) becomes too small, and the plasma may not be stably maintained. Referring to FIG. 12, interlayer capacitors (2131 to 2134) may also be included between layer antennas (2110) included in an antenna structure (2100) having reduced inductance according to the present disclosure. The interlayer capacitors (2131 to 2134) are positioned between adjacent layer antennas (2110) as described in FIG. 4, thereby electrically connecting the adjacent layer antennas (2110). Meanwhile, the layer antenna (2110) has a first end and a second end, the first end of the N-th layer antenna (2110) is connected to the second end of the N-1st layer antenna (2110) or the first terminal of the RFG (1000), and the second end of the N-th layer antenna (2110) is connected to the first end of the N+1st layer antenna (2110) or the second terminal of the RFG (1000).Based on this, with reference to FIGS. 11 and 12, unlike FIGS. 4 and 5, in the antenna structure (2100) for increasing the exhaust gas decomposition rate, when viewed from the inlet (2210) (or outlet (222)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)), the first ends of the layer antennas (2110) are all positioned at the same location, and the second ends of the layer antennas (2110) are all positioned at the same location. The antenna structure (2100) is configured such that when viewed from the inlet (2210) (or outlet (222)) toward the outlet (2220) (or inlet (2210)). In other words, in an antenna structure (2100) for increasing the exhaust gas decomposition rate, when viewed from the inlet (2210) (or outlet (222)) of the discharge tube (2200) toward the outlet (2220) (or inlet (2210)), the antenna structure (2100) can be arranged such that all first ends overlap and all second ends overlap. In other words, the angle between the virtual central axis of the discharge tube (2200) and the virtual first lines connecting the first ends of each layer antenna (2100) can be 0 degrees, and the angle between the virtual central axis and the virtual second lines connecting the second ends of each layer antenna (2100) can also be 0 degrees. This is because each of the layer antennas (2110) of the antenna structure (2100) is composed of turn antenna segments having the same radius, so there is no possibility of a potential difference occurring between different turns, so arcing caused by a potential difference between turns does not occur, and the potential distribution between adjacent layer antennas (2110) is the same, so arranging the antenna structure (2100) as described above does not cause a potential difference between the layer antennas (2110), and thus arcing can be minimized.

[0201] In addition, when the layer antenna (2110) is configured with only one turn as in FIG. 11, the interlayer capacitors (2131 to 2134) can be placed between turns having the same radius as adjacent layer antennas, so the complexity of the connection is lower than when the interlayer capacitors (2131 to 2134) are placed between turns having different radii as in FIG. 4 to 5, so that the layer antennas (2110) can be densely arranged enough to supply a strong induced electromotive force inside the discharge tube (2200) without being placed by pulling them out in a direction away from the discharge tube (2200) or distributed along the outer surface of the discharge tube (2200).

[0202] [3] Switch damage caused by overcurrent before plasma ignition

[0203] The applicant appropriately reduced the inductance of the antenna structure (2100) as described above [Method 1 for increasing the exhaust gas decomposition rate while preventing switch damage] so that high power could be applied to the load (2000) even after the plasma is ignited. However, before the plasma is ignited, there is no increase in impedance due to the plasma, but if the inductance of the antenna structure (2100) is reduced, the total impedance of the load (2000) becomes very small, and the current flowing through the antenna structure (2100) and the inverter (1300) increases significantly.

[0204] When such a greatly increased current flows within the inverter (1300), the switches included in the inverter (1300) may be damaged. That is, if the inductance of the antenna structure (2100) is reduced, a current exceeding the maximum allowable current of the switches flows through the switches, causing the switches to be destroyed and consequently causing the problem of having to replace the inverter (1300).

[0205] To solve this, a method may be considered in which the number of turn antenna segments included in the layer antenna (2110) is increased to increase the inductance of the antenna structure (2100) before plasma ignition, or the number of layer antennas (2110) included in the antenna structure (2100) is increased, and after plasma ignition, the number of turn antenna segments and / or the number of layer antennas (2110) is decreased again. However, this is a method that is practically impossible as it involves changing the structure of the antenna structure (2100) while power is being supplied to decompose the exhaust gas.

[0206] Therefore, even though the antenna structure (2100) has low inductance, a method is needed to lower the current magnitude before plasma ignition so that the switch of the inverter (1300) is not damaged, while increasing the current magnitude after plasma ignition so that sufficient power is supplied to the load (2000).

[0207] [4] Method 1 for increasing exhaust gas decomposition rate while preventing switch damage - Method of controlling output voltage using multiple output voltage patterns

[0208] FIG. 13 briefly illustrates the full-bridge inverter structure of FIG. 7, and FIG. 14 shows the output voltage output from the inverter (1300) according to the operation control of the four switches (S1 to S4) included in the inverter (1300). In FIG. 13, Vin, an input DC voltage input from the rectifier (1200), is applied to node (1650), and Vref, a reference voltage, is applied to node (1670) as a ground node. Here, the reference voltage may be 0, but may have other values. Meanwhile, when a high-level voltage is applied to the switch, the switch is turned on, and when a low-level voltage is applied to the switch, the switch is turned off.

[0209] Referring to FIGS. 13 and 14, when switches S1 and S3 are turned on and switches S2 and S4 are turned off, the inverter (1300) outputs +Vin (i.e., positive voltage). On the other hand, when switches S2 and S4 are turned on and switches S1 and S3 are turned off, the inverter (1300) outputs -Vin (i.e., negative voltage). Also, when switches S1 and S4 are turned on and switches S2 and S3 are turned off, the inverter (1300) outputs 0[V]. Also, when switches S2 and S3 are turned on and switches S1 and S4 are turned off, the inverter (1300) outputs 0[V]. Hereinafter, for convenience of explanation, the operation in which switches S1 and S3 are turned on and switches S2 and S4 are turned off is referred to as the first switch operation, the operation in which switches S2 and S4 are turned on and switches S1 and S3 are turned off is referred to as the second switch operation, and the operation in which switches S1 and S4 are turned on and switches S2 and S3 are turned off, or switches S2 and S3 are turned on and switches S1 and S4 are turned off is referred to as the third switch operation.

[0210] Output voltage patterns can be created by using the output voltage of the inverter (1300) according to the operation of switches as described in FIGS. 13 and 14. This will be examined through FIG. 15.

[0211] Referring to FIG. 15 (a), the inverter (1300) can output an AC voltage by repeating the first switch operation and the second switch operation to alternately output +Vin and -Vin. However, if a third switch operation is added as in FIG. 15 (b), a section in which 0[V] is output can be created. Meanwhile, such switching operations occur in half-cycle units of the driving frequency of the output voltage, and thus, +Vin, -Vin, and 0[V] can be output in half-cycle units of the driving frequency.

[0212] Referring to FIG. 15(b), the period during which +Vin or -Vin is output by the first switch operation or the second switch operation during the half-cycle of the driving frequency of the output voltage is called the Powering Duration (P-period). Additionally, the period during which 0[V] is output by the third switch operation during an integer multiple of the half-cycle of the driving frequency of the output voltage, and during which neither +Vin nor -Vin is output, is called the Freewheeling Duration (F-period). That is, the Powering Duration is fixed at the half-cycle of the driving frequency, but the Freewheeling Duration may be the same as or longer than the half-cycle of the driving frequency. Furthermore, the meaning of 0[V] being output may mean that neither +Vin nor -Vin is output during the Freewheeling Duration. Additionally, the meaning of 0[V] being output may mean that no voltage is output during the Freewheeling Duration.

[0213] The output voltage pattern may be an appropriate combination of the power ring section and the freewheeling section. For example, the output voltage pattern may consist only of the power ring section or only of the freewheeling section. Alternatively, the output voltage pattern may consist of at least one power ring section and at least one freewheeling section.

[0214] Since the magnitude of the current flowing through the antenna structure (2100) and the inverter (1300) during the freewheeling period is reduced exponentially based on the time constant of the load (2000), the longer the total length of the freewheeling period included in the output voltage pattern, the more the current flowing through the antenna structure (2100) and the inverter (1300) can be reduced when the inverter (1300) applies an output voltage to the antenna structure (2100) according to the output voltage pattern. Accordingly, damage to the switches (S1 to S4) included in the inverter (1300) can be prevented.

[0215] Conversely, the shorter the total length of the freewheeling section included in the output voltage pattern, the greater the current flowing through the antenna structure (2100) and the inverter (1300) when the inverter (1300) applies output voltage according to the output voltage pattern. Accordingly, more power can be provided to the plasma inside the discharge tube (2200) to increase the exhaust gas decomposition rate.

[0216] As described above, multiple output voltage patterns having different total lengths of freewheeling sections may be stored in a memory included in the RFG (1000) or in an external memory. In this case, the multiple output voltage patterns may be combinations of a powering section and a freewheeling section for the same unit output time. Alternatively, the multiple output voltage patterns may be combinations of a powering section and a freewheeling section for different output times.

[0217] Additionally, if the length of one freewheeling section is too long, the current flowing through the antenna structure (2100) during the freewheeling section becomes too small, and the power supplied to the plasma is reduced, which may cause the plasma to turn off. Accordingly, the length of one freewheeling section may be less than or equal to a predetermined upper limit length that prevents the plasma from turning off.

[0218] FIGS. 16 and 17 illustrate a method for controlling the current flowing through the inverter (1300) and the antenna structure (2100) using the plurality of output voltage patterns. Referring to FIG. 16, an ignition gas is supplied to the discharge tube (2200) (S1610). At this time, the ignition gas may be nitrogen as described in Example 1, but is not limited thereto, and may be an inert gas such as helium or argon, or oxygen. When the ignition gas is supplied, the controller (1500) selects output voltage patterns among the plurality of output voltage patterns stored in the memory of the RFG (1000) or an external memory, wherein the total length of the freewheeling section is greater than or equal to a first length, and controls the switch operation of the inverter (1300) according to the output voltage pattern having a length greater than or equal to the first length (S1630). At this time, the first length may be the minimum length such that the current flowing through the switches of the inverter (1300) becomes less than the maximum allowable current, even though the load (2000) has an impedance less than the relatively low first impedance, as described in [2] Design of the antenna structure (2100) for reducing the inductance of the antenna structure (2100)], and the plasma is not yet ignited.

[0219] In other words, an output voltage pattern in which the total length of the freewheeling section is greater than or equal to the first length may be an output voltage pattern in which the current flowing through the switches of the inverter (1300) becomes less than or equal to the maximum allowable current even before plasma ignition, despite the antenna structure (2100) having an inductance less than or equal to a predetermined upper limit value.

[0220] The controller (1500) can determine whether the plasma has been ignited periodically or non-periodically (S1650). For example, the controller (1500) can detect the magnitude of the output current flowing from the inverter (1300) to the antenna structure (2100). Here, the magnitude of the output current may be the RMS value of the output current or the peak value of the output current. At this time, since the sensor module (1400) described in FIG. 6 is a current for detecting the current phase information of the output current, it may not be suitable for detecting the magnitude of the output current to determine whether the plasma has been ignited. Therefore, the RFG (1000) may further include a separate current sensor (not shown) capable of detecting the magnitude of the output current in addition to the sensor module (1400). Alternatively, the separate current sensor may be an external device.

[0221] When the plasma is ignited, the impedance of the load (2000) increases, and the magnitude of the output current may decrease. Therefore, the controller (1500) can determine that the plasma has been ignited if the magnitude of the sensed output current is less than a predetermined value. That is, if the magnitude of the sensed output current is greater than or equal to the predetermined value, it can determine that the plasma has not yet been ignited. Alternatively, the controller (1500) can estimate the output current by measuring the input DC current of the inverter (1300). Generally, the peak value of the output (AC) current is greater than the magnitude of the input DC current of the inverter (1300). It is as large as that. Therefore, considering this, it is possible to measure the input DC current and determine whether the output current is less than the previously determined value.

[0222] Alternatively, the controller (1500) can determine whether plasma ignition has occurred by measuring the amount of power supplied to the load (2000). For example, as the impedance of plasma ignition increases, the amount of power supplied to the load (2000) decreases. Accordingly, the controller (1500) can determine that plasma ignition has occurred if the output power of the inverter (1300) is less than a predetermined amount of power, or if the power reduction per unit time is greater than or equal to a predetermined reduction amount. On the other hand, the controller (1500) can determine that plasma ignition has not yet occurred if the output power is greater than or equal to the predetermined amount of power, or if the power reduction per unit time is less than or equal to the predetermined reduction amount.

[0223] Meanwhile, the controller (1500) can measure the output power of the inverter (1300) through a separate power measurement sensor. The separate power measurement sensor may be included in the RFG (1000) or may be an external device for measuring power. Meanwhile, the controller (1500) may estimate the output power by measuring the input DC voltage and input DC current. This is because, generally, the input power of the inverter (1300) has a value similar to the output power. Meanwhile, as described above, the controller (1500) may calculate the output power based on the output current measured through a separate current sensor and the output voltage output according to the current switch operation of the inverter (1300).

[0224] If the controller (1500) determines that the plasma has not yet been ignited, it returns to step S1630 and can control a plurality of switches (S1 to S4) according to an output voltage pattern in which the total length of the freewheeling section is greater than or equal to the first length. At this time, the controller (1500) may control the switch operation of the inverter (1300) according to an output voltage pattern identical to the previously selected output voltage pattern. Alternatively, the controller (1500) may select an output voltage pattern different from the previously selected output voltage pattern and control the switch operation of the inverter (1300) according to the newly selected output voltage pattern. For example, even if a specific output voltage pattern with a total length of the freewheeling section longer than the previously selected output voltage pattern is selected, if it is determined that the current flowing through the inverter (1300) is less than the maximum allowable current of the switch, the controller (1500) may select the specific output voltage pattern and control the switch operation of the inverter (1300) according to the specific output voltage pattern. At this time, it is obvious that the total length of the freewheeling section included in the specific output voltage pattern must also be greater than or equal to the first length.

[0225] When it is determined that the plasma has been ignited, the controller (1500) selects an output voltage pattern among the plurality of output voltage patterns in which the total length of the freewheeling section is less than the first length, and can control the switch operation of the inverter (1300) according to the selected output voltage pattern (S1670). Since the impedance of the load (2000) has increased due to the plasma being ignited, even if the total length of the freewheeling section is less than the first length, the current flowing through the switch of the inverter (1300) will be less than the maximum allowable current. In addition, the output voltage pattern selected by the controller (1500) may be an output voltage pattern that causes the exhaust gas decomposition rate to be greater than the expected decomposition rate. For example, the controller (1500) can select an output voltage pattern in which the total length of the freewheeling section is less than the second length, and by controlling the switch operation of the inverter (1300) according to the selected output voltage pattern, the exhaust gas can be decomposed at a rate greater than the expected decomposition rate. That is, the second length is the total length of the freewheeling section capable of decomposing the exhaust gas at a rate greater than the expected decomposition rate. At this time, the second length may be less than or equal to the first length. For example, the output voltage pattern selected by the controller (1500) after plasma ignition may be an output voltage pattern consisting only of a powering section without a freewheeling section, which maximizes the exhaust gas decomposition rate so that the exhaust gas decomposition rate becomes greater than the expected decomposition rate.

[0226] Meanwhile, according to FIG. 16, when the exhaust gas is decomposed, the timing at which the exhaust gas is supplied to the discharge tube (2200) may vary. For example, as can be seen in FIG. 16, the exhaust gas may be supplied to the discharge tube (2200) at least one time among before S1610, between S1610 and S1630, between S1630 and S1650, and between S1650 and S1670.

[0227] FIG. 17 is intended to illustrate another example using one of the multiple output voltage patterns. Since S1701, S1703, and S1705 in FIG. 17 are identical to S1610, S1630, and S1650 in FIG. 16, redundant descriptions are omitted, and the operation of the RFG (1000) after the plasma is ignited is described.

[0228] When the controller (1500) determines that the plasma has been ignited, it selects an output voltage pattern among the plurality of output voltage patterns in which the total length of the freewheeling section is less than the first length (S1707), and can control the switch operation of the inverter (1300) according to the selected output voltage pattern (S1709). The controller (1500) determines whether the selected output voltage pattern is an output voltage pattern that includes only a power ring section without a freewheeling section (S1711), and if the selected output voltage pattern is not an output voltage pattern that includes only a power ring section, the controller (1500) selects an output voltage pattern in which the total length of the freewheeling section is shorter than the previously selected output voltage pattern (S1713), and returns to step S1709 to control the switch operation of the inverter (1300) according to the selected output voltage pattern. For example, the output voltage pattern selected by the controller (1500) in step S1713 may be an output voltage pattern in which the length of the freewheeling section is shorter by a preset section than the previously selected output voltage pattern.

[0229] The controller (1500) repeats steps S1709 to S1713, and when an output voltage pattern including only a power ring section without a freewheeling section is selected, thereafter the switch operation of the inverter (1300) can be controlled according to the output voltage pattern including only the power ring section (S1715). Accordingly, maximum energy is delivered to the load (2000), so that the exhaust gas decomposition rate is maximized, and the exhaust gas can be decomposed beyond the expected decomposition rate.

[0230] Previously, in the embodiment mentioned in FIG. 16, it was stated that if it is determined that the plasma has been ignited, an output voltage pattern including only the power ring section can be selected to maximize the exhaust gas decomposition rate. However, in this case, the current reduced by the ignition of the plasma may increase rapidly as it is immediately changed to an output voltage pattern including only the power ring section, which may place a burden on the inverter (1300) switches due to current spikes, etc. Accordingly, the burden applied to the switches of the inverter (1300) can be prevented by gradually reducing the total length of the freewheeling section as shown in FIG. 17. In addition, considering that the speed at which the plasma changes is several milliseconds, if the process of S1709 to S1713 is repeated very quickly (e.g., several microseconds to tens of microseconds), the exhaust gas decomposition rate can be increased such that there is almost no difference in effect compared to changing to an output voltage pattern including only the power ring section. Meanwhile, when the plasma is ignited, the impedance of the load (2000) increases, so that even if the controller (1500) controls the switch operation of the inverter (1300) according to an output voltage pattern that includes only a power ring section without a freewheeling section, a current less than or equal to the maximum allowable current of the switch flows through the switch, thereby preventing damage to the switch.

[0231] Meanwhile, step S1711 can be substantially omitted. For example, the controller (1500) can continuously change the output voltage pattern so that the total length of the freewheeling section included in the output voltage pattern becomes progressively shorter, without substantially determining whether the selected output voltage pattern includes only the power ring section, and according to a predetermined algorithm, until the output voltage pattern is finally changed to include only the power ring section without the freewheeling section.

[0232] Meanwhile, the point in time when the exhaust gas is supplied may be at least one of the following: before S1701, between S1701 and S1703, between S1703 and S1705, and between S1705 and S1707, similar to what is described in FIG. 16.

[0233] Meanwhile, the control operation of the controller (1500) that makes the driving frequency of the output voltage follow the resonant frequency of the load (2000) as described in the above-described [Plasma Ignition Method in Exhaust Gas Decomposition Device] can be performed in the same way in [Method 1 for Increasing Exhaust Gas Decomposition Rate While Preventing Switch Damage]. That is, the control operation of the controller (1500) that makes the driving frequency of the output voltage follow the resonant frequency of the load (2000) even after plasma ignition can be performed, and additionally, it can be performed together with the operation of controlling by changing a plurality of output voltage patterns in [Method 1 for Increasing Exhaust Gas Decomposition Rate While Preventing Switch Damage]. Accordingly, the length of the half-cycle of the driving frequency used to define the powering section and the freewheeling section can be continuously varied as the driving frequency is continuously changed to follow the resonant frequency of the load (200) that is variable.

[0234] [5] Method 2 to increase exhaust gas decomposition rate while preventing switch damage - Method of operating in half-bridge mode before plasma ignition and switching to full-bridge mode after plasma ignition

[0235] Below, with reference to FIGS. 18 and 19, we will examine another switch operation control method for controlling the current flowing through the switch depending on whether plasma is ignited.

[0236] Referring to FIG. 18, FIG. 18 (a) shows an example in which a controller (1500) controls the switch operation of an inverter (1300) in a so-called half-bridge operation mode. Referring to FIG. 18 (a), when controlling in a half-bridge operation mode, the controller (1500) can control switches S1 and S2 to be selectively turned on or turned off, while controlling switch S3 to be turned on and switch S4 to be kept off among the four switches included in the inverter (1300). Alternatively, when controlling in a half-bridge operation mode, the controller (1500) can control switches S1 and S2 to be selectively turned on or turned off, while controlling switch S4 to be turned on and switch S3 to be kept off. Alternatively, when controlling in half-bridge operation mode, the controller (1500) can control switch S1 to be kept on and switch S2 to be kept off, while selectively controlling switch S3 and switch S4 to be turned on or turned off. Alternatively, when controlling in half-bridge operation mode, the controller (1500) can control switch S2 to be kept on and switch S1 to be kept off, while selectively controlling switch S3 and switch S4 to be alternately turned on or turned off.

[0237] Meanwhile, FIG. 18 (b) shows that the controller (1500) controls the switch operation of the inverter (1300) according to the full bridge operation mode. Since the full bridge operation mode is identical to the switch operation of the inverter (1300) described in the examples above, such as [2] Configuration of RFG (1000) and FIG. 14, a detailed description is omitted.

[0238] When the controller (1500) controls the switch operation of the inverter (1300) according to the half-bridge operation mode, as seen in FIG. 18 (a), the inverter (1300) outputs a value of 0[V] or +Vin, or outputs a value of 0[V] or -Vin. That is, in the half-bridge operation mode, the inverter (1300) does not output one of the voltages of +Vin and -Vin. Therefore, the voltage output by the inverter (1300) in the half-bridge operation mode is half the voltage output by the inverter in the full-bridge operation mode, and accordingly, the current flowing through the switches of the load (2000) and the inverter (1300) can also be reduced by half. Accordingly, when the controller (1500) controls the switch operation of the inverter (1300) according to the half-bridge operation mode, the antenna structure (2100) has an inductance below a predetermined upper limit value as described in [2] Design of the antenna structure (2100) for reducing the inductance of the antenna structure (2100), and since the plasma is not yet ignited, the load (2000) has an impedance below a relatively low first impedance, so the current flowing through the switches of the inverter (1300) can be made to be below the maximum allowable current.

[0239] FIG. 19 illustrates the operation of an exhaust gas decomposition device (100) using the above-described half-bridge operation mode. Referring to FIG. 19, an ignition gas is supplied to a discharge tube (2200) (S1910). At this time, the ignition gas may be nitrogen as described in [Method for Plasma Ignition in an Exhaust Gas Decomposition Device], but is not limited thereto; it may be an inert gas such as helium or argon, or oxygen. When the ignition gas is supplied, the controller (1500) controls the switch operation of the inverter (1300) according to the half-bridge operation mode (S1930). Afterward, the controller (1500) determines whether the plasma has been ignited periodically or non-periodically (S1950). Since S1950 is identical to S1650 described above, a detailed description is omitted.

[0240] When the controller (1500) determines that the plasma has not yet been ignited, it maintains the half-bridge operation mode, and when it determines that the plasma has been ignited, it controls the switch operation of the inverter (1300) according to the full-bridge operation mode (S1970). As can be seen in FIG. 18 (b), when the controller (1500) controls the switch operation of the inverter (1300) according to the full-bridge operation mode, +Vin and -Vin are alternately output, thereby maximizing the exhaust gas decomposition rate so that the exhaust gas decomposition rate becomes greater than the expected decomposition rate. Additionally, when the plasma is ignited, the impedance of the load (2000) increases due to the ignited plasma, so the current flowing through the switch of the inverter (1300) can be less than the maximum allowable current. Meanwhile, the point in time when the exhaust gas is supplied may be at least one of the following: before S1910, between S1910 and S1930, between S1930 and S1950, and between S1950 and S1970, similar to what is described in FIG. 16.

[0241] [6] Combination of methods 1 and 2 to increase exhaust gas decomposition rate while preventing switch damage

[0242] Meanwhile, methods 1 and 2 for increasing the exhaust gas decomposition rate while preventing damage to the switch do not necessarily have to be performed separately, but can be performed in combination. For example, even if the controller (1500) controls the switch operation of the inverter (1300) according to the half-bridge operation mode before plasma ignition, if a current exceeding the maximum allowable current of the switch still flows through the switch, the controller (1500) can select an output voltage pattern among a plurality of output voltage patterns that is set for the half-bridge operation mode and has a total length of the freewheeling section less than or equal to a predetermined third length, and control the switch operation of the inverter (1300) according to the selected output voltage pattern.

[0243] FIG. 20 is a diagram illustrating the freewheeling section in the half-bridge operation mode. Referring to FIG. 20, the powering section (P- section) in the half-bridge operation mode is a section in which +Vin or 0[V] is selectively output, or -Vin or 0[V] is selectively output during half a cycle of the driving frequency of the output voltage. However, in the half-bridge operation mode, depending on whether the switch is kept on or off, +Vin is output and -Vin is not output during the powering section, or -Vin is output and +Vin is not output. Meanwhile, the freewheeling section (F- section) in the half-bridge operation mode is a section in which 0[V] is output for N times (N is an integer greater than or equal to 2) of the driving frequency of the output voltage. Referring to FIG. 20, the power ring section in the half-bridge operation mode refers to a section that is half a cycle of the driving frequency regardless of the output voltage, and the freewheeling section refers to a section where the output voltage is 0[V] and is more than twice the half cycle of the driving frequency.

[0244] The controller (1500) selects an output voltage pattern in which the total length of the freewheeling section is less than or equal to the third length according to the half-bridge operation mode before plasma ignition, and controls the switching operation of the inverter (1300) according to the selected output voltage pattern, and after plasma ignition, controls the switching operation of the inverter (1300) according to the half-bridge operation mode, but selects an output voltage pattern in which the total length of the freewheeling section exceeds the third length, and controls the switching operation of the inverter (1300) according to the selected output voltage pattern. That is, it can be understood that all operations in Solution 2-1 are performed in the half-bridge operation mode. However, even in this case, in order to increase the exhaust gas decomposition rate above the expected decomposition rate, the controller (1500) changes to the full-bridge operation mode after a certain period of time after plasma ignition, and finally controls the switching operation of the inverter (1300) according to an output voltage pattern that includes only the powering section without the freewheeling section, so that maximum power can be supplied to the load (2000). Additionally, the controller (1500) may select an output voltage pattern that includes only the power ring section while gradually reducing the total length of the freewheeling section as described in FIG. 17.

[0245] Alternatively, the controller (1500) may select an output voltage pattern in which the total length of the freewheeling section is less than or equal to the third length according to the half-bridge operation mode before plasma ignition, and control the switching operation of the inverter (1300) according to the selected output voltage pattern, and control the switching operation of the inverter (1300) according to the full-bridge operation mode after plasma ignition. At this time, the output voltage pattern selected in the full-bridge operation mode may be an output voltage pattern in which the total length of the freewheeling section exceeds the fourth length. Here, the fourth length may be equal to, smaller than, or longer than the third length described above. This is because, in full-bridge operation, the voltage range increases by a factor of 2 compared to the voltage range in half-bridge operation, so even if the total length of the freewheeling section included in the output voltage pattern is longer than the third length, the current flowing through the switches of the antenna structure (2100) and the inverter (1000) can be substantially increased. However, even in this case, the controller (1500) may control the switch operation of the inverter (1300) according to an output voltage pattern that includes only a power ring section without a freewheeling section after a certain time has passed since the plasma was ignited, in order to increase the exhaust gas decomposition rate above the expected decomposition rate, so that maximum power can be supplied to the load (2000). In addition, the controller (1500) may select an output voltage pattern that includes only the power ring section while gradually reducing the total length of the freewheeling section as described in FIG. 17.

[0246] Alternatively, the controller (1500) may select an output voltage pattern in which the total length of the freewheeling section is less than the first length in full-bridge operation mode before plasma ignition, and control the switching operation of the inverter (1300) according to the selected output voltage pattern, and after plasma ignition, select an output voltage pattern in which the total length of the freewheeling section exceeds the third length in half-bridge operation mode, and control the switching operation of the inverter (1300) according to the selected output voltage pattern. However, even in this case, in order to increase the exhaust gas decomposition rate above the expected decomposition rate, the controller (1500) may switch to full-bridge operation mode after a certain period of time after plasma ignition, and finally control the switching operation of the inverter (1300) according to an output voltage pattern that includes only the power ring section without the freewheeling section, so that maximum power can be supplied to the load (2000). In addition, the controller (1500) may select an output voltage pattern that includes only the power ring section while gradually reducing the total length of the freewheeling section as described in FIG. 17.

[0247] Accordingly, the controller (1500) can control the switching operation of the inverter (1300) by appropriately combining a half-bridge operation mode, a full-bridge operation mode, and a plurality of output voltage patterns according to at least one of the power and impedance of the load (2000) that changes before and after plasma ignition, and the current flowing through the antenna structure (2100).

[0248]

[0249] [Another structure of RFG included in the exhaust gas decomposition device according to the present disclosure]

[0250] Meanwhile, in the example described above, an exhaust gas decomposition device (100) operating according to the RFG structure described in FIGS. 6 and 7 was described. Below, we intend to describe the structure of an RFG that can be used in the above-described Embodiments 1 and 2, although it has a structure different from the RFG described in FIGS. 6 and 7. That is, the RFG and the components included in the RFG mentioned in Embodiments 1 and 2 can be interpreted as being replaced by the RFG and the components included in the RFG described below. In the following description, components and functions that are not specifically mentioned should be understood as being the same as those described in FIGS. 6 and 7.

[0251] [1] RFG connecting two or more inverters in parallel

[0252] FIG. 21 shows an RFG in which two or more inverters are connected in parallel. Compared to FIGS. 6 and 7, inverter (1301) and inverter (1302) are connected in parallel instead of inverter (1300). However, each of inverter (1301) and inverter (1302) is a full-bridge inverter that has the same components as inverter (1300) and the same connection relationships between the components. Also, if the reference number in FIG. 21 is the same as the reference number in FIG. 6 and 7, it is the same component. For example, the controller (1500) is the controller (1500) in FIG. 21 as well.

[0253] Referring to FIG. 21, the inverter (1301) and the inverter (1302) can be connected in parallel between the rectifier (1200) and the load (2000). Additionally, a controller (1500) provides the same control signal to each of the inverter (1301) and the inverter (1302). That is, the controller (1500) controls each of the inverter (1301) and the inverter (1302) equally. Through this, the sum of the output currents output from each of the inverters (1301, 1302) flows into the antenna structure (2100). That is, even if each of the inverters (1301, 1302) provides an output current that is half the size of the output current provided by the inverter (1300) to the load (2000) as in FIG. 6 and 7, a current of the same size can be provided to the load (2000). For example, to supply 100A of current to the load (2000), the inverter (1300) must output 100A of current, but the inverter (1301) and the inverter (1302) each need to output 50A of current, so the amount of current flowing through the switch can be reduced and damage to the switch can be prevented.

[0254] In this way, two or more inverters can be connected in parallel to reduce the current flowing through each inverter. For example, three, four, five, or more inverters can be connected in parallel using a parallel connection method as shown in FIG. 21.

[0255] [2] Half-bridge inverter structure

[0256] FIG. 22 shows an RFG (1000) using a half-bridge inverter (1800) instead of the full-bridge inverter (1300) described in FIG. 6 and 7. Except for the fact that a half-bridge inverter (1800) is used, the RFG (1000) of FIG. 22 is identical to the RFG (1000) described in FIG. 6 and 7. The half-bridge inverter (1800) uses a capacitor (C) instead of switches S3 and S4 of the full-bridge inverter (1300). DCPlace ).

[0257] When using the half-bridge inverter (1800) of FIG. 22, the RFG (1000) in the [plasma ignition method in an exhaust gas decomposition device] to the [exhaust gas decomposition device for maximizing exhaust gas decomposition rate] described above can be understood as the RFG (1000) of FIG. 22. In other words, some examples of the [plasma ignition method in an exhaust gas decomposition device], [method 1 for increasing exhaust gas decomposition rate while preventing switch damage], and [combination of methods 1 and 2 for increasing exhaust gas decomposition rate while preventing switch damage] can be implemented with the half-bridge inverter (1800). However, other examples of [method 2 for increasing exhaust gas decomposition rate while preventing switch damage] and [combination of methods 1 and 2 for increasing exhaust gas decomposition rate while preventing switch damage] cannot be implemented with the half-bridge inverter (1800). For example, [Method 2 for increasing exhaust gas decomposition rate while preventing switch damage] cannot be implemented by the half-bridge inverter (1800) because the half-bridge inverter (1800) cannot perform a full-bridge operation mode, and [Combination of Method 1 and 2 for increasing exhaust gas decomposition rate while preventing switch damage] also cannot be implemented by the half-bridge inverter (1800) in an embodiment in which the operation mode is switched from the half-bridge operation mode to the full-bridge operation mode after plasma ignition. On the other hand, [Method 2 for increasing exhaust gas decomposition rate while preventing switch damage] and [Combination of Method 1 and 2 for increasing exhaust gas decomposition rate while preventing switch damage] can be implemented by the half-bridge inverter (1800) in an embodiment in which one of a plurality of output patterns is selected periodically / non-periodically to control the switch operation of the inverter (1300) without switching the operation mode.

[0258] FIG. 23 is a diagram for defining the power ring section and the freewheeling section in an example in which a controller (1500) controls the switch operation according to a plurality of output voltage patterns in an RFG (1000) including a half-bridge inverter (1800). Referring to FIG. 23, the power ring section (P-section) is a section in which +Vin / 2 or -Vin / 2 is selectively output during half a cycle of the driving frequency of the output voltage. The freewheeling section (F-section) is a section in which +Vin / 2 or -Vin / 2 is output for N times (N is an integer greater than or equal to 2) of half a cycle of the driving frequency of the output voltage, and whether the output voltage during the freewheeling section is +Vin / 2 or -Vin / 2 can be determined in advance. Referring to FIG. 23, in half-bridge operation mode, the power ring section refers to a section that is half a cycle of the driving frequency regardless of the output voltage, and the freewheeling section refers to a section that is at least twice the half cycle of the driving frequency regardless of the output voltage.

[0259] However, when using a half-bridge inverter (1800), the range of the output voltage is half that of a full-bridge inverter (1300), so the power applied to the load (2000) for exhaust gas decomposition after plasma ignition can be reduced to 1 / 4 compared to a full-bridge inverter (1300), so the exhaust gas decomposition rate can be lower than when using a full-bridge inverter (1300).

[0260] Meanwhile, the half-bridge inverter can be implemented not only as the half-bridge inverter (1800) of FIGS. 22 to 23, but can also be implemented as the half-bridge inverter (1900) described in FIGS. 24 to 25. FIGS. 24 to 25 show an RF generator (1000) including a half-bridge inverter (1900) of a different type from the half-bridge inverter (1800). The half-bridge inverter (1900) of FIGS. 24 to 25 includes switch S1 and switch S2 among the four switches (S1 to S4) included in the full-bridge inverter (1300), and may be implemented in a form where the switch S3 part is shorted and the switch S4 part is open as in FIG. 24 (a), or in a form where the switch S3 part is open and the switch S4 part is shorted as in FIG. 25 (a).

[0261] Meanwhile, referring to FIG. 24 (b), in the case of a half-bridge inverter (1900) as in FIG. 24 (a), when switch S1 is turned on and switch S2 is turned off, a voltage of +Vin is output, and when switch S2 is turned on and switch S1 is turned off, 0[V] is output. This is the same as the controller (1500) controlling switch S1 and switch S2 to be selectively turned on or turned off while controlling switch S3 to always be turned on and switch S4 to always be turned off in the half-bridge operation mode disclosed in [Method 2 for increasing exhaust gas decomposition rate while preventing switch damage].

[0262] On the other hand, referring to FIG. 25 (b), in the case of a half-bridge inverter (1900) as in FIG. 25 (a), when switch S2 is turned on and switch S1 is turned off, a voltage of -Vin is output, and when switch S1 is turned on and switch S2 is turned off, 0[V] is output. This is the same as the controller (1500) controlling switch S1 and switch S2 to be selectively turned on or turned off while controlling switch S4 to always be turned on and switch S3 to always be turned off in the half-bridge operation mode disclosed in [Method 2 for increasing exhaust gas decomposition rate while preventing switch damage].

[0263] The operation of the half-bridge inverter (1900) of FIG. 24 and the half-bridge inverter (1900) of FIG. 25 is the same as that of the half-bridge inverter (1800) described above, so a detailed description is omitted. However, when applying the contents of the half-bridge inverter (1800) to the half-bridge inverter (1900) of FIG. 24, the power ring section is interpreted as +Vin / 2 and 0[V] being selectively output during half a cycle of the driving frequency of the output voltage, and the freewheeling section is interpreted as 0[V] being output during N times (N is an integer greater than or equal to 2) of the driving frequency of the output voltage. Likewise, when applying the content of the half-bridge inverter (1800) to the half-bridge inverter (1900) of FIG. 25, the power ring section is interpreted as -Vin / 2 and 0[V] being selectively output during half a cycle of the driving frequency of the output voltage, and the freewheeling section is interpreted as 0[V] being output during N times (N is an integer greater than or equal to 2) of the driving frequency of the output voltage.

[0264]

[0265] Meanwhile, the exhaust gas disclosed in the above-described application may be unreacted metal precursors, but is not limited thereto. For example, the exhaust gas includes byproduct gas, which is a byproduct generated by the partial reaction of a metal precursor, and reaction product, which is a product formed by the reaction of a metal precursor inside a chamber that is not completely discharged and remains inside the piping. Additionally, the exhaust gas may further include purge gas and carrier gas. For example, in the present disclosure, the metal precursor may include zirconium compounds, tungsten compounds, titanium compounds, silicon compounds, alinium compounds, and hafnium compounds.

[0266]

[0267] [Experiment to increase exhaust gas decomposition rate]

[0268] The applicant performed various experiments to solve the problems mentioned in Examples 1 and 2 during the process of making the exhaust gas decomposition device (100) described above, and found that the problems mentioned above are solved according to the solutions described in Examples 1 and 2.

[0269] [Experimental Example #1] An experiment to ignite a nitrogen-based plasma by controlling an inverter so that the driving frequency of the output voltage tracks the resonant frequency of the load in phase.

[0270] 1. Experimental Objective

[0271] This is an experiment to see if plasma can be ignited by using in-phase frequency tracking from before plasma ignition, even if nitrogen is supplied as the ignition gas to the ICP plasma device.

[0272] 2. Experimental Design

[0273] The ICP plasma device is an ICP plasma device of Intocore, consisting of a cylindrical discharge tube and an antenna structure arranged to surround the cylindrical discharge tube. RFGs are connected to both ends of the antenna structure to supply AC voltage to the antenna structure through the RFGs. The discharge tube used in this experiment had an outer diameter of 200 mm, a thickness of 10 mm, and a height of 351 mm, and was manufactured of quartz. The antenna structure consisted of 20 layer antennas, and each layer antenna was a single-turn antenna composed of one turn antenna segment. (Total of 20-turn antenna structures) Adjacent layer antennas were connected by interlayer capacitors, which were fixed capacitors. The RFG was a full-bridge inverter RFG capable of frequency variation between 1.5 MHz and 2.0 MHz. The AC power supply providing AC voltage to the rectifier of the RFG was a three-phase power supply with 208VAC, 60Hz, and 60Arms, and accordingly, the magnitude of the output pulse of the RFG was approximately 300V. In addition, the system clock of the FPGA used in the RFG was 200MHz. For the exhaust gas decomposition experiment of this experiment, a dry pump with a performance of 100,000 L / min was used as the pump fluidically connected to the downstream end of the discharge tube.

[0274] 3. Experimental Method

[0275] The flow rate of nitrogen supplied to the discharge tube was classified into 10 slm, 15 slm, and 20 slm, respectively, and when each flow rate was supplied, the pressure inside the discharge tube was changed using a pump, and it was observed whether the plasma was ignited using nitrogen.

[0276] At this time, the starting driving frequency of the output voltage output by the RFG was the same as the source resonant frequency of the antenna structure, and as mentioned in Example 1 above, the driving frequency of the output voltage was controlled to track the resonant frequency of the entire ICP plasma device to check whether the plasma was ignited. At this time, the range of the delay time between the output voltage and the output current for determining whether the driving frequency sufficiently tracked the resonant frequency was 5ns. That is, the RFG was set to determine that the driving frequency sufficiently tracked the resonant frequency if the absolute value of the delay time was within 5ns.

[0277] 4. Experimental Results

[0278] Table 1 below shows whether the nitrogen plasma ignites despite pressure changes when the nitrogen flow rate is 10 slm, Table 2 shows whether the nitrogen plasma ignites despite pressure changes when the nitrogen flow rate is 15 slm, and Table 3 shows whether the nitrogen plasma ignites despite pressure changes when the nitrogen flow rate is 20 slm.

[0279] Nitrogen Flow Rate 10 slm Pressure (Torr) Ignition Success 0.3 Success 0.5 Success 1 Success 1.5 Success 2 Success

[0280] Nitrogen Flow Rate 15 slm Pressure (Torr) Ignition Success 0.5 Success 1 Success 1.5 Success 2 Success

[0281] Nitrogen Flow Rate 20 slm Pressure (Torr) Ignition Success 0.5 Success 1 Success 1.5 Success 2 Success

[0282] As can be seen in Tables 1 to 3 above, when the driving frequency of the output voltage was controlled to track the resonance frequency of the entire ICP plasma device regardless of nitrogen flow rate and pressure, it was found that the plasma ignited even when nitrogen gas was used as the ignition gas. 5. Interpretation of Experimental Results

[0283] Accordingly, it was found that even before plasma ignition, if the driving frequency of the output voltage is controlled to track the resonance frequency of the entire ICP plasma device, the electromotive force or energy delivered inside the discharge tube increases, and thus plasma can be ignited even when using a gas that requires high energy for plasma ignition, such as nitrogen, as the ignition gas.

[0284] [Experiment Example #2] Experiment on whether current can be maintained by controlling the total length of the freewheeling interval included during a unit time interval

[0285] 1. Experimental Objective

[0286] We intend to investigate whether it is possible to control the pattern of the output voltage as in [Method 1 for increasing exhaust gas decomposition rate while preventing switch damage] so that a current below the maximum allowable current of the switch flows through the RFG switch before plasma ignition, and then increase the current again after it is reduced due to plasma ignition.

[0287] 2. Experimental Design

[0288] Except for the experimental design and antenna structure of Experimental Example #1, everything was identical. The antenna structure in Experimental Example #2 consisted of 7 layer antennas, and each layer antenna was a single-turn antenna composed of 1 turn antenna segment (total 7-turn antenna structure). In addition, the maximum allowable current of the switch included in the RFG was 60 Arms.

[0289] 3. Experimental Method

[0290] 256 output voltage patterns were established with different total lengths of freewheeling intervals included in the output voltage patterns during the same unit period of 666ns. The output voltage pattern levels were labeled such that the shorter the total length of the freewheeling interval, the higher the output voltage pattern level. That is, the 256th output voltage level, which has the highest output voltage pattern level, is an output voltage pattern that includes only a power ring interval without a freewheeling interval, and the 1st output voltage pattern is the output voltage pattern with the longest total length of the freewheeling interval among the multiple output voltage patterns.

[0291] The ignition gas was nitrogen, and at the start of plasma ignition, the output voltage was applied to the ICP plasma device according to the output voltage pattern 1, and the level was gradually increased, and the change in output current was measured as the level increased.

[0292] 4. Experimental Results

[0293] As can be seen in Fig. 26, when voltage was applied to the ICP plasma device according to the output voltage pattern of level 1 before plasma ignition, an output current of less than 30 Arms was measured, which was lower than the maximum allowable current. Subsequently, the level of the output voltage pattern was gradually increased, and the output current rose to about 40 Arms, but it was still lower than the maximum allowable current.

[0294] Meanwhile, it can be seen that the current decreases even when the level of the output voltage pattern is increased from the time the internal state of the discharge tube becomes unstable due to plasma ignition, but when the level of the output voltage pattern is continuously increased after the plasma is ignited, it can be seen that the output current recovers to nearly 40 Arms.

[0295] 5. Interpretation of Experimental Results

[0296] Accordingly, it can be seen that even if the inductance of the antenna structure is reduced, the current flowing through the ICP plasma device can be controlled below the maximum allowable current of the switch by controlling the output voltage pattern.

[0297]

[0298] [Experiment Example #3] Experiment observing how much power increases when the inductance of an antenna structure is reduced

[0299] 1. Experimental Objective

[0300] This is an experiment to observe how much the applied power increases by measuring the power applied to the ICP plasma device while gradually decreasing the number of turns and layers of the antenna structure.

[0301] 2. Experimental Design

[0302] It is identical to Experiment Example #2. However, the antenna structure was replaced by changing the number of layers and turns according to the purpose of Experiment Example #3.

[0303] 3. Experimental Method

[0304] The power measured at 66 Pa, 133 Pa, and 200 Pa was examined while replacing each different antenna structure. During this process, the nitrogen ignition gas supplied to the discharge tube was fixed at 10 slm. The following five antenna structures were used in the experiment.

[0305] 1) A 20-turn antenna structure comprising 20 layer antennas, each layer antenna comprising 1 turn antenna segment.

[0306] 2) A 15-turn antenna structure comprising 15 layer antennas, wherein each layer antenna comprises 1 turn antenna segment.

[0307] 3) A 12-turn antenna structure comprising 12 layer antennas, wherein each layer antenna comprises 1 turn antenna segment.

[0308] 4) A 10-turn antenna structure comprising 10 layer antennas, wherein each layer antenna comprises 1 turn antenna segment.

[0309] 5) A 7-turn antenna structure comprising 7 layer antennas, wherein each layer antenna comprises 1 turn antenna segment.

[0310] 4. Experimental Results

[0311] Antenna Structure Turns Pressure (Pa) 20 Turns 15 Turns 12 Turns 10 Turns 7 Turns 6 RF Power (W) 30 15 31 67 35 17 37 59 61 49 Frequency (MHz) 1.5 73 61.6 42 51.7 24 81.7 99 91.6 15 Impedance (Ω) 20.9 71 9.8 71 8.1 11 6.4 89.7 81 33 RF Power (W) 36 11 390 84 27 14 51 26 68 Frequency (MHz) 1.5 6 75 1.6 33 1.7 0 861.7 79 91.6 16 Impedance (Ω) 17.1 71 5.9 14.8 11 3.7 89.2 28 RF Power (W)41614495480051397298 Frequency (MHz)1.56481.62881.70331.77331.645 Impedance (Ω)15.0213.7912.9812.248.47

[0312] Looking at Figure 27 and Table 4 above, it was observed that when the number of turns was reduced from 20 to 10, the increase in RF power was not significant, but when the number of turns was changed from 10 to 7, the RF power increased rapidly. In addition, it was observed that the RF power applied to the ICP plasma increased as the number of turns was reduced, and when comparing the 20-turn antenna structure and the 7-turn antenna structure, it was observed that the applied RF power increased by about 2 times to approximately 3000W.

[0313] 5. Interpretation of Experimental Results

[0314] When using the 20-turn antenna structure and the 10-turn antenna structure, respectively, the difference in power applied to the ICP plasma device was not very large. On the other hand, it was observed that the power applied to the ICP plasma device using the 7-turn antenna structure was about 1.6 times higher than the power applied to the ICP plasma device using the 10-turn antenna structure. This is because as the number of turns decreases, the coupling between the antenna structure and the plasma decreases; when changing from the 10-turn antenna structure to the 7-turn antenna structure, the coupling decreases rapidly, causing the actual resistance of the load to decrease rapidly, and consequently, it is presumed that the power applied to the ICP plasma device increased rapidly.

[0315] Accordingly, it can be seen that if the number of turns included in the antenna structure is reduced so that the inductance is less than or equal to a certain inductance, the power applied to the ICP plasma device increases rapidly, and sufficiently high power can be applied to the ICP plasma device.

[0316]

[0317] [Experimental Example #4] Final exhaust gas decomposition rate

[0318] 1. Experimental Objective

[0319] Finally, we intend to see whether exhaust gas can be decomposed at an expected decomposition rate of 90% or more when an exhaust gas decomposition device is implemented according to the embodiments disclosed in this disclosure.

[0320] 2. Experimental Design

[0321] It is identical to Experiment Example #2.

[0322] 3. Experimental Method

[0323] Nitrogen gas was used as the ignition gas and supplied at a flow rate of 10 slm. Regardless of whether plasma ignition occurred before or after, the driving frequency of the output voltage was controlled to track the resonant frequency of the entire ICP plasma device within a range of 5 ns. Additionally, the output voltage pattern level was set to the lowest level, 1, before plasma ignition, and after plasma ignition, the output voltage pattern level was gradually increased to 256 in increments of 4.2 µs; once the level 256 was reached, the output voltage pattern was maintained. Furthermore, ZAC gas was used as the exhaust gas, with a flow rate of 0.1 slm. The flow rate of oxygen introduced into the discharge tube for ZAC gas oxidation was 4 slm.

[0324] 4. Experimental Results and Interpretation of Experimental Results

[0325] Through the above experiment, it was confirmed that the power supplied to the ICP plasma device exceeded 8000W, resulting in a ZAC decomposition rate of 90% or higher. This exceeds the expected decomposition rate of 90%, indicating that exhaust gases can be significantly decomposed to prevent the deposition of metallic materials on pumps and piping. Therefore, it was confirmed that designing and controlling an exhaust gas decomposition device according to an embodiment of the present disclosure can significantly prevent the deposition of metallic materials on pumps and piping, thereby increasing the replacement cycle of pumps and piping. Furthermore, it is expected that maintenance costs for semiconductor processes can be drastically reduced through this.

[0326]

[0327] The features, structures, effects, etc. described in the embodiments above are included in at least one embodiment of the present invention and are not necessarily limited to only one embodiment. Furthermore, the features, structures, effects, etc. exemplified in each embodiment may be combined or modified and implemented in other embodiments by a person skilled in the art to which the embodiments belong. Accordingly, details regarding such combinations and modifications should be interpreted as being included within the scope of the present invention.

[0328] Furthermore, although the embodiments have been described above, this is merely illustrative and does not limit the invention. Those skilled in the art will understand that various modifications and applications not exemplified above are possible within the scope of the essential characteristics of the embodiments. In other words, each component specifically shown in the embodiments may be modified and implemented. Differences related to such modifications and applications should be interpreted as being included within the scope of the invention as defined in the appended claims.

Claims

1. An exhaust gas decomposition apparatus for decomposing a metal precursor discharged from a semiconductor process chamber using plasma, wherein A discharging tube having an inlet and an outlet, and providing an internal space between the inlet and the outlet where the plasma is induced—wherein, the metal precursor is introduced into the internal space through the inlet—; An antenna disposed to surround the discharge tube and supplied with RF power to induce the plasma in the internal space of the discharge tube—wherein, the antenna has an inductance less than or equal to a predetermined value such that, after the plasma is ignited, an electromotive force is provided in the internal space of the discharge tube to cause the metal precursor to decompose at a rate greater than the expected decomposition rate—; and It has a first output terminal and a second output terminal, and applies a first voltage or a second voltage to the antenna through the first output terminal and the second output terminal, and includes an RFG (Radio Frequency Generator; RFG) - wherein the first voltage is defined as the first potential of the first output terminal being higher than the second potential of the second output terminal, and the second voltage is defined as the second potential being higher than the first potential -; The above RFG is, An inverter having a plurality of switches and applying the first voltage or the second voltage to the antenna according to the operation of the plurality of switches; and A controller that controls the operation of the plurality of switches; including, The above controller is, Sensing the output current provided from the inverter to the antenna, Based on the above-mentioned sensed output current, one of a plurality of output voltage patterns is selected, wherein the first voltage and the second voltage are patterns output from the inverter, and The operation of the plurality of switches is controlled such that the first voltage or the second voltage is applied to the antenna according to the selected output voltage pattern, wherein each of the plurality of output voltage patterns includes at least one power ring section that outputs the first voltage or the second voltage and at least one freewheeling section that does not output both the first voltage and the second voltage. The above controller is, In response to the sensed output current indicating that the plasma is ignited prior to ignition, a first output voltage pattern is selected in which the total length of the freewheeling section within the output voltage pattern is greater than or equal to a first length, and the operation of the plurality of switches is controlled according to the selected first output voltage pattern—wherein, the selected first output voltage pattern causes the current flowing through each of the plurality of switches prior to ignition of the plasma to be less than or equal to the maximum allowable current of each of the plurality of switches, despite an inductance less than or equal to a predetermined value—, Characterized by selecting a second output voltage pattern in which the total length of the freewheeling section within the output voltage pattern is less than the first length from the point in time when the sensed output current indicates that the plasma has been ignited, and controlling the operation of the plurality of switches according to the selected second output voltage pattern. Exhaust gas decomposition device.

2. In Paragraph 1, If the above-mentioned sensed output current is greater than or equal to a predetermined value, it indicates that the plasma has not yet been ignited, and If the above-mentioned sensed output current is less than the above-determined value, it indicates that the plasma has been ignited. Exhaust gas decomposition device.

3. In Paragraph 1, The above controller is, The method is further characterized by controlling the operation of the plurality of switches according to the second output voltage pattern, while repeatedly replacing the second output voltage pattern such that the total length of the freewheeling section included in the second output voltage pattern gradually decreases from the point in time when the sensed output current indicates that the plasma has been ignited. Exhaust gas decomposition device.

4. In Paragraph 3, The above controller is, The method is further characterized by controlling the operation of the plurality of switches according to the second output voltage pattern being replaced, while repeatedly replacing the second output voltage pattern until it is replaced with the second output voltage pattern composed of the power ring section without the freewheeling section. Exhaust gas decomposition device.

5. In Paragraph 4, The above controller is, After the second output voltage pattern without the freewheeling section is selected, the operation of the plurality of switches is controlled according to the second output voltage pattern without the freewheeling section to cause the metal precursor to decompose at a rate greater than the expected decomposition rate. Exhaust gas decomposition device.

6. In Paragraph 5, Even if the second output voltage pattern without the freewheeling section is selected, the current flowing through each of the plurality of switches becomes less than or equal to the maximum allowable current due to the increase in impedance caused by the ignited plasma. Exhaust gas decomposition device.

7. In Paragraph 1, The above expected decomposition rate is 90% or higher, Exhaust gas decomposition device.

8. In Paragraph 1, The above exhaust gas decomposition device is, An additional gas supply pipe for supplying oxygen to the internal space of the discharge tube to oxidize the metal precursor; and A trap further comprising a trap fluidically connected to the above outlet to capture a metal precursor oxidized by oxygen. Exhaust gas decomposition device.

9. In Paragraph 1, The inlet of the above discharge tube is fluidically connected to the piping of the semiconductor process chamber through which the metal precursor is discharged, and The diameter of the above discharge tube corresponds to the diameter of the above pipe, Exhaust gas decomposition device.

10. In Paragraph 1, The above antenna is, It is composed of a plurality of layers corresponding to each of a plurality of planes perpendicular to the virtual central axis of the above discharge tube, and Each of the above plurality of layers includes a turn antenna segment having a certain radius from the virtual central axis, wherein The combined inductance of the turn antenna segments included in each of the plurality of layers is less than or equal to the predetermined value, Exhaust gas decomposition device.

11. In Paragraph 1, The above controller is, Controlling the operation of the plurality of switches by inputting a control signal that controls the operation of the plurality of switches to the inverter, wherein The above controller is, Assuming a virtual signal obtained by shifting the above control signal by a predetermined time, Calculate the delay time between the above virtual signal and the above output current, Based on the above delay time, a driving frequency corresponding to a resonant frequency that varies according to at least one of the antenna, the discharge tube, and the plasma is determined, and Further characterized by controlling the operation of the plurality of switches so that the first voltage or the second voltage is output based on the above driving frequency. Exhaust gas decomposition device.

Images