JP5198616B2 - Plasma processing equipment - Google Patents

Description

  The present invention belongs to semiconductor manufacturing technology. In particular, the present invention relates to a plasma processing apparatus suitable for plasma processing of a semiconductor wafer using plasma and a method for controlling the plasma processing apparatus.

  As semiconductor devices have been highly integrated in recent years, circuit patterns have been increasingly miniaturized, and the required processing dimension accuracy has become increasingly severe. Also, the wafer diameter has been increased to 300 mm for the purpose of reducing the manufacturing cost of semiconductor devices, but for the purpose of increasing the yield, plasma is made uniform in a wide range from the center of the wafer to the vicinity of the outer periphery, and high quality is achieved. Therefore, it is required that uniform processing can be performed. At the same time, it is required to reduce contamination of the wafer by a metal material or the like caused by reduction of foreign matters or scattering due to sputtering.

  In order to achieve such an object, conventionally, in a parallel plate type plasma generator composed of an upper metal plate electrode and a lower wafer (operating as an electrode), each of the upper electrode and the lower electrode (wafer) is provided. A technique is known in which a high frequency bias having the same frequency is applied, a high frequency phase between the biases is controlled, and plasma is made uniform (for example, refer to Patent Document 1).

  In addition, by controlling the phase of the high frequency between the upper and lower biases, there is a technique for reducing charging damage by improving the uniformity of plasma and reducing the amount of foreign matter generated, while either upper or lower electrode always operates as ground. It is known (see, for example, Patent Documents 2 and 3).

  Apart from this, the fact that the high-frequency transmission line affects the high-frequency voltage / current and phase difference is different in the high-frequency waveform at the output part of the high-frequency matching unit and the wafer, and therefore to obtain information on the potential of the wafer. In addition, it is known that a wafer potential probe method for directly measuring a wafer potential is effective (see, for example, Patent Document 4).

Japanese Patent Laid-Open No. 2001-127045 JP 2002-184766 A JP 2004-111432 A JP 2001-338917 A

  However, although the conventional technique has means for detecting and adjusting the high-frequency phase difference between the upper and lower biases and it is good to control the phase difference to 180 degrees, how to achieve the object by phase control? There is a problem that it is not clarified whether to guarantee. For example, the plasma uniformity is improved, but the means for ensuring the plasma uniformity by phase control is unknown. The same applies to the reduction of foreign matter generation and charging damage.

  As a second problem, there is a problem that even if the phase difference is controlled to 180 degrees by the conventional method, it cannot be guaranteed that the phase difference of the voltage actually appearing on the upper and lower electrodes is 180 degrees. In these methods, phase information is obtained at a matching unit of an application circuit that applies a high frequency bias to the upper and lower electrodes. However, the high frequency applied to the upper and lower electrodes is a bias high frequency circuit for generating plasma that passes through a high frequency filter or the like for the bias high frequency application circuit to the upper electrode after the output unit of the matching unit. The high-frequency transmission line to the upper electrode exists as a load of the upper electrode and is unique to the upper electrode. On the other hand, with regard to the bias high frequency application circuit to the lower electrode, naturally, the structure in which the lower electrode is mounted on the wafer and the wafer temperature is controlled exists in the transmission path of the high frequency bias, and the electrostatic chucking circuit of the wafer transmits the high frequency bias. There is a feature that a machine part such as a wafer transfer mechanism, a cover for protecting them, and the like existing as a load in the path are present as a ground of the high-frequency bias transmission path. That is, although phase information is obtained by each matching unit of the upper and lower electrodes, there is no guarantee that the phase difference of the voltage generated at the electrodes is controlled because the high-frequency transmission path to the subsequent electrodes is different.

Furthermore, this problem is also affected by the fact that the impedance of the high-frequency transmission path leading to the ground viewed from the upper electrode is not the same as the impedance of the high-frequency transmission path leading to the ground viewed from the lower electrode. The upper electrode sees the vacuum vessel wall and lower electrode and the electric circuit connected to it as part of the load as ground, while the lower electrode uses the vacuum vessel wall and upper electrode and the electric circuit connected to it as part of the load. to see. As described above, the structure of the upper electrode and the lower electrode and the electric circuit connected to them are not the same, so the high-frequency impedances of these loads are naturally different. Further, the range of the wall of the vacuum vessel seen by the upper electrode and the wall of the vacuum vessel seen by the lower electrode are not the same range because the positions where they are attached are facing each other. Further, plasma during processing is generated in the vicinity of the upper electrode and diffused to the periphery of the lower electrode and the characteristics thereof change. In addition, since the processing of the wafer proceeds in the lower electrode, reaction products generated in large quantities from the wafer greatly change the plasma state in the vicinity of the lower electrode.

  Accordingly, the impedance determined by the plasma density and electron temperature near the upper electrode is not the same as the impedance determined by the plasma density and electron temperature near the lower electrode. Of course, these impedances are different because the upper and lower electrodes see a ground that is not in the same range through these non-equal impedance plasmas.

  Further, as described in Patent Document 2, when a magnetic field is applied to a part or the whole of the vacuum vessel, there is no completely uniform magnetic field. The impedance of each high frequency path leading to ground is different. Also, when processing various conditions of plasma (density, electron temperature, high-frequency power that causes it, gas type / gas pressure, positional relationship between upper and lower electrodes, magnetic field conditions, etc.) when processing wafers, Depending on conditions, the impedance of each high-frequency path from the upper and lower electrodes to the ground differs.

  Since the phase shift due to the power path is due to stray capacitance and coil components existing in the power path, the phase shift increases as the frequency increases. Although depending on the configuration of the apparatus, the effect of phase shift is generally significant at a frequency of 1 MHz or higher.

  From the above, it can be said that the phase difference between the voltages generated at the upper and lower electrodes should be determined by the entire high-frequency circuit passing through each electrode and the high-frequency frequency. It is that it cannot be controlled.

  The third problem stems from the fact that the voltage waveform of the high-frequency bias generated at the upper and lower electrodes is not a sine wave, and the optimum value control cannot be performed by the conventional control method. It is well known that an electrically non-linear region called an ion sheath is formed on the front surface of an electrode exposed to plasma. This non-linearity means that the current and the voltage are not in a proportional relationship. When the voltage is positive, the electron current flows into the electrode, and the voltage-current characteristic of the electron current is determined by the electron diffusion coefficient, and usually the voltage. It becomes an exponential function. Conversely, when the voltage is negative, an ionic current flows into the electrode, but this current is almost constant without depending on the voltage due to the space charge limiting current. When a high frequency is applied to this electrode, current does not flow in proportion to the high frequency voltage, and the voltage waveform also includes distortion harmonics from sine waves.

  What is important in the present invention is that the characteristics of the plasma are different on the front surfaces of the upper and lower electrodes, and hence the characteristics of the ion sheath are different. As a result, the amount of high-frequency distortion (order and magnitude of the harmonics) at the upper and lower electrodes. ) Is different, that is, the waveform of each electrode voltage is different. Actually, as described in Patent Document 2 and Patent Document 3, the measured values of the maximum electrode voltage gate inter-oxide voltage, δVdc, and outer peripheral ion saturation current density, which are physical quantities serving as phase control indices, are characteristic. The optimum phase difference of 180 degrees is not the optimum value, and there is a shift. Most of the causes of this deviation are that the phase difference of each electrode cannot be controlled because the impedance of the above-described high-frequency propagation path is different, and that the current / voltage waveform of the electrode is distorted due to harmonics. The cause can be determined.

  The fourth problem is that the apparatus for improving the wafer processing by such phase control is limited to a parallel plate type plasma source having upper and lower electrodes. In industry, various plasma sources such as an inductively coupled plasma source and an ECR plasma source are used. However, since these plasma sources do not have electrodes that are capacitively coupled with two independent plasmas, this method cannot be applied.

  In order to solve the above problems, the first object of the present invention is to solve the problems 1 to 3 above, and to provide a plasma processing apparatus and a plasma processing apparatus having a function of using a signal optimal for the purpose of phase control. It is to provide a control method.

  In order to cope with the fourth problem, the second object of the present invention has the same function for a plasma source having no electrode capacitively coupled to two independent plasmas such as an inductively coupled plasma source. A plasma processing apparatus and a method for controlling the plasma processing apparatus are provided.

The present invention includes a vacuum vessel in which plasma is generated in a vessel, a first high-frequency power source, a second high-frequency power source, a third high-frequency power source provided outside the vacuum vessel, and the first high-frequency power source. A coiled antenna provided outside the vacuum vessel to which the first high-frequency voltage is supplied, and a sample placed on the upper surface and provided in the vacuum vessel to which the second high-frequency voltage from the second high-frequency power source is supplied An Faraday shield provided between the coiled antenna and the vacuum vessel and supplied with a third high-frequency voltage from a third high-frequency power source; a second high-frequency voltage of the second electrode; A plasma processing apparatus comprising: a phase control device that controls a phase difference of a third high-frequency voltage from the third high-frequency power source;
Voltage detecting means for detecting a high-frequency potential propagating in the plasma on the inner wall surface of the vacuum vessel; and plasma emission detecting means for detecting a light emission state of the plasma in the vacuum vessel, the voltage detecting means The phase difference between the second high-frequency voltage and the third high-frequency voltage having the same frequency is controlled using a signal source that is a combination of the high-frequency voltage detected by the plasma light emission and the plasma light detected by the plasma emission detection means. It was made to do.

  According to the present invention, in the phase control of each voltage waveform of the third high-frequency power and the second high-frequency power that applies a bias to the processing substrate, the phase difference in each electrode that radiates each power to the plasma is accurately controlled. And the energy of ions impinging on the wall and the amount of material generated from the wall as a result of the interaction between the ion and the wall can be controlled from a minimum to a maximum. As a result, in an apparatus for processing a processing substrate with plasma, it is possible to provide an optimum vacuum container environment for each processing. Furthermore, this function can be realized not only in an apparatus using a parallel plate type plasma source but also in all apparatuses that generate plasma.

1 is a schematic sectional view showing the overall structure of a plasma processing apparatus according to a first embodiment of the present invention. The graph which shows the relationship between a high frequency electric potential, a plasma space electric potential, and a sheath voltage. The schematic sectional drawing which shows the whole structure of the plasma processing apparatus of the 2nd Example of this invention. The graph which shows the relationship between ion energy and phase difference. The schematic sectional drawing which shows the whole structure of the plasma processing apparatus of the 3rd Example of this invention. The graph which shows the relationship between the light-emission amount of aluminum, and a phase difference, and the graph which shows the relationship between the light-emission amount of carbon monoxide, and a phase difference. The graph which shows the relationship between the energy of ion, and the light-emission amount of aluminum and carbon monoxide. The schematic sectional drawing which shows the whole structure of the plasma processing apparatus of the 4th Example of this invention. The schematic sectional drawing which shows the whole structure of the plasma processing apparatus of the 5th Example of this invention. The schematic sectional drawing which shows the whole structure of the plasma processing apparatus of the 6th Example of this invention.

The first object of the present invention can be achieved by incorporating a signal optimal for ensuring the purpose of phase control into a phase control circuit. As a specific first example, for the purpose of controlling the phase difference between the voltages of the upper and lower electrodes, a probe that directly measures the phase difference is connected to the upper and lower electrodes, and the obtained signal is processed by calculation. This can be achieved by using a control signal for phase control.
As a second specific example, the purpose of suppressing the spatter and damage of the wall of the vacuum vessel and reducing the foreign matter resulting from them is that from the upper and lower electrodes by a voltage probe installed around the target vacuum vessel. This can be achieved by extracting a high-frequency signal propagating through the plasma and controlling the phase so that the voltage is minimized or set to an appropriate value.
Furthermore, as a third specific example, when it is desired to remove deposits adhering to the vacuum vessel wall or the like, or when it is not desired to remove deposits adhering to the wall or the wall, it is emitted from the wall into the plasma. By measuring the amount of light of a plurality of specific wavelengths generated by a plurality of substances to be emitted, the phase of the light of the substance to be removed is emitted, or the light of the substance not desired to be removed is not emitted. This can be achieved with proper control.
The second object of the present invention is to provide an electrode capacitively coupled with plasma independent of the wafer in a plasma source having only one electrode capacitively coupled with independent plasma such as an inductively coupled plasma source. Can be achieved.
That is, the present invention includes a vacuum vessel in which plasma is generated in a vessel, a first high-frequency power source, a second high-frequency power source, a third high-frequency power source, and a first high-frequency power source provided outside the vacuum vessel. The first electrode provided in or outside the vacuum container to which the first high-frequency voltage is supplied, and the vacuum on which the sample is placed on the upper surface and the second high-frequency voltage is supplied from the second high-frequency power source A plasma processing apparatus comprising: a second electrode provided in a container; and a phase control device that controls a phase difference between a second high-frequency voltage and a third high-frequency voltage from a third high-frequency power source, In addition to the two electrodes, an RF radiation unit capacitively coupled to the plasma generated in the vacuum vessel is configured, and a signal source other than a phase signal is used at the same frequency between the second electrode and the RF radiation unit. Second high frequency voltage and third high frequency voltage Of so as to control the phase difference. Furthermore, in the plasma processing apparatus, the RF radiating unit is configured by either a parallel plate electrode antenna, a focus ring, a Faraday shield, or another electrode, and a signal source other than the phase signal is a high-frequency signal. Either voltage, plasma light, or a combination of high-frequency voltage and plasma light was used. Further, in the plasma processing apparatus, the RF radiating unit is configured by either a parallel plate electrode antenna and a focus ring and another electrode combination, or a combination of a Faraday shield and a focus ring and another electrode. did.
The present invention includes a vacuum vessel, a first high frequency power source, a second high frequency power source, a third high frequency power source, a first high frequency voltage from the first high frequency power source, and a third high frequency power source from the third high frequency power source. A first electrode provided in a vacuum vessel to which a high frequency voltage of 3 is supplied, and a vacuum vessel in which a sample is placed on the upper surface and a second high frequency voltage from a second high frequency power supply is supplied. A plasma processing apparatus comprising a provided second electrode and a phase control device that controls a phase difference between the second high-frequency voltage and the third high-frequency voltage, wherein the second high-frequency voltage and the third high-frequency voltage are: First phase detection means for detecting the phase of the third high-frequency voltage of the first electrode, and second phase detection means for detecting the phase of the second high-frequency voltage of the second electrode having the same frequency And based on outputs of the first phase detection means and the second phase detection means And to control the phase difference between the second RF voltage and the third RF voltage.
The present invention includes a vacuum vessel, a first high frequency power source, a second high frequency power source, a third high frequency power source, a first high frequency voltage from the first high frequency power source, and a third high frequency power source from the third high frequency power source. A first electrode provided in a vacuum vessel to which a high frequency voltage of 3 is supplied, and a vacuum vessel in which a sample is placed on the upper surface and a second high frequency voltage from a second high frequency power supply is supplied. A plasma processing apparatus comprising a provided second electrode and a phase control device that controls a phase difference between the second high-frequency voltage and the third high-frequency voltage, wherein the second high-frequency voltage and the third high-frequency voltage are: A voltage detecting means for detecting a high-frequency potential having the same frequency and propagating in the plasma on the inner wall surface of the vacuum vessel, and a second high-frequency voltage and a third high-frequency voltage from the detected voltage of the voltage detecting means; A phase difference calculating means for calculating a phase difference, So that the output of the phase difference calculation means is minimum or a specific value, and to control the phase difference between the second RF voltage and the third RF voltage.
The present invention includes a vacuum vessel, a first high frequency power source, a second high frequency power source, a third high frequency power source, a first high frequency voltage from the first high frequency power source, and a third high frequency power source from the third high frequency power source. A first electrode provided in a vacuum vessel to which a high frequency voltage of 3 is supplied, and a vacuum vessel in which a sample is placed on the upper surface and a second high frequency voltage from a second high frequency power supply is supplied. A plasma processing apparatus comprising a provided second electrode and a phase control device that controls a phase difference between the second high-frequency voltage and the third high-frequency voltage, wherein the second high-frequency voltage and the third high-frequency voltage are: Plasma emission detection means for detecting the emission state of the plasma in the vacuum vessel having the same frequency is provided, and the phase difference between the second high-frequency voltage and the third high-frequency voltage is controlled based on the plasma emission state. did.
The present invention is provided in a vacuum container, a first high-frequency power supply, a second high-frequency power supply, a third high-frequency power supply, and a vacuum container supplied with a first high-frequency voltage from the first high-frequency power supply. The first electrode, the second electrode provided in the vacuum vessel on which the sample is placed on the upper surface and supplied with the second high-frequency voltage from the second high-frequency power source, and the second electrode are placed. A focus ring that is provided around a sample to be supplied with a third high-frequency voltage from a third high-frequency power source, and a phase control device that controls a phase difference between the second high-frequency voltage and the third high-frequency voltage. A plasma processing apparatus, wherein the second high-frequency voltage and the third high-frequency voltage have the same frequency, and comprise plasma light emission detecting means for detecting the light emission state of the plasma in the vacuum vessel, based on the state of the plasma light emission Second high frequency voltage and third high frequency voltage Of so as to control the phase difference.
The present invention is provided outside a vacuum container to which a first high-frequency voltage is supplied from a vacuum container, a first high-frequency power supply, a second high-frequency power supply, a third high-frequency power supply, and a first high-frequency power supply. A first electrode, a second electrode provided in a vacuum vessel on which a sample is placed and a second high-frequency voltage is supplied from a second high-frequency power source, and the first electrode and the vacuum vessel And a Faraday shield that is supplied with a third high-frequency voltage from a third high-frequency power source, and a phase control device that controls the phase difference between the second high-frequency voltage and the third high-frequency voltage. In the processing apparatus, the second high-frequency voltage and the third high-frequency voltage have the same frequency, and are provided with plasma light emission detecting means for detecting the light emission state of the plasma in the vacuum vessel. 2 high frequency voltage and 3rd high frequency voltage And to control the phase difference.
The present invention is provided outside a vacuum container to which a first high-frequency voltage is supplied from a vacuum container, a first high-frequency power supply, a second high-frequency power supply, a third high-frequency power supply, and a first high-frequency power supply. A first electrode, a second electrode provided in a vacuum vessel on which a sample is placed and a second high-frequency voltage is supplied from a second high-frequency power source, and the first electrode and the vacuum vessel A third electrode provided between the Faraday shield and the third high-frequency voltage supplied from the third high-frequency power source and capacitively coupled to the plasma provided in the vacuum vessel; the second high-frequency voltage; And a phase control device for controlling the phase difference of the three high-frequency voltages, wherein the second high-frequency voltage and the third high-frequency voltage have the same frequency, and the light emission state of the plasma in the vacuum vessel is determined. Plasma emission detection means for detecting And to control the phase difference between the second RF voltage and the third RF voltage on the basis of the state.
The present invention relates to a vacuum vessel in which plasma is generated in a vessel, a first high frequency power source, a second high frequency power source, a third high frequency power source, and a first high frequency power source provided from the first high frequency power source. A first electrode provided inside or outside the vacuum container to which one high-frequency voltage is supplied, and a vacuum container to which a second high-frequency voltage is supplied from a second high-frequency power source on which a sample is placed. A control method for a plasma processing apparatus, comprising: a second electrode provided; and a phase control device that controls a phase difference between a second high-frequency voltage and a third high-frequency voltage from a third high-frequency power source. Supplying a predetermined power from one high-frequency power source to ignite plasma; and supplying a predetermined power from the second high-frequency power source and the third high-frequency power source after confirming the plasma ignition, respectively. , The second high frequency power supply and At power application start of the high frequency power source 3, with a preset mode without phase control is fixed to a predetermined phase angle, the alignment operation is to start the phase control after stabilization.
The present invention relates to a vacuum vessel in which plasma is generated in a vessel, a first high frequency power source, a second high frequency power source, a third high frequency power source, and a first high frequency power source provided from the first high frequency power source. A first electrode provided in or outside the vacuum container to which one high-frequency voltage is supplied, and a vacuum container in which a sample is placed on the upper surface and a second high-frequency voltage from the second high-frequency voltage is supplied And a phase control device for controlling a phase difference between the second high-frequency voltage and a third high-frequency voltage from a third high-frequency power source. A step of controlling the phase difference between the high-frequency voltage and the third high-frequency voltage of the second electrode by using a signal source other than the phase signal at the same frequency between the RF radiation portion capacitively coupled to the plasma and the second electrode. And removing the deposited film in the vacuum vessel Time, so as to control the phase difference according to the amount of the deposited film.
The overall structure of the plasma processing apparatus according to the present invention will be described below with reference to the schematic cross-sectional view shown in FIG. The plasma processing apparatus has a vacuum vessel 1 made of a conductor such as aluminum, and an antenna 2 and a shower plate 3 are attached to the inside thereof by respective shower plate support flanges 5 and 6. The vacuum vessel 1 is grounded. Between the antenna 2 and the shower plate 3, a gas is supplied from the atmosphere side via the process gas introduction pipe 7, and is jetted into the vacuum vessel 1 through a number of micro holes formed in the shower plate 3. It has become. Here, the antenna 2 is naturally formed of a conductor, but the shower plate 3 may be formed of a conductor / dielectric or semiconductor. When the shower plate 3 is formed of a conductor or a semiconductor, the shower plate 3 operates as an antenna that directly transmits power to plasma. The process gas is exhausted by the vacuum exhaust device 11 through the exhaust duct 10 and is maintained at a pressure suitable for the process.

  By supplying a predetermined power of a predetermined frequency from the first high-frequency power source 4 to the antenna 2 through the matching unit 28, a plasma 9 having characteristics suitable for the process is generated. The magnet (electromagnet) 8 may be used to control the plasma characteristics, but may not be used.

  The structure described here is a so-called capacitively coupled discharge. However, in the present invention, as long as plasma having characteristics suitable for the process can be generated, the discharge system is ECR discharge, inductively coupled discharge as described later. Other methods may be used. Of course, the structure of the discharge section must be suitable for the discharge system, and it is not necessary to stick to the structure described here.

  A wafer (processing substrate) 12 is carried into the vacuum vessel 1 from the atmosphere side through a carry-in port (not shown), and is placed on the processing electrode 14 by an appropriate delivery device (not shown). An upper surface of the processing electrode 14 has an electrostatic adsorption film (not shown) having appropriate characteristics, and is electrostatically adsorbed by a voltage applied from the DC power source 23 through the choke coil 22 from the outside. Further, the processing electrode 14 has a coolant groove 15 inside, and is supplied with a coolant from the outside (not shown), and the wafer 12 is adjusted to a temperature suitable for the process. A focus ring 13 is provided on the outer periphery of the wafer 12, and a susceptor 16 is provided for protecting the processing electrode 14 from the plasma 9. The processing electrode 14 is attached to the vacuum vessel 1 via an insulating base 17 and an electrode base flange 18. The processing electrode 14 can be moved up and down by an appropriate vertical mechanism (not shown), and the interval between the wafer 12 and the shower plate 3 may be made variable. The present invention can be applied regardless of the presence or absence of the vertical mechanism.

  High frequency power is applied to the processing electrode 14 from the second high frequency power source 21 through the conductor protected by the insulating tube 19 and the matching unit 20, and a high frequency bias is applied to the wafer 12. As described above, the first high-frequency power source 4 and the matching unit 28 do not need to follow the configuration shown in this figure. For example, the output is applied to the processing electrode 14 in parallel with the second high-frequency power source 21 and the matching unit 20. May be. Further, the first high-frequency power source 4 is not provided, and the second high-frequency power source 21 may be used for both generating the plasma 9 and applying a high-frequency bias to the wafer 12.

  A signal output from the oscillator 30 passes through the phase controller 26, is output from one of the outputs, and is input to the second high-frequency power source 21. The other output of the phase controller 26 is a signal having a controlled phase difference with respect to the signal going to the second high-frequency power source 21, and is input to the third high-frequency power source 25. The output of the third high frequency power supply 25 is mixed with the output of the first high frequency power supply 4 by the mixer 27 through the matching unit 24 and applied to the upper electrode 2 (or the shower plate 3).

  The mixer 27 not only mixes the outputs of the first high-frequency power supply 4 and the third high-frequency power supply 25, but also the output of the first high-frequency power supply 4 enters the third high-frequency power supply 25 and its matching unit 24. It has a filter function to prevent this. Furthermore, the mixer 27 has a filter function that prevents the output of the third high-frequency power supply 25 from entering the first high-frequency power supply 4 and its matching device 28. As described above, a high-frequency bias whose phase is controlled can be applied to the upper electrode 2 (or shower plate 3) and the lower processing electrode 14 (or wafer 12) facing each other at the same frequency.

  Since the first high frequency power source 4 generates plasma, a frequency of about 13.56 MHz or more about 2.4 GHz can be used. On the other hand, the main purpose of the frequencies of the second high-frequency power source 21 and the third high-frequency power source 25 is to apply a bias rather than to generate plasma. Usually, a frequency of 13.56 MHz or lower is used. Available.

  The phase controller 26 is connected to the signal of the wafer potential probe 31 that directly measures the potential of the wafer 12 that is the lower electrode and the signal of the high voltage probe 32 that directly measures the potential of the antenna 2 that is the upper electrode. Since this signal includes signals of the first high-frequency voltage and the third high-frequency voltage, the signal of only the third high-frequency voltage is extracted by the filter. The phase controller 26 performs phase control of signals output to the second high frequency power supply 21 and the third high frequency power supply 25 by detecting a phase difference between these two probes. At this time, the phase shift caused by the signal transmission path between the wafer potential probe 31 and the high voltage probe 32 must be accurately calibrated. For this reason, the phase controller 26 has a function of canceling this phase shift.

  By using the above functions and configuration, the phase between the upper and lower electrodes can be controlled with high precision, and either of the upper and lower electrodes always operates as ground, improving the uniformity of plasma and reducing the amount of foreign matter generated. It is possible to control the functions such as reducing damage to a higher level than in the prior art.

  Next, a wafer processing sequence using the plasma processing apparatus of FIG. 1 will be described. First, the vacuum container 1 is evacuated by the vacuum evacuation device 11 so that the degree of vacuum is below a predetermined level. At this time, it is confirmed that the gas leakage amount from the process gas introduction pipe 7 and the atmosphere is sufficiently low in the vacuum container 1. At the same time, it is confirmed that the amount of outgas from the components in the vacuum vessel 1 is sufficiently low. In this state, the processing substrate (wafer) 12 is mounted on the processing electrode 14. Next, a process gas is introduced into the vacuum vessel 1 from the process gas introduction pipe 7, and the vacuum exhaust capacity is adjusted to control the pressure optimal for the process.

  After the above process start conditions are satisfied, first, plasma 9 is ignited by supplying predetermined power from the first high-frequency power source 4. At this time, the matching operation start point is optimized using a preset function in advance so that the operation of the matching unit 28 is stable and the time during which the reflected wave from the antenna 2 is sufficiently small is minimized. After the operation of the matching unit 4 is stabilized and the ignition of the plasma 9 is confirmed, predetermined power is supplied from the second high-frequency power source 21 and the third high-frequency power source 25, respectively. As described above, after the ignition of the plasma 9 is confirmed, the power is supplied from the second high-frequency power source 21 and the third high-frequency power source 25 because the load impedance is high in the absence of plasma, so that the antenna 2 and the processing substrate 12 are supplied. This is because there is a high possibility that a high voltage will be generated and an abnormal discharge or device malfunction due to noise will occur. It is desirable that the power application start timings of the second high frequency power supply and the third high frequency power supply be the same. This shortens the processing time of the processing substrate by minimizing the rising operation time of the first, second, and third high-frequency power supplies, and the processing substrate plasmas within the rising time of the high-frequency power supply (that is, before the start of the process). To minimize the deposition of reaction products from the plasma on the processing substrate, and to achieve the effect of phase-controlled bias application from the start of processing as much as possible It is an object.

  When the second high-frequency power source 21 and the third high-frequency power source 25 are turned on simultaneously, the matching operations of the matching units 20 and 24 affect the matching operations, and the matching operation time may be prolonged. This corresponds to the case where the plasma characteristics, that is, the load impedance of the high-frequency power source is greatly changed by the matching operation. In this case, it is necessary to shift the power supply start timings of the second high frequency power supply and the third high frequency power supply. In this case, priority is given to the start of power supply of the second high-frequency power source 21 that applies a bias to the processing substrate 12. This is because priority is given to reducing damage to the processing substrate and processing itself.

  At the start of power application of the second high-frequency power source and the third high-frequency power source, the phase controller 26 does not perform phase control but fixes the phase angle to a predetermined value in the preset mode. This is because the phase angle changes greatly during the matching operation of the matching units 20 and 24, and there is no guarantee that correct phase control operation can be performed even if phase control is performed during the matching operation. Of course, in order to minimize the time from the start of operation of the matching units 20 and 24 to the stabilization, the matching start position of each matching unit must be optimized using a preset function.

  The phase controller 26 starts phase control after confirming that the power application of the high-frequency power supply is stabilized, that is, after confirming that the matching operation of the matching units 20 and 24 is stable. In order to perform a stable phase control operation, the phase angle in the preset mode is fixed to a desired phase angle measured in advance. By adopting the above sequence, the start-up of the high-frequency power source and the phase control operation can be minimized, and the processing of the processing substrate 12 can be started quickly.

  Next, how the high frequency bias is applied to the electrodes will be described with reference to FIG. The phenomenon described below occurs qualitatively in the upper and lower electrodes 2 and 14 in the same manner. FIG. 2 shows the relationship between the RF potential on the wafer surface and the plasma space potential (Vs). The Peak-to-Peak voltage Vpp and the self-bias voltage Vdc of the RF potential waveform are defined in the figure. The self-bias voltage Vdc is a potential at which the sum of the electron current and ion current flowing from the plasma into the electrode becomes exactly zero during one high frequency period, and is a time average value of the surface voltage of the electrode. On the other hand, the plasma space potential takes a constant value (defined as Vs0 in FIG. 2) while the RF voltage is sufficiently low. However, when the RF potential becomes equal to or higher than the plasma space potential, it rises by being dragged to the RF potential. At the maximum value of the RF potential, it becomes almost equal to the RF potential. The reason why the plasma space potential is increased by being dragged to the RF potential is that a large amount of electrons are extracted from the plasma due to the increase of the RF potential, so that the plasma becomes ion rich. Here, the ratio of Vdc to Vpp (Vdc / Vpp) is determined by the frequency of the high frequency and the impedance of the earth as viewed from the wafer. When the energy distribution function of electrons in the plasma is Maxwell distribution, it is always 0 <Vdc / Vpp. <0.5. Vdc / Vpp approaches 0 as the frequency of the high frequency increases and as the impedance of the ground portion increases. In other words, Vdc / Vpp is a value unique to the device and further unique to the upper and lower electrodes.

  From the description of FIG. 2, it can be seen that when a high frequency bias is applied to the electrodes, there is a moment when the space potential of the plasma rises within one period of the high frequency. This rise in plasma space potential naturally propagates through the plasma and reaches ground. That is, a part of the high frequency bias propagates in the plasma as a change in plasma space potential. The propagation waveform at this time cannot be at least a sine wave, and includes a bias frequency and its harmonics, as compared with the plasma space potential waveform of FIG. An increase in the plasma space potential accelerates ions emitted from the plasma to the ground at the ground potential. Accelerated ions cause wall spattering and chemical reaction, and the walls are scraped, causing foreign matter. On the contrary, when a reaction product adheres to the wall, it is possible to suppress the adhesion of the reaction product by using sputtering or chemical reaction by the ions.

  As shown in FIG. 1, when a high frequency bias is applied to each of the upper and lower electrodes 2 and 14, the rise in plasma space potential generated on the surface of each electrode is superimposed in the plasma. Therefore, it is possible to control the rise in plasma space potential by controlling the phase of the high-frequency bias of both electrodes. This situation is explained in detail in US Pat.

  In the present invention, based on this fact, the high frequency propagating in the plasma is used for bias phase control. In order to suppress the generation of foreign matter, it is necessary to minimize the energy of ions accelerated toward the ground. According to the explanation so far, this condition is equivalent to the minimum increase in plasma potential due to the high frequency bias. This is utilized in the second embodiment of the present invention, which is schematically shown in FIG.

  Here, the microelectrode 40 is inserted in the vicinity of the vacuum vessel wall, thereby receiving the high frequency that propagates the plasma. The signal generated in the microelectrode 40 includes the high frequency from the first high frequency power source 4, the second high frequency power source 21, and the third high frequency power source 25, its harmonics, plasma floating potential, and noise components. In order to detect the high frequency from the second high frequency power supply 21 and the third high frequency power supply 25 from this signal, there are an analog filter and a digital filter (such as a fast Fourier transform method). In FIG. The configuration used is shown.

  A reference signal is given from the oscillator 30 to the lock-in amplifier 41. The lock-in amplifier is preferably a two-phase lock-in amplifier. The output of the lock-in amplifier is equivalent to the energy of ions incident on the wall, as can be seen from the above description. The output of the lock-in amplifier is input to the phase controller 29 as a control signal.

  The phase controller 29 has a function of minimizing the output of the lock-in amplifier 41, that is, controlling the phase so as to minimize the rise of the plasma potential, or controlling the phase so as to have a specific value. . Controlling to a specific value means controlling the energy of ions emitted from the plasma 9 to the vacuum vessel wall to a certain state, and means controlling the sputtering and chemical reaction of the wall surface by ions.

  FIG. 4 shows the dependency of the output of the lock-in amplifier 41 on the phase difference in the output of the phase controller 29. At this time, the frequency of the second high-frequency power source 21 and the third high-frequency power source 25 is 4 MHz, and each power is 1 kW. It can be seen that the output of the lock-in amplifier 41 is minimized when the phase difference is around 170 degrees. The reason why the minimum value is not obtained when the phase difference is 180 degrees is that the high frequency emitted from the second high frequency power source passes through the matching unit 20, the processing electrode 14, the wafer 12, and the plasma 9 and reaches the microelectrode 40 ( That is, the impedance of the circuit is different from the phase shift until the high frequency from the third high frequency power source passes through the matching unit 24, the mixer 27, the antenna 2 and the plasma 9 and reaches the microelectrode. In general, the minimum value is not 0V. As described above, this reached the microelectrode 40 because the impedance of the path through which the high frequency from the second high frequency power supply propagates is different from the impedance of the path through which the high frequency from the third high frequency power supply propagates. This is because the voltage amplitudes at the time are not the same, and there is no guarantee that the voltages will be zero even if the voltages are superimposed. Another reason is that the amount of harmonics generated in the ion sheaths on the upper and lower electrode surfaces is different, and the lock-in amplifier cannot completely remove odd-order harmonics due to its operating principle.

  By adopting the configuration and method as shown in FIG. 3, a function of reliably controlling the impact of ions on the wall can be obtained. In addition, there is a merit that it is not necessary to use advanced techniques such as measuring high-frequency high-frequency waves at parts such as the wafer 12 and the antenna 2 and detecting the phase between the two high-frequency waves. In addition, a small signal easily output from the microelectrode 40 can be processed by a high S / N signal processing system called a lock-in amplifier 41, and the reliability of the signal processing is increased. Since it is a direct current voltage, there is a merit that signal handling is easy and reliable.

  From the viewpoint of controlling ion bombardment on the wall of the vacuum vessel, a method of detecting a substance released from the wall can be considered. The third embodiment of the present invention has been made from this viewpoint, and an outline thereof is shown in FIG. The substance emitted from the wall is collisionally excited by electrons in the plasma, and emits light having a wavelength unique to the substance. By condensing this light on the spectroscope 52 by the collimator lens 51 through the window 50, the intensity of the wavelength specific to the substance can be obtained from the spectroscope 52. This emission intensity can be used as a control signal for the phase controller 29.

  Using the configuration of FIG. 5, oxygen plasma (electron density: about 3 × 10 16 cm −3, electron temperature: about 2.5 eV) was generated. When aluminum whose surface is anodized is used for the vacuum vessel 1, the phase of the phase controller 29 is changed, and the amount of luminescence of aluminum generated by ion sputtering of the wall of the vacuum vessel 1 is shown in FIG. Shown in It can be seen that when the phase difference is decreased from 180 degrees, the light emission amount of aluminum increases abruptly when the phase difference is 80 degrees or less. Thus, it can be seen that the energy of the ions colliding with the wall exceeded the alumite sputtering threshold when the phase difference was 80 degrees or less. From the above, the sputtering of the aluminum wall can be controlled by controlling the phase difference to, for example, 90 degrees or more.

  When the wafer is processed with plasma using a CF-based gas (for example, C 4 F 8) in the plasma processing apparatus of FIG. 5, CF-based polymer is deposited as a reaction product on the wall of the vacuum vessel 1. This deposit is well known to become a foreign substance on the wafer. With this deposit deposited on the wall of the vacuum vessel 1, the oxygen plasma was generated to clean the wall. At this time, carbon (C) of the deposited film on the wall reacts with oxygen to generate carbon monoxide (CO). As a result, the deposited film on the wall is decomposed and vaporized into gas components such as carbon monoxide, carbon dioxide and fluorine, and the deposited film can be removed.

  The dependence of the carbon monoxide emission intensity on the phase difference of the phase controller 29 is shown in FIG. Unlike the case of the emission (396 nm) intensity of aluminum (Al) in FIG. 6A, the emission of carbon monoxide is observed regardless of the phase difference value. This is because carbon monoxide is generated by a chemical (combustion) reaction with oxygen. For example, carbon monoxide is generated by reaction with oxygen radicals in plasma, even if the energy of ions is small. doing. However, when the phase difference is decreased from 180 degrees, the emission intensity of carbon monoxide increases as in the case of FIG. This is because the chemical reaction with oxygen is promoted because the temperature of the deposited film rises as the ion energy increases.

  Here, when combined with the result of FIG. 4, the phase difference on the horizontal axis of FIG. 6 is equivalent to the output of the lock-in amplifier, and can be finally replaced with the energy of ions. The result is shown in FIG. As can be seen from FIG. 7, when carbon monoxide or aluminum is generated from the wall as a result of the interaction with the plasma, the generation amount and the presence / absence of the generation have ion energy dependency. Here, if the result of FIG. 7 is used, it is possible to effectively remove the deposited film when the deposited film is formed on the wall. When there is a large amount of deposited film on the wall, the phase difference is controlled so that the output of the lock-in amplifier becomes an arrow A in FIG. Thereby, the deposited film can be efficiently removed at high speed. As the removal of the deposited film proceeds, the wall background is exposed. Under this condition, light emission of aluminum, which is a wall material, appears, so that the exposure of the wall surface can be detected immediately. When the exposure of the wall background is detected, the phase difference is controlled so that the output of the lock-in amplifier becomes an arrow B in order to avoid scraping the wall material. Thereby, removal of a deposited film can be performed at high speed, without shaving the ground of a wall. When the removal of the deposited film further proceeds, the deposited film decreases, and the emission intensity of carbon monoxide decreases. When the emission intensity becomes a certain value or less, it can be determined that the removal of the deposited film is completed. That is, it is possible to determine the end point of the deposited film removal. When the deposited film removal is completed, the phase difference can be set to an arbitrary value according to the subsequent purpose. For example, the phase difference can be set so that the condition of the arrow C that minimizes the ion impact on the wall is satisfied. This deposited film removal step is desirably performed at an arbitrary interval set in advance. For example, it can be arbitrarily set according to the number of samples to be processed, and can be arbitrarily set according to the sample processing time. May be.

  The relationship between the ion energy and the output of the lock-in amplifier 41 depends on the signal processing system such as the microelectrode 40 and the lock-in amplifier 41. That is, the value is unique to the device. The relationship between the light emission amount and the phase difference (or the output of the lock-in amplifier) in FIGS. 6 and 7 also depends on the processing conditions such as the configuration of the wafer 12 to be processed and the gas used. Therefore, by collecting data as shown in FIGS. 4, 6, and 7 in advance, it is possible to obtain a control signal optimum for the apparatus and the processing process.

  As described above, by combining the signal equivalent to the ion energy shown in FIG. 3 and the signal showing the amount of the generated substance from the wall shown in FIG. It can be seen that the substances generated from can be controlled. This is equivalent to controlling the surface condition of the wall, and the present invention makes it possible to provide a vacuum vessel wall environment that is optimal for a process for processing a wafer.

  The ion energy dependency in the material generation from the wall is not limited to carbon monoxide or aluminum, and any material has a dependency specific to the material. In addition, the dependence inherent in this substance also varies depending on the plasma used. Therefore, the most effective wall surface control can be performed by selecting an optimal signal when processing a wafer or removing a wall deposition film. Naturally, this function can be used, for example, in evacuation after the vacuum vessel is opened to the atmosphere to accelerate the release of gas components such as moisture and nitrogen adhering to the wall, shortening the evacuation time of the apparatus, As a result, functions such as efficient removal of gas components such as fluorine and chlorine adsorbed on the wall and efficient removal of process-inhibiting factors in the next treatment are included.

  The signal equivalent to the ion energy described here is not limited to the high-frequency voltage signal shown in FIG. 3, but other signals such as an ion energy analyzer or plasma floating potential as long as the signal is equivalent to the ion energy. Such signals can also be used. Furthermore, as the signal indicating the generated substance from the wall, not only the emission intensity of the substance but also other signals such as a signal of a mass analyzer can be used as long as it is a signal indicating the generated substance from the wall.

  Furthermore, by simultaneously providing a device that outputs a signal equivalent to the ion energy shown in FIG. 3 and a device that outputs a signal indicating the amount of the generated substance from the wall shown in FIG. Can be done.

  The functions described above can be realized without being limited to the case where phase control is performed on the upper and lower electrodes of the parallel plate type plasma source shown in FIG. The necessary condition is that two independent high-frequency radiating parts (RF radiating parts) capacitively coupled to the plasma are required, and phase-controlled high frequency is applied to the two high-frequency radiating parts (RF radiating parts). If applied, the function of the present invention can be realized. The high-frequency radiation section (RF radiation section) here is applied with high-frequency power, and in an equivalent electric circuit related to a high frequency including the high-frequency radiation section, the plasma is included as a part of the load of the high-frequency radiation section. It is a structure. Even if the structure is not intended to be capacitively coupled with plasma, it is included in the high-frequency radiation section as long as it is coupled with plasma by stray capacitance. According to this definition, the wafer 12 shown in FIG. The following invention pays attention to the fact that many high-frequency radiation portions exist in the plasma processing apparatus in addition to the wafer and the plasma generation electrode. FIG. 8 shows an example in which the output of the third high-frequency power source is applied to the focus ring 13 beside the processing substrate 12 instead of the antenna 2. When the focus ring 13 is a conductor or a semiconductor, the effects of the phase control described so far can be realized with the configuration of FIG. Further, even when the focus ring 13 is a conductor or a semiconductor and the conductor and the semiconductor are covered with an insulator, as long as the impedance of the insulator is small and the output of the third high-frequency power source is radiated to the plasma 9, The effect of phase control can be realized.

  FIG. 9 shows a so-called inductively coupled plasma source in which a power radiation system including a coiled antenna 61 and a Faraday shield 62 is partitioned from a vacuum part by a dielectric window 60. Even in this configuration, since the Faraday shield 62 is capacitively coupled to the plasma 9, the power of the third high-frequency power supply 25 is applied to realize the effect of phase control. As can be seen from this example, if there is an antenna (electrode) that radiates the power of the third high-frequency power source 25, the effect of phase control can be realized. In this case, the parallel plate type capacitive coupling as shown in FIG. The effect of phase control can be realized by any plasma source other than the type plasma source.

  For example, as shown in FIG. 10, even if the same inductively coupled plasma source as shown in FIG. The effect of phase control can be realized. This high-frequency radiation part (RF radiation part) may be constituted by an antenna of a parallel plate electrode, a focus ring, a Faraday shield, or another electrode. Moreover, you may comprise by the combination of the antenna of a parallel plate electrode, a focus ring, and another electrode, and the combination of a Faraday shield, a focus ring, and another electrode.

The present invention can be applied to general substrate processing and processing apparatuses using plasma. Such processing apparatuses and processes include a plasma CVD apparatus and a CVD process, a plasma ashing apparatus and an ashing process, a plasma sputtering apparatus and a sputtering process, and the like. Further, the present invention can be applied regardless of the plasma generating means. Such plasma generating means include an inductively coupled plasma generating device, a capacitively coupled plasma generating device, and an Electron.
Includes a cyclotron resonance type plasma generator, a helicon type plasma generator, a surface wave discharge type plasma generator, and a magnetron type plasma generator. The processing substrate includes a semiconductor wafer, a dielectric wafer typified by quartz, an LCD substrate, and a conductive wafer typified by AlTiC.

1: vacuum container, 2: antenna, 3: shower plate, 4: first high frequency power supply, 5: shower plate support flange, 6: antenna support flange, 7: process gas introduction pipe, 8: magnet (electromagnet), 9 : Plasma, 10: Exhaust duct, 11: Vacuum exhaust device, 12: Processing substrate (wafer), 13: Focus ring, 14: Electrode (processing electrode), 15: Refrigerant groove, 16: Susceptor, 17: Insulating base, 18 : Electrode base flange, 19: Insulating tube, 20: Matching device, 21: Second high frequency power source, 22: Choke coil, 23: DC power source, 24: Matching device, 25: Third high frequency power source, 26: Phase control 27: mixer, 28: matching device, 29: phase controller, 30: oscillator, 31: wafer potential probe, 32: high voltage probe, 40: microelectrode, 41: lock-in amplifier , 50: Window 51: collimator lens, 52: spectroscope, 53: insulating pipe, 60: dielectric window 61: the coil antenna, 62: Faraday shield, 63: insulating pipe, 64: electrode.

Claims (1)

A vacuum container where the plasma in the container is produced, the first and the high frequency power source, the second RF power supply and the third RF power source, a first frequency from the first high frequency power source provided outside the vacuum vessel a coil antenna in which the voltage provided to the vacuum vessel outside Ru is supplied, and an electrode provided in the vacuum container in which the sample on the upper surface is the second high frequency voltage from being placed the second high-frequency power is supplied , a Faraday shield third RF voltage from the third RF power is supplied is provided between the coil antenna and the vacuum chamber, the second high-frequency voltage of the second electrode 3 A plasma processing apparatus comprising: a phase control device that controls a phase difference of a third high-frequency voltage from the high-frequency power source;
Voltage detecting means for detecting a high-frequency potential propagating in the plasma on the inner wall surface of the vacuum vessel; and plasma emission detecting means for detecting a light emission state of the plasma in the vacuum vessel, the voltage detecting means using a signal source is a combination of the detected plasma light by the detected high-frequency voltage the plasma emission detecting means, the control of the phase difference between the second high-frequency voltage and the third high-frequency voltage is the same frequency A plasma processing apparatus.

Images