KR20060064047A - Plasma processing method and apparatus - Google Patents

Abstract

[PROBLEMS] In a plasma processing, the range of a supply frequency to between electrodes is set such that a stable discharge and a high output efficiency can be achieved. [MEANS FOR SOLVING PROBLEMS] In a plasma processing apparatus, there are provided a secondary coil (22b) of a transformer (22) that boosts a power supply voltage; and an electrode circuit (1) comprising a pair of electrodes (11, 12) opposed to each other. There is also provided a solid dielectric (13) on the opposing surface of at least one of the electrodes (11,12). The electrode circuit (1) constitutes an LC series resonant circuit comprising a leakage inductance of the secondary coil and a capacitance of the electrodes. The supply frequency to the electrode circuit (1) is set between the resonant frequency during non-discharging and the resonant frequency during a period when the spacing (10p) between the electrodes can be regarded as a conductor.

Description

Plasma processing method and apparatus {PLASMA PROCESSING METHOD AND APPARATUS}

TECHNICAL FIELD The present invention relates to a method and apparatus for forming a plasma by glow discharge or the like in a pressure (approximately atmospheric pressure) environment near atmospheric pressure, for example, and for surface treating a target such as a semiconductor substrate.

Various methods have been proposed for plasma treatment in a pressure-treated atmosphere near atmospheric pressure. In this type of method, an electric field is applied between a pair of electrodes to cause atmospheric glow discharge and plasma treatment of the processing gas. The plasma-processed processing gas is brought into contact with a to-be-processed object such as a semiconductor substrate to perform surface treatment such as film formation or etching.

For example, patent documents; In Japanese Unexamined Patent Application Publication No. 10-36537, a pair of electrodes are disposed to face each other, and a solid dielectric is provided on the opposite surface of at least one electrode. Then, a pulse electric field is applied between the electrodes under a pressure near atmospheric pressure to form a glow discharge plasma. The preferred pulse frequency for stable plasma treatment is in the range of 0.5 Hz to 100 Hz.

Patent literature; In Japanese Unexamined Patent Application Publication No. 2001-284099, a condition in which a glow discharge is possible regardless of the atmosphere gas species is performed. The ratio of the capacitance and the feeding frequency of the dielectric film coated on the electrode is 1400 pF / (m 2 · ㎡) or less. have.

Patent literature; In Japanese Patent Laid-Open No. 10-154598, the electric field strength between electrodes is set at 0.1 to 10 kV / mm, the frequency is set at 0.5 to 100 kHz, etc., but these values are limited to the case where the applied electric field is a pulse wave, and the like It does not apply to continuous waves.

In general, a power supply for applying an electric field between electrodes boosts a primary voltage with a transformer, and supplies a secondary voltage after boosting to an electrode. For example, when the peak-to-peak voltage (Vpp) when the distance between electrodes is 1 mm is about 5 kV, it only glows in a lean gas atmosphere, and the output itself is not sufficient. When the peak-to-peak voltage Vpp when the distance between electrodes is 1 mm is 10 to 20 kV, glow discharge can be caused even in an atmosphere other than lean gas such as air or nitrogen. The frequency is fixed at a numerical value (e.g., 10 Hz) to ensure the stability of the output. However, power supply efficiency is low and sufficient output cannot be obtained. When the peak-to-peak voltage Vpp exceeds 20 kV, arc discharge occurs.

In the above conventional methods, various numerical values such as frequencies for generating a glow discharge in an approximately atmospheric pressure environment have been proposed in various ways, but these values mean only when the waveform of the electric field, the kind of the processing gas, etc. are predetermined. Not universal In addition, the output efficiency of the power supply device is not taken into account, so that the loss may increase. Power supply efficiency from a power supply, stability of output, or high processing capacity for an object to be processed can be ensured, and conditions that can be universally applied have not yet been established.

The inventors earnestly studied and considered in view of the above circumstances. That is, an electrode structure composed of a pair of opposing electrodes can be considered to be a series connection of the impedance of the interelectrode space and the capacitance of the solid dielectric provided on the opposing surface of at least one electrode. In general, a transformer is interposed between the electrode structure and the power supply, and a hot electrode is connected to the secondary coil of the transformer. Since the transformer has a leakage inductance, it can be said that the LC series resonant circuit is constituted by this and the electrode structure. As is known, in the LC series resonant circuit, the output can be maximized when the power source is driven at the resonant frequency. On the other hand, in the electrode structure, the capacitance of the solid dielectric can be easily obtained, but the impedance of the inter-electrode space varies depending on the state of the plasma or the like, and it is not easy to directly interpret it.

By the way, if it is not discharged, the impedance of the space between electrodes is also constant. At this time, the resonance frequency can be calculated by knowing the physical properties such as the permittivity of the processing gas, and of course, can be easily obtained by actual measurement.

When discharge starts, it is considered that the impedance of the inter-electrode space is lowered, and the resonance frequency becomes smaller than during non-discharge.

In addition, when the arc discharge occurs through the glow discharge, the inter-electrode space can be regarded as a conductor, so the impedance of the entire electrode structure can be considered to be as large as that of the solid dielectric. The resonance frequency at this time can be calculated by calculation. In addition, measurement can also be performed by using the equivalent circuit which substituted the space | interval between electrodes with a conductor.

As a result of these studies, the following findings were obtained. That is, stable glow discharge can be obtained by setting the energizing frequency to the electrode within the range between the resonance frequency at the time of non-discharge and the resonance frequency when the space between the electrodes is regarded as a conductor. In addition, a frequency capable of high output may be present within this range.

The 1st characteristic of this invention is based on this knowledge,

An electrode circuit 1 including a pair of electrodes 11 and 12 facing each other and an inductor 22b is provided, and at least one electrode is provided with a plasma processing apparatus provided with a solid dielectric 13 on an opposite surface thereof. The process gas is introduced into the space 10p between the said electrodes, and it feeds to the said electrode circuit 1, and performs a plasma process,

The power supply frequency fs to the electrode circuit 1 at the time of the process is the resonance frequency f _r1 at the time of non-discharge (hereinafter referred to as "the first resonance frequency" suitably) and the space between the electrodes. Is set between the resonant frequencies f _r2 (hereinafter referred to as "second resonant frequencies" as appropriate) when it can be regarded as a conductor.

Also,

A device for converting a processing gas into plasma by feeding power from the power source 21 to perform plasma processing,

The power supply may include a pair of electrodes 11 and 12 and an inductor 22b provided with a solid dielectric 13 on at least one opposing surface while forming a space 10p in which processing gases are introduced to face each other. The electrode circuit 1 fed from 21;

The power supply frequency fs from the power supply 21 to the electrode circuit 1, the resonance frequency f _r1 at the time of non-discharge of the interelectrode space 10p, and the interelectrode space 10p are conductors. It is provided with the frequency setting part 23 set between the resonance frequencies f _r2 when it can be considered.

As a result, a stable discharge state can be obtained, and the frequency range in which the peak frequency at which the high output efficiency is present can be set universally.

Here, a current is generated by adjusting the feed frequency fs to the electrode circuit 1 between the first resonance frequency f _r1 and the second resonance frequency f _r2 while causing discharge in the interelectrode space 10p. It is preferable to determine the frequency f _PEAK to be a peak and to set the feed frequency fs at or near the peak frequency f _PEAK to execute the processing. Thereby, high output efficiency can be obtained reliably.

The pair of electrodes 11 and 12 are connected in series between the capacitance component Cp of the interelectrode space 10p filled with the processing gas and the capacitance component Cd of the solid dielectric 13. One resonant frequency f _r1 may be calculated. The second resonance frequency f _r2 may be calculated using only the pair of electrodes 11 and 12 as the capacitance component Cd of the solid dielectric 13.

Instead of the calculation, an electric field having an amplitude less than a threshold value at which discharge occurs due to power supply to the electrode circuit 1 is applied between the electrodes, and the power supply frequency is adjusted so that the frequency at which the current peaks is defined as the first resonance frequency ( f _r1 ). Further, the pair of electrodes 11 and 12 are brought into contact with each other with a solid dielectric 13 interposed therebetween to eliminate the inter-electrode space 10p (circuit diagram on the right in FIG. 3) to adjust the feed frequency, thereby providing a current. The frequency at which the peak becomes a peak may be the second resonant frequency f _r2 .

The plasma processing apparatus is configured to boost the voltage from the power supply 21 to the transformer 22 to feed the electrode circuit 1, and constitute the inductor by the leakage inductance L of the transformer 22. It is desirable to.

The electrode circuit 1 is constituted by applying a real inductor Lx, Ly or a capacitor Cx, Cy to an inductor composed of the leakage inductance L and a capacitor composed of the pair of electrodes 11 and 12. You may also Thereby, the 1st and 2nd resonant frequencies f _r1 and f _r2 can be adjusted, and also the setting range of the feed frequency fs at the time of a process can be adjusted.

Further, the inventors have made a thorough study and found that there is a constant relationship between the amount of shift of the feed frequency with respect to the natural vibration frequency during discharge, the power efficiency and the stability of the output.

That is, as shown by the solid line in Fig. 12, in the region R3 where the deviation of the feed frequency fs with respect to the natural vibration frequency f ₀ at the time of discharge (estimated) is very large in both positive and negative directions. The efficiency is very low, and furthermore, it is flat with little change for the value of the feed frequency fs. Here, the ratio (Vpp / V) of the input voltage V and the peak-to-peak voltage Vpp of the hot electrode 11 is used as an index of power supply efficiency. In the constant region R2 on the side closer to the natural vibration frequency f ₀ than the flat region R3, the slope is drawn such that Vpp / V increases as the natural vibration frequency f ₀ approaches. In the regions R2 and R3, Vpp / V is almost uniquely determined with respect to the feed frequency fs. That is, Vpp / V does not fluctuate in time and is stable.

Slope area (R2) than when the area (R1) in a range closer to the natural frequency (f _0), go to increase with the passage of Vpp / V The time. However, the elevation is relatively gentle and controllable. Further, when the natural frequency area (R0) of a predetermined range including a (f ₀₎ the natural frequency (f ₀₎ becomes close to, the Vpp / V is the inability to control the momentary surge. The broken line in FIG. 12 shows Vpp / V immediately after the start of discharge in the slow fluctuation region R1 and the instantaneous fluctuation region R0. This Vpp / V is peaked at the natural vibration frequency f ₀ .

The natural oscillation frequency f ₀ varies depending on the discharge state and is difficult to specify. However, since the natural vibration frequency f ₀ can be considered to depend on the applied voltage between the electrodes 11 and 12, it is estimated by experiment or the like at the same applied voltage. It is possible to do

The second feature of the present invention is based on the above findings,

By supplying electric power to the electrode circuit 1 constituting the LC circuit including the pair of electrodes 11 and 12 and the inductor 22b, an electric field is applied to the space 10p at approximately normal pressure between the electrodes to discharge the electric charge. Is a method of causing plasma treatment,

A step of preliminarily estimating the natural vibration frequency f ₀ of the electrode circuit 1, that is, the LC circuit, at the time of discharge in the approximately normal pressure space 10p between the electrodes;

A setting step of shifting and setting the feed frequency fs to the LC circuit 1 from the estimated natural vibration frequency f ₀ ;

The main processing step of performing plasma processing by feeding power at the set frequency fs is performed.

Thereby, the general setting method of the range of the power supply frequency fs which can ensure stability while raising power efficiency can be provided.

In this second aspect, the atmosphere gas may be a lean gas such as helium or argon, or a gas other than a lean gas such as air or nitrogen. The feed waveform to the LC circuit 1 may be a continuous wave such as a sine wave or a square wave or an intermittent wave such as a pulse wave.

Here, the approximately normal pressure (pressure near atmospheric pressure) refers to the range of 1.013 × 10 ⁴ to 50.663 × 10 ⁴ Pa, and in consideration of ease of pressure adjustment and simplified device configuration, preferably 1.333 × 10 ⁴ to 10.664 × 10 ⁴ pa, more preferably 9.331 × 10 ⁴ to 10.397 × 10 ⁴ Pa.

In the step of estimating the natural vibration frequency f ₀ , after the power supply is temporarily supplied to the LC circuit 1, attenuation vibration is generated in the LC circuit 1 by cutting the power supply. The voltage applied to the electrodes 11 and 12 in the initial stage is approximately equal to the set applied voltage in the present processing step so that the initial frequency of the damping vibration is measured, and the measured natural vibration frequency is measured. (f ₀ ) may be used (see FIG. 11). Thereby, the natural vibration frequency at the time of discharge can be estimated reliably. The temporary feeding may be an intermittent wave or a continuous wave. In the case of intermittent wave feeding, the power cut is made by turning off each wave element of the intermittent wave. Therefore, it is only necessary to measure the initial frequency of the attenuation oscillation occurring in the LC circuit during the pause period between one wave element of the intermittent wave and the next wave element. The pause period is preferably a length at which the attenuation sufficiently converges. In the case of continuous wave feeding, the continuous wave may be turned off and the initial frequency of the attenuation vibration thereafter may be measured. The waveform of each wave component or continuous wave of the said intermittent wave can be variously selected, for example, square wave, sine wave, or triangular wave may be sufficient. The intermittent wave may be a pulse wave.

Further, in the estimation step, the frequency at the point where the input / output ratio is maximized by sweeping the feed frequency while making the applied voltage between the electrodes approximately equal to the set applied voltage in the present processing step is obtained. It may be as in the natural frequency (f _0).

In the setting step, the feed frequency fs is preferably set at least from the region R0 at which the input / output ratio at the periphery of the estimated natural vibration frequency f ₀ changes instantaneously. This can prevent the output from congestion. From the standpoint of the stability of the output, the input / output ratio around the estimated natural oscillation frequency f ₀ is shifted and set not only in the region in which the input / output ratio fluctuates in time, that is, the instantaneous variation region R0, but also in the gentle variation region R1. It is desirable to. In other words, it is preferable to set the input / output ratio in the regions R2 and R3 that are stable in time. However, in the case of a short time process, it can set to the said gentle fluctuation area | region R1.

More preferably, in this main processing step, the feed frequency fs is set in the slope region R2 in which the input / output ratio is stable in time and increases or decreases according to the feed frequency fs. More preferably, the feed frequency fs is a boundary between the time-varying regions R0 and R1 in the regions R2 and R3 where the input / output ratio is stable in time, that is, the slope region R2 and the gentle fluctuation region. Set at the boundary of (R1). Thereby, the stability of an output can be ensured and power efficiency can be made high.

Power supply to the LC circuit 1 is performed by boosting the output voltage of the inverter 21a with the transformer 22, and the transformer 22 constitutes the inductor component 22b of the LC circuit 1. It is desirable to do it.

In addition, when the direct current is converted into the alternating current from the inverter 21a and the voltage is boosted by the transformer 22 to feed the LC circuit 1,

When the peak-to-peak voltage (Vpp) between the electrodes (11, 12) and the direct current input voltage (V) ratio (Vpp / V) is taken as the "input and output ratio", and when performing the estimation process or the setting process It may be used as a parameter of.

As shown in Fig. 12, according to the experiments of the inventors, the deviation of the feed frequency fs is instantaneously fluctuated within a range of about 25% of the estimated natural vibration frequency f ₀ (fs = 0.75f ₀ to 1.25f ₀ ). Region R0, and the range of ± about 25% to ± 50% (fs = 0.5f ₀ to 0.75f ₀ , 1.25f ₀ to 1.5f ₀ ) is the gentle fluctuation region R1, and ± about 50% to ± 80% of range (fs = 0.2f ₀ To 0.5f ₀ , 1.5f ₀ to 1.8f ₀ ) are the slope region R2 with stable input / output ratio, and the range (fs ≤ 0.2f ₀ , fs ≥ 1.8f ₀ ) of ± about 80% or more was the flat region R3. . Therefore, in order to prevent at least the output of the congestion in the present process, it will be moved ± set at least about 25% of the estimated natural frequency (f ₀₎ to the power supply frequency (fs) in the setting step. In order to reliably ensure the reliability, it is preferable to set ± move at least about 50% of the estimated natural frequency (f _0). In order to ensure stable and good power efficiency, it is preferable to set it to ± about 50% (fs = 0.5f ₀ , 1.5f ₀ ) of the estimated natural vibration frequency f ₀ .

Instead of the estimated natural vibration frequency f ₀ , the power supply frequency fs at the time of processing may be set based on the input / output ratio (for example, Vpp / V). For example, after making the applied voltage between the electrodes 11 and 12 equal to the set applied voltage at the time of processing under the normal pressure, the feed frequency fs is swept, so that the feed frequency fs and the input / output ratio A preliminary step of preliminarily obtaining a relationship with the step, a setting step of setting the feed frequency (fs) at the time of processing in a range such that the input / output ratio is less than or equal to a predetermined percentage (for example, about 70%) with respect to the maximum value thereof; The main processing step of performing plasma processing may be performed by feeding power at the set frequency fs. Thereby, at least instantaneous fluctuation area | region R0 can be avoided.

Since the present invention is to perform plasma processing by glow discharge instead of corona discharge or the like, it is preferable that the electrode is shaped to form an equal electric field. It is preferable that the part where an electrode (or dielectric material) discharges is planar shape (this planar part is called a "discharge surface" hereafter). Moreover, it is preferable that the distance between a pair of electrodes is substantially constant (discharge surfaces of a pair of electrodes are parallel). Thereby, arc discharge by electric field concentration can be prevented, and uniform glow discharge can be generated. 0.5 mm or more and 20 mm or less are preferable, and, as for the distance between discharge surfaces, 1 mm or more and 7 mm or less are more preferable. Although the discharge surface may be a curved surface, the larger the radius of curvature is preferable (R = 5 mm or more), and the plane is more preferable. Moreover, it is preferable that the discharge surface is smooth (smooth). Unevenness or bumps are undesirable because the flames stand out. As an electrode structure that satisfies these conditions, a roll concave composed of a parallel plate electrode type in which a pair of flat electrode faces in parallel, a roll (cylindrical) electrode, and a concave electrode having a cylindrical concave along its peripheral surface And a surface electrode type, and a coaxial cylindrical electrode type composed of a pair of cylindrical electrodes forming a coaxial inside and outside.

1 shows an embodiment according to a first aspect of the present invention, which is a schematic circuit diagram of an atmospheric pressure plasma processing apparatus.

2 is an equivalent circuit diagram of the electrode circuit of the apparatus.

3 is an illustration of a method of measuring a second resonant frequency in the above apparatus.

4 is a circuit diagram showing a modification of the electrode circuit of the apparatus.

Fig. 5 is a circuit diagram showing another modification of the electrode circuit of the above apparatus.

6 is a graph showing measurement results of a relationship between frequency and current according to the first embodiment.

7 is a graph showing a measurement result of the relationship between the output and the light emission intensity of the plasma according to the first embodiment.

8 is a graph showing measurement results of the relationship between the processing conditions and the processing capacity (contact angle for each conveyance speed) according to the first embodiment.

9 is a graph showing a measurement result of a relationship between processing conditions and processing capacity (contact angle for each conveyance speed) according to the third embodiment.

Fig. 10 shows an embodiment according to the second aspect of the present invention, which is a schematic circuit diagram of an atmospheric pressure plasma processing apparatus.

FIG. 11 is a waveform graph of the inverter output voltage V1 and the electrode voltage V2 for estimating the natural vibration frequency by the attenuation waveform in the apparatus of FIG.

12 is a graph showing the relationship between the input / output ratio and the power supply frequency during discharge.

[Description of the code]

1: electrode circuit (LC circuit)

10: electrode structure

10p: interelectrode space

11: hot electrode

12: Earth electrode

13: solid dielectric

20: field applying device (power supply device)

21: alternating power

21a: inverter

22: trance

22a: primary coil

22b: secondary coil (inductor)

23: frequency setting unit

Lx, Ly: Real Inductor

Cx, Cy: Real Capacitors

fs: feed frequency

f _PEAK : peak frequency

f _r1 : first resonant frequency

f _r2 : second resonant frequency

f ₀ : estimated natural frequency

R0: instantaneous fluctuation range

R1: Slow fluctuation range

R2: Stable Slope Area

R3: stable flat area

EMBODIMENT OF THE INVENTION Hereinafter, embodiment of this invention is described with reference to drawings.

[Embodiment According to the First Feature]

As schematically shown in FIG. 1, the atmospheric pressure plasma processing apparatus includes an electrode structure 10 and an electric field application device (power supply device) 20. The electrode structure 10 is composed of a pair of electrodes 11 and 12 facing each other. A solid dielectric 13 is provided on at least one opposing surface of the pair of electrodes 11 and 12. Although only the earth electrode 12 is provided here, it may be provided in the hot electrode 11 and may be provided in both the electrodes 11 and 12. The processing gas is introduced into the space 10p between the electrodes 11 and 12 (between the hot electrode 11 and the solid dielectric 13 of the earth electrode 12) by a processing gas introduction unit (not shown).

The electric field applying device 20 has an alternating power supply 21 and a transformer 22. The alternating power source 21 includes, for example, a rectifying unit for rectifying commercial alternating current to a DC voltage, and an inverter (see symbol 21a in FIG. 10) for switching the DC voltage to output an alternating voltage of a desired frequency. A frequency setting unit 23 is connected to the inverter of the alternating power source 21, and the frequency setting unit 23 can set and adjust the output frequency of the alternating voltage, that is, the feed frequency fs to the electrode circuit 1. It is supposed to be. The output may be a continuous wave such as a sine wave or an intermittent wave such as a pulse wave.

The transformer 22 has the primary coil 22a connected to the inverter of the alternating power supply 21 and the secondary coil 22b connected to the electrode 11, and boosts the output voltage of the alternating power supply 21, and the electrode 11 ) To be supplied.

As a result, an alternating electric field is applied to the interelectrode space 10p to cause glow discharge, thereby causing the process gas from the process gas introduction portion to be plasmalized (including activation, ionization, and the like). The plasma-processed processing gas is brought into contact with an object to be processed, such as a semiconductor substrate, to thereby surface-treat the object. In addition, this process is performed under the pressure (approximately normal pressure) of atmospheric pressure vicinity.

The electrode circuit 1 is configured by the secondary coil 22b and the electrode structure 10 of the transformer 22. The transformer 22 has a leakage inductance L. In addition, the electrode structure 10 can be regarded as a capacitor. Therefore, the electrode circuit 1 can be considered as an LC series resonant circuit. The resonance frequency f _r is represented by the following equation.

[Equation 1]

Where L is the leakage inductance of the coil 22b and C is the capacitance of the electrode structure 10.

2 is an equivalent circuit of the electrode circuit 1. The electrode structure 10 is a series connection of the impedance component Zp in the interelectrode space 10p and the capacitance component Cd in the solid dielectric 13. The impedance component Zp in the interelectrode space 10p is represented by the parallel connection of the capacitance Cp and the resistance R of the interelectrode space 10p. The capacitance Cd of the solid dielectric 13 is determined by the dimensional shape and the dielectric constant such as the thickness and the cross-sectional area of the solid dielectric 13 and can be easily calculated.

When no discharge occurs in the interelectrode space 10p (at the time of non-discharge), R = ∞ in the equivalent circuit. Therefore, the capacitance [C (= C ₁ )] of the electrode structure 10 becomes as follows.

[Equation 2]

The capacitance Cp of the interelectrode space 10p at the time of non-discharge is easily calculated based on the dimensional shape such as the thickness and the cross-sectional area of the space 10p and the physical properties such as the dielectric constant of the processing gas filled in the space 10p. can do. Furthermore, the capacitance C at the time of non-discharge can be easily calculated by Formula (2).

In addition, the resonance frequency f _r (= f _r1 ) of the electrode circuit 1 at the time of non-discharge is represented by the following equation.

[Equation 1 (a)]

By these formulas (1a) and (2), the resonance frequency f _r1 of the electrode circuit 1 at the time of non- _discharge can be calculated easily. Hereinafter, the resonant frequency f _r1 at the time of non- _discharge is suitably called "the first resonant frequency f _r1 ".

On the other hand, when arc discharge occurs in the interelectrode space 10p, the interelectrode space 10p can be regarded as a conductor. At this time, R = 0 in the equivalent circuit of FIG. Therefore, the capacitance [C (= C ₂ )] of the electrode structure 10 is

[Equation 3]

C ₂ = Cd

Becomes In addition, the resonance frequency f _r (= f _r2 ) of the electrode circuit 1 during arc discharge is expressed by the following equation.

[Equation 1 (b)]

By these formulas (1b) and (3), the resonance frequency f _r2 of the electrode circuit 1 at the time of arc discharge can be calculated easily. Hereinafter, the resonant frequency f _r2 at the time of arc discharge (when the interelectrode space 10p is regarded as a conductor) is appropriately referred to as "second resonant frequency f _r2 ". The second resonant frequency f _r2 is smaller than the first resonant frequency f _r1 . In other words,

[Equation 4]

f _r2 <f _r1

to be.

In addition, the 1st and 2nd resonant frequencies f _r1 and f _r2 can also be obtained from actual measurement. That is, the output voltage of the alternating power supply 21 is set so that the electric field of the amplitude below the threshold which discharge generate | occur | produces between the electrodes 11 and 12 is applied. Then, the output frequency is scanned to measure the current at the primary side (or secondary side) of the transformer 22. The frequency at which this current measurement peaks is the first resonance frequency f _r1 (see FIG. 6).

In addition, as shown in FIG. 3, the pair of electrodes 11 and 12 are brought into close contact with each other by the thickness of the interelectrode space 10p, thereby contacting with the solid dielectric 13 interposed therebetween, thereby eliminating the interelectrode space 10p. An electrode structure 10X is made. Thereby, it can be made circuit equivalent to the arc discharge state (state where the interelectrode space 10p can be regarded as a conductor). As described above, the output frequency is scanned to measure current. The frequency at which this current measurement peaks is the second resonance frequency f _r2 (see FIG. 6).

When performing the plasma surface treatment by the plasma processing apparatus, the frequency setting unit 23 determines the output frequency of the alternating power source 21, that is, the magnitude of the feed frequency fs to the electrode circuit 1 for the calculation or measurement described above. It adjusts so that it may enter between the 1st and 2nd resonant frequencies f _r1 and f _r2 obtained. That is, it is adjusted to be within the range of the following equation.

[Equation 5]

f _r2 <fs <f _r1

As a result, stable glow discharge can be caused in the interelectrode space 10p, and favorable plasma surface treatment can be performed.

In addition, there is always a frequency f _PEAK at which the output efficiency peaks within a range satisfying equation (5) (see Fig. 6). That is, the following equation holds.

f _r2 <f _PEAK <F _r1

By setting the feed frequency fs to this peak value f _PEAK , very good output efficiency can be obtained. In addition, since reaction in the surface of a to-be-processed object may deteriorate when reaction progresses too much in the interelectrode space 10p, in that case, the feed frequency fs may be shifted from a peak and set.

The upper and lower limits of the feed frequency fs, that is, the values of the first and second resonant frequencies f _r1 and f _r2 can be arbitrarily changed. For example, as shown in FIG. 4, the inductor Lx of the real body or the capacitor Cx of the real body are interposed in series at the front end or the rear end of the electrode structure 10 of the electrode circuit 1, or FIG. As shown in Fig. 5, the real inductor Ly and the real capacitor Cy are provided in parallel with the electrode structure 10. Thereby, the 1st and 2nd resonant frequencies f _r1 and f _r2 can be moved, and also the frequency setting range can be changed. Of course, the modification of the electrode circuit 1 is not limited to what was described in FIG. 4 and FIG. 5, A various circuit structure can be employ | adopted.

The frequency setting range represented by the above formula (5) can be applied universally regardless of the output waveform, the type of processing gas, the processing contents, the device configuration, or the like. In other words, the output waveform may be a pulse, a sine wave or a square wave. Moreover, it can apply widely to various plasma surface treatments, such as film-forming, etching, washing | cleaning, ashing, surface modification, etc., and the kind of processing gas and apparatus structure are not limited. The so-called remote type of disposing the object to be disposed outside the interelectrode space 10p and the so-called direct method of disposing inside the interelectrode space 10p can be applied. Alternatively, the present invention is not limited to the atmospheric pressure plasma treatment in the vicinity of atmospheric pressure, and can also be applied to the reduced pressure plasma treatment.

[First Embodiment]

The inventors measured the first and second resonant frequencies f _r1 and f _{r2 in the} same plasma processing apparatus as in Fig. 1 by the method of the above embodiment. That is, the output voltage of the power supply 21 was 50V, and the electric field between the electrodes 11 and 12 was less than the threshold of discharge. In addition, the frequency was scanned and the primary current of the transformer 22 was measured. As shown by the dashed-dotted line in Fig. 6, the peak of the current was nearly 65 Hz (= f _r1 ).

In addition, when the two electrodes 11 and 12 were brought into contact with each other as shown in Fig. 3, the current was measured. As shown by the dashed-dotted line in Fig. 6, the peak of the current was nearly 20 mA (= f _r2 ).

In addition, the current value of FIG. 6 is normalized and shown as the maximum value in each measurement as 100 (FIG. 7 described later also is shown). In addition, the current value in the step of obtaining the resonant frequencies f _r1 and f _r2 is also weak in the peak, and is considerably smaller than that in the glow discharge process described later.

Then, the voltage of the power supply 21 was 250V, introduce | transducing the alternating electric field between the electrodes 11 and 12, introducing 100% of nitrogen gas as process gas into the interelectrode space 10p. And the relationship between frequency and current was measured. As a result, as shown by the solid line in Fig. 6, the peak of the current appeared at 55 mA (= f _PEAK ). It was confirmed by this that the relational expression f _r2 <f _PEAK <f _r1 shown in the above formula (6) was established. In addition, the primary side current when fs = 55 mA was 9.2 A, and the input power, that is, the output, was 2300 W. It was 12 W / cm <2> in conversion per unit area of an electrode.

In addition, as shown in Fig. 7, the light emission intensity of the discharge increases in proportion to the output and reaches a maximum when f _PEAK = 55 mW, and a very good and stable glow discharge was confirmed.

In the range of 20 kV to 65 kPa in which the relational expression f _r2 <fs <f _r1 shown in equation (5) was satisfied, stable discharge was obtained in the entire region of the interelectrode space 10p. At 65 mA (= f ₁ ) or more and 20 mA (= f ₂ ) or less, it was difficult to obtain a desired discharge.

In addition, the cleaning ability (contact angle and conveyance speed) of the glass was compared with the condition A of the output 2500 W, the frequency of 55 Hz, and the condition B of the output of about 1/2 W of the frequency of about 1/2 W. The conveyance speed of glass was made into 2 types, 1 m / min and 2 m / min. In addition, under condition (A), an alternating electric field converted from direct current was applied, whereas under condition (B), a pulse electric field was applied. As a result, as shown in FIG. 8, it was confirmed that condition (A) had an equivalent contact angle at twice the conveyance speed with respect to condition (B), and the processing capacity was almost proportional to the output.

Second Embodiment

Using the same apparatus as in the first embodiment, the treatment gas was measured for the relationship between frequency and current instead of argon gas. As a result, almost the same results as in Fig. 6 were obtained. In the range of 20 kV to 65 kPa in which the relational expression f _r2 <fs <f _r1 shown in equation (5) was satisfied, stable discharge was obtained in the entire region of the interelectrode space 10p. At 20 kPa or less, the output shifted to spark discharge. Above 65 mW, transition to arc discharge could not be performed in a moment and stable discharge could not be performed.

Third Embodiment

In the apparatus of the first resonant frequency f _r1 = 190 Hz and the second resonant frequency f _r2 = 75 Hz, using nitrogen gas as the processing gas and measuring the frequency f _PEAK as in the first embodiment. It was. As a result, it was confirmed that f _PEAK = 150 kPa and the relational expression f _r2 <f _PEAK <f _r1 shown in the formula (6) was established. Needle discharge was obtained at 190 Pa or more. The discharge did not occur at 75 Pa or less.

In addition, the glass cleaning ability under the condition (C) of the alternating electric field converted from 2000 W, the frequency of 150 kHz, and the direct current, and the output of 1200 W, the frequency of 30 kHz, the condition of the pulse electric field (D) of about 1/2 Contact angle and conveyance speed). As a result, as shown in Fig. 9, the condition (C) became an equivalent contact angle at twice the conveyance speed with respect to the condition (D), and it was confirmed that the processing capacity was almost proportional to the output.

[Embodiment According to the Second Feature]

Next, an embodiment according to the second feature will be described. About the structure which overlaps with the said 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted suitably.

As shown in Fig. 10, the electrodes 11 and 12 of the atmospheric pressure plasma processing apparatus of this embodiment are arranged in an atmospheric air atmosphere. The atmosphere gas may be nitrogen instead of air or a lean gas such as helium or argon. The processing gas is introduced into the space 10p between the electrodes 11 and 12 by a processing gas introduction unit (not shown). The thickness of the interelectrode space 10p is 1 mm, for example. A solid dielectric 13 is provided on at least one opposing surface of the pair of electrodes 11 and 12, but the illustration is omitted.

The power supply device (field applying device) 20 has an inverter 21a and a transformer 22. The inverter 21a switches the DC voltage V to convert to AC.

The transformer 22 has a primary coil 22a connected to the inverter 21a and a secondary coil 22b connected to the electrode 11, and boosts the output voltage of the inverter 21a to supply it to the hot electrode 11. do. As a result, for example, an alternating voltage of Vpp = 10 kV and a feeding frequency fs is applied to the interelectrode space 10p having a thickness of 1 mm, and atmospheric glow discharge occurs, thereby converting the processing gas from the processing gas introduction portion into plasma. Thus, the atmospheric pressure plasma surface treatment of the object to be treated is performed.

As described in the above first embodiment, the electrode circuit 1 composed of the secondary coil 22b (leakage inductance L) of the transformer 22 and the electrodes 11 and 12 (capacitor) has an LC series resonant circuit. It consists.

The conductance of the interelectrode space 10p is zero at the time of non-discharge, but exhibits a value other than zero at the time of discharge, and which also varies depending on the discharge state. Therefore, the natural vibration frequency at the time of discharging the electrode circuit 1, that is, the LC circuit, varies depending on the discharge state.

On the other hand, the discharge state is basically changed by the applied voltage between the electrodes. Therefore, it can be considered that the natural vibration frequency at the time of discharge depends on the applied voltage.

In the atmospheric pressure plasma processing apparatus having the above configuration, a procedure of setting the power supply frequency fs to the electrode circuit 1 during processing will be described.

[Estimate process of unique vibration frequency]

Prior to the actual plasma processing (hereinafter referred to as "main processing"), the natural vibration frequency f ₀ during the main processing (at the time of discharge) of the electrode circuit 1 is estimated in advance. Examples of the estimation method include the following sweep equation, attenuation wave equation, and the like.

(Sweep estimation method)

The feed frequency fs is adjusted in the range of 0 to several hundred kHz while adjusting the input voltage V of the inverter 21a so that the voltage applied to the electrode 11 is maintained at the magnitude (Vpp = 10 kV) during the main processing. Sweep Then, the input / output ratio (Vpp / V) is calculated from the adjustment value of the voltage (V). As a result, data as shown in FIG. 12 can be obtained. The frequency f ₀ when the input / output ratio Vpp / V becomes maximum is estimated as the natural vibration frequency of the electrode circuit 1. In this swept formula, attention is paid to output congestion in the vicinity where Vpp / V becomes the peak, that is, in the instantaneous fluctuation region R0.

(Attenuation Waveform Estimation Method)

11, and inputs a voltage (V ₁₎ of the intermittent wave shape of the frequency (f ₁₎ on the transformer (22) from the inverter (21a). Each file element of the intermittent wave (V ₁₎ is a square wave of the short period (1 / f _1A). The amplitude of each wave element is determined by the input voltage V of the inverter 21a. Due to the feeding of the intermittent wave voltage V ₁ , the vibration voltage V ₂ is generated in the electrode 11 of the electrode circuit 1. This voltage V ₂ vibrates at the same frequency as the wave element during the period in which each wave element of the intermittent wave voltage V ₁ is output. On the other hand, from the moment when the wave element is turned off, the voltage V ₂ is attenuated while vibrating at the frequency unique to the electrode circuit 1. The amplitude of each wave element of the intermittent wave V ₁ , i.e., so that the initial peak-to-peak voltage V ₂ pp of this attenuation vibration becomes equal to the peak-to-peak voltage Vpp (= 10 kV) during the main processing, i.e. The input voltage V of the inverter 21a is set.

Thereby, the inter-electrode space 10p can be made into the same discharge state as at the time of normal treatment at least in the initial stage of attenuation vibration, and can be made into the same natural frequency as at the time of this process. The frequency f ₀ of the initial stage of the damping vibration, in particular, is measured. In other words, the time (period: 1 / f ₀ ) for one cycle of the voltage V ₂ from the moment when the wave element of the intermittent wave V ₁ is turned off is measured. This is estimated as the natural vibration frequency f ₀ at the time of this processing.

The above measurement is preferably repeated every time the wave component of the intermittent wave V ₁ is turned off and the average thereof is taken. Thereby, estimation precision can be improved. The stop period t ₁ [= (1 / f ₁ )-(1 / f _1A )] between one wave element of the intermittent wave (V ₁ ) and the next wave element is set so that the attenuation is sufficiently convergent. It is preferable. Thereby, overlapping with the next attenuation wave can be avoided.

The estimated natural vibration frequency f ₀ is, for example, 100 Hz to 120 Hz.

[Setting process]

Next, on the basis of the estimated natural vibration frequency f ₀ , the power supply frequency fs at the time of this processing is set. Most preferably, it is set such that the magnitude of ± 50% of the estimated natural frequency (f _0). In other words,

fs = f ₀ × (1-0.5) = 0.5f ₀ ... (Eq. 7)

Or fs = f ₀ × (1 + 0.5) = 1.5f ₀ … (Eq. 8)

Set to. This is the frequency at which the boundary with the fluctuating region R1 in the stable slope region R2 is used. For example, when f ₀ = 120 ms, fs = 60 ms.

In the main processing step, atmospheric pressure plasma processing is performed while feeding at the set frequency fs. Thereby, output can be stabilized and power efficiency can be improved further.

In addition, as shown in Fig. 12, from the viewpoint of the stability of the output, the feed frequency fs is not limited to exactly ± 50% of the estimated natural vibration frequency f ₀ , and may be set beyond this. That is, what is necessary is just the inside of the stable areas R2 and R3 shown by following formula (9) and (10).

fs ≤ f ₀ × (1-0.5) = 0.5f ₀ ... (9)

Or fs ≥ f ₀ × (1 + 0.5) = 1.5f ₀ ... (Eq. 10)

However, it is preferable that the range of transfer is only about +/- 80%. If it moves more than this, power efficiency will become low too much and it will become difficult to obtain a desired output. That is, as shown by the following formulas 11 and 12, the flat region R3 is excluded from the stable region.

fs? f ₀ × (1-0.8) = 0.2f ₀ ... (Eq. 11)

Or fs ≤ f ₀ × (1 + 0.8) = 1.8f ₀ ... (Eq. 12)

When the equations 9 to 12 are put together, the setting range of the power supply frequency fs, which can secure the stability of the output and also increase the power efficiency, becomes the stable slope region R2 represented by the following equations (13) and (14).

0.2f ₀ ≤ fs ≤ 0.5f ₀ … (Eq. 13)

Or 1.5f ₀ ≤ fs ≤ 1.8f ₀ ... (Eq. 14)

In addition, when the processing time is short (for example, about 10 minutes to 10 minutes), the feed frequency fs may be shifted by at least ± 25% or more with respect to the estimated natural vibration frequency f ₀ , and more than ± 50% It may be set into the side of the oscillation frequency (f _0). That is, you may set in the slow fluctuation | variation area | region R1 shown by following formula (15) and (16).

0.5f ₀ &lt; fs &lt; 0.75f ₀ [= f ₀ × (1-0.25)]... (Eq. 15)

Or 1.5f ₀ &gt; fs &gt; 1.25f ₀ [= f ₀ × (1 + 0.25)]... (Eq. 16)

In this gentle fluctuation region R1, the power efficiency is very high and a large output can be obtained. The input / output ratio rises over time, but the rate is moderate, and it does not rise quickly. Therefore, if the processing is turned off after a short time, the inverter 21a and the electrodes 11 and 12 will not be destroyed.

By shifting the feed frequency fs by at least ± 25% or more with respect to the estimated natural vibration frequency f ₀ , the region R0 (0.75f ₀ <fs <1.25f ₀ ) where the input / output fluctuates momentarily can be avoided. It is possible to prevent the element and the electrodes 11 and 12 of the inverter 21a from being destroyed by the large current.

In the above description, the estimated natural vibration frequency f ₀ is used as a reference. Alternatively, the feed frequency fs may be set based on the input / output ratio Vpp / V.

In detail, as a preliminary step, the relationship between the feed frequency (fs) and the input / output ratio (Vpp / V) is first determined. The method is substantially the same as the above-mentioned "sweeping formula". That is, while adjusting the primary side voltage V of the inverter 21a so that the voltage applied to the electrode 11 is maintained at the magnitude (Vpp = 10 kV) at the time of this processing, the feed frequency fs is set from 0 to several hundred kHz. Sweep in the range of. Then, the voltage V is measured, and the input / output ratio Vpp / V with respect to the feed frequency fs is calculated to be data.

Next, as a setting step, the power supply frequency fs at the time of this process is set in the range which the said input-output ratio Vpp / V becomes 70% or less with respect to the maximum value, for example. As a result, at least the instantaneous fluctuation region R0 can be avoided, and destruction of the electrodes 11 and 12 and the inverter 21a can be prevented.

The present invention can be used, for example, in surface treatment techniques such as cleaning, film formation (CVD) and etching of semiconductor substrates in semiconductor manufacturing processes.

Claims (22)

An electrode circuit including a pair of electrodes and an inductor opposed to each other, and using a plasma processing apparatus provided with a solid dielectric on opposite surfaces of at least one of the electrodes and introducing a processing gas into the space between the electrodes; It is a method of supplying the electrode circuit and performing a plasma treatment, Resonance when the feed frequency to the electrode circuit at the time of processing is regarded as the resonance frequency (hereinafter referred to as "the first resonance frequency") at the time of non-discharge and the space between the electrodes can be regarded as a conductor. And a frequency (hereinafter referred to as "second resonance frequency"). The frequency at which the current becomes a peak is obtained by adjusting a feeding frequency to an electrode circuit between the first resonance frequency and the second resonance frequency while causing discharge in the interelectrode space. The plasma processing method characterized by setting a feed frequency in the vicinity and performing a process. The plasma of claim 1, wherein the pair of electrodes are connected in series between a capacitance component of an interelectrode space filled with a processing gas and a capacitance component of a solid dielectric to calculate the first resonance frequency. Treatment method. The plasma processing method according to claim 1, wherein the second resonant frequency is calculated using only the pair of electrodes as a capacitance component of a solid dielectric. The plasma processing method according to claim 1, wherein an electric field having an amplitude below a threshold at which discharge occurs is applied between the electrodes and the frequency thereof is adjusted so that the frequency at which the current peaks is set as the first resonant frequency. . The plasma processing method according to claim 1, wherein a feeding frequency is adjusted in a state in which the pair of electrodes are in contact with a solid dielectric, and the frequency at which the current peaks is set as the second resonance frequency. 2. The plasma processing method according to claim 1, wherein the plasma processing device is configured to step up a voltage from a power supply into a transformer and to feed the electrode circuit, and the inductor is configured by a leakage inductance of the transformer. 8. The electrode circuit according to claim 7, wherein the electrode circuit is configured by applying a real inductor or a capacitor to an inductor composed of the leakage inductance and a capacitor composed of the pair of electrodes, thereby adjusting the first and second resonance frequencies. Plasma processing method characterized in that. A device for converting a processing gas into a plasma by powering from a power supply and performing a plasma processing, An electrode circuit fed from the power supply, including a pair of electrodes and an inductor provided with a solid dielectric on at least one opposing surface while forming a space in which processing gases are introduced therebetween; And a frequency setting section for setting a power supply frequency from the power supply to the electrode circuit between a resonance frequency at the time of non-discharge of the inter-electrode space and a resonance frequency at the time where the inter-electrode space can be regarded as a conductor. Plasma processing apparatus, characterized in that. By feeding an LC circuit including a pair of electrodes and an inductor, a method is performed by applying an electric field to a space of approximately normal pressure between the electrodes, causing a discharge, and performing plasma treatment. A step of preliminarily estimating the natural vibration frequency of the LC circuit at the time of discharge in the substantially normal pressure space between the electrodes; A setting step of shifting and setting the feed frequency to the LC circuit from the estimated natural vibration frequency; A plasma processing method comprising performing the main processing step of performing a plasma processing by feeding power at a set frequency. 11. The method according to claim 10, wherein after the power supply is temporarily supplied to the LC circuit in the estimation step, attenuation vibration is generated in the LC circuit by cutting the power supply, and further, between the electrodes in the initial stage of the damping vibration. The applied voltage is made approximately equal to the set applied voltage in the present processing step, the initial frequency of the attenuation vibration is measured, and the measured value is the estimated natural vibration frequency. The frequency at the point where the input / output ratio is maximized by sweeping the feed frequency while making the applied voltage between the electrodes in the estimating step approximately equal to the set applied voltage in this processing step. And using the estimated natural vibration frequency as the estimated natural vibration frequency. The plasma processing method according to claim 10, wherein, in the setting step, the feed frequency is set to be shifted from an area in which the input / output ratio around the estimated natural vibration frequency fluctuates instantaneously. The plasma processing method according to claim 10, wherein, in the setting step, the feed frequency is set to be shifted from an area where the input / output ratio around the estimated natural vibration frequency varies in time. The plasma processing method according to claim 10, wherein in the setting step, the feed frequency is set to a region where the input / output ratio is stable in time and increases or decreases in accordance with the feed frequency. The plasma processing method according to claim 10, wherein in the setting step, the power supply frequency is set to a boundary with a region that changes in time in a region where the input / output ratio is stable in time. The plasma processing method according to claim 10, wherein, in the setting step, the feed frequency is shifted and set by ± 25% or more of the estimated natural vibration frequency. The plasma processing method according to claim 10, wherein, in the setting step, a feed frequency is shifted and set by at least about 50% of the estimated natural vibration frequency. The plasma processing method according to claim 10, wherein in the setting step, a feed frequency is set to ± 50% of the estimated natural vibration frequency. The plasma processing method according to claim 10, wherein the feeding to the LC circuit is performed by boosting the output voltage of the inverter with a transformer, wherein the transformer constitutes an inductor component of the LC circuit. 11. The method of claim 10, wherein the power supply to the LC circuit is made by converting a direct current from an inverter to an alternating current and boosting it with a transformer. A plasma processing method, characterized in that the peak-to-peak voltage (Vpp) between the electrodes and the input / output ratio (Vpp / V) of the direct current input voltage (V) are used as parameters when the estimation process or the setting process is performed. By feeding an LC circuit including a pair of electrodes and an inductor, a method is performed by applying an electric field to a space of approximately normal pressure between the electrodes, causing a discharge, and performing plasma treatment. A preliminary step of sweeping the feed frequency after equalizing the applied voltage between the electrodes with the set applied voltage at the time of processing to obtain a relationship between the feed frequency and the input / output ratio; A setting step of setting a power supply frequency to the LC circuit in a range such that the input / output ratio becomes a predetermined percentage or less with respect to its maximum value; A plasma processing method comprising performing the main processing step of performing a plasma processing by feeding power at a set frequency.

Images