JPH03204925A - Plasma processor - Google Patents

Abstract

PURPOSE:To enable various kinds of plasma processing to be performed avoiding the damage or pollution to a substrate, etc., of a work as well as the contamination of processing atmosphere by a method wherein the frequency of the first high-frequency power supply connected to the second electrode further connected to the first electrode is specified to be higher than the frequency of the second high-frequency power supply. CONSTITUTION:A high-frequency power supply 110 outputting the high-frequency power in the second frequency is connected to the second electrode 104 through a matching circuit 108 while another high-frequency power supply 111 outputting the high-frequency power in the first frequency higher than the frequency fed to the second electrode 104 is connected to the first electrode 107 through another matching circuit 109. The first electrode 107 is fed with the high frequency power in the first frequency (100-250MHz) to provide plasma between electrodes while the second electrode 104 is fed with the second frequency (10-50MHz) lower than the first frequency to control the self bias of the second electrode 104. Accordingly, the energy of ions with which the first electrode 107 is irradiated is sufficiently lowered so that the electrode surface may not be sputtered thereby enabling the contamination of a substrate surface to be avoided.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention relates to an apparatus used for performing various plasma processes, such as reactive ion etching (RIE), plasma chemical vapor deposition (PCVD), etc. It is.

[Prior Art] In recent years, in order to proceed with various processes at low temperatures, plasma is generated in a reduced pressure container, and integrated circuit substrates and the like are coated in the plasma atmosphere. A: Many equipment technologies have been developed to perform various physical processes. For example, various thin films of integrated circuits (conductive thin films such as Au, W, Ta, p
oly-3i.

Semiconductor thin film such as St, or SiO2, 5L3N 4
．． RIE (Reacti
veIon Etching) method, other An (CH
s)3, AnH(CH3)2 etc. as a raw material gas,
All (CHs)2 and AJ2 in plasma
There is a pCVD film formation method in which Afl is decomposed into (CH3), adsorbed onto a substrate, and then Afl is deposited by a surface reaction.

The RIE method is one that generates excited active species in a vacuum container, such as CF4F2, CCj24Cρ2,
A gas such as CF2 CR2 (hereinafter referred to as excited active species source gas) is introduced, and direct current or high frequency power is applied to the susceptor as a means for holding the substrate to cause a glow discharge to generate plasma, and the generated plasma is generated in the plasma. This is a method in which ions and excited active species are simultaneously applied to the surface to be etched to perform physical and chemical etching.This method allows anisotropic etching while maintaining a high selectivity with respect to the photoresist, which is the mask material. can be realized.

Even in glow discharge caused by high-frequency input, the substrate surface is negatively biased (this is called self-bias) with respect to the plasma in terms of direct current, but ions accelerated by the potential difference between this self-bias voltage and the plasma potential are released onto the substrate surface. The surface of the substrate is etched by the action of the excited active species adsorbed on the surface of the substrate.

FIG. 5 shows a schematic diagram of the cross-sectional structure of a typical active ion etching apparatus used conventionally. 503 is a substrate having a surface to be etched, such as a semiconductor wafer or a substrate made of glass, quartz, metal, etc., and 504 is a susceptor electrode. High frequency power is supplied to the susceptor electrode 504 via a matching circuit, and the vacuum container (chamber) 505 is normally grounded for safety. Here, a high frequency power source (RF power source) having an output frequency of 13.56 MHz is normally used. Note that there are many structures in which a flat electrode is provided above and opposite to the susceptor electrode 504.

In the actual device, in addition to the above configuration, an exhaust unit for evacuation and gas exhaust in the vacuum container 505,
Although it is provided with an inlet for gas into the interior of 05, a mechanism for taking in and taking out the base 503, etc., these are omitted in the figure to simplify the explanation.

Base body 503 such as a semiconductor wafer and susceptor electrode 504
Due to the RF power applied to the susceptor electrode 504, the surface of the susceptor electrode 504 is subjected to a direct negative self-bias with respect to the plasma, and ions accelerated by this voltage act on the surface of the substrate, promoting surface reactions, and causing a reaction on the substrate. The surface to be etched is etched.

[Problems to be Solved by the Invention] In the case of the RIE apparatus described above, in general, when increasing the etching rate, it is necessary to increase the plasma density by increasing the high frequency power.

However, in conventional devices, when the radio frequency power is increased, the self-bias of the electrode is also increased, and at the same time, the plasma potential is also increased. As a result, the substrate is irradiated with ions having large energy that are accelerated by the voltage difference between the large self-bias and the plasma potential. Therefore, the following problems arise.

(2) If the irradiated ion energy increases, the etching of the resist will also fail, causing a change in pattern dimensions, and as a result, accurate microfabrication will no longer be possible. In particular, this phenomenon will be noticeable in future highly integrated devices in which the thickness of the resist will be approximately 0.5 μm or less.

■Because ions with high energy are irradiated,
This damages the underlying material, leading to a decline in the performance and reliability of devices made of such material. In particular, this will cause serious problems such as increased leakage current and deterioration of withstand voltage.

■Since the plasma potential is usually around +50 to +100■, ions determined by the plasma potential collide with the inner surface of the chamber.This high-energy ion collision causes sputtering of the inner surface of the chamber, and the material forming the chamber, e.g. Fe.

Ni, Cr, Cu, etc. contaminate the substrate surface. That is,
This is contamination of the substrate surface of the chamber constituent materials due to high-energy ion collisions. If the substrate surface is contaminated with such heavy metals, defects may occur on the substrate surface during the next high-temperature process, or leakage current may increase, resulting in a significant deterioration of device characteristics.

In addition, in the conventional device, the frequency of the high frequency power source is 13.
I am using a 56MHz one, but 13.56MHz
When the plasma excitation frequency is low, the self-bias of the direct current generated at the electrode becomes negative and large even if the chamber gas pressure and high-frequency power are constant. Figure 3(a) shows an Ar
Gas pressure: 5xlO”3 Torr, high frequency power: 50W
This shows the current/voltage characteristics when In the figure, the horizontal axis represents the negative DC voltage applied to the electrode, and the vertical axis represents the current flowing through the electrode. A negative value of the current means that electrons are flowing into the electrode, and a positive current means that positive ions are flowing into the electrode. The negative voltage when the current is O corresponds to the self-bias of the electrode. This is because high frequency power is normally supplied to the electrodes via a capacitor and no direct current flows through them.

As can be understood from Figure 3(a), the self-bias of the electrode is determined by the frequency of the high-frequency power being 14MHz to 40.68MHz.
-400V and -26 when z and 100MHz, respectively.
0V, -90V.

That is, even if the electrode structure, gas pressure, and power are held constant, the negative self-bias of the electrode becomes progressively smaller as the frequency increases.

FIG. 3(b) shows the details.

That is, the Ar gas pressure in the chamber is 7×10
rr, shows how the self-bias of the electrode changes when the high frequency of the high frequency power for plasma excitation is changed from 10 MHz to 210 MHz when the high frequency power is 100 W, the electrode spacing is 3 cm%, and the electrode diameter is 10 cm. As the frequency increases, the negative self-bias decreases rapidly. In FIG. 3(b), the plasma potential is also shown, and even if the frequency changes from 10 MHz to 210 MHz, this plasma potential almost always increases by +20 MHz.
It is maintained at V.

As LSIs become ultra-fine and highly integrated, the aspect ratios of contact holes and via holes gradually increase. That is, it is required to etch thin and deep holes with good controllability and good reproducibility. It is necessary to set the gas pressure in the etching chamber low (eg, on the order of 10-' Torr) to lengthen the mean free path of the molecules. In order to generate sufficiently high-concentration plasma and increase throughput even when the gas pressure is low, it is desirable that the frequency of discharge excitation be high. However, it is not desirable that the wavelength of the discharge excitation frequency be shorter than the diameter of the susceptor electrode 504. This is because higher-order mode discharge occurs, and plasma with a uniform density is not excited within the electrode, making it impossible to obtain uniform etching performance.

That is, in the conventional apparatus, it is not possible to independently and directly control the plasma density, that is, the ion irradiation amount and irradiation ion energy, and the conditions such as the pressure, flow rate, and high frequency power of the excited active species source gas cannot be controlled. It is necessary to control the situation indirectly by combining them appropriately.

Furthermore, in addition to the above-mentioned RIE apparatus, devices that should be configured to be able to process objects to be processed at high speed without causing damage to objects other than the objects to be processed in plasma include a PCVD apparatus, 02 plasma resist asher, Examples include dry cleaning equipment, but in the past, these equipment had been individually designed and produced, even though their basic parts had common usage conditions. At the same time, it also had the disadvantages of (1), (2), and (3) mentioned above.

The above-mentioned problems were discovered by the present inventor, and the present inventor conducted extensive research to solve the above-mentioned problems occurring in conventional devices, and finally found a solution.

The present invention allows substrate etching and film formation on the substrate without damaging or surface contaminating the substrate (substrate), and even though the structures such as chambers and electrodes are the same,
By changing the gas introduced and the plasma excitation frequency, the present invention can be applied to etching and film formation, and the purpose is to provide a low-cost, high-performance plasma processing device with excellent productivity.

[Means for Solving the Problems] The present invention aims to provide a versatile device that can be applied to various plasma processes while avoiding the damage and surface contamination described above and achieving high-speed processing. In a plasma processing apparatus configured to generate plasma between two opposing electrodes installed in the container and to process a workpiece in the plasma, first and second electrodes formed in a flat plate shape, a protective member made of a material stable at least against the plasma and provided to cover the first electrode, and a protective member provided on the second electrode. a holding means for attaching the object to be processed; a first high-frequency power source connected to the first electrode; a second high-frequency power source connected to the second electrode; and a desired gas in the container. The first high frequency power source is characterized in that the frequency of the first high frequency power source is set higher than the frequency of the second high frequency power source.

[Function] For example, when used in an RIE apparatus, a substrate on which a thin film to be etched is formed as an object to be processed is mounted on a second electrode in a container, and the pressure in the container is reduced and a predetermined gas supply means is applied. Depending on the thin film to be etched, chlorine gas,
Introduce fluorine gas, mixed gas, etc. The first frequency (100 to 250 MHz) is applied to the first electrode.
) is supplied to generate plasma between the electrodes, and the second electrode is supplied with high frequency power at a second frequency (10 to 50 MHz) lower than the first frequency. control self-bias. That is, the generated plasma density and the amount of ion irradiation applied to the substrate are controlled by the high frequency power of the first frequency supplied to the first electrode.

On the other hand, the energy of ions incident on the substrate surface is controlled by the self-bias of the high-frequency power of the second frequency supplied to the second electrode.

Since the high frequency power supplied to the first electrode plays a role in generating plasma, the power is usually large. However, since the frequency is increased, the negative self-bias of the tenth electrode can be made sufficiently small. Therefore, the ion energy irradiated to the first electrode is sufficiently small so that the surface is not sputtered and the substrate surface is not contaminated. Second
Since the negative self-bias induced in the electrode controls the ion energy irradiated onto the substrate surface to an optimal value, there are of course no problems with damage or contamination.

When used in a PCVD apparatus, a substrate on which a deposited film is to be formed is held on the second electrode. The magnitude relationship between the first frequency and the second frequency is set in the same manner as in the RIE apparatus, but the gas introduced into the container is, for example, in the case of Si film formation, SiH4,5i).
12Cl12゜5i2H, etc., in the case of 5in2 film formation S
i A mixed gas of H4 and 02 or Si2H6 and 02 is introduced. In this case as well, damage to the substrate to be processed and contamination of the processed object can be prevented for the same reasons as described for RIE.

Furthermore, the present invention can also be applied to resist ashers, where problems of damage and contamination of the substrate surface are unavoidable in conventional techniques. For example, photoresists essential for fine pattern processing are usually H
zS in a wet process using a mixture of SO4, H, and O2
However, when used as a mask material for ion implantation, the resist is hardened by high-energy ion irradiation, so it cannot be removed by a normal wet process. Therefore, it was necessary to generate 03 and 0 radicals using 02 plasma and remove the ion-implanted resist using ion energy.

When used as a resist asher, the previous RIE or PCV
As mentioned in the explanation of D, place the substrate on the second electrode,
By controlling the self-bias of the second electrode using high-frequency power at the second frequency, the resist can be removed without damaging or contaminating the substrate surface.

As described above, although the condition settings may be slightly changed during use, it can be widely applied to various plasma processing devices.

[Example] Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1(a) shows a first embodiment in which the present invention is applied to an RIE apparatus for etching the surface of a substrate. Here, a case will be described in which a thin film formed on a semiconductor substrate is etched.

In a vacuum container (chamber) 105, an upper flat electrode 107 and a lower flat susceptor electrode 104 are arranged to face each other, and the vacuum container 105 is made of metal and is connected to ground. . The inner surface of the vacuum container 105 is made of a material that is stable against plasma of corrosive gases such as fluorine-based or chlorine-based gases, that is, coated with an oxide film, nitride film, or fluoride film to prevent corrosion even when exposed to the plasma. It is being said. The electrode 107 includes a base material 102 made of a conductive material, 5iOz formed on the surface of the base material 102,
Si3 N4. A protective layer 101 as a protective member made of AJ2203, Aj2N, or the like.

The protective layer 101 is for preventing the base material 102 from being etched by plasma generated by discharge, and is made of, for example, Si, 5in2, quartz, SiC, Si,
Na, AA203. Made of AIIN and other materials.

Furthermore, a passive film that may be composed of a fluoride having a substantially stoichiometric ratio exhibits excellent etching resistance, and the formation of the passive film can be performed, for example, as follows. Just go. That is, the base material (e.g. stainless steel,
After finishing the surface of the base material (made of nickel, nickel alloy, aluminum alloy, or other metal or alloy) using electrolytic polishing technology, etc., to a mirror-like finish without any process-altered layer,
Baking is performed at a predetermined temperature in a high-purity inert atmosphere to remove adsorbed water. After baking, fluoride with high-purity fluorine at a predetermined temperature, and then heat-treat in a high-purity inert atmosphere at a temperature slightly higher than the temperature during fluorination to satisfy approximately stoichiometric composition. A passive film is formed on the base material.

Note that if the protective layer 101 is made of Sr,
Even if the protective layer 101 is etched, the susceptor electrode 104
Since St will be mixed into the upper substrate 103, the influence on the substrate 103 can be minimized.

A high frequency power source 110 that outputs high frequency power at a second frequency f2 is connected to the susceptor electrode 104 via a matching circuit 108. This embodiment shows an example in which high frequency power of 100 MHz is supplied. For potential control of the susceptor electrode 104, the second frequency f2 is preferably 10
~50MHz is suitable. Further, the electrode 107 is supplied with a first frequency f which is a higher frequency than the frequency f2 supplied to the susceptor electrode 104 via a matching circuit 109.
A high frequency power source 111 that outputs one high frequency power is connected. This embodiment shows an example in which high frequency power of 250 MHz is added. Note that, as will be described in detail later, it is desirable that the two frequencies f+ and f2 are not in an integral multiple relationship.

Furthermore, only the first high frequency (250 MHz in this example) and the second high frequency (100 MHz in this example) are manually applied to the electrode 107 and the susceptor electrode 104, respectively.
ator) 112. 113 is provided. That is, the first high frequency wave f1 is short-circuited to the ground in the susceptor electrode 104, and the second high frequency wave f2 is short-circuited to the ground at the susceptor electrode 104.
It is shorted to ground at 02. The electrode 107
and the band eliminator 11 used in 104
2° 113 can basically have a configuration like the tank circuit 102b shown in FIG. 1(b), L, .
The parallel circuit of CI is f, = (1/2π(at r=-))
The impedance reaches its maximum at the resonant frequency (Figure 1 (
C)), other frequencies are almost always short-circuited, so the predetermined frequency (in this case f+=250MH
Only the high frequency of z) can be selected and supplied to the electrodes.

It goes without saying that the configuration of FIG. 1(b) shown here merely shows the basic principle and may be modified for various improvements. For example, FIG. 1(d) is an example of an improvement.

The circuit 102b is grounded in terms of direct current through the inductance L1, but it is placed in a floating state (f) in terms of direct current.
For example, if you want to use
) It is sufficient to add a capacitor C1 as shown in 102d to cut the DC component. In this case, the value of C needs to be a sufficiently large value to satisfy f,.L+>1/flC- so that the resonant frequency of the circuit 102d does not deviate from the frequency f1.

In this case, for the frequency to = (x/2rc (at π7)), the impedance of the series circuit of L, , C, is O.
So, the frequency is f. A short circuit occurs due to high frequency. This frequency f0 is applied to the susceptor electrode 104 at a frequency f2
By setting it equal to , it is possible to effectively prevent the high frequency of frequency f2 from being superimposed on the electrode 107. That is, even if the electric field of the high-frequency power entering the susceptor electrode 104 is vertically terminated from the susceptor electrode 104 to the electrode 107, the electrode 107 is short-circuited to the ground for the high frequency of frequency f2. The voltage of 107 does not vary with the power of frequency f2.

The above was about Band Eliminator-112, but
If the band eliminator 113 has a similar configuration, the voltage of the susceptor electrode 104 will not vary depending on the frequency f supplied to the electrode 107. That is, in the circuit of FIG. 1 ((1), the inductance L
l is the inductance L2. In order to generate ions to form plasma, capacitor CI is used as capacitor C2+ and capacitor C5 is used as capacitor C82.
The discharge of the excited active species source gas introduced at step 05 is performed by a high frequency wave having a frequency f1.

Even if the power at the frequency f+ is increased in order to increase the ion density, the voltage of the susceptor electrode 104 will not be affected.

Similarly, the frequency f supplied to the susceptor electrode 104
The same can be said about the high frequency power in item 2. This is because even if the high frequency power of frequency f2 is changed, the power of f2 is short-circuited to the ground in electric Fi 107.

An example thereof is shown in FIG. 1(e), in which the distance between the first and second electrodes is 3 cm.

Its diameter is 100m and the gas pressure is 7X10-3T.

In the state of rr, f, = 100MHz, the human power is 1
The DC self-bias of the first electrode and the second electrode is plotted when the power is kept constant at 50 W and f2=30.40°50 MHz when the power is varied. 1st
The self-bias of the electrode is about -25V? It is independent of the frequency and power supplied to the second electrode. The potential of the second electrode is about 10V when there is no high frequency input, but
As the high frequency power of frequency f2 increases, it decreases linearly, and becomes a negative voltage above a certain power. The lower the frequency f2, the greater the change in self-bias voltage for the same power change. In any case, depending on the potential of the opposing electrodes,
It is evident in FIG. 1(e) that the DC potential (self-bias) of the electrodes can be controlled by the radio frequency power and its frequency without significantly affecting the electrode.

With the above configuration, the electrode 107 and the susceptor electrode 104 can effectively prevent the high frequency waves to be supplied to the other from being superimposed, and can supply only the high frequency waves that should be supplied to each. It becomes possible to easily and accurately control the bias plasma density and the irradiated ion energy.

Note that the cylindrical magnet 106 provided on the back surface of the electrode 107
As a result, a magnetic field approximately parallel to the surface of the electrode 107 is generated, and the electrons are wrapped around this magnetic field and perform cyclotron motion. When a vertical high-frequency electric field exists between the picture electrodes 107 and 104, energy is effectively imparted to the cyclotron-moving electrons, and the high-frequency power effectively generates high-density plasma. Therefore, in this device, the electric fields of the two input high-frequency powers are set to be almost perpendicular and terminate at the susceptor electric field 8i104 and electric field Fi107, respectively.

Note that 106 is a permanent magnet for magnetron discharge. In reality, electromagnets using ferromagnetic materials are preferred. Furthermore, the apparatus is provided with an exhaust unit that evacuates the inside of the vacuum container 105, a mechanism for introducing gas, and a mechanism for roughly loading and unloading the substrate 103, but these are omitted to simplify the explanation.

In the device of this embodiment, unlike the conventional device, an electrode 107 is provided in addition to the susceptor electrode 104.
By supplying a high-frequency power source with high power to 07, high-density plasma can be generated, and high-speed etching can be performed. However, if a high frequency wave with a large amount of power is supplied to the electrode 107, the self-bias will also become large, and there is a risk that the electrode will be sputter-etched. In order to prevent such etching, the frequency f1 of the high frequency power supply 111 supplied to the electrode 107 should be changed to the frequency f2.
In addition, a protective layer 1 is applied to the surface of the base material 102 of the electrode 107.
01 is set.

On the other hand, the self-bias generated in the susceptor electrode 104 can be controlled by the power and frequency of the high-frequency power source 110, as shown in FIG. 1(e). The power and frequency of 110 may be selected and supplied to the susceptor electrode 104.

After all, if the apparatus of this embodiment is used, high-density plasma is generated by the high-frequency power supplied to the electrode 107 (plasma density, that is, ion density is controlled by the power), and ions are irradiated onto the substrate surface. Since the energy can be controlled to a desired value by the high frequency power of frequency f2 supplied to the susceptor electrode 104, high-speed RIE can be performed while preventing damage to the substrate 103 and the like.

Next, the influence of the high frequency power and frequency supplied to the electrode 107 and the susceptor electrode 104 will be described.

FIG. 2 shows a circuit configuration for measuring the current and voltage characteristics of the electric Fi 1104 using the apparatus shown in FIG.

The high frequency filter 203 connected to the chamber 8i104 is, for example, the band eliminator 1 shown in FIG. 1(b).
02b, the impedance is high only at the frequency f2 of the high frequency supplied to the susceptor electrode 104, and it is configured so that it is almost short-circuited for frequencies that deviate from that frequency. A DC power supply 201 and an ammeter 202 are connected in series.

A capacitor 206 is connected in parallel to the connection point between the high frequency filter 203 and the ammeter 202 in order to short-circuit the DC power supply 201 and the ammeter 202 in terms of high frequency.

In this state, for example, Ar gas is introduced into the vacuum container 105 for 50 minutes.
Figure 3 shows a graph of the relationship between the DC voltage V applied to the electrode 104 and the resulting current flowing when the voltage is introduced at a pressure of
a). In this case, the frequency of the high frequency power source 110 is variable, for example, 14 MHz, 40.68 MHz, and 10 MHz.
The frequency is changed to three frequencies: 0M) (2). Note that the current through which positively charged ions flow into the electrode 104 is taken as a positive value.

For example, looking at the characteristics of 100MHz, the DC voltage V
is about -95■ (this value is taken as the self-bias voltage VSa), the DC current I=O, and when V>VSB, I
< O, V < V sate is written as I>0. The self-bias voltage VsB is a DC bias voltage generated when high-frequency discharge is performed while the electrode 104 is in a floating state. That is, when the electrode 104 is at this potential, the numbers of ions and electrons flowing into the electrode 104 from the plasma are equal, so they cancel each other out, and the DC current becomes O.

On the other hand, due to the externally applied DC bias voltage, electrode 1
When the potential of 04 is controlled, a current flows.

For example, if the relationship of V > V s s is established between the DC voltage (2) and the self-bias voltage VSB, more electrons will flow, resulting in (2) <O.

On the other hand, in the case of the relationship of V x V, sB, the potential barrier to electrons increases and the number of electrons flowing in decreases, so the ionic current becomes larger and a positive current flows. Furthermore, when the DC voltage V is increased in the negative direction, the current value is saturated when V=vo, and becomes an approximately constant value. This is equal to the current value for ions only.

From the above, the slope of the IV characteristic curve near V=VsB corresponds to the width of the electron energy distribution.

In other words, a large slope means that the width of the electron energy distribution is narrow. As is clear from FIG. 3(a), the energy distribution at 100 MHz is about 1/10 smaller than that at 14 MHz. On the other hand, the width of the ion energy distribution is △E1°. When the width of the electron energy distribution is ΔE, there is a substantially proportional relationship between the two, so it can be said that the width of the ion energy distribution is similarly reduced to about 1/10.

Furthermore, v5! Although the value of l is the same high frequency power of 50W, it is 100MHz compared to 400V in the case of 14MHz.
In this case, the absolute value is about -95V, which is less than 1/4. When the power is lowered to 5W with toOMHz discharge, VS
The value of a decreases to -25V. In other words, by controlling frequency and power, self-bias can be controlled over a wide range.

In the conventional RIE method, the underlying substrate is damaged and the characteristics of the device are deteriorated, and this is due to the following reasons.

In the conventional example, the electrode 107 has a low frequency of 13.56 MHz.
Due to the discharge tee, lv,,bl=400V~5o
oov, and the ions accelerated by this high voltage collided with the substrate.

However, in the first embodiment of the present invention, the electrode 1°7 has 2
Since the discharge is performed using a high frequency of 50MHz, ΔElo is lower than the conventional case of 13.56MHz. is 1
/20 or less. In the device of the present invention, the discharge is maintained by high-frequency power at a frequency f, which is applied to the electrode 107, thereby generating a high-density plasma, and at a frequency f, which is higher than the frequency f2 which is supplied to the susceptor electrode 104. (250MH
z), the ion energy distribution width in the generated high-density plasma is also small (the number of ions having energy different from the average energy value is small). Furthermore, as will be explained later, the magnetic circuit is designed to make the magnetic field strength in the direction parallel to the electrodes as strong as possible, so the self-bias voltage is less than -30V when 50W of high-frequency power is input, and the plasma Density is approximately 1
It has been improved by more than 0 times. According to Fig. 1(e), when the high frequency power is 100W and the self-bias is about -10V at f+=210MHz, f1 is 250MHz.
Then, the self-bias is less than =5V.

The self-bias of the electrode 107 is as low as −5 V or less, and since it has the protective layer 101, the base material 102 of the electrode 107 is not sputtered at all. Therefore, it is extremely easy to control the high-frequency power or frequency f2 applied to the susceptor electrode 104 to a level so that the self-bias does not damage the substrate, and the frequency f can be adjusted so as to obtain the desired etching rate. By setting the power, ions with high energy that would damage the substrate surface will not be irradiated, and high-speed and highly selective etching can be performed without damaging the thin film, resist film, or underlying substrate. It becomes possible to do so.

That is, the self-bias voltage VSa becomes lower as the frequency of the high-frequency power source increases and as the high-frequency power decreases. Therefore, the frequency and power to be supplied to the susceptor electrode 104 may be selected to provide the ion energy and ion dose necessary for high-speed etching without damaging the quality of the thin film or underlying substrate.

On the other hand, since high-frequency power with a frequency of 250 MHz is applied to the electrode 102, a small self-bias voltage is generated, and since the protective layer 101 is formed, the base material 102 can be prevented from being etched. . Furthermore, in the embodiment shown in FIG. 1, a permanent magnet 106 is attached, which causes a magnetron discharge near the electrode 107 (electrons wind around magnetic lines of force and move in a cyclotron, receiving energy from a high-frequency electric field and generating a neutral The excited active species source gas molecules are efficiently ionized), increasing the ion concentration and further increasing the etching rate.

As described above, according to the dual frequency excitation RIE apparatus according to the present invention, it is possible to etch high-quality thin films and substrates with high selectivity without causing damage to the substrate while maintaining a high etching rate.

It is also possible to control the energy of ions flowing into the susceptor electrode 104 by applying a DC bias voltage to the susceptor electrode 104 as shown in FIG. By applying such a DC bias voltage, the susceptor electrode 1
The method of controlling the potential of 04 and, as a result, the surface potential of the substrate, is effective when the thin film or substrate (substrate) to be etched is a conductive material.

Above, only the case where the frequencies of the high-frequency power supplied to the electrode 107 and the susceptor electrode 104 are set to 100 MHz and 250 MHz, respectively, has been described, but it goes without saying that the frequency selection is not limited to this.

In short, in the case of the RIE apparatus, the first frequency f1 supplied to the electrode 107 may be higher than the second frequency f2 supplied to the susceptor electrode 104. The actual value differs depending on the purpose, and may be determined by taking into account the required etching rate, the shape of the formed film covering the stepped portion, etc. Further, the material to be etched is not limited to an insulating material, and may be a conductive material.

Further, the structure of the magnet 106 installed on the back surface of the electrode 107 is not limited to that shown in FIG.

For example, FIG. 4(a) shows a second embodiment of the present invention, and in the case of this embodiment, a strong raceway magnet 40
9 is provided to perform scanning in order to improve the uniformity of the magnetic field. In this case, the scanning system 410 of 65 stones 409
It is advantageous to provide the reaction system outside the vacuum vessel 105 to prevent the reaction system from being contaminated by dust generated from mechanical operations.

Furthermore, a magnet is also installed on the susceptor electrode 104 side to perform RI.
The efficiency of E may be increased. Further, the magnet used here may be fixedly attached like the magnet 106 shown in FIG. 1, or it may be movable like the magnet 409 attached to the scanning system 410. It doesn't matter.

Further, in order to further reduce damage to the substrate 103, it is also possible to take the following method, for example. For example, S
When etching an absolute M film such as S i O 2 formed on the surface of the substrate 103 such as i, firstly, while the film of about several μm is being formed, the R supplied to the susceptor electrode 104 is etched.
In this method, the RF power is increased to perform high-speed etching, and the RF power is decreased just before the surface of the substrate 103 begins to be exposed. In this way, after the substrate 103 begins to be exposed, etching can be performed in a sufficiently low self-bias state, so that damage to the substrate surface can be reduced to almost zero.

If the kinetic energy of the ions irradiated onto the surface of the substrate 103 is too large, any material will be damaged. Damage begins to occur in the material when the kinetic energy of the irradiated ions becomes slightly larger than the critical energy for damage generation, which is determined in relation to the interatomic bonding force of each material. The bonding force between atoms is usually greater in insulators than in semiconductors. Substrate 10
The energy of the irradiated ions may be determined in consideration of 3 and the properties of the insulating material.

FIG. 4(b) shows a third embodiment, which shows a method in which damage to the substrate 103 is eliminated and the energy of ions irradiated onto the surface of the substrate 103 can be freely selected.

The difference from the first embodiment shown in FIG. 1(a) is that f2. The difference is that it is possible to manually switch between two different frequencies called f3, and the band eliminator 401 is also changed accordingly. 402 and 403 are LC resonance circuits, each with f2. It has a resonant frequency of fs.

f2=1/(7 at 2πF7) fs=1/(T at 2yrFr7) The band eliminator 401 in which two resonant circuits 402 and 403 are connected in series has f2. Since the impedance is large only for the two frequencies of f3 and is substantially short-circuited for other frequencies, the susceptor electrode 10 is selectively connected only to these two types of high frequencies.
It has the function of supplying power to 4.

For example, let f 1 = 250 MHz and f 2 = 10
0 MHz, f3 = 40 MHz. For example, during the initial formation of a film of about 0.5 to 1 μm, if the frequency of the high frequency wave applied to the susceptor electrode 104 is f3 (40 MHz), the self-bias will be as shown in FIG. 1(e). Like O~-100V
, and a great etching effect can be obtained. When there are about 100 people on the surface, change the frequency to f, (100M
Hz) to etch a thin (eg, 10 to 100) film. In this way, when the substrate surface begins to be exposed, the substrate surface is irradiated with ions at a small self-bias value (approximately -10 to -20 V) corresponding to 100 MHz, so that almost no damage to the substrate occurs.

Such a method becomes particularly important when controlling the flatness of the surface shape of a thin film deposited by the RIE method. This is because by changing the frequency, the most effective etching ion energy can be controlled, and the optimum energy value can be selected without causing damage to the substrate 103.

Here f2. Although only the case of two different frequencies of f3 has been described, for example, f2. f3. f.

It goes without saying that these three values may be used.

However, in this case, the first applied frequency f4 is f4>
As for fz and fa, it is important to use the ones with the highest frequencies later on to minimize damage.

In addition, when using a plurality of frequencies, these include the discharge excitation frequency f+, f+, f2. f3゜・・・・
It is desirable to select them so that they do not have a harmonic relationship with each other.

The discharge space is nonlinear, so f+, f2 . fs,
This is because the harmonics of ... may be superimposed in a completely different state depending on the discharge conditions, and the setting of the conditions is no longer unique.

Note that the same effect can be achieved by using a resonant circuit shown in FIG. 4(C) in place of the resonant circuits 402 and 403 in FIG. 4(b). However, in FIG. 4(C), it is necessary to satisfy Cs > C2, Cs.

Next, the concept of high performance, which is common to various processes and is performed by creating plasma between opposing parallel plate electrodes, which is a basic component of the above-mentioned RIE apparatus, will be explained.

The two requirements for improving the performance of the discharge plasma process are (1) no damage to the substrate surface, and (2) no contamination of the substrate surface due to sputtering of the vacuum vessel or electrode material. . Of course, there are other specific requirements for high-performance etching and film-forming, such as high-speed etching and high-speed film formation, and the realization of high-density plasma with as little high-frequency power as possible. Nor.

In order to realize requirements (1) and (2), the plasma potential of the plasma formed by discharge must be at a value that does not sputter the vacuum container or electrode material, that is, +30V.
Hereinafter, the voltage is desirably required to be +20V or less. Although a vacuum container is normally used in a grounded state, the energy of ions incident on the inner surface of the vacuum container is equivalent to the plasma potential. Normally, a negative DC voltage is applied to the electrode 102 and the susceptor electrode 104 by the supply of high-frequency power, so positively charged ions are incident, but their energy is controlled to an energy value with a desired purpose. Ru. In any case, it is an essential condition that the plasma potential formed between the opposing electrodes be suppressed to a range of about +5 to +20V. The energy of each individual ion irradiating the substrate surface has the following effects: etching,
There are optimum values for each substrate surface material depending on the purpose of film formation. The individual ion energy is adjusted to the optimum value for each material by adjusting the high frequency power of frequency f2 supplied to the susceptor electrode 104, and by adjusting the self-bias voltage -V, (V) of the susceptor electrode 104, vo ,=
v, +Vs. This is because if ions do not collide between the plasma and the substrate surface, the ion energy irradiated onto the substrate surface is determined by the potential difference between the plasma potential and the substrate surface potential.

However, ■. P: optimal ion irradiation potential, Vp: plasma potential, -VS + self-bias of the susceptor electrode.

Note that such setting conditions can be applied only when the potential of the plasma formed between the opposing electrodes is suppressed to a low positive voltage. That is, VP<V. P must hold true. This is because the self-bias realized by applying high frequency power to the susceptor electrode 104 always acts in the negative voltage direction.

Therefore, the plasma potential VP is set to a low positive voltage such that Vsp>Vp and Vop>Vp are satisfied. However, Vgp is the sputtering start voltage of the vacuum container and electrode material.

As a conclusion from the above, improving the performance of plasma application equipment requires lowering the plasma potential to a lower positive voltage (Vsp>Vp).

V op > V p ). The reason why the plasma potential increases with positive voltage is mainly that negatively charged electrons, whose mass is lighter than ions, escape from the plasma space, and positively charged ions become excessive, causing the plasma to become positively charged. Depends on what you have. In other words, in order to maintain the plasma potential at a low positive potential, it is necessary to prevent electrons from escaping from the plasma space as much as possible. At the same time, it is important that the high-frequency power causes discharge and ionization to occur as effectively as possible.

Next, the DC magnetic field distribution and high frequency electric field distribution that realize these conditions will be explained using FIG. 6. Figure 6 (
a) shows a flat electrode 107 and a susceptor electrode 10 facing each other.
A DC magnetic field distribution 601 (dotted line) and a high-frequency electric field distribution (solid line) 602 for 4 are shown. FIG. 6(a) shows an example of an ideal state. That is, the opposing picture electrodes 10
7. A DC magnetic field exists parallel to the pole plates of 104, and a high frequency electric field exists perpendicularly between the pole plates. Electrons existing between the pole plates are wrapped around a DC magnetic field and move in a circular motion (cyclotron motion).
do. Since a high-frequency electric field exists in the direction of the circularly moving electrons, energy is efficiently converted from the electric field to the movement of the electrons. Since the energized electrons are trapped by cyclotron movement between the electrode plates,
It efficiently collides with neutral molecules and atoms and ionizes them. Electrode 107. Due to high frequency input, the susceptor electrode 104 normally has a self-bias of a negative voltage. Therefore, negatively charged electrons do not enter the picture electrode. Therefore, electrons are confined between the two electrodes in the vertical direction. However, since the parallel ends of the picture electrode are a space, electrons flow out from the ends.In order to suppress the escape of electrons in the lateral direction, the method shown in Fig. 6(b) is required. The strength of the DC magnetic field B may be distributed as follows. That is, the strength of the DC magnetic field B is made constant with respect to the distance r from the center of the pole plate to the vicinity of the end of the pole plate, and the magnetic field strength is increased near the end. As a result, electrons are reflected in areas where the magnetic field strength is strong and are confined in areas with a constant magnetic field strength.

FIG. 7 shows a fourth embodiment to which the concept shown in FIG. 6 is applied. Components that are the same as those of the first embodiment shown in FIG. 1 are given the same numbers and redundant explanations will be omitted.

High frequency power at frequency f for exciting a discharge between the electrodes is supplied through coaxial connector 710. 716
is an inner conductor that guides high frequency power to the electrode 102, and 71
2 is the outer conductor of the coaxial cable formed in a tapered shape, and is a vacuum conductor 105 made of metal such as A1 alloy, stainless steel, Ti, etc.
It is connected to the. In the first embodiment shown in FIG. 1, the DC magnetic field was formed by the permanent magnet 106, but the
In the second embodiment shown in the figure, it is formed by an electromagnet. 71
Reference numeral 5 indicates a magnetic material having a high magnetic permeability μ and high saturation magnetic flux density and constitutes the electromagnet, and 714 indicates an electric wire for supplying direct current. Since the electromagnet is completely surrounded by the inner conductor 716 and the electrode base material 102, it is not exposed to the electric field or magnetic field of the high frequency f1.

Frequency f that controls the self-bias of the susceptor electrode 104
The high frequency power of No. 2 is supplied via the coaxial connector 711. 717 is an inner conductor of the coaxial cable, and 713 is an outer conductor. In addition, inductance L and capacitor C2
The series circuit of inductance L2 and capacitor C2 is a circuit for short-circuiting high frequencies of frequencies f+ and f2, respectively.

708 and 709 are insulating substrates constituting these short circuits, and are made of, for example, Teflon-impregnated insulators. The circuit that short-circuits the inner conductors 716, 717 and the outer conductors 712, 713 is formed in a conical shape to match the cylindrical coaxial configuration. FIGS. 8(a) and 8(b) show examples of the short circuit. The short circuit 1.2 has a hole 805, 806 in the center thereof for inserting the inner conductor 716, 717, and the substrate is formed into a disk shape from a Teflon-impregnated insulator. In the example shown in FIG. 8, four series resonant circuits are arranged radially at angular intervals of 90 degrees. 801, 803 are inductances, 802,
804 is a high frequency capacitor such as a multilayer ceramic. The shaded area is the Cu thin film left on the insulator substrate. The thin film usually has a thickness of about 35 to 70 μm. The thickness of the insulating substrate is 1 to 3 mm, depending on the high frequency power.
That's about it. In FIG. 8(a), the inductance 801
An inductance with a straight line is used, and the capacitor is a chip capacitor. In FIG. 8(b), the inductance 803 uses a coil formed by winding an electric wire the required number of turns, and the capacitor 804 uses a flat plate capacitor.

The explanation returns to FIG. 7. Since high-frequency power, especially power at frequency f1 that forms a discharge between the electrodes, is efficiently confined between the electrodes 102. The susceptor electrodes 104 are made of insulating ceramics 706 and 707, respectively, floating from the vacuum container 105. The distance from the electrodes to the vacuum container is far compared to the electrode spacing. This is to terminate most of the electric field of the high-frequency power of frequency f incident on the electrode 102 at the susceptor electrode 8i104.9 After the high-frequency current of frequency f terminates at the susceptor electrode 104, the electric field of the high-frequency power of frequency f, which is incident on the electrode 102, is terminated at the inner conductor 717. Short circuit (LL, CI
), flows out to the outer conductor 712 via the vacuum vessel 105.

The electrode spacing is usually about 2 to 10 cm, although it depends on the gas pressure. Since the electrode area is set larger than the substrate 103, the diameter of the sea urchin as the substrate 103 is 6 inches and 8 inches.
inch, 10 inches, the diameter of each electrode is at least 20 cm.

It needs to be larger than 25cm or 30cm.

FIG. 9 shows a fifth embodiment, which is relatively close to the actual structure. In the case of this embodiment, the picture electrode 107.1
Since the 040 spacing is narrow, most of the high frequency electric field is confined between the opposing electrodes.

If the short circuit for the high frequency power of the frequency f supplied to the susceptor electrode 104 is insufficient, as in the sixth embodiment shown in FIG.
A short circuit may be provided directly between the 05 and 05. The key point of the configuration of this embodiment is to provide as strong a magnetic field as possible between two opposing electrodes.

As in the embodiment shown in FIGS. 9 and 10, the coil 722
In the case of an electromagnet composed of the magnetic material 715 and the magnetic material 715, the distribution of magnetic lines of force becomes a distribution that spreads downward as shown in FIG.

As in the seventh embodiment shown in FIG. 12, the picture electrode 104.
If superconductors or superconducting thin films 731 and 732 exhibiting perfect diamagnetic properties are provided on the back surface of 107, magnetic lines of force will not leak outside of these superconductors 731 and 732, and will exist only between the two electrodes. .

When it is necessary to cool the substrate 103, for example, an extremely large magnetic field confinement effect can be produced by coating the back surface of the electrode with an oxide superconductor that exhibits a superconducting phenomenon at a liquid nitrogen temperature to a thickness of about 1 μm or more by sputtering. obtain.

The eighth embodiment shown in FIG. 13 shows such a bn magnetic field confinement effect.

Similarly, in order to confine the magnetic field between both electrodes and generate a strong parallel magnetic field, the electromagnets (721, 721,
723) may be provided.

FIG. 14 shows a ninth embodiment constructed based on this principle. Electromagnet (715°722), (7
21, 723) are both substantially surrounded by inner conductors 716, 717 for high frequency power supply. coil 7
The wire for supplying current to 22,723 is the inner conductor 7
16,717 and is drawn out to the outside.

In the youngest embodiment 9, it would be even more preferable if the electrodes 102 and 104 were coated with a completely diamagnetic superconductor. In this embodiment, in the vacuum vessel in which discharge plasma is formed, the vessel main bodies 706 and 707 are made of ceramic, and the outer vessel 105' is made of metal. The outer container 105° serves to ground and conduct high frequency current. With this configuration, the discharge between the electrode 102 and the vacuum vessel, which was observed in the apparatus of each embodiment shown in Fig. 1.9.10, is eliminated, and the high frequency power is almost confined between the electric current, the il O2, and the 104. As a result, high-density plasma can be formed between the electrodes with a small amount of high-frequency power.

It goes without saying that the electromagnets surrounded by the inner conductors 716 and 717 may be formed by permanent magnets. The material constituting the permanent magnet usually has a low relative magnetic permeability of 4 to 5 or less.

Therefore, in the ninth embodiment shown in FIG.
22, 723 are removed to form donut-shaped perfect diamagnetic superconductors 751, 752 as in the 10th embodiment shown in FIG.
It is a good idea to insert it. In this case, superconductors 751, 752
The magnetic bodies 715 and 721 fitted with the magnetic bodies 715 and 721 are made of permanent magnets.

The above has described an RIE apparatus that completely suppresses sputter contamination of the chamber material and causes no damage to the substrate. I
C11a, CHt C112, CCfl 4, etc.),
Fluorine-based (F2, CH2F2, CF4, StF,
etc.) and mixed gas systems (CFzCu2 etc.) are used,
Carrier gases Ar, He, and m addition gases H2 and 02 are added.

From the standpoint of uniform etching, the wavelength of the output frequency f1 of the high frequency power source used for high frequency discharge is required to be at least twice the diameter of the wafer. Preferably, the frequency f1 is about 100 MHz (wavelength 3 m) to IGHz (wavelength 30 cm).

However, for example, when microwaves of 2.45 GHz are used, the wavelength of the electromagnetic waves becomes smaller than the diameter of the wafer serving as the substrate, which is not preferable because it may cause variations in the amount of etching.

Although the embodiments of the present invention have been described above mainly regarding etching of 5in2 and Si films, it is of course not necessary to limit the invention to this. For example, PSG film, BPSG film, ASG film, silicon nitride film, ρ203 film, AρN film, Au, W, M
It may also be used for etching films and substrates made of O, Ta, Ti, or alloys thereof.

Further, the excited active species source gas may be appropriately selected depending on the type of thin film to be etched. For example, in the case of poly-3i thin film, Cf22, CCf:14. CCf22F2.
CIL2 etc., in the case of Si thin film, Cf22. CCl2F
3. CF4 etc., in the case of 5in2 thin film, CF4 /H2
, C2F6, etc., in the case of An thin film, CCf1a, S
i Cf14. BCIIs, Cl12, etc., in the case of MO thin film, W thin film, Ti thin film, Ta thin film, etc., F2, Cf
12, CF4, etc. may be used as appropriate.

It is also effective to add v N2 , o, , N2 as an additive gas.

Further, the substrate 103 on which these are formed is not limited to an insulating material, but may be a conductive material or a semiconductor material.

Furthermore, it goes without saying that the present invention can also be applied to etching polymeric materials such as polyimide films and resists. Furthermore, it goes without saying that the substrate to be etched is not limited to semiconductor wafers. It can also be used for sputter etching other than reactive ion etching.

Next, the apparatus having the configuration of each of the above embodiments is equipped with the above RIE
In addition, it can be easily used for PCVD, dry cleaning, resist ashing, resist dry development, etc. by changing some usage conditions.

First, plasma CVD (PCVD) is 5ili
For the film, raw material gases such as SiH4, Si2 H6, 5iH2Cf2, etc. or gases such as Ar, He, 82, etc. are added to this, and for A1 film formation, H2+All (CH3)
3, H2+AJ2HCCHs) 2 etc. gas,
For 5in2 film formation, S i H4 + 02°S i H2
CA2 +02, etc., for Si3N4 film formation, S i H
4 Supply gases such as +NH and +H2. 1.7, 9,
In the embodiments shown in FIGS. 10, 12, 14, and 15, such raw material gas is discharged by a high frequency power source f and brought into a plasma state. A high-density plasma is formed between the electrodes, but at high frequency f.

is maintained at a high frequency of 150 to 250 MHz, the self-bias appearing on the electrode 102 is as low as 10 to -2.sup., and the electrode is not sputtered. Further, the ion energy for irradiating the substrate surface necessary for film formation is controlled by high frequency power at a frequency tz (for example, 10 to 80 MHz) lower than fl. The irradiation ion energy is controlled by the power of f2 to the optimum value required for film formation, and the irradiation ion density is
It is controlled by the power of fl. For example, in the case of Si film formation, if the gas to be supplied is (Ar+S i H4), then A
It is important to adjust the mixing ratio of r and 5fH4. In particular, activation of the Sf surface by ion irradiation is a decisive factor in forming a high-quality Si film at a low temperature of about 400° C. from room temperature. For example, this is because one or more ions are usually required to be irradiated with optimal energy while one St atom is placed in a regular lattice position. For example, one Si atom is irradiated with 10 or 50 ions.

Usually, the amount of Ar is set to be sufficiently larger than that of SiH4. The same applies to other film formations. The ions that irradiate the substrate surface do not need to be atoms or molecules that directly contribute to film formation. It is more suitable for high-quality film formation if the atoms and molecules that contribute to film formation are completely different from the ions irradiated on the substrate surface, since the amount of ion irradiation of the substrate and the film formation rate can be controlled independently.

Oh, resist! ! ! As mentioned above, J separation is usually performed by a wet process using a mixed solution (H2SO4+H2O2), but the resist that has undergone the ion implantation process is not dissolved in the mixed solution (H2So4+H2O2). Therefore, it is removed by a strong oxidation reaction in oxygen (O2) plasma.

However, conventional equipment suffers from damage caused by high-energy ion irradiation and metal contamination of the substrate surface due to sputtering on the inner surface of the chamber, making it impossible to remove the resist.

However, the device of the present invention (No. 1.7, 9, 10° 12.1)
4. 15), it is possible to completely control the oxygen plasma to remove the resist without any damage or metal contamination. If a small amount of Cl12 is added to O2, the metal components contained in the resist will also be removed at the same time. When removing the resist, the Si surface is thinly oxidized by 02 plasma, but the thin oxide film is thinly oxidized by N2. 0.6 in Ar
It can be easily removed by vapor phase etching with HF gas mixed in at a concentration of about 1.5%. The Si surface from which the oxide film has been removed is terminated with fluorine.
It is easily removed by (Ar+H2) plasma accelerated to about eV.

Next is dry cleaning, but organic contamination is 1 to 15 eV.
It is cleaned by o2 ions and o3 which are accelerated to a certain degree.

A thin oxide film (SiO
z), as mentioned above, is N2. 0.5-0.6 in Ar
% of HF gas. Metal component is 1 to 15
It can be removed by C11.2 ions accelerated to eV.

The device of the present invention is fully applicable.

[Effects of the Invention] According to the present invention, in a plasma processing apparatus configured to generate plasma in a container that can be depressurized and to process a workpiece in the plasma, first and second electrodes each formed in a flat plate shape; a protective member made of a material stable at least against the plasma and provided to cover the first electrode; a holding means for attaching the object to be processed on the second electrode; a first high-frequency power source connected to the first electrode; a second high-frequency power source connected to the second electrode; and the container. and a gas supply means for introducing a desired gas into the RIE, and the frequency of the first high-frequency power source is set higher than the frequency of the second high-frequency power source. Various plasma processes such as plasma chemical vapor deposition, resist asher dry cleaning, etc. can be performed without damaging or contaminating the substrate of the processed object, and without contaminating the processing atmosphere. It is possible to provide high quality semiconductor devices.

In addition, without changing the basic structural components,
By making slight changes to specific setting conditions, such as the output frequency of the high-frequency power source and the type of gas introduced, it can be applied to various plasma process equipment, making it possible to standardize each equipment and improving the quality of semiconductor devices. Enables consistent and uniform manufacturing operations.

Furthermore, since each device has common components,
Manufacturing, management, maintenance, etc. of component parts becomes easier, and it can contribute to improving the performance of the entire device.

[Brief explanation of drawings]

FIG. 1(a) is a schematic configuration diagram of a device showing a first embodiment of the present invention, FIG. 1(b) is a circuit diagram showing an example of the band eliminator shown in FIG. C) is shown in Figure 1(b)
) graph showing the resonance characteristics of the band eliminator, 1st
FIG. 1(cl) is a circuit diagram showing another example of the band eliminator shown in FIG. 1(b), and FIG. 1(e) shows the first. A graph showing changes in the potential of the second electrode, FIG. 2 is a schematic diagram of the configuration of an apparatus for measuring the current-voltage characteristics of the electrode, and FIG. A graph showing an experimental example, FIG. 3(b) is a graph showing a change in self-bias voltage with respect to a change in frequency, and FIG. 4(a) is a schematic configuration diagram showing a second example. (b) is a schematic configuration diagram showing the third embodiment, and FIG. 4(C) is a circuit diagram showing another example of the band eliminator. FIG. 5 is a schematic diagram showing a schematic configuration of a conventional example. Fig. 6 is a distribution diagram of a parallel plate electrode structure and a high frequency electric field and a DC magnetic field, Fig. 7 is a sectional view of a main part showing a fourth embodiment of the present invention, and Fig. 8 shows an example of a short circuit. The circuit diagram, FIG. 9 is a sectional view showing the fifth embodiment of the present invention, and FIG.
11 is a magnetic field distribution (line of magnetic force) diagram, FIG. 12 is a sectional view showing the seventh embodiment of the present invention, and FIG. 13 is the eighth embodiment. 14 is a cross-sectional view showing the ninth embodiment of the present invention, and FIG. 15 is a diagram showing the distribution of magnetic lines of force when a superconducting thin film is provided on the back surface of the electrode. FIG. 7 is a sectional view showing a tenth embodiment. 101... Protective layer (protective member), 102... Base material,
103... Substrate (workpiece), 104... Susceptor electrode (second electrode), 105... Vacuum container, 107...
... Electrode (first electrode), 110... Second high frequency power source 111... First high frequency power source. 1st! Figure (a) Figure (d) Point diagram (b) Figure (C) Figure 3 Figure (b) Frequency (MHzl Parameter High frequency power A "Gas pressure 00W 7X10-3Torr Figure (b) r Figure (C) ) Figure (a) Figure (b Figure 8 Figure 8 (a) Short circuit diagram (b) Short circuit diagram Figure 11 Figure 2

Claims (1)

[Claims] (1) In a plasma processing apparatus configured to generate plasma in a container that can be depressurized and to process an object to be processed in the plasma, a plurality of flat plate-shaped devices are provided facing each other in the container. the first and second electrodes formed, a protective member made of a material that is stable at least against the plasma and provided to cover the first electrode, and a material to be processed on the second electrode. a first high-frequency power source connected to the first electrode, a second high-frequency power source connected to the second electrode, and a desired gas introduced into the container. 1. An apparatus for plasma processing, characterized in that the frequency of the first high-frequency power source is higher than the frequency of the second high-frequency power source.