KR20060029621A - Plasma generator and method with adjustable duty cycle and high frequency drive circuit - Google Patents

Abstract

Reactive circuits are disclosed as part of a method and system for generating high density plasma that does not require the use of a dynamic matching network to directly drive a plasma exhibiting dynamic impedance. The reactive circuit exhibits a small overall reactance when the plasma reactance is the first plasma reactance and exhibits a reactance that does not exceed the reactance specified at the second plasma reactance. The first plasma reactance and the second plasma reactance span substantially a portion of the expected dynamic plasma reactance range. The first plasma reactance and the second plasma reactance may correspond to, for example, a high expected plasma reactance limit and a low expected plasma reactance limit, or the first plasma reactance may correspond to an average expected plasma reactance.

Description

Plasma Production Device and Method and RF Driver Circuit with Adjustable Duty Cycle

Reference to related application

This application claims priority under U.S. Provisional Application No. 60 / 480,338 (filed Jun. 19, 2003) under U.S. Patent Act No. 119, and this application also claims U.S. Patent Application No. 10 / 419,052 (2003). Continuation-in-part of the application dated April 17, which also enjoys priority under US Provisional Application No. 60 / 328,249 (filed October 9, 2001) under section 119 of the US Patent Act. Partially filed in US Patent Application No. 10 / 268,053, filed Oct. 9, 2002. All these applications are hereby incorporated by reference in their entirety.

The present invention relates generally to the design and implementation of a plasma generation system. More particularly, the present invention relates to high frequency amplifiers, antennas, and effective circuit connections that interface amplifiers and antennas for the generation of plasma.

Plasma is generally considered to be one of four states of matter, followed by solid, liquid and gaseous states. In the plasma state, the basic components of the material exist in substantially ionized form. This form is particularly useful for many applications due to its increased reactivity, energy, and suitability for the formation of directional beams.

Plasma generators are widely used in the manufacture of electronic components, integrated circuits, and medical equipment, and in the operation of various goods and machines. For example, a plasma may be induced, for example, by (i) chemical reaction or sputtering from a source, followed by the deposition of a desired layer of material, (ii) by etching the material with high precision, (i) by plasma or by plasma. It is widely used to sterilize objects by free radicals and (iii) to change the surface properties of the material.

Plasma generators based on high frequency (“RF”) power supplies are often used in experimental and industrial environments because they provide ready-to-use plasma sources and are portable and repositionable. Such plasma generators couple RF radiation to the gas, typically under a reduced pressure (and density) that ionizes the gas. In any RF plasma generation system, the plasma appears to vary in load at the antenna terminal typically driven by the RF power supply as the process conditions change. These varying process conditions include changes in operating gas and pressure that affect the amount of load seen at the antenna terminal. In addition, the amplitude of the RF drive waveform itself also affects plasma temperature and density, which in turn also affects antenna load. Thus, antenna / plasma coupling is seen as a non-linear and nonlinear load on the RF power supply.

A typical RF power supply has an output impedance of about 50 ohms, which results in the most efficient coupling of the RF power supply to a load that matches the 50 ohms impedance. Due to inadvertent changes in plasma magnetic inductance, effective resistance, and mutual inductance with the antenna, impedance matching is sometimes caused by the readjustment of the plasma element and sometimes by the plasma to obtain satisfactory energy transfer from the RF power source to the plasma generated. Is done. To achieve this, an adjustable impedance matching network or "matching box" is typically used to compensate for changes in load impedance due to changes in plasma state.

A typical matching box has two independently adjustable components, one for adjusting the series impedance and the other for adjusting the parallel impedance. These adjustable elements must be coordinated with each other to achieve optimal power delivery to the plasma. It is no surprise that correct adjustment of these elements is often a difficult task. Typically, readjustment typically requires complex feedback circuits due to manual / mechanical tasks / devices that adjust one or more element values as a change in plasma impedance and limited automation possibilities.

It is well known that the application of a sufficiently large electric field to a gas ionizes the gas by separating electrons from the positively charged nucleus in the gas atoms and forms a substance such as a conductive fluid known as a plasma. Coupling a high frequency electric and magnetic field to a gas via an antenna creates an induced current in this ionized gas. This in turn increases the electrical conductivity by further ionizing the gas, which in turn increases the efficiency with which the field of the antenna is coupled to the charged particles in the gas. This also leads to an increase in induced currents, which in turn results in substantial ionization of the gas and gradual electrical breakdown by various mechanisms. The efficiency of the RF coupling depends on the specific RF field and / or wave used. Specific types of RF fields and waves that are efficient for generating large volumes of plasma are described below.

Whistler wave is a right-hand-circularly-polarized electromagnetic wave (sometimes referred to as R-wave) that can propagate in an infinite plasma contained in a static magnetic field B _o . If these waves are generated in a finite plasma, such as a cylinder, then the presence of boundary conditions (ie the fact that the system is not infinite) is dependent on the left circular polarization mode (L-wave), with an electrostatic contribution to the field of the entire wave. To exist at the same time. This "bounded whistler" is known as a helicon wave. Boswell, RW, Plasma Phys. 26, 1147 (1981). Their interesting and useful attributes include the following: (1) generates and sustains relatively high density plasma with greater efficiency than other RF plasma generation technologies, and (2) the number of particles per cubic centimeter, even with relatively small devices with only a few kW of RF input power. Plasma density Np of up to 10 ¹⁴ , (3) stable and static plasma in most cases, (4) high degree of plasma uniformity, and (5) a wide pressure range from 1 mTorr to tens of mTorr of moisture. Generation of plasma across. Significant plasma enhancement associated with helicon mode excitation is observed in the relatively low B _o field, which is easily and economically achieved using inexpensive components.

Significant increase of the plasma density (Np) and uniformity can be achieved by the low-field m = +1 helicon wave excitation of R- within the B _o <150 G relatively small chamber. For example, this can be achieved by using the following antenna. In other words, the antenna is similar and thus coupled to one or more helicon modes and field patterns occupying the same volume as the antenna field. Appropriate settings of the combined conditions include the applied magnetic field _Bo , RF frequency (F _RF ), density (N _p ) itself, and physical size.

Some antenna designs for coupling RF power to a plasma are disclosed in US Pat. Nos. 4,792,732, 6,264,812, and 6,304,036. However, these designs are relatively complex and often require custom components that increase system acquisition and maintenance costs. In addition, not all of these designs are suitable for efficient generation of the helicon mode disclosed herein as the preferred mode.

RF power sources typically accept an external RF signal as input or include RF signal generation circuitry. In many process applications, this RF signal has a frequency of about 13.56 MHz. This RF signal is amplified at the output of the power and coupled to the gas / plasma via an antenna in the plasma generator for generation of the plasma.

Amplifiers, including RF amplifiers suitable for RF power supplies, are typically classified into many classes depending on their intended application and performance characteristics such as efficiency, linearity, amplification degree, impedance, and the like. Heat dissipation must be provided to dissipate heat, which in turn increases the size of devices that use inefficient amplifiers, so the amount of heat dissipated by heat is an important consideration in power amplification. The output impedance exhibited by the amplifier is a matter of classification as it sets an inherent limit of power dissipated by the amplifier.

A typical RF amplifier is designed to exhibit a standard output impedance of 50 ohms. Since the voltage across the output terminal of the amplifier or the current through it is non-zero, their product provides a calculation of the power wasted by the amplifier.

Similar to the power dissipation in the switch, this can be reduced by introducing a phase difference between the voltage and the current appearing at the output terminal of the amplifier. Unlike conventional amplifiers, the switch provides two states: short circuit, ie on or low circuit, which corresponds to a low impedance, or off, which corresponds to infinite (or at least very large) impedance. In a switched mode amplifier the amplifier element acts as a switch under the control of the signal to be amplified. By making these signals appropriately, one can introduce a phase difference between current and voltage, for example, via a matching load network to shift the phase to minimize power dissipation in the switch element. In other words, if the current is high, the voltage is made low or zero. And this is also the case. US Patent Nos. 3,919,656 and 5,187,580 disclose various voltage / current relationships for reducing or minimizing power dissipation in switch mode amplifiers.

U. S. Patent No. 5,747, 935 discloses a switching mode RF amplifier and matching load network in which harmonics of the fundamental wave are short-circuited so that the RF power is more stabilized in consideration of the variation in plasma impedance while the impedance shown at the desired frequency is high. These matching networks add complexity of operation to the switch mode power supply rather than excluding the dynamic matching network. Such load matching networks are also not very sensitive to frequency because they rely on strong selectivity in narrow frequency bands that are close to the base.

U. S. Patent No. 6,432, 260 discloses the effective neutralization of the reactive element of plasma impedance using a switched element in an impedance matching network so that the dynamic complex impedance of the plasma appears close to the resistance value. This causes the power supply to respond only to changes in resistance in the plasma because only these changes are seen by the power supply. Dynamic plasma resistance controls the power delivered to the plasma.

When the plasma impedance is a small part of the impedance seen by the RF power source, the change in plasma impedance is relatively less severe. Thus, if the variation range of the plasma impedance is a small part of the overall impedance seen by the power supply, the plasma can be driven by an RF power supply without an intervening dynamic matching network. Overpowering the plasma inductance with enough high-power drivers can guarantee some efficiency. As a result, a matching network is required when the dynamic plasma impedance is a significant part of the overall impedance seen by the RF power supply.

U.S. Patent Nos. 6,150,628, 6,338,226, 6,486,431 and 6,552,296 disclose constant current switching mode RF power supplies comprising inductive elements in series with a plasma load. This plasma is mainly driven as the secondary side of an iron or ferrite core transformer whose primary side is driven by an RF power supply. In this configuration, it is disclosed that no dynamic impedance matching network is required. The current through the plasma is maintained at approximately the value of the initial inductor current to regulate the power based on the size of the load.

Also disclosed in these patents are various plasma ignition methods, including high voltage pulses, ultraviolet light and capacitive coupling, which also contribute to limiting the change in plasma impedance by avoiding large impedance changes encountered during plasma ignition. .

Further designs are known using plasma as the secondary side of the transformer, where the primary side and the secondary side are relatively weakly coupled through a shared core. Taylor invented a plasma generation technique for cleaning the interior of a toroidal vacuum chamber using a process plasma, and in 1973 the device was built. The circuit as the primary side of the transformer used a matching network consisting of tokamak air-core Ohmic Heating (OH) winding and fixed C1 and C2. Several similar designs are known for using iron core transformers to drive other tokamaks. These designs typically operate in the frequency range of 1-50 kHz.

(여기서, M_1p는 1차측 인덕턴스 L₁과 플라즈마 인덕턴스 L_p간의 상호 인덕턴스)가 매우 작기 때문이다. 결과적으로, 변압기 1차측 단자에서 보여지는 유도 부하의 변화가 더 작다. 반면에, 플라즈마가 실질적으로 직접 구동될 때, 예컨대, 전류띠(current-strap)를 통해

(여기서, M_{ant -plasma}는 안테나 인덕턴스 L_ant와 플라즈마 인덕턴스 L_plasma 간의 상호 인덕턴스)는 작지 않고, 결과로서 플라즈마 임피던스에서의 변화는 RF 전원에 의해 보여지는 부하 임피던스에서 상대적으로 더 큰 변화를 나타낸다. 이 변화는 전형적으로 전력 전달을 위해 RF 전원의 50 오옴 임피던스에 합리적인 매칭을 제공하기 위해 가변 매칭 네트워트의 사용을 요구한다.In designs similar to Taylor's design, changes in plasma impedance do not have a significant effect on the load of the driver, which is a parameter.

(Where M _1p is the mutual inductance between the primary inductance L ₁ and the plasma inductance L _p ) is very small. As a result, the change in inductive load seen at the transformer primary terminal is smaller. On the other hand, when the plasma is driven substantially directly, for example through a current-strap

(Where M _{ant -plasma} is the mutual inductance between the antenna inductance L _ant and the plasma inductance L _plasma ) is not small and as a result the change in plasma impedance represents a relatively larger change in the load impedance seen by the RF power source. This change typically requires the use of a variable matching network to provide a reasonable match to the 50 ohm impedance of the RF power supply for power delivery.

Changes in plasma impedance at the primary winding of the antenna terminal or coupling transformer when the plasma is driven directly, i.e. without using a core for substantially coupling the plasma secondary to the primary winding connected to the RF power source. Is considerable. This configuration has been coupled to a plasma or plasma / antenna combination through a dynamic matching network to continuously adjust in response to plasma impedance changes.

The challenges faced by the design of an efficient plasma generator include the need for an antenna that is easy to maintain and configure, the elimination of expensive and limited matching networks to directly couple RF power to the nonlinear dynamic impedance exhibited by the plasma, and efficient Includes the need for a frequency-sensitive RF power supply.

An improved design for efficiently coupling one or more RF power supplies to a plasma is disclosed. Also disclosed is a method and system for generating a plasma with the aid of an RF power source, without using a dynamic matching network for coupling the RF power source to the plasma. The dynamic matching network here requires adjusting the impedance in response to the dynamic impedance represented by the plasma.

Instead of a dynamic impedance matching network, the reactive network couples the RF power source to the antenna-plasma combination. This reactive network is selected such that at least a first plasma impedance value, substantially a resistive load, appears in the RF power supply. Furthermore, at the second plasma impedance value, it is desirable that the reactance seen by the RF power supply be selected to be approximately equal to that of the RF power supply itself so as to cover a substantial portion of the expected dynamic plasma reactance range. Accordingly, a method of designing a reactive circuit capable of eliminating a dynamic matching circuit between a plasma and an RF power source is disclosed. In addition, a reactive circuit suitable for a plasma generator operating at about 13.56 MHz is disclosed. This disclosed method is also applicable to designing reactive circuits that operate at many frequencies other than 13.56 MHz.

In addition to plasma impedance, other considerations may also be considered in the design of the reactive circuit. For example, a phase difference can be designed in the switched power supply, the RF power source, which can improve the efficiency of the power supply by reducing the resistance loss in the switches. These additional conditions may typically require three or more reactance element values determined to provide the desired operation.

An exemplary plasma generation system includes at least one plasma source, at least one high frequency power source, a static magnetic field, and a reactive network. The at least one plasma source has an antenna comprising a plurality of loops, each loop having a loop axis, and the plurality of loops are arranged such that each loop axis is substantially orthogonal to the common axis with respect to the common axis. The at least one high frequency power source is for driving a plurality of loops in quadrature and is coupled via the antenna to a plasma load driven in a circular polarization mode, preferably in a helicon mode. The static magnetic field is formed along the common axis, and the reactive network couples a switching amplifier to the antenna loop.

The high frequency power supply consists essentially of a Class A amplifier, substantially a Class AB amplifier, substantially a Class B amplifier, substantially a Class C amplifier, substantially a Class D amplifier, substantially a Class E amplifier, and substantially a Class F amplifier It is preferable to include at least one crowd. In an embodiment, they are connected to the primary side of the transformer to reduce the drive impedance to low values. More preferably, the high frequency power supply includes a push-pull type D amplifier having a relatively low output impedance.

Preferably, the high frequency power source exhibits a low output impedance. Often this low output impedance is significantly less than the standard 50 ohms. This output impedance is preferably within a range selected from the set consisting of less than about 0.5 Ohms, less than about 2 Ohms, less than about 3 Ohms, less than about 5 Ohms, less than about 8 Ohms, less than about 10 Ohms, and less than about 20 Ohms. . This output impedance is preferably less than 5 ohms, more preferably 0.5 to 2 ohms, and most preferably less than 1 ohm. In a preferred embodiment, the output impedance is about 12 ohms. The use of such a low impedance driver in conjunction with the disclosed circuitry to connect the driver to the current strap of the antenna eliminates the need for a matching box, thus reducing the complexity of the circuit and reducing the failure and high cost of the plasma processing system. Eliminate the cause.

Another advantage of the disclosed system is that the voltage applied to the antenna can be significantly greater prior to plasma formation, thereby enhancing the ability to generate plasma at various operating conditions. Once the plasma is formed, the voltage is reduced to a lower level that sustains the plasma.

Antenna according to the phase between the element and the Bo value, the system operation, or the magnetic induction coupling plasma as helicon source; may operate in (magnetized inductively coupled plasma MICP), or B _o = operating at 0 to ICP. Furthermore, it has been observed that when using a conventional plasma source, it works efficiently and robustly in pressure situations that are very difficult to access or could not be used properly (eg, Po is about 100 mTorr). The current in the antenna element is rapidly "locked" in quadrature excitation mode when the conditions for neutral pressure P ₀ , input power P _RF , and externally applied axial magnetic field B _o are correct. appear. This advantageously allows the plasma to fill the chamber approximately uniformly, thus providing the ability to create uniform process conditions.

In addition, this combination of the antenna system and the RF generator provides a state in which plasma parameters change over a much larger range than reported from other sources without any element of a dynamic matching network (eg, medium voltage for approximately one minute of cycles). Po can be lowered from 100 mTorr to 5 mTorr and back to 100+ mTorr), creating and maintaining a plasma.

Another advantage of this disclosed system is that it can be used "instant-on" as a plasma source without a matching network. This feature can be used to provide additional control for the process used. In particular, the amplitude of the RF power generating the plasma can be modulated between two (or more) levels, such as 30% and 100% or a full on-off scheme (0% to 100%). This modulation can happen quickly, for example at frequencies of several kilohertz, and can achieve several purposes. For example, the average RF power can be reduced as a result of the decrease in the average plasma density. "Immediate" operation can produce 50 liters of volume plasma with an average RF input power of 5W.

In addition, modulation can be used to control the spatial distribution of the working gas in the reaction chamber. The plasma alters the distribution of the working gas and thus contributes to the nonuniformity of the flow of active chemicals or radicals. By modulating the duty cycle of plasma generation, the characteristics of the neutral gas flow at plasma off (or when the power level is reduced) can be adjusted and control the uniformity of the process. Since the plasma startup time is typically RF application within 10-20 microseconds, the duty cycle can be controlled at frequencies as high as tens or hundreds of kHz.

These and other features of the invention are described below in conjunction with the following illustrative figures.

The following illustrative drawings are provided to better illustrate various embodiments of the invention and are not intended to limit the scope of the claims.

1 shows a plasma source chamber having two sets of antenna elements.

2 illustrates an adjustable circuit with an RF power source coupled to the antenna.

3 shows a second adjustable circuit in which an RF power source is coupled to the antenna.

4 shows a third adjustable circuit in which an RF power source is coupled to the antenna.

FIG. 5 shows a circuit in which an RF power amplifier is coupled to an antenna current strap.

6 shows a second circuit in which an RF power amplifier is coupled to an antenna current band.

7 shows a third circuit in which an RF power amplifier is coupled to an antenna current band.

8 illustrates a simplified model of an RF power amplifier, antenna current band, and plasma.

9 is an overall equivalent circuit diagram of the model shown in FIG. 8.

10 is a diagram illustrating the frequency response of a plasma source in the absence of plasma.

11 is a diagram illustrating a frequency response of a plasma source in a state where plasma is present.

12 is a diagram illustrating a feedback configuration for controlling a plasma source.

13 illustrates a reactive network for coupling an RF power source to a plasma.

14 is a diagram of devices selected as a reactive network that eliminates the need for a dynamic matching network in accordance with an exemplary embodiment of the present invention.

First, referring to the drawings, FIG. 1 shows a plasma source chamber having two sets of antenna elements. This antenna design includes two orthogonal one or multiple winding loop elements 105, 110, 115, and 120 arranged about a common axis. Antenna elements 105, 110, 115, and 120 are driven by RF power source A 125 or B 130, respectively, as shown. Each antenna loop may be coupled to different RF power supplies for driving antenna elements in quadrature or via a phase splitter to the same RF power supply. Preferably the loops in the antenna consist of 8 gauge Teflon coated wire, although uncoated copper wire or other conductor may be used.

1 shows two sets of orthogonal two-element Helmholtz-coil-like loop antennas, where loop elements 105 and 115 correspond to one set and loop element 110. And 120) corresponds to the second set. The loop elements are conformally wrapped around the insulating cylinder 135 such that the magnetic field created when current flows through them substantially crosses the axis of the cylinder. Each set of opposing elements is connected in series according to the Helmholtz method. The wires interconnecting the opposing loop elements are preferably arranged such that adjacent segments flow current in the opposite direction in order to promote the elimination of stray fields associated with them, but this is not essential to the operation of the apparatus. The antennas are energized so that the current in both orthogonal branches is nearly equal and there is a phase difference of 90 degrees to create a magnetic field close to the rotating transverse magnetic field.

In the example of helicon mode plasma, the static axis B _o -field is made, for example, by a simple electromagnet. This field flows along the axis of the cylinder. The direction of this static field causes the rotating transverse field to resemble that of the m = +1 helicon wave. In practice, the amplitude and direction of the currents that make up the external field can be adjusted to alter the performance of the plasma generator. The overall amplitude of the required field is typically in the range of 10-100 gauss for the parameters discussed herein, although other ranges may be employed for other size sources. Once the optimal amplitude and direction of the static field is selected, they usually do not require further adjustment.

In combination with this, the static field and the RF field of the antenna elements create an m = +1 helicon mode inside the isolation cylinder and sustain this plasma discharge. In addition, the helicon mode may not be directly excited by adjusting to shift the static magnetic field or by not applying a field at all. This operation, of course, creates a plasma but is not as efficient as normal helicon mode. Of course, a static field may be applied later to improve the operation of the plasma source / generator.

It is also possible, for example, to achieve the same state as in Figure 1 as a whole using loop antennas and / or squat bell jars wound multiple times instead of a single loop. Although not necessarily required, the bell jar is preferably adjusted to have a gap of 1/2 inch or less with the antenna frame.

An example of a plasma source setup is as follows. That is, the quartz bell jar has a straight cylindrical cross section of 15 cm height with an approximately 12 inch inner diameter (as standard K.J. Lesker 12 x 12) and a hemispherical top of 6 inches radius. The jaws are placed on a vacuum chamber approximately 12 inches inside and 8 inches high (excluding the plasma source portion). The antenna consists of two sets of opposing, full, approximately rectangular, two-time continuous loop antenna elements, which wrap the bell jar at intervals of approximately 1/8 inch to 1/2 inch at all points. In each element the turns are connected in series, and the two elements are also connected in series in each set so that the fields are added. In this example, each set has a self-inductance of approximately 10 micro henrys, and the mutual inductance between the two sets is less than 1 micro henry. The vertical and horizontal cross sections of the antenna loop are approximately 25 cm and 20 cm long, respectively, and consist of 8-gauge Teflon coated wire. In alternative embodiments, a single wound copper conductor may be employed in place of a one or two wound teflon coated wire. The specific embodiments described herein for creating a transverse rotation field are not intended to limit the scope of the claimed invention unless expressly stated.

Conventional RF power supplies and dynamic matching structures (see FIGS. 2-4) can be used to generate antenna current in the antenna described above. Furthermore, the circuits of FIGS. 2-4 are compatible with many of the disclosed methods. Some of these methods include providing a low output impedance to the RF power supply; And adjusting the reactance coupling the RF power source to the antenna such that the resonant frequency in the absence of plasma is a desired RF frequency. Low output impedance can be understood by referring to the circuit's cue value ("Q") with and without plasma. The "Q" value in the absence of a plasma will be five to ten times or more than when a plasma is present. Note that, unlike known circuits, the combination of the RF power source and the antenna does not have to be readjusted by changing the reactance in response to the plasma's impedance change in the presence of the plasma.

The RF power supply 200 in FIG. 2 may be a commercial 2 MHz, 0-1 kW generator, and is connected to quadrature / hybrid circuits via a 50 ohm coaxial cable at port “A” 125 shown in FIG. 1. The " +45 degrees " and " -45 degrees " stages of the quadrature / hybrid circuit are connected with respective L-type capacitive matching networks consisting of adjustable capacitors 205, 210, 215, and 220. At each operating frequency, the reactance of capacitor 225 is about 100 ohms, and the reactance on either side of transformer 230 is about 100 ohms when the other side is open. As shown in FIG. 2, a single RF power supply 200 is a passive power splitter (quadrature / hybrid circuit) and four adjustable to match two separate antenna inductances 235 and 240. Can be used with the adjusting elements 205, 210, 215, and 220.

Another embodiment shown in FIG. 3 employs two separate RF power sources 305 and 310 and is thus connected to inductances 335 and 340 via adjustable capacitors 315, 320, 325, and 330, respectively. It is completely separated by two antenna power circuits. This configuration is advantageous in that each RF power supply can operate at full power, thus doubling the amount of input power compared to a single RF power supply, and the phase and amplitude ratios between the antennas can be adjusted. The power sources 305 and 310 typically operate with approximately the same amplitude and 90 degrees of phase difference, but the amplitude and / or phase difference may be varied to change the nature of the excitation mode. For example, by operating them at different amplitudes, an elliptically polarized plasma helicon mode may be maintained rather than a fully circularly polarized mode.

The third embodiment shown in FIG. 4 has a passive resonant circuit comprising an inductor / antenna inductance 405 and a set of adjustable capacitors 410, with adjustable capacitors 415 and 420 connected to the antenna inductance 425. FIG. The other end is driven by the RF power supply 400 together with the dynamic matching circuit having a. This configuration helps the passive side to excite an elliptic helicon mode of the same nature in which the passive side has approximately 90 degrees of phase difference with the driving side, thus providing many advantages for the configuration of only one RF power supply and dynamic matching network.

In this example shown, the working gas is argon at a pressure ranging from 10 mTorr to 100 mTorr. The static axis field is manually set at 0-150 G and is formed with a radius of approximately 9 inches by a coil located outside the bell jar / antenna assembly.

Plasma operation at approximately 75 mTorr exhibits at least three different modes. First, when _Bo <B _critical when P _RF is less than or near 200 W, a bright mode in which the plasma is concentrated near the periphery of the bell jar is observed. Where B _o is the axial magnetic field and B _critical is the threshold of the axial field to excite the plasma using the helicon mode. Similarly, P _EF and P _threshold refer to the RF power supplied to the antenna and the threshold power described below. In this mode, the RF antenna current is not in quadrature but rather tends to be as large as 180 degrees out of phase. Second, a dull-glow-discharge-like mode with uniform density / glow at higher powers and dark spaces approximately 1-2 cm thick along the walls of the bell jar there is seen, which is when the B _o> B _critical but P _EF <P _threshold. In this case, the RF current shows a solid quadrature, and appears to be fixed at approximately 90 degree phase shift immediately after the plasma is formed. Third, a clear plasma is formed at P _EF > P _threshold and B _o > B _critical , resulting in a more even radial distribution than mode (1), and the antenna current tends to be fixed in quadrature again. The third type represents an efficient mode of operation, which is known to be very difficult to access with known plasma sources at medium pressure. However, each of these types can be applied to a plasma process.

On the other hand, in the interest of a simplified power circuit, the conventional RF power supply and the matching network shown in FIGS. 2-4 can be excluded.

In a preferred embodiment, the RF power circuit directly drives the antenna current strip using the configuration as shown in FIG. The RF amplifier shown in FIG. 5 is preferably one of many types of RF amplifiers having a low output impedance (ie, push-pull output stage) known in the art. Transistors 505 and 510 are driven in a push-pull fashion by an appropriate circuit 500, as is known to those skilled in the art. In this configuration only one or the other transistor at any given time is typically energized with a duty cycle of 50% or less. The outputs of the transistors are combined to produce a complete signal.

Preferably, power semiconductors such as transistors 505 and 510 are operated in a switching mode at the output stage. In Figures 5-7 they are represented by FETs, but may be, for example, bipolar transistors, IGBTs, vacuum tubes, or other suitable amplifiers. Examples of switching modes are provided by class D operation. In this mode, alternating output devices are quickly switched on and off in opposite cycles of the RF waveform. Ideally, the output elements are in a fully on state with no voltage drop or a completely off state with no current flow, resulting in no power dissipation. Class D operation is therefore ideally 100% efficient. However, this calculation assumes on-impedance switching of the spirit with infinitely fast switching time. Actual implementations typically exhibit near 90% efficiency.

The RF driver is then coupled directly to the antenna current band 520 via a fixed or variable reactance 515, preferably via a capacitor. This coupling reactance value is preferably such that the resonance frequency of the coupling reactance and the circuit of the antenna is approximately equal to the RF operating frequency in the absence of plasma.

Another configuration of the output stage of this circuit is shown in FIG. 6A, which includes a transformer 620 with a driver 600 and transistors 605 and 610, followed by or integrated with a push-pull stage. Provide electrical isolation. Transformer 620 may be selectively arranged to convert the output impedance of the push-pull stage to a low impedance if it is too high. The capacitor 615 is configured to resonate with an induction circuit formed of the transformer 620 and the antenna current band 625 at a desired drive frequency. A similar embodiment is shown in FIG. 6B where capacitor 615 is used for direct current removal, and capacitor 630 resonates with a series circuit of the leakage inductance of the transformer and the inductance of current band 625.

7 shows another RF power and antenna current band configuration. An inductor 725 with a tab in the center coupled to the direct current power input is connected to the output stage with push-pull driver 700 and transistors 705 and 710. It is isolated by a transformer 720. Again, at any point in time only one or the other transistor is energized, typically with a duty cycle of 50%. The circuits of FIGS. 5-7 are provided only as illustrative examples. Any well-known push-pull stage or other configuration that provides a low output impedance can be used in place of the alternative.

The RF power supply may also be a symmetric antenna (Nagoya Type III or a variant thereof, such as a Boswell-type paddle-shaped antenna) or an asymmetric antenna (e.g. right-hand helical, It can be used with any helicon antenna, such as the twisted-Nagoya-III antenna specification, or any other non-helicon inductive coupling specification.

The RF power source may be amplitude modulated with various duty cycles to provide reduced or zero plasma density times interspersed between times of higher plasma density. This modulation of plasma density can be used to affect the uniformity of the working gas and consequently the uniformity and hydrodynamics of the process. Thus, a plasma having a more spatially uniform distribution can be generated by selecting an appropriate modulation scheme for the plasma generation system.

In general, a plasma generation system may include a substantially Class A amplifier, substantially Class AB amplifier, substantially Class B amplifier, substantially Class C amplifier, substantially Class D amplifier, substantially Class E amplifier, or substantially A high frequency power supply based on operation as a low class F amplifier or any subcombination thereof can be used. This power source coupled with the antenna for excitation in helicon mode is suitable for the generation of high density plasma. Further, for the non-switching amplifier as shown in Figs. 2 to 4, an intermediate stage for converting the RF power impedance into a low output impedance may be employed to approximate the efficient operation of the switching amplifier based on the embodiments described herein. have.

In an inductively coupled plasma source, the antenna current band is located near the area where the plasma is formed, which is usually outside of the insulated vessel. From the circuit's point of view, the antenna element forms the primary side of the non-ideal transformer, and the plasma is the secondary side. In the equivalent circuit shown in FIG. 8, the inductor 810 is represented as a sum element representation of all inductances of the current strip and wiring, including, for example, the output transformer of the driver shown in some embodiments. The element in box P represents the plasma, the inductor 820 represents the magnetic inductance of the plasma, and the impedance 815 represented by the effective resistance represents the plasma disappearance. M represents the mutual inductance between the antenna and the plasma. Transistor driver 800 is represented by a square wave voltage source. Capacitance 805 is adjusted so that when the system is installed, the resonant frequency of the circuit approximately matches the desired operating frequency. In other embodiments using fixed capacitors, the RF frequency can be adjusted to achieve the same effect.

To describe the operation of this system, the entire system can be represented as shown in FIG. In FIG. 9, all inductors are represented as inductance 905, all capacitors as capacitance 910, and all missing elements as resistors 915, with the amplifier ideally serving as an RF voltage source (ie output impedance 0) It works.

In the absence of plasma, R is small since there is little loss, and the circuit of FIG. 9 shows a narrow resonance response to a change in frequency as shown in FIG. This provides one of the advantages of the operation of this circuit. That is, it is possible to apply a high voltage to the antenna with a relatively very small power input, thus facilitating the initial breakdown of the gas in the reaction chamber. Once the plasma is formed, the damping in the system considerably widens the resonance peak and reduces the Q value of the overall circuit, as shown in FIG. The center frequency of resonance can be shifted depending on the plasma state, but this shift can be negligible compared to the width of the resonance response when a plasma load is present. Thus, when operating with a plasma load, the circuit is relatively insensitive to changes in operating conditions and does not require readjustment. This is illustrated in Figure 11, where the overall system resonance is slightly shifted in frequency and the Q value is sufficiently reduced, but the operation of the system is still efficient. In addition to the reduction of the Q value of the circuit, the voltage applied to the plasma is self-regulated to significantly decrease for the absence of plasma. In some embodiments, it may be somewhat advantageous to adjust the operating frequency of the RF drive from the exact resonance in the absence of the plasma to actually slightly shift to one or the other depending on the shift in the resonance frequency when the plasma is formed.

The level of power input into the plasma can be controlled by various techniques such as adjusting the level of direct current supply at the RF output stage. In one embodiment, the supply voltage may maintain a constant power relative to the plasma source in response to a change in the sensed plasma load. As shown in FIG. 12, the plasma load may be sensed in the following manner for adjustment by the DC supply regulator 1230. That is, for example, by monitoring the voltage from the DC supply 1215 by the voltage sensor 1200, by monitoring the DC current into the RF / plasma system by the current sensor 1205, and their products RF By using the approximation of amplifier efficiency in the previously measured module 1210 to calculate the net power from the amplifier 1220 to the plasma 1225, the plasma load can be sensed. The amplifier efficiency for the gain module 1235 can be measured at different output levels, for example by monitoring the thermal load at various points in the system, and digitally stored to calculate the change in efficiency with the output level. Alternatively, RF power and current can be measured and their in-phase product can be evaluated to calculate the actual power lost in the plasma.

The sensing of the plasma can also be extended to the sensing of spatial uniformity by indirect or direct sensing through variations in voltage or current. The change in the duty cycle according to this variation can control the spatial distribution of the plasma. Furthermore, modulation of the duty cycle may enable control over average input power that enhances the efficiency of plasma generation. The feedback configuration of FIG. 12 also makes it possible to switch between the two or more power levels described above.

"Low" impedance, as used herein, means that the series resonant circuit shown in FIG. 9 has a Q value that is five to ten times higher than the plasma-free state, with a Q value in the absence of plasma. do. That is, the amplifier output impedance is small enough that the energy lost during half of the output is much less than that stored in the reactance element. This condition can be mathematically represented as Z_out << (L / C) ^1/2 , where L and C are the total values shown in FIG. 9. If this condition is maintained, the RF amplifier will approach its operation as a voltage source.

For example, low resistance for the low impedance of an RF power supply generally refers to a resistance of less than 10 ohms, preferably less than 6 ohms, more preferably less than 4 ohms, and most preferably less than 1 ohm. .

However, not all embodiments of the present invention require that elements in a reactive circuit coupling an RF power source to an antenna / plasma be selected based on the resonant frequency of the circuit in the absence of plasma. There may be several other conditions that allow for an efficient coupling of the dynamic impedance of the plasma while also allowing proper specification of a reactive circuit that does not require a dynamic matching circuit.

Although referred to as variable impedance, the plasma reactance can be explained by the fact that a limit is expected between the upper and lower limits. Thus, a high expected plasma reactance component and a low expected plasma reactance can be specified. For example, this specification may reflect a 1σ deviation from the expected mean value. Many other similar specifications are possible, indicating the possibility that the plasma impedance actually deviates from specified limits. Indeed, instead of the high plasma reactance, it is also possible to specify a value that is not symmetrically located at the low expected plasma reactance. Furthermore, even though a particular plasma impedance does not follow a normal distribution, some vowels of the plasma show a normal distribution of combined impedances throughout.

Similarly, some collections of RF power supplies also result in a normal distribution in both frequency and time. Thus, proper selection of the reactance network ensures that the change in the plasma reactance actually matches well with the change in the RF power supply, by matching them at two values of the expected plasma reactance.

By specifying this plasma reactance and the lowest or very low plasma reactance, the recognition or evaluation of the value at which the change in plasma impedance is greatest, a method of specifying components in the reactive circuit can be reached.

, 병렬 임피던스는

이다. 입력측에서 본 임피던스는 Z₁+Z₂이다. 여기서 L1은 2.1μH로 주어지고, C_a가 약 81.6pF으로, C_b가 약 376pF으로 조절되며, 이 복소수는 적절한 성분에 의해 조절되어 14+j12.6 오옴이 된다.For example, the circuit illustrated in FIG. 13 (Source: “3kW and 5kW Half-bridge Class-D RF Generators at 13.56 MHz with 89% Efficiency and Limited Frequency Agility”, Directed Energy Inc. © 2002, document number 9300-0008 Rev. 1, viewed June 10, 2004, at the web address http://www.ixysrf.com/pdf/switch_mode/appnotes/3ap_3_5kw13_56mhz_gen.pdf, incorporated herein by reference, the impedance at the RF jack is 50 ohms , C _a = C ₁ + C ₂ , C _b = C ₃ + C ₄ , the series impedance is

Parallel impedance

to be. The impedance seen from the input side is Z ₁ + Z ₂ . Where L1 is given as 2.1 μH, C _a is adjusted to about 81.6 pF, C _b is adjusted to about 376 pF, and this complex number is controlled by the appropriate component to be 14 + j12.6 ohms.

For example, in a capacitively coupled system using at 13.56 MHz to provide an RF bias to a substrate in a semiconductor processing chamber, an exemplary plasma antenna combination may, for example, have a resistance R _p of about 1 to 15 ohms and about -8 to-. 25 ohms reactance X _p . Thus, it is generally difficult to hook the circuit of FIG. 13 to this antenna / plasma combination. The large imaginary component of this impedance will allow the transistor switching circuit to operate safely at a portion of the desired peak supply voltage of about 700 to 800V, such as about 250V (greater than 200). The peak output voltage is given by V _supply / 2X | H | (where | H | is the magnitude of the transfer function of the system) and will be in the range of 28 to 83V at various plasma conditions in the 250V example.

로 되도록 하는 커패시터를 직렬로 추가함으로써 주어진 동작 주파수에서 제로 또는 그에 가까운 레벨로 조절될 수 있다. 유사하게, 트랜지스터나 모스펫과 같은 상당한 출력 커패시턴스를 가지는 출력 소자를 사 용하는 구동기 회로에서, 출력 커패시턴스(예컨대 어떤 시방서에서 C_oss)에 기인하는 소실은 인덕터 부하를 간단히 제공함으로써 감소될 수 있다. 이는 그 커패시턴스에 축적된 전하 때문으로, 적절하게 조정된 인덕터 부하는 이 전하를 소실함이 없이 커패시턴스를 방전한다. When operating at a given frequency, the overall impedance can be adjusted by adding an inductor (with positive reactance) or capacitor (with negative reactance) in series with the impedance. For example, if there is stray inductance L due to the terminal, etc., the overall impedance is adjusted by adjusting the capacitance.

By adding a capacitor in series that allows it to be adjusted to zero or near the given operating frequency. Similarly, in driver circuits using output devices with significant output capacitance, such as transistors or MOSFETs, the loss due to output capacitance (eg, C _oss in some specifications) can be reduced by simply providing an inductor load. This is due to the charge accumulated in the capacitance, so that a properly adjusted inductor load discharges the capacitance without losing this charge.

14 illustrates a typical exemplary reactive circuit 1400 suitable for coupling a high frequency power supply 1405 to a capacitively driven plasma or antenna-plasma combination. This circuit relates, for example, to a capacitively coupled driver for RF bias of a substrate in a semiconductor processing plasma, but the principle of determining the value of each component also applies to an inductive coupling system. Exemplary generic reactive circuit 1400 may be tuned using a capacitor or inductor or both. For example, the reactances of capacitors 1415 and 1425 may be selected at about 500 pF, respectively, to be approximately equal to the minimum plasma reactive component. Inductors 1410 and 1420 are adjusted to satisfy the following two conditions. I.e. a) the imaginary part of the total load seen at the transistor output stage is at its maximum magnitude, i.e., the upper limit of the expected plasma reactance, and b) the plasma reactance is the minimum magnitude, i. The imaginary part is adjusted to optimize the operation of the high frequency power supply, for example, to +12 ohms in the circuit described in the Direct Energy reference mentioned above.

The impedance seen at the transistor drive stage is Z _load = Z ₁₄₁₀ + Z ₁₄₁₅ + (Z ₁₄₂₀ + Z ₁₄₂₅ ) || Z _p . Here Z ₁₄₁₀ represents the impedance of the inductor ₁₄₁₀ of FIG. 14, and Z _p represents a value of the expected plasma reactance. That is, the driver assumes that the capacitor 1410 and the inductor 1415 are added in series to the parallel combination of the plasma impedance 1440 and the capacitor 1425 + inductor 1420 in series.

In a high frequency power supply which works most preferably when driving a load with an output reactance of +12 ohms, case "a" corresponds to Im (Z _load ) of about 0 ohms at a plasma reactance X _p of about -25 ohms. Case “b” corresponds to Im (Z _load ) of about 12 ohms at a plasma reactance Xp of about −8 ohms. These conditions result in a pair of equations that can be solved by setting R _p to a low value, say about 1 ohm, because this level of plasma reactance causes a large change in the load seen by the RF driver. These two equations can be solved for the values of inductance 1415 and 1420 of the support. Under the conditions described above, in this exemplary embodiment the value of inductance 1420 is about 345 nH and the value of inductance 1415 is about 185 nH, which means that Im (Z _load ) is about each in conditions "a" and "b". Let it be 0 ohms and about 11.9 ohms. More complex calculations should take into account stray inductance, coil inductance, etc., due to non-ideal effects.

If subtracting two comparable values results in a large error, other choices may be taken that take the value of the capacitors 1410 and 1425, for example smaller values that improve the tolerance. Additionally, instead of modifying the values of the capacitors 1410 and 1425 and adjusting the values of the inductors 1415 and 1420, the values of the inductors 1415 and 1420 may be modified and the values of the capacitors 1410 and 1425 may be adjusted. have. It should further be noted that the overall impedance of any series or parallel combination of reactive components is an important amount, and that particular values or specific geometries of L and C may be used in the circuits described above. As an example, the series combination of inductor L and capacitor C may have a reactance of about 5.9 ohms when L = 345nH and C = 500pF or when L = 620.5nH and C = 250pF. These values can be adjusted to meet other constraints of the system, such as the need to have a high (or low) impedance at the second harmonic.

Alternative output transistor stages can be operated at different impedances of the reactive load including some capacitive load. Then, the condition that Im (Z _load ) is about 0 ohms can be specified at some intermediate point rather than the low or high limit of the expected plasma reactance. Thus, the imaginary part of the total load seen at the transistor output stage is small at this specified magnitude of plasma reactance, i.e., at the limit of the specified plasma reactance. Furthermore, the specified plasma reactance may be outside the expected operating range. However, this specification can lead to higher output currents. In addition, adding a resistance pass in parallel to the capacitor 1425 can improve the performance of the reactive circuit. Thus, the reactive circuit may also include a resistance element.

의 범위 내의 인덕턴스가 플라즈마 리액턴스의 변화에 대한 전달 함수 H의 응답 감도를 저감하는 데에 사용될 수 있다. 한편, 음의 상호 인덕턴스는 이 감도를 높일 수 있다.In another aspect, a nonlinear resistive or reactive element can be used for the purpose of reducing the impedance change seen in an RF power supply. In another aspect, inductors 1415 and 1420 may be arranged such that the mutual inductance is small, whether positive or negative. Positive mutual inductance M _{1415 , 1420}

An inductance in the range of can be used to reduce the response sensitivity of the transfer function H to a change in plasma reactance. On the other hand, negative mutual inductance can increase this sensitivity.

These methods for coordinating or setting up a reactive network offer several advantages in addition to eliminating the need for dynamically adjusting matching circuits. For example, adjusting at one plasma reactance within the range of reactance values for plasma matching is matched to the operation of the amplifier, thus providing the transistors with the reactive impedance needed to fully operate at high voltages. Furthermore, the other end of the plasma range, but the reactance seen at the output stage is small and the overall load is also small, enabling operation at low supply voltages of high currents and the reactance present in the transistors becomes less important. In addition, this specification allows a reasonable amount of power from the RF power source to the plasma to be delivered over a wide range of plasma reactances. In another aspect, such a design makes it possible to use a large number of output stages coupled in parallel, for example.

Often the specification for an RF power supply is an output voltage for application to the antenna terminal with an RF input voltage level that is adjusted to produce the desired output voltage as required for various plasma operating conditions. The test for the transfer function H = V _plasma / V _in is carried out with the system "voltage transfer function" or the ratio of the output voltage to the input voltage H = V _plasma / V _in = [(Z ₁₄₁₀ + Z ₁₄₁₅ ) || Z _p ] / Z _load .

As described above for the adjustment, this transfer function, if not entirely, has a resonance characteristic of more than one magnitude of H over the actual operating range. | H | is approximately 75 at X _p of -25 ohms (case A described above, at R _p of about 1 ohm) approximately 1.5 at X _p of -8 ohms (case "b" described above, still about 1 To ohm's R _p ). At higher plasma reactance, eg, R _p of about 4 ohms, | H | varies from approximately 21 to 1.6. Thus, a reactance network well chosen to operate at the lowest expected plasma reactance ensures with high certainty that the change in plasma impedance is smaller at higher values of plasma reactance.

Although some of the foregoing descriptions have been made with the term resonant frequency of a reactive network, it is often desirable to drive a high frequency power source with a slight deviation in the direction in which the frequency spreads as there is a plasma than in the absence of the plasma. It should be noted that This ensures stable and efficient operation at the frequency of interest.

The systems and methods described above have the advantage of enabling dielectric breakdown in gas and generating plasma from the fact that the high Q value of the circuit in the absence of plasma induces a high voltage to the antenna element with relatively low power requirements. to provide. The voltage in the absence of this plasma can be controlled to provide programmed dielectric breakdown of the working gas, and once the plasma is formed, the current induced in the plasma lowers the high voltage to load the system and induce dielectric breakdown, thus You can avoid stress on the system.

The circuit configuration described above only requires a fixed capacitance C and therefore does not require a variable adjustment element such as a mechanically adjustable capacitor. However, in some preferred embodiments, various circuits can be built using, for example, variable capacitors adjusted to match the system resonance to the desired operating frequency. However, this does not require real time impedance matching at the plasma operating point. This matching is useful for countering the effects of mechanical vibration or aging that can cause drift of the L-C resonant frequency.

In one embodiment, the operating frequency is adjusted to compensate for small deviations from resonance, while large deviations are compensated for by mechanical adjustment of the capacitor. In another embodiment, the adjustment is made by adjusting the capacitor. In a preferred (adjusted) embodiment, this adjustment takes place when it is automated and the source is offline. In another aspect, the adjustment is made part of process control, for example to fine tune the process conditions, such that the above described configuration reduces the number of adjustable elements as low as in the embodiment with adjustable adjustment elements.

It will be apparent to those skilled in the art that the present invention described above has many variations and alternative implementations without departing from the spirit and spirit of the invention. Such modifications are intended to fall within the scope of the claims set forth below. For example, a transformer may be provided with a conventional amplifier for impedance matching for low impedance. In addition, although the present invention eliminates the need for a dynamic matching circuit, it is also within the scope of the present invention to use any dynamic matching circuit in combination with the reactive circuits disclosed herein to alleviate other stringent requirements placed on the dynamic matching network. It is intended to be. Therefore, it is to be understood that the claims include such modifications and variations and equivalents thereof. Furthermore, all documents referred to herein are incorporated by reference in their disclosure and suggestion as a whole.

Claims (40)

A method for driving a dynamic plasma impedance by eliminating the need for a dynamic matching circuit, Providing a high frequency power source having a low output impedance; Between the high frequency power supply and the plasma, first and second reactances are selected such that a substantially resistive load appears at the RF power supply at a first plasma reactance and a reactance specified by the RF power supply at a second plasma reactance is shown. Providing a reactive network; And Controlling the average input power. The method of claim 1, wherein the first plasma reactance and the second plasma reactance values span a substantial fraction of the dynamic plasma reactance. 2. The method of claim 1, wherein the first and second plasma reactance values correspond to a high expected plasma reactance limit and a low expected plasma reactance limit, respectively. The method of claim 1, wherein the reactive network is effective when the plasma reactance is low. The method of claim 1, wherein the plasma resistance is an element of a set consisting of a set of about 1 to 5 ohms, a set of less than 1 ohm, and a set of less than 10 ohms. The method of claim 1, Calculating the low expected plasma reactance limit and the high expected plasma reactance limit; And Calculating a low plasma reactance for which the reactance network is required to be valid. The method of claim 1, wherein the specified reactance seen by the RF power source is about 12 ohms. 8. The method of claim 7, wherein modulating the duty cycle of the high frequency power supply is further provided for neutral gas flow. 2. The method of claim 1, further comprising sensing a spatial distribution of the plasma and modulating the duty cycle in response thereto. 10. The method of claim 9, further comprising: sensing a spatial distribution of the plasma and in response to modulating the duty cycle of the supply time of the neutral gas flow to modulate the spatial distribution of the reactive elements constituting the plasma or neutral gas. Method comprising a. The method of claim 1 wherein the average input power is applied at an average density of greater than about 1 watt per 10 liters of volume. The method of claim 1, wherein the average input power allows selection between a plurality of output power levels. 13. The method of claim 12, wherein the at least one output power level is selected from the set consisting of about 5 Watts, about 10 Watts, about 5-10 Watts, about 10-50 Watts, and about 5 Watts to about 25 kW. Way. The method of claim 13, wherein the plasma power is rapidly switched between two or more levels. 15. The method of claim 14, wherein the plasma power is switched from about 30 percent of the total power to about 100 percent. The method of claim 1, wherein at least one frequency for modulation of the duty cycle is at least about 1 Hz, at least about 10 Hz, at least about 100 Hz, at least about 500 Hz, at least about 1000 Hz, at least about 5000 Hz, at least about And from a set consisting of 10,000 Hz, and at least about 100,000 Hz. 2. The method of claim 1, wherein the average input power is controlled by one or more pulse width modulations or changes in the DC power supply within the RF power supply. High frequency power means for providing high frequency power; And At least one reactive circuit for interfacing the high frequency power means to a plasma of constantly varying impedance; At a low expected plasma reactance limit, the at least one reactive circuit exhibits a small total reactance when the plasma reactance is at a high expected plasma reactance limit and exhibits a reactance that does not exceed a specified reactance. 19. The system of claim 18, wherein the specified limit is similar to the reactance of the high frequency power means. The set of claim 18, wherein the specified limit is a set smaller than about 0.5 ohms, a set smaller than about 2 ohms, a set smaller than about 3 ohms, a set smaller than about 5 ohms, a set smaller than about 8 ohms, about 10 ohms. A system selected from the set consisting of a smaller set, a set smaller than about 20 ohms, and a set that is about 12 ohms. 19. The system of claim 18, wherein the reactive circuit further comprises a transformer to provide DC isolation between the power supply and the antenna. 22. The apparatus of claim 21, wherein the high frequency power means comprises: a substantially Class A amplifier, substantially a Class AB amplifier, substantially a Class B amplifier, substantially a Class C amplifier, substantially a Class D amplifier, substantially a Class E amplifier And at least one element of an assembly consisting essentially of a Class F amplifier. 23. The system of claim 22, wherein said high frequency power means comprises a push-pull circuit. 24. The system of claim 23, wherein the push-pull circuit comprises at least one transistor that operates substantially in accordance with one of class D, class E, and classer F modes. 19. The system of claim 18, wherein the plasma generator generates a capacitively coupled mode ("E-mode") plasma. 19. The system of claim 18, further comprising an antenna having a loop of at least one current strap positioned proximate to a plasma source chamber and coupling said reactive circuit to said plasma. 27. The system of claim 26, wherein the current strip is a main coupler that couples power to the plasma. In the method for supplying power to a plasma exhibiting a dynamic plasma impedance with a high frequency power supply, Coupling the high frequency power source to at least one antenna via at least one reactive circuit; And wherein said at least one reactive circuit exhibits a small total reactance when said plasma reactance is a first plasma reactance and exhibits a reactance that does not exceed a specific reactance at a second plasma reactance. 29. The method of claim 28, wherein the first plasma reactance and the second plasma reactance values span substantially a portion of an expected dynamic plasma reactance range. 29. The method of claim 28, wherein the first plasma reactance and the second plasma reactance respectively correspond to a high expected plasma reactance limit and a low expected plasma reactance limit. 29. The method of claim 28, wherein the first plasma reactance corresponds to an average expected plasma reactance. A method of designing a reactive circuit that eliminates the need for a dynamic matching circuit between a plasma and an RF power source, Providing a low output impedance to the high frequency power source, And wherein said reactive circuit exhibits a small total reactance when said plasma reactance is a first plasma reactance, and exhibits a reactance that does not exceed a reactance specified in a second plasma reactance. 33. The method of claim 32, wherein the first plasma reactance and the second plasma reactance values span substantially a portion of an expected dynamic plasma reactance range. 33. The method of claim 32, wherein the first and second plasma reactance values correspond to high expected plasma reactance limits and low expected plasma reactance limits, respectively. 33. The method of claim 32, wherein the first plasma reactance corresponds to an average expected plasma reactance. 33. The method of claim 32, further comprising: adjusting, via feedback, power coupled to the plasma; Determining an output voltage and a current from a DC supply to the high frequency power supply; Calculating the measured efficiency of the high frequency power supply; And changing the output of the DC supply in response to the product of the output voltage and output current from the DC supply. A reactive circuit for interfacing high frequency power means having a low output impedance to a plasma having a dynamic impedance, And wherein the reactive circuit exhibits a small overall reactance when the plasma reactance is the first plasma reactance and exhibits a reactance that does not exceed the reactance specified in the second plasma reactance. 38. The method of claim 37, wherein the first plasma reactance and second plasma reactance values span substantially a portion of an expected dynamic plasma reactance range. 38. The method of claim 37, wherein the first plasma reactance and the second plasma reactance values correspond to high expected plasma reactance limits and low expected plasma reactance limits, respectively. 38. The method of claim 37, wherein the first plasma reactance corresponds to an average expected plasma reactance.

Images