|Publication number||USRE40264 E1|
|Application number||US 10/439,245|
|Publication date||Apr 29, 2008|
|Filing date||May 14, 2003|
|Priority date||Dec 4, 1995|
|Publication number||10439245, 439245, US RE40264 E1, US RE40264E1, US-E1-RE40264, USRE40264 E1, USRE40264E1|
|Inventors||Daniel L. Flamm|
|Original Assignee||Flamm Daniel L|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (35), Referenced by (5), Classifications (11), Legal Events (9)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This present application is a continuation-in-part of U.S. application Ser. No. 60/058,650 filed Sep. 11, 1997, and a continuation-in-part of U.S. application Ser. No. 08/567,224 filed Dec. 4, 1995, now abandoned which are hereby incorporated by reference for all purposes.
This invention relates generally to plasma processing. More particularly, one aspect of the invention is for greatly improved plasma processing of devices using an in-situ temperature application technique. Another aspect of the invention is illustrated in an example with regard to plasma etching or resist stripping used in the manufacture of semiconductor devices. The invention is also of benefit in plasma assisted chemical vapor deposition (CVD) for the manufacture of semiconductor devices. But it will be recognized that the invention has a wider range of applicability. Merely by way of example, the invention also can be applied in other plasma etching applications, and deposition of materials such as silicon, silicon dioxide, silicon nitride, polysilicon, among others.
Plasma processing techniques can occur in a variety of semiconductor manufacturing processes. Examples of plasma processing techniques occur in chemical dry etching (CDE), ion-assisted etching (IAE), and plasma enhanced chemical vapor deposition (PECVD), including remote plasma deposition (RPCVD) and ion-assisted plasma enhanced chemical vapor deposition (IAPECVD). These plasma processing techniques often rely upon radio frequency power (rf) supplied to an inductive coil for providing power to produce with the aid of a plasma.
Plasmas can be used to form neutral species (i.e., uncharged) for purposes of removing or forming films in the manufacture of integrated circuit devices. For instance, chemical dry etching is a technique which generally depends on gas-surface reactions involving these neutral species without substantial ion bombardment.
In a number of manufacturing processes, ion bombardment to substrate surfaces is often undesirable. This ion bombardment, however, is known to have harmful effects on properties of material layers in devices and excessive ion bombardment flux and energy can lead to intermixing of materials in adjacent device layers, breaking down oxide and “wear out,” injecting of contaminative material formed in the processing environment into substrate material layers, harmful changes in substrate morphology (e.g. amophotization), etc.
Ion assisted etching processes, however, rely upon ion bombardment to the substrate surface in defining selected films. But these ion assisted etching processes commonly have a lower selectivity relative to conventional CDE processes. Hence, CDE is often chosen when high selectivity is desired and ion bombardment to substrates is to be avoided.
In generally most, if not all, of the above processes maintain temperature in a “batch” mode. That is, the temperature of surfaces in a chamber and of the substrate being processed in such chamber are controlled to be at a substantially a single value of temperature during processing.
From the above it is seen that an improved technique, including a method and apparatus, for plasma processing is often desired.
The present invention provides a technique, including a method and apparatus, for fabricating a product using a plasma discharge. One aspect of the present technique relies upon multi-stage etching processes for selectively removing a film on a workpiece using differing temperatures. It overcomes serious disadvantages of prior art methods in which throughput and etching rate were lowered in order to avoid excessive device damage to a workpiece. In particular, this technique is extremely beneficial for removing resist masks which have been used to effect selective ion implantation of a substrate in some embodiments. In general, implantation of ions into a resist masking surface causes the upper surface of said resist to become extremely cross-linked and contaminated by materials from the ion bombardment. If the cross-linked layer is exposed to excessive temperature, it is prone to rupture and forms contaminative particulate matter. Hence, the entire resist layer is often processed at a low temperature to avoid this particle problem. Processing at a lower temperature often requires excessive time which lowers throughput. Accordingly, the present invention overcomes these disadvantages of conventional processes by rapidly removing a majority of resist at a higher temperature after an ion implanted layer is removed without substantial particle generation at a lower temperature.
In another aspect, the present invention provides a process which utilizes temperature changes to achieve high etch rates while simultaneously maintaining high etch selectivity between a layer which is being pattered or removed other material layers. An embodiment of this process advantageously employs a sequence of temperature changes as an unexpected means to avoid various types of processing damage to the a device and material layers. A novel inventive means for effecting a suitable controlled change in temperature as part of a process involves the use of a workpiece support which has low thermal mass in comparison to the heat transfer means. In an aspect of this invention, a fluid is utilized to change the temperature of a workpiece. In another aspect, the thermal capacity of a circulating fluid is sufficiently greater than the thermal capacity of the workpiece support that it permits maintaining the workpiece at a substantially uniform temperature.
Still another aspect of the invention provides an apparatus for etching a substrate in the manufacture of a device using different temperatures during etching. The apparatus includes a chamber and a substrate holder disposed in the chamber. The substrate holder has a selected thermal mass to facilitate changing the temperature of the substrate to be etched. That is, the selected thermal mass of the substrate holder allows for a change from a first temperature to a second temperature within a characteristic time period to process a film. The present apparatus can, for example, provide different processing temperatures during an etching process or the like.
The present invention achieves these benefits in the context of known process technology. However, a further understanding of the nature and advantages of the present invention may be realized by reference to the latter portions of the specification and attached drawings.
The etch apparatus includes a chamber 12, a feed source 14, an exhaust 16, a product support check or pedestal 18, an inductive applicator 20, a radio frequency (“rf”) power source 22 to the inductive applicator 20, wave adjustment circuits 24, 29 (WACs), a radio frequency power source 35 to the pedestal 18, a controller 36, an agile temperature control means 19 , and other elements. Optionally, the etch apparatus includes a gas distributor 17.
The chamber 12 can be any suitable chamber capable of housing a product 28, such as a wafer to be etched, and for providing a plasma discharge therein. The chamber can be a domed chamber for providing a uniform plasma distribution over the product 28 to be etched, but the chamber also can be configured in other shapes or geometries, e.g., flat ceiling, truncated pyramid, cylindrical, rectangular, etc. Depending upon the application, the chamber is selected to produce a uniform entity density over the pedestal 18, providing a high density of entities (i.e., etchant species) for etching uniformity.
The product support chuck can rapidly change its temperature in ways defined herein as well as others. The wafer is often thermally coupled to the support check which permits maintaining the wafer temperature in a known relationship with respect to the chuck. Coupling will often comprise an electrostatic chuck or mechanical clamps, which apply a pressure to bring the product into close proximity with the support check, which enables a relatively good thermal contact between the wafer and support chuck. The support chuck and wafer are often maintained at a substantially equal temperature. A pressure of gas is often applied through small openings in the support chuck behind the wafer in order to improve thermal contact and heat transfer between the wafer and support chuck.
The present chamber includes a dome 25 having an interior surface 26 made of quartz or other suitable materials. The exterior surface of the chamber is typically a dielectric material such as a ceramic or the like. Chamber 12 also includes a process kit with a focus ring 32, a cover (not shown), and other elements. Preferably, the plasma discharge is derived from the inductively coupled plasma source that is a de-coupled plasma source (“DPS”) or a helical resonator, although other sources can be employed.
The de-coupled source originates from rf power derived from the inductive applicator 20. Inductively coupled power is derived from the power source 22. The rf signal frequencies ranging from 800 kHz to 80 MHz can be provided to the inductive applicator 20. Preferably, the rf signal frequencies range from 5 MHz to 60 MHz. The inductive applicator (e.g., coil, antenna, transmission line, etc.) overlying the chamber ceiling can be made using a variety of shapes and ranges of shapes. For example, the inductive applicator can be a single integral conductive film, a transmission line, or multiple coil windings. The shape of the inductive applicator and its location relative to the chamber are selected to provide a plasma overlying the pedestal to improve etch uniformity.
The plasma discharge (or plasma source) is derived from the inductive applicator 20 operating with selected phase 23 and anti-phase 27 potentials (i.e., voltages) that substantially cancel each other. The controller 36 is operably coupled to the wave adjustment circuits 24, 29. In one embodiment, wave adjustment circuits 24, 29 provide an inductive applicator operating at full-wave multiples 21. This embodiment of full-wave multiple operation provides for balanced capacitance of phase 23 and anti-phase voltages 27 along the inductive applicator (or coil adjacent to the plasma). This full-wave multiple operation reduces or substantially eliminates the amount of capacitively coupled power from the plasma source to chamber bodies (e.g., pedestal, walls, wafer, etc.) at or close to ground potential. Alternatively, the wave adjustment circuits 24, 29 provide an inductive applicator that is effectively made shorter or longer than a full-wave length multiple by a selected amount, thereby operating at selected phase and anti-phase voltages that are not full-wave multiples. Alternatively, more than two, one or even no wave adjustment circuits can be provided in other embodiments. But in all of these above embodiments, the phase and anti-phase potentials substantially cancel each other, thereby providing substantially no capacitively coupled power from the plasma source to the chamber bodies.
In alternative embodiments, the wave adjustment circuit can be configured to provide selected phase and anti-phase coupled voltages coupled from the inductive applicator to the plasma that do not cancel. This provides a controlled potential between the plasma and the chamber bodies, e.g., the substrate, grounded surfaces, walls, etc. In one embodiment, the wave adjustment circuits can be used to selectively reduce current (i.e., capacitively coupled current) to the plasma. This can occur when certain high potential difference regions of the inductive applicator to the plasma are positioned (or kept) away from the plasma region (or inductor-containing-the-plasma region) by making them go into the wafer adjustment circuit assemblies, which are typically configured outside of the plasma region. In this embodiment, capacitive current is reduced and a selected degree of symmetry between the phase and anti-phase of the coupled voltages is maintained, thereby provided a selected potential or even substantially ground potential. In other embodiments, the wave adjustment circuits can be used to selectively increase current (i.e., capacitively coupled current) to the plasma.
As shown, the wave adjustment circuits are attached (e.g., connected, coupled, etc.) to ends of the inductive applicator. Alternatively, each of these wave adjustment circuits can be attached at an intermediate position away from the inductive application ends. Accordingly, upper and lower tap positions for respective wave adjustment circuits can be adjustable. But both the inductive applicator portions below and above each tap position are active. That is, they both can interact with the plasma discharge.
A sensing apparatus can be used to sense plasma voltage which is used to provide automatic turning of the wave adjustment circuits and any rf matching circuit between the rf generator and the plasma treatment chamber. This sensing apparatus can maintain the average AC potential at zero or a selected value relative to ground or any other reference value. This wave adjustment circuit provides for a selected potential difference between the plasma source and chamber bodies. These chamber bodies may be at a ground potential or a potential supplied by another bias supply, e.g., See
In this embodiment, the wave adjustment circuits are adjusted to provide substantially zero AC voltage at one point on the inductive coil (refer to point 00 in FIG. 2A). This embodiment also provides substantially equal phase 70 and anti-phase 71 voltage distributions in directions about this point (refer to 00-A and 00-C in
In this embodiment, it is desirable to reduce or minimize capacitive coupling current from the inductive element to the plasma discharge in the plasma source. Since the capacitive current increases monotonically with the magnitude of the difference of peak phase and anti-phase voltages, which occur at points A and C in
The discharge source helical resonator 53 can be constructed using conventional design formulae. Generally, this helical resonator includes an electrical length which is a selected phase portion “x” (A to 00 to C) of a full-wave helical resonator. The helical resonator wave adjustment circuits are each selected to jointly comprise a portion (2π-x) of full-wave helical resonators. Physical parameters for the wave adjustment helical resonators can be selected to realize practical physical dimensions and appropriate Q, ZO, etc values. In particular, some or even all of the transmission line parameters (Q, ZO, etc.) of the wave adjustment circuit sections may be selected to be substantially the same as the transmission line parameters of the inductive applicator. The portion of the inductive plasma applicator helical resonator, on the other hand, is designed and sized to provide selected uniformity values over substrate dimensions within an economical equipment size and reduced Q.
The wave adjustment circuit provides for external rf power coupling, which can be used to control and match power to the plasma source, as compared to conventional techniques used in helical resonators and the like. In particular, conventional techniques often match to, couple power to, or match to the impedance of the power supply to the helical resonator by varying a tap position along the coil above the grounded position, or selecting a fixed tap position relative to a grounded coil end and matching to the impedance at this position using a conventional matching network, e.g., LC network, π network, etc. Varying this tap position along the coil within a plasma source is often cumbersome and generally imposes difficult mechanical design problems. Using the fixed tap and external matching network also is cumbersome and can cause unanticipated changes in the discharge Q, and therefore influences its operating mode and stability. In the present embodiments, the wave adjustment circuits can be positioned outside of the plasma source (or constrained in space containing the inductive coil, e.g., See FIG. 2A. Accordingly, the mechanical design (e.g., means for varying tap position, change in the effective rf power coupling point by electrical means, etc.) of the tap position are simplified relative to those conventional techniques.
In the present embodiment, rf power is fed into the lower wave adjustment circuit 59. Alternatively, rf power can be fed into the upper wave adjustment circuit (not shown). The rf power also can be coupled directly into the inductive plasma coupling applicator (e.g., coil, etc.) in the wave adjustment circuit design, as illustrated by FIG. 2B. Alternatively, other applications will use a single wave adjustment circuit, as illustrated by FIG. 2C. Power can be coupled into this wave adjustment circuit or by conventional techniques such as a tap in the coil phase. In some embodiments, this tap in the coil phase is positioned above the grounded end. An external impedance matching network may then be operably coupled to the power for satisfactory power transfer efficiency from, for example, a conventional coaxial cable to impedances (current to voltage rations) existing between the wave adjustment circuit terminated end of the applicator and the grounded end.
A further embodiment using multiple inductive plasma applicators also is provided, as shown in FIG. 2D. This embodiment includes multiple plasma applicators (PA 1, PA2. . . PAn). These plasma applicators respectively provide selected combinations of inductively coupled power and capacitively coupled power from respective voltage potentials (V1, V2. . . Vn). Each of these plasma applicators derives power from its power source (PS1, PS2. . . PSn) either directly through an appropriate matching or coupling network or by coupling to a wave adjustment circuit as described. Alternatively, a single power supply using power splitters and impedance matching networks can be coupled to each (or more than two) of the plasma applicators. Alternatively, more than one power supply can be used where at least one power supply is shared among more than one plasma applicator. Each power source is coupled to its respective wave adjustment circuits (WAC1, WAC2. . . WACn).
Generally, each plasma applicator has an upper wave adjustment circuit (e.g., WAC1a, WAC2a. . . WACna) and a lower wave adjustment circuit (e.g., WAC1b, WAC2b. . . WACnb). The combination of upper and lower wave adjustment circuits are used to adjust the plasma source potential for each plasma source zone. Alternatively, a single wave adjustment circuit can be used for each plasma applicator. Each wave adjustment circuit can provide substantially the same impedance characteristics, or substantially distinct impedance characteristics. Of course, the particular configuration used will depend upon the application.
For instance, multiple plasma applicators can be used to employ distinct excitation frequencies for selected zones in a variety of applications. These applications include film deposition using plasma enhanced chemical deposition, etching by way of ion enhanced etching or chemical dry etching and others. Plasma cleaning also can be performed by way of the multiple plasma applicators. Specifically, at least one of the plasma applicators will define a cleaning plasma used for cleaning purposes. In one embodiment, this cleaning plasma can have an oxygen containing species. This cleaning plasma is defined by using an oxygen discharge, which is sustained by microwave power to a cavity or resonant microwave chamber abutting or surrounding a conventional dielectric vessel. Of course, a variety of other processes also can be performed by way of this multiple plasma applicator embodiment.
This present application using multiple plasma applicators can provide a multi-zone (or multi-chamber) plasma source without the use of conventional mechanical separation means (e.g., baffles, separate process chambers, etc.). Alternatively, the degree of interaction between adjacent zones or chambers can be relaxed owing to the use of voltage potential control via wave adjustment circuits. This plasma source provides for multiple plasma source chambers, each with its own control via its own plasma applicator. Accordingly, each plasma applicator provides a physical zone region (i.e., plasma source) with selected plasma characteristics (e.g., capacitively coupled current, inductively coupled current, etc.). These zones can be used alone or can be combined with other zones. Of course, the particular configuration will depend upon the application.
In the present embodiments, the wave adjustment circuit can be made from any suitable combination of element(s) such as various types of transmission lines, circuits, etc. These transmission lines include conventional solid or air dielectric coaxial cable, or ordinary, repeating inductor/capacitor discrete approximations to transmission lines, and others. These types of transmission lines are co-axial transmission lines, balanced parallel transmission lines, so called slow wave transmission lines with a spiral inner conductor (e.g., selected portions of a helical resonator, etc.), and others. Individual lumped, fixed, or adjustable combinations of resistors, capacitors, and inductors (e.g., matching networks, etc.) also can be used in place of transmission line sections for the wave adjustment circuit. These general types of wave adjustment circuits are frequency dependent, and can be termed frequency dependent wave adjustment circuits (or FDWACs).
Frequency independent elements also can be used as the wave adjustment circuits. These wave adjustment circuits can be termed frequency independent WACs (or FIWACs). Frequency independent wave adjustment circuits include degenerate cases such as short-circuit connections to ground or an infinite impedance (i.e., open circuit), and others. Frequency independent wave adjustment circuits can be used alone, or in combination with the frequency dependent wave adjustment circuits. Alternatively, the frequency dependent wave adjustment circuits can be used alone or in combination with other wave adjustment circuits. Other variations, alternative constructions, and modifications also may be possible depending upon the application.
With regard to operation of the wave adjustment circuits, various embodiments can be used, as illustrated by FIG. 2E. The wave adjustment circuits are used to select a wave length portion to be applied in the plasma applicator. In some embodiments, the average rf plasma potential is maintained close to ground potential by providing substantially equal phase 90, 81 and anti-phase 91, 82 capacitively coupled portions of the inductive applicator. This can occur in multi-wave embodiments 92, full-wave embodiments 93, half-wave multiple embodiments, quarter-wave multiple embodiments, or any other embodiments 94.
In alternative embodiments, it is desirable to maintain an elevated source plasma voltage relative to ground potential to induce a controlled ion plasma flux (or ion bombardment) to the product substrate (or any other chamber bodies). These embodiments are provided by selecting distinct electrical lengths for each of the wave adjustment circuit sections such that the capacitive coupled current from a phase section of the inductive plasma applicator is in excess of capacitive coupled current from the anti-phase portion. In these embodiments, the wave adjustment circuit provides a deliberate imbalance between the phase and anti-phase of the coupled voltages. In some embodiments 97, this occurs by shifting the zero voltage nodes along the process chamber axially, thereby achieving a bias relative to the plasma discharge. As shown, the phase 95 is imbalanced relative to its anti-phase 96. In other embodiments 99, one phase portion 84 is imbalanced by way of a different period relative to its complementary phase portion 85. Other embodiments are provided where the source plasma voltage is lower relative to ground potential. In the embodiments where imbalance is desirable, the potential difference between the phase and anti-phase potential portions is reduced (or minimized) when the amount of sputtering (e.g., wall sputtering, etc.) is reduced. The amount of sputtering, however, can be increased (or maximized) by increasing the potential difference between the phase and anti-phase potential portions. Sputtering is desirable in, for example, sputtering a quartz target, cleaning applications, and others. Of course, the type of operation used will depend upon the application.
Current maxima on an inductive applicator with distributed capacitance (e.g., helical resonator transmission line, etc.) occur at voltage minima. In particular, conventional quarter-wave helical resonator current is substantially at a relative maximum at its grounded end of the coil, and to a lesser extend in the nearby coil elements. Therefore, partial inductive coupling of power, if it occurs, will tend to be at this grounded end. In conventional half-wave helical resonators, inductive coupling tends to occur at each of the two grounded ends.
In the present invention, substantially anti-symmetric phase and anti-phase inductive half-wave and other fractional wave applicator sections support substantially more inductive coupling at a selected rf voltage node, e.g.,
In a specific embodiment, the power system includes selected circuit elements for effective operation. The power system includes an rf power source. This rf power source can be any suitable rf generator capable of providing a selected or continuously variable frequency in a range from about 800 kHz to about 80 MHz. Many generators are useful. Preferably, generators capable of operating into short and open-circuit loads without damage are used for industrial applications. One example of a suitable generator is a fixed frequency rf generator 28.12 MHz-3 kW CX-3000 power supply made by Comdel, Inc. of Beverly, Mass. A suitable variable frequency power supply arrangement capable of the 3 kW output over an 800 kHz to 50 MHz range can be made by driving an IFI Model TCCX3500 High Power Wide Band Amplifier with a Hewlett Packard HP116A, 0-50 MHz Pulse/Function Generator. Other generators including those capable of higher or lower power also can be used depending upon the application.
Power from the generator can be transmitted to the plasma source by conventional coaxial cable transmission line. An example of this transmission line is RG8/U and other higher temperature rated cable (e.g., RG1151U, etc.) with a coaxial TEFLON™ dielectric. In some embodiments, power is fed to conventional end-grounded half-wave helical resonators by positioning a movable tap on the helical coil and connecting a power source between the tap and the ground. In other embodiments, matching networks can be introduced between the coaxial cable power feed and the helical coil tap for flexibility. The matching network will depend on the selected wave configuration and wave adjustment circuits. In a balanced half-wave helical resonator embodiment, for example, the ends of the resonator coil can be terminated with wave adjustment circuits which substantially have zero susceptance. In particular, the wave adjustment circuit is designed as an open circuit by making no electrical connections to the ends of the coil, or establishing an electrical equivalence thereof. Alternatively, the ends of the coil are isolated by high series impedance chokes, thereby maintaining DC coupling to a fixed reference potential. These types of wave adjustment circuits are frequency independent and are “degenerate” cases. In these embodiments, the rf power is provided such that the phase and anti-phase current flows above and below the electrical midpoint (i.e., zero voltage node, etc.) of the coil. This provides for substantially balanced phase and anti-phase current flow from the power source stabilizing desired operation in coil voltages above the midpoint of the coil, and also provides substantially equal phase and anti-phase voltages.
The embodiments described above also can be applied to other plasma processing applications, e.g., PECVD, plasma immersion ion implantation (PIII), stripping, sputtering. For instance,
The product 119 having an upper surface 130 is placed into the present CVD apparatus for deposition, e.g., plasma enhanced chemical vapor deposition (PECVD), and others. Examples of deposition materials include a dielectric material such as a silicon dioxide (SiO2), a phosphosilicate glass (PSG), a borophosphosilicate glass (BPSG), a silicon nitride (Si3N4), among others.
In one embodiment, the deposition occurs by introducing a mixture comprising organic silane, oxygen, and an inert gas such as helium or argon according to the present invention. The organic silane can be any suitable organic silicate material such TEOS, HMDS, OMCTS, and the like. Deposition is also conformal in selected instances. As for the oxygen, it includes a flow rate of about 1 liter/per minute and less. A relative flow rate between the organic silane such as TEOS and oxygen ranges from about 1:40 to about 2:1, and is preferably less than about 1:2 in certain applications. A deposition temperature of the organic silane-oxygen layer ranges from about 300 to about 500° C., and can also be at other temperatures. Pressures in the range of 1 to 7 Torr are generally used. Of course, other concentrations, temperatures, materials, and flow rates can be used depending upon the particular application.
This chamber also includes a wave adjustment circuit 127. The wave adjustment circuit 127 is used to provide a helical coil operating with capacitive coupling to selected phase and anti-phase voltages. This portion 127 of the wave adjustment circuit coil also is shielded 140 to prevent rf from interfering with the plasma discharge or external elements, e.g., equipment, power, etc. The coil shield 140 is made of a conductive material such as copper, aluminum, or the like. In one embodiment, an operating frequency is selected and the wave adjustment circuit is adjusted to short circuit the upper end of the helical applicator coil to ground 124. This provides a helical coil operating at approximately a full-wave multiple and has substantially equal phase and anti-phase sections. This full-wave multiple operation provides for balanced capacitance of phase 151 and anti-phase 153 voltages along the coil 132 adjacent to the plasma source. Full-wave multiple operation reduces or even substantially eliminates the amount of capacitively coupled power from the plasma source to chamber bodies (e.g., pedestal, walls, wafer, etc.) at or close to ground potential.
In the present embodiment, the wave adjustment circuit 127 is a variable coil portion 128 of a spiral transmission line, which is selectively placed outside the outer shield 133. Accordingly, when the wave adjustment circuit is adjusted to become a short circuit, the plasma source “sees” only a selected full-wave multiple comprising substantially equal phase 151 and anti-phase 153 of the entire instantaneous AC voltages 134, 135. In this embodiment, stress of the deposited oxide film is often tensile, which can be undesirable.
Alternatively, the wave adjustment circuit 127 provides a helical resonator operating at selected phase and anti-phase voltages that are not full-wave multiples. This wave adjustment circuit provides for a selected amount of capacitive coupling from the plasma source to the chamber bodies. Stress of the deposited oxide film in this embodiment can be made to be zero or slightly compressive. In some embodiments, the oxide films can be deposited with an rf plasma potential of several hundred volts between the plasma source and the substrate to decrease the tendency of the oxide film to absorb moisture. This can occur by adjusting the wave adjustment circuit to add in a small section of transmission line outside of the source and correspondingly shortening the applicator coil (by moving the lower point at which the applicator coil is short-circuited and thereby decreasing the inductance of the applicator coil and electrical length of the helical resonator 126 (e.g., spiral transmission line, etc.) ). Of course, the selected amount of capacitive coupling will depend upon the application.
In this embodiment, the wave adjustment circuits rely upon open circuits (i.e., zero susceptance). Power transfer can be effected with a balanced feed such as an inductively-coupled push-pull arrangement with means such as coupled inductors. Techniques for constructing these coupled inductors are described in, for example, “The ARRL Antenna Book,” R. D. Straw, Editor, The American Radio Relay League, Newington, Conn. (1994) and “The Radio Handbook,” W. I. Orr, Editor, Engineering Ltd, Indiana (1962), which are both hereby incorporated by reference for all purposes. In one embodiment, a ferrite or powdered iron core “balun” (balanced-unbalanced) toroidal transformer (i.e., broadband transmission transformer, broadband transformer, etc.) 401 can be used to provide balanced matching from a conventional unbalanced coaxial transmission line. Techniques for constructing toroidal baluns are described in, for example, “Transmission Line Transformers,” J. Sevick, 2nd Edition, American Radio Relay League, Newington, Conn. (1990). The toroidal transformer is coupled between the rf power source 122 and the coil 132. The midpoint 406 between the phase 405 and anti-phase voltage on the coil is effectively rf grounded, hence it may be convenient to directly ground this midpoint of the inductive application in some embodiments for stability. This permits alternate operation in which power may be coupled into the inductive applicator (e.g., coil, etc.) with a conventional unbalanced feed line tapped on one side of the center. Push-pull balanced coupling ignites the plasma more easily than conventional unbalanced coil tap matching and generally is easier to adjust in selected applications.
Referring to the helical resonator embodiments operating at substantially equal phase and anti-phase potentials,
In an alternative operating mode, rings of plasma caused by inductively coupled plasma current are visible near top and bottom extremes of the inductive applicator, as illustrated by FIG. SC. This operating mode is generally for a full-wave 517 inductive coupling coil operated at a very high power, e.g., maximum power input to the inductive applicator is often limited by thermal considerations and breakdown. The rings 513, 515 of current in the plasma discharge are simulated by maximum coil current areas corresponding to voltage minima at the top and bottom shorted ends of the coil. Under these high power conditions, subordinate current rings are detectable and some excitation is often visible in the intermediate regions. This excitation is partially caused by capacitively driven currents within the discharge coupled to the voltage maximum and voltage minimum positions along the inductive applicator.
Alternatively, subordinate inductive plasma current rings at the top and bottom ends 513 of the resonator do not appear with limited input power. The coil current and inductive flux fall beyond the ends of the inductive applicator so that a single inductive ring 515 in the center portion is more stable, provided that the conductivity of the plasma is large enough to support a single current ring at a specified input power.
In alternative applications using high power operation, no secondary plasma current rings may be desirable. These applications often have substantially minimum internal capacitive coupling. In these applications, the inductive applicator (e.g., coil) abutting the vacuum vessel may be shortened from a full wave to an appropriate length such that only the central current maxima exists on the coil abutting the plasma source and the potential difference between maximum and minimum voltage on the applicator is substantially reduced. The present application is achieved by stabilizing the desired waveform along the applicator by appropriate impedance wave adjustment circuits.
An effective conventional method employed to avoid plasma potential shift in conventional commercially available inductive sources is to shield the plasma from the electrical fields on the inductive coupling element (commonly a multi-turn coil) by inserting a grounded conductive member between the inductive driving element and the plasma discharge tube. Shielding is, however, cumbersome and inconvenient and has serious disadvantages in practice. Shields couple to inductive applicator elements and can cause wide excursions in the natural resonance frequency, which are not predicted by conventional analytical design formulae. This often results in laborious trail and error and iterative mechanical designs to achieve a desired resonance.
Another disadvantage of shielding is that shields often make it difficult to achieve initial ignition of the plasma since shields generally exclude capacitive electric fields in the plasma discharge tube. Those with skill in the art will recognize that in practice, substantial ionization (e.g. charge carriers) must be present in the tube volume before in inductive plasma current can be sustained. Thus, it is often said that a capacitive plasma must be ignited first in order to sustain an inductive plasma. This is not completely true since other external means of generating ionization (e.g. an intense photoionizing source, electron beams, etc.) can be used to generate the ionization which is prerequisite to sustaining an inductive plasma. However, a capacitive discharge is often a convenient and cost-effective means to ignite some plasma which is a prerequisite to starting and sustaining an inductive discharge.
In general, wave adjustment circuits are employed to substantially diminish capacitive coupling between a plasma source and an inductive applicator. If most capacitive coupling is removed, it may be difficult to ignite a plasma. However, wave adjustment circuits also provide a means to overcome the difficulty with igniting a plasma in the absence of capacitive coupling under steady state operating conditions. This means is provided by electrically, mechanically, or electromechanically tuning the wave adjustment circuits prior to the time of desired plasma ignition in a manner which generates an additional imbalance and capacitive coupling to the discharge volume. Of course, the characteristics of the plasma as a load to the applicator will dynamically change during ignition. The wave adjustment circuit can also be continuously tuned under feedback control during ignition in order to provide a desired voltage coupling and diminish undesired transients during plasma breakdown.
Insertion of the shield close to high voltage RF point in a network (such as the voltage maximum points in a helical resonator or the high potential driven side of a TCP coil) also causes large displacement currents to flow through the capacitance between the shield and coil. This high potential difference is also a potential cause of damaging rf breakdown across the air gap, hence the gap may require protection by inconvenient solid or liquid dielectric insulation. The displacement current flow causes power loss and requires that higher power RF generating equipment be used to compensate for the power loss. Coupling loss in the plasma source structure is also undesirable from the standpoint of thermal control. These limitations are overcome by the present invention using the wave adjustment circuits, an inductive applicator of selected phase length, and other elements.
The side-view diagram shown as “SIDE-VIEW A” illustrates a plurality of zones 603, which correspond to each of the concentric zones 608A, 608B, 608C, and 608D. Each zone is separated, in part, from each other by a baffle 605. Each baffle extends from a lower region of the substrate holder toward an upper portion of the substrate holder, but does not touch the upper surface 601 of the substrate holder. In preferred embodiments, the baffles do not touch the upper surface. Accordingly, the temperature of the baffles, which may be different from the fluid, does not create an uneven temperature distribution along the upper surface 601. Additionally, the upper region 604, which is an open region within the substrate holder, provides fluid communication between each of the zones. Fluid communication in connection with the upper surface also provides an even distribution of temperature along the upper surface.
The fluid can be used to heat or cool the upper surface of the substrate holder. In a specific embodiment, the fluid generally should have a relatively high thermal conductivity and large heat capacity. The fluid should also be substantially non-corrosive, easy to transport, and can withstand a relatively large temperature range and still maintain its fluid characteristics. Additionally, the fluid should be able to be pumped and be substantially non-reactive with the substrate holder material itself. The fluid can be any commercial heat transfer fluid selected for the desirable temperature range. As noted, the substrate holder and upper surface cools down or is heated up by way of the fluid. The fluid can traverse through the zones and can absorb thermal energy or release thermal energy by an external heat transfer device such as the one described below, but can be others.
In a specific embodiment, the substrate holder also includes a plurality of heating elements 607. The heating elements can selectively heat one or more zones in a desirable manner. As shown in the “SIDE-VIEW B” diagram, each of the heating elements can be directed to a single zone 603, which has an adjacent baffle 605. Alternatively, the heating elements can be directed to multiple zones or other specific regions of the substrate and in particular the backside of the substrate according to some embodiments. The heating elements can be any suitable device for supplying heat energy to the fluid. The heat can be supplied by single or in combination using radiation, conduction, and convention. As merely an example, the heating element can be a resistive heating unit, an infrared heating unit, and others. Of course, the type of heating unit used depends highly upon the application. Alternatively, the heating unit can also be replaced by cooling units.
The present invention provides a substantially uniform temperature distribution along the upper surface 601 of the substrate holder. In a specific embodiment, the uniformity of the temperature is within one Degree Celsius along the entire surface which comes in contact with the object such as the wafer. Preferably, the wafer temperature also tracks the temperature uniformity of the upper surface. In other embodiments, the uniformity of the of the temperature is within one-tenth of a Degree Celsius along the entire surface which comes in contact with the object such as the wafer. Preferably, the wafer also has a uniform temperature distribution along the wafer.
The substrate holder has an upper surface, which holds an object in a secure manner during processing. The upper surface is generally made of a suitable material that has desirable heat transfer characteristics. In a specific embodiment, the upper surface is made using a low thermal mass, high conductivity material. As merely an example, the upper surface can be a diamond-like or diamond film overlying a copper or copper-like substrate. Of course, the type of surface used depends upon the application.
In a specific embodiment, the substrate holder also has temperature sensing units 615 such as the one shown in “SIDE-VIEW C.” The temperature sensing unit can be any suitable unit that is capable of being adapted to the upper surface of the substrate holder. Alternatively, the temperature sensing unit can measure the temperature of the fluid or lower surface of the substrate holder. As merely an example, the temperature sensing unit is a “fluro-optic” “fluoroptic” sensor unit made by a company called Luxtron in Santa Clara, California. Alternatively, the sensing unit can be a an edge band band edge IR sensor or the like. The sensing unit is capable of measuring a variety of spatial locations along the upper or lower surface of the substrate holder. The substrate holder can be implemented using a variety of systems for heating and/or cooling applications such as the one described below, but can be others.
Fluid from the fluid reservoir is pumped from the reservoir through the heating unit, which selectively sets the temperature of the fluid. The fluid leaves the heating unit at the selected temperature and traverses through the substrate holder, which can be similar to the one shown above, but can also be others. The fluid temperature selectively transfers energy in the form of heat to the wafer holder to a desirable temperature. Fluid leaves the substrate holder and traverses through node 721 and branch 719. Fluid traverse through branch 719 and node 717. Fluid leaves node 717 and returns back to the fluid reservoir via line 715.
In an alternative embodiment, the fluid can also be cooled using a heat exchanger 723. The fluid leaves the substrate holder and enters node 721. Fluid then enters the heat exchanger and traverses through loop 725. Thermal energy in the form of heat transfers to fluid in loop 723, which is cooler in temperature and draws heat away from the heat in the fluid in loop 725. In a specific embodiment, cooling fluid 727 enters and leaves the heat exchanger.
In a specific embodiment, system 700 operates in a manner to program a process temperature of the substrate holder. In this process, the reservoir with a suitable heat transfer fluid is maintained at a temperature below the desired process temperature. The fluid is circulated through the substrate holder or wafer chuck by the pump. The fluid line downstream of the pump is equipped with the an electrical heater which is capable of heating the fluid to a desired temperature. The desired fluid temperature is determined by comparing the desired wafer or wafer chuck set point temperature to a measured wafer or wafer chuck temperature (this measurement can performed with a thermocouple, thermistor, pyrometer, fluor optic fluoroptic sensor or other sensing means). If the measured temperature of the wafer of chuck or chuck is below the desired temperature, a suitable control algorithm such as a proportional controller or a proportional-integral-derivative (i.e., PID) controller algorithm increases the temperature by supplying more power to the heater.
The temperature of fluid emerging from the chuck is also measured (normally there will be a small temperature difference since there is heat exchange between the fluid and chuck). If this temperature is above the desired fluid temperature the fluid stream is diverted to a heat exchanger via control valves 1 and 2, which are respectively at nodes 721 and 717. If this temperature is less than or equal to the desired fluid temperature the fluid stream can be made to bypass the heat exchanger, or optionally can be heated in a heat exchange system. Since the temperature of commonly used heat exchangers cannot be changed rapidly, the heat exchanger is usually maintained at a single temperature sufficiently below the lowest desired fluid temperature to achieve the lowest desired temperature to be attained. The heat exchanger, fluid flow rate, coolant-side fluid temperature, heater power, chuck, etc. should be designed using conventional means to permit the heater to bring the fluid to a setpoint temperature and bring the temperature of the chuck and wafer to predetermined temperatures within specified time intervals and within specified uniformity limits.
In a preferred embodiment, the present invention uses a microprocessor based system to oversee the operations of the system described above. The microprocessor based system can have input and output ports, which are coupled to each of the elements, e.g., pump, heater, fluid reservoir, substrate holder. The microprocessor based unit selectively turns ON and/or OFF one or more of the elements to control the temperature of the substrate holder to provide a uniform distribution of temperature across the surface of the substrate holder.
In an alternative embodiment,
The temperature is programmed by use of this invention to achieve better control of the etching process as well as to permit stripping the resist in the same chamber and controlled control by the same process program which is used for stack etching.
A process according to the present invention can be briefly outlined as follows:
At the end of the breakthrough step, at time BB, the control program increases within several seconds to a higher steady state value at time C B. The tungsten silicide is etched at this temperature until this layer is breached at random locations on the wafer. This endpoint is conveniently observed by a change in the slope of intensity of an optical light emission from the plasma such as optical emission at 530 nm (point C in FIG. 10). The complete removal of the unmasked tungsten silicide areas is similarly signaled by a change in light emission such as that shown at point D (at time D all “patches” of the tungsten silicide are “cleared” from unmasked polysilicon areas; the signal begins to rapidly decease at time D because at constant temperature, polysilicon consumes chlorine more rapidly than tungsten silicide (e.g. a faster etch rate) and optical emission at this wavelength originates from a chlorine species.
Since it is not practical to change chuck temperature, at this point the etch rate would increase rapidly. As a consequence it can often be difficult to detect and terminate the polysilicon etching step when the thin oxide layer is reached. Another problem associated with the use of a single temperature for both silicide and polysilicon layers is that chlorine etching processes often undercut (etch along the mask direction, sideways- e.g. the etch is partly isotropic) silicon at the elevated temperatures suitable for a low residue tungsten silicide etch. Therefore it is highly desirable and advantageous to reduce the etching temperature during the polysilicon etch. The wafer temperature is gradually reduced at point DD D in order to achieve a slower and more anisotropic polysilicon etching step. The temperature necessary to etch tungsten silicide and polysilicon during this temperature programmed sequence are schematically compared in the FIG
At point H the stack etch is complete and the plasma is extinguished. The flow of feed gas comprising Cl2 for etching is stopped and a flow of O2 is started. A plasma discharge in oxygen at elevated temperatures will strip resist rapidly. At time I the wafer temperature is increased rapidly to a selected value in the range 180-220° C. and a plasma is ignited to remove the resist. After a selected interval the plasma is extinguished again and the chuck temperature rapidly lowered. These sequences of steps merely provide an example of the present invention. Other examples can also occur and the description above should not be limiting in any manner.
While the invention has been described with reference to specific embodiments, various alternatives, modifications, and equivalents may be used. In fact, the invention also can be applied to almost any type of plasma discharge apparatus. This discharge apparatus can include an apparatus for plasma immersion ion implantation or growing diamonds, TCPs, and others. This discharge apparatus can be used for the manufacture of flat panel displays, disks, integrated circuits, diamonds, semiconductor materials, bearings, raw materials, and the like.
Therefore, the above description should not be taken as limiting the scope of the invention which is defined by the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5179264||Dec 13, 1989||Jan 12, 1993||International Business Machines Corporation||Solid state microwave powered material and plasma processing systems|
|US5294778||Sep 11, 1991||Mar 15, 1994||Lam Research Corporation||CVD platen heater system utilizing concentric electric heating elements|
|US5320982 *||Jul 2, 1991||Jun 14, 1994||Hitachi, Ltd.||Wafer cooling method and apparatus|
|US5556204 *||May 19, 1994||Sep 17, 1996||Hitachi, Ltd.||Method and apparatus for detecting the temperature of a sample|
|US5571366||Oct 20, 1994||Nov 5, 1996||Tokyo Electron Limited||Plasma processing apparatus|
|US5609720 *||Sep 29, 1995||Mar 11, 1997||Lam Research Corporation||Thermal control of semiconductor wafer during reactive ion etching|
|US5645683||Feb 7, 1995||Jul 8, 1997||Nec Corporation||Etching method for etching a semiconductor substrate having a silicide layer and a polysilicon layer|
|US5667631||Jun 28, 1996||Sep 16, 1997||Lam Research Corporation||Dry etching of transparent electrodes in a low pressure plasma reactor|
|US5695564||Aug 3, 1995||Dec 9, 1997||Tokyo Electron Limited||Semiconductor processing system|
|US5700734||Jun 24, 1996||Dec 23, 1997||Nec Corporation||Process of fabricating field effect transistor having reliable polycide gate electrode|
|US5705433 *||Aug 24, 1995||Jan 6, 1998||Applied Materials, Inc.||Etching silicon-containing materials by use of silicon-containing compounds|
|US5756401||Mar 8, 1993||May 26, 1998||Fujitsu Limited||Process for the etching of polycide film|
|US5770099||Jul 18, 1995||Jun 23, 1998||Applied Materials, Inc.||Plasma etch apparatus with heated scavenging surfaces|
|US5863376||Jun 5, 1996||Jan 26, 1999||Lam Research Corporation||Temperature controlling method and apparatus for a plasma processing chamber|
|US5925212||Sep 5, 1995||Jul 20, 1999||Applied Materials, Inc.||Apparatus and method for attaining repeatable temperature versus time profiles for plasma heated interactive parts used in mass production plasma processing|
|US5939831||Nov 13, 1996||Aug 17, 1999||Applied Materials, Inc.||Methods and apparatus for pre-stabilized plasma generation for microwave clean applications|
|US5948283||Jun 28, 1996||Sep 7, 1999||Lam Research Corporation||Method and apparatus for enhancing outcome uniformity of direct-plasma processes|
|US5965034 *||Oct 28, 1996||Oct 12, 1999||Mc Electronics Co., Ltd.||High frequency plasma process wherein the plasma is executed by an inductive structure in which the phase and anti-phase portion of the capacitive currents between the inductive structure and the plasma are balanced|
|US6008139||Jun 17, 1996||Dec 28, 1999||Applied Materials Inc.||Method of etching polycide structures|
|US6033478||Nov 5, 1996||Mar 7, 2000||Applied Materials, Inc.||Wafer support with improved temperature control|
|US6042901||Feb 20, 1996||Mar 28, 2000||Lam Research Corporation||Method for depositing fluorine doped silicon dioxide films|
|US6048798||Jun 5, 1996||Apr 11, 2000||Lam Research Corporation||Apparatus for reducing process drift in inductive coupled plasma etching such as oxide layer|
|US6068784||Apr 1, 1993||May 30, 2000||Applied Materials, Inc.||Process used in an RF coupled plasma reactor|
|US6077357 *||May 29, 1997||Jun 20, 2000||Applied Materials, Inc.||Orientless wafer processing on an electrostatic chuck|
|US6087264 *||Jan 15, 1997||Jul 11, 2000||Samsung Electronics Co., Ltd.||Methods for patterning microelectronic structures using chlorine and oxygen|
|US6090303||Dec 5, 1996||Jul 18, 2000||Applied Materials, Inc.||Process for etching oxides in an electromagnetically coupled planar plasma apparatus|
|US6140612||Jun 7, 1995||Oct 31, 2000||Lam Research Corporation||Controlling the temperature of a wafer by varying the pressure of gas between the underside of the wafer and the chuck|
|US6165311||May 13, 1996||Dec 26, 2000||Applied Materials, Inc.||Inductively coupled RF plasma reactor having an overhead solenoidal antenna|
|US6167834||Aug 13, 1992||Jan 2, 2001||Applied Materials, Inc.||Thermal CVD/PECVD reactor and use for thermal chemical vapor deposition of silicon dioxide and in-situ multi-step planarized process|
|US6391148||Jan 12, 2001||May 21, 2002||Tegal Corporation||Cobalt silicide etch process and apparatus|
|US6486069||Dec 3, 1999||Nov 26, 2002||Tegal Corporation||Cobalt silicide etch process and apparatus|
|US20010003676||Jan 12, 2001||Jun 14, 2001||Tegal Corporation||Cobalt silicide etch process and apparatus|
|EP1236226A2||Nov 21, 2000||Sep 4, 2002||Tegal Corporation||Cobalt silicide etch process and apparatus|
|JPS5976876A *||Title not available|
|WO2001041189A2||Nov 21, 2000||Jun 7, 2001||Tegal Corp||Cobalt silicide etch process and apparatus|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US8105935 *||Mar 19, 2008||Jan 31, 2012||Fujitsu Semiconductor Limited||Method of manufacturing a semiconductor device|
|US8226769||Apr 26, 2007||Jul 24, 2012||Applied Materials, Inc.||Substrate support with electrostatic chuck having dual temperature zones|
|US8663391||Jul 23, 2012||Mar 4, 2014||Applied Materials, Inc.||Electrostatic chuck having a plurality of heater coils|
|US20070258186 *||Apr 26, 2007||Nov 8, 2007||Applied Materials, Inc||Substrate support with electrostatic chuck having dual temperature zones|
|US20080236614 *||Feb 29, 2008||Oct 2, 2008||Hitachi High-Technologies Corporation||Plasma processing apparatus and plasma processing method|
|U.S. Classification||216/59, 156/345.53, 438/715, 156/345.52, 216/67, 438/714, 216/74, 216/68, 204/192.32|
|International Classification||H01L21/302, H05H1/00|
|Nov 24, 2008||REMI||Maintenance fee reminder mailed|
|May 17, 2009||REIN||Reinstatement after maintenance fee payment confirmed|
|Dec 22, 2009||FPAY||Fee payment|
Year of fee payment: 8
|Dec 22, 2009||SULP||Surcharge for late payment|
|Feb 22, 2010||PRDP||Patent reinstated due to the acceptance of a late maintenance fee|
Effective date: 20100222
|Dec 24, 2012||REMI||Maintenance fee reminder mailed|
|Mar 18, 2013||SULP||Surcharge for late payment|
Year of fee payment: 11
|Mar 18, 2013||FPAY||Fee payment|
Year of fee payment: 12
|Oct 13, 2015||IPR||Aia trial proceeding filed before the patent and appeal board: inter partes review|
Free format text: TRIAL NO: IPR2015-01768
Opponent name: LAM RESEARCH CORPORATION
Effective date: 20150818
Free format text: TRIAL NO: IPR2015-01766
Opponent name: LAM RESEARCH CORPORATION
Effective date: 20150818
Free format text: TRIAL NO: IPR2015-01764
Opponent name: LAM RESEARCH CORPORATION
Effective date: 20150818
Free format text: TRIAL NO: IPR2015-01759
Opponent name: LAM RESEARCH CORPORATION
Effective date: 20150818